I/I E> R.AR.Y OF THE UNIVERSITY or ILLINOIS G30.7 *o.354-3G3 Cop. 2. W3R1CUITURE The person charging this material is responsible for its return to the library from which it was withdrawn on or before the Latest Date stamped below. Theft, mutilation, and underlining of books are reasons for discipli- nary action and may result in dismissal from the University To renew call Telephone Center, 333-8400 UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN APR 2 8 2 15 LI6I O-1096 A Nonacid Babcock Method for Determining Fat in Ice Cream By O. R. OVERMAN and O. F. GARRETT UNIVERSITY OF ILLINOIS AGRICULTURAL EXPERIMENT STATION BULLETIN 360 CONTENTS PAGE REVIEW OF LITERATURE 393 PLAN OF THE INVESTIGATION 395 RESULTS OF THE INVESTIGATION 399 PROCEDURE FOR NONACID METHOD 403 NONACID METHOD COMPARED WITH OFFICIAL METHOD 405 CONCLUSIONS AND OBSERVATIONS 405 LITERATURE CITED.. .406 Urbana, Illinois December, 1930 Publications in the Bulletin series report the results of investigations made or sponsored by the Experiment Station A Nonacid Babcock Method for Determining Fat in Ice Cream By O. R. OVERMAN, Assistant Chief in Dairy Chemistry, and O. F. GARRETT, First Assistant A -\ T ACCURATE and rapid method for the determination of fat /\\ / ^\in ice cream is necessary in order quickly to standardize and j( )\ to control ice-cream mixes. Most ice-cream plants have the regular Babcock equipment for testing milk and cream. An accurate and inexpensive method which will utilize the regular Babcock testing apparatus will be invaluable in the ice-cream industry. This bulletin describes an attempt to devise a method for the deter- mination of fat in ice cream in which the Babcock apparatus can be used and which it is believed will prove more satisfactory than the rapid methods now in use. Review of Literature Many procedures have been developed for the determination of fat in ice cream. Some of these have used Babcock apparatus. Most of the methods have depended on the action of an acid or a mixture of acids for the liberation of the fat. The most serious objection to the use of acid is that it reacts with the cane sugar to form a fluffy, charred material which rises into the neck of the test bottle and interferes with the reading of the fat column. This is especially true of sulfuric acid. Attempts have been made to use acids other than sulfuric, but the results obtained have not been very satisfactory. Probably the most successful of the acid methods have been those in which sulfuric acid was used in mixture with some other acid or acids. Certain alcohols, also, have been used with the acids. Some of these methods have given fair results with plain ice cream but have been unsatisfactory when used on fruit and chocolate ice creams. There are a number of modifications of the Babcock method for the determination of fat in ice cream. One of the early methods used a mixture of equal parts of hydrochloric and acetic acids. White 25 * reported a method in which he used sulfuric acid after diluting the ice cream with water to three times its volume. The methods developed by Grigsby, Brinsmaid (both reported by Hortvet 7 *), Utt, 23 * Ross, 18 * Lichtenberg, 12 * and Halverson 5 * used mixtures of sulfuric and acetic 393 394 BULLETIN No. 360 [December, acids. Halverson's method required a special bottle \vith a stopcock at the bottom for drawing off the sugar solution. Francis and Morgan 3 * modified the sulf uric-acetic acid method by adding a few drops of nitric acid. This method seems to give very good results with plain and vanilla ice creams. Le Cointe 10 * reported a method in which a small amount of amyl alcohol was added to the mixture of ice cream and sulf uric acid. In another method, 2 * used rather extensively, ethyl alcohol was mixed with the ice cream previous to the addition of sulfuric acid. It is doubtful if any of these methods have proved wholly successful on all kinds of ice cream. Reagents other than acid have been used for the rapid determina- tion of fat in milk, cream, and other dairy products. In 1888 Short 19 * reported a method in which two solutions were used. No. 1 was com- posed of 250 grams of caustic soda, 300 grams of caustic potash, and 1,809 grams of w r ater. No. 2 