.‘2! 'F'i-AJ‘V‘I.” .v._».n. ....1' H_,:g:g::5,_,j:i:Z:_:_z:;_;,__:_::__:__, LIBRARY Michigan State University PLACE IN RETURN Box to remove this checkout from your record. TO AVOID FINES return on or before date due. ,MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 2/05 p:/ClRC/DaleDue.indd-p.1 THE FREEZING UUT UF VANILLA FLAVURS FRUM ICE CREAM. THE FREEZING UUT UF VANILLA FLAVJRS FRUM ICE CREAM. Submitted to the faculty of the Michigan State College in partial fulfillment of the requirements for the degree of Master of Science. by Ariel Cardon Merrill m 1928 ACKNuWLEDGMENTS. The writer wishes to express his indebtedness to Professors u.E. Reed and P.s. Lucas of the Dairy Depart- ment, for their valuable suggestions and help in outlining this work. He is also indebted to members of Foote and Jenks Extract Company for help and information received from them. Introduction................................... vanillaOOOCOOOOOOOOOOO0.00... The PlantOOOOOOOOOOOOOO Pollination............ Distribution...00.000.000.00.0.00.00.00.00 Classification and Distribution Chart.......... Gathering and curing................. ...... .0... Grading and Packing................. ..... . ..... The Manufacture of Vanilla Extract....... Chemistry of Vanilla.................... ....... Modified Vanilla Flavors......... ...... .. ...... Tonka Beans............ Review of Literature......... Experimental Work........... 0... ..... COOOOOOOOOOOOOO Object 0f Experiment......0..OOOOOOOOOOOOOOCOOO Plan of Experiment............................. mothOd of ProceedurGOOCOOOOOOOOOOOOCOOOOOOO DiagramNo. 1.0.0.0000000000000.00.00.000. Collecting of Experimental Data............ Experimental Data..................... TableNOO10000000000000.0.000000900000000 Discussion of Table No. I 1......OOO...0.0.0 oofimmrrumm A) re u) up n) to r4 rd h‘ r4 as i4 h‘ \o a) CD -4 a— C) -< «a -q \fl 2- \N l4 Page Table No. 2.00.0.0...OOOOOOOOOOOOOOOOOOO0.000.0. 31 Discussion of Table No. 2....................... 32 Table No. 3..................................... 33 Discussion of Table No. 3....................... 34 Table No. 4........................... ..... ..... 36 Discussion of Table No. 4....................... 38 Table No. 5..................................... 39 Discussion of Table No. 5....................... 41 Table No. 6..................................... 43 Discussion of Table No. 6....................... 44 Table No. 7..................................... 45 Discussion of Table No. 7....................... 46 Table No. 8..................................... 47 Discussion of Table No. 8....................... 48 Summary..................... ............. ............ 49 Conclusions.......................................... 51 BibliograthOO0.0.0.0....OOOOOOOOOOOOOOOOOO0.0.0.0... 52 .- ... .. 0.. 'c 0.. '0 INTRODUCTION. Ice cream, like many other foods, is not especially palatable without the addition of flavor. For this reason the organic acids, esters, and essential oils in the form of flavors have been used with it. They add the necessary zest-giving propertiescand contribute to its digestability. Ice cream and refrigeration are very closely associa- ted. It is impossible, of course, to keep ice cream in a frozen state unless refrigerated. The use of modern refrig- eration methods has caused the long storage of ice cream to become common. As a result of comparatively long storage many complex- ities have arisen. One of these is the effect of storage upon the delightful flavors used in ice cream. This is especially true of vanilla flavors. It is with a desire to study the effect of low tempera- tures and cold storage upon vanilla flavors in ice cream that this thesis has been undertaken. . --— , . . . I . I O . t . . A! . f .f . L i -.‘ ‘ .u , I. VANiLLA The vanilla bean is the fruit of an orchid, which is grown in South America and the West Indies (1). It is from the dried and cured pod of this plant that vanilla extract is made. THE PLANT. The plant itself is a vine, and although there are per- haps fifty different species, the vanilla planifolia (Andrews) is the one usually cultivated (2). This particular species is a native of Southeastern Mexico, in the valley of Mazantla. Vanilla beans from this valley are by far the most highly prized because of their delicate, pleasing flavor. This re- markable flavor is undoubtably due to the environment under which the vine grows. The ideal conditions for the growth of the vanilla plant includes temperatures of from 60°F. to 90°F., and rainfall that is evenly distributed over all seasons (3). In that portion of the tropics where it thrives, the mean temperature is about 80°F., with an average rain- fall of approximately 38 inches. Similar conditions are found in many places in the equatorial belt, but like many other plants, when moved from their natural habitat for trans-plantation in another country, they under-go modifica- tion. -3- During growth the vanilla plant shoots forth long slen- der tendrils of aerial rootlets, which cling to the object supporting it. The idea is held that the vine reouires the dense