THE COMPOSITION OF CHICKEN LIPIDS AND COOKING OIL AS INFLUENCED BY COOKING, REUSED COOKING OIL AND FROZEN STORAGE OF FRIED CHICKEN I Thesis for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY WHU ~TA LEE 1972 2- » 1' ~ .. v.-“‘:~'~\‘xj-{ “e‘z'Jfi‘Frf-j‘ ""‘114‘v'.‘r“1.‘..", I. - ‘ ~ ~ , ‘ max-eemeem.rev-ma, ., . I u . .I LIBRARY Michigan Static University This is to certify that the thesisentitled The. Cameosifion of Chicken Lipiis Mi Cookmg OII as influenced. b3 CCOkinjt Reused. Cooking Oil. MA Puget} Wmae 0F Rag Char/Inn presented by LUhu-To. Lee. has been accepted towards fulfillment of the requirements for JD.” fidegreein—EJLLEmed Human N (“If i 011 //5&LLW7\ Major profm 0-7639 ‘ I 2'" amomc av ‘;I_‘ I HDAB&SUNS' “ sonxsmemme, - LIBRARY SINGERS I SPRIIEPORI. IICKIGI'II \; - I Jr.) \— TILE com | mm as}; PRCZEII ST Che after he fatty aciI adducting acid deal that hYdl Simultam Dr‘Olongm The mcle m P331111 tic highly u: D°1¥unga DROSphat ABSTRACT THE COMPOSITION OF CHICKEN LIPIDS AND COOKING OIL AS INPLUENCED BY COOKING, REUSED COOKING OIL. AND FROZEN STORAGE OF FRIED CHICKEN. by Whu - Ta Lee Chemical changes in commercial corn oil were studied after heating for 6. 2h. 36 and #8 hours at 200.0. Free fatty acids. viscosity. peroxide values. and non-urea adducting fraction increased. and iodine values and linoleic acid decreased with heating times. These results indicated that hydrolysis, oxidation. and polymerization took place simultaneously and proceeded at an accelerated rate during prolonged heating times. The predominant fatty acids. in both uncooked chicken muscle and skin fats were oleic acid. linoleic acid. and palmitic acid. Phospholipids from muscle and skin contained highly unsaturated fatty acids and large amounts of polyunsaturated fatty acids. Phosphatidylcholine. phosphatidylethanolamine. and sphingomyelin were the major compone CI rachic‘ in phos losses heat - double occurre DI used, I and ne‘~ in phos IIIOSphc fatty ; more as FAi dew “Sued Ur IiPids . ionths ’5 ir 88 re hc it ir. k g — — Whu-Ta Lee components of phospholipids. Chicken pieces were cooked in fresh corn oil. and arachidonic (203“). eicosatrienoic (20:3) and linolenic acids in phospholipids of chicken fat decreased considerably. The losses of unsaturated fatty acids were presumed due to the heat - induced oxidative deterioration of the unsaturated double bonds during cooking. It appears that oxidation occurred first with phospholipids. During the cooking process. in which reused corn oil was used. unsaturated fatty acids decreased in both phospholipids and neutral lipids. The losses of these acids were greater in phospholipids than in neutral lipids. probably because the phospholipids contain higher amounts of highly unsaturated fatty acids. Phosphatidylethanolamine and sphingomyelin were more sensitive to changes during cooking in reused corn oil. Fat deteriorations were greater in the chicken cooked in resued corn oil than in fresh corn oil. Unsaturated fatty acids declined in total lipids. neutral lipids. and phospholipids of uncooked chicken after three Inonths storage. and further losses occurred after six months etbrage. Decreases of unsaturated fatty acids were greater in phospholi; chorus con lipids dur all compo: three mom and lysoph storage. throughout Increase i found afte- Unsa‘. and ”cap? SIX IIIOI'IIZI'IS aCids Were total 0r Storage We I‘E‘Jsed C01 contents <- Getaimed , com”: 12 Fat . Whu-Ta Lee phospholipids than in total lipids or neutral lipids. Phos- phorus content of phospholipids indicated a decline in phospho- lipids during frozen storage. In muscle lipids. losses of all components of phospholipids were found during the first three months storage. and further losses of phosphatidylcholine. and lysophosphatidylcholine were found after six months storage. These results indicated that oxidation occurred throughout storage. A decrease in phosphatidylcholine and an increase in lysOphosphatidylcholine in skin lipids were found after storage. indicating that lipolysis occurred. Unsaturated fatty acids in total lipids. neutral lipids. and phospholipids of cooked chicken declined after three and six months frozen storage. The losses of unsaturated fatty acids were more pronounced in the phospholipids than in the total or neutral lipids. Greater losses of such acids after storage were found in the chicken fats which were cooked in reused corn oil than in fresh corn oil. Decreased phosphorus contents of phospholipids indicated that phospholipids declined during frozen storage. and were influenced by ‘cooking in reused corn oil. Fat deterioration during frozen storage was greater in “he chick an c Optimum . . I 011 Is 01 tan product : oon corn oil am contribu e u Whu-Ta Lee the chicken cooked in reused corn oil than in fresh corn oil. Optimum quality of chicken pre - cooked in heated corn oil is obtained by using fresh corn oil and consuming the product soon after cooking. Cooking in reheated or reused corn oil. and increasing the frozen storage periods. contribute to quality deterioration. TIE COI'POSITI IIWIENCED E' STORAGE O? F'. in part Departme THE COMPOSITION OF CHICKEN LIPIDS AND COOKING OIL AS INFLUENCED BY COOKING. REUSED COOKING OIL. AND FROZEN STORAGE OF FRIED CHICKEN by Whu - Ta Lee A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Food Science and Human Nutrition 1972 The w for his kim graduate pr. Sincere Lyman Bratz. Iarkakis fo: the guidancI The au Parents who. 0f ETaduate 13 8180 app (74470 ACKNOWLEDGMENTS The author expresses his gratitude to Dr. Lawrence Dawson for his kindness. helpfulness. and guidance during his graduate program. Sincere thanks are expressed to Dr. John Boezi. professor Lyman Bratzler. Dr. LeRoy Dugan Jr.. and to Dr. Pericles Markakis for the advice and encouragment given as members of the guidance committee. The author wishes to thank his wife. Yu-Mei. and his parents whose moral support was a major force in the persuance of graduate studies. The typing of this manuscript by Yu-Mei is also appreciated. ii LIST OF TABL‘ LIST 0? FIGS: INTRODUCTION Objectives. LITERATURE F. 1. Campos 20 Chemici Oil du Nutrit Foods - Effect Qualit Frozer g: Extracr TABLE OF CONTENTS LIST _ OF TABIES O O O O O O O O O O O O O LISTOFFIGURESoooeoooeeoeo INTRODUCTION 0 O O O O O O O O O O O 0 Objectives . . . . . . . . . . . . . LITERATUREREVIEW........... 1. 2. \OCD'Q C‘kfl 3w 0 Composition of Chicken Fats . . . Chemical Reactionswhich Occur in Cooking Oil during Heating. and their Effects on Nutritional Value and Quality of Fried Foods . . . . Effect of Cooking on Chicken Fats . . . Quality Stability of Fried Chicken during Frozen Storage . . . . . . . . . Extraction of Lip ids . . . Separation of Phospholipids from Neutral Lip ids . . Gas-L1quid Chromatography e e e e e e e e Methylation of Lipids . . . Phosphorus Determination of Phospholipids EXPERIMENTAL PROCEDURE . . .. . . . . . 1. 2. 3. 5. 6. 7. Changes in Corn Oil during Cooking with COtton Balls 0 e e e e e e e e 0 Preparation of Chicken . . . . . Cooking Procedure . . . . . . . Lipid Extraction and Purification . . . Fatty Acid Composition of Chicken Fats and COOking 011 e e e e 0 Separation of Phospholipids from Neutral Lipids o e e e 0 Determination of Phosphorus Content of PAGE so V so v1 . 1 ..7 . 8 . 15 . 22 .25 . 30 .31 . 32 - 33 . 35 . 36 .b6 ChickenLlpids.........o......‘¥8 iii RESULTS AND I 1. Change during 2. Campos Cookir. (1). (2). (3). (A). (5). 30 Change Fat du MARY AND PROPOSAL p0: LITERATURE APPENDIX . Chemical Gotten E TABLE OF CONTENTS ---- Continued page RESULTS AND DISCUSSION . . . . . . . . . . . . . . . . 53 1. Changes in Fatty Acid Composition of Corn Oil during Cooking with Cotton Balls . . . . . . . . 53 2. Compositional Change in Chicken Fat during COOking and Frozen Storage 0 e e e e e e e e e e 56 E1.TatallipidSeeeeeeeeeeeee0-56 2.PhOSphOlipidBe-esees-000000067 (3.Neutrallipld8... eeoeeoeee78 (A . Classification of phospholipids . . . . . .86 (5 . Phosphorus determination of phospholipids .86 3. Changes in Fatty Acid Composition of Cooking Fat during COOking With ChiCken e e e e e e e e .96 SUMMARY AND CONCLUSIONS . . . . . . . . . . e . . . . .98 PROPOSAL FOR FUTURE RESEARCH . . . . . . . . . . . . .106 LITERATURE CITED . . . . . . . . . . . . . . . . . . .107 APPENDIX . . . . . . . . . . . . . . . . . . . . . . .117 Chemical Changes in Corn Oil during Cooking with CottonBalls....................117 iv fable 2. 3. 5. 7. LIST OF TABLES Table Page 1. Changes in fatty acid composition of corn oil during cooking with cotton balls . . . . . . . .55 2. Changes in fatty acid composition of chicken muscle total lipids during cooking and frozenstorageeeeeesuccess-000058 3. Changes in fatty acid composition of chicken skin total lipids during cooking and frozen storage . . . . . . . . . . . . . . . . .66 u. Changes in fatty acid composition of chicken muscle phospholipids during cooking and frozen Storageeeeeeeeeeeeeeeeeeeee68 5. Changes in fatty acid composition of chicken skin phospholipids during cooking and frozen storage....................77 6. Changes in fatty acid composition of chicken muscle neutral lipids during cooking and frozen Storageeeoeeeeeeeeee000000079 7. Changes in fatty acid composition of chicken skin neutral lipids during cooking and frozen Storageeeee00.000000000000085 8. Identified classes of phospholipid . . . . . . .87 9. Changes in phospholipid content of chicken muscle lipids during cooking and frozen storage.89 10. Changes in phospholipid content of chicken skin lipids during cooking and frozen storage . 95 V Rare 2. 3. Changd heatir Changq in cor The fj 1n co L LIST OF FIGURES Figure Page 1. The stages of oxidation of a fat . . . . . . . 19 2. Diagrammatic illustration of a thin-layer chromatogram showing how on Rf value is measured and CfilCUlated e e e e o e e e e e e 0&9 3. Changes in free fatty acid and viscosity in corn 011 during heating 0 e e e e e e e e e e 118 h. Changes in absorbance of corn oil during heating 0 e e e e e e e e e e e e e e e e e o 120 5. Changes in iodine values and peroxide values in corn 011 during heating 0 e e e e e e e e 0122 6. The formation of non-urea adducting fraction in corn oil during heating . . . . . . . . . .12A vi The co: increased 1‘: in197o. D: of all red YT FOUndS in 1g Situation. 1 Values of 10 increased ai red Mats ir DOUltr-y ind; Genetic utilizationl resistaan. Rowth rates took 1N WEE} I NTRODUCTI ON The consumption of broilers in the United States has increased from 23.h pounds per capita in 1960. to 37.3 pounds in 1970. During this same period. the per capita consumption of all red meats in the United States increased from 160.9 pounds in 1960. to 186.3 pounds in 1970 (Poultry and Egg Situation. 1971). When 1960 consumption data are assigned values of 100. the per capita consumption of broilers in 1970 increased almost 60 percent. whereas the consumption of all red meats increased only 16 percent. This expansion of the poultry industry was influenced by many factors. Geneticists have deve10ped strains with improved feed utilization. growth rates. egg production. and disease resistance. Nutritionists have contributed by increasing growth rates and decreasing feed conversions. In 193A. it took in weeks and h to #2 pounds of feed to produce a broiler. By 1961. it required only 8 weeks and 2 to 25 pounds of feed Fer pound < tions. 1m} improved a. picking. f. have incre Poult and serve sensory pr a source c fat contai Proteins E 1959). T] “$in di. ”awning; ideal 1‘00 old 990131 their Wei during co fits well SChOOIS . The 2 per pound of gain to produce broilers under commercial condi- tions. Improved processing and distribution have accompanied improved efficiency of production. Developments in automatic picking. faster chilling and improved packaging and handling have increased the product shelf life. Poultry meat is economical. quick and easy to prepare and serve and has a number of desirable nutritional and sensory properties. Meat from poultry is low in calories. is a source of both saturated and unsaturated fatty acids. the fat contains essential fatty acids (Katz. et al.. 1966) and the proteins are a good source of essential amino acids (Scott. 1959). The meat fibers are tender. easy to chew or grind. easily digested and the flavor is mild and blends well with seasonings and other foods. Therefore. poultry meat is an ideal food for infants. young children. adolescents. adults. old peOple. convalescents. and those attempting to control their weight. Because of its high meat yield. low shrinkage during cooking and ease of cooking and serving. poultry meat fits well on the menu of restaurants. hotels. hospitals. schools. and institutions. The National Commission on Food Market (National Commision on Food Karl as value is closer to u of poultry ‘ apparent su in recent y market abou (Broiler In Labor final Produ in the deve reduce labc the labOr 1 in retail 1 3 on Food Market. 1966) showed that margins of profits increase as value is added to the product or as the product is brought closer to ultimate consumption. Therefore. further processing of poultry meat is one way to increase product value. The apparent successful expansion of Kentucky Fried Chicken Inc. in recent years is an example of this. and they expect to market about 750 million pounds of cooked chicken in 1972 (Broiler Industry . 1971) . Labor costs. which represent one of the major costs of a final product. have increased greatly. An important objective in the development of precooked frozen chicken has been to reduce labor costs. and this process eliminates or reduces the labor required for cutting up and repackaging Operations in retail stores. cooking Operation in branch shape. and reduces home preparation time. Thus there appears good potential for increased use of precooked frozen chicken. Fried chicken that has been frozen and reheated can be distinguished from freshly fried chicken by a slight diffe- rence in texture that seems to be recognized as tenderness rather than juiciness (Hanson. 195A). During storage. fried chicken f 1r: a "warmed 0' I chicken darLI et al.. 195‘ storage and blens invol fried chickI solving theI Oxidat ation of me 31-. 195s). may affect I‘elated to Le chicken first loses its "freshly cooked" flavor. then develops a ”warmed over” flavor. and eventually a rancid flavor (Hanson et al.. 1959). Besides a gradual change in flavor. fried chicken darkens in color in both skin and meat following storage and may appear less moist. These are the main pro- blems involved in quality stability of the precooked frozen fried chicken. and the future of these products depends on solving these problems. Oxidative rancidity is a major cause for flavor deterior- ation of meat during storage (Time and Watts. 1958; Turner et al.. 195A). Therefore. the lipids present in.muscle tissues may affect flavor quality. and can be reaponsible for problems related to product stability. The susceptibility of natural fat to oxidative rancidity depends largely upon its degree of unsaturation and its fatty acid composition. Poultry fat has a total unsaturation of 60 - 70 percent (Chang and Watts. 1952). thus poultry meat tends to become rancid faster than beef or lamb. Hanson et a1. (1959) found that flavor changes deve10p at a faster rate in muscle than skin. Katz et a1. (1966) showed a higher phospholipids content in muscle than in skin. and that phospholipids contained more long - chain polyuns- aturated fat have been cc correlate wi Changes presumably I and oxidizil due to a de- The cor Stahdpoint. effective 1 animals fed Oanmt be p food8. lino hm °1°8e1y acids can 1: some or the carbon Chai have 10we r 5 aturated fatty acids than did neutral lipids. Phospholipids have been considered as an important component which may correlate with flavor deterioration. Changes in texture of meat occur during frozen storage. ' presumably brought about by a reaction between the proteins and oxidizing fatty acids. Discoloration may also result due to a deterioration of meat proteins and fats. The composition of fat is important from a nutritional standpoint. A polyunsaturated fatty acid. linoleic acid. is effective in curing dermatitis and restoring growth of young animals fed a diet devoid of or very low in fat. Since it cannot be produced by the body and must be provided from foods. linoleic acid is considered an essential fatty acid. Two closely related fatty acids. linolenic and arachidonic acids can be synthesized from linoleic acid. They perform some of the same functions as linoleic acid. The shorter carbon chain length fatty acids and unsaturated fatty acids have lower melting points. are more easily emusified and hydrolyzed by bile salt and enzyme - lipase. so they are easier and more rapidly absorbed as nutrients. Precooked frozen fried chicken is prepared during high temperature How the hig this process of precooke' in this are Variou and POlymer (Carlin et oil are cor reactions , 0f time co: °°°kin€ oi; (temp in c. in chicken influEnCe d studies or Chem} ho'eVer . r curfew“t the effec 6 temperature cooking processes. which promote fat oxidation. How the highly unsaturated chicken fats will be affected by this process is very important for studying quality stability of precooked frozen chicken. since there are few references in this area. Various chemical reactions such as hydrolysis. oxidation. and polymerization may occur in cooking oil during heating (Carlin et al.. 195“). The free radicals formed in cooking oil are considered as the initiating agents for these chemical reactions. and the amounts formed may be related to the length of time cooking oils are heated or reheated. Absorption of cooking oil. or subtitution of part of the moisture might occur in chicken during cooking. The compositional changes in chicken fats during cooking and frozen storage may be influenced by reuse of cooking oil. and are important in studies on the stability of precooked frozen chicken. Chemical changes in cooking oils have been reported. however. most studies utilized cooking conditions quite different from practical deep fat frying. Before studying the effects of reused corn oil on chicken fat. the selection of specifi I prelinir changes ir using dee; of cherries before se] The s 1. Chenic heatir fresh durim OOrn . of specific heating times of reused corn oil was essential. A preliminary experiment was designed to study the chemical changes in heated corn oil during simulated cooking conditions using deep fat frying. An.evaluation of the types and rates of chemical reactions which occur in cooking oil was necessary before selection of apprOpriate chicken cooking treatment. The specific objectives of this study were to evaluate: 1. Chemical changes which occur in cooking oil during heating. 