H 4—7" fi—v‘w . .. .5 THE EFFEVCTWENESS OF NITROGEN-AND SULFUR-CHELATENG COMPOUNDS EN INHI'BITENG THE DEVELOPMENT OF OXIDIZED FLAVOR IN MILK Thai: for the Dagru, of M. 5. MICHIGAN STATE UNIVERSETY Peter Fraser Pierpo-nt 1961 .. ‘~-.‘--."‘u .' _ a. ":‘ HESS LIBRARY Michigan State University 2' -‘ ,,~\w;l'mr!f ABSTRACT THE EFFECTIVENESS OF NITROGEN- AND SULFUR—CHELATING COMPOUNDS IN INHIBITING THE DEVELOPMENT OF OXIDIZED FLAVOR IN MILK by Peter Fraser Pierpont The present investigation was designed to compare the effectiveness of the calcium-disodium salt of ethylenediamine- tetraacetic acid (EDTA), carboxymethylmercaptosuccinic acid (CMMS), and the monolauryl ester of carboxymethylmercapto- succinic acid in inhibiting the development of the metal- induced oxidized flavor in milk. Mixed herd milk, used in all trials was pasteurized either at 1430 F. for 30 min. or at 1610 F. for 15 sec. Copper in concentrations of 0.1 and 0.5 p.p.m. was added to the milk either before or after pasteurization. The chelating agents, at concentrations equivalent to l, 5, 10, 100,:KDO, and 4349 times the amount stoichiometrically necessary to chelate the added copper, were added to the milk immediately following the addition of the copper. Qualitative and quantitative evaluations of the control and treated milk were obtained by organoleptic examination and from the determination of thiobarbituric acid and peroxide values. 2 Peter Fraser Pierpont The calcium-disodium salt of ethylenediaminetetraacetic acid, when present to the extent of 100 or 1000 times the amount stoichiometrically necessary to chelate the added copper, was very effective in inhibiting the oxidized flavor development in the milk. When added in concentrations of less than 100 times the amount stoichiometrically necessary to chelate the added copper, the EDTA seemed to have had little effect in inhibiting the development of the oxidized flavor. Carboxymethylmercaptosuccinic acid failed to prevent completely the development of the oxidized flavor in the presence of added copper under the experimental conditions. The monolauryl ester of carboxymethylmercaptosuccinic acid, possessing both lipophilic and hydrophilic characteristics, was ineffective also in completely preventing the development of the oxidized flavor in the copper-treated milk. THE EFFECTIVENESS OF NITROGEN- AND SULFUR-CHELATING COMPOUNDS IN INHIBITING THE DEVELOPMENT OF OXIDIZED FLAVOR IN MILK BY PETER FRASER PIERPONT A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Food Science 1961 G/slf‘i ii ACKNOWLE DGME NTS The author wishes to express his sincere appreciation to Dr. C. M. Stine, Associate Professor of Food Science, for his patience and guidance in directing this study and graduate program, and for his counsel and advice in preparing this manuscript. The author is deeply indebted to Dr. G. M. Trout, Profes- sor of Food Science,for his meticulous reading and subsequent suggestions in preparing this manuscript. Grateful acknowledgment is due to Dr. H. A. Lillevik, Associate Professor of Chemistry, for his careful reading of the manuscript and services as a committee member. The author sincerely appreciates the financial support of the Dairy Industries Supply Association, Inc. and the funds and facilities provided by Michigan State University. The author is most grateful to his wife, Mary, for her assistance in preparing this manuscript and her inspiration and encouragement throughout the course of this graduate program. iii TABLE OF CONTENTS Page INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . 1 REVIEW OF LITERATURE . . . . . . . . . . . . . . . . . . 3 Types of Oxidized Flavor 3 Source of Oxidized Flavor 4 Detection of Oxidized Flavor 5 Metallic Contamination 8 Retarding or Inhibiting Development of the Oxidized Flavor 12 Effect of Aeration and Deaeration 12 Elimination of Metallic Contamination 12 Segregation 13 Condensing 13 Effect of Ascorbic Acid 13 Effect of Homogenization 14 Effect of Antioxidants 15 Metal Inactivators 17 Effect on Edible Oils 18 Effect on Milk l9 EXPERIMENTAL PROCEDURE . . . . . . . . . . . . . . . . . 21 Determination of Thiobarbituric Acid Values 23 Determination of Peroxide Values 24 Copper Determinations EXPERIMENTAL RESULTS . . . . . . . . . . . . Treatment of Holder Pasteurized Milk Effect of Calcium-disodium Salt of EDTA Effect of CMMS Effect of MLCMMS Treatment of HT-ST Pasteurized Milk Copper Determinations DISCUSSION . . . . . . . . . . . . . . . . . . SUMMARY AND CONCLUSIONS . . . . . . . . . . . LITERATURE CITED . . . . . . . . . . . . . . . iv Page 26 3O 3O 30 4O 50 61 62 71 76 78 10. TABLES Page Influence of EDTA on the Stabilization of Flavor in Copper-treated Milk Pasteurized at 143° F., for 30 Minutes. (Additions made before pasteurization). . . 36 Influence of EDTA on the Stabilization of Flavor in Copper-treated Milk Pasteurized at 143° F. for 30 Minutes. (Additions made after pasteurization). . . 37 Influence of EDTA on the Stabilization of Flavor in Copper-treated Milk Pasteurized atl43O F. for 30 Minutes. (Additions made after pasteurization) . . . 39 Influence of CMMS on the Stabilization of Flavor in Copper-treated Milk Pasteurized at 1430 F. for 30 Minutes. (Additions made before pasteurization) . . 46 Influence of CMMS on the Stabilization of Flavor in Copper-treated Milk Pasteurized at 143° F. for 30 Minutes. (Additions made after pasteurization) . . . 47 Influence of CMMS on the Stabilization of Flavor in Copper-treated Milk Pasteurized at 143° F. for 30 Minutes. (Additions made after pasteurization) . . . 49 Influence of MLCMMS on the Stabilization of Flavor in Copper-treated Milk Pasteurized at 1430 F. for 30 Minutes. (Additions made before pasteurization) . 56 Influence of MLCMMS on the Stabilization of Flavor in Copper-treated Milk Pasteurized at 143° F. for 30 Minutes. (Additions made after pasteurization). . 57 Influence of MLCMMS on the Stabilization of Flavor in Copper-treated Milk Pasteurized at 143° F. for 30 Minutes. (Additions made after pasteurization). . 59 Influence of EDTA on the Stabilization of Flavor in Copper-treated Milk Pasteurized at 1610 F. for 15 Seconds. (Additions made before pasteurization). . . 63 ll. 12. 13. Influence Seconds. Influence of CMMS on the Stabilization of Flavor in Copper-treated Milk Pasteurized at 1610 F. for 15 Seconds. Influence of CMMS on the Stabilization of Flavor in Copper-treated Milk Pasteurized at 1610 F. for 15 Seconds. of EDTA on the Stabilization of Flavor in Copper-treated Milk Pasteurized at 1610 F. for 15 (Additions made'after pasteurization) (Additions made before pasteurization). (Additions made after pasteurization) vi Page 65 67 69 10. vii FIGURES Page Thiobarbituric acid values of fluid milk heated to 143° F. for 30 Minutes with additions made before the heat treatment . . . . . . . . . . . . . . . . . 32 Peroxide values of fluid milk heated to 143° F. for 30 minutes with additions made before the heat treatment . . . . . . . . . . . . . . . . . . . . . 33 Thiobarbituric acid values of fluid milk heated to 1430 F. for 30 minutes with additions made before the heat treatment . . . . . . . . . . . . . . . . . 34 Peroxide values of fluid milk heated to 1430 F. for 30 minutes with additions made before the heat treatment . . . . . . . . . . . . . . . . . . . . . 35 Thiobarbituric acid values of fluid milk heated to 143° F. for 30 minutes with additions made before the heat treatment . . . . . . . . . . . . . . . . . 42 Peroxide values of fluid milk heated to 1430 F. for 30 minutes with additions made before the heat treatment . . . . . . . . . . . . . . . . . . . . . 43 Thiobarbituric acid values of fluid milk heated to 143° F. for 30 minutes with additions made before the heat treatment . . . . . . . . . . . . . . . . . 44 Peroxide values of fluid milk heated to 1430 F. for 30 minutes with additions made before the heat treatment . . . . . . . . . . . . . . . . . . . . . 45 Thiobarbituric acid values of fluid milk heated to 1430 F. for 30 minutes with additions made before the heat treatment . . . . . . . . . . . . . . . . . 52 peroxide values of fluid milk heated to 143° F. for 30 minutes with additions made before the heat treatment . . . . . . . . . . . . . . . . . . . . . 53 11. 12. viii Page Thiobarbituric acid values of fluid milk heated to .1430 F. for 30 minutes with additions made before heat treatment . . . . . . . . . . . . . . . . . . 54 Peroxide values of fluid milk heated to 1430 F2 for 30 minutes with additions made before the heat treatment . . . . . . . . . . . . . . . . . . . . . 