shade of a tropical forest, probably owing to the fact that it's support, when growing wild, was the forest tree. The usual support of the plant when cultivated is also the tree, principally because trees are plentiful and easy to ob- tain. The aerial rootlets condense needed moisture from the air; none is derived from their host. The main rootlets penetrate the soil in search of nourishment. The fleshy stem of the vine winds about the tree upon which it climbs, with it‘s long flat shining leaves hanging conspicuously from vine and branches. The plants bear when three years old and con- tinue bearing thirty years or more. EgLLINATIUE; Under modern cultivation the blossom of the vanilla plant is artificially pollinated. The original cultivators depended entirely upon the unreliable results of insects for this function and the effect was that some vines over-pro- duced and some under-produced with resulting loss in both quantity and quality. A French planter (4) succeeded, in 1812, in trans- planting cuttings of Mexican vanilla to the Island of Bour- bon (later named Reunion.) In this new environment the plant -4- did not produce satisfactorily, as the natural pollination agencies, bees and humming-birds, were lacking. This need was met by Professor Morren (5) of Liege University who discovered, 1837, a method by which the flowers might be artificially pollinated. This process is comparatively a simple one and is accomplished by taking a portion of the pollen from the male flowers, by means of a wood splinter, and placing it upon the female flower. As the flowers open during the night and close by mid-day, this work is carried out early in the morning. Best results are obtained if the flower is pollinated on the first day of it's opening. The flowers that are not successfully pollinated soon wither and die. It is, therefore, easy to estimate at a glance the degree of yield of the plant. This pollination process can be readily controlled according to the age and strength of the vine. DISTRIBUTLUN. From the island of Bourbon cuts of vanilla planifolia have been taken to various parts of the equatorial belt,and to the islands of the Indian ocean. The following classifica- tion as compiled by A.Adams Lund (6) gives a very complete idea of the distribution of commercial vanilla. CLASSIFICATIUN AND DISTRIBUTIUN UF CUMMERCIAL VANILLAS Botannical classifica- tion Kinds Distribution A Mexican Mexico Grows wild in the states of Vera Cruz and uaxaco. The bulk of the cultivated vanilla comes from the cantons of Papantla and Mazantla (Vera Cruz). South French- Vanilla American fleet Indies. Crown to a lesser extent plani- Guadeloupe in other West Indian Is- folia. Dominica lands, Central America, Martinique and tropical portions of South America. Javas Dutch- East Indies. Chiefly, the Island of Ceylon Island of Java. Ceylon (Br.) Bourbons Madagascar (French) Numerous islands in the Indian ucean, adjacent to the east coast of Africa. The Comores (French) Reunion (French) Seychelles (British) Mauritius (British) Vanilla Tahiti (French) The Friendly Croup. Tahitiana Tahiti~ Taiarafa (French) South Pacific ucesn. This chart needs very little, or no explanation, but it is well to add the following figures (7) which show the production of the various kinds of beans under the most favorable conditions. Mexican---------- 250,000 pounds Bourbon --------- -l,000,000 Tahiti- ---------- 220,000 N Quandeloupe------ 50,000 n Java--- ---------- 37,000 N All others ------- 12,000 N - . _ . - - . . - _ , — --,_ . _ .. . _ , _ - fl 4- -- vs.— 1 v r v -_ 7- .V — ‘— . ~- 0 1 . .. . 1 ~ I . . ‘_ .. . _. L n -- . , .. A V , O - - I ,- - ‘ 'c \ - . e ‘ U . , v.‘ e . . o . _ . . . n _ . . v ‘ ‘ r \ C C n . ' . . '( C _ .4 '\ r ‘ ' ‘ ' ' I . ,_ y ,, _ . i..y_. . . V 2 z i ,,4- ., ”a. - A... ... .F . . a e ' ‘ n ‘ ' .- 7' C . u . v , . o u ‘1' c ~ . e r v o . —_ a i . I , .- . u -__- __ _ -. .—.—-_o .- _-,__ .,, - - a O _... _ o o ‘u .--. .. \ ‘ ..,4., - -5- QATHERING AND CUPING THE VANILLA BEAN. The pods mature in six or seven months after the flower- ing period, becoming slightly hard and turning yellow at the lower end. The entire crop does not attain maturity at once, but the gathering season extends over a period of several months. This is a critical period in the development of the high-grade bean, much eXperience being necessary to detect the prOper state of maturity for gathering. If the bean is not fully matured its flavor is of an inferior quality and the bean is susceptible to mold. If the bean is over-matured it splits during the curing process and must be sold at a reduced price. The pod at the time of gathering has an of- fensive odor and the task of gathering them is not a pleasant one. The characteristic flavor and aroma is developed dur- ing the curing process. These offensive-smelling pods are taken to curing stations where they are weighed, stemmed, and graded. There have been differences of opinion among eXperts as to the best method of curing the beans and as a result many planters have their own methods. In all cases, however, the bean under-goes