THE DETERMINATION OF CAROTENE IN BUTTER FAT BY HAROLD M. BARNETT (From the Research Division of S. M. A. Corporation, Cleveland) (Received for publication, December 18, 1933) Since the discovery that carotene is converted into vitamin A in the animal organism, the importance of accurate determinations of the carotene content. of butter has greatly increased. The method outlined by Palmer ((1) p. 259) was of value in showing in a general way the amount of carotene in butter fat, but it is doubtful if this was expected to be used as a standard method. That it has been so interpreted, however, isshown by referencesin theliterature such as those of Rogers (2) and of Wilbur, Hilton, and Hauge (3). The Palmer method assumes carotene to impart the same color intensity to oils such as butter fat as an equal proportion of carotene gives to solvents such as petroleum ether. As previously pointed out by Goldblatt and Barnett (4), such an assumption is in error since the color intensity of carotene in oils was found to be several times that in petroleum ether. In first attacking this problem, an attempt was made to work out a ratio between the color of carotene, as determined colorimetrically in petroleum ether, and the color of an equal quantity of carotene in various edible oils, allowance being made for the original color of the oils. Although several oils were found to give the same ratio, this conversion factor could not be applied to all the oils tested and was, therefore, of questionable application to butter fat. A test was then performed in which some butter fat was decolorized with activated carbon and various amounts of pure carotene dissolved in portions of it. Calorimetric determinations were made to determine the ratio between the actual carotene added and the amount as determined by the Palmer method. Results were inconsistent and it was found that the carotene rapidly decomposed after addition to the decolorized butter fat. This is in 269
Carotene in Butter Fat confirmation of the recent findings of Shrewsbury and Kraybill (5) published while this paper was in the process of preparation. Either an antioxidant had been removed or an oxidation catalyst added during the carbon treatment. Attention was then turned towards the development of a spectrophotometric method for the determination of the carotene content of butter fat. One spectrophotometric method has already been developed and described by Schertz (6) for the determination of carotene in solvents such as alcohol and petroleum ether. My determinations showed, however, that this method could not accurately be applied to butter fat, because the transmittancy of light at various wave-lengths was different for carotene in oil solutions than for an equal concentration in alcohols or ethers. And because of the low percentage of carotene in butter fat, it is impractical to dilute samples of butter fat sufficiently with these solvents to make the effect of the oil on the spectrophotometric determinations negligible. Furthermore, the dilution required would depend on the carotene content of the butter fat to be tested and the error would not be consistent. It seemed well, therefore, to pick an oil as a diluent and solvent for the butter fat determinations. The oil selected for this purpose was refined coconut oil, as it has very little color and has an advantage over other white fats, such as lard, in having a lower melting point. For the spectrophotometric determinations a model C Keuffel and Esser color analyzer (7) was used. The absorption cells were 3 cm. in length. The carotene used was prepared from carrots and carefully purified by recrystallizing several times from hot petroleum ether. Its purity was checked by calorimetric and spectrophotometric determinations. This purified carotene was preserved in evacuated ampules, fresh ampules being opened as needed. Because of the small amounts of carotene required to fortify the butter fat and other oils used, stock solutions containing 25 mg. of pure carotene in 50 gm. of refined coconut oil were prepared. Precautions were taken against decomposition of carotene and fresh solutions and tests for loss were made frequently. Because of the instability of carotene in decolorized butter fat, as mentioned above, the determinations which furnish t,he basis
H. M. Barnett 261 for the new method here developed were made on fresh butter fat of known history and low in carotene content. To portions of this butter fat diluted with coconut oil were added various amounts of the stock carotene solution described above, so that fortifications of 0.75, 1.50, and 2.50 mg. of carotene per kilo had been made. As a blank in each case, some of the original butter fat was diluted with coconut oil in the same proportion as the fortified butter fat. In this way a compensation was made in each reading for the absorption of the original butter fat and, therefore, absorption measurements were solely those due to added carotene. As an example of the method used, readings were made for 1.50 mg. of carotene per kilo of butter fat on solutions prepared as follows: 0.18 gm. of stock carotene solution (1 mg. of carotene in 2 gm. of oil) was diluted with 19.82 gm. of butter fat and this in turn was diluted 2:l with coconut oil. As a blank, some of the unfortified butter fat was diluted 2:l with coconut oil. This dilution