55 INTRODUCTION The United States market—milk industry is a multi-billion dollar business. Of the total amount spent for dairy products in 1959, approximately 64 percent was spent on fluid milk and cream. Despite this tremendous volume of market-milk processed and consumed annually in the United States, the per capita consumption might be even greater if all the milk were free of flavor defects. Off-flavors in milk have been classified according to their causes into seven classes as follows: a) bacteria; b) feed; c) absorbed; d) composition, e) processing: f) chemical changes; and g) foreign. Knowing the causes of off flavors, one can attack the flavor problems with greater assurance of overcoming them. For several decades the major flavor defect in milk, asso- ciated with chemical flavor deterioration, was the metal-induced oxidized flavor. However, when homogenizing beverage milk became an accepted practice, this specific off flavor was, in large part, inhibited. NOnetheless, the problem of light— induced oxidized flavor became more serious. This light- induced oxidized flavor defect in homogenized milk has been inhibited almost completely by the use of paper and/or colored- glass milk containers. While the development of the metal-induced oxidized flavor is inhibited by homogenization and by elimination of copper equipment, the metal-induced, cardboardy, oxidized flavor is still a problem and threat to the dairy industry. Recently, studies on improving the oxidative stability of vegetable oils have shown that some coordination com- pounds, known as chelating agents, are very effective inhibi- tors of the oxidative deterioration in the oils. A limited amount of work has been reported on the application of these chelating agents in inhibiting the oxidized flavor in beverage milk. The main purpose of this study was to compare the effectiveness of some of the chelating agents in inhibiting the development of the oxidized flavor in fluid milk. A secondary objective was to determine the optimum concentration of these chelating agents found effective and to observe whether the different methods of pasteurization had any effect on their activity. To these ends the studies reported herein were undertaken. REVIEW OF LITERATURE The most important single flavor defect in beverage milk is the oxidized flavor (27). Since this defect in milk was first recognized by Golding and Feilman (18), it has become a constant, almost insurmountable problem to the dairy industry. Comprehensive reviews on the subject are available (6, 22, 34, 52). In the studies reported herein, an extensive review of the entire subject matter wasrmt;attempted unless the literature pertained specifically to metal-induced oxidized flavor in fluid milk. Types of oxidized flavor. Oxidized flavor in fluid milk may be classified into one of two general categories, namely as to being a) light or b) metal induced. Thurston (57), working with metal-induced oxidized flavor, reported that all milk could be grouped into a) spontaneous, b) susceptible, or c) nonsusceptible types. Immediately cooled milk that develops oxidized flavor upon storage without added copper is termed spontaneous. Roadhouse and Henderson (44) observed that this spontaneous development of oxidized flavor occurred most frequently in late winter and early spring. Milk that develops the oxidized flavor when small amounts of copper are added is termed susceptible. Reports indicate that individual cows vary considerably in respect to the minimum amount of copper required to cause this off-flavor. Milk that will not develop the oxidized flavor after 2 or 3 days of storage at 40 degrees F., when 1 p.p.m. copper is added, is termed non-susceptible. Milk exposed to artificial or natural light for a period of 10 min. to 1 hr. will develop an oxidized flavor. This flavor is sometimes termed an activated flavor (52). Pont (39) reported that this flavor is composed of two elements. He described these elements as an oily-tallowy flavor in the fat, and a scorched-gluey flavor in the non-fat constituents, the latter becoming stronger with longer exposure. However, he has also observed that the scorched-gluey flavor in the non-fat constituents is adsorbed on the fat globule membrane. The composite flavor of these two elements, which he termed "cardboard," is then carried by the fat globule. Homogenized milk is known to be more susceptible to the light-induced flavor defect (52). This greater susceptibility is thought to be a result of the increased area of protein surface absorbed on the fat globule. Source of oxidized flavor. Early workers believed that the oxidized flavor of beverage milk was caused by the oxidation of the fat in the milk. Pont (40) reported that this theory was discarded because of the inability to substantiate sufficient supporting evidence. At that time, the chemical methods of measurements employed were the Kreis test, iodine number and iodometric peroxide tests. Pont (41) concluded that these methods of analysis were not delicate enough to evaluate the small amount of oxidation which accompanied the first signs of an off-flavor. However, by using peroxide determinations, he collected evidence which demonstrated that fat oxidation always accompanied the oxidized flavor. Present theories indicate that the phospholipids serve as the origin of the oxidized flavor in fluid milk. Evidence has been obtained (27) to show that, while the iodine value of the milk fat has undergone little change, the iodine value of the phospholipid has been lowered considerably. Pont (40) pointed out that phospholipids were an unlikely source of the flavor. From his observations, he concluded that the compounds responsible for the oxidized flavor were steam volatile. These steam volatile compounds are now believed to be low-molecular-weight, cleavage products from the secondary oxidation products of highly unsaturated fatty acids. Many of these compounds have recently been identified by Day and Lillard (l3). Detection g: oxidized flavor. Ever since oxidized flavor was first recognized, the flavor has been readily detectable by organoleptic procedures. Due to the human element involved, numerous descriptions and scales of measurement have been reported in the literature. The more common terms used to describe the oxidized flavor are ”cappy," "cardboard," "metallic," "oily," "oxidized," "papery," and "tallowy." In addition to a description of the flavor, the intensity of the off-flavor is important also. A number of designations have been given for indicating the intensity of the oxidized flavor. The most prevalent classification used appears to be: - no oxidized flavor, I questionable oxidized flavor, + slight oxidized flavor, ++ distinctly oxidized flavor, +++ pronounced oxidized flavor. Chemical measurement 9: oxidized flavor. From a practical standpoint, the most direct approach to the measurement of milk fat oxidation is through organoleptic examination. However, because of the uncontrollable variables associated with organoleptic procedures, a number of chemical methods have been developed to measure the fat deterioration (27). Of these numerous chemical methods, the two most commonly used are peroxide value and thiobarbituric acid value. a) Peroxide value. Of all the chemical methods available, the peroxide value analysis developed by Hills and Thiel (25) appears to be the best technique for observing the oxidative deterioration of fat (27). The development of a de-emulsifi- cation reagent by Pont (42) made this ferric thiocyanate method adaptable for tracing milk fat oxidation. He observed that this reagent gave results which more nearly duplicated the results of more tedious, but reliable, methods. The method employed by Hills and Thiel (25) was further modified by Stine _£._l. (51). The combined effects of these improvements have made the peroxide value method a very accessible tool in measuring milk fat deterioration. In addition, the peroxide value method has been shown to be very sensitive, giving results correlating closely with those from organoleptic examination. b) Thiobarbituric acid value. The method of using 2- thiobarbituric acid reagent to measure fat oxidation has come into extensive use in recent years. Bernheim gt 31. (4) were the first to show that 2-thiobartiburic acid produced a red pigment when reacted with the oxidation products of unsaturated fatty acids. Further investigation by Wilbur et a1. (59) showed that the red color reaction was characteristic of specific fatty acid oxidation. Very little or no color was observed when other non-specific fatty acids were treated with 2-thiobarbituric acid. Patton and Kurtz (38) made the fundamental investigation of the 2-thiobarbituric acid reaction as a means of measuring milk fat oxidation. By spectral analysis, they showed that when 2-thiobarbituric acid was heated with oxidized milk fat, it reacted with malonic dialdehyde to give the characteristic red pigment. After intensive study of the factors that influenced the test, Dunkley and Jennings (14) developed a simple procedure for applying the 2-thiobarbituric acid test to milk. Their results showed the reproductability of results by the test as well as the influence of certain contributing factors on the test. Close correlations between 2-thiobar- bituric acid values and organoleptic detections of oxidized flavors have been observed (7, 14). Metallic contamination. The fact that contamination of fluid milk with traces of heavy metals induces rapid oxidation with a subsequent shortening of keeping quality, is well established. Golding and Feilman (18) were probably the first to observe that milk contaminated with metallic ions developed an off-flavor. They described this off-flavor as "mealy." The primary cause of the off-flavor development in their investigation was copper contamination of a worn cooler from exposed copper. Their experiments clearly showed that the copper acted upon the milk fat in the presence of air, and that small quantities of copper, ranging from 1 to 100 p.p.m. went into solution in the milk. Briggs (5), working with a large variety of heavy metals, observed that vanadium was the most active metal in inducing autoxidation in milk fat. Investigation by Davies (12), however, showed that when traces of metal were incorporated into butter in the amounts of 2 to 50 p.p.m., copper has the most effective catalytic action. These fundamental investi- gations lead to the present belief that cupric ions, even when present in a fraction of a p.p.m., are by far the most effective catalysts of oxidation in milk fat. As yet, the complete chemical mechanism of oxidized flavor development is not completely understood. Present reports indicate that metallic catalysts have their function through the decomposition of hydroperoxides to produce additional free radicals. Holman _E.