a curing or fermentation process much the same as tea. This is accomplished by placing them between the folds of blankets and alternating the sunning and sweat- ing processes. During these processes they turn to a dark -7- brown color, or even black, and have a soapy or waxy surface. In all localities artificial drying has now largely re- placed exposure to the sun, as it shortens the time required for curing, and gives a more uniform product. GRADING AND PACKING VANILLA. After the beans have been properly cured they are examined for splits and defective pods which are sorted out (8). The former are known as "splits" and the latter as "cuts" and are sold at a reduction, although they are better, sometimes, than the whole beans. The better and remaining portions are then graded according to length and aroma. Then they are ready to be bundled and shipped. A bundle, if made up of the best beans, is about 10 inches long and resembles the shape of a plump ear of corn. In Mexico, forty of these bundles are packed in a tin con- tainer, which is packed in a wooden case, usually of sweet- smelling cedar. In some cases the "cuts" and "splits" are bundled, but they are usually packed loose in large boxes. In Mexico the beans are taken to shipping points, by pack-mules or by train, and sent on their way to all parts of the world. Some of the beans are used direct in chocolate manufacture or in soaps and perfumes. The greater portion of the crop, however, comes to the United States where it is, for the greater part, made into extracts. -8- THE MANUFACTURE OF VANILLA EXTRACT; Chopping the beans, extracting their flavor in an alco- holic solution, and aging, are the only steps in the manu- facture of pure vanilla extract. This seems very simple, yet extract manufacturers have spent their lives trying to perfect the process. "Vanilla extract is the flavoring extract prepared from vanilla beans, with or without sugar or glycerin, and con- tains in one hundred cubic centimeters the soluble matter from not less than ten grams of the vanilla bean." This is the official definition (9), as given by the United States Department of Agriculture, with which all vanilla extracts must comply. The first problem to be considered when preparing to make an extract is the selection of the bean. The Mexican bean is regarded as supplying the highest quality of flavor, a mild, mellow, pleasing flavor. The flavor of the Bourbon bean is slightly more pronounced and for that reason is used with the Mexican bean to some ex- tent, to add to its distinctiveness and to reduce the cost. The two are so very much alike that it is quite impossible for the layman to distinguish the flavor of one from the other. The Tahiti is of inferior quality. The beans of the first crop are superior to those of the succeeding crops. -9- The ”splits" and "cuts" find their way into cheaper extracts. Finally there are the wild beans, known as vanillons. They find a use in the market, although they are of-very inferior quality. In the making of the extract, the beans are first finely chopped. This is usually done by a machine made for that pur- pose and the fineness is dependent upon the opinion of the in- dividual manufacturer. A serious problem in making extracts is the solvent to be used for digesting the bean. There are many solvents in~ eluding chloroform, carbon tetrachloride, benzene, ether, acetone, methyl alcohol, ethyl alcohol, water, and others of less importance. It has been found by experiment (10) that the last two mentioned when used in combination make the best 'solvent. Most of the others will extract the fixed oil, but not the valuable water soluble constituents of the bean. Then too, the solvents other than the last two will impart an objectional odor or flavor to the extract and some of them are decidedly injurious to the human body, when taken internal- 1y. The ratio of alcohol to water (11) in the menstruum is important and for satisfactory results it must be between thirty-five and sixty-five per cent. The color of the ex- tract becomes darker as the percentage of alcohol increases, . ' 1 . . '4‘ . . ’- 4 , J I . . a . . ‘ 1 . . - ‘ I E U ‘ . . . r s \ ‘ f , . H ~'\ . A , . , , r , . I ' ‘ ‘ ' O. o l ‘ ' . f ‘ ‘ ' V l ' I ‘ ‘ ‘ , a 0 ~ ‘ . . .. , Q I .-. . , , , a - e . - a. K I . ‘1' a ‘ ‘ e v \ . ‘ . ' . - ‘ e u. - v ' v \ . . o 0 . . i 0‘ .