with coconut oil served three purposes: (1) it lowered the absorption of light by the blank, (2) it lessened the tendency for the butter fat to solidify in the tubes while measurements were being made, and (3) it gave a ratio of butter fat to coconut oil of an order similar to that expected in making spectrophotometric determinations of the carotene content of various butter fats. The results obtained on the three different concentrations of carotene in butter fat are shown plotted in Chart I. Only those transmittancies obtained for wave-lengths 455 to 500 rnp are plotted, since the intensity of light was too low for accuracy in the lower wave-length region and curves plotted for wave-lengths above 500 mp tended more towards the horizontal direction, which would mean greater errors in reading results from Chart I. Each point represents the average of five or more readings at that particular wave-length. In cases where two or more lines fell very close together, an average was taken and one line plotted. It is interesting to note that the absorption of carotene in butter fat is practically the same over the spectrum range of 475 to 490 mp. No doubt there is a definite absorption band in this region, but with the spectroscope slit open as for these determinations, such a band was not discernible. Also, in plotting the concentration of carotene in oil against the --log,, transmittancies, it was found that the points do not fall on a straight line as when similar meas-
Carotene in Butter Fat uremcnts are made on carotene in organic solvents such as petroleum ether (6). Thus it would appear that Beer s law does not apply to dilute solutions of carotene in oils. 0 0.5 1.0 1.5 a.0 25 3,o MGS. CRRDTENE PER KILO CHART I. Transmittnncies of varying concentrations of carot,ene in butter fat (3 cm. cell) at different wave-lengths. For use in the determination of the carotene content of butter fats or other oils in which carotene is the only pigment of consequence. To test the value of Chart I in determining the carotene content of various butt er fa.ts, five lots of butter of known history and
H. M. Rarnett 263 nat,ural color were obtained. Different breeds of cows on different rations are represented in this group. The butters were melted; the curd, salt, and water allowed to settle; a portion of each butter fat filtered twice and refrigerated. Spectrophotometric determinations were made on each butter fat diluted with coconut oil, coconut oil also being used as the blank. The dilutions were not t,he same in all cases, because it was desired to make readings from the chart at various concentrations. In Table I are shown both the transmittancies of these butter fat solutions at various wave-lengths and the carotene content in each case as taken from Chart I. It may be noted that in no case does a det,ermination at any of the ten wave-lengths show a deviation from the mean of over 7 per cent, and in most cases determinations are quite close to the mean. By multiplying the average concentration by the dilution, the carotene content of each individual butter fat is shown. The lots vary from 7.3 to 13.0 mg. per kilo of butter fat. All of these butter fats were from cows receiving more or less green pasture and, therefore, do not cover the complete butter color range. In Table I are also shown the results obtained when applying the same methods to other dilute solutions of carotene in oils. Lot 6 was a solution of carotene in Wesson oil which had been used for biological tests. It was prepared to contain 11.3 mg. of carotene per kilo. The determination by the new method showed 11.6 mg. per kilo. Lot 7 was pure carotene dissolved in some oleomargarine. It was made up to contain 10.0 mg. of carotene per kilo and the average results as taken from the butter fat chart are 9.9, mg. per kilo. It would appear from these two determinations that the method worked out for carotene in butter fat could be applied very well to determinations of the carotene content of other oils, that is where carotene is the only pigment of consequence present. In the determination of the carotene content of butter fat by this method, it is suggested that the sample be diluted with refined coconut oil in such a manner that the carotene content lies between 0.75 mg. and 2.5 mg. per kilo and that determinations be made at the various wave-lengths shown in Chart I, coconut oil being used as a blank. An average of these determinations should give more accurate results than when readings are taken
TABLE I Determinations of Carotene Content of Butter Fats and of Other Oils by New Spectrophotometric Method is given in per cent. is given in mg. of carotene per kilo. Lot NO. Wave-length in mp 455 / 460 / 466 / 4i0 j 475 j 480 / 485 / 490 / 495 j 500 1 - ;e - Dilution 1 luttcr fats Crystalline caretent in oils 10.7 120.1 22.3 125.2 h6.7 26.6 26.0 26.9 30.7 37.9 0.94 0.96 0.97 0.95 0.98 0.97 1.00 0.98 0.99 1.00 0.9i 2.6 12.8 14.1 17.0 17.6 18.4 18.1 19.4 24.1 31.0 1.35 1.32 1.35 1.30 1.35 1.30 1.33 1.27 1.21 1.23 1.3C 1.1 10.8 13.9 15.3 15.0 15.1 15.9 19.5 26.8 1.44 1.46 1.48 1.49 1.50 1.50 1.45 1.43 1.41 1.46 8.9 9.0 10.9 12.1 13.4 12.5 12.3 13.8 16.6 23.7 1.65 1.64 1.59 1.62 1.62 1.68 1.69 1.58 1.59 1.56 1.62 4.3 4.1 5.4 6.4 6.7 6.3 6.7 8.8 13.7 2.51 2.58 2.51 2.42 2.36 2.44 2.36 2.29 2.25 2.41 5.1 4.5 6.2 7.1 6.8 6.4 6.9 9.0 14.9 2.30 2.46 2.34 2.29 2.34 2.42 2.33 2.27 2.13 2.32 6.2 5.9 6.8 8.2 9.8 10.0 9.6 9.9 12.2 18.3 2.05 2.12 2.08 2.03 1.93 1.91 1.95 1.92 1.93 1.87 1.98 7:l 9:l 4:l 7:l 4:l 4:l 4:l 13.00 b 7.30 3 12.96 12.05 5 ct- 11.60 9.90 T