§l- (26) report that by this mechanism of hydroperoxide decomposition metallic salts promote oxida- tion. They propose that since many organic peroxides are capable of taking electrons from and yielding electrons to metallic cations, metallic salts possibly could function by the following mechanism: + ++ . C11 + ROOH _—'>' CU. + R-O: + OH ++ — , + Cu + ROO: R-OO + Cu ——+ The metallic catalyst could thus maintain a concentration of active hydroxyl radicals. a) Effects of natural copper. Varying amounts of copper are known to appear in milk naturally. King and Dunkley (28) have observed that as much as 0.2 p.p.m. natural copper is common to early lactation milk. During the succeeding lacta- tion period, they report that from 0.02 to 0.04 p.p.m. natural 10 copper may be expected in the milk. They found that a highly significant correlation existed between the concentration of natural copper and the incidence and intensity of spontaneous oxidized flavor when the cows were on a hay-grain concentrate diet. These research workers concluded that natural copper in milk is an important catalyst of oxidized flavor. Jenness and Patton (27) have reported that natural copper in milk is to a large extent in the form of complexes with protein. Investigations by Tappel (55) have shown that copper- proteins, formed by the binding of copper ions to con albumin or caseinate, are more effective catalysts for linoleate oxidation than is copper alone. He reasoned that copper is more effective as a protein complex because of the increased ease with which copper forms the complex and the greater stability of the intermediate complex of linoleate peroxide- copper protein. b) Effects of copper contamination. Contamination of milk during processing by trace metals, and more specifically by copper, has long been recognized as a catalyst to the formation of the oxidized flavor (24, 27, 44, 49). However, with copper and copper-tinned equipment virtually obsolete in the dairy industry, the problem of copper contamination would seem to be of historic interest only. This reasoning was refuted by the recognition that the often—used "white—metal" 11 and nickel alloys, which contain copper could induce the oxidized flavor. Henderson and Roadhouse (23) demonstrated that little copper goes into solution from alloys containing tin and zinc. Consequently, alloys that contain these two elements do not influence the flavor of milk as much as alloys which are void of these two elements. They verified that copper was the source of contamination in these alloys. This fact was substantiated by showing that even when pure strips of nickel, lead, and zinc were exposed to milk, no oxidized flavor developed. Lusas 33 31. (36) pointed out that the problem of copper— induced oxidized flavor in milk from "white-metal" systems is still very prominent. They reported that the average increase in the trace metal content of milk processed in a "white- metal" HT—ST pasteurization system was as follows: copper - 0.095 p.p.m., iron- 0.0 p.p.m., nickel— 0.04 p.p.m. The milk processed by the "white-metal" system consistently developed the oxidized flavor upon storage. They concluded that the increase in copper content was probably responsible for the development of the oxidized flavor. From this evidence, trace metal-induced oxidized flavor is obviously a problem unless all processing equipment is either stainless steel or enameled. 12 Retarding 9£_Inhibiting Development Qf_the Oxidized Flavor Effect Qf_aeration and deaeration. Aeration and deaeration have been acclaimed by a number of workers as methods for in- hibiting the development of oxidized flavor. Greenbank (21) was able to retard the development of oxidized flavor in fluid milk by aeration. He demonstrated also that aeration was a suitable method for improving the copper tolerance of fluid milk. Dalhe and Palmer (11) were probably the first to conclude that the removal of oxygen from beverage milk could prevent the development of the oxidized flavor. Greenbank (20) demon- strated that deaeration can protect against oxidized flavor development even when relatively high concentrations of copper are present. Roadhouse and Henderson (44) have reported that deaerated fluid milk produced no oxidized flavor in seven days even when contaminated with 1.0 p.p.m. copper ions. Elimination 9f_metallic contamination. The fact is well established that much of the metal—induced flavor development in milk could be eliminated if metallic contamination did not occur. The passing of milk through bronze, copper, brass, or "white-metal" pipes and fittings induces the development of the oxidized flavor. The elimination of this type of proces- sing equipment and the employment of stainless steel or enamel processing equipment, greatly minimizes the problem. l3 Segregation. The separation of spontaneous milk and milk of low copper tolerance from the milk supply is another method of preventing metal—induced oxidized flavor in milk. Greenbank (19, 20) reported that milk of low copper tolerance can be detected by observing the increase in the oxidation- reduction potential of the milk after copper has been added. Condensing. Corbett and Tracy (10) showed that milk condensed under vacuum to a concentration of 2 to 1 prevents the development of the oxidized flavor in the condensed product. They observed that when this condensed milk was reconstituted to the original solids concentration, the oxidized flavor was yet inhibited. In addition, these research workers have reported that 4 percent milk made from condensed skimmilk and 32 percent cream, does not develop the metal-induced oxidized flavor in the presence of as much as 3 p.p.m. added copper. They concluded that the inhibiting action was probably due to the high heat treatment the pro- ducts receive during the condensing process. This conclusion is in agreement with the observation of Bernhart and Linden (3) who reported that high temperature heat treatment of fluid milk reduces the pro—oxidant effect of the copper present in the milk. Effect 9; ascorbic acid. Krukovsky and Guthrie (31, 32) have demonstrated that ascorbic acid oxidation is an essential l4 link in the reactions which result in the development of the oxidized flavor. They reported that inhibition of the oxidized flavor can be accomplished by quick and complete oxidation of the ascorbic acid to dehydroascorbic acid prior to pasteurization. They observed also that only partial oxidation of ascorbic acid to dehydroascorbic acid stimulates the development of the oxidized flavor. These workers indi- cated also that when milk had been completely depleted of ascorbic acid by quick oxidation to dehydroascorbic acid, and pasteurized thereafter, the oxidized flavor development was inhibited, even with added copper. This observation is in accord with the reports of Jenness and Patton (27) who claim that when conditions are favorable, the amount of ascorbic acid in milk enhances the development of the oxidized flavor. Jenness and Patton reported also that when fluid milk was fortified with ascorbic acid in the amount of 50 to 100 mg. per liter, the ascorbic acid acted as an antioxidant. Similar observations have been reported by Ritter (43). Effect of homogenization. According to the literature, one of the best known practical methods of inhibiting the development of the oxidized flavor is homogenization. Tarassuk and Koops (56) reported disagreement on the postulates explaining why homogenization inhibits the development of the oxidized flavor. These research workers have shown that the 15 concentration of phospholipids and copper-protein complexes on the fat globule is decreased proportionally to the homo- genization pressure. They concluded that this decrease in concentration per unit of newly formed fat globule surface area is the main factor in retarding the development of the oxidized flavor by homogenization. Effect 9; antioxidants. One of the most straightforward methods of preventing milk fat oxidation is through the use of antioxidants. While in most areas such addition is pro— hibited by law, considerable research has been done evaluating the effectiveness of antioxidants in retarding the development of the oxidized flavor. Ritter (43) showed that hydroquinone inhibited the develop- ment of the oxidized flavor when added to milk. Garrett (17) observed that divalent manganese, when added to milk con- taminated with iron or copper, completely inhibited the development of the oxidized flavor. Garrett reasoned that the iron or copper contaminant was able to oxidize the divalent manganese to a higher valence, and in this way prevented reaction with the milk fat. Russell and Dahle (45) reported that dried milk, when added to fluid milk can act as an antioxidant. They based the effectiveness of such a treatment on the a) processing f“ ~ treatment of the product, b) type of product, c) method of 16 adding the product, and d) amount of the product added, to the fluid milk. The effectiveness of phenolic inhibitors as antioxidants in dry lard and in aqueous solution at pH 7.5 were observed by Lehmann and Watts (35). They reported that nordihydro— guararetic acid (NDGA) was the most effective phenolic inhibitor in dry hard lard. They observed that butylated hydroxyanisole‘wasthe most attributable phenolic inhibitor in an aqueous system at pH 7.5. Reports by Stull t.