- 0 ‘ ' ‘ 1 | .i .i I - a ' . y . . , .. . Q ‘ r g I . ‘ . ' i . ‘ i a _ '4' . r - .. ‘ o A. l l - D . - ’ ‘ . . . ' ‘ - r ' ** . _ x . . . -10- until the alcohol content reaches about sixty per cent. After this concentration is reached, the color becomes lighter until, with ninety-five per cent alcohol, very little color is extracted. Any percentage alcohol above seventy per cent will precipitate considerable fixed oil in a fine cloud, which cannot be filtered out. If the men- struum contains more than 65% water, percolations become difficult because of dissolved gums. The Research Commit- tee of the Extract Manufacturers Association (12) recommends 47% to 50% alcohol. A menstruum such as the one recommended by the Research Committee will extract nearly all of the vanillin, volatile oils, water, soluble constituents other than gum, and most of the color and resins. It will not disolve the undesirable gum, fat, and wax. The extraction is never complete but it is not advisable to exhaust the flavoring material in the beans, as the expense of so-doing is prohibitive. There are several methods of extraction in common use. Maceration, which is the simplest and oldest, is performed by merely mixing the chopped beans with the menstruum and allowing it to soak for an indefinite period, from 24 hours to as many months. The material is then transferred to a percolater and the extract is completed. -11- In the displacement method, the beans are placed in a basket and suspended Just below the surface of the solvent. The menstruum becomes saturated and sinks to the bottom per- mitting fresh menstruum to come in contact with the beans. This process requires considerable time. The circulatory method is the one most extensively used. Some of the features of the displacement and maceration methods are often combined with it. The menstruum is pumped from beneath a screen or membrane supporting the chopped beans in a tank and is returned to the tank at the top, thus, causing a circulation of the menstruum through the beans. The menstruum is then filtered or centrifuged to remove sedi- ment. The sugar, or glycerin, or both, are then added to it and prOper dilution made for the finished product. Regardless of the method of manufacture the extract is usually stored away for aging. This is sometimes done in large glass lined tanks but more often in wooden kegs or tanks. During the aging the vanilla seems to acquire a more mellow and pleasing flavor. It is claimed by some manufacturers that the wood helps in this mellowing process and by others that it does not: thus two types of storage tanks are used. CHEMISTRY OF VANILLA; Vanillin is the principal flavoring constituent of vanilla beans (13). Other than vanillin the beans contain -12- volatile oils, sugars, gums, fats, wax, resins, acids, color- ing matter and mineral constituents. The bean contains from 1% to 3% vanillin. It is the only constituent whose chemist- ry has been thoroughly worked out. Vanillin does not occur in the green beans (14), but is developed during the curing process from the glucoside conif- erin, which is split up, probably by an enzyme, into conifer- yl alcohol and glucose. The former is oxidized during the curing process to vanillin. Chemically, vanillin is methyl-protocatechuic aldehyde. It crystallizes out in the form of white needles and has a rather sharp taste and characteristic odor.' It is water sol- uble and acts as a weak acid when in solution. Vanillin, in the presence of ferric chloride, is colored slightly blue- violet. This color reaction is taken advantage of as a quali- tative test. The other constituents of vanilla, which go to make up its flavor and odor, have defied all efforts at identification and classification. This is also true of the coloring matter. The gums, fats, and waxes are not desirable in the extract, therefore, little work has been done on their chemistry. Tahiti vanilla beans contain anisyl alcohol and piperon- al, and in this respect, differ from Mexican, Bourbon, South American and Java beans. Vanillons, like Tahiti beans con- -13- tain anisyl alcohol and piperonal. MCDIFIED VANILLA FLAVCRS. Vanilla flavors might be classed under four groups: pure extracts, compounds, imitations, and concentrates. The pure extracts of vanilla are those extracted from the beans according to the legal standards given above. Compounds are those containing vanillin, coumarin, or both, mixed with the pure extract. The imitations are those con- taining all vanillin and all coumarin, or a mixture of both. Concentrates consist of those which are of increased flavor- ing strength and might be either a compound or pure extract. A concentrated extract has become very popular in the past few years because of the economy involved. If an ex- tract is concentrated to six or eight times that of a single strength extract a great saving is effected in alcohol, or solvent used, labor, containers, transportation charges, and cost to the consumer. Where the concentrated product is made by distillation in a vacuum there is undoubtably a great loss in the volatile constituents of the extract, which results in an inferior product. However, there are some manufacturers who claim to produce a concentrate without the application of heat. Where this is possible, it is of course, highly desireable to use the concentrated extract. TQNKA BEANS. The Tanks. bean is sometimes associated with vanilla but it has neither the flavor nor true taste of pure vanilla. Its chief flavoring constituent is coumarin, a substance found in many other plants. Thus an extract with a Tonka blend is much the same as one reinforced with coumarin. Co- umarin adds to the intensity of the flavor but it is coarse and usually leaves a bad after-taste in the mouth. It also acts much the same as saccarin, deadening the sense of taste so that the unpleasent after-taste is not especially notice- able to those who habitually use it. laced because of its cheapness but its use by tobacco manufact- urers has caused its price to reach that of the best grade of Vanilla at times . The Tonka bean was first \ . v . I I n . g ' ' ‘ ‘ v - ' I > - . ' ' . 