was a mixture of equal parts of commer- cial sulfuric and acetic acids. Thorner, 22 * in 1892, reported a method for the determination of fat in dairy products in which either an aqueous or an alcoholic potash solution could be used. Treatment of the milk with this solution was followed by treatment with acetic acid. In the Nahm 15 * milk test, reported in 1894, milk was boiled with a mixture composed of potassium hydroxid, ethyl alcohol, amyl al- cohol, and 30-percent ammonia solution, and the fat was read off on a scale. In the lactanalyt 6 * test, milk was warmed with an alkaline solu- tion, alcohol, and ether. Leze, 11 * in 1899, reported a method in which a mixture of caustic potash, ammonia, ethyl alcohol, and amyl alcohol were used. The Sichler 20 - 26 * sinacid method, first reported in 1903, which is patented and has been modified many times, has been used for testing milk and cream. Trisodium phosphate and sodium citrate were first used in this method, but these chemicals were replaced with sodium hydroxid, sodium sulf ate, and sodium potassium tartrate (Rochelle salt). Isobutyl alcohol was used. Gerber 4 ' 21 * developed and reported a method in 1906 known as the Gerber "Sal" method. Milk was mixed in a butyrometer with a solu- tion consisting of sodium chlorid, sodium hydroxid, sodium potassium tartrate, and a small amount of coloring matter. Isobutyl alcohol was added. The mixture, after being heated in a water bath, was centri- fuged and the fat read. Wendler, 24 * in 1910, patented a nonacid method which was known 1930] A NONACID BABCOCK METHOD FOR DETERMINING FAT IN ICE CREAM 395 as the "Neusal" method. A mixture of neutral salts consisting of sodium citrate and sodium salicylate with a small amount of methylene blue was used. Isobutyl alcohol was added separately. In 1915 Pinkhoff 17 * reported a method for the determination of fat in milk without the use of sulfuric acid. Milk was warmed with amyl alcohol and a solution composed of sodium hydroxid, sodium phosphate, sodium citrate, and sodium chlorid. The butyrometers were centrifuged and the fat read. Miwa 14 * patented a method in 1918 in which a mixture of iso- butyl alcohol, a benzoate, a tartrate, a citrate, ethyl alcohol, a dye soluble in fats, and a water soluble dye was used. Hoyberg 8 ' 9 * first reported his method in 1920, but it has been modi- fied several times since. In this method milk was heated in a butyrome- ter with isobutyl alcohol and a solution of sodium hydroxid and sodium potassium tartrate. No centrifuge was used. In 1920 Oshida 16 * reported a method in which the reagent consisted of 3.5 cc. ethyl alcohol, 5.6 cc. water, 3.5 cc. amyl alcohol, 1.0 cc. glycerol, .005 gram methylene blue, and .0025 gram sudan III. Maghano and Porzio, 13 * in 1927, modified Hoyberg's method by using Fehling's alkali solution (60 grams sodium hydroxid, 173 grams sodium potassium tartrate, made up to 500 cc. with water) and a mix- ture composed of 45 parts methyl alcohol and 55 parts isobutyl alcohol. In 1914 Bradbury 1 * adapted the Short method to the determina- tion of fat in ice cream and condensed milk. The Wendler method (reported by Hortvet 7 *) was tried on ice cream but did not seem to give consistent results. According to the directions for this method a 9-gram t sample in a Babcock milk-test bottle was warmed with 19 cc. of a solution composed of 125 grams sodium hydroxid, 25 grams sodium potassium tartrate, and 25 grams sodium chlorid made up to 500 cc. with water. Ten to twelve drops of isobutyl alcohol were added to the mixture of ice cream and reagent. Plan of the Investigation In the present investigation a study of the acid methods was not attempted. A search was made for some nonacid reagent or reagents which could be used to replace acids. Alkali and salt solutions of vary- ing concentrations and combinations were studied. The effects of various organic solvents with these solutions were also studied. The general procedure followed in studying the different solutions was as follows: A solution of the desired composition was prepared. 