§£- (53, 54) indicated that NDGA was an effective antioxidant in milk. These authors observed that NDGA in concentrations of 0.001 to 0.01 percent was equally effective, with or without an added synergist, in retarding the development of the oxidized flavor. Trout and weinstein (58) showed that NDGA was effective also in inhibiting the activated flavor in fluid milk. These research workers reported also that neither hydroquinone nor d-tocopherol gave complete protection against the development of the activated flavor. Chilson _E a1. (8), while working with propyl gallate, showed that 20 mg. of this inhibitor per liter of freshly pasteurized milk, would prevent the development of the oxidized flavor for at least 14 days. They found also that the propyl gallate was equally effective, with or without added copper. 17 Metal inactivators. According to Schwab t 31. (47) coordination complexes have been recognized for over 100 years. A satisfactory explanation as to the nature of these complexes was not available until Werner proposed his coor- dination theory (26) in 1891. Dutton _t_al. (15) investigated the use of polycarboxylic acids and polyhydric alcohols in improving the stability of oils. With the knowledge that salts and esters of organic acids are inactive, they realized that free carboxyl groups are necessary if the acids are going to improve the stability of the oils. Among the four carbon dicarboxylic acids studied, they observed that the activity of the acids increased as the number of hydroxyl groups increased. In addition, Dutton _t‘al. reported that the introduction of citric acid and sorbitol increased the stability of oils containing pro-oxidant metals. This increase in stability was compatible with the known metal scavenging ability of citric acid and sorbitol. Krum and Fellers (33) pointed out that the disodium salt of ethylenediaminetetraacetic acid combines with heavy metals in wines to form soluble complexes of high stability. They recommended that the disodium salt of ethylenediaminetetra- acetic acid be added at the rate of 8 times the amount of metal present in order to prevent cloudiness in the wines. 18 Effect gg_edib1e oils. In recent years, considerable work has been done with metal inactivators in the field of edible oils. Schwab gt_al, (47) studied the effect of different nitrogen coordination compounds in improving the oxidative stability of soybean oil. They reported that ethylenediamine— tetraacetic acid was a very effective inhibitor of oxidation in the presence of both added iron and copper. These authors also found that while imino alpha or beta dicarboxylic acids have varying degrees of effectiveness as an inhibitor of oxidation, chelidamic acid is a very effective metal deactiva- ting agent for both copper and iron. Schwab _3 a1. (48) studied the effectiveness of a number of sulfur coordination compounds in improving the oxidative stability of soybean oil. They reported that while many improved the oxidative stability of soybean oil, some of the sulfur compounds imparted an off-flavor. One of the compounds that improved the flavor of the oil as well as the oxidative stability was carboxymethylmercaptosuccinic acid (CMMS) . They observed that CMMS was very effective against both iron and copper contamination. Evans _Eh_l. (16) showed CMMS to be one of the most effective metal inactivators examined for use in glyceride oils. Reports indicate that CMMS has an extremely low order of toxicity. l9 Cooney ggial. (9) observed that CMMS was not an active metal inactivator against trace metal contamination in unheated vegetable oils. They suggested that the trace metals were held within complexes of unknown structure, which would make the trace metals unavailable for complexing by metal inactivators until after the oil had been heated. These research workers demonstrated that while the trace metals were not available as an uncomplexed ion, they were still a very strong pro— oxidant in unheated oils. Effect gg_mi;k. A limited amount of data has been reported showing the effectiveness of coordination compounds or chelating agents in inhibiting oxidative deterioration of milk fat. Arrington and Krienke (1, 2) were probably the first to call attention to the effectiveness of various salts of ethylenediaminetetraacetic acid in inhibiting the development of the oxidized flavor in beverage milk. They observed that the various salts of ethylenediaminetetraacetic acid were approximately equal in effectiveness in inhibiting the oxidized flavor. They found that the minimum quantity required for complete protection against the oxidized flavor appeared to be about 5 times the amount of added copper calculated on a molar basis. They noted too that sodiumdiethy1dithiocarbamate appeared to be a more effective inhibitor than ethylenediamine- tetraacetic acid but that the sodiumdiethyldithiocarbamate had 20 a tendency to impart an off—flavor to the milk. They con- cluded that while these chelating agents were equally effective when added before or after the added copper, pasteurization did not alter their effectiveness. King and Dunkley (29) reported on the role of ethylene- diaminetetraacetic acid in inhibiting the development of the oxidized flavor in milk labeled with Cu64. Their results showed that a greater percentage of the copper was chelated from the washed fat globules than from the skimmilk. They concluded that ethylenediaminetetraacetic acid inhibited the development of the oxidized flavor in fluid milk by removing the catalytic copper from some key site on the fat globule membrane. 21 EXPERIMENTAL PROCEDURE Mixed herd milk from the Michigan State University dairy herd was used throughout the studies reported herein. The milk was pasteurized either at 1430 F. for 30 min. or at 1610 F. for 15 sec. in stainless steel containers. During both the heating and the cooling periods the milk was stirred gently using stainless steel utensils. Following the pasteurization exposure, the milk was cooled immediately to 400 F. All experimental samples were stored at 350 F. in half gallon glass milk bottles which had been washed and sanitized in a commercial bottle washer. Oxidized flavor development in the stored milk was induced by adding ionic copper to the milk in the form of an aqueous solution of copper sulfate. The concentration of copper added was either 0.1 or 0.5 p.p.m. Chelating agents were used to sequester the copper ions. Those used were a) the calcium-disodium salt of ethylene- diaminetetraacetic acid (EDTA), b) carboxymethylmercapto- succinic acid (CMMS), and c) the monolauryl ester of carboxy- methylmercaptosuccinic acid (MLCMMS). A sample of "Food Grade" EDTA was obtained from the Dow Chemical Company, Midland, Michigan. Samples of CMMS and MLCMMS were obtained from Evans Chemetics, Incorporated, New York. 22 The chelating agents were dissolved in ion-exchanged water which was subsequently redistilled in Pyrex glassware. The addition of the chelating agent solutions to the milk was made either immediately before pasteurization or immedi- ately after the cooling period. All additions of the chelating agents were made directly after the addition of copper. The chelating compounds were added at concentrations equivalent to l, 5, 10, 100, 1000, and 4349 times the amount stoichiometrically necessary to chelate the added copper. These concentrations appear throughout the results as a number X the amount of added copper, which is defined as the multiple of the amount stoichiometrically necessary to chelate the added copper. The development and intensity of the oxidized flavor were determined organoleptically at 0, 2, 4, and 8 days following treatment of the samples. A panel of experienced judges, working independently, was used for this purpose. The intensities of the oxidized flavors were recorded according to the following scale: - no oxidized flavor ? questionable oxidized flavor + slight oxidized flavor ++ pronounced oxidized flavor +++ very strong oxidized flavor 23 Determination g: Thiobarbituric Acid Values Thiobarbituric acid (TBA) values were determined at 0, 2, 4, and 8 days after treatment by the procedure of Dunkley and Jennings (14). a) Reagents. A solution of 0.025 M 2-thiobarbituric acid in one molar phosphoric acid was prepared by mixing equal volumes of 0.05 M 2-thiobarbituric acid and 2 M phosphoric acid. The extraction mixture of isoamyl alcohol and pyridine was prepared by mixing two volumes of A. R. isoamyl alcohol: with one volume of A. R. pyridine. / b) Procedure. Ten milliliters of milk was pipetted into a 50 m1. centrifuge tube and 5 m1. of TBA reagent was added. After thorough mixing, the tube was placed in a boiling water bathlbr 10 min. Upon removal from the water bath, the tube was cooled immediately in cold water. After cooling, 15 m1. of the extraction mixture was added to the tube. The tube was stoppered and shaken vigorously for at least 30 sec. after which it was centrifuged for 5 min. at 3,000 r.p.m. in a centrifuge having a 16 in. peripheral diameter. Part of the clear solvent layer was removed from the tube by means of a pipette, and transferred to a cuvette. The absorbancy of the solvent layer was determined at 535 mu in a spectophotometer. 24 Determination g: Peroxide Values Peroxide values were also determined at 0, 2, 4, and 8 days after treatment by the method of Hills and Thiel (25), t al. (51) and Pont (42). as modified by Stine a) Reagents: 1) De-emulsification reagent. The de-emulsification reagent was prepared by dissolving 50 g. of sodium citrate, 50 g. of sodium salicylate, and 86 ml. of n-butanol in 450 m1. of distilled water. All chemicals were A. R. grade and the n-butanol was redistilled. 