INDUSTRIAL AND ENGINEERING
106
,
PREPARATION OF LEADSALTS T o clarify the situation, lead salts were prepared by the following procedures: METHOD A. For this preparation 3 grams of ~ s o l i acid c were dissolved in 400 CC. Of hot 95 pel’ Cent alcohol. TO this Solution were added 2.5 grams of crystalline lead acetate dissolved in 30 cc. of hot water. The hot mixture was set aside to cool gadually and finally placed in a refrigerator. The lead ursolate separated in a crystalline condition and was filtered out on a Buchner funnel. It was washed with water and alcohol, again with water and alcohol, and finally with ether. It was then air-dried overpight before being desiccated in an Abderhalden drier at 111’ C. 0.2426 gram substance
Found P b Pb(CsdH1108)) requires, P b
0 0650 gram PbSO4
18.3
18.5
METHOD 33. In this preparation the method as described by Cohee and St. John was literally followed, except that half quantities were used. 0.2032 gram substance
Found, P b
0.0470 gram 15.8
Method A differs from method B in that a molecular quantity of lead acetate was used and that the product was
CHEMISTRY
Vol. 7, No. 2
allowed to separate under conditions favoring crystallization. the product was subsequently treated to eliminate both unreacted lead acetate and ursolic acid. The analytical results confirmed the authors’ previous finding. The analytical result for method B, a repetition of the experiment by Cohee and St. John, indicated an impure but did not duplicate their finding. Since method A gave an analytically pure product, whereas method B did not, the authors interpret these observations as a confirmation of their criticism.
LITERATURE CITED (1) Cohee and St. John, IXD. ENG.CHIM.,26, 781 (1934). (2) Dodge, J. Am. Chem. Soc., 40,1932-4 (1918). (3) Hutchinson and Pollard, J. Chem. Soc., 69,213 (1896). (4) Kamm, o., “Qualitative Organic Analysis,” 2nd ed., New York, John Wiley & Sons,1932. (5) Sando, J. Biol. C h m . , 56, 463 (1923). (6) Treadwell and Hall, “Analytical Chemistry.” 5th ed., Vol. 2, New York, John WiIey & Sons, 1923. RECEIVED January 30, 1935.
Determination of Lactose in Mixed Feed D. A. MAGRAW AND C. W. SIEVERT, American Dry Milk Institute, Inc., Chicago, Ill.
T
HE consumption of dry and sugar manufacture, the oil A method for the determination of lactose in m i k products in mixed meals resulting from the expresent-&y complex, mixed feed has been defeeds has shown a traction of oils from various meat veloped and shown to be accurate to ~ 0 . 2 per 5 nomenal i n c r e a s e during the products, many kin& of fish cent Of lactose* Corrections of blanks and losses 15 to 18 years since the f i s t use meals, whale guano, mineral mahave been incorporated. of these products by feed manuterials of various sorts, molasThe method entails no exceptionally specialized ses, and sometimes flavoring f a c t u r e r s . This increase has been particularly rapid during laboratory technic, materials. The meat packinghouse bythe last 5 or 6 years. The application of this method to the deterIt is of interest to feed manuproducts may be used separately, Of dry skim milk Percentages in mixed f a c t u r e r s a n d f e e d control or may be present in a mixture feeds is indicated, and application to other dairy sold as meat scrap. Evaporated officials to be able to find out tank water or “stick” is a soluble fhe amount of milk used in SOproducts in mixed feeds is suggested. protein material which is hard to called milk mashes and other f e e d m i x t u r e s featuring milk precipitate, but which may be as an ingredient. Qualitative methods for the detection of found in connection with tankage or other meat products. milk in feed mixtures do not give the desired information, The nitrogenous material in the various protein concentrates is and former quantitative methods have not been accurate not necessarily in combination as protein a t all times. Upon when applied to the present complex feed mixtures that are partial hydrolysis some of the amino-acid components of proteins are set free, causing difficulty in precipitation, on the market. The literature dealing with the determination