3% BULLETIN No. 360 [December, TABLE 1. COMPOSITION OF SOLUTIONS GROUPED ACCORDING TO EFFICIENCY IN LIBERATING THE FAT FROM ICE-CREAM Mix Sol. MiBC. TriMd. phos. Sod. ammon. phos. Sod. pot. tartrate Sod. citrate Sod. acetate Sod. salicyl. Water Ammon. hydrox. Group 1 Solutions which failed to liberate the fat 1 gramt 90NaOH grant grant grant gramt gramt 150 gramt ee. Made up ee. 2 3 80 NaOH 200 200 to 1000 4 150 " 5 100 ' 6 300 NatHPCh " 7 100 200 1 000 8 100 200 700 300 9 200 200 1 000 10 200 200 1 000 11 100 100 200 1 000 12 100 100 200 1 000 13 100 100 200 1 000 14 100 100 200 700 300 15 100 200 100 1 000 16 100 200 100 700 300 17 100 200 100 1 000 18 100 200 100 700 300 19 200 200 1 000 20 200 200 700 300 21 200 200 700 300 22 200 200 700 300 23 400 1 000 24 400 700 300 25 400 700 300 Group 2 Solutions which liberated an appreciable amount of fat 26 27 28 90NaOH /90 NaOH \20 NaS04 80 NaOH 150 150 140 Made up to 1000 29 30 70 NaOH 60 NaOH 130 120 31 50 NaOH 110 32 33 34 35 36 37 80 NaOH 70 NaOH 60 NaOH SONaOH 100 NaOH / 90 NaOH 120 100 80 70 170 38 \150Borax 90 NaOH 150 ,. 39 1 100 NaOH ii 40 \170 Borax 1 110 NaOH ,. 41 42 43 \200 Borax .2 Methylene blue .2 Methylene blue 300 500 80 100 80 100 " 44 400 200 45 200 200 1 000 46 200 200 i 666 47 200 200 700 300 48 200 200 1 000 49 200 200 700 300 50 100 200 100 1 000 51 100 200 100 700 300 52 200 100 100 700 300 53 200 200 1 000 54 200 100 100 1 000 55 200 100 100 700 300 56 150 150 1 000 67 100 200 100 1 000 58 100 200 100 700 300 59 100 100 200 1 000 1930] A NONACID BABCOCK METHOD FOR DETERMINING FAT IN ICE CREAM 397 TABLE 1. Continued Sol. Misc. Trbod. phoe. Sod. ammon. phoa. Sod. pot. tartrate Sod. citrate Sod. acetate Sod. salicyl. Water Ammon. hydros. 60 grams gram 100 grams 100 grami grant grami grams 200 cc. 700 cc. 300 61 200 200 700 300 62 200" 200 700 300 63 100 100 200 1 000 64 100 10Q 200 700 300 65 100 100 200 700 300 66 100 100 200 700 300 67 100 100 200 1 000 68 100 100 200 700 300 69 100 200 100 1 000 70 100 100 200 1 000 71 100 100 200 700 300 - 72 JOO 200 100 700 300 73 100 200 100 1 000 74 100 200 100 700 300 75 100 200 100 1 000 76 100 i 200 100 700 300 77 Variable Variable 1 000 78 200 200 1 000 70 200 200 700 300 80 200 200 1 000 81 200 200 700 300 82 175 175 000 83 150 150 000 84 100 100 000 85 200 200 700 300 86 175 175 000 87 150 150 000 88 100 100 000 89 200 200 000 90 400 000 91 400 700 300 92 400 1 000 93 400 700 300 94 400 1 000 95 400 700 300 96 400 1 000 97 Variable Variable Variable Variable Variable Group 3 Solutions which appeared to liberate all the fat 98 '200 100 Made up 99 100 200 tol ( 000 100 150 50 150 < 101 .- 200 100 700 300 102 ' 300 200 Made up 103 200 200 tolOOO 1 000 104 200 200 700 300 105 100 200 100 1 000 106 .... ' 100 200 100 700 300 107 200 100 100 1 000 108 100 100 100 100 1 000 109 100 100 100 100 700 300 110 200 200 700 300 111' Variable Variable Variable Variable 112 200 100 100 1 000 113 200 100 100 700 300 114 200 100 100 1 000 115 200 100 100 700 300 116 200 100 1 000 117 200 100 700 300 118 175 175 1 000 119 100 200 100 1 000 120 100 200 100 700 300 121 100 100 200 1 000 122 100 100 200 700 300 123 100 100 200 1 000 124 100 100 200 700 300 125 100 200 100 1 000 126 100 200 100 700 300 127 100 200 100 700 300 128 100 200 100 1 000 398 BULLETIN No. 360 [December, TABLE 1. Concluded Sol. Misc. Trisod. pboe. Sod. ammon. phos. Sod. pot. tartrate Sod. citrate Sod. acetate Sod. sal icyl. Water Ammon. hydro*. 129 grams grams 100 grams grams 200 grams grams grams 100 et. 1 000 ct. 130 100 200 100 700 300 771 Variable Variable 1 000 131 200 200 700 300 132 200 200 1 000 133 200 200 1 000 134 400 700 300 135 400 1 000 136 150 100 100 1 000 97 Variable \Y.ri:iMr Variable Variable V;iri:i't.;( 'There are 19 aqueous solutions of mixtures of trisodium phosphate and sodium acetate, the variation in the araoun t of each salt being from 50 to 200 grams per liter of water. Five of the solutions belong in Group 2, and 14 in Group 3. There are 61 aqueous and aqueous-ammoniacal solutions of a mixture of trisodium phosphate, sodium acetate, and sodium salicylate. The phosphate varies from 50 to 350 grams, the acetate from 25 to 200 grams, and the salicylate from 10 to 100 grams per liter of solvent. The amount of ammonium hydroxid in the aqueous-ammoniacal solutions varies from 25 to 300 cc. per liter of solvent. Eighteen of these solutions