2. Compositional changes in chicken fat during cooking with fresh corn oil. and reused corn oil. 3. Compositional changes in raw and cooked chicken fats during frozen storage. including chicken cooked in reused corn oil. Fats a and are exc 1?? of raw concentrate mocalorie energy as a (Guthrie. 1 Steres GXCe b“? is der Terroi two grouPS: This distin 0f inadefine to mmiSh LITERATURE REVIEW 1. Composition of Chicken Fats Fats are one of the major components in the animal body. and are exceeded only by water and protein. Fats comprise 17% of raw chicken (Watts and Merril. 1963). and serve as a concentrated source of energy. Each gram of fat. provides 9 kilocalories per gram - two and a quarter times as much energy as an equal weight of either carbohydrate or protein (Guthrie. 1967). Fats represent the form in which the animal stores excess energy; thus the amount of fats in an animal body is determined by the energy balance of the animal. Terroine (Guthrie. 1967) classified the body fat into two groups: the ”constant element” and the "variable element". This distinction was based upon the fact that during periods of inadequate food intake the variable element is drawn upon to furnish energy for body processes while the constant element remains intact to preserve the essential structures of the body. The constant element represents that part which I l is essentia consists pr Phosph are lipids then yield phospholipi and a gener Hz-C-C I H ~c-c “2-0-: 9 is essential as a constituent of functioning cells. and consists primarily of phospholipids and sterols. Phospholipids. also called phosphatides and phospholipins. are lipids containing phosphorus. Upon hydrolysis. all of them yield fatty acids and phosphoric acid. and some phospholipids are called lecithins. or phosphatidylcholine. and a general formula is as follows: o Hz-C-O-G-R1 (CH3)3 i’ H -C-O-C-R2 N+ o Hz-C-O-F-O-(CH2)2-N+-(CH3)3 CH2 0- CHZOH Lecithin Choline R1 and R2 represent the residues of the molecules of fatty acids. The various members of the group contain different combinations of fatty acids. Cephalins are phospholipids which are similar to the lecithins except that the choline is replaced by amino - ethyl alcohol (hydroxy- ethyl amine. or ethanolamines) or by serine. After "I! . _.fl La... decaJ phOS} quan‘ Earl: lin I attac sphir Struc 10 decarboxylation. phosphatidyl serine may be converted to phosphatidylethanolamine. Sphingomyelins yield. on complete hydrolysis. equimolar quantities of sphingosine. choline. phosphate. and fatty acid. Early studies. following the initial isolation of sphingomye- lin by Thudichum (188k). demonstrated that the fatty acid was attached by an amide linkage to the primary amino group of sphingosine. Sphingomyelin. therefore. has the following structure a H H ?- ex; cn3-(cnz)12-cn a cn-t-t—cuzo-P-ocnzcnz- CH3 on NH 0 CH3 080 R Sphingomyelin Phospholipids have been found in chicken muscle and skin (Katz et al.. 1966. Peng. 1965). Katz et al. (1966) indicated white meat from chicken muscle. the lowest in total lipids. contained almost equal amounts of neutral and phospholipids. dark meat had twice as much lipid as did white meat. contained ll 79 percent neutral lipids and 21 percent phospholipids. and skin fat. high in total lipids. contained 98 percent neutral lipids and only 2 percent phospholipids. As the total lipid content in tissues increased. the ratio of neutral to phospholipids increased. The composition of phospholipids from poultry fats have been studied by several workers (Katz et al.. 1966: Peng. 1965; Miller et al.. 1962; Issacks et al.. 196A: Marion et al.. 1967). The predominant fatty acids of phospholipids are palmitic. stearic. oleic. linoleic. and arachidonic acids. These acids account for 75 percent of the total fatty acids. The fatty acid composition of phospholipids varies among the different tissues. Although fatty acid composition from the white and dark meat phospholipids are similar. they differ from that found in skin and depot fat. The most apparent difference in the phospholipid fatty acid composition of the different tissues is a variation in the arachidonic acid which is a twenty carbon chain acid containing four unsaturated double bonds at the 5. 8. 11 and 1h carbons. The percentage of arachidonic acid is lower in ID» Ir , .- the lip by chc C03 12 the lipid rich tissues (skin and depot fat) than in the muscle lipids. The type of phospholipids from poultry fate was studied by Peng (1965). and Davidkova and Khan (1967). Phosphatidyl- choline and phosphatidylethanolamine. are the predominant components of the phospholipids from chicken muscle. and the lesser components are phosphatidylserine. sphingomyelin. lysOphosphatidylcholine. and phosphatidylinositol. Kuksis et al.. (1969) found that rat heart. kidney. and plasma contained different amounts of lecithins with a different placement of fatty acids. Positional distribution of saturated and unsaturated fatty acids on egg lecithin have been studied by many workers (Tattrie. 1959: Dehaas et al.. 1960. Hanahan et al.. 1960) who have found that there is specific location of the saturated and unsaturated fatty acids on lecithins. The saturated fatty acids of lecithin are generally located at the.£3—(C-1)-ester position and unsaturated fatty acids are found preferentially located at the B-(C-2)-ester position. Peng (1965) studied the positional distribution of chicken phospholipids and showed that the poi po: pct {‘81 1‘85 an: is I SPIN» Amalie...“ CEI‘ die (Is: beer. 13 P°1yun8aturated fatty acids were located primarily at the 9 position and the saturated fatty acids mainly at the.£’ position. The variable element is the much larger group and represents the fat which has been deposited as an energy reserve. This depot fat consists principally of triglycerides and some diglycerides. monoglycerides and fatty acids. A considerable amount of triglycerides in animal tissues is formed directly from the absorbed dietary fat through the lymphatic system without following the biosynthetic pathway. Machlin et al. (1962) indicated that dietary fat affected and reflected the composition of body fat in the chicken. Generally. body tissues tends to assume the fatty acid composition of the fat in the diet (Marion et al.. 1967. Marion and Woodroof. 1963. Machlin et al.. 1962). Trigly- cerides in blood plasma and adipose tissue reflect more closely dietary fatty acids than do the phospholipid fractions (Issacks et al.. 196“). The composition of neutral lipids from poultry meat have been studied by many workers (Katz et al.. 1966. Peng. 1965. l \h n I ' ‘ ., raw-J" y: .- a u _ anou: the I pain. acid totaj aCCOI Quan. Such fatt3 REESE each t1 tat 1h Issacks et al.. 196% Machlin et al.. 1962). The neutral lipids from various tissues show similarity in the relative amount of the fatty acids. The predominant fatty acids in the neutral lipids are 016 fatty acids (palmitic and paLmitoleic) and 018 fatty acids (stearic. oleic. and linoleic acid). These C16 and 018 fatty acids amount to 9“ percent of total fatty acids in the neutral lipids. Palmitic acid accounts for a major portion.of the saturated fatty acids. A large percentage of the unsaturated acids are oleic and linoleic acids which amount to about 60 percnet. The major difference between phospholipids and neutral lipids from poultry meat in amounts of fattyacids as reported by Marion and Woodroof (1963). is that phospholipids contain appreciable quantities of long chain and highly unsaturated fatty acids such as 20-. 22-. 2a. polyunsaturated fatty acids. while fatty acids of chain length greater than 20 carbonsare present in low quantities in neutral lipids. Davidkova and Khan. (1967) separated and quantitated each component of chicken fat from muscle by TLC and found that phospholipids and triglycerides were the major components. "' "O.“‘A Q~';3_J‘I 830W WEN diglz (Char fatt: Ippre fat. fat. oils reac' Oils AldrI 0113 or a Slow 15 amounting to 89 percent of the total lipids. Minor components were cholesterol. cholesterol esters. free fatty acids. and diglycerides. Poultry fat has a total unsaturation of 60-70 percent (Chang and Watts. 1952. Davenport. 196A) . The predominant fatty acids are palmitic. oleic. and linoleic acids. Appreciable amounts of arachidonic acid are present in muscle fat. However. there is very little arachidonic acid in skin fate 2. Chemical Reactip ns Which OccurE in Cooking Oil dur ng Heat ng. and their Effects on Nutritional Value and Quality of Fried Foods. Carlin et al. (195A) described the thermal reactions of oils used for deep fat frying. Three general chemical reactions. all heat-induced. take place simultaneously when oils are used at elevated frying temperatures; namely. hydrolysis. oxidation and polymerization. Lantz and Carlin (1938) reported the rapid hydrolysis of oils when held at temperature of 3809F(193°C) in the presence of a constant flow of steam. Hydrolysis was reported to be slow during the first 20 hours of the heating period but that free fats shap. dougj 375 1 evim redu. 16 proceeded at a rapidly accelerating rate after the free fatty acid level reached 0.5 to 1.0%. This observation coincides with actual production experiences in the manufacture of fried foods in which large volumes of water must be evaporated. These authors. like Arenson and Heyl (19h3) also reported that the more highly saturated hydrogenated fats deve10ped free fatty acids at more rapid rates than did the saturated fats. Goodman and Block (1952) reported drastic changes in the shape. crust character. and fat absorption properties of doughnuts fried in fats subjected to prolonged heating at 375°?(19f’c). Thermally formed substances in the fat were evidenced by increased free fatty acid contents and greatly reduced iodine numbers. Carlin and Lannerud (19h1) reported a rapid drop in the oxidative stability of fats subjected to deep fat frying temperatures and a subsequent reduction in keeping quality of potato chips. Also noted were foaming tendencies of frying fat. a decrease in iodine number. an increase in saponification number. and marked increase in refractive index. T1. deve food food whic rath oxid Valu and- from the incr. bond Pan- the < “UGEr 17 Carlin et al. (195“) reported changes resulting in foam development. changes in.physical properties of the fried foods. loss of rancidity resistance in both fat and fried foods. color development and objectionable flavor development which were associated with thermally induced oxidative change rather than hydrolysis. Johnson and Kummerow. (1957) showed that the thermal oxidation of corn oil at 20090 caused a decrease in iodine value and carbonyl value and an increase in refractive index and viscosity. The linoleic acid content of the oil decreased from 53% to 30% during a Zh-hr. treatment. However. during the treatment. the monounsaturated acid content of the oil increased from 26% to 39.9%. suggesting that only one double bond of the linoleic acid was involved in part of the reactions. Perkin (1967) described the various stages involved in the oxidation of a typical fat as shown in Fig. 1. A fat or oil containing an unsaturated fatty acid. will undergo an induction period in which no detectable reaction occurs. The peroxide content of fat then begins to rise: and. 5“? J -Mwm£ afh per« mat the for: (191 ten; and fern cont PPOC oxic imp: 1963 1330 was Rats 18 after reaching a maximum. it eventually declines. while the percentage of oxygen in the oils gradually increases. Evenp tually polymerisation,increases. as does the viscosity of the oils. Degradation reactions also take place. resulting in formation.of volatile compounds. Crampton et a1. (1951). Pashke et al. (1952). and Powers (19h9) showed that. when fats and oils were heated at temperatures 250-300°C in inert atmospheres. molecular weight and viscosity increased. resulting partially from the formation of polymers through the Dials-Alder type of condensation. Cyclic monomers were also believed to be produced under these conditions (Paschke and Wheeler. 1955). The peroxides formed as labile first products of oxidation have been shown to destroy or damage physiologically important enzymes (Wills. 1961) and protein (Desai and Tappel. 1963) as well as mitochondria. microsomes and probably lysozomes (Tappel et al.. 1962). Air-oxidized soybean oil was fed to weanling rats to examine the peroxide effect. Rats on diets containing oxidised oil with peroxide value of 19 pee a co compwemxo so mmmMpm was made 4mg quh 0 :fl :oszo omp< cowhxo soflpmemmmoo Tllllllllllllll+llf hpwmoomfl _ . — — _ _ — _ ---———---_—----- compwmoasooon “ _ H _ _ _ _ . _ OUHXOHWQ weapmNfiaeshHom ‘ Ln-_nuu_-,___---_--__m-__- . I ’ nowpmauo oeflxomo >—--O1fi‘-——-———-———— ————-—.—-—-—- ..mwo>om schema cowposucH 20 100 performed as well as those receiving the control diet. Less weight. but substantial growth. was achieved on a diet containing oil with a peroxide value of #00. practically no weight gain at a peroxide value of 800 and loss of weight with death in 3 weeks resulted from ingestion of oil with a peroxide value of 1200. Peroxide groups appeared to be largely destroyed in the process of absorption from the intestine (Andrew at al.. 1960; Glavind and Tryding. 1960). Low levels of peroxide do not seriously influence growth. The peroxides are labile substances. which do not accumulate in strongly heated fats. and are able to decompose or react further by oxidation. fission or dehydration. or by polymerization to oxygen containing dimer or higher polymers. The volatile fission products. consisting mainly of aliphatic carbonylic compounds. are toxic (Kaunitz et al.. 1956) but they are normally present only in comparatively small quantities and their intense rancid odor and flavor are likely to prevent consumption at a dangerous level. Kritchevsky et al. (1961). and Kritchevsky and Tapper (1961). showed that a cholesterol-corn oil diet becomes more 21 atherogenic for rabbits when the oil was heated at 200°C for about 20 minutes. Heated fat is also more atherogenic for chickens (Nishida et al.. 1958). The chemical effect of the heating in this instance is a hydrolysis of the triglyceride of the oil. and the fatty acids thus released have been implicated as the cause of increased atherogenicity from the diet. The greater atherogenicity of cholesterol diets containing free fatty acid has been reported by others (Rona et al.. 19593 Meraill. 1960). Johnson and Kummerow. (1957) and Fleischman et al. (1963) showed that cooking oil loses appreciable quantities of linoleic acid after 2“ hours at cooking temperature. Linoleic acid is effective in curing dermatitis and restoring the growth of young animals fed a diet devoid of or very low in fat. Since it cannot be produced by the body. linoleic acid is considered as an essential fatty acid. The need for essential fatty acid. about 1% of the total calories. is provided by linoleic acid. Crampton et al. (1951. 1952) found that polymerized linseed oil heated at 275°C for 12 hours under carbon dioxide caused a high incidence of death when fed to rats. and they CO} dir ml' ha 80' OT Stt foc tre hol 0t! 22 concluded that the main toxic materials were cyclic monomers. dimers. and higher polymers. with the former causing the greatest toxicity. In recent years. special attention in nutritional studies has been given to the polymerized materials in damaged fats. It is apparent that loss of nutritive value of lipids is not significant when changes due to oxidation. thermal stress. or other factors are not great. therefore. any processing or storage procedure should avoid excess or extreme stress on a food system. This implies the use of mild processing treatments. minimum time and temperature heat stress. short holding periods. protection from oxidation. and use with other foods which insure a diet with all necessary nutrients (Dugan. 1968). 