2) Benzene-methanol solvent. A mixture of 70 volumes of thiophene free benzene and 30 volumes of C. P. methanol was employed. The benzene was redistilled and the methanol was dried by refluxing for 4 hours with magnesium ribbon (5 g. per liter), followed by distillation. 3) Ferrous chloride solution (approximately 0.014 M). Hydrated barium chloride (0.4 9.) dissolved in 50 ml. of water was added slowly, with stirring, to a solution of 0.5 g. of hydrated ferrous sulfate in 50 ml. of water. Finally, 2 ml. of 10 N hydrochloric acid was added. The precipitated barium sulfate was allowed to settle, and the clear solution was decanted into a brown glass bottle and stored in a refrigerator. All reagents were as free as possible of ferric iron. 25 4) Ammonium thiocyanate solution. Thirty grams of A. R. ammonium thiocyanate was dissolved in water and made to a volume of 100 ml. The solution was transferred to a brown glass bottle and stored in a refrigerator. b) Procedure. Thirty milliliters of milk was added to a 9 g. Babcock cream test bottle, followed by 15 ml. of the de-emulsification reagent. After gentle agitation to assure thorough mixing of the reagent and the milk, the test bottle was heated in a water bath at 70° C. for 10 min., and was centrifuged in a Babcock centrifuge for l min. The fat was brought into the neck of the flask by the gentle addition of hot distilled water. The test bottle was centrifuged again for l min. The fat was tempered by immersing the test bottle to the top of the fat line in a water bath at 45° C. for 5 min. A 0.5 ml. sample of the tempered butter oil was removed from the cream test bottle with a 0.5 ml. Ostwald—Folin pipette and placed in a 10 m1. standard taper volumetric flask. After the addition. Q E“ ‘-W) (J ‘1 5)) ’,I 03 006— ,r” E x \ o E g . + .1 0 .° 93% -- 0 o . e e 0 E5 -~\\\‘~‘~ 0(14 __ -_° ~—.<> 0.02 J 1 l J 2 4 6 8 STORAGE (days at 35° F. ) Figure 1. Thiobarbituric acid values of fluid milk heated to 1430 F. for 30 minutes with additions made before the heat treatment. 33 (124 —— OJB —- ," O.|2 PEROXIDE VALUES (milliequivalents 02/kg far) . . . . Control -—- ‘.lppm Cu + on EDTA 0.06 — A IOOX EDTA 0 IOOOX EDTA 0 4349x EDTA 0.0 1 1 1 J 2 4 6 a STORAGE (days at 35° F. ) Figure 2. Peroxide values of fluid milk heated to 143° F. for 30 minutes with additions made before the heat treatment. 01r— 0.08 - ODl>+ THIOBARBITURIC ACID VALUES (absorbancy or 535 my) 0.04 *- o.02 1 2 34 Control .5 ppm Cu IOX EDTA IOOX EDTA l000X EDTA 4349X EDTA 1 1 o 4 6 8 STORAGE (days of 35° F.) Figure 3. Thiobagbituric acid values of fluid milk heated to 143 F. for 30 minutes with additions made before heat treatment. PEROXIDE VALUES (milliequi valents O2 / kg. far) Figure 4. 0.24 0.|8 .0 up 0.06 0.0 35 I I I I I I h— I I I I I aura-D” + + .00 00000000000000 11% C1 )-- o o o o COHIVOI A IOOX EDTA —-- .5 ppm Cu 0 I000)( EDTA + “3" EDTA 0 4349x EDTA l l 1 CJ 2 4 6 8 STORAGE (days at 35° F.) Peroxide values of fluid milk heated to 143° F, for 30 minutes with additions made before the heat treatment. 36 TABLE 1 . . . a Influence 9: EDTA 9g the Stabilization 9: Flavor 1 Coppe - treated Milk Pasteurized a; 1430 E. for 39_Minutes. (Additions made before pasteurization) Copper Storage (days) (p.p.m.) 0 2 4 8 Control none - - _ _ none - - _ _ Added Copper 0.1 - + + + 0.5 _ ? + + Added copper plus 10 X EDTA 0.1 - + ? ? 0.5 - - - - Added copper plus 100 X_§QTA_ 0.1 - - - - 0.5 - - - — Added copperpplus 1000 X EDTA 0.1 - - - - 0.5 - - — — Added copper plus 4349 X EDTA 0.1 - - - foreign 0.5 - foreign foreign foreign - no oxidized flavor, ? questionable oxidized flavor,-+ slight oxidized flavor, ++ pronounced oxidized flavor, +++ very strong oxidized flavor. 37 TABLE 2 Influence gf_EDTA 9p the Stabilization 9: Flavor i3 Copper- treated Milk Pasteurized g5 143° F. for 39_Minutes. (Additions made after pasteurization) Storage Thiobarbituric Peroxide Flavor (days) acid valuesa valuesb scoresc Control 0 0.048 0.041 - 2 0.040 0.047 - 4 0.052 0.050 - 8 0.057 0.187 - 0.1 p.p.m. added copper 0 0.048 0.041 — 2 0.060 0.119 ++ 4 0.071 0.109 ++ 8 0.070 0.231 + 0.1 p.p.m. added cgpper plus 1 X EDTA 0 0.048 0.041 - 2 0.070 0.131 + 4 0.068 0.109 + 8 0.063 0.214 + 0.1 ppp.m. added copper plus 5 X EDTA 0 0.048 0.041 - 2 0.067 0.111 ? 4 0.065 0.109 + 8 0.060 0.178 ? 0.1Ap.p.m. added copper plus 10 X EDTA 0 0.048 0.041 - 2 0.053 0.114 ? . 4 0.058 0.113 + 8 0.054 0.177 - TABLE 2 (Continued) 38 Storage Thiobarbituric Peroxide Flavor (days) acid values valuesb scoresC 0.1 p.p.m. added copper plus 100 X EDTA 0 0.048 0.041 - 2 0.047 0.092 - 4 0.047 0.084 ? 8 0.047 0.153 - 0.1 p.p.m. added copperpplus 1000 X EDTA 0 0.048 0.041 - 2 0.042 0.063 - 4 0.040 0.050 - 8 0.038 0.106 - 0.1pp.p.m. added copper plus 4349 X EDTA 0 0.048 0.041 - 2 0.042 0.048 foreign 4 0.035 0.050 foreign 8 0.039 0.119 foreign aAbsorbancy at 535 mu. bMilliequivalents of 02/kg. fat. °- no oxidized flavor, ? questionable oxidized flavor, + slight oxidized flavor, ++ pronounced oxidized flavor, +++ very strong oxidized flavor. TABLE 3 39 Influence 9; EDTA 2a the Stabilization 9: Flavor lg Copper— treated Milk Pasteurized g; 1430 F. for 39_Minutes. (Additions made after pasteurization) Storage Thiobarbituric Peroxide Flavor (days) acid valuesa valuesb scoresC Control 0 0.049 0.132 - 2 0.047 0.122 - 4 0.063 0.112 — 8 0.059 0.149 - 0.5 p.p.m. added copper 0 0.049 0.132 - 2 0.062 0.030 + 4 0.079 0.152 + 8 0.080 0.290 ++ 0.5_p;p.m. added copper plus 100 X EDTA 0 0.049 0.132 - 2 0.055 0.122 ? 4 0.061 0.111 — 8 0.053 0.119 - 0.5 p.p.m. added copper plus 1000 X EDTA 0 0.049 ' 0.132 — 2 0.053 0.128 ? 4 0.056 0.118 - 8 0.050 0.113 - 0.5 p.p.m. added copper plus 4349 X EDTA 0 0.049 0.132 - 2 0.050 0.134 foreign 4 0.043 0.104 foreign 8 0.033 0.121 foreign aAbsorbancy at 535 mu. bMilliequivalents of 02/kg. fat. °- no oxidized flavor, ? questionable oxidized flavor, + slight oxidized flavor, ++ pronounced oxidized flavor, +++ very strong oxidized flavor. 40 Effect g§_carboxymethylmercaptosuccinic acid (CMMS) gp_ inhibition 9: oxidized flavor development. The results of adding CMMS to fluid milk before pasteurization are shown in Figures 5, 6, 7, and 8. Data in these figures show the results of the chemical measurement of fat oxidation when the concen— tration of the additions of CMMS was based either on 0.1 or on 0.5 p.p.m. added copper. Obviously, CMMS did not improve the oxidative stability of the milk fat under any of the con- ditions studied. In some instances the CMMS appeared to promote the oxidative deterioration of the milk fat. However, the available data do not lend themselves to more than an indication of this reaction. The most effective concentrations of CMMS were found to be those which were equal to 10 or 100 X the amount of added copper. The flavor scores, recorded in Table 4, show that CMMS, at any of the concentrations studied, did not improve the flavor of the copper-treated milk. A foreign flavor was noted when additions of CMMS were made at 1000 and at 4349 X the amount of added copper. Undoubtedly, this foreign flavor was due to the relatively large amount of CMMS added to the milk. While the lower concentrations of CMMS appeared to be more effective inhibitors of the oxidized flavor department, as shown by chemical analysis, they seemed to have had no influence on the development of the oxidized flavor when measured by organoleptic analysis. 41 When the additions of the CMMS and copper were made gfpgg pasteurization, there was only a slight increase in the effectiveness of the CMMS as an inhibitor of the oxidized flavor. These data are shown in Tables 5 and 6. The lower concentrations of CMMS again appeared to be the most effective in improving the oxidative stability and flavor of the fluid milk. THIOBARBITURIC ACID VALUES (absorbancy at 535 mu) Figure 5. 42 0.1 — /<) .I25 0 0.08 0.06 - ° ° ° Control --- .I ppm Cu 0.04 e- + IOX CMMS A IOOX CMMS D IOOOX CMMS O 4349x CMMS ' 2 4 6 8 STORAGE (days at 35‘ E) Thiobarbituric acid values of fluid milk heated to 1430 F. for 30 minutes with additions made before the heat treatment. PE ROX IDE VALUES (milliequivalents 02/ kg. fat) Figure 6. 0.24 e 0 .422 o .4I2 0.l8 0.|2 0.06 - D .333 D .3IB b 0................O ' ' ' ' Control --- .Ippm Cu IOX CMMS IOOX CMMS IOOOX CMMS Onl>+ 4349X CMMS 0.0 l L l 2 4 6 STORAGE (days at 35° F.) peroxide values of fluid milk heated to 143° F. for 30 minutes with additions made before the heat treatment. 43 .542 0 .43I D 44 0l—— 0 .I280 0.08 e m u: 2) 4“ §é o 6% '3 . . , . ° 9 E . . - ' o to . - <1 {3 , . - ' 9 ._ 0.06 .° - t 0 ,' E a -' t 0 3'2 3 .- < s . m g . Q '8 .° ° - ° 0 Control ::.\— .' h --- .l ppm Cu 0.04 + IOX MLCMMS A Ioox MLCMMS D IOOOX MLCMMS 0 4349X MLCMMS 0.02 1 l . 1 J 2 ' 4 6 8 STORAGE (days at 35° F. ) Figure 9. Thiobarbituric acid values of fluid milk heated to 143° F. for 30 minutes with additions made before the heat treatment. 53 .3l l A .248 +- .260 - U ‘A’///,/’//° ./ / OI8 F /’ ‘N // + to. I . e ,/ co 0" I lu : I / 3 N .1 o I § I e *3 ’ " Q) I o S 1: °°'2 * , . Q g , o ‘ . O . . O a: q) I . . o ' m z: I . o Q.== . E l ' \— I .0 I, O .. . o . - Control I 3. __... .l ppm Cu 0.06 L— I’ - ,' + IOX MLCMMS . . .' A IOOX MLCMMS ' ° 0 IOOOX MLCMMS O 4349X MLCMMS (10 L l l J 2 4 6 8 STORAGE (days at 35° F. ) Figure 10. Peroxide values of fluid milk heated to 143° F. for 30 minutes with additions made before the heat treatment. 