of lactose or Some of the oil meals contain sugars which are not fermilk solids in mixed feeds is very limited. Two papers were mented by ordinary yeast, but which do reduce copper. It published in 1928. Coe (2) describes a method for the esti- has been necessary to devise an application of enzymes so that mation of dried buttermilk from the lactose content of feed, such sugars are changed into fermentable form, This must and names cottonseedmeal and soy bean meal as interfering be done without attacking or changing the lactose present. with the accuracy of the results obtained by his method. Davis (3) published seven methods to be used in determinMETHOD ing the amount of dry buttermilk in feed mixtures. Some APPARATUSAND REAGENTS.Centrifuge and suitable sediof these methods are quantitative and others are qualitative, ment tubes, and I G 4 Jena fritted-glass filtering crucibles. Animal diastase (Armour & Co., Pharmaceutical Dept., ChiDIFFICULTIES CAUSEDBY PRESENTFEEDMIXTURES cago, Ill.), invertase-melibiase scales (Wallerstein Laboratories, It has been possible to determine lactose in simple feed Wall St., New York, N. Y.),baker’s yeast, solution of saturated neutral lead acetate, solution of 5 per cent mercuric chloride, somixturesconsisting of ground grabs, mill feed, certain types lution of 20 per cent phosphotungstic acid, and hydrogen sulfide. of meat scrap, and dairy Products, by following Some Of the PROCEDURE. Place 16.25 grams (weighed to within 0.03 gram) methods previously recommended. hesent feed mixtUreS, of the well-mixed feed in a small beaker and extract the fat with a however, are apt to be exceedingly complicated. It is not out 100-cc. portion of ethyl ether, allowing it to stand for a short time. when the material has settled out completely, decant of the ordinary to find mash feeds containing from twelve to the ether. Allow t o stand in a warm place to evaporate remaintwenty different ingredients. Such ingredients the ing ether. Transfer the sample to a 300-cc. volumetric flask with various ground grains, the common grain by-products Pro- 200 CC. of distilled water, and digest in a hot water bath with duced in the milling of grain, the by-products of corn starch occasional shaking for a period of 30 minutes. Cool and fill
March 15, 1935
ANALYTICAL EDITION
t o volume with distilled water, then centrifugalize o f f the insoluble material. Place 150 cc. of this solution in a 200-cc. flask. To this add 100 mg. of animal diastase, 75 mg. of invertase-melibiase scales, and 1.5 to 2 grams of east. After fermentation at a temperature of 24.4" to 26.67' C. 66' to 80" F.) for a period of 40 to 48 hours, make tip to volume and centrifugalize to separate the yeast. Boil down 190 cc. of the liquid to about 25 or 30 cc. and wash into a 100-cc. volumetric flask by the aid of hot distilled water. To this add 10 cc. of saturated neutral lead acetate. Make up to volume with distilled water, and remove the precipitate by centrifugalizing. To 50 cc. of the resulting liquid in a 100-cc. volumetric flask add 2.5 cc. of a 5 per cent solution of mercuric chloride and allow t o stand for a short time (20 to 30 minutes) with repeated shakings, after which add 5 cc. of a 20 per cent solution of phosphotungstic acid. Make up to volume with distilled water, and separate the precipitate by means of the centrifuge. If the resulting solution has any suspended preci itate, filter through a dry filter paper before carrying further. iaturate the resulting liquid with hydrogen sulfide gas and filter. Next pipet 50 cc. of the clear, colorless solution into a 300-cc. Erlenmeyer flask, and boil until no hydrogen sulfide is given off. Then make up volume t o 50 cc. and determine the lactose by the Munson & Walker method. Consistently better results are possible with the Jena fritted-glass filtering crucible I G 4 than with an asbestos mat in a Gooch crucible. After thoroughly washing the precipitate, dissolve it with 5 cc. Qf 1to 1nitric acid, and determine the copper by the volumetric sodium thiosulfate method (1). The milligrams of lactose are interpolated from the Munson and Walker table (1).