3. Effect of Cooking on Chicken Fats Research literature indicates that cooking of meat may affect certain characteristics of eating quality. such as tenderness. juiciness. color. and flavor. Goodwin et al. (1962) reported that whole turkeys roasted to an internal temperature of 55°C(13f’F) were significantly less tender as 23 shown by higher shear values than turkey cooked to an internal temperature of 77°C(171°F) or above. The water lost during meat cooking is assumed to be derived from both intercellular spaces and from the muscle fibers themselves. Any fibrillar water loss. due to coagulation of muscle fiber protoplasm. would be expected to be reflected in fiber shrinkage or decrease in cell diameter with changes in tenderness and/or cooking method. however. all studies are not in agreement. It is generally recognized that some intramuscular fat is lost from meat during cooking. Fat from endomysial and perimysial fat cells moves into the surrounding spaces. Chang and Watts. (1952) reported that the higher peroxide ‘value in meat after cooking indicated that certain amounts of fats were oxidized during cooking. Heat is one of the accelerators of lipid oxidation. It is known that primary oxidation of lipids involves oxidation at the double bonds and the carbon Jito a double bond to form hydroperoxides as primary oxidation products. These undergo a variety of scissions and dismutations to form a wide spectrum of carbonyl 2h compounds. hydroxy compounds. and short chain fatty acids. The formation of fatty acid hydroperoxides has little effect on quality since the products are essentially colorless. odorless and flavorless. Further reactions have a marked effect. Pippen et al. (1963) described an obvious difference in aroma between raw and cooked chicken. therefore. it was of interest to determine how cooking affected the chemical composition of chicken volatiles. He found the volatiles from uncooked chicken were very small and essentially odorless. whereas volatiles from cooked chicken were relatively large and odorous. N-hexanal and n—2.h-decadienal were the predominant components of the volatiles. To evaluate the possible role of the individual carbonyls with regard to odor. the carbonyls were regenerated. and the methyl ketones were described as oily or minty: the alkanals as meaty or turkey-like; the 2-enals had a strong oxidized broth-like odor: and the 2.h-dienal had a strong. painty. nutmeg or spicy odor (Bassette and Day. 1960). The variability in amounts and concentrations of these carbonyls 25 were sufficiently high to make them detectable by taste (Day et al.. 1963). In addition to carbonyls. the sulfur containing compounds from muscle protein (Mecchi et al.. 196“) and nonevolatile components of raw meat such as glucose. fructose. ribose. amino acids. amines. inosine. and monOphosphate (Koehler and Jacobson. 1967) are considered as precursors of chicken flavor. The effects of cooking on the composition of chicken fate is not yet completely understood. and is important in nutritional evaluation. h. Quality Stability of Fried Chicken during Frozen Storage Many workers (Hanson et al.. 1950: Hanson and Fletcher. 1958; Lineweaver et al.. 1952) reported that frozen fried chicken was considerably less stable than other frozen precooked poultry products. although the flavor changes in fried chicken were qualitatively the same as those observed in other frozen precooked poultry products. The first observed change was the occurrence of "warmed over” flavor or slight staleness. and eventually an objectionable rancid flavor. oc ca re Dr 81 D: N Pr 0:: Yo- 0c: 26 The flavor deterioration of cooked meat has been related to the oxidation of intramuscular lipids (Time and Watts. 1958; Turner et al.. 1951»). The thiobarbituric acid (TBA) test has been used to determine extent of oxidation (Younathan and Watts. 1960). The TBA values increased and off-flavor and odor developed with storage time and/or temperature in precooked meat (Cash and Carlin. 1968; Chang et al.. 1961a Keskinel et al.. 1960). Tappel (1952) reported that undesirable changes in color. flavor and nutritive value occur as meat constituents. such as pigments. other proteins. carbohydrates and vitamins. change in texture and rehydratability. Interaction between lipid oxidation and protein was thought to occur during the processing and storage of dried foods (Koch. 1962). The role of tissue lipid in rancidity development was postulated by Time and Watts (1958). They noted a rapid flavor deterioration in cooked meats during storage and proposed that this change in flavor resulted from the oxidation of highly unsaturated protein-bound phospholipids. Younathan and Watts. (1960) showed that less oxidation occurred in the neutral lipid fraction separated from rancid L "I I311 -. ."_. y ,. w -‘..‘ 27 cooked pork than in a total lipid extract or in the fraction referred to as phospholipids. Hanson et al. (1959) indicated that flavor changes of frozen fried chicken develop at a faster rate in.meat than in skin. The lability of the phospholipids is a result of their highly unsaturated fatty acid content and high levels of linoleic and arachidonic acids (Katz et al.. 1966; Peng. 1965). Phospholipids may also exist in closer contact with tissue catalysts of oxidation than do the triglycerides. thus increasing their tendency to oxidize (El-Gharbawi and Dugan. 1965). Much effort has been devoted to identification of catalysts in animal tissues responsible for the oxidation of unsaturated lipids. The accelerating effect of hemoglobin and other iron porphyrins on the oxidation of lipids is a generally accepted phenomenon. and hemoproteins have been implicated as the major prooxidants in meat and meat products (Tappel. 1952: Younathan and Watts. 1959: Tappel. 1953). Watts (195“) reported that heme catalyzed lipid oxidation results in destruction of the pigments. as well as oxidation of the fatty tissue. 28 Ferric hemochromogen is considered as the active catalytic form of the muscle pigments (Younathan and Watts. 1959: Tappel. 1953). In cooked meat. the pigment is in the active denatured ferric hemochromogen form. accounting for rapid initiation of lipid oxidation. Some evidence indicates that trace metal impurities present in salt account for its effects on lipid oxidation (Lea. 1939). however. there is evidence for a direct role of sodium chloride in initiating fat oxidation.(Ellis et al.. 1968). Salt is believed to catalyze the oxidation of the stored triglycerides (Watts. 1962). and the effect of sodium chloride on fat oxidation depends on the level of free moisture in the system (Chang and Watts. 1950). The effect of Nacl on oxidation has been attributed to the action of the reactive chloride ion on lipids. or to a modification of the hemo- protein catalysts of lipid oxidation (Ellis et al.. 1968). Several workers (Berry and McKerrigan. 1958; Caddie et al.. 1959: Pippen et al.. 1963) showed that carbonyl compounds. cleavage products of oxidized fatty acids. have been associated with off flavor. Pippen et al. (1963) studied 29 the volatile fractions from fresh and rancid chicken by gas chromatography and showed that the size of the 2.h-decadienal peaks in rancid chicken.was greater. He suggested that hexanal. 2.4-decadienal. and probably other carbonyl compounds contributed to rancid flavor of chicken. He noticed that these compounds had a desirable odor reminescent of cooked fat or of fried chicken. However, 2.9-decadienal. on exposure to air at room temperature. first developed stale odors and then rancid odors. Thus. this compound may contribute to desirable aroma but may also be an.immediate precursor of stale and rancid odors. Fresh fried chickens had lowest content of all classes of carbonyls. and the concentration of total carbonyls increased during storage time. Patton et al. (1959) pointed out that the principal precursor of carbonyls was linoleic acid. Dimiok and MacNeil. (1970) reported that the changes in fatty acid composition of the residue phospholipids during storage suggested that linoleic and arachidonic acids are the probable substrates in autoxidative deterioration. 30 5. Extraction of Lipids Preparation of a representative lipid sample from animal tissue requires two distinct processes: (a). extraction of the lipid and (b). purification of the lipid extract. i.e.. removal of unwanted nonlipid material from the lipid sample. To extract the lipids from animal tissue. the tissue must be severely altered and interactions between lipids and proteins or other molecules minimized. Tissue membranes may thereby pass through many thermotropic and lyotropic mesophases depending on extraction condition (Landbrooke and Chapman. 1969). Some of the lipid may not be bound to other solutes and may dissolve quite freely in a ”good” solvent. Other nonlipid material may associate with lipids quite soluble in a particular solvent. For this reason. many authors wash the lipid extract with aqueous solutions (Folch et al.. 1957). The combination of non polar and polar solvents has been considered suitable for lipid extraction from animal tissue. Entenman (1957) reviewed no less than 11 different 31 procedures to extract animal material under a variety of conditions and solvent combinations and Radin (1969) concluded that chloroform-methanol appears to be a good method for extracting wet tissue. Bloor (1914. 1915) used ethanol ether (3:1) and Folch et al. (1957) used chloroform-methanol (2:1) which are the most common methods used at the present time. 6. Separation of Phospholipids from Neutral lipids Chromatography on columns ofeilicioacid has been extensively applied to the separation of the polar lipids present in animal tissues (Hanahan et al.. 1957: Skipski et al.. 1952: O'Brien and Benson. 1969). The separation of phospholipids by TLC has been recognized as a fast and satisfactory method. Skipski et al. (1965) reported that animal fats could be separated in the following order: cholesterol ester. triglyceride. free fatty acid. diglyceride. monoglyceride. and phospholipid by TLC with a non polar solvent system. Phospholipids remained at the bottom and could be scraped from the glass plate. 32 7. Gas-Liquid Chromatography Gas-liquid chromatography (CLC) has been recognized as the most significant analytical innovation in the lipid field. The method was introduced by James and Martin (1952). who used a liquid phase of silicone grease. with or without 10% stearic acid. to separate normal saturated acids up to 012. which were detected by titrimetry. Cropper and Heywood (1953) extended the method to acids up to 622 by chromatographing the more volatile methyl esters. which were detected with katharometer. James and Martin (1956) introduced the gas density balance as a detector for methyl esters. and with the kpiezon hydrocarbon greases as stationary phases. they obtained separations of straight-chain and branched chain saturated and unsaturated esters of C10 - 018 acids. They also observed the first indications that geometric and positional isomers of unsaturated acids might be separated. Orr and Callen (1958). introduced the first of the polar polyester liquid phases. These greatly shortened analysis times and gave fairly complete resolution of the esters of common.fatty acids. including polyunsaturated acids. The 33 polar polyesters. including ethylene glycol succinate and adipate. and diethylene glycol succinate and adipate polyesters. are the most widely used for fatty acid esters and related compounds. The elution characteristics of individual compounds were first listed as absolute retention times or volumes. but these are so dependent on conditions that they were soon replaced by relative retention times. i.e.. the time of elution of individual compounds relative to some other component-usually palmitate or stearate. The introduction of the concept of “carbon numbers” (Woodford and Van Cent. 1960) or ”equivalent chain-length values” (Miwa et al.. 1960). further facilitated the description of elution characteristics and identification of individual components. These are derived by plotting the retention times against their chainlength on semilogarithmic graph paper and by interplotting on the straight line so obtained the retention times of other components. 8. Ibthylation of Lipids Two general approaches have been.used for methylation of I - an ‘ ~_' 319,3", 3a lipids. The most commonly used method involves liberation and isolation of fatty acids from lipids by saponification. acid hydrolysis. or enzymatic hydrolysis. The esters of the fatty acids may then be prepared by a variety of methods which required acid catalysis of the esterification reaction. The use of anhydrous methanol (Stoffel et al.. 1959) boron trifluoride in anhydrous methanol (Metcalf and Schmitz. 1961) diazomethane (Ropper and Ma. 1957) and sulfuric acid and methanol (Boyle and Ludwig. 1962: Rogozinski. 196%) have been used with varying success. A number of problems have been encountered using these methods such as safety hazard involved in working with diazomethane. long reaction time. and conjugation of double bonds. The second approach is direct formation of esters by interesterification reactions brought about by basic catalysis (Luddy et al.. 1960: Smith and Jack. 195“). McGinnis and Dugan (1965) introduced a rapid low temperature method which is accomplished by forming the sulfuric acid complex of the lipid in ethyl other at the temperature of dry ice-acetone bath. which results in direct formation of methyl esters of fatty acids from the decomposition of the complex with methanol. 35 9. Phosphorus Determination of Phospholipids Numerous procedures. including gravimetric. titrimetric. and colorimetric. have been introduced for phosphorus determination. Of these. colorimetric procedures are usually preferred for their rapidity and adaptability to microanalysis. The method suggested by Bell and Doisy (1920) and developed by Fiske and Subbarrow (1925). appears to be the one most generally accepted. Later Parker and Peterson (1965) introduced the method whereby the samples were scraped directly from TLC plate including silica gel. EXPERIMENTAL PROCEDURE 1. Changes in Corn 011 during Cooking with Cotton Balls (1). Preparation of samples: Six hundred grams of Miesel (brand) commercial corn oil were placed in a beaker and heated on an electric hot plate to 200°C and maintained at this temperature. Oil was heated for 6 hours. cooled overnight. then reheated 6 hours. and this process repeated through a total of #8 hours of heat. Moist cotton balls. previously washed thoroughly first with ethanol and then with redistilled hexane (Kawada et al.. 1967). weighing 1 g. and containing 80% by weight of water. were fried in the oil every hour. After 2 min. of cooking. the cotton balls contained approximately 2.95 moisture and 91.h% oil. Forty grams of oil were sampled at 1/2. 1 and 6 hours during the first period. then at the end of 29. 36 36 (2). (3). (u)- (5). (6). (7)- 37 and #8 hours of heating. Samples were cooled to room temperature. and kept in vacuum desiccator at 0°C. Free fatty acid (AOAC Official Method): Peroxide value (AOAC Official Method): Iodine value (Hanus Method - AOAC Official Method): Viscosity: Ostwald Viscometer was used for this analysis. Temperature bath maintained at 38°C. constant within O.1°C, Color: Samples were measured with Beckman DB spectrophotometer at wavelength 550:uu Samples were measured directly (undiluted) in a quartz cell of 1-mm path length with hexane as a reference. Non urea adduct forming fraction: Twenty-five g. of the oil were saponified by refluxing with 125 ml. of 5% alcoholic potassium hydroxide for 2 hrs. After cooling. 100 ml. water were added and the saponification mixture was acidified with two 100 ml. portions of ethyl ether. The combined extracts were washed with water to remove all mineral acid and dried over sodium sulfate. The solvent was then removed 38 by evaporation under low temperature and reduced pressure. Twenty g. of the fatty acids were weighed into a 500 ml. flask and dissolved into 200 ml. of methanol containing 80 g. urea. The flask was heated until a clear solution was obtained and then stored overnight in a refrigerator. The contents were then filtered through a Buchner funnel and the precipitate was washed with 50 ml. portions of cold ethyl ether. The filtrate was evaporated to dryness on a flask evaporator. This fraction was called the non-adduct forming fraction. To the precipitate. 100 ml. water and 10 g. sodium chloride were added. The fatty acids were released by heating and then extracted with ethyl ether. The resulting fatty acids were again treated with urea and methanol (5 g. urea and 10 ml. methanol for each g. of fatty acids). The fatty acids were then recovered from the urea inclusion products as described earlier. 2. Preparation of Chicken Seven week old male broilers were obtained from a commercial farm and were processed in the University poultry laboratory. The birds were killed. defeathered. and 8V1 wit 39 eviscerated in the usual manner. and chilled in water mixed with crushed ice for three hours. The birds were cut into portions identified as breasts. thighs. drumsticks. and wings. packaged in Cryovac bags and stored at 0°C. (1). (2). (3)- 3. Cooking Procedure Eggdmilk dip: Eight large eggs were blended for 1 minute in a Waring blendor. two “15.8 gram (1h.5 ounces) cans evaporated milk. and 1.9 liters (two quarts) cold water were added and mixed well immediately before use. Breading: A basic mixture of 11.“ kgs (25 pounds) of breading. obtained by mixing all purpose wheat flour and potato flour (5 to 1 ratio). 1.5 kgs (3.25 pounds) salt. and 738.