4 o . . ’ r ' . ’ ' ‘ E s ' . ‘ . ‘ . V O A ' r F. - - a A . V e .. A l' ,, . - . r (I. . ‘ e ‘ , r I- i u. U . , I ' L P ‘ V . .II - . ‘ . . r ’ ‘ T . .- o l ‘ fl. -‘ I ‘ h . - . . I n ' ' s ‘ g I. a ‘ ~ r ' A ’ i ' - _ M V . w A ’ - . \ "f . " \ - ‘ . I A . (I -‘ j 4 I ‘ . I ‘\ -'-. u ‘ . r . . ‘ ’ I .r . —‘ \ . - - .' ‘ ‘ . . I H - . k . g _ V. . I. 'l . ' ‘ ‘ . . ‘ .- ‘ e ' ' ._ ‘ I g ' r . , - 1‘ i ' ' r .- ’ r ‘ . I. . -. ' r r , c- . . a. \ ‘ T ‘ I A I .u- ' ' A -‘ > n \ ‘ -A C - A A . '1 I ‘ .‘ A ‘ , - I r - v . . I I v 1 . ’ - . ~ ' . \ " ' V ' ‘ I i- . 1 . ~ , n H b I in. i- \' - ' I ‘ . v - , . ‘ or ‘N . . I, N . I. ~. ~ . I f -. ’1 T . -- . - ' l ' ' t . | -’ W ’ ' ’ h w e .1 g _. r. \ v v 7 . . . ' ‘ - fi . REVIEW VF LITERATURE. The ice cream score card allows 40 points for flavor. This indicates that the relative importance of flavor to quality is great. In spite of this, there has been very little experimental work carried out, by the research depart- ments of the ice cream industry, which has been confined to flavors. I.B. Combs, H.B. Pierce and W.F. Borst (15) of the Pennsylvania State College did work on the use of true and imitation vanilla extracts in ice cream. They selected ten samples of vanilla at random and made a chemical analysis of them. They then tasted the extracts in an attempt to place them by taste as to their desirability and whether or not they were true or imitation extracts. Their conclusions were as follows: 1. ”Tasting extracts will not enable the average ice cream manufacturer to determine the difference between a true and an artificial extract. 2. "Tasting samples of ice cream will not enable the aver- age ice cream manufacturer to tell whether or not a true or an artificial extract has been used in the ice cream. 3. ”The experiments tend to show that true flavors per- sist better than the artificial. . . . < , , I ~ ‘ .A ' " O - o - 1 O . ,. c -_ ,- . 0-. O O O O Q . I ’ . u I . s O l ‘ g n ‘ a , g . - . t C . , l' . . .' I . v 1 ' -\ I 1. s . . . g . ‘ _ ‘0 0 " ‘ .‘ a N - O - U _ " . 4‘ " . . I ~ 0 _' x O -15- 4. "The data obtained by chemical analysis can tell little with regard to the flavoring ability of the two general types of extracts." The Research Committee of the Extract Manufacturer's Association of the United States (16) reported an eXperi- ment on the "Freezing Out of Vanilla Extracts in Ice Cream" at their 1924 convention, in which they used a uniform mix flavored with true and imitation vanilla extracts. They sum up their work as follows: "It is shown that there is not enough change in the flavor of ice cream ingredients during storage of reasonable length to materially affect the character or strength of vanilla flavor itself. In other words, vanilla does not freeze out of ice cream." In both of these experiments the opinion is that vanil- la extract does not freeze out of ice cream. The workers at Pennsylvania State College seem to think the true ex- tracts pereist better, while the Research Committee were of the opinion that the imitation extracts improved with from 7-13 days aging. -17- EXPERIMENTAL WURK. UBJECT UF EXPERIMENT. The makers of cheap vanilla flavors are undoubtably responsible for the propaganda regarding the freezing out of vanilla from ice cream. This conception has grown until there are many who firmly believe that genuine vanilla flavors freeze out while, on the other hand, there are many who are Just as firmly convinced that they do not. ' The purpose of this experiment is to determine as far as possible whether or not the delicacy or intensity of vanilla flavors are in any way altered by the storage con- ditions to which ice cream is subjected. PLAN uF EXPERIWENT. Vanilla flavors representing as many common combina- tions as thought advisable were secured for this experiment. Two forms of concentrates were used; one a true vanilla bean extract, and the other, a true vanilla bean extract reinforced with coumarin. These concentrates were six times the stand- ard strength as defined by law, which requires that the ex- tractive material from 13 1/3 ounces of vanilla beans be used for a gallon of extract. Two standard strength vanilla compounds were chosen; one a pure extract of vanilla beans reinforced with vanillin . a ' . l u . . . o f . ' v t e 0 ~ . I C C --._. _ --. .-s- --.