belong in Group 2, and 43 in Group 3. There are 46 aqueous and aqueous-ammoniacal solutions of mixtures of trisodium phosphate and sodium salicylate. The phosphate varies from 150 to 400 grams and the salicylate from 50 to 200 grams per liter of solvent. The amount of ammonium hydroxid in the aqueous-ammoniacal solutions varies from 25 to 300 cc. per liter of solvent. Nine grams of the prepared sample of ice cream were weighed into a Bab- cock milk-test bottle. The solution and the organic solvent were added, the latter being measured exactly. The mixture was placed in a shallow water bath heated with a jet of steam, and was shaken occasionally while heating. The bottle was centrifuged and the fat column read as in the regular Bab- cock procedure for milk. The Roese-Gottlieb method as given by the Association of Official Agricultural Chemists was used for comparison. Three extractions with ether were made in each case. Duplicate determinations were made on each sample. Eleven chemical compounds sodium hydroxid, sodium sulfate, borax, disodium phosphate, trisodium phosphate, sodium ammonium phosphate, sodium potassium tartrate, sodium citrate, sodium acetate, sodium salicylate, and ammonium hydroxid were used in making up 261 solutions. Table 1 shows the number and composition of the solu- tions studied in this investigation. Effects that ten different organic compounds had on the separation of the fat were studied. The compounds were n-butyl alcohol, isobutyl alcohol, acetone, toluol, chloroform, carbon tetrachlorid, amyl acetate, glycerol, amyl alcohol, and ethyl alcohol. In addition, seventeen various mixtures of acetone with amyl alcohol, n-butyl alcohol, or isobutyl alcohol were tried. Unsatisfactory results were obtained with acetone, toluol, chloroform, carbon tetrachlorid, amyl acetate, glycerol, and the acetone-alcohol mixtures. The butyl alcohols and amyl alcohol gave the greatest promise of producing good results. 1930] A NONACID BABCOCK METHOD FOR DETERMINING FAT IN ICE CREAM 399 Results of the Investigation The efficiency of the salt solutions in liberating the fat from the other components of the ice cream is indicated by the groupings in Table 1. Group 1 includes solutions which failed to liberate any ap- preciable amount of fat. In most cases the ice cream was curdled. Solutions 1 to 25 inclusive fall in this group. Group 2 includes solu- tions which liberated a considerable portion of the fat but failed to TABLE 2. FAT CONTENT OF ICE-CREAM MIXES AS DETERMINED BY USE OF SOLU- TIONS NUMBERED 77 CONTAINING BOTH TRISODIUM PHOSPHATE AND SODIUM ACETATE IN VARYING CONCENTRATION (Fat in sample 13.61 percent determined by the Official method; .8 cc. n-butyl alcohol was used in each test) Solution Composition of solution 1 Tests Description Trisodium phosphate Sodium acetate Orig. Dup. Aver. grams grams perct. perct. perct. 77a 200 200 13. 8 13.8 13.8 Clear fat column 77b 200 150 13.6 Il6 13.6 Clear fat column 77c 200 100 13.6 13.6 13.6 Very slightly curdy 77d 200 50 13 2 13.4 13.3 Slightly curdy 77e 175 100 13 6 13.6 13.6 Very slightly curdy 77f 175 50 13.6 13.8 13.7 Slightly curdy t 150 150 100 50 13.6 !3.6 13.6 13.6 13.6 13.6 Very slightly curdy Slightly curdy 77i 150 200 13.2 13.6 13.4 Slightly curdy 775 100 200 11.6 9.8 Curdy 77k 50 200 8.0 8.0 Curdy 771 100 175 13.2 13.2 13.2 Slightly curdy 77m 50 175 11.2 11.8 Curdy 77n 100 150 13.2 13.4 13 '.3 Slightly curdy 77o 50 150 12.8 12.8 12.8 Curdy 77p 150 200 13.6 13.6 13.6 Very slightly curdy 77q 175 175 13.6 13.6 13.6 Clear fat column 77r 100 100 13.0 13.0 13.0 Slightly curdy 'One liter of distilled water was used in making each solution. produce complete liberation. In most cases a considerable amount of undissolved curd appeared at the bottom of the fat column. Solutions 26 to 97 inclusive are in this group. Group 3 includes all solutions which appeared to liberate the fat completely. Very little undissolved curd appeared and the fat columns were, in most cases, clear and sharply divided from the aqueous layer. Solutions 98 to 136 inclusive fall in this group. This group also includes some of Solutions 77 and 97. In Group 3 Solutions 77 (trisodium phosphate and sodium acetate), 97 (trisodium phosphate, sodium acetate, and sodium salicylate), and 111 (trisodium phosphate and sodium salicylate) seemed most promis- ing. They were given a thoro trial on various kinds of ice creams and ice-cream mixes. 