“ gram (26 ounces) commercial seasoning was used in these experiments. Cooking: A Mies Commercial Pressure Fryer. Model C. was used in these experiments. 12.2 kgs (27 pounds) of Miesel (brand) commercial corn oil was placed in the cooker. no then preheated to 205°C (400°F). Four out up birds were dipped in the egg-milk mixture for 10 seconds to wet each piece evenly. drained. and breaded. When the temperature of the cooker reached 205°C. the pieces were added piece by piece and cooked for about 1 minute depending upon the color. When the chicken.was crown. the cooker lid was fastened shut and the pressure regulated to 15 psi. and the chicken was cooked for 9.5 minutes. Immediately after cooking. the pressure was released and the pieces removed. placed on a wire rack. and then transferred to a warming oven at 70°C to drain and darken in color. The pieces were held in the warming oven approximately 15 minutes. then repacked in a heat sealed polyethylene bag. Four groups of cut-up-birds were cooked during the first day. The temperature of corn oil in the cooker was maintained at 205°C (400°F) for 6 hours a day. cooled overnight. then reheated 6 hours. and this process repeated through a total of 48 hours. Four groups of cut up pieces were cooked at the end of 24 hours and 42 hours of heating times and 90 gram of (.0 luv-Iqhhfi | .o _ 41 corn oil was sampled at the end of 6. 30 and 48 hours of heating. Moist cotton balls. which were previously washed thoroughly first with ethanol and then with redistilled hexane. weighing 20 g. and containing 80% of water. were cooked every hour when chickens were not cooked. After 2 minutes of cooking. the cotton balls contained approximately 2.9% moisture and 91.4% of oil. One and four tenths kgs (3 ‘ pounds) of fresh corn oil were added each day before cooking. Twelve groups of chicken pieces including 3 groups uncooked cut-up-birds. 3 groups cooked with fresh corn oil. 3 groups cooked with 24 hour heated corn oil. and 3 groups cooked with 42 hour heated corn oil. were packaged in Cryovac bags and frozen at -37°C. All groups of samples were stored at -18°C for 3 or 6 months. The samples of corn oil including fresh corn oil. and corn oil heated for 6. 30 and 48 hours were kept in a vacuum desiccator and stored at 0°C. 4. Lipids Extraction and Purification Each group of chicken containing four cut-up-birds was ta )0! 38( gr! 1!...5 (i.e.wh . . 42 thawed at room temperature at the end of three months or six months storage. Muscle and skin samples were obtained from each group following separation of muscle and skin. and were ground prior to lipid extraction. Total lipids were extracted with chloroform-methanol- water 8:4:3 (v/y/y) according to the Folch method (Folch et al.. 1957). A 100 g. ground sample was blended in a Waring blendor for 2 minutes with 100 ml. chloroform. 200ml. methanol. and 4 ml. water. One hundred ml. chloroform were added and the mixture blended for 30 minutes after which 100 ml. water were added and the mixture blended for an additional 30 minutes. The resulting mixture separated by centrifuging at 4.000 rpm for 15 minutes. the supernatant was poured into a 1.000 ml. separatory fUnnel and the residue was blended again for one minute in 100 ml. chloroform. filtered. rinsed with 180 ml. chloroform and 40 ml. methanol. and collected in the separatory funnel. The filtrates in the separatory funnel were held at room temperature overnight to facilitate separation. A biphase liquid system was formed. The non-lipid components were absorbed by the upper phase of water and methanol. lipids 43 - moved to the lower phase of chloroform and some methanol. The lipids were collected and dried over sodium sulfate. then concentrated in a rotary vacuum evaporator at low temperature. Finally the total lipids were kept in a vacuum desiccator and held at 0°C. Lipids from fresh uncooked and cooked birds were obtained in the same manner. 5. Fatty Acid Composition of Chicken Fats and Cooking Oil (1). Methylation of lipids: Methylation of lipids was performed by a rapid low temperature method introduced by McGinnis and Dugan (1965)- The methyl esters of fatty acid were transferred. with a syringe. to a small tube. made from N0. 4 glass tubing by heat sealing one end. and stoppered with parafilm. This sample was kept at 0°C in the dark for not more than 12 hours prior to GLC analysis. (2). Gas-liquid chromatography: Fatty acid methyl esters were separated in an F a M 44 (model 810) dual column gas chromatograph. equipped with a flame ionization detector. A 72x1/4” copper column was packed with 15 percent diethylene glycol succinate and 3 percent phosphoric acid as liquid phase and chromosorb-W as solid support. Helium was used as carrier gas at a flow rate at 35 ml/min. The hydrogen flame was fed by using 60 ml/min hydrogen and 170 ml/min compressed air. The column.was stabilized by running nitrogen gas at 200°C for two days until a smooth base line was obtained. The volume of sample should be less than.2«ul for each injection to avoid shutting off detector flames. Column temperature was maintained at 190°C. with detector temperature at 26090 and injector at 250°C. Rubber plugs for sample inlet should be changed after each ten injections. otherwise gas may leak and peak height may be reduced. Preparation of column: Three grams of phosphoric acid were dissolved in distilled water and mixed with 82 g. acid washed. 80-100 mesh. chromosorb-W. The compounds were mixed in a round bottom flask. on a Rinco rotator. and dried in an oven. Fifteen g. of diethylene glycol succinate (DECS) dissolved in chloroform “5 were added to the dried mixture and enough chloroform to assure prOper mixing was added. After mixing and evaporating on the Rinco evaporator. the residual chloroform was removed in an air oven. and the dried mixture was ready for packing. Approximately 12 g. of DECS were packed in a 72x1/4“ copper tube with the aid of an electrical vibrator and vacuum pump. Class wool plugs were placed in both ends of the column to prevent loss of packing. Identification of fatty acid methyl esters on the chromatogram was made by direct comparison of retention time of each peak with the retention time of standards passed through the same column under the same conditions. When standards were not available. peaks were tentatively identified by semilogarithmic plots of corrected retention times. which were obtained by subtracting retention time of the air peak from those of sample peaks. against carbon number. and the best straight line joining them was constructed. A value corresponding to the retention time of any other peak could then be read from the graph. and gave the chain length from the same lipid class. such as. saturated. monoenes. 46 dienes. or trienes. The percentage of each fatty acid ester was calculated by dividing the area of each individual peak by the total area of all peaks. 6. Separation of Phospholipids from Neutral Lipids Thin layer chromatography: Supporting plates: The supporting layers were pieces of glass plate 20x20 cm and 5x20 cm. The glass plates were washed with detergent. rinsed with distilled water. and dried before use. Slurries: Thirty grams of silica gel C powder (Merck) were weighed into a mortar. 60 ml. distilled water were added. the mixture was carefully mixed for approximately 30 seconds. after which 3 to 5 additional ml. distilled water were added. followed by additional mixing. Layer casting: The Stahl-Desaga applicator was used. Thsthickness of casting layer was adjusted to 250.&as Well cleaned glass plates were arranged on a plastic holder and the applicator was seated properly on one end of glass plates. Slurry was poured into the applicator as soon as it was well 4? mixed. The applicator was carefully moved along the glass plates at a constant speed. The casting plates were allowed to dry at room temperature then kept in a desiccator. Sample spotting: The sample. dissolved in chloroform. was applied along the bottom of the plate in several spots at a distance of 0.5 cm from each other. The plate was allowed to stand until the spotting sample was dried. Development: Before initiating chromatography. the developing chamber was saturated with solvent. This was done by lining the walls of the chamber with filter paper (half way around and almost to the top). The paper was wetted with solvent before the chromatography was started. The solvent system was a combination of hexane. diethyl ether. and acetic acid in a volume ratio of 80:20:1. Ascending chromatography was used in a saturated chamber. This procedure allowed the solvent to rise to within .5 cm of the t0p of adsorbent. Average running time was 30 minutes. After development. the solvent was allowed to evaporate from the plate and the spots were then made visible in iodine vapor. Phospholipids remained at the origin because a non 48 polar solvent was used. The phospholipids were scraped off the plate as soon as possible to avoid deterioration and washed into a small test tube with the solvent containing chloroform and methanol at a ratio of 2 to 1 (v/y) through the filter paper. After the phospholipids were completely scraped off the plate. the remaining neutral lipids were scraped. and washed into small tubes with ethyl ether. The purified phospholipids and neutral lipids were subjected to methylation for CLC analysis. 7. Determination of Phosphorus Content of Chicken Lipids (1). Classification of phospholipids: Thin layer chromatography: Each sample. dissolved in chloroform containing 5 to 25 41.8 of phospholipids. was applied on a plate along with standard materials obtained from Applied Science Laboratories Inc.. Each plate was developed in a chamber saturated with the solvent containing chloroform. methanol. acetic acid. and water with the ratio of 25:15:4:2 by volume. After the development of the plate. the solvent was allowed 49 to evaporate and the spots were made visible in iodine vapor. The Rf value of each spot was obtained by dividing the distance moved by the solvent front into the distance moved by the compound (measured to the center of the spot). Both values were measured from the origin (See Fig 2). --------K-_- Solvent Front Origin Fig. 2. Diagrammatic illustration of a thin-layer chromatogram showing how on Rf value is measured and calculated. Identification of each spot was made by comparing the Rf value of the unknown epot with the Rf value of a standard. 50 (2). Phosphorus content: (a) . Standard curve A solution was made of 0.4393 g. dry C. P. potassium dihydrogen phosphate in water and diluted to one liter. This solution contained 0.1 mg. of phosphorus per ml. Further dilution provided a concentration of 0.01 mg. per ml. The various amounts of KHzPOu solution were pipetted into 30 ml. Kjeldahl digetion flasks. followed by 0.9 ml. perchloric acid (phosphorus free. 70-72%). and distilled water to a total volume to 5.9 ml. Two glass beads were added to each flask. The flask contents were digested using a low heat setting (4-5 position) on an electrically heated digestion rack (Laboratory Construction Company. model A) and the flasks were shaken occasionally. After approximately 30 minutes of digestion. the digestion fluid became colorless. the excess perchloric acid was evaporated rapidly by increasing the heating rate and introducing a piece of glass tubing into the neck of the Kjeldahl flask. the other end of the tubing was 51 connected to the water pump fume head. Following digestion. the flasks were removed from the heating rack and allowed to cool. and 5 ml. distilled water. 1 ml. of 2.5% ammonium molybdate. 1 ml. of 10% ascorbic acid. and additional 2 ml. of distilled water were added. The flasks were heated in boiling water for 5 minutes. A spectronic 20 (Bausch & LombIncorporated. Rochester. New York.) was set at 820 1m and adjusted to read zero with the blank as a reference. The optical density of each solution was recorded and the standard curve was plotted with optical density vs. phosphorus content. (b). Phosphorus content of chicken fats Samples were dissolved in chloroform. applied as narrow streaks on TLC plate. 1.5 cm from bottom with a microliter syringe containing approximately 200.ug phospholipids. The plates were developed in a chamber saturated with combination of chloroform. methanol. acetic acid. and water with the ratio of 25:15:4:2 by volume. This solvent was allowed to rise to within 0.5 cm of the tap 52 of the adsorbent. Average running time was 1.5 hours. After the plates were dried. they were exposed to iodine vapor. and each row of spots was immediately outlined. Each row of spots was scraped directly into a 30 ml. Kjeldahl digestion flask. An adjacent area of blank silica gel corresponding in size and position to the areas containing phospholipid were also scraped into digestion flasks. in which 0.9 m1. of perchloric acid and glass beads were added. After complete digestion and cooling. 5 ml. distilled water. 1 ml. of 2.5% ammonia molybdate. 1 ml. of 10% ascorbic acid. and 2 ml. distilled water were added and the mixture was boiled for 5 minutes. Optical density of each solution was recorded. Finally. phosphorus content was determined from the standard curve. RESULTS AND DISCUSSION 1. Changes in Fatty Acid Composition of Corn Oil during Cooking with Cotton Balls This experiment was designed to precede actual chicken cooking and analyses. to determine the influences of heating time on extent and rates of chemical reactions in corn oil. Specific heating times of reused corn oil were essential to the establishment of appropriate chicken cooking treatments. Many commercial cooking oils and fats have been used for fried foods. Corn oil is one of the more popular cooking oils. and has a fatty acid composition appropriate for this study. Cooking conditions. similar to practical deep fat frying procedures were performed as follows: Commercial corn oil. heated to 200°C. was used to cook moist cotton balls instead of chicken (Kawada et al.. 1967). Oil samples were taken after 0.5. 1. 6, 24. 36 and 48 hrs. during cooking. Corn oil 53 54 was found to contain 58.6% linoleic acid. 25.3% oleic acid. 13.7% palmitic acid. 2.4% stearic acid. and trace amounts of linolenic acid (Table 1). The fatty acid composition of corn oil during heating showed a decreasing amount of linoleic acid (C18,2). and increasing amounts of oleic (C18'1) and palmitic acids (016:0)' Linoleic acid decreased at a slow rate during the first 24 hours heating period. then it proceeded at a more rapid rate. Oleic acid increased slowly throughout the heating process. Palmitic acid remained essentially unchanged through the 24 hours heating period. then increased gradually. Stearic acid remained relatively constant. The percentage of all fatty acids was calculated as 100 percent. therefore. when one of these fatty acids decreased. one or more of the other fatty acids would increase on a relative basis. The increase in oleic and palmitic acids resulted from the decreasing linoleic acid during heating. therefore. it is concluded that the change in fatty acid of corn oil during heating was mostly due to the decrease in linoleic acid. 55 .mccon cansoc no bones: . muonamo no monssz .m «.8 «.8 01% :31 2.3 9% 0.8 3.3:: Hence m.ma a.ma o.oa m.wH 0.3H a. a H.wH .psm Haves :: :: :: u: :: some» comm» m.wH 06m 0.0m min QR a.mm ER 9mm «.3 N.~ o.on ~.m~ m.w~ m.w~ a.mm n.m~ a.ma Sm eta. ~.~ :4 ate. a; {N Ema a.ma n.0a m.ma m.aa m.~a o.ea a.ma c.0a “some some oAnv psoouem we mm am e a m.o o «memos seems a.mugv mesa» weaves: r maamn coppoo and: wswxooo washes ago Shoo mo cowpamoasoo macs apps“ a“ commune .H capes 56 The chemical reactions which occur in cooking oil include hydrolysis. oxidation. and polymerization. These reactions were demonstrated by the increase in free fatty acids. the decrease in iodine values and in linoleic acid content and by formation of NUAF (Appendix). The mechanisms of these reactions are still not clear. However. the formation of heat—induced free radicals and the highly unsaturated double bonds present in oil have been considered as the reactive components. Linoleic acid contains two double bonds separated by a single methylene group. It has been known for many years that this separated methylene group is extremely vulnerable to attack. The rates of autoxidation of linoleic acid and oleic acid present either as acid form or triglyceride form are 27 to 1. Therefore. the decrease in linoleic acid in corn oil was expected. 2. Compositional Change in Chicken Fat during Cooking and Frozen Storage ( 1 ) . Total lipids: (a). Total muscle lipids: The major fatty acids from uncooked chicken 57 muscle fat were oleic. linoleic. and palmitic acids. and accounted for 79.3% of the total. The remainder were minor fatty acids such as stearic. palmitoleic. linolenic. and arachidonic acids (20:4). The total unsaturation was 70.1% (Table 2). Chang and Watts (1952) reported that chicken :nuscle fat contained 45.2% oleic acid. 18.1% linoleic acid. (3.98% linolenic acid. 0.58% arachidonic acid (20:4). and 230.3% saturated acid. They found higher oleic acid but lower Ilinoleic acid than.were found in this study. Dietary fat taffects and reflects the composition of body fat in chicken :18 reported by Marion and Woodroof(1963). and by Machlin et Gal. (1962). Marion and Woodroof (1963) reported that chicken nnuscle fat contained 27.7% oleic acid and 23.6% linoleic acid vvhen the chicks were fed a basal ration. However. composition (changed to 24.7% oleic acid and 38.7% linoleic acid when czhicks were fed with additional 6% corn oil. and further (:hanged to 34.5% oleic acid and 13.8% linoleic acid when chicks Were fed a basal diet plus 6% beef tallow. Thus. chicks fed ijgth corn oil. which was high in linoleic acid. contained high linoleic acid in the muscle fat. and when fed beef tallow. .mcson cansoe mo genes: . msocnmo mo mecssznnp .mason a: mom cosmos admsofl>oua ado spec .o .mmson am you moves: hamsofi>oma Hwo choc .m .Hwo shoe seems .