-- ..v..,r.r'\ ," .' , . . l O “ . u. . u' ' I D . . ’ ,— -18- and the other, a pure extract of vanilla beans reinforced with coumarin. In addition to those already mentioned, four pure ex- tracts of standard strength were chosen; three were pure ex- tracts of vanilla beans, and the other, a pure extract of Tonka beans. One of the three pure extracts was made en- tirely from Mexican beans, another was made from Bourbon beans and third was half Mexican and half Bourbon beans. The compositions of all these flavors are given in Table No. 1. At different times ice cream mixes were made, as uniform- ly as possible, and flavored with these various flavors. The mix compositions are given in Table No. 2. Samples of ice cream were taken in one pint Sealright cartons and placed in the hardening room for storage. These samples were taken from the hardening room at intervals and scored for flavor, using a freshly frozen series each time to check them with. The theory has been advanced that it is the volatile constituents of vanilla flavors that freeze out of ice cream. In accordance with this theory an attempt was made to volatil- ize some of these flavors at low temperatures. Another theory which has been advanced by some is that the vanilla flavors lose some of their strength during the freezing process by volatilization. This is a second reason for the volatilization experiment. The set-up for this experi- ment is shown in Diagram No. l. The question which naturally arises is whether the effect of the low temperatures, or the effect of the ice cream in- gredients upon the flavors cause them to leave ice cream. In an attempt to answer this question samples of the flavors were placed in cold storage for long periods, and were brought out occasionally for "heat shocking." They were eventually used to flavor batches of ice cream which were scored against fresh- ly frozen check batches for flavor. "Heat shock", as used in this paper, is taken to mean the subjecting of flavors or ice cream, which has been stored at temperatures around 00F., to room temperatures for periods of various duration. An experiment was conducted to test the theory that it is "heat shocking" of ice cream and not long storage at low temperatures that causes flavor to freeze out. Samples of ice cream were taken from the hardening room at intervals and subjected to room temperature for periods of various lengths. After the samples had been treated in this manner they were scored for flavor against a freshly frozen series. -eo- An analysis was made on all ice cream mixes used in this eXperiment. This was done so that uniformity of composition of mixes might be more accurately controlled and so that any irregularity which might occur could be accounted for. This analysis was made by the Mojonnier method for fat, solids not fat, and total solids. A record was kept of the existing temperatures in the hardening room. A recording thermometer was used to obtain this record. The maximum and minimum temperatures were tabu- lated from the thermometer charts. This was done so that a knowledge might be had of the temperatures under which the samples were stored and so that any irregularities might be traced to their source. The overrun was kept as nearly constant as possible by the use of the Mojqnnier Overrun Tester. A variation in over- run would affect the coldness of the ice cream as registered in the mouth. Any other effect different overruns might have upon the flavor would be eliminated by maintaining uniformity. METHOD OF PROCEEDURE. The flavors were made according to the formulas given in Table No. 1. This was done so that the flavors used would be of known composition. They were chosen to be representative of practically all vanilla flavors. -21- The mix chosen for this work contained as near as pos- sible, under creamery methods, 12% fat, 10% milk solids not fat, 14% sugar, 0.50% gelatin, and 36.5% total solids. It was made from sweet cream and was pasteurized at 145°F. for 30 minutes. It was homogenized at l20°F. at a pressure of 2,000 pounds after which it was put in 10 gallon cans and stored at 35°F. for 36 hours. The mix was removed from the cooler and frozen in 25 pound batches using 60cc. of ex- tract in all batches except the concentrates where 10cc. per 25 pounds of mix was used. The flavor was added after the mix was placed in the freezer. The freezer was washed after every batch to prevent the carrying over of any flavor from one batch to another. One batch was frozen without any flavor with each series, to serve as a check in scoring. Samples were taken in one pint Sealright paper cartons for scoring. The date and the number representing the flavor was placed on the top and bottom of each carton for identi- fication. They were then placed in storage. These samples were taken from the hardening room after 2 to 5 days storage and scored against a freshly frozen batch for flavor. They were then returned to storage for periods varying in length from 26 to 174 days after which they were scored and returned to storage again. The third period varied V -22.. from 53 to 214 days from the first storage. These periods of various lengths were used so that a study might be made for many combinations of storage periods. In all cases the scoring was done against a freshly frozen series. This was done to eliminate the personal-taste element which varies from time to time. Samples from Series No. 12 were taken from the harden- ing room and were eXposed to room temperature for 3 hours to "heat shock" them. Two days later they