CALCULATION To obtain the number of grams of dry material used in the aliquot, Formula 1 is used: 150 190 50 50 X 16.25 = 2.003grams (1) (300)xXxX100x(100 - 1) These figures are the dilutions which are made in the method and the (300 - 8) and (100 - 1)allow for a correction of 9 cc. for volumes of precipitates. As 2.003 grams are very nearly 2.000 grams and well within experimental error, 2.000 grams have been used in all further calculations. To obtain the percentage of lactose in the feed, Formula 2 may be used : (
t G ";PI
X 100 = per cent of lactose in feed
(2) grams of lactose determined 0.005 = correction for blank 0.90 = lactose factor for fermentation loss 2 = weight of dry material in aliquot used as derived in Formula 1
X
.-
When dry skim milk is the dairy product used in the feed mixture, it is possible to compute its percentage from the percentage of lactose present. On hundreds of samples of dry skim milk from all parts of the country, examined in this laboratory, the lactose content was found t o be from 48 to 52 per cent. The average lactose content is approximately 50 per cent. The formula for calculating the amount of dry skim milk in feed is as follows: Per cent Of lactose = per cent of dry skim milk in feed .5 0 0.50 = 50 per cent = per cent of lactose in dry skim milk
If dried buttermilk is the dairy product, then the lactose oontent of that material will have to be substituted in the above formula in place of the lactose content of dry skim milk. Dried buttermilk, especially when neutralizer is used, has a widely varying lactose content. Coe (2) has reported lactose contents of from 8 t o 40 per cent. I n recent years less variation has been noted and 36 per cent is suggested as a fair average value. The formula then becomes: Per cent Of lactose = per cent of dried buttermilk in feed
0.36
107
If dried whey is used as the dairy product in the feed mixture, the proper factor should be inserted in a similar formula. The lactose content of whey varies considerably, depending largely on the source of the whey. There is no adequate or reliable method for determining the exact type of dairy product used in an unknown feed mixture. Only when the exact type of dairy product used is known may the lactose percentage be convertible into definite percentage of milk product.
EXPERIMENTAL PREPARATION OF SAMPLES. In pursuing this work, it was necessary to make samples containing as many different meat, fish, and oil-meal products as possible, so that practically all combinations that are used in poultry mash feeds, in pig feeds, hog feed, concentrates, calf meals, etc., would be represented. These were hand-mixed in the laboratory. Samples of mash and other feeds made by reputable manufacturers of open formula feeds were also used. Formulas were selected which contained no milk. These were checked for blank determinations and were also used after definite amounts of milk had been added. Various mineral mixtures were also added to feed mixtures in order to study their effect. The following materials were used as feed ingredients of mixtures studied in the development of this method: Ground Grains. Corn, wheat, oats, barley, milo, kafir, sunflower seed, buckwheat peas, hemp, common beans, millet, rye, soy beans. Grain Products. Wheat bran, wheat middlings, wheat flour middlings, red dog flour, second clear flour, wheat germ meal, corn feed meal, hominy feed, corn gluten feed, corn gluten meal, barley feed, oat feed, clipped oat by-product, grain screenings, dried brewers' grains, dried corn distillers' grains, locust-bean meal, anise seed, and foenugreek. Oil Meals. Corn-oil cake meal, linseed-oil meal, cottonseed-oil meal, soy bean-oil meal, sesame-oil meal, coconut-oil meal, and peanut meal. Other Feed Ingredients. Alfalfa meal, alfalfa-leaf meal, dried beet pulp, cane (blackstrap) molasses, and charcoal. Meat Products. Meat cracklings, meat scraps, tankage, blood meal, and blood flour. Fish Products. White fish meal, menhaden fish meal, cod liver meal, sardine meal, and whale guano. Vitamin D Oils. Cod liver oil, concentrated cod liver oil, and sardine oil. Minerals. Ground oyster shell, ground limestone, steamed bone meal, spent bone black, rock phosphate, tri- and dicalcium phosphate, epsom salts, Glauber's salts, copperas, iron oxide, potassium iodide, common salt, sulfur, and wood ashes. PROTEIN ELIMINATION. A number of protein precipitants were tried, both singly and in various combinations, before a suitable procedure was finally adopted. The final combination which was adopted as outlined in the foregoing procedure does not entail any close manipulation of conditions such as acidity, temperature, pH, etc., and results in a crystal clear, colorless solution. It is important to have the solution very slightly acid before the protein precipitations for it has been found that in alkaline solution some lactose may be lost in the procedure. SUGARELIMINATION. It is necessary to eliminate all reducing sugars which interfere with the determination of the true percentage of lactose present in the feed mixture. Since the ingredients which are used in a complicated mixed feed contain widely varying types of sugars that are classified as reducing sugars and sugars nonfermentable by ordinary yeast, several enzymes must be used to bring about their fermentation from the original solution. One of these sugars is raffinose, which is found in several of the ingredients of feed mixtures. On fermenting with ordinary yeast, it breaks down into levulose and melibiose, which is nonfermentable but is a reducing sugar. Melibiose was finally eliminated by the use of invertase-melibiase scales.