<::m 58 0.00 0.00 0.00 0.a0 0.00 0.00 0.00 0.00 0.00 0.00 0.H0 “.00 .p0020 H0000 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 .000 H0900 .. a- u- 0.“ u- -- u- 0.0 u- -- u: 0.0 0.00 -- u- u- -u n- u- -u n- -u -u 0.e H.0 0.00 0.0 0.0 0.H 0.0 0.a 0.0 0.0 0.0 a.“ 0.0 0.H m.H 0.00 0.00 0.a0 0.00 0.0a «.30 0.00 0.00 H.00 0.00 0.00 0.00 0.00 0.0H 0.00 0.00 m.«0 0.00 0.00 0.00 0.H0 0.00 0.00 0.00 0.Hm 0.00 H.0H 0.: 0.0 0.0 0.0 0.: 0.0 0.0 0.0 0.0 «.0 0.m 0.0 0.0a we: we: me# New Me: 0.0 Com Mom 3cm Ben as Com H~Wfi 0.00 0.HN 0.0a 0.00 0.00 0.00 0.00 0.00 m.a0 0.00 0.H~ 0.00 0.0a 0.“ «.0 0.0 0.0 0.0 0.0 0.0 0.0 .0 0.0 m.a H.a 0.0a 0.s N.H 0.0 0.0 a.“ 0.0 0.0 m.“ 0.0 N.H 0.“ a.H 0.:H Amend amen oquv Pceoaom omcaom 09900 0 0 < :00 0 0 < 300 0 0 4. 300 00205008. :0 o m o A.oEV osfip owmuopm owmsopm meson“ one mswxooo weaken mcwafia Hmpop caches sexowso mo sowvfimanoo whom 099mm :H newsmno .N magma 59 which was high in oleic acid. had an increased amount of oleic acid in muscle fat. Since heating accelerates chemical changes in lipids. an evaluation of the effects of heating on the composition of chicken fat is necessary for nutritional evaluation and quality stability analysis of fried chicken for further storage. Losses of arachidonic (020,4). eicosatrienoic (020.3). (oleic. stearic. and palmitoleic acids were found in total rnuscle lipids after chicken was cooked in fresh corn oil (Table 2). However. linoleic acid increased. Chang and Watts (1952) reported similar results. During the cooking procedure. there are several physical sand chemical changes which may occur in chicken. The Iphysical changes may involve losses of water and some chicken fat but gains of some cooking oil. The chemical changes fianlude oxidation. hydrolysis. and polymerization. The losses c>f unsaturated acids such as arachidonic and eicosatrienoic acids. provide evidence for the occurrence of chemical reactions during cooking. The increase in linoleic acid during cooking. may be explained as an actual increase by 60 the absorbance of corn oil. which contains 58% linoleic acid. and/or actual losses of other acids. which on a relative basis. results in an increase in linoleic acid. Studies on heated corn oil (Appendix) showed that various chemical reactions did occur during cooking. Also. three stages in the rate of chemical reactions. were found. during the first 24 hours. from 24 hours to 36 hours. and after 36 hours of heating. From these results. three heating times were selected to preheat the corn oil used to cook chicken. namely 0. 24. and 42 hours. and the effects of reused corn oil on chicken fat during cooking and frozen storage were determined. Total muscle lipids from the chicken cooked in fresh corn oil contained 1.0% eicosatrienoic acid (20:3). 1.6% linolenic acid. and 31.3% oleic acid. However. the total muscle lipids from chicken cooked in corn oil previously heated for 24 hours contained 1.2% linolenic acid. 30.7% oleic acid. and without any detectable amount of eicosatrienoic acid (20:3). The total muscle lipids from the chicken cooked in Corn oil previously heated was similar in both the 24 and 42 61 hours groups. except the former fats had a slightly lower oleic acid content (Table 2). Eicosatrienoic (20:3) and oleic acids declined but linoleic acid increased during the time chicken was cooked with reused corn oil. The occurrence of chemical reactions in heated corn oil is postulated to be initiated by the formation of heat - induced free radicals. and the rate of reactions may depend upon the concentration of free radicals formed. In other ‘words. the corn oil may contain more free radicals when heated for 42 hours than 24 hours. or than fresh corn oil. ‘When chicken.was cooked in these reused corn oils. it was (expected that the chemical reactions would occur at a faster rate. Thus the increased losses of eicosatrienoic and oleic tacids which occurred in chicken muscle fat when cooked in reused corn oil were expected. The apparent increase of linoleic acid may be due to the decreases of other acids on a relative basis . The cut-up-birds were packed in polyethylene bags. four birds per bag. then frozen at -37°C. and stored at -18°C for three or six months. 62 After three months storage . samples showed normal flavor and appearance without any observable deterioration. However. after six months storage. the cooked samples appeared darker in color. dried out. slightly toughened in texture. and less juicy. A rancid flavor was noticed at the moment the bag was opened. and a slight brown color deterioration in the dark meat was found. Uncooked samples showed a yellowish color on the skin and a slightly browned color in the dark meat. Cooked samples and uncooked samples were studied separately. The effects of frozen storage on the fatty acid composition of total muscle lipids from uncooked birds are shown in Table 2. The percentage of polyunsaturated acids including arachidonic (20:4). eicosatrienoic (20:3). linolenic. and linoleic acids decreased during the first three months of storage. After six months. linoleic acid further decreased. while oleic acid and linolenic acid decreased slightly. The total unsaturated fate of fresh chicken muscle was 70.1%. and decreased to 64.6% during three months. and to 61.3% after 313! months storage (Table 2). The decrease in unsaturation of fatty acids in chicken muscle during storage. agrees with results of Keskinel et al. 63 (1964). They reported oxidative deterioration during storage of raw beef. lamb. pork. and turkey. Various meat showed different degrees of oxidation which could be attributed to two major factors: the nature of lipids. especially the degree of unsaturation of fatty acids present. and heme catalysis. Quality of fried chicken decreased during six months storage. Some cooking oil was absoer by chicken pieces during cooking. and this absorbed reused corn oil. (which may contain various amounts of free radicals) may'have affected the stability of chicken. When chicken was cooked in fresh corn oil. the eicosatrienoic (20:3). and linolenic acids in total muscle fat decreased during three months storage. while linoleic acid continued to decrease during six months storage (Table 2). Decreasing unsaturated fatty acids during storage indicate oxidation of muscle fat. When chicken was cooked in corn oil previously heated for 24 hours. linolenic and linoleic acids decreased during both three and six months storage. When birds were cooked in corn oil previously heated for 42 hours. 64 linoleic acid decreased during three months storage. with a further decrease in linoleic acid and linolenic acid after six months storage (Table 2). This indicated that greater losses of polyunsaturated fatty acids occurred in the muscle fat from chicken cooked in reused corn oil than in fresh corn oil after storage. The absorption of reused corn oil in chicken. which contained more free radicals than fresh corn oua HHo Chou .m .Hao shoe smash .<::m 66 =70~bJfiO“O:tb. “01 Ge wed N“ e o c o o . fiOHédeo N mm 0 C O. O O . viOWDdWfiCand N MM N “V‘s-(@039! O I O O O O «I O\O\O\OO\o.-: 01 -&v4 \0 [\HMOxgxm 3mm mm H Igmcomo o.m0 0.0w 0 O O O Q Q INNMdO N mm ‘N CNOchaam O O O I . IONMr-OCOO mm \0 announaqt N e d'V\ m0 Noswmomo 03 -3v4 Amend xmoa vov esoonem o m < 3mm 01a) C 0 "Ma cac- tees. <3C>Chitfifl\b43 N mm O\NV\H\OHHU\ cncuuavnouaowa C O O O O O . CDOWDC\FKDOMH MM N 0.00 0.00 O O O I o . HOWQQCDCOH (“H N ourwfiraruwcwox 3mm m .Pdmfid HmPoa .000 H0000 0.0a 0.0a H.0H 0.0a H.0H 0.00 0.0a 0.0H 000000 00000 «escapees» ”no A.oec was» omesopm .cmmnovm cououm cam mcwxooo wsfimsc moaned Hope» same message no sodpdmocsoo chem haven SH newssso .0 oases 67 fatty acid composition of total skin fats from the cooked birds was a loss of polyunsaturated fatty acids (Table 3) especially linoleic acid. A more pronounced loss of polyunsaturated acids occurred in muscle fats than in skin fats. which may be due to the effects of absorbed fat. (2 ) . Phospholipids . (a). Muscle phospholipids: The predominant fatty acids in phospholipids from uncooked chicken muscle were linoleic. stearic. palmitic. arachidonic and oleic acids. which totaled 82.1% (Table 4). The total polyunsaturated fatty acids was about 45%. and arachidonic acid (20:4). was highest at 15.3%. The highly unsaturated fatty acid content and high levels of arachidonic acid and linoleic acid are characteristic of phospholipids. The composition of phospholipids from chicken have been reported by many workers (Katz et al.. 1966: Peng. 1965: Issacks et al.. 1964: Marion et al.. 1967) showing agreement with these results. .mcsop cansoc mo Mensa: . ascends mo 009252 .9 .mnsos 00 you 00000: hamsoa>oua ado suoo .0 .0050: 00 you 00900: hamsoa>ona ado smoo .m .Hao shoe £0000 .<::0 68 0.00 0.00 0.0m 0.0m 0.M0 0.00 0.m0 0.00 0.M0 0.00 0.00 0.00 .0000: 00000 0.00 0.00 0.0 0.0 0. n 0.00 0. 0 0.00 0. n 0.00 0.00 0.00 .000 00000 0.0 0.0 0.0 0.00 0.0 0.0 0.0 0.0 0.0 0.0 0.00 0.00 0.00 .. -u n- -- -u s: n- -- u- u- 0.0 0.0 0.00 n: :: :: I: :: H.o 00000 06009 :: :: 0.0 o.m n.0a 0.00 0.00 0.00 0.00 0.0 0.00 .00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.MH 0.00 0.00 .00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 m. H 0.0 0.0 0.0 0.00 0.0 0.0 0.00 0.00 0.00 0.0 0.0 0.0 .0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.00 -- u- u- u: u- u: u: 0.0 0.0 0.0 0.0 0.0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.0 0.0 0.00 0.00 0.0 0.0 0.0 0.0 0.00 -- u- -a u- u- -- :u u- 0.0 0.0 0.0 u: 0300000 0.0 0.0 0.0 0.0 0.0 00000 0.0 0.0 0.0 0.0 0.0 m.0 0.00 0.0 0.0 0.0 0.0 u- :. 00000 00009 0.0 0.0 0.0 .0 0.00 0.0 0.0 -- 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.00 A0000 amen oncv pseomem 900060 00000 < :00 0 0 < :00 0 0 4 300 0000000000 000 m c 5.080 0500 ewmuopm owdhopm scsoum 0:0 wsaxooo weamsc mcwaaaosamosa odomss message mo sowpfimoasoo 0000 awash :0 newsman .0 manna 69 Phospholipids were identified as constant elements by Terroine's classification (1920). However. it has been found that the so called "constant element” was not exact. since the composition of phospholipids may vary with dietary fat and protein (Marion and Woodroof. 1963: and Marion et al.. 1967). They pointed out that the most marked changes due to diet in the long chain fatty acids was in the substitution of 20:5 and 22:6 for 22 :4 when menhaden oil was fed. The fatty acids such as 2031, 20:2. 22:3. 22:“. which were reported by Katz et al. (1966) were not found in this study. presumably due to the different diet fed. Losses of arachidonic (C20,4). eicosatrienoic (020,3). linolenic (C180). and stearic acids were found in the muscle Phospholipids after the chicken was cooked in fresh corn oil. Linoleic acid and oleic acid increased, and a new unidentified Short chain acid also was found (Table 21,), The corn oil used for this study did not contain phospholipids. therefore. the absorption of corn oil by chicken pieces during cooking would not interfere with the composition of chicken phospholipids. The changes in 70 composition of muscle phospholipids during cooking is presumed to be due to chemical reactions. Losses of arachidonic. eicosatrienoic, and linolenic acids in chicken muscle phospholipids are probably due to the oxidative deterioration which occurs during cooking. One double bond present in the unsaturated acids is first attacked to form hydroperoxides as primary products. These. undergo a variety of scission and dismutation reactions to form a wide spectrum of carbonyl compounds. hydroxy compounds. and short chain fatty acids. The formation of carbomrl compounds may contribute to the aroma of cooked chicken (Bassette and Day. 1 960). The unknown short chain acid found after cooking was probably the product formed from the scission and dismutation reactions of hydroperoxides. Hydrolysis may also occur in chicken fat when heat is applied. The increased amounts of linoleic and oleic acids are probably related to the decreased amounts of higher polyunsaturated acids. since the percentage 01‘ all fatty acids amounted to 100%. When some fatty acids decreased in percentages. other fatty acids must increase on a percentage basis . 71 The muscle phospholipids from the chicken cooked in fresh corn oil contained 10.9% arachidonic (20th). 0.6% eicosatrienoic (20:3). and 0.6% linolenic acids (Table h). However. the muscle phospholipids from the chicken cooked in corn oil previously heated for 2h hours contained 9.7% arachidonic acid. and without any detectable amounts of eicosatrienoic acid and linolenic acid. The arachidonic acid (2034) further declined to 6.0% when the chicken.was cooked in corn oil previously heated for #2 hours. Obviously. the effects of cooking in reused corn oil on chicken muscle phospholipids was the greater losses of arachidonic acid (203“). eicosatrienoic acid (20:3). and linolenic acid. than from chicken cooked in fresh corn oil. The formation of free radicals and the presence of unsaturated acids and oxygen are the initiative agents for oxidative deterioration. The presence of free radicals in reused corn oil. the highly unsaturated fatty acids in chicken.phospholipids. and the heat from the cooking process. stimulate the oxidation reaction. Total unsaturated muscle phospholipids from uncooked 72 birds decreased during storage. Losses of arachidonic (20sh). eicosatrienoic (20:3). linolenic. linoleic. and stearic acids were found after the first three months of storage. Losses of linoleic and oleic acids were found after six months storage. however. the percentage of arachidonic acid (20th) increased during the period. Increasing amounts of short chain fatty acids were also found after storage (Table h). Decreases of unsaturated fatty acids in muscle phospholipids indicated that severe oxidation occurred during storage. The oxidative deterioration is supposed to be catalyzed by heme protein (Younathan and Watts. 1960). The ferric heme pigments are formed from the oxidation of meat pigments. oxymyoglobin and myoglobin. This lipid oxidation could be reduced by anaerobic packaging. which should reduce oxidation of myoglobin. and the enzyme activity in meat may establish anaerobic conditions quickly and completely to reduce formation of metmyoglobin (Greene. 1969). Arachidonic acid appeared to decrease during the first three months storage. and increase during the second three months storage. These unexpected changes may have been a 73 result of low numbers of samples analyzed or chance selection of a non-representative sample. or to preferential lipolysis. Arachidonic (20:4). eicosatrienoic (20:3). linolenic. palmitoleic. and stearic acids in muscle phospholipids from chicken cooked in fresh corn oil declined during three months storage. Further losses of linoleic and oleic acids were found after six months storage (Table a). The greater loss of unsaturated fatty acids in phospho- lipids than in total lipids indicated a more serious oxidation in the phoSpholipids fractions than in the total lipid extract. Rancid flavor was noticed after six months of storage. and pronounced losses of arachidonic. linoleic. and oleic acids were found. Changes in these acids are probably responsible for this quality deterioration. These acids are first oxidized to form hydroperoxides. then. undergo a variety of scission and dismutations to form a wide spectrum of carbonyl compounds. hydroxyl compounds. and short chain fatty acids. The hydroperoxides are essentially colorless. odorless and flavorless and have little effect on quality. However. 'the formation of carbonyl compounds are considered to be associated with rancid flavor (Berry and McKerrigan. 19588 7h Gaddis et al.. 1959). Much effort has been devoted to identification of catalysts in cooked animal tissues responsible for the oxidation of unsaturated lipids. The accelerating effect of hemoglobin and other iron porphyrins on the oxidation of lipids is a generally accepted phenomenon. and the hemo- proteins have been implicated as the major prooxidants in meat and meat products (Tappel. 1952: 19533 Younathan and Watts. 1959). Ferric hemochromogen is postulated to be an active catalytic form of the muscle pigments (Tappel. 1953; Younathan and Watts. 1959). In cooked meat. the pigment is in the active denatured ferric hemochromogen form. accounting for the rapid imitation of lipid oxidation. Tappel et al. (1962) proposed a mechanism for the prooxidant activity of hemes based on their known ability to decompose lipid peroxides. In this theory. free radicals. resulting from the peroxide scission. initiate new reaction chains. Other meat components have attributed catalytic roles in lipid oxidation. Some metals. especially ferrous from. present in meat in trace amounts. are efficient lipid oxidation catalysts. Sodium chloride has a puzzling effect 75 on oxidative changes in meat. Salt is believed to catalyze the oxidation of the stored triglycerides (Watts. 