400 BULLETIN No. 360 [December, D 1 H J x a j (J 4> i^ H SB it CQ O c CO Ed ^ .o u a: J3 - o I t S^^i nooooooooooQOOoooooor- 5 o o * ^- * cc O *"*> a o OS - O O O +++++++++++++++++ 1 +++ < o> o c-i o t~ oo at o e m >o o *> o * * * * 4>999V8aw9 1930] A NONACID BABCOCK METHOD FOR DETERMINING FAT IN ICE CREAM 401 Table 2 shows the composition of the various solutions of trisodium phosphate and sodium acetate and the results obtained on a sample of unfrozen plain mix. The solution containing 200 grams trisodium phosphate and 150 grams sodium acetate (77b) was chosen as the best of the lot because it seemed as efficient in liberating the fat as the more highly concentrated solutions and it is more economical be- cause of the smaller amount of the more expensive salt sodium acetate. Ammonium hydroxid was not added directly to the salt solution since it causes a large part of the salts to crystallize. However, am- monia seems to be necessary for the test. A second solution was made of 75 parts ammonium hydroxid, 35 parts n-butyl alcohol, and 15 parts ethyl alcohol by volume. Ammonium hydroxid and n-butyl alcohol are not miscible in these proportions, but on the addition of the ethyl alcohol a homogeneous solution is formed. A 2.5 cc. portion of this solution was used in each test. A comparison between Solution 111 (200 grams trisodium phos- phate and 200 grams sodium salicylate) with and without ammonia and Solution 97 (150 grams trisodium phosphate, 100 grams sodium acetate, and 25 grams sodium salicylate) with and without ammonia is shown in Table 3. In this comparison .7 cc. n-butyl alcohol was used in each determination. A comparison between Solution 111 with ammonia and Solution 77b (200 grams trisodium phosphate and 150 grams sodium acetate) is shown in Table 4. In this comparison .7 cc. of n-butyl alcohol was used with Solution 111, while 2.5 cc. of the alcohol-ammonia mixture were used with Solution 77b. Solution 77b was finally chosen as the best because it seems to give results most nearly approaching the correct values, does not change on long standing, remains in solution at ordinary temperatures, and is less expensive in that sodium acetate is a cheaper chemical than sodium salicylate. The effect of the length of time of heating in the water bath before centrifuging is shown in Table 5. Fifteen minutes seems sufficient for most of the plain ice creams, but a longer time is usually required for the fruit and especially the chocolate ice creams. Denatured grain alcohol may be used instead of 95-percent ethyl alcohol in making up the alcohol-ammonia mixture. Ethyl alcohol was denatured according to formulas 1, 2, 3, 4, 5, and 7 given in "Dena- 402 BULLETIN No. 360 [December. TABLE 4. COMPARISON OF FAT PERCENTAGES OBTAINED WHEN TESTING ICE- CREAM SAMPLES WITH SOLUTIONS 111 AND 77b (.7 cc. n-butyl alcohol was used with Solution 111 and 2.5 cc. of the alcohol-ammonia mixture were used with Solution 77b) Sample Official fat Solution 111 Solution 77b Orig. Dup. Aver. Dif. Orig. Dup. Aver. Dif. perct. perct. perct. perct. perct. perct. perct. 64 8.94 8.6 9.0 8.80 - .14 9.0 9.0 9.00 + .06 65 9.96 9.8 10.2 10.00 + .04 10.0 10.0 10.00 + .04 66 10.84 11.0 11.0 11.00 + .16 10.9 10.9 10.90 + .06 67 11.92 12.3 12.0 12.15 + .23 12.1 12.0 12.05 + .13 68 13.25 13.8 13 4 13.60 + .35 13.4 13.2 13.30 + .05 69 15.13 15.2 15.4 15.30 + .17 15.2 15.2 15.20 + .07 70 16.27 16.4 15.6 16.00 - .27 16.4 16.4 16.40 + .13 71 18.01 18.0 18.0 18.00 - .01 18.1 18.1 18.10 + .09 72 19.02 19.2 19.2 19.20 + .18 19.1 19.1 19.10 + .08 73 23.44 23.6 23.6 23.60 + .16 23.8 23.6 23.70 + .26 74 11.20 10.4 10.4 10.40 - .80 11.0 11.0 11.00 - .20 75 11.83 11.8 12.0 11.90 + .07 11.8 11.8 11.80 - .03 76 12.34 12.5 12.6 12.55 + .21 12.5 12.6 12.55 + .21 77 12.41 12.4 12.5 12.45 + .04 12.6 12.6 12.60 + .19 78 10.21 10.4 10.4 10.40 + .19 10.4 10.5 10.45 + .24 79 10.67 10.6 10.7 10.65 - .02 10.7 10.7 10.70 + .03 80 11 10 11.0 11.2 11 10 11 2 11.2 11.20 + .10 82 2.32 .8 .9 .85 -1.47 2.3 2.3 2.30 - .02 83 13.83 13.6 13.7 13.65 - .18 13.8 13.8 13.80 - .03 84 11.08 10.4 10.6 10.50 - .58 11.1 11.1 11.10 + .02 85 10.29 10.0 10.0 10.00 - .29 10.4 10.3 10.35 + .06 86 9.09 9.1 9.0 9.05 - .04 9.1 9.2 9.15 + .06 87 10.70 10.6 10.5 10.55 - .15 10.8 10.8 10.80 + .10 88 11.84 11.4 11.4 11.40 - .44 11.8 11.9 11.85 -1- .01 89 10.28 10.4 10.5 10.45 + .17 10.4 10.4 10.40 + .12 90 14.08 14.0 13.9 13.95 - .13 14.2 14.2 14.20 + .12 Average.. 