108
INDUSTRIAL AND E N G I N E E R I N G CHEMISTRY
It was also found necessary to put in a small amount of animal diastase to speed up the starch hydrolysis, shorten the time of fermentation, and completely destroy reducing sugars other than lactose. FERMEKTATION. During the experimental work, it was found that fermentation time and temperature played a large part in the accuracy of the results. Many different temperatures and lengths of fermentation time were tried with varying results. This led to a %day fermentation in a cabinet, with controlled heat of 24.4' to 28.67' C. (76' to 80" F.) which produced accurate checks, and consequently was used in all following work. Higher temperature than this produced lactose loss. MOLD. During the study of fermentation times and temperatures, in a number of cases growth was found on the top of the fermented solutions, and under a microscope it was seen to consist of molds, In all instances where this substance was present the results were low. Whether these molds were carried through from the original feed material, or had been seeded along the line was not determined, but probably the contamination occurred in the laboratory. If samples show mold, they should be discarded and fresh samples started. LACTOSEDETERMINATIONS. It was necessary to try several different methods for the determination of lactose. The volumetric thiosulfate method (I) gave more satisfactory checks than gravimetric or direct determination of cuprous oxide. More satisfactory checks were also obtained by using a Jena fritted-glass filtering crucible, rather than an asbestos mat in a Gooch crucible, apparently because of a better retention of the copper oxide. The authors' findings on this were also substantiated by several workers in other laboratories. Colorimetric methods were not studied in this work. CORRECTIONS Feeds are made of moderately insoluble material. I n many cases in this procedure, quite voluminous precipitates occur and therefore a correction in volumes should be made. It was found on measuring that the total volume of precipitates from the first extraction of the original feed was 7 cc. The next largest precipitate was that from the phosphotungstic acid precipitation which amounted to 1 cc. The remainder of the precipitates were calculated to be approximately 1 cc., making a total correction for precipitates of 9 cc. In the calculations for this method, this 9 cc. correction hasbeenmade in Formula 1. The weight of the original sample of the dry material has been calculated so that the dry material represented in the aliquot taken for the lactose determination is 2 grams, thus simplifying further calculations. By using this method on over 100 samples of feeds containing no milk products, results were obtained representing approximately 0.25 per cent of lactose. It was found from these adjustments that 0.25 per cent should be subtracted from the results obtained, and this correction has been allowed for in the calculations under Formula 2, where 0.005 gram is subtracted from the grams of lactose determined in the feed. On carrying the fermentation for two days at 24.4" to 26.67' C. (76' to 80" F.), there was some loss in the lactose. The cause of this loss was not determined but, in carrying a series of over 20 milk samples through the method, a loss of approximately 10 per cent of the total lactose was found. These results indicated that a 10 per cent lactose loss should be accounted for. This correction has been made in Formula 2, and is represented by the 0.90 in the formula. After all the above corrections have been made, the results show this method can be depended upon to * 0.25 per cent of lactose. Among the many ingredients which are used in the com-