1962). Losses of arachidonic. stearic and palmitoleic acids were found in the muscle phospholipids of birds cooked in corn oil previously heated for an hours and stored three months. After a six month period. considerable losses of linoleic and oleic acids. and a slight loss of arachidonic (20:h) were found (Table fl). A similar loss of unsaturated acids in the muscle phospholipids of birds cooked in corn oil previously heated for #2 hours was also found after storage. Thus. the greatest loss of unsaturated fatty acids was found in the chicken cooked in corn oil previously heated for #2 hours. followed by that heated for 2h hours. and lowest in chicken cooked in fresh corn oil. In other words. the reused corn oil caused more loss of unsaturated fatty acids during storage. and may have contributed to the stability changes. (b). Skin phospholipids: The predominant fatty acids of phospholipids from uncooked chicken skin fats were oleic. linoleic. palmitic. and arachidonic acids (203“) and amounted to 8h.2% (Table 5). 76 The major difference between muscle and skin phospholipids was that muscle contained higher amounts of polyunsaturated acids including arachidonic acid (20:4). eicosatrienoic acid (20:3). linolenic acid. and linoleic acid. The percentage of arachidonic was lower in the lipid rich tissues (skin) than in the muscle lipids. The results obtained showed general agreement with other workers (Katz et al. 1966; Peng. 1965: Issacks et al.. 196a; Marion et al.. 1967) except for the higher percentage of oleic acid in skin phospholipids. Changes in composition of skin phospholipids after the chicken was cooked in fresh corn oil were decreases in arachidonic (20th). linolenic (18:3). oleic. and palmitic acids. and increases in linoleic acid and some short chain acids (Table 5). Lower amounts of arachidonic (201“) and linolenic acids were found in chicken cooked in reused corn oil than in that cooked in fresh corn oil (Table 5). Fat deteriorations were greater in skin phospholipids from the chicken cooked in reused corn oil. Changes in fatty acid composition of chicken skin phospholipids during Table 5 . cooking and frozen storage Storage time (mo.) Raw Raw Raw Oil treatmenta Percent (GLC peak area) Fatty acidsb 77 MO MHNommr—I I PINI--IOt--Iu---I¢"'\NI.N I I N Nd’ NONNNOH‘D‘H Hr-Ir-IOV-IOMNCD I I H N3 O\O O\V\V\1\0\O O \O O O O O O O O O 0 IO OHOGDr-II-INNO H N“ NMHWHWOCDV‘H HOMN MWBWQOH N (“r-I Md): O\-&®O\C\® I OOOOM-IOMNQ I I H Nd 0:} mac Od‘OM—INt-‘HO I IOOCQCONMHHO H N“ \OIN HNr-‘ICDOM-IQN [oomoooooooo IOO OQr-IHNI-IOOM OI-I N3 OKNMO Md {\MOWD IOONo-IM-S'BVNOON N MN 0 \O\O\OI-I\OV\I-I:IV\OO IMHOMHONNOHM u-I N-‘i'P Od’WWQVMDdl-IQM IHOOWOONNNOM H N“ N0\l\\00\\n cocoa-uh . O O HOOr-ILNN I-II'NMI-Ifi u-I Nd‘ co dH‘OdNCQI-fim I IN ImwNv—IQWOZDI-IH H (fir-I l-I OOOOOHOOHNM NMVMO‘ONQDGDCDQO t-II-Il-Ir-Ir-Ir'Ir-Ir-It-Ir-IHN 9 2 29. Total unsat. Total sat. Corn oil previously heated for 2# hours. : number of double bonds. Corn oil previously heated for #2 hours. Number of carbons Fresh corn oil. a-A. Bo Co b. 78 Unsaturated fatty acids in skin phospholipids from uncooked birds including arachidonic and oleic acids. decreased during storage (Table 5). Losses in unsaturated fatty acids. including arachidonic and linolenic acids. were found in skin phospholipids from cooked birds after frozen storage. Greater losses of such unsaturated fatty acids was also found in skin phospholipids in chicken cooked in reused corn oil than in the phospholipids in chicken cooked in fresh corn oil (Table 5). (3). Neutral lipids: (a). Muscle neutral lipids: The predominant fatty acids in neutral lipids from uncooked chicken muscle were palmitic. oleic. and linoleic acids and amounted to 79.6%: 27.9% of the muscle neutral lipids were polyunsaturated (Table 6). These results agreed with those of Katz et al.. 1966: Peng. 1965; Issacks et al.. 196#: and Machlin et al.. 1962 on the predominant fatty acids of neutral lipids from muscle fat. which were reported as palmitic. oleic. and linoleic acids. .00000 000000 00 00900: . 0009000 00 000052 .0 .0050: N0 000 00000: 0H0200>000 H00 0000 .o .0000: am 000 000000 mamsoa>00n H00 0000 .m 79 .H00 0000 s000m .<::0 0.00 .00 0.00 0.00 0. a m.ma 0.m0 0.00 0. 0 0.00 0.00 0.00 .00000 00000 0.00 .00 0.00 0.00 m. m 0.00 0.00 a.mm 0. 0 0.00 0.00 0.00 .000 H0000 :: u: u: n: u: n: u: :: :: :: m.o 0.0 0.00 n: n: u: u: u: :: n: :: :: u: 0.0 0.0 0.00 ~.H m.H o.H m.a 0.0 0.0 0.0 H.H 0.0 m.a 0.0 0.H m.ma 0.00 0.0m 0.00 .00 0.0m o.nm 0.0n 0.00 0.0m H.0m a.mm 0.00 0.00 0.0m 0.0m 0.0m a.mm 0.0m 0.00 0.00 0.00 0.0m 0.0m 0.00 0.00 0.00 0.0 0.0 0.0 0.0 0.0 m.0 0.0 0.0 H.m 0.0 0.0 0.0 0.00 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 m. m.0 0.0 0.0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 .00 0.00 0.00 0.00 0.0 0.0 0.0 0.0 :: :n u: 0.0 00000 00000 0.0 0.0 0.00 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.00 :: u: :: :: :: u: a: u: u: :- u: 0.0 a.ma -: u: :: -: n: :: u: u: u: a- :: n.o 0.00 A0000 0000 oAuV 0000000 0.3000 00000 0 m 0 :00 0 m 0 300 0 m 0 :00 0000000000 000 0 m o A.osv 0000 0000000 000030 00000H H000sms 0aomss 0000000 0o 00000000200 0000 00000 :0 0000000 0000000 000000 000 0000000 .0 00000 80 The neutral lipids in animal tissues are formed both from the biosynthetic pathway and are directly formed from absorbed dietary fat through the lymphatic system (Weiss and Kennedy. 1956), therefore. dietary fat may affect and reflect the composition of neutral lipids. Issacks et al. (196k) reported that triglycerides reflected more closely the dietary fatty acids than did the phospholipids. The major difference between phospholipids and neutral lipids in amounts of fatty acids was that phospholipids contained almost 1.5 times as much polyunsaturated fatty acids as neutral lipids. Of particular interest was the amount of arachidonic acid (zouh) of which 15% was found in phospho- lipids but only 0.5% was found in neutral lipids. The changes in muscle neutral lipids during cooking in fresh corn oil were mainly increases in linoleic acid and a slight decrease in arachidonic (ZOsh). eicosatrienoic (20:3). linolenic (18:3). oleic and stearic acids (Table 6). These changes. are presumed to be due to chemical reactions. physical absorption of corn oil. and loss of rendered fat from chicken. 81 When the comparison between the decrease of polyunsaturated fatty acids in muscle phospholipids and in muscle neutral lipids was made, the decrease in neutral lipids was very slight. This suggests that the deterioration during cooking occurred faSter in phospholipids. This may result from phospholipids having high unsaturated fatty acids, particularly high levels of arachidonic and linoleic acids. A comparison between the chicken cooked in corn oil previously heated for 24 hours with the chicken cooked in fresh corn oil showed that arachidonic (203“) and linolenic acids decreased. Again. a comparison between the chicken cooked in corn oil previously heated for #2 hours and that heated for 2h hours was made and showed that linolenic and linoleic acids decreased (Table 6). Obviously. the most severe deterioration was found in neutral lipids from the chicken cooked in corn oil previously heated for 42 hours, followed by that heated for 2h hrs. Chicken cooked in fresh corn oil was least affected. Greater losses of polyunsaturated acids was indicated in phospholipids than in neutral lipids due to cooking in reused corn oil. This shows that phospholipids are more 82 readily susceptible to stress than neutral lipids and are more important in determining meat quality. Unsaturated fatty acids from the uncooked birds. including arachidonic (20xh), eicosatrienoic (20:3). linolenic. and linoleic acids. in muscle neutral lipids decreased during storage (Table 6). Comparing phospholipids and neutral lipids. it is obvious that the loss of unsaturated fatty acids were more pronounced in the phospholipids fraction than in the neutral lipids. The oxidation of tissue lipids seems to occur in two stages: the phospholipids are oxidized first. and the neutral lipids oxidized later. It is not known why the phospholipids are more susceptible to oxidation than neutral lipids. but a possible explanation may lie in the fact that the fatty acids of the complex lipids are more unsaturated than those of trigly- cerides. Also. the former may be more closely associated with the iron-containing heme compounds of meat tissue. which may act as prooxidants (El-Gharbawi and Dugan. 1965). Losses in polyunsaturated fatty acids including arachidonic (20:4), and eicosatrienoic acids (20:3) were 83 observed in muscle neutral lipids from the chicken cooked in fresh corn oil during frozen storage (Table 6). The decrease in these polyunsaturated fatty acids was greater in phospholipids than in neutral lipids during storage. This indicates that chemical deteriorations apparently first occur in phospholipids. The effects of reused corn oil on neutral lipids during storage is shown in Table 6; a more pronounced decrease of linoleic acid was indicated. Obviously. more serious fat deteriorations occur in chickens cooked in reused corn oil than in fresh corn oil during storage. and phospholipids were more affected by reused corn oil than were neutral lipids. (b). Skin neutral lipids: The major fatty acids in skin neutral lipids from uncooked chicken were oleic, palmitic. and linoleic acids (Table 7). The relative amounts of fatty acids in neutral lipids from skin and muscle were similar. The former lipids had slightly higher percentages of oleic and palmitic acids but a lower percentage of linoleic 8h acid. A pronounced increase in linoleic acid, 18.0% to 41.5%, was found in skin neutral lipid after cooking in fresh corn oil. whereas. the other fatty acids decreased. This pattern of change may be partly due to chemical reactions in chicken fat. and may also be from the absorption of corn oil and rendering of fat from chicken during cooking. Cooking chicken in reused corn oil from the chicken skin neutral lipids resulted in a loss of linoleic acid. Slight losses of linolenic and linoleic acids in skin neutral lipids from uncooked birds were found during storage (Table 7). This seems to suggest that the skin neutral lipids are more stable than the muscle neutral lipids. Changes in skin phospholipids were more pronounced than in skin neutral lipids. Linoleic acid content of skin neutral lipids from cooked birds, declined during frozen storage. The loss of unsaturated fatty acids in the skin neutral lipids from chicken cooked in reused corn oil was greater than from chicken cooked in fresh corn oil (Table 7). Since reused meson mansoc mo hopssc . msonnmo mo nonssz .p .muso: «a now posse: hamSow>oun Hwo shoe .0 .wuson am pom poems: mausoa>mna Hwo :hoo .m .HHo shoe :mopm .m «nanosonhoapanamonaoumq saaohsomswcmm osddogoflhoaaannuonm ocauouasenpaenuosm andsuaosdnpoaavapdsnmosm acanaaogamonn mas“: o o H 5o.o 00.0 0o.o « 0H.o 0a.o 0H.o m 0«.o 0«.o 0«.o s 0:.o 5:.0 5s.o 0 o0.o as.o 5:.o 0 «5.0 05.o 5 vandaosnuonn no commune madauovus udnflao:Amona unawaonamonn pogo assesses «0 cm can» «one uflouaa.uoum no .02 .csgaflognuonm do ”sundae emaeapcmuH .0 manna 88 sphingomeyelin were the predominant components of phospho- lipids from uncooked muscle fat and amounted to 85.2% of phospholipids. The high amounts of phosphatidylcholine and phosphatidylethanolamine is characteristic of chicken.muscle fat (Table 9). These results are in agreement with those reported by Davidkova and Khan (1967) except for the slightly lower phosphatidyl serine and slightly higher sphingomyelin. The total phosphorus content of phospholipids declined from 10.71 to 6.81 mg/g fat. or approximately h mg (Table 9) of phosphorus was lost during cooking in fresh corn oil. In other words. about 100 mg of’phospholipids (4 mg x 25) lest per 1 gram of fat. The loss of phospholipids may be due to both the chemical deterioration and physical rendering of fats from muscle during cooking. The chemical deterioration of phospholipids may be characterized as autoxidation. hydrolytic decomposition. lipid ”browning” reactions. and lipid-protein.co-polymerization reactions. Each component of phospholipid declined during cooking. among them. phosphatidylcholine declined greatest in.quantity. followed by sphingomyelin. lysophosphatidylcholine. 9 00 .mpg N3 you coeds: hamsoa>opn ago shoe .0 .mp3 3N mom mouse: hamsOfi>oun Hwo cmoo .m .Hwo cuoo smeum .m saddens 00000 0000» «0.0 0s.0 ««. 0H.0 3«.0 00.0 35.0 00.0 3«.0 05.0 nH00apagsmogsowza 00.0 a«.0 0«.0 00.0 0H.0 0H.0 00.0 00.0 moss» «H.0 00.0 00.H 00H0s50000000 3H.N 3m.~ 00.N m3.m 0m.« m0.« mm.m m5.n mo.m -.m «N.m 00.3 mcHaonoahofivmngmonm nu nu nu uu un uu uu moss» some» moss? moss» 00.0 oswpomahofipmsamonm esfismao 5H.H 5H.H 5m.a 00.m Hm.a 3m.H 00.H 00.N mm.H 5«.« Hm.« 00.« ussnveahofipmnnmonm mossy moss» 0H.0 HH.0 53.0 33.0 30.0 00.0 «0.0 30.0 03.0 03.0 mofiaflaosnmosa has“: memsaaoss Avmm M\m wev vcepsoo masonmmonm among 90 mommmao o m < 3mm o m < sum 0 m < 30m upsesvsoup H00 0 m .J o A.osv med» ewsuopm ommpovm canopy 0cm wCaxooo mcfisso mownwa saunas soxodno mo psepnoo nflafiaosgmonn cw mewsmno .m wands 9O phosphatidylserine. and phosphatidylethanolamine (Table 9). It appears that oxidation and polymerization are the main reactionsinvolved during cooking. Greater loss of phospholipids was found in the chicken cooked in corn oil previously heated for #2 hours than in fresh corn oil. The losses of phosphatidylethanolamine. sphingomyelin. phosphatidylcholine (Table 9) indicated that phosphatidylethanolamine and sphingomyelin are more sensitive to cooking in reused corn oil. Since lysophOSphatidylcholine is the hydrolytic product of phosphatidylcholine (Hanahan et al.. 195#). its increase and the decrease in phosphatidyl- choline strongly showed that hydrolysis occurred in.muscle phospholipids during cooking in reused corn oil. More serious fat deterioration occurred when chicken.was cooked in corn oil previously heated for #2 hours. than in fresh corn oil or that heated for 2# hours. Fresh uncooked chicken muscle fats contained 10.71 mg of phosphorus per gram of fat. This decreased to 7.05 mg after three months storage. and to 6.30 mg after six months storage (Table 9). This indicates that phospholipids were broken 91 down during storage. The mechanism of phospholipid deterioration during frozen storage is not clear. It is likely that several complicated reactions may be involved. They include oxidation. lipid "browning” reactions. lipid- protein co-polymerization reactions and lipolysis or enzymatic degradation. Changes in quantities of each component from muscle phospholipids. showed decreases of phosphatidylethanolamine. phoSphatidylserine. phosphatidyl- choline. sphingomyelin. lysophosphatidylcholine. and minor phospholipids. but increases in the component which may include non-lipid phosphorus. peroxides and hydroxy compounds during the first three months. indicating oxidative deterio- ration. Phosphatidylcholine and lysophosphatidylcholine continued to decrease. whereas phosphatidylethanolamine and Sphingomyelin did not decrease after three months of storage. Lysophosphatidylcholine decreased during storage. This disagrees with the work of Davidkova and Khan (1967) who reported an increase. The samples studied by Davikova were vacuum packed and oxidative degradation of the phospholipids in the samples was slow. with most deterioration due to 92 lipolysis. However. the samples in this study were stored in polyethylene bags without vacuum sealing. and thus oxidative deterioration of phospholipids coupled with lipolysis were presumed to occur. The total phosphorus in muscle fat from the birds cooked in fresh corn oil was 6.81 mg per gram. and decreased to 5.91 and #.#8 mg after three and six months storage respectively (Table 9). The mechanisms for the destruction of phospholipids during frozen storage are not clear. The reactions involved may be oxidation. lipid ”browning". and lipid-protein co-polymerization. The changes in each component during storage showed a loss of phosphatidylethanol- amine during the first three months. and losses of phosphatidylethanolamine. phosphatidylcholine. sphingomyelin. and lysophosphatidylcholine during the second three months. Since phosphatidylethanolamine has been related to browning deteriorations. its decrease may have affected the dark color in chicken found after storage. A greater decrease in total phosphorus from the chicken cooked in reused corn oil than in fresh corn oil during storage. is shown in Table 9. Phosphatidylethanolamine. 93 phosphatidylcholine. lysophosphatidylcholine. and sphingomyelin decreased continuously through storage. This may indicate oxidative deterioration. In summary. chicken cooked in reused corn oil was less stable during storage than that cooked in fresh corn oil. (b). Skin phospholipidsn The predominant components of phospholipids from uncooked skin fats were phosphatidylcholine. sphingomyelin. and phosphatidylethanolamine. Skin fats had a high level of phosphatidylcholine and sphingomyelin instead of phosphatidylcholine and phosphatidylethanol- amine as found in muscle fats (Table 10). In contrast to muscle phospholipids. the total phosphorus of skin.phospholipids increased from 0.50 to 0.73 mg/g fat after cooking in fresh corn oil (Table 10). Except for sphingomyelin. increases of other components were found. The apparent increase of phospholipid in skin lipids is presumed to be due to the loss of rendered fat which mainly was neutral lipids. 9# Total phospholipid content of skin lipids decreased slightly in the chicken cooked in the corn oil previously heated for #2 hours (Table 10). The decrease in phosphatidylcholine but increases in the component which may include non-lipid phosphorus. peroxides. hydroxy compounds and other molecules with strong polar groups. indicates oxidative deterioration. Thus chemical reactions which occur in chicken may vary with different tissues and oil treatments. Fresh uncooked chicken skin fats contained 0.50 mg lipid phosphorus per gram. After three and six months storage. lipid phosphorus decreased to 0.#8 and 0.29 mg respectively. The loss of phospholipids was due mainly to a loss of phosphatidylcholine and sphingomyelin. coupled with a minor loss of phosphatidylethanolamine. Since lysophosphatidyl- choline is a hydrolytic product of phosphatidylcholine. lysophosphatidylcholine increased slightly and phosphatidyl- choline decreased. This suggests that some degree of lipolysis occurred in skin phospholipids during storage (Table 10). The hydrolytic enzymes could be affected by many .00: «3 you 00900: 0H03o0>00n A00 snoo .0 .00: 3m you 00900: hamso0>oun 00o 2000 .m .H0o £000 £0000 .m os0aono 00.0 30.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 n0000000000000000 50.0 00.0 50.0 00.0 50.0 00.0 00.0 00.0 00.0 00.0 00.0 30.0 0000020000000 0«.0 0«.0 ««.0 «0.0 0«.0 0«.0 0«.0 0«.0 «0.0 50.0 50.0 3«.0 0:00000000000000000 nu nn nn nn nn nn nn nn nn nn soapy cosy? 0:0000H000psnnmonm 2.00:3.”0 00.0 00.0 00.0 50.0 00.0 00.0 00.0 00.0 00.0 00.0 «0.0 00.0 n00000000000000000 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 «0.0 «0.0 00.0 nu 0000000000000 00:02 000000000 0900 w\a wsv psopCoo msponmmonm nmonn no mommsao 0 0 < 300 0 0 < 300 0 0 0 300 0000200000 000 0 m o 0.080 e80» ommsopm omsuopm sououm 0cm ws0xooo w:0050 000900 20x0 sexo0co mo vampsoo 00Q00onnmosn s0 mowssno .00 00909 96 factors. such as the build up of substrates. oxidative reactions resulting in inhibition and concentration of solutes during freezing and frozen storage. These factors may affect the mode of attack or the orientation of substrates. which in turn. affect the selectivity and rate of hydrolysis (Braddock and Dugan. 1972). Chemical changes in chicken during storage may vary with type of packaging and different tissues. Phosphatidylcholine. sphingomyelin. and lysophosphatidyl- choline in skin lipids from cooked birds decreased (Table 10). indicating specific chemical changes during storage. 3. Changes in Fatty Acid Composition of Cooking Fat during Cooking with Chicken Linoleic acid decreased during prolonged heating. from 59.5% in fresh corn oil. to 50.0% in oil heated for #8 hours (Table 11). These results showed that linoleic acid declined slightly less during cooking of chicken than when the corn oil was cooked with moist cotton balls. Possible explanations aren (1). fresh corn oil was added daily. (2). a pressure 97 cooker was used. (3). chicken fats were released during cooking. and these fats contained highly unsaturated fatty acids. which oxidized faster than corn oil. Table 11. Changes in fatty acid composition of corn oil during cooking with chicken Heating times (hrs.) Fatty acidsa 0 6 30 #8 Percent (GLC peak area) 1#:0 -- 0.# 0.3 0.7. 15:0 -- trace -- -- 16:0 1#.8 15.# 16.7 19.2 1631 ." 0.2 0.8 1.8 1830 1.8 2.2 1.6 1.6 18:1 2#.0 2#.1 25.3 26.? 18:2 59.5 57.5 55.2 50.0 18:3 trace 0.2 -- ~- TOtal sat. 16.6 18.0 18.6 21.5 Total unsat. 83.5 82.0 81.3 78.5 a. Number of carbons: number of double bonds. -_ .n. ‘0 n SUMMARY AND CONCLUSIONS Chemical changes in heated commercial corn oil were determined after 6. 2#. 36 and #8 hours at 200°C. Moist cotton balls were fried in the corn oil at frequent intervals. Increases in free fatty acids. viscosity. absorbance values. peroxide values. and non-urea adducting fraction. and decreases in iodine value and loss of linoleic acid were related to heating times. These results indicate that hydrolysis. oxidation. and polymerization took place simultaneously during the course of heating. Three stages in rate of chemical reactions were observed. During the first stage (2# hours of heating) the chemical reactions occurred at a slow rate. During the second stage. (2# to 36 hours) the chemical reactions proceeded at a more rapid rate. and during the third stage (after 36 hours) the chemical reactions proceeded at the fastest rate. These results were used to establish cooking oil treatments for subsequent experiments. 98 99 The fatty acid composition of chicken fats from uncooked birds was evaluated. The predominant fatty acids. in both muscle and skin fats were oleic. linoleic and palmiticacids. The differences in composition.between.muscle and skin.fats were that muscle fats contained highly unsaturated acids including arachidonic (20:3) and eicosatrienoic acids (20:3). however. there were no detectable amounts of these acids present in skin fats. Phospholipids were separated from neutral lipids by thinplayer chromatography. The predominant fatty acids in both muscle and skin phospholipids were oleic. linoleic. arachidonic and palmitic acids. Stearic acid was also a major component in muscle phospholipids. The muscle phospho— lipids contained greater amounts of polyunsaturated acid including arachidonic (20:10. eicosatrienoic (20:3). linolenic and linoleic acids than skin phospholipids. Characteristics of phospholipids found were the highly unsaturated fatty acid content and high levels of arachidonic and linoleic acids. The predominant fatty acids in both muscle and skin neutral lipids were oleic. linoleic and palmitic acids. The neutral lipids from various tissues were similar in the 100 relative amounts of fatty acids. The amounts of polyunsaturated fatty acids in neutral lipids were approximately half of that found in.phospholipids. The classes of phospholipids found in.muscle and skin were phosphatidylethanolamine. phosphatidyl serine. phosphatidylcholine. sphingomyelin. and lysophosphatidyl- choline. Phcsphatidylcholine. phosphatidylethanolamine. and sphingomyelin.were the major components. The fat composition of chicken cooked in fresh corn oil was evaluated. The fatty acid analysis of total fats and neutral lipids. including muscle and skin. showed an increase in linoleic acid but a decrease inumost other fatty acids after cooking. Corn oil. used as cooking oil. contained 58$ linoleic acid. Therefore. the changes in fatty acid composition in total fats and neutral lipids were presumably due mostly to the absorption of corn oil. loss of rendering fat from chicken tissues. and chemical reactions. The fatty acid analysis of phospholipids showed pronounced losses of arachidonic (20n#). eicosatrienoic (20:3) and linolenic acids in muscle phospholipids and pronounced losses of arachidonic. linolenic. and oleic acids in skin 101 phospholipids during cooking in fresh corn oil. The loss of unsaturated fatty acids in phospholipids was presumed due to the heat induced oxidative deterioration of the unsaturated double bonds during cooking. It appears that oxidative deterioration occurred first with phospholipids. The total phosphorus content in muscle lipids declined from 10.71 to 6.81 mg/g fat during cooking. however. the total phosphorus in skin lipids increased from 0.50 to 0.73 mg/g fat. This suggested that some phospholipids in skin ‘ fats came from the release of muscle phospholipids. and various chemical reactions which occurred during cooking. The effects of reused (Heated) corn oil on the composition of chicken fats during cooking were studied. Chickens were cooked in fresh corn oil. and corn oil previously heated for 2# hours. or #2 hours. Arachidonic (20:#). eicosatrienoic (20:3) and linolenic acids were lower in both muscle and skin phospholipids from the chicken cooked in reused corn oil than those cooked in fresh corn oil. The decreases in unsaturated fatty acids were greater in phospholipids than in neutral lipids. This suggests that phospholipids in chicken tissues are more 102 sensitive to change when the chicken is cooked in reused corn oil. probably due to the fact that phospholipids contain higher amounts of highly unsaturated fatty acids. Lower values for phosphatidylethanolamine. sphingomyelin. and phosphatidylcholine were found in muscle phospholipids when chickens were cooked in reused corn oil than in those cooked in fresh corn oil. This suggests that phosphatidyl- ethanol and sphingomyelin are more sensitive to change in reused corn 011 during cooking. and fat deterioration during cooking was greater in the chicken cooked in reused corn oil than in fresh corn oil. Uncooked chickens had a yellowish skin color and slightly browned pigmented dark meat after six months frozen storage. The values for arachidonic (20:#). eicosatrienoic (20:3). linolenic. and linoleic acids were lower in the muscle total fats and muscle neutral lipids of chickens after three months of frozen storage than in fresh chickens. Further losses of linolenic. linoleic and oleic acids were found after six months storage. Skin total fats and skin neutral lipids had lower values of linolenic acid after three months storage than 103 fresh. and a further loss of linoleic acid was found after six months storage. The loss of such unsaturated fatty acids was greater in muscle fats than in skin fats. Arachidcnic (20:#). eicosatrienoic (20:3). linolenic. linoleic and stearic acids in muscle phospholipids were lower after three months storage than in fresh tissues. ~Further losses of linoleic and oleic acids were found after six months storage. Similar changes were found in skin phospholipids. Total phosphorus decreased from 10.71 to 6.30 mg/g fat in muscle fats. and from 0.50 to 0.29 mg/g fat in skin fats after six months storage. All components of phospholipids in muscle fats decreased during three months storage. and further decreases in phosphatidylcholine. lysophosphatidyl- choline were found after six months storage. These indicate that oxidative deterioration occurred during storage. Decreases in phosphatidylcholine and sphingomyelin and an increase in lysophosphatidylcholine were found in skin fats after storage. which indicates that lipolysis was involved. Precooked chickens were darker in color. dehydrated. slightly tougher in texture. less juicy. and had a rancid 10# flavor after six months of frozen.storage. How these changes relate with chicken fat. and the effect of reused corn oil on chicken fats during storage were studied. Unsaturated fatty acids declined in both muscle and skin fats during three and six months frozen.stcrage. A greater loss of unsaturated fatty acids was found in the chicken.fats which were cooked in reused corn oil than in those cooked in fresh corn oil. A.more pronounced loss of unsaturated fatty acids was found in.muscle fats than in skin fats. Unsaturated fatty acid losses were more pronounced in the phospholipids than in the total or neutral lipids. Phospholipids contain.more highly unsaturated fatty acids than neutral lipids. and the former may be more closely . associated with the ironpcontaining heme compounds of meat tissue. Decreases of arachidonic and linoleic acids in phospholipids may be responsible for the develOpment of rancid flavor from precooked chicken after six months storage. Phosphorus content of phospholipids decreased during frozen storage. The mechanism for the deterioration of phospholipids during storage is still not clear. It may 105 possibly include various reactions such as oxidation. lipid "browning" reactions. and lipid-protein co-polymerization. The changes in each component were a loss of phosphatidyl- ethanolamine during the first three months. and the losses of phosphatidylethanolamine. phosphatidylcholine during the second three months. Since phosphatidylethanolamine has been related to browning deteriorations. a decrease of phosphatidylethanolamine may have affected the dark color in chicken found after storage. Fat deterioration during frozen storage was greater in the chicken cooked in reused corn oil than in fresh corn oil. The use of reused corn oil resulted in less stable products. PROPOSAL FOR FUTURE RESEARCH Fat deterioration has been found in chicken due to cooking. frozen storage. and cooking in reused cooking 011. They were losses of unsaturated fatty acids and phospholipids. To find out how to prevent fat deterioration in chicken is the objective for future research. They are preposed as follows: (1). Vacuum or nitrogen gas packed. (2). The use of antioxidants. (3). The use of coating materials such as ice coating on chicken pieces before frozen storage. 106 LITERATURE CITED Andrew, J. S., W. H. Griffith, and R. A. Stein, 1960. Toxicity of air-oxidized soybean oil. J. Nutr. 70:199. Arenson, S. W., and B. G. Heyl, 1943. Influence of various liquid and solid vegetable shortenings upon doughnut physical characteristics. Oil and Soap 20(8):l49. Bassette, R., and B. A. Day, 1960. Regeneration of carbonyl compounds from 2,4-dinitrophenylhydrazones with sulfuric acid. J. Am. Oil Chemists' Soc. 37:482. Bell, R. D., and B. A. Doisy, 1920. Rapid colorimetric methods for the deterioration of phosphorus in urine and blood. J. Biol. Chem. 44:55. Berry, N. W., and A. A. MbKerrigan, 1958. Carbonyl compounds are a criterion of flavor deterioration in edible fats. J. Sci. Food Agr. 9:693. Bloor, W. R., 1914. A method for determination of fat in small amounts of blood. J. Biol. Chem. 17:377. Bloor, W. R., 1915. A method for determination of lecithin in small amounts of blood. J. Biol. Chem. 22:133. Boyle, J. J., and B. H. Ludwig, 1962. Analysis of fatty acids of continuously cultured mammalian cells by gas-liquid chromatography. Nature 196:893. Braddock, R. J., and L. R. Dugan, Jr.. 1972. Phospholipid changes in muscle from frozen stored Lake Michigan Coho Salmon. J. Food Sci. 37:426. Broiler Industry, 1971. KFC: Will growers have chance to share in its profits ? Broiler Industry 10:19. Carlin, G. T., and B. Lannerud, 1941. Rancidity problems in baked and fried foods. Oil and Soap 18(3):60. Carlin, G. T., R. P. Hopper, and B. N. Rockwood, 1954. Some factors affecting the decomposition of frying fats. Food Technol. 8:161. Cash, D. B., and A. F. Carlin, 1968. Quality of frozen boneless turkey roasts precooked to different internal temperatures. Food Technol. 22:1477. / 0 175108 109 Chang, 1., and B. M. Watts, 1950. Some effects of salt and moisture on rancidity in fats. Food Res. 15:313. Chang, 1., and B. M. Watts, 1952. The fatty acid content of meat and poultry before and after cooking. J. Am. Oil Chemists' Soc. 29:334. Chang, P. Y., M. T. Younathan, and B. M. Watts, 1961. Lipid oxidation in precooked beef preserved by refrigeration, freezing, and irradiation. Food Technol. 15:168. Crampton, B. W., R. H. Common, F. A. Farmer, H. F. Wells, and D. Crawford, 1951. Studies to determine the nature of the damage to the nutritive value of some vegetable oils from heat treatment. 11. Investigation of the nutritiousness of the products of thermal polymerization of linseed oil. J. Nutr. 44:177. Crampton, E. W., R. Common, F. A. Farmer, F. W. Berryhill, and L. Wiseblatt, 1952. Studies to determine the nature of the damage to the nutritive value of some vegetable oils from heat treatment. J. Nutr. 44:177. Cropper, F. R., and A. Heywood, 1953. Analytical separation of the methyl esters of the 012-0 fatty acids by vapour-phase chromatography. Nature 17§?1101. Davenport, J. B., 1964. The phospholipids of pigeon and ox skeletal muscle. Biochem. J. 90:116. Davidkova, B., and A. W. Khan, 1967. Changes in lipid composition of chicken muscle during frozen storage. J. Food Sci. 32:35. Day, E. A., D. A. Lillard and M. W. Mbntgomery, 1963. Autoxidation of milk lipids. III. Effect on flavor of the additive inter- actions of carbonyl compounds at subthreshold concentrations. J. Dairy Sci. 46:292. DeHaas, G. H., I. Mulder and L. L. M. VanDeenen, 1960. On the specificity of phospholipase A. Biochem. Biophys. Res. Comm. 3:287. Desai, I. D., and A. L. Tappel, 1963. Damage to proteins by peroxidized lipids. I. Lipid Res. 4:204. Dimiok, P. S., and J. H. MacNeil, 1970. Poultry product quality. 2. Storage time-temperature effects on carbonyl composition of cooked turkey and chicken skin fractions. J. Food Sci. 35:186. 110 Dugan, L. R.,Jr., 1968. Processing and other stress effects on the nutritive value of lipids. World Review of Nutrition and Dietetics. 9:181. El-Gharbawi, M. I., and L. R. Dugan, Jr., 1965. Stability of nitro- geneous compounds and lipids during storage of freeze—dried raw beef. J. Food Sci. 30:817. Ellis, R., G. T. Currie, F. E. Thornton, N. C. Bollinger, and A. M. Gaddis, 1968. Carbonyl in oxidizing fat. 11. The effect of the prooxidant activity of sodium chloride on pork tissue. J. Food Sci. 33:555. Entenman, C., 1957. General procedures for separating lipid components of tissue. In Methods in Enzymology, Vol. 3, S. P. Clowich and N. Kaplan, Eds., Academic Press, New York, p 299. Fiske, C. H., and Y. Subbarrow, 1925. The colorimetric determination of phosphorus. J. Biol. Chem. 66:375. Fleischman, 1., A. Fitzgerald, B. A. Caldwell, and G. Eastwood, 1963. Studies on cooking fats and oils. J. Amer. Diet. Ass'n. 42:394. Folch, J., M. Lees, and G. M. 8. Stanley, 1957. A simple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem. 226:497. Gaddis, A. M., R. Ellis, and G. T. Currie, 1959. Carbonyls in oxidizing fat. 1. Separation of steam volatile monocarbonyls into classes. Food Res. 24:283. Glavind, J., and N. Tryding, 1960. On the digestion and absorption of lipoperoxides. ACTA Physiol. Scand. 