I-1. M. Barnett 265 at only one or two wave-lengths. Also, through the use of a wider spectral region, one can tell with some degree of certaimy whether foreign colors have been added to the butter fat. This method was developed with full knowledge that the spectrophotometer is not a common instrument in biological and dairy laboratories. For this reason a conversion factor has been worked out which appears to give fairly accurate results when applied to determinations made by the Palmer calorimetric met hod for the determination of the carotene content of butter fat. In t,he case of the butter fats, and also the oils colored with pure carotene, TABLE Iftrtio hrfween Carotene Content oj Butter Pats and Othw Oils As lktrrm~nd b!y (ivlorimetric and Xpectrophotometric Methods 13ut~trr fak Crystdline carotene in oils - 1 37.2 2 24.0 3 38.3 4 23.5 5 25.7 6 25.8 7 30.9 32lorim etrr ty. per I 26.7 41.3 25.9 42.2 35.6 II nry. pm icy. 28.7 44.4 27.8 45.4 41.5 38.0 40.9 32.1-34.5 ny. per lq. 7.76 13.00 7.30 12.96 12.05 11.60 9.90 0.27 0.29 0.26 0.29 0.29 0.28 Average 0.28 0.29 shown in Table I, a calorimetric determination was made on each of the undiluted oil solutions with 0.2 per cent potassium dichromate solution set at 30 mm. as a standard. The average reading in each case is shown in Table II. From these average readings the concentration of carotene in mg. per liter was calculated through the use of Palmer s chart ((I) p. 260). These results in turn were transferred to mg. per kilo and compared with the results obtained by the new spectrophotometric method. By dividing the concentrations obtained by the new method by those obtained by the calorimetric method, conversion factors were secured to be applied to results by the calorimetric method.
Carotene in Butter Fat The variations between the conversion factors obt ained for t)he various lots of butter fat and for the solutions of pure carotene in oils were not large. The average of seven such comparisons gives a conversion factor of 0.28 which it is thought may be applied very satisfactorily to results obtained by the calorimetric method where a spectrophotometer is not available. As a further check on the conversion factor obtained by the method given above, two samples of butter fat were carefully compared calorimetrically with standard dichromate solution set at 30 mm. The one lot was fortified with pure carotene in an amount equivalent to 4 mg. per kilo and to the other were added 6 mg. per kilo. Then calorimetric determinat ions were again made. By dividing the amount actually known to have been added in each case by the amount found by the difference in t,he two calorimetric determinations before and after the fortification, factors of 0.26 and 0.29 respectively were obtained, thus confirming the factor of 0.28 obtained above. It should be clearly understood that neither the spectrophotometric nor the calorimetric method here developed is applicable to the determination of the carotene content of butters containing commercial butter colors such as annatto or yellow AB and yellow OB. Research is now in progress, however, designed to develop methods not only for the detection and identification of artificial butter colors but also for a quantitative determination of the proportion of the total color of any commercial butter which is due to pigments other than carotene. SUMMARY 1. A new spectrophotometric method for the determination of the carotene content of butter fat has been developed. It is based upon transmittancies at different wave-lengths of varying amounts of pure carotene dissolved in butter fat. 2. Determinations by the new method have been made on five samples of butter fat. Results ranged from 7.3 to 13.0 mg. of carotene per kilo of butter fat. 3. The new method was found applicable -to other dilute solutions of carotene in oil. 4. A comparison of determinations of the carotene content of various butter fats and of other oils by calorimetric and spectro-
H. M. Barnett 267 photometric methods was made. A factor of 0.28 was obtained which may be applied to calorimetric determinations where a spectrophotometer is not available. BIBLIOGRAPHY 1. Palmer, L. S., Carotinoids and related pigments, American Chemical Society monograph series, New York (1922). 2. Rogers, L. A., Fundamentals of dairy science, New York, 98 (1928). 3. Wilbur, J. W., Hilton, J. H., and Hauge, S. M., J. Dairy SC., 16,155 (1933). 4. Goldblatt, H., and Barnett, H. M., Proc. Sot. Exp. Biol. and Med., 30, 201 (1932). 5. Shrewsbury, C. L., and Kraybill, H. R., J. Biol. Chem., 101, 701 (1933). 6. Schertz, F. M., J. Agric. Research, 26, 383 (1923). 7. Keuffel, C. W., J. On&al Sot. America, 11, 403 (1925).
THE DETERMINATION OF CAROTENE IN BUTTER FAT Harold M. Barnett J. Biol. Chem. 1934, 105:259-267. Access the most updated version of this article at http://www.jbc.org/content/105/2/259.citation Alerts: When this article is cited When a correction for this article is posted Click here to choose from all of JBC's e-mail alerts This article cites 0 references, 0 of which can be accessed free at http://www.jbc.org/content/105/2/259.citation.full.h tml#ref-list-1