Vol. 7. No. 2
mercial feed mixes of today, only cane (blackstrap) molasses has given any difficulty in the above procedure. The trouble is due to the sugar "glutose" which is contained to the extent of 3 to 7 per cent in cane molasses, and is nonfermentable by ordinary yeast. At the present time, no enzyme or method of procedure has been found to eliminate such interference, but this is not important because very few feeds contain large amounts of molasses in combination with milk. It was found in the experiments that the presence of 5 per cent molasses indicated approximately 0.2 per cent of lactose, making it possible to make this correction on feeds containing molasses. RESULTS Some of the results which have been obtained in this laboratory on known samples are listed in Table I. These results have been calculated into percentage of dry skim milk, as that was the dairy product used in making the TABLEI. RESULTS ON KNOWN SAMPLES TYPEOF
(All corrections made) DRY S K I M FEED MILKADDED
Egg mash Milk egg mash Starter and broiler ration
%
None 4
None 3.97 3.81 4.29 8.23 8.09 8.09 12.87 13.34 12.64 4.64 0.29 4.23 4.63 8.23 8.40 6.12 6.15 9.05 None 5.95 4.62 None 7.50 12.29 4.14
8 12.5
Hog supplement Mixture (prepared in a state feed control laboratory)
5 None 4.3 8.26
Mash 6 per cent
Egg mash Laying mash Growing mash
DETERMINBID
%
Turkey starter
Hog supplement H o g meal
DRY S X I M MILK
6 8.9 None 5.94 4.75 None 7.45 11.86 4.46
samples. All the known samples were prepared outside the laboratory, and their dry skim-milk content was unknown to the laboratory until after the determination had been made and reported. Many other known samples were analyzed with equally satisfactory results. DISCUSSION This method is fairly long and entails a considerable amount of work and time. The authors believe, however, that it has been simplified as far as possible in keeping with the nature, variety, and number of feed ingredients used, or possibly used, in feed mixtures a t present. No difficult or exceptionally closely controlled condition is necessary in this method except that of fermentation temperature. The actual pH of the solution of any point of this procedure is of small consequence, except that a t all times a solution alkaline to litmus must be avoided. Excessively high acidities which might interfere with fermentation should be avoided, but such conditions hardly come up in feed mixtures. ACKNOWLEDGMENTS The authors are indebted for collaboration, assistance, and constructive comments to E. C. Thompson, director of laboratory, Borden Co., New York, N. Y.; A. H. Johnson, chemist, National Dairy Products Corp. Laboratory, Baltimore, Md.; and A. 31. Besemer, chief of Research Laboratory, Golden State Co., Ltd., San Francisco, Calif. They are also indebted to H. R. Kraybill of Purdue University,
March 15, 1935
ANALYTICAL EDITION
W. C. Rose of the University of Illinois, and C. H. Bailey of the University of Minnesota for their helpful suggestions in pursuing this work. The authors wish to thank W. Ndte of the American Dry Milk Institute, Inc., for his help in this manuscript, and especially to preciation to E*Copeland of this laboratory for his painstaking analytical work and whole-hearted cooperation.
109
LITERATURE CITED (1) Assoc. Official Agr. Chem., Official and Tentative Methods, 1930. (2) Coe, M.R.,J.Assoc. OfficialAgr. Chem., 11,251 (1928). (3) Davis, A. B., Ibid., 11,410 (1928). RECEIVED July 3, 1934. Presented before the Division of Agricultural and Food Chemistry at the 88th Meeting of the Ameriaan Chemical Pociety, Cleveland, Ohio, September 10 to 14, 1934.