12.3096 12.2115 - .0981 12.3846 -1- .0750 TABLE 5. RESULTS OBTAINED ON SAMPLE OF ICE CREAM BY NONACID METHOD WHEN TIME OF HEATING VARIED FROM 10 TO 90 MINUTES (Fat in sample 14.31 percent determined by the Official method) Heating time Original test Duplicate Average Difference minutei perct. perct. perct. 10 14.0 14.0 14.00 - .31 15 14.4 14.4 14.40 + .09 20 14.4 14.4 14.40 + .09 25 14.4 14.3 14.35 + .04 30 14.4 14.4 14.40 -1- .09 40 14.4 14.4 14.40 + .09 50 14.4 14.4 14.40 + .09 60 14.5 14.6 14.55 + .24 70 14.6 14.6 14.60 + .29 80 14.6 14.6 14.60 + .29 90 14.6 14.4 14.50 + .19 turization of Industrial Alcohol," Senate Document No. 195, 69th Congress, Second Session, Appendix to Regulations No. 61, Com- pletely Denatured Alcohol Formulae. Alcohol-ammonia mixtures were made up with the various de- natured alcohols. The results obtained on an ice-cream mix are shown in Table 6. The denaturants seem to have no harmful effect on the 1930] A NONACID BABCOCK METHOD FOR DETERMINING FAT IN ICE CREAM 403 TABLE 6. RESULTS OBTAINED ON SAMPLE OF ICE CREAM BY NONACID METHOD WHEN DENATURED GRAIN ALCOHOL WAS USED IN THE ALCOHOL- AMMONIA MIXTURE (Fat in sample 13.58 percent determined by Official method) Denaturing formula No. Original test Duplicate Average 1... perct. 13 6 perct. 13 6 perct. 13 60 2 13 6 13 7 13 65 3 13 6 13 8 13 70 4 13 6 13 6 13 60 5 13 7 13 9 13 80 7. . 13 6 13 7 13 65 95% alcohol 13.6 13.7 13.65 method. The denatured alcohol found on the retail market is usually prepared by one of these formulas and should prove quite satisfactory for use in this test. Procedure for Nonacid Method The method as finally adopted requires two reagents. Reagent A, the alcohol-ammonia mixture, is made up as follows: 75 cc. of C.P. ammonium hydroxid 35 cc. of n-butyl alcohol 15 cc. of 95-percent ethyl alcohol or denatured grain alcohol This reagent should be kept in a glass-stoppered bottle. Reagent B is made up as follows: 200 grams of trisodium phosphate 150 grams of sodium acetate 1 liter of water ' Trisodium phosphate and sodium acetate of commercial grade are satisfactory. Clean tap water may be used instead of distilled water. The reagent remains in solution at ordinary temperatures but will partially crystallize on becoming cold. If crystallization takes place the crystals should be dissolved by warming before the reagent is used. The reagent does not seem to deteriorate on long standing. In addition to the regular Babcock equipment a shallow water bath provided with some means for heating and a burette graduated to .5 cc. are necessary. A 2.5-cc. pipette may be used instead of the burette. The heat may be supplied by a gas burner or a jet of steam. As a result of this investigation, the following procedure is pre- sented for the determination of fat in ice cream: 1. Weigh 9 grams of the well-mixed sample of ice cream into a Babcock milk-test bottle (8- or 10-percent) or a 20-percent ice-cream test bottle. 2. Add exactly 2.5 cc. of Reagent A from a burette or pipette. Mix thoroly. 3. Add 9 to 10 cc. of Reagent B and again mix thoroly. The reagent may be measured in an ordinary 9-cc. acid measure or in a 10-cc. pipette. 4. Place the test bottle in a shallow water bath and heat the bath to boiling, 404 BULLETIN No. 360 [December, TABLE 7. PERCENTAGE OF FAT IN 54 SAMPLES OF ICE CREAM AS DETERMINED BY OFFICIAL METHOD AND BY NONACID BABCOCK METHOD Simple Kind of ice cream Official method Nonacid method Original Duplicate Average Difference Samples from University of Illinois Creamery 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 1 2 3 4 5 6 7 8 9 13 14 15 16 23 perct. 8.940 9.956 10.840 11.920 13.245 15.125 16.270 18.010 19.020 23.440 11.205 11.830 12.335 12.405 10.205 10.665 11.100 14.305 2.320 13.830 11.080 10.290 9.090 10.700 11.840 10.280 14.075 13.300 13.525 13.935 13.815 11.095 13.580 13.860 12.415 10.935 11.645 12.065 11 780 12.190 12.910 11.440 11.270 8.285 13 580 perct. 