49:97. Goodman, A. H., and Z. Block, 1952. Problems encountered in the commercial utilization of frying fats. J. Am. Oil Chemists' Soc. 29(12):616. Goodwin, T. L., V. D. Bramblett, G. E. Vail, and W. J. Stadelman, 1962. Effects of end—point temperature and cooking rate on turkey meat tenderness. Food Technol. 16(12):101. Greene, 8. B., 1969. Lipid oxidation and pigment changes in raw beef. J. Food Sci. 34:110. Guthrie, H. A., 1967. Introductory Nutrition. The C. V. Mosby Company, St. Louis. p 39. Hanahan, D. J., M. Rodbell, and L. D. Turner, 1954. Enzymatic formation of monopalmitoleyl- and monopalmitoyllecithin (lysolecithins). J. Biol. Chem. 206:431. 111 Hanahan, D. J., J. C. Dittmer, and E. Warashina, 1957. A column chromatographic separation of classes of phospholipids. J. Biol. Chem. 228:685. Hanahan, D. J., H. Brockerhoff, and E. J. Barron, 1960. The site of attack of phospholipase A on lecithin. J. Biol. Chem. 235:1917. Hanson, H. L., M. Winegarden, M. B. Harton, and H. Lineweaver, 1950. Preparation and storage of frozen cooked poultry and vegetables. Food Technol. 4:430. Hanson, H. L., 1954. Proc. Inst. Am. Poultry Ind., Fact Finding Conference. Hanson, H. L., and L. R. Fletcher, 1958. The time-temperature tolerance of frozen foods. XII. Turkey dinners and turkey pies. Food Technol. 12:1. Hanson, H. L., L. R. Fletcher, H. Lineweaver, 1959. Time-temperature tolerance of frozen foods. XVII. Frozen fried chicken. Food Technol. 13:221. Harris, N. D., and R. C. Lindsay, 1972. Flavor changes in reheated chicken. J. Food Sci. 37:19. Issacks, R. E., R. E. Davies, T. M. Ferguson, R. Resister, and J. R. Couch, 1964. Studies on avian fat composition. 1. Effect of dietary fat on the fatty acids of the triglyceride and phospholipid fraction of the blood plasma and adipose tissue lipids of the laying hen. Poultry Sci. 43:105. Johnson, 0. C., and F. A. Kummerow, 1957. Chemical changes which take place in an edible oil during thermal oxidation. J. Am. Oil Chemists' Soc. 34:407. James, A. T., and A. J. P. Martin, 1952. Gas-liquid partition chromatography: The separation and micro-estimation of volatile fatty acids from formic acid to dodecanoic acid. Biochem. J. 50:679. James, A. T., and J. P. Martin, 1956. Gas-liquid chromatography: The separation and identification of the methylesters of saturated and unsaturated acids from formic acid to n- octadecanoic acid. Biochem. J. 63:144. Katz, M. A., L. R. Dugan, Jr., and L. E. Dawson, 1966. Fatty acids in neutral lipids and phospholipids from chicken tissue. J. Food Sci. 31:717. 112 Kaunitz, H., C. A. Slanetz, R. E. Johnson, H. B. Knight, D. H. Sounders, and D. Swern, 1956. Biological effects of the polymeric residues isolated. J. Am. Oil Chemists' Soc. 33:630. Kawada, T., B. G. Krishnamurthy, B. D. Mbokerjee, and S. 5. Chang, 1967. Chemical reactions involved in deep fat frying. 11. Identification of acidic volatile decomposition products of corn oil. J. Am. Oil Chemists' Soc. 44:131. Keskinel, A., J. C. Ayres, and H. E. Snyder, 1964. Determination of oxidative changes in raw meats by the 2—thiobarbituric acid method. Food Technol. 18:223. Koch, R. S., 1962. Dehydrated foods and model systems. In Lipids and their Oxidation, AVI Publishers, Westport, Connecticut. Koehler, H. H., and M. Jacobson, 1967. Characteristic of chicken flavor-containing fraction extracted from raw muscle. J. Agr. Food Chem. 15(4):407. Kritchevsky, D., J. Langan, J. Makowitz, J. W. Berry, and D. A. Turner, 1961. Cholesterol vehicle in experimental atherosclerosis. Effects of absence or presence of fatty vehicle. J. Am. Oil Chemists' Soc. 38:74. Kuksis, A., W. C. Breckenridge, L. Marai, and O. Stachnyk, 1969. Melecular species of lecithins of rat heart, kidney and plasma. J. Lipid Res. 10:25. Lantz, C. W., and G. T. Carlin, 1938. Stability of fats used for deep fat frying. Oil and Soap 15(2):38. Landbrooke, B. D., and D. Chapman, 1969. Thermal analysis of lipids, protein and biological. Chem. Phys. Lipids 3:304. Lea, C. H., 1939. "Rancidity in Edible Fats", Chemical Publishing Company, New York, p 214. Lea, C. H., 1965. Chemical and nutritional aspects of oxidized and heated fats. Chem. and Ind. Feb. 244. Idxumweaver, H., J. H. Anderson, and H. L. Hanson, 1952. Effect of antioxidant on rancidity development in frozen creamed turkey. Food Technol. 6:1. Lowe, B., S. Pradhan, and J. Kastelie, 1958. The free fatty acid content and the smoke point of some fats. J. Home Ec. 50:778. Luddy, F. B., R. A. Barford, and R. W. Riemenschneider, 1960. Direct conversion of lipid components to their fatty acid methylesters. J. Am. Oil Chemists' Soc. 37:447. 113 Machlin, L. J., R. S. Gordon, J. Marr, and C. W. Pope, 1962. Effect of dietary fat on the fatty acids composition of eggs and tissues of the hen. Poultry Sci. 41:1340. Marion, J. B., and J. G. Woodroof, 1963. The fatty acid composition of breast, thigh and skin tissues of chicken broilers as influenced by dietary fat. Poultry Sci. 42:1202. Marion, J. B., T. S. Boggess, Jr., and J. G. Woodroof, 1967. Effect of dietary fat and protein on lipid composition and oxidation in chicken muscle. J. Food Sci. 32:426. McGinnis, G. W., and L. R. Dugan, Jr., 1965. A rapid low temperature method for preparation of methylesters of fatty acids. J. Am. Oil Chemists' Soc. 42:305. Mecchi, E. P., E. L. Pippen, and H. Lineweaver, 1964. Origin of hydrogen sulfide in heated chicken muscle. J. Food Sci. 29:393. Metcalf, L. D., and A. A. Schmitz. 1961. The rapid preparation of fatty acid esters for gas chromatographic analysis. Anal. Chem. 33:363. Meraill, J. H., 1960. Effects of an unsaturated and saturated lipid on experimental cholesterol atherosis in rabbits. Proc. Soc. Exp. Biochem. Med. 105:268. Miller, E. C., H. Menge, and C. A. Denton, 1962. A composition of fatty acid content of skin of chickens and turkeys. Poultry Sci. 41:1667 (Abstract). Miwa, T. K., K. L. Mikolajczak, F. R.Earle, and I. A. Wolff, 1960. Gas chromatographic characterization of fatty acids identification constants for mono- and dicarboxylic methylesters. Anal. Chem. 13:1739. National Commission on Food Marketing, 1966. Organization and competition in poultry and egg industries. Tech. Study No. 27. Washington, D.C. Nishida, T., F. Takenaka, and F. A. Kummerow, 1958. Effect of dietary protein and heated fat on serum cholesterol and beta lipo- protein level, and on incidence of experimental athero- sclerosis in chickens. Circulation Res. 6:194. O'Brien, J. S., and A. A. Benson, 1964. Isolation and fatty acid composition of plant sulfolipid and galactolipids. J. Lipid Res. 5:432. Orr, C. H., and J. E. Callen, 1958. Separation of polyunsaturated fatty acid methylesters by gas chromatography. J. Am. Chemists' Soc. 80:249. 114 Parker, F., and N. F. Peterson, 1965. Quantitative analysis of phospho- lipids and phospholipid fatty acids from silica gel thin-layer chromatograms. J. Lipid Res. 6:455. Paschke, R. F., J. E. Jackson, and D. D. Wheeler, 1952. Thermal polymerization of drying oils. Ind. Eng. Chem. 44:1113. Paschke, R. P., and D. H. Wheeler, 1955. Cyclization of eleostearic acid. J. Am. Oil Chemists' Soc. 32:473. Patton, S., I. J. Barnes, and L. E. Evans, 1959. N-Deca-2,4—diena1, its origin from linoleate and flavor significance in fats. ~ J. Am. Oil Chemists' Soc. 36:280. Peng. C. Y., 1965. Some studies on the composition and structure of phospholipids in chicken muscle. Ph.D. Thesis, Michigan State University. Perkin, G. B., 1967. Formation of non—volatile decomposition products in heated fats and oils. Food Technol. 21:125. Pippen, E. L., M. Nonaka, F. T. Jones, and F. Stit, 1963. Gas chromatography of chicken and turkey volatiles. The effect of temperature, oxygen and type of tissue on composition of the volatile fraction. J. Food Sci. 28:334. Poultry and Egg Situation, 1971. Economic Research Service, USDA, June, p 14. Powers, P. 0., 1949. Mechanism of the oxidation of drying oils. Ind. Eng. Chem. 41:304. Radin, R. S., 1969. In Methods in Enzymology, Vol. 14, J. M. Iowenstein, Ed., Academic Press, New Yerk, p 245. Rogozinski, M., 1964. A rapid quantitative esterification technique for carboxylic acids. J. Gas Chrom. 2:136. Rona, C., C. I. Chappel, and R. Gaudry, 1959. Aggravation of cholesterol atherosis in rabbits by free unsaturated fatty acid. Canada J. Biochem. Physiol. 37:479. Ropper, R., and T. S. Ma, 1957. Diazomethane as a reagent for micro- synthesis. Michromchem. J. 1:245. Rouser, G., A. N. Siakotos, and S. Fleischer, 1966. Quantitative analysis of phospholipids by thin-layer chromatography and phosphorus analysis of spots. Lipids 1:85. 115 Sahasrabudhe, M. R., and V. R. Bhalerao, 1963. A method for the determination of the extent of polymerization in frying fats and in fats. J. Am. Oil Chemists' Soc. 40:711. Scott, M. L., 1959. Composition of turkey meat. III. Essential amino acids composition. J. Am. Dietetic Assoc. 35:27. Skipski, V. P., R. F. Peterson, and M. Barclay, 1952. Separation of phosphatidylethanolamine, phosphatidyl serine and other phospholipids by thin-layer chromatography. J. Lipid. Ree. 3:467. Skipski, V. P., A. F. Smolowe, R. C. Sullivan, and M. Barclay, 1965. Separation of lipid classes by thin-layer chromatography. Biochim. Biophys. Acta 106:386. Smith, L. M., and E. L. Jack, 1954. The unsaturated fatty acids of milk fat. I. Methylester fractionation and isolation of mono- ethenoid constituents. J. Dairy Sci. 37:380. Stoffel, W., F. Chu, and E. H. Ahrens, Jr., 1959. Analysis of long- chain fatty acids by gas-liquid chromatography. Anal. Chem. 31:307. Tappel, A. L., 1952. Linoleate oxidation catalyzed by hog muscle and adipose tissue extracts. Food Res. 17:550. Tappel, A. L., 1953. Linoleate oxidation catalysts occurring in animal tissues. Food Res. 18:104. Tappel, A. L., 1962. Hematin compounds and lipoxidase as biocatalysts, In "Lipids and Their Oxidation", AVI Publishing Co., Westport, N. J., p 122. Tattrie, N. H., 1959. Positional distribution of saturated and unsaturated fatty acids on egg lecithin. J. Lipid Res. 1:60. Terroine, E. F., 1920. Contribution a la connaissance de la physiologie des substances grasses et lipoidiques, Ann. Sci. Nat. 2001. 4(10), 5:397. Thudichum, J. W. L., 1884. "A Treatise on the Chemical Composition of the Brain", Bailliere, Tindall and Cox. Time, M. I., and B. M. Watts, 1958. The protection of cooked meats with phosphates. Food Technol. 12:240. Turner, E. W., W. D. Paynter, E. J. Montie, M. W. Bessert, G. M. Struck, and F. C. Olson, 1954. Use of the 2-thiobarbituric acid reagent to measure rancidity in frozen pork. Food Technol. 8:326. 116 Vail, G. B., and R. Hilton, 1943. Edible fats and oils. Two chemical characteristics. J. Home Ec. 35:43. Watts, 8. M., 1954. Oxidative rancidity and discoloration in meat. In "Advances in Food Research", Academic Press, Inc.. New York p 1. Watts, B. M., 1962. Meat products. In "Lipids and their oxidations", AVI Publishing Co., Westport, p 202. Watts, B., and A. Merrill, 1963. Composition of foods - raw, processed and prepared, USDA Handbook No. 8, Washington, D. C., U. S. Department of Agriculture. Weiss, S. B., and E. P. Kennedy, 1956. The enzymatic synthesis of triglycerides. J. Am. Chem. Soc. 78:3550. Wills, E. D., 1961. Effect of unsaturated fatty acids and their peroxides on enzymes. Biochem. Pharmac. 7:7. Woodford, F. P., and C. M. VanGent, 1960. Gas—liquid chromatography of fatty acid methylester: the "carbon-number" as a parameter for comparison of columns. J. Lipid Res. 1:188. Younathan, M. T., and B. M. Watts, 1959. Relationship of meat pigments to lipid oxidation. Food RES. 24:728. Younathan, M. T., and B. M. Watts, 1960. Oxidation of tissue lipids in cooked pork. Food Res. 25:538. APPENDIX APPENDIX Chemical Changes in Corn Oil during Cooking with Cotton Balls Free fatty acid content of corn oil increased with increased cooking times (Fig. 3). It increased gradually during the first 2# hours of heating. then increased rapidly. The formation of free fatty acids in cooking oil during heating has been reported by many workers (Carlin et al.. 195#n Lantz and Carlin..1938: Marion and Woodroof. 1963: Lowe et al.. 1958: Vail and Hilton. 195#). Results similar to those obtained in this research were reported by Johnson and Kummerow. 1957. The formation of free fatty acids during heating is the :result of hydrolysis. Most commercial cooking oils consist of'a.large percentage of triglycerides and only small percentages of free fatty acids. diglycerides and monoglycerides after the refining process. Hydrolysis may occur in triglyceride. diglyceride. and or monoglyceride 117 Free fatty acid (%) .7 .5 .# .3 .2 118 A J 12 18 24 30 36 Heating time (hrs.) x: Free fatty acid a: Viscosity Fig. 3. Changes in free fatty acid and viscosity in corn oil during heating. #2 #8 1#.0 12.0 ’ 10.0 8.0 6.0 2.0 ( estod ) KitsoostA .l‘iil‘ J. 119 molecules when cooking foods in oil. Viscosity of cooking oil also increased with heating time (Fig. 3). Viscosity increased slowly during the first 2h hours of heating. then proceeded more rapidly. and after 36 hours of heating. the viscosity increased very rapidly. These results are similar to these reported by Johnson and Kummerow (1957). The increasing viscosity probably results from various polymerization reactions. Some workers (Crampton et al.. 1951: Paschke et al.. 1952: Powers 19159) have shown that this reaction is due partially from the formation of polymers through the Dials-Alder type of condensation. and the formation of oxypolymers. The color changes in corn oil, as absorbance. during heating are shown in Fig. #. Absorbance values increased slowly during the first 2“ hours of heating. then proceeded more rapidly. and after 36 hours of heating. the absorbance values increased very rapidly. Visual appearance of dark color occurred when absorbance values were greater than 0.05. The reason for color changes in corn oil during cooking is due to that chromophoric materials are formed as a O\ O O O O O O 0 UP“ 0 N O Absorbance (21000) at 500 ’IML O 5...: O O 120 J J l L I V 6 12 18 24 30 36 1&2 #8 Heating time (hrs.) Fig. 14. Changes in absorbance of corn oil during heating -vli . Jr\ j/‘Iln' JIII all 121 consequence of the polymerization reactions. The changes in iodine values are shown in Fig. 5. Iodine values declined gradually during the first 24 hours of heating. then declined more rapidly. The iodine value of fresh corn oil was 120. and after #8 hours of heating it dropped to 103. These results are similar to these reported by Johnson and Kummerow. (1957). by Perkin (1967), by Lea (1965) . The decreasing iodine value of corn oil during cooking indicated a decrease in the number of unsaturated linkages. This is evidence that oxidation and/or polymerization occurred in the oil during heating. The peroxide values ofcorn oil (fig.5) increased gradually during the first 2h hours of heating. This suggests that oxidation proceeded at an accelerated rate during prolonged heating times. Since peroxide values decreased after 36 hours of heating. this indicates a decrease of substrate and that secondary reactions occurred with the hydrOperoxides undergoing a variety of scission and dismutation reactions to form a wide spectrum of compounds. The formation and isolation of polymer compounds in Iodine values 130 125 120 115 110 105 " 100 122 sentsa eptxoaeg AL 1 a 1 I 6 12 18 24 30 36 #2 #8 Heating tine (hrs.) x I Iodine values. a : Peroxide values. Fig. 5. Changes in iodine values and peroxide values in.corn oil during heating. 123 cooking oil during heating has been reported by Crampton et al. (1951. 1952). Crampton et a1. (1951) determined that cyclic monomers. dimers. and polymers were the components considered as toxic agents in cooking oil. The amounts of non-urea adducting fraction increased greatly during heating (Fig. 6). Sahasrabudhe and Bhalerao (1963) studied the formation of NUAF in corn oil and reported 0.5%. 3.5%. 8% and 17.5% of NUAF after 0. 8. 16 and 24 hours of heating. The value at 24 hours appeared higher than in this experiment. probably due to the difference in cooking conditions. Based on analysis of the changes in free fatty acid liberation. viscosity. color. iodine values. peroxide values. non-urea adducting fraction. and acid composition in corn oil during heating. there are obviously three stages which can be found according to the rate of chemical reactions. The first stage is the period of the first 24 hours of heating in.which free fatty acids. viscosity. absorbance values. peroxide values. and NUAF increase at a slow rate. but iodine values and linoleic acid decline slightly. Thus. hydrolysis. oxidation. and polymerization all occur at a slow rate in N O H 0 Non urea adducting fraction.(%) H \A U! Figs 60 124 j 12 18 2h 30 36 42 #8 Heating time (hrs.) The fernation.of nonpurea adducting fraction in.corn oil during heating. 125 this period. The second stage is the period from 2“ hours to 36 hours of heating time. The chemical reactions proceed at a more rapid rate. and it is presumed that the increased formation of free radicals speeds up various chemical reactions. The last stage is defined as the heating period after 36 hours. in.which decomposition of peroxides. viscosity. absorbance values. and NUAF increase at greatest rate. indicating that oxidation and polymerization proceed at an acclerated rate. WNW L H I“ S‘II‘ R“ EI' V“ WIN ‘Wl'iifimfilim 1293 03196 53 3