54 0.l 0.08 i (n kl :3 4 § .4 e E" L) < a 0.06 2 a E a. t 8 a; D o: 2 <1 8 Control 8 e -' ‘ "' ° --- .5 m Cu 33 PP 004 + on MLCMMS IOOX MLCMMS E D IOOOX MLCMMS l ogugi J l l J 2 4 6 8 STORAGE (days at 35' F. ) Figure 11. Thiobarbituric acid values of fluid milk heated to 1430 F. for 30 minutes with additions made before heat treatment. 55 0.24 O.I8 2 e 3' fl3~e. 3 0‘“ § :2 E %» O.|2 x .> o a E e 0- E E . . . . Control --- .5ppm Cu 0.06 + IOX MLCMMS A IOOX MLCMMS D IOOOX MLCMMS 0.0 l l J J 2 4 6 8 STORAGE (days at 35° F.) Figure 12. peroxide values of fluid milk heated to 143° F. for 30 minutes with additions made before the heat treatment. 56 TABLE 7 Influence pf MLCMMS 93 the Stabilization g: Flavoralig Copper- treated Milk Pasteurized 3; 143° F. for ;Q_Minutes. (Additions made before pasteurization) Storage (days) Copper (p-p-m-) 0 2 4 8 Control none - - - - none - — - — Added copper 0.1 - + + ++ 0.5 - + + ++ Added copper plus 10 X MLCMMS 0.1 — + + + 0.5 — ++ + ++ Added copper plus 100 X MLCMMS . — + ++ ++ 0.5 _ ++ ++ +++ Added copper plus 1000 X MLCMMS 0.1 — ++ ++ +++ 0.5 - foreign foreign foreign Added copperpplus 4349 X MLCMMS 0.1 - foreign foreign foreign °- no oxidized flavor, ? questionable oxidized flavor, + slight oxidized flavor, ++ pronounced oxidized flavor, +++ very strong oxidized flavor. TAB LE 8 57 Influence g£_MLCMMS‘gg the Stabilization 2; Flavor 1g Copper- treated Milk Pasteurized g; 1430 F, for §Q_Minutes. (Additions made after pasteurization) Storage Thiobarbituric Peroxide Flavor (days) acid valuesa valuesb scoresC Control 0 0.025 0.004 — 2 0.038 0.012 - 4 0.038 0.033 — 8 0.035 0.037 - 0.1_p.p.m. added copper 0 0.025 0.004 — 2 0.041 0.023 + 4 0.047 0.050 + 8 0.039 0.050 + 0.1 p.p,m. added cppper plus 5 X MLCMMS 0 0.025 0.004 — 2 0.048 0.042 + 4 0.060 0.052 + 8 0.046 0.042 + 0.1 p p.m. added copper plus 10 X MLCMMS 0 0.025 0.004 - 2 0.040 0.039 - 4 0.042 0.033 + 8 0.045 0.039 + 0.1 ppp.m. added copper plus 100 X MLCMMS 0 0.025 0.004 - 2 0.055 0.094 + 4 0.058 0.092 ++ 8 0.055 0.084 ++ 58 TABLE 8 (Continued) Storage Thiobarbituric Peroxide Flavor (days) acid valuesa valuesb scoresc 0.14p.pgm. added copper plus 1000 X MLCMMS 0 0.025 0.004 - 2 0.098 0.113 ++ 4 0.093 0.113 ++ 8 0.078 0.122 ++ 0.1 ppp,m. added copper plus 4349 X MLCMMS 0 0.025 0.004 - 2 0.098 0.094 foreign 4 0.102 0.088 foreign 8 0.065 0.096 foreign aAbsorbancy at 535 mu. bMilliequivalents of OZ/kg. fat. °- no oxidized flavor, ? questionable oxidized flavor, + Slight oxidized flavor, ++ pronounced oxidized flavor, +++ very strong oxidized flavor. 59 TABLE 9 Influence gf.MLCMMS gg the Stabilization 9; Flavor i Co er- treated Milk Pasteurized g3 1430 F. for 39_Minutes. (Additions made after pasteurization) Storage Thiobarbituric Peroxide Flavor (days) acid valuesa valuesb scoresC Control 0 0.025 0.004 - 2 0.038 0.012 — 4 0.038 0.033 - 8 0.035 0.037 — 0.5 p.p,m. added copper 0 0.025 0.004 - 2 0.042 0.031 + 4 0.053 0.063 + 8 0.063 0.176 +++ 0.5 p.p.m. added copper plus 57X MLCMMS 0 0.025 0.004 - 2 0.042 0.035 - 4 0.053 0.084 + 8 0.060 0.201 + 0.5 p.p.m. added copper plus 10 X MLCMMS 0 0.025 0.004 - 2 0.038 0.035 - 4 0.051 0.096 + 8 0.065 0.193 ++ 0.5 p,p.m. added coppergplus 100_X MLCMMS 0 0.025 0.004 - 2 0.054 0.082 + 4 0.071 0.168 ++ 8 0.090 0.269 ++ 60 TABLE 9 (Continued) Storage Thiobarbituric Peroxide Flavor (days) acid valuesa values scoresC 0.5 p.p.m. added copper plus 1000 X MLCMMS o 0.025 0.004 - 2 0.050 0.065 +t 4 0.066 0.105 +++ 8 0.077 0.164 +++ °Absorbancy at 535 mu. bMilliequivalents of Oz/kg. fat. °- no oxidized flavor, 2 questionable oxidized flavor, + slight oxidized flavor, ++ pronounced oxidized flavor, +++ very strong oxidized flavor. 61 Treatment 9; HT-ST Pasteurized Milk Many of the experimental trials conducted on holder pasteurized milk were carried out also on aliquot portions of the milk pasteurized at 161° F. for 15 sec. Again, addi— tions of 0.1 or 0.5 p.p.m. added copper were made either before or after the pasteurization treatment. Data obtained on these studies are reported in Tables 10, ll, 12, and 13. From the data reported therein, one observes no significant differences in the effectiveness of the chelating agents in inhibiting the oxidized flavor due to the pasteurization exposures. The comparison of data in Tables 10 and 11 with those in Tables 1 and 2 and Figures 1 and 2 show the difference in the two pasteurization treatments when EDTA was employed as the chelating agent. From these data, one may conclude that when the milk was pasteurized 161° F. for 15 sec. and additions were made pgigp to pasteurization, the effectiveness of EDTA was slightly enhanced. When the additions were made pipe; pasteurization, the effectiveness of the EDTA in inhibiting the oxidized flavor did not appear to be altered by the dif- ferent pasteurization treatments. The comparison of data in Tables 12 and 13, with those in Tables 4 and 5, and in Figures 5 and 6 illustrates the dif- ference in the two pasteurization treatments when CMMS was 62 used as the chelating agent. From these data one may conclude that the difference in the pasteurization treatment had very little influence on the effectiveness of the CMMS in inhibiting the development of the oxidized flavor. This fact was observed to be true when additions were made either before or pipe; the pasteurization treatment. Since there appears to be no significant difference in the effectiveness of the chelating agents studied from the dif— ferent pasteurization treatments, only the data from experi— mental trials pasteurized at 161° F. for 15 sec. to which EDTA and CMMS had been added at concentrations based on 0.1 p.p.m. added copper are presented (Tables 10, 11, 12, and 13). Copper Determinations The milk used in these trials was safeguarded throughout the studies against copper contamination. The milk contacted nothing but stainless steel containers and utensils. However, copper determinations were run routinely on the milk samples to insure that contamination of the whole processed milk had not occurred. The copper content of the processed milk, used in this research, was found to range from 40 to 60 ug. per liter. This range is well within that reported for good quality milk (27). 63 TABLE 10 Influence we: EDTA 23 the Stabilization of Flavor i3 Copper- treated Milk Pasteurized 33 1610 E. for 15 Seconds. (Additions made before pasteurization) Storage Thiobarbituric Peroxide Flavor (days) acid valuesa valuesb scoresC Control 0 0.057 0.132 - 2 0.078 0.149 - 4 0.060 0.201 - 8 0.089 0.162 - . 0.1 p.p.m. added copper 0 0.057 0.132 - 2 0.091 0.177 + 4 0.069 0.241 + 8 0.097 0.217 ++ 0.1 p.p.m. added copper plus 1 X EDTA 0 0.057 0.132 - 2 0.084 0.176 + 4 0.069 0.241 + 8 0.092 0.215 + 0.1 p.p.m. added copper plus 5 X EDTA 0 0.057 0.132 - 2 0.088 0.176 ? 4 0.069 0.227 + 8 0.092 0.188 + 0.1 p.p.m. added copper plus 10 X EDTA 0 0.057 0.132 - 2 0.082 0.180 - 4 0.069 0.227 + 8 0.085 0.208 - 64 TABLE 10 (Continued) _7 Storage Thiobarbituric Peroxide Flavor (days) acid valuesa valuesb scoresC 0.1 p.p.m. added copper plus 100 X EDTA 0 0.057 0.132 - 2 0.069 0.172 - 4 0.052 0.227 - 8 0.060 0.185 - 0.1 p.p.m. added copper plus 1000 X EDTA 0 0.057 0.132 - 2 0.045 0.172 - 4 0.043 0.227 - 8 0.050 0.175 - 0.1 pep.m. added coppereplus 4349 X EDTA 0 0.057 0.132 - 2 0.051 0.206 - 4 0.043 0.280 — 8 0.058 0.190 - aAbsorbancy at 535 mu. bMilliequivalents of OZ/kg. fat. °— no oxidized flavor, ? questionable oxidized flavor, + slight oxidized flavor, ++ pronounced oxidized flavor, +++ very strong oxidized flavor. 65 TABLE 11 Influence 9: EDTA 93 the Stabilization 2: Flavor ig Copper- treated Milk Pasteurized e; 1610 F. for 15 Seconds. (Additions made after pasteurization) Storage Thiobarbituric Peroxide Flavor (days) acid valuesa valuesb scores Control '0 0.050 0.079 — 2 0.051 0.088 - 4 0.042 0.133 - 8 0.065 0.143 - 0.1 p,p.m. added copper 0 0.050 0.079 - 2 0.078 0.138 + 4 0.074 0.197 ++ 8 0.090 0.203 ++ 0.1pp.p.m. added copper plus 1 X EDTA 0 0.050 0.079 - 2 0.097 0.138 + 4 0.078 0.197 ++ 8 0.089 0.196 + 0.1 pep.m. added coppereplus 5 X EDTA 0 0.050 0.079 - 2 0.089 0.138 + 4 0.078 0.193 ++ 8 0.089 0.214 + 0.1 p.p.m. added cepper plus 10 X EDTA 0 0.050 0.079 - 2 0.092 0.138 ? 4 0.078 0.199 + 8 0.080 0.227 ? 66 TABLE 11 (Continued) 4 r Peroxide Flavor Storage Thiobarbituric valuesb scores (days) acid valuesa C 0.14p.p.m. added copper plus 100 X EDTA 0 0.050 0.079 - 2 0.081 0.128 - 4 0.063 0.180 - 8 0.060 0.220 - 0.1 p.p.m. added copper plus 1000 X EDTA 0 0.050 0.079 - 2 0.068 0.097 - 4 0.043 0.147 - 8 0.053 0.149 - 0.1 p.p.m. added copper plus 4349 X EDTA 0 0.050 0.079 - 2 0.055 0.096 - 4 0.040 0.123 - 8 0.039 0.134 foreign aAbsorbancy at 535 mu. bMilliequivalents of OZ/kg. fat. °- no oxidized flavor, ? questionable oxidized flavor, + slight oxidized flavor, ++ pronounced oxidized flavor, +++ very strong oxidized flavor. 67 TABLE 12 Influence g: CMMS gp_the Stabilization pf Flavor 13 Copper- treated Milk Pasteurized ep_l61°°§. for 15 Seconds. (Additions made before pasteurization) Storage Thiobarbituric Peroxide Flavor (days) acid valuesa values scores Control 0 0.057 0.132 - 2 0.078 0.149 - 4 0.060 0.201 - 8 0.089 0.162 - 0.1 p.p.m. added copper 0 0.057 0.132 - 2 0.091 0.177 + 4 0.069 0.241 + 8 0.097 0.217 ++ 0.1 p.p.m. added cepper plus 1 X CMMS 0 0.057 0.132 - 2 0.080 01219 + 4 0.069 0.265 + 8 0.079 0.217 + 0.1 pep.m. added copper plus 5 X CMMS 0 0.057 0.132 - 2 0.080 0.228 + 4 0.070 0.277 + 8 0.080 0.234 + 0.1 p.p.m. added copper plus 10 X CMMS 0 0.057 0.132 - 2 0.080 0.212 + 4 0.069 0.267 ‘ + 8 0.081 0.226 + 68 TABLE 12 (Continued) Storage Thiobarbituric Peroxide Flavor (days) acid valuesa valuesb scoresc 0.1 p.p.m. added copper plus 100 X CMMS 0 0.057 0.132 - 2 0.081 0.218 + 4 0.070 0.269 + 8 0.093 0.244 + 0.1 .p.m. added copper plus 1000 X CMMS 0 0.057 0.132 - 2 0.072 0.235 foreign 4 0.078 0.301 foreign 8 0.107 0.262 foreign 0.14p.p.m. added cgppereplus 4349 X CMMS 0 0.057 0.132 - 2 0.093 0.353 foreign 4 0.088 0.403 foreign 8 0.122 0.377 foreign aAbsorbancy at 535 mu. bMilliequivalents of 02/kg. fat. °— no oxidized flavor, ? questionable oxidized flavor, + slight oxidized flavor, ++ pronounced oxidized foavor, +++ very strong oxidized flavor. 