Determination of Copper in Organic Matter A Note on Ansbacher's Method OLIVESHEETS,ROBERT W. PEARSON,AND MARVINGIEGER Mississippi Experiment Station, State College, Miss. Instead of placing the crucible containing the copper sulfide in N ACCURATE method for the determination of small glass triangle over a crystallizing dish, it is placed in a 50-cc. amounts of copper is of great importance to investi- aErlenmeyer flask, the top of which has been cut off so that the gators in the field of nutritional anemia, and is essential crucible will fit into it to a depth of about 1.25 om. (0.5 inch). A to the correct interpretation of results obtained from investi- lip is also made in one side of the flask for pouring and rinsing out gations on the role of copper in hemoglobin regeneration. the copper nitrate and sulfate solution. The use of the Erlenmeyer flask decreases the danger from copper contamination, and Unreliable methods for the determination of copper are loss of the sample due to the crucible tipping over. doubtless partially responsible for the conflicting opinions To dissolve the cop er sulfide and evaporate the copper nitrate which exist in this field of research. and sulfate solution t i e Erlenmeyer flask is placed on an alumiFor several years the Department of Home Economics has num water bath which rests on an electric hot plate having aluminum top and sides. The hot plate is placed in a metal-free been trying out copper methods in an attempt to find one hood lined with asbestos sheet rock. Evaporation to dryness can which would prove satisfactory in the analysis of food materials. usually be completed on the water bath, but it may be necessary Modifications of the xanthate (IO), the Biazzo (7), and the car- to place the flask directly on the hot plate for a few minutes. bamate method as described by Callan and Henderson (3) It is also desirable, as Ansbacher suggests, t o have a glass plate were all tried out, but the results obtained were not reliable. over the crucible and flask to exclude copper contamination. INTERFERING ELEMENTS. The authors have encountered (A recent modification of the carbamate method has been reported, 8, which the authors have not tried.) The only no difficulty with interfering elements in the foods analyzed. method which has thus far given consistent results is one in Ansbacher found that other metals which might occur in which chromotropic acid is used. This method was de- biological materials, including those of the second group, did veloped by Ansbacher (a),but has not as yet been adopted by not interfere with the method (5). However, if an excess of sulfur is present the final solution is many investigators in the United States. It is applicable to a wide variety of biological materials. The authors have used acid instead of neutral. When fuming nitric acid is used in it successfully for the analyses of vegetables and sirups, and wet ashing, a part of it may remain after digestion, and when Ansbacher and collaborators for the analyses of fruits, vege- hydrogen sulfide is passed through the solution considerable tables, milk, eggs, oysters, meat, and other animal products free sulfur is precipitated. This forms sulfuric acid in the (1, 2, 4). The chief difficulty in the use of this method is the subsequent treatment with fuming nitric acid, producing an determination of the end point when titrating with the acid residue on evaporation. When an unknown quantity of chromotropic reagent. The authors have been able to over- acid remains in the final solution, it is impossible to determine come this difficulty, and in addition have developed certain the quantity of ammonia to add without the use of an indicator. If the solution is too acid or too alkaline, low results details of technic which it seemed worth while to report. are obtained. Sulfuric and perchloric acids alone are, thereMETHOD fore, used in wet ashing, although nitric acid may also be used if care is taken to remove all of it after the digestion is comI~ESTRUCTIONO F ORGANIC MATTER. If the material to be analyzed contains sufficient copper so that a small sample (1to pleted. This may be done by cooling the digest, adding 3 5 grams) may be used, it can be wet ashed. If a larger sample volumes of water, and concentrating by rapid boiling. PREPARATION OF CHROMOTROPIC REAGENT.The sodium is required, dry ashing is preferable. With certain materials which are difficult to ash, such as sirups, a combination of wet salt of the nitroso-chromotropic acid is made by dissolving 0.35 gram of chromotropic acid (Eastman Kodak Co. No. and dry ashing was found to be most satisfactory. 1613) in 5 cc. of water and adding 0.21 gram of sodium carbonFor this method the weighed sample is placed in a silica or platinum dish and ashed in an electric muffle. The initial ate, 0.5 cc. of 2 N sodium nitrite, and a slight excess of dilute temperature should be below 100" C. The heat is rapidly acetic acid. Otherwise, the method described by Ansbacher increased until the temperature reaches 450' C. If a stream of for making up and standardizing the chromotropic reagent is air is drawn through the furnace, less time is required for ashing followed. and a more complete ash is obtained. The time re uired is about DETERMINATION OF COPPER. When titrating with the 2.5 hours, varying with the nature of the materij The ash is washed into a 300-cc. Kjeldahl flask with copperlfree distilled chromotropic reagent, a colorimeter was first used to deterwater; the dish is washed with 10 cc. of concentrated sulfuric mine the end point, since the change in color from a purple to a acid, which is added to the ash. The latter is digested for 10 brown tinge could be detected by this means when it could not minutes, adding 1 cc. of 60 per cent perchloric acid if necessary to be detected by the naked eye. When the eye becomes clear the solution, and transferred to a 300-cc. Erlenmeyer flask. trained to observe the slight color change, it is not necessary PRECIPITATION OF COPPERAND SOLUTION OF COPPER to use the colorimeter. In standardizing the chromotropic reagent as well as in the SULFIDE.The procedure is the same as that described by determination of copper in the unknown, the authors use a Snsbacher with the following modifications:
