ANALYTICAL EDITION
November 15, 1941
the factor
(s)'"
as above, we find 3.40 microamperes as
the diffusion current of europium a t a potential of - 1.6 volts. Thus the true diffusion current of ytterbium is 6.00 - 3.40 or 2.60 microamperes. The results obtained for ytterbium, in the mixture of europium oxide and ytterbium oxide described above, were 42.0 and 43.1 per cent, respectively, in two determinations with different concentrations of rare earth. The summations of europium oxide and ytterbium oxide were 97.0 and 97.8 per cent, respectively. Since the sample was known to contain only small amounts of other rare earths, it is believed that the results for ytterbium are essentially correct. If no correction had been made for the effect of decreasing drop time on the europium diffusion currents, the summations would have been 94.1 and 94.7 per cent, respectively. The tangent method of measuring wave heights would likewise lead t o erroneously low results for ytterbium because of the unsymmetrical nature of the second wave. The magnitude of the correction due to changes of drop time will, of course, increase with a n increasing ratio of europium to ytterbium.
Summary Ammonium chloride is a suitable indifferent electrolyte for the determination of europium and ytterbium. I n 0.1 N ammonium chloride the half-wave potentials of europium and ytterbium were found to be -0.671 and -1.415 volts with reference to the saturated calomel electrode. The polarographic method for europium gives results agree-
829
ing within 3 per cent with the volumetric method using the Jones reductor. Conditions favoring the successful polarographic determination of ytterbium are discussed. Reliable results can be obtained if well-defined reduction waves resembling those of pure ytterbium are obtained. Satisfactory results were obtained with yttrium earth ores containing about 5 per cent ytterbium. Ores of low ytterbium content, particularly those