were exposed at room temperature for 6 hours. They were placed in the hardening room again for two days and were then scored. Samples of extracts, in amounts equivalent to that used for each batch, were carefully measured into test tubes and placed in the hardening room. After 15 days the samples were brought out and eXposed to room temperature for 4 hours. They were again placed in the hardening room for 20 days when the four-hour exposure period was again repeated. They were then replaced in the hardening room for 24 hours, after which they were taken out and allowed to thaw. They were then used to flavor ice cream Series No. 13. Samples of the all Mexican bean, all Bourbon bean, vanil- la reinforced with coumarin, and vanilla reinforced with vanillin, were volatilized by forcing air into them through -23- a folin tube. The samples which were in test tubes were immersed in a water bath. The bath was kept at a constant temperature by means of ice, which was kept in the bath dur- ing the volatilization of the flavors. The air was supplied from a pressure tank and the process continued for 8 hours. After this was completed the flavors were used in batches of ice cream which were scored for flavor. The following diagram shows the set-up for this experiment. -24- Connecting tube Air from clamp Exhaust tube Test tube Folin tube a _ \ _ _ ......II-” Jig“... .,. . >1 .15... 92. a. -fiv“ . L V5... .. «is 2% O Illal -..: . 44.. .71 E pressure tank Test tubes Figure No.1. est tube Folin tube t 6 .1. t u 0 k r r. .1 O A 0 ..Lu dubifldfl .11., ”MW!“ e bu v Air intake . Q1%-———-LiquiH mm Hm MO 0H Hm mm bH mm Hm mb mmHMMm SpHH 4 Kr Hm bH mm Hm mm d. Kr mm bH mm Hm 0 Mo Hm Hm mm mm bH mm Hm mm mmHmmm SPOH MO Hm mm mm Acooefipsoo- m .02 mammm .44. k.» Hm 10 N .14 \ Knr. Hm Hm pH Hm $5 mMHMMm gum LC) 03 bH Om Hm mm mMMOom Hm bH Om Hm m H1 No H MO M MO m .02 Hm Hm Hm Hm Hm Hm m .03 0H 0H 0H mH mH mH b .02 t0 mm mm mm Hm Hm Hm .03 mm ma mm mm mm mm m .03 .oz (\1 '\2 CR N 01 N V“ mm mm mc bH pH pH 0H 0H 0H m .02 mm mm .mm Hm Hm Hm m .02 Hm Hm Hm Hm Hm Hm H .02 nepmm m ommnopm me when me mbH m me mm mmHmmm gum mmHmmm app -41- Dlfiflflfifilufl_uE_IABLE_Nu-5a In these scorings a change in intensity or delicacy of the vanilla flavor was the only thing considered. After a score was once established for any sample it was changed when a difference in its vanilla flavor was noted. In no single case was a change noted in any of the scorings for that particular sample. In several cases the score was raised or lowered a point for a sample but that score re- mained constant for all three scorings. This change was made when those who were doing the scoring thought the flavor had a position too high or too low in relation to other flavors. In every case the scoring was done by the same Judges. This was necessary in an eXperiment of this kind, where a change was the important thing sought, and not the desirability. In the second scoring of these samples the texture had started to change to a rough, sandy nature,_but in most cases this was not noticeable. In the third scoring the texture was decidedly rough and sandy. In some cases it was rather dry and an old flavor was present but the intensity of the vanilla flavor remained the same. The following terms were used to describe the various flavors: No. 1 Vanilla concentrate (pure extract)--— pleasing, but . . 14 .. . . as _ , r . _ - . r . f o. - r . l . m - . _ - . . . . . u . - . e x . D u- t a n . . 1 ‘ . o ' e . - u z a . l . e n . . . . , o f -42- rather light. No. 2 Vanilla concentrate (compound) --- slightly coarse, not entirely pleasing. No. 3 Vanilla reinforced with coumarin --- harsh, rough, rank, coarse, Tonka flavor. No. 4 ,Vanilla reinforced with vanillin --- rather desir- able, lacks mildness, not characteristic. No. 5 All Mexican bean extract--- mellow, pleasing, delicious, excellent, very desirable. No. 6 , All Bourbon bean extract --- lacks delicacy, not especially pleasing, slightly coarse, rather good. No. 7 All Tonka bean extract --- harsh, raw, rank, not like vanilla, repulsive. No. 8 Mexican and Bourbon bean extract--- rather good, not totally pleasing, lacks mellowness, good but not especial- ly noticeable. Batch No. 9 was frozen without any flavor each time to serve as a check. In a few cases an old flavor could be de- tected. O.” , .. ‘9 e _ I . ‘ . . I ~ "‘ .‘ 93 e . . . r . . . A . . r - . ‘x ' K ' 0 I . . .-. - . f. . I. . \ . ' a-.. ' ' ' . . . . , . ... . ., ‘ 5. ,, - 0 ~ o x I! I -- " s - - e ‘ . r . - 1' - -' 1* y! Y ‘ . , I, ‘ ,f‘c . . .7. .‘ .. . ,- . - _- . r . . ". ' I I . . I r ' - I v ' (- , . - O a .. , . Q . e , . r . e -43- TABLE No.6. ”Heat Shocking” Ice Cream Samples. Batch No. Score Decrease 1 20 1 2 21 1 3 15 2 4 20 2 5 24 l 6 20 l 7 15 1 8 20 1 very 9 rancid |. . fli' .i -44- DISCUSSIUN 0F TABLE N0. 