69 TABLE 13 Influence g: CMMS gp_the Stabilization 9: Flavor 13 Copper- treated Milk Pasteurized e; 1610 E. for 15 Seconds. (Additions made after pasteurization) _— _— Storage Thiobarbituric Peroxide Flavor (days) acid valuesa valuesb scoresc Control 0 0.050 0.079 - 2 0.051 0.088 - 4 0.042 0.133 - 8 0.065 0.143 — 0.1 p.p.m. added copper 0 0.050 0.079 — 2 0.078 0.138 + 4 0.074 0.197 ++ 8 0.090 0.203 ++ 0.1 p,p.m. added copper plus 1 X CMMS 0 0.050 0.079 - 2 0.070 0.146 + 4 0.080 0.218 ++ 8 0.092 0.229 + 0.l_p,p.m. added copper plus 5X CMMS 0 0.050 0.079 - 2 0.070 0.155 + 4 0.069 0.218 ++ 8 0.090 0.216 + 0.1pp.p.m. added copper plus 10 X CMMS 0 0.050 0.079 - 2 0.062 0.159 + 4 0.079 0.218 + 8 0.079 0.265 - 70 TABLE 13 (Continued) Peroxide Flavor valuesb scores Storage Thiobarbituric (days) acid valuesa c 0.1 p.p.m. added copper plus 100 X CMMS 0 0.050 0.079 - 2 0.070 0.168 + 4 0.073 0.214 + 8 0.080 0.265 - 0.1_p.p:m. added copper plus 1000 X CMMS 0 0.050 0.079 - 2 0.077 0.178 foreign 4 0.092 0.237 foreign 8 0.091 0.309 foreign 0.1 p.pem. added copper plus 4349 X CMMS 0 0.050 0.079 - 2 0.069 0.227 foreign 4 0.092 0.298 foreign 8 0.087 0.355 foreign °Absorbancy at 535 mu. bMilliequivalents of 02/kg. fat. °— no oxidized flavor, ? questionable oxidized flavor, + slight oxidized flavor, ++ pronounced oxidized flavor, +++ very strong oxidized flavor. 71 DISCUSSION The Calcium-disodium Salt of Ethylenediaminetetraacetic Acid (EDTA) While working with three different salts of ethylenediamine- tetraacetic acid, namely the disodium salt, the tetra sodium salt, and the calcium-disodium salt, Arrington and Krienke (l, 2) found that the quantity of ethylenediaminetetraacetic acid required for complete protection against the oxidized flavor was about five times the amount of added copper on a molar basis. King and Dunkley (29) reported that 2 g. of "EDTA" per liter was effective in preventing the oxidized flavor even with 0.1 p.p.m. added copper. In the work reported herein, an attempt was made to duplicate both of the recom- mended concentrations. Since King and Dunkley did not report the type of "EDTA" used in their studies, the assumption was made that they used the tetracarboxylic acid form. On the basis of the molecular weight of ethylenediaminetetraacetic acid, a concentration of 4349 X the amount of added copper appeared to approximate that used by King and Dunkley. The data obtained in the studies reported herein do not substantiate, in toto, those secured by the previous investi- gations. The concentration of EDTA necessary to inhibit completely the oxidized flavor development without imparting a foreign flavor to the milk was inbetween those concentrations 72 reported by other workers. EDTA, added either at 100 or at 1000 X the amount of added copper, was found in the current study to inhibit completely the development of the oxidized flavor in fluid milk stored for 8 days at 35° F. Furthermore, these concentrations did not impart any foreign flavor to the milk. By the end of an 8-day storage period, EDTA, at the levels indicated, appeared to have lost some of its inhibitory action on the development of the oxidized flavor. For some reason or other, the oxidative deterioration of the samples at 8-days storage was greater than at earlier periods in the storage life. This is in agreement with the findings of Stine (50), who studied the disodium salt of ethylenediaminetetraacetic acid as a possible inhibitor of oxidative deterioration in dry whole milk. He found that while the disodium salt of ethylene— diaminetetraacetic acid initially was effective in inhibiting oxidation of dry whole milk, the effectiveness of the disodium salt of ethylenediaminetetraacetic acid decreased proportionally with the length of the storage period. However, since beverage milk is usually marketed within 8 days after processing, EDTA, from a practical standpoint, might still be a very effective inhibitor of the oxidized flavor in fluid milk. 73 While many of the salts of ethylenediaminetetraacetic acid are known to have a very low order of toxicity, the calcium-disodium salt of ethylenediaminetetraacetic acid is the only salt presently approved as a food additive by the Federal Food and Drug Administration. Carboxymethylmercaptosuccinic acid (CMMS) Investigation by Evans _£H_1. (16) showed that CMMS was one of the most effective metal inactivators for use in gly- ceride oils. While they based the effectiveness of CMMS on peroxide tests and organoleptic evaluations, they observed that CMMS was unstable to thermal deodorization temperatures. Cooney e£_el, (9) reported that CMMS was not an active metal inactivator in unheated vegetable oils. They suggested that the trace metals are held within a complex of unknown structure which makes them unavailable for complexing by metal inacti- vators until after heating. Stine (50) found that CMMS was a more effective metal inactivator than EDTA in improving the oxidative stability of dry whole milk when preheating treatments of 190° F. for 10 min. were applied to the fluid whole milk. However, as our results indicate, CMMS was not an effec- tive inhibitor of the oxidized flavor in beverage milk at either of the pasteurization treatments employed. The possibility 74 exists that the pasteurization treatments employed did not sufficiently heat the milk to activate the metal scavenging ability of the CMMS reported by other workers. The heat treatments used by all these workers were considerably greater than any heat treatments employed in the present research. Due to the effectiveness of EDTA in inhibiting the oxidized flavor in the beverage milk studied, it does not appear likely that the complexes of unknown structure described by Cooney _E.§£- (9) could be interfering with the availability of the metallic ions. However, there would appear to be some correlation between the heat treatment the product receives and the metal inactivating ability of CMMS. As pointed out by Evans e3 3;. (16), indications are that CMMS has an extremely low order of toxicity. The Monolauryl Ester g: Carboxymethylmercaptosuccinic Acid (MLCMMS) While there has been a limited amount of work done on the application of MLCMMS as a metal inactivator in natural fats, one might reasonably believe that MLCMMS could prove to be an effective inhibitor of the oxidized flavor in fluid milk. Since CMMS is a water—soluble compound, logic dictates that CMMS would be rather ineffective in chelating metals should the metallic ions be dissolved in the lipid phase. For this reason, in MLCMMS the carboxyl group beta to the sulfur atom 75 of CMMS has been esterified with lauric acid to yield the fat- soluble monoester of CMMS. Such a compound as MLCMMS, with both hydrophilic and lipophilic characteristics, might be expected to chelate effectively metals in the fat itself. However, as reported in this thesis, MLCMMS failed to inhibit the development of the oxidized flavor in fluid milk. However, milk fat, composed mainly of triglycerides, is present in milk as discrete globules on which are adsorbed molecules of many surface-active materials such as phospholipids, proteins, and enzymes. The lipophilic character of MLCMMS was of no advantage in protecting the milk fat from oxidation. Possibly the MLCMMS was unable tr>che1ate the metallic ions at a critical site or dissolve itself in the fat globules because of the adsorbed membranes surrounding the globules. Thus, if the lipophilic characteristics of MLCMMS were unable to function, the effectiveness of MLCMMS in inhibiting the oxidized flavor in fluid milk would not be expected to be greater than that of CMMS. 76 SUMMARY AND CONCLUSIONS Samples of fluid milk were treated with either 0.1 or 0.5 p.p.m. added copper either before or ef£e£_pasteurization at 143° F. for 30 min. or at 161° F. for 15 sec. Immediately a_f£e_£ adding the copper, the chelating agents were added at concen- trations equivalent to l, 5, 10, 100, 1000, and 4349 times the amount stoichiometrically necessary to chelate the added copper. The chelating agents used were the calcium-disodium salt of ethylenediaminetetraacetic acid (EDTA); carboxymethyl- mercaptosuccinic acid (CMMS), and the monolauryl ester of carboxymethylmercaptosuccinic acid (MLCMMS). Samples were analyzed by organoleptic examination and by determination of thiobarbituric acid and peroxide values at 0, 2, 4, and 8 days. EDTA, added at concentrations of 100 or 1000 times the amount stoichiometrically necessary to chelate the added copper, was an effective inhibitor of the copper-induced oxidized flavor in the milk. EDTA added at a concentration of 4349 times the amount stoichiometrically necessary to chelate the added copper appeared to contribute an off—flavor to the milk. However, EDTA, added at concentrations lower than 100 times the amount stoichiometrically necessary to chelate the added copper did not give complete protection against the oxidative deterioration of the milk fat. 77 CMMS failed to prevent completely the development of the oxidized flavor of the fluid milk in the presence of added copper under the experimental conditions. MLCMMS, possessing both lipophilic and hydrophilic characteristics, also failed to inhibit completely the development of the oxidized flavor in the copper—treated milk. Both CMMS and MLCMMS, at concentrations of 1000 and 4349 times the amount stoichiometrically necessary to chelate the added copper, often imparted an off—flavor to the milk. 78 LITERATURE CITED Arrington, L. R., and W. A. Krienke. 1953. Effects of chelating compounds upon oxidized flavor of milk. (Abst.) Jour. Dairy Sci. 33: 571. Arrington, L. R., and W. A. Krienke. 1954. Inhibition of the oxidized flavor of milk with chelating compounds. Jour. Dairy Sci. 31: 819-824. Bernhart, F. W., and Elizabeth Linden. 1950. Effect of heat treatment on the pro-oxidant activity of copper in milk. Jour. Dairy Sci. 33:166-175. Bernheim, F., Mary L. C. Bernheim, and K. M. Wilbur. 1948. Reaction between thiobarbituric acid and the oxidation products of certain lipids. Jour. Biol. Chem. 315:257-264. Briggs, L. M. 1931. The autoxidation of butterfat. I. Factors influencing the reaction. Jour. Dairy Res. 3361-69. . Brown, W. C., and L. M. Thurston. 1940. A review of oxidation in milk and milk products as related to flavor. Jour. Dairy Sci. 33: 629-685. Cerna, E. 1956. Thiobarbituric acid test for the determination of oxidative changes in butterfat. Prumysl. Potravin..1:182-184. (In Czech.) orig. not seen; cited in Dairy Sci. Abst. 3§;781. Chilson, W. H., W. H. Martin, and C. H. Whitnah. 1950. Use of propyl gallate to defer development of oxidized flavor in market milk. Jour. Dairy Sci. 335925-928. Cooney, Patricia M., C. D. Evans, A. W. Schwab, and J.C. Cowan. 1958. Influence of heat on oxidative stability and on 10. ll. 12. 13. 14. 15. 16. 17. 79 effectiveness of metal inactivating agents in vegetable oils. Jour. Amer. Oil Chem. Soc. 33: 152-156. Corbett, W. J., and P. H. Tracy. '1940. Effect of condensing on the development of oxidized flavor. Jour. Dairy Sci. 33;209-214. Dahle, C. D., and L. S. Palmer. 1937. The oxidized flavor in milk from the individual cow. Penn. Agr. Expt. Sta. Bul. 347. 30 pp. Davies, W. L. 1932. The inactivation of lipase in dairy products by traces of heavy metal salts. Jour. Dairy Res.'3:254-263. Day, E. A., and D. A. Lillard. 1960. Autoxidation of milk lipids. I. Identification of volatile monocarbonyl compounds from autoxi- dized milk fat. Jour. Dairy Sci. 23:585-597. Dunkley, W. L., and W. G. Jennings. 1951. A procedure for application of the thiobarbituric acid test to milk. Jour. Dairy Sci. 34:1064— 1069. Dutton, H. J., A. W. Schwab, Helen A. Moser, and J. C. Cowan. 1948. The flavor problem of soybean oil. IV. Structure of compounds counteracting the effect of pro- oxidant metals. Jour. Amer. Oil Chem. Soc. 235385—388. Evans, C. D., A. W. Schwab, and Patricia M. Cooney. 1954. Metal inactivation in edible oils by carboxy- methylmercaptosuccinic acid. Jour. Amer. Oil Chem. Soc. 3359-12. Garrett, O. F. 1941. Some factors affecting the stability of certain milk properties. V. Interrelation of certain metals and metallic ions and the development of oxidized flavor in milk. Jour. Dairy Sci. 34; 103—109. l8. 19. 20. 21. 22. 23. 24. 25. 26. 27. Golding, J. 1905. Greenbank, 1938. Greenbank, 1938. Greenbank, 1940. Greenbank, 1948. Henderson, 1940. Herrington, 1948. 80 , and E. Feilmann. Taint in milk due to contamination by copper. Soc. Chem. Indus. Jour. 34:1285-1286. G. R. Detecting milk that may become oxidized. (Abst.) Jour. Dairy Sci. 33:143. G. R. The relation of oxidation-reduction potential to oxidized flavor in milk. (Abst.) Jour. Dairy Sci. 33:144. G. R. Variation in the oxidation—reduction potential as a cause for the oxidized flavor in milk. Jour. Dairy Sci. 33:725-744. G. R. The oxidized flavor in milk and dairy products: A review. Jour. Dairy Sci. 33:913-933. J. L., and C. L. Roadhouse. The influence of "white-metal" copper-nickel alloys on the flavor of milk. Jour. Dairy Sci. 33:215-220. B. L. Milk and Milk Processing. McGraw Hill Book Co. Inc., New York. 343 pp. plus vii. Hills, G. L., and C. C. Thiel. 1940. Holman, R. 1954. Jenness, R. 1959.- The toxic thiocyanate method of estimating" peroxide in the fat of butter, milk, and dried milk. Jour. Dairy Res. 33:340-353. T., W. 0. Lundberg, and T. Malkin. Progress lg the Chemistry 9; Fats and Other Lipids. Vol. II. Academic Press, Inc., New York. 347 pp. plus Vii. , and S. Patton. Principles g£_Dairy Chemistry. John Wiley & Sons, Inc., New York. 446 pp. plus viii. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 81 King, R. L., and W. L. Dunkley. 1959. Relation of natural copper in milk to incidence of spontaneous oxidized flavor. Jour. Dairy Sci. 435420-427. King, R. L., and W. L. Dunkley. 1959. Role of a chelating compound hithe inhibition of oxidized flavor. (Abst.) Jour. Dairy Sci. £35897. Krol, B. M., and P. C. denHerder. 1956. A routine method for the determination of copper in milk and in dried milk. Neth. Milk and Dairy Jour. 3:56-62. Krukovsky, V. N., and E. S. Guthrie. 1945. Ascorbic acid oxidation a key factor in the inhibition or promotion of the tallowy flavor in milk. Jour. Dairy Sci. 335565-579. Krukovsky, V. N., and E. S. Guthrie. 1946. Vitamin C, hydrogen peroxide, copper and the tallowy flavor in milk. Jour. Dairy Sci. 32:293-306. Krum, J. K., and C. R. Fellers. 1952. Clarification of wine by a sequestering agent. Food Technol. 35103-106. Lea, C. H. 1953. Oxidation defects of milk and dairy products. 13th Internatl. Dairy Cong., The Hague. 35 1037-1048. Lehmann, Barbara T., and Betty M. watts. 1951. Antioxidants in aqueous fat systems. Jour. Amer. Oil Chem. Soc. 3§;475-477. Lusas, E. W., E. W. Bird, and W. S. Rosenburger. 1956. The possibility of copper induced oxidation of milk in stainless steel—white metal systems. Jour. Dairy Sci. 33:1487-1499. Martell, A. E., and M. Calvin. 1952. Chemistry g£_the Metal Chelate Compounds. 2nd. ' Ed. Prentice Hall, Inc., New York. 613 pp. plus xi. 38. 39. 40. 41. 42. 43. 44. 45. 46. 82 Patton, S., and G. W. Kurtz. 1951. 2-thiobarbituric acid as a reagent for detecting milk fat oxidation. Jour. Dairy Sci..3g:669-674. Pont, E. G. 1951. Flavor changes and some associated chemical reactions in whole milk exposed to sunlight. Austral. Jour. Appl. Sci. 3:326-338. Pont, E. G. 1952. Studies on the origin of oxidized flavour in whole milk. Jour. Dairy Res. 33:316-327. Pont, E. G. 1953. Fat oxidation in milk. 13th Internatl. Dairy Cong., The Hague. .3:1049-1051. Pont, E. G. 1955. A de-emulsification technique for use in the peroxide test on the fat of milk, cream, concentrated and dried milks. Austral. Jour. Dairy Technol. 33:72-74. Ritter, W. 1937. Oxidative changes in milk, cream, and butter. Mitt. aus dem Geb. der Lebensmtl. Untersuch. u. Hyg. 333145 (1935). (In German) orig. not seen; cited in Dairy Sci. Abst. 33:147. Roadhouse, C. L., and J. L. Henderson. 1950. The Market-Milk Industry. 2nd. Ed. McGraw Hill Book Co. Inc., New York. 716 pp. plus xviii. Russell, D. M., and C. D. Dahle. 1943. The prevention of oxidized flavor in milk and ice cream by the use of concentrated milk products. Jour. Dairy Sci. 3§;25-35. Sandell, E. B. 1959. Chemical Analysis. 22;..;;;. Colorimetric Determinations g: Traces g; Metals. 3rd Ed. Inter-science Publications, Inc., New York. 1032 pp. plus xxii. 47. 48. 49. 50. 51. 52. 53. 54. Schwab, A. W., Cowan. J. C. 1953. Schwab, A. W., Evans. C. D. 1953. Sommer, H. 1952. Stine, C. 1957 O Stine, C. 1954. Stull, J. 1953. Stull, J. 1951. Stull, J. 1951. 83 Patricia M. Cooney, C. D. Evans, and The flavor problem of soybean oil. XII. Nitrogen coordination compounds effective in edible oil stabilization. Jour. Amer. Oil Chem. Soc. 39:177-182. Helen A. Moser, Rosemary S. Gurley, and The flavor problem of soybean oil. XIII. Sulfur coordination compounds effective in edible oil stabilization. Jour. Amer. Oil Chem. Soc. 395413-417. H. Market Milk and Related Products. 3rd. Ed. Author, Madison, Wis. 750 pp. plus xiv. M. . A study of lipid oxidation in dry whole milk. Thesis for degree of Ph.D., Univ. of Minn., Minneapolis. (Microfilmed) M., H. A. Harland, S. T. Coulter, and R. Jenness. A modified peroxide test for detection of lipid oxidation in dairy products. Jour. Dairy Sci. 31:202-208. W. The effect of light on activated flavor development and on the constituents of milk and its products: A review. Jour. Dairy Sci. 3351153-1164. W., E. O. Herreid, and P. H. Tracy. A study of the effect of nordihydroguaiaretic acid on the oxygen absorption of the phos- pholipid fraction of milk. I. The effect of concentration of antioxidant. Jour. Dairy Sci. 345181-186. W., E. O. Herreid, and P. H. Tracy. A study of the effect of nordihydroguaiaretic acid on the oxygen absorption of the phos- pholipid fraction of milk.~ II. Effect of synergists, metallic catalysts and pH. Jour. Dairy Sci. 33:187-194. 55. 56. 57. 58. 59. 84 Tappel, A. L. 1955. Catalysis of linoleate oxidation by copper— proteins. Jour. Amer. Oil Chem. Soc. 33:252-254. Tarassuk, N. P., and J. Koops. 1960. Inhibition of oxidized flavor in homogenized milk as related tothe concentration of copper and phospholipids per unit of fat globule surface. Jour. Dairy Sci. 33:93-94. Thurston, L. M. 1937. Oxidized flavor in milk. Milk Dealer 33(6): 112-118. Trout, G. M., and B. R. weinstein. 1953. The solar-activated flavor of homogenized milk. 13th Internatl. Dairy Cong., The Hague. 3: 1061-1066. Wilbur, K. M., F. Bernheim, and O. W. Shapiro. 1949. The thiobarbituric acid reagent as a test for the oxidation of unsaturated fatty acids by various agents. Arch. Biochem. 32:305-313. MICHIG TATE UNIVERSITY LIBRARIES 1111qu HI 74589 llllllrl 3