9.0 10.0 10.9 12.1 13.4 15.2 16.4 18.1 19.1 23.8 11.0 11.8 12.5 12.6 10.4 10.7 11.2 14.4 2.3 13.8 11.1 10.4 9.1 10.8 11.8 10.4 14.2 13.2 13.6 14.1 13.8 11.2 13.8 14.1 12.3 11.0 11.8 12.0 11.6 12.2 13.0 11.4 11.2 8.4 13.6 perct. 9.0 10.0 10.9 12.0 13.2 15.2 16.4 18.1 19.1 23.6 11.0 11.8 12.6 12.6 10.5 10.7 11.2 14.4 2.3 13.8 11.1 10.3 9.2 10.8 11.9 10.4 14.2 13.3 13.5 14.1 13.9 11.2 13.9 14.0 12.3 11.0 11.7 12.0 11.6 12.2 13.0 11.4 11.3 85 13.7 perct. 9.00 10.00 10.90 12.05 13.30 15.20 16.40 18.10 19.10 23.70 11.00 11.80 12.55 12.60 10.45 10.70 11.20 14.40 2.30 13.80 11.10 10.35 9.15 10.80 11.85 10.40 14.20 13.25 13.55 14.10 13.85 11.20 13.85 14.05 12.30 11.00 11.75 12.00 11.60 12.20 13.00 11.40 11.25 845 13.65 (+) (-) .060 ... .044 ... .060 ... .130 ... .055 ... .075 ... .130 ... .090 ... .080 ... .260 ... .... 205 .... 030 .215 ... .195 ... .245 ... .035 ... .100 ... .095 ... .... 020 .... 030 .020 ... .060 ... .060 ... .100 ... .010 ... .120 ... .125 ... .... 050 .025 ... .165 ... .035 ... .105 ... .270 ... .190 ... .... 115 .065 ... .105 ... .... 065 .... 180 .010 ... .090 ... .... 040 .... 020 .165 ... .070 ... Plain mix Plain mix Plain mix Chocolate Vanilla Caramel Pistachio nut Fruit Fruit Fruit Plain mix Vanilla Chocolate Chocolate Fruit Vanilla Fruit Vanilla Vanilla Vanilla Vanilla Vanilla Fruit Vanilla Mint Chocolate Fruit Vanilla Fruit Chocolate Fruit Fruit Chocolate Chocolate Fruit . . Vanilla Samples from commercial ice-cream plants 1 10 11 12 17 18 19 20 21 22 Vanilla 11.435 11.085 10.060 10.770 9.705 9.150 14.935 14 225 12.695 11.4 11.3 10.2 10.8 9.8 9.2 15.0 14.3 12.7 11.4 11.3 10.3 10.8 9.8 9.2 15.0 14.3 12.8 11.40 11.30 10.25 10.80 9.80 9.20 15.00 14.30 12.75 035 .215 .190 .030 .095 .050 .065 .075 .055 Chocolate Fruit Vanilla Fruit Chocolate Vanilla Fruit Chocolate Summary of results from all samples 1 Averag Averag samp Averag Averag e of 54 samples 12.2223 13.7267 10.4493 11.0455 .... 12.2898 13.8000 10.5367 11.0364 .0675 .... .0733 .. .0873 .... 0091 e of 25 plain mix and vanilla iles e of 15 fruit samples e of 11 chocolate samples. . 'The authors wish to express their appreciation to the following named companies for the courtesy shown by them in furnishing samples: Snow and Palmer, Bloomington; John T. Cunningham Ice Cream Co., Chicago; and J. D. Rosiell Co., Peoria. Many samples used in the development of the method were obtained from the Champaign Ice Cream Co., Champaign, Illinois. 'Samples 76, 77, and 3 are not classified. 1930] A NONACID BABCOCK METHOD FOR DETERMINING FAT IN ICE CREAM 405 continuing the heating for several minutes. Shake the contents of the bottle two or three times while heating. 5. Usually at the end of 15 to 30 minutes the fat will separate and form a clear yellow layer on top of the liquid. The heating must not cease until the fat layer has definitely separated from the dark portion of the liquid and has become clear. 6. After all the fat has separated, place the test bottle in the centrifuge and whirl 5-2-1 minutes, adding hot water as in the regular Babcock milk test except that the water must not be softened with acid. 7. Place the bottle in a hot-water bath at 130-140 F. for 5 minutes. 8. Read the test, measuring from the bottom of the fat column to the top of the upper meniscus. 9. If an 8- or 10-percent milk-test bottle is used, multiply the reading by two. Nonacid Method Compared With Official Method Determinations of fat by the method just described were made on fifty- four samples of ice cream and the results compared with those obtained by using the Official Roese-Gottlieb method (Table 7). De- termined by the Official method, the fat in these samples varied from 2.32 percent in the case of a sherbet to 23.44 percent in the case of a high- fat ice crearo. The solids-not-fat varied from approximately 8 to 14 percent. The samples included plain, unfrozen mixes and ice creams containing whole fruit, fruit sirups, various flavorings and extracts, caramel, nuts, mint candy, and chocolate. The whole fruits, nuts, and other solid particles were removed from the samples before testing by passing the liquid portion thru cheesecloth. Of the results on