6. The Judges questioned very much whether the intensity of the vanilla flavor was decreased or not. In some cases it seemed the delicacy had been slightly lowered and it was _ upon this basis that the scores were out. It will be seen from No. 9 that the ice cream mix had develOped some bad flavors and odors. Their presence would naturally have a tendency to subdue the delicacy of the vanilla flavor. Unless one was tasting these samples especial- ly for vanilla flavor the rancid flavors would be so pro- nounced that it would be difficult to detect the presence of the vanilla flavor. This ice cream, of course, would not be saleable. There- fore, consideration of it from a practical point of view is not important. Never-the-less, some of the propaganda con- cerning the freezing out of vanilla may be traced to this source . -45- .TABLE NO.Z. "Heat Shocking“ Flavors. Batch No. Score 1 21 2 22 3 17 4 22 S 25 6 21 7 16 8 21 -45- DlSCUSSIUN 0F TABLE N0. 7. This ice cream was also scored against a freshly frozen series. No difference in intensity or delicacy of flavor could be detected in any case. This condition would tend to show that none of the delicacy could be driven off from vanilla flavors at low temperatures. It leads to the fact that low temperatures do not affect vanilla flavors when in a free state. If there is any loss of vanilla flavor in ice cream while in storage it would be an action of the mix con- stituents upon the flavor. That this is not the case is shown by the experimental data given in Table No. 5. -47- TABLE no.8. Volatilized Flavors. Batch No. Score 3 17 4 22 5 25 ~48- DiSCUSSiUN 0F TABLE N0. 8. These were scored against a series which was frozen especially for that purpose. There was no change in vanilla flavor in any case. This would indicate that those portions of vanilla flavor which make up its intensity and delicacy are not volatile at low temperatures. The theory that vanil- la volatilizes while in the freezer would not be true accord- ing to this experiment. EXA'LL’LLIH‘fiLZMA‘.“ ..JJL‘IJ ._ if.“ --.-o -49- SUMMARY. The delicacy and intensity of the flavor was not in- jured or improved by storage, in any single case, even though the samples were stored as long as 214 days. This also means that no particular kind of flavor seemed to have the ability to improve or to lose its flavoring characteristics. This was shown by the fact that they all maintained their original score through-out the storage period. That "heat shocking" the ice cream does not alter the intensity or delicacy of the flavor is evident from the data obtained in this eXperiment. However, the increase in intens- ity of flavors from the other ingredients of the ice cream, due to decomposition, has a tendency to subdue the vanilla flavor. Long storage and subjection to "heat shocks" does not affect the ability of vanilla flavors to stay in ice cream. This is evident from the scores on the flavor of ice cream made from extracts which had been subjected to these treat- ments. Those portions of vanilla flavors which go to make up its distinctive flavoring characteristics are but slightly, if at all, volatile at temperatures as low as 00F. This was -50- shown by the scores of ice cream which were made from flavors which had air bubbled through them for twelve hours at that temperature. CoNCLUSIuNS. l. Vanilla flavors do not freeze out of ice cream during long storage at temperatures between -15°F. and.e15°F. 2. None of the flavors seemed to improve their delicacy and intensity, when subjected to low temperatures, but had the same ability to retain their original score. 3. Vanilla flavors do not freeze out of ice cream when subjected to "heat shocks." 4. Vanilla flavors do not lose any flavoring strength when subjected to long storage at a temperature of 00F., and "heat shocked." ‘ 5. Vanilla flavors are volatile only to an immeasurable degree, if at all, when subjected to 00F. temperature. (1) (2) (3) (4) (5) (6) (7) -52- BIBLIOGRAPHY. Chace, E.M. Bureau of Chemistry, U.S.D.A. Year Book 1908. The Twentieth Edition of the 0.8. Dispensitory. Joseph Burnett Company. Vanilla. Page 26. Lund, A.Adams. Manufacturing Confectioner. Volume 6, N0. 9 Page 38. Sept. 1926. Lund, A.Adams. Manufacturing Confectioner. Volume 6, No. 9 Page 39. Sept. 1926. Lund, A.Adams. Manufacturing Confectioner. Volume 6, No. 9 Page 35. Sept. 1926. Gnadinger, C.B. --- Chemist, McLaughlin, Gormley, King Co. Vanilla. (8) (9) (10) (11) (12) (13) (14) Lucas, P.S. Quality Flavors Facilitate Ice Cream Merchandising. Dairy Products Merchandising. oct. 1926. Circular 136. U.S. Dept. Agr. Standards of Purity for Food Products. Wilson, J.B. and Sale, J.W. Suitability of Various Solvents for Vanilla Beans. Industrial and Engineering Chemistry. Volume 17. Pages 506-7. May 1925. Problem in Vanilla Percolation. Spice Mill. Volume 44. Pages 1470-72, 1473. Tea and Coffee Trade Journal. Volume 122. July 1924. Chamberlain, Joseph, Scudder, Ph.D. organic Chemistry. (a text book) Gnadinger, C.B. --- Chief Chemist, McLaughlin, Gorm- ley, King Co. Vanilla. Pages 34, 35, 36. AI (15) (16) -54- Combs, W.B., Pierce, H.B., and Borst, W.F. Journal of Dairy Science. Volume 7, N0. 6. Nov. 1924. Page 585. Minutes of the Fifteenth Annual Meeting of the Extract Manufacturers Association of the United States, held in Boston, Mass. 1924. 31293 02670 . . . v .......