fifty-four samples tested by the nonacid method, 64.82 percent came within .1 percent of the Roese-Gottlieb values, 24.07 percent fell between .1 and .2 percent, and 11.11 percent between .2 and .3 percent ; that is, 88.89 percent of the tests showed a difference of less than .2 percent from the Roese-Gottlieb values while only 11.11 percent showed differences greater than .2 percent. The range of differences was from .205 to + -270 percent. The average difference was -f- .0675 percent for all samples. Conclusions and Observations 1. Butter fat can be separated quantitatively from ice cream with- out the use of sulfuric acid. 2. Babcock test bottles and other Babcock apparatus can be used in a volumetric method for determining fat by nonacid reagents. 3. The method presented is fairly rapid, sufficiently accurate for commercial use, and is inexpensive. 4. The operator must become completely familiar with the method in order to achieve success with it. 5. A large number of determinations may be made concurrently. 406 BULLETIN No. 360 Literature Cited 1. BRADBURY, C. M. The alkali method for the determination of fat in ice cream and condensed milk. Va. Dairy and Food Circ. 42. 1914. Thru Chem. Abs. 10, 941. 1916. 2. FISHER, R. C., and WALTS, C. C. A comparative study of methods for deter- mining the percent of butterfat in dairy products. Jour. Dairy Sci. 8, 54. 1925 3. FRANCIS, C. K., and MORGAN, D. G. The determination of fat in certain milk products. Okla. Agr. Exp. Sta. Bui. 114. 1917. Thru Expt. Sta. Rec. 37, 507. 1917. 4. GERBER, N. New rapid method for the determination of fat in milk with- out the use of acid. Rev. Gen. Lait 5, (14), 318. 1906. Thru Expt. Sta. Rec. 17, 1134, 1907. 5. HALVERSON, J. O. The modified Babcock for fat in sweetened dairy products ice cream. Indus, and Engin. Chem. 5, 403. 1913. 6. HILLS, J. L. The lactanalyt, a new milk tester. Vt. Agr. Exp. Sta. Rpt. 161. 1894. Thru Expt. Sta. Rec. 8, 89. 18%. 7. HORTVET, J. Report on dairy products. Jour. Assoc. Off. Agr. Chem. 2, 238. 1917. 8. HOYBERG, H. M. U. S. Patent 1,329,183, Jan. 27, 1910. Chem. Abs. 14, 991. 1920. 9. - - U. S. Patent 1,521,953, Jan. 6, 1925. Chem. Abs. 19, 685. 1925. 10. LE COINTE, P. P. A rapid method for the determination of fat in ice cream. Canad. Chem. and Metall. 7, 208. 1923. Thru Chem. Abs. 17, 3723. 1923. 11. LEZE, R. Determination of fat in milk. Ann. Chim. Analyt. 4, 371. 1899. Abs. in Chem. Centbl. 1, (1), 69. 1900. Thru Expt. Sta. Rec. 11, 904. 1899. 12. LICHTENBERG, H. F. A rapid method for the determination of fat in ice cream. Indus, and Engin. Chem. 5, 786. 1913. 13. MAGHANO, A., and PORZIO, J. Practical method for the determination of fat in milk. Lait 8, 714. 1927. Thru Chem. Abs. 22, 284. 1928. 14. MIWA, Z. Reagent for the determination of fat in milk. Japanese Patent 32,726, May 24, 1918. Chem. Abs. 12, 2220. 1918. 15. NAHM, A. N. The Nahm milk test. Milch Ztg. 23, (35), 555. 1894. Thru Expt. Sta. Rec. 6, 474. 1894. 16. OSHIDA, T. Reagent for the determination of fat in milk. Japanese Patent 36,282, April 28, 1920. Chem. Abs. 15, 1583. 1921. 17. PINKHOFF, J. The determination of fat in milk without the use of sulfuric acid. Pharm. Weekbl. 52, 754. 1915. Thru. Chem. Abs. 10, 1559. 1916. 18. Ross, H. E. Testing ice cream for fat. Ice Cream Trade Jour. 7, 56. 1911. 19. SHORT, F. G. A new method for determining fat in milk. Wis. Agr. Exp. Sta. Bui. 16. 1888. 20. SICHLER, A. German Patent 184,639, Dec. 11, 1903. Chem. Abs. 1, 2435. 1907. 21. SIEGFELD, M. The salt method. Molk. Ztg. 20, (14), 371. 1906. Thru Expt. Sta. Rec. 17, 1134. 1907. 22. THORNER, W. Determination of fat in milk and dairy products. Chem. Ztg. 16, 1101. 1892. Thru Expt. Sta. Rec. 4, 289. 1892. 23. UTT, C. A. A. The determination of fat in ice cream by the Babcock method. Indus, and Engin. Chem. 7, 773. 1915. 24. WENDLER, O. A new sal method for estimating fat in milk. Milch Ztg. 39, (20), 230. 1910. Thru Expt. Sta. Rec. 23, 614. 1910. 25. WHITE, J. W. A method for the determination of fat in ice cream. Penn. Sta. Rpt. 243. 1910. 26. WINDISCH, K. The sinacid and salver processes for fat determination in milk. Milchw. Zentbl. 5, 344. 1909. Thru Chem. Abs. 3, 2987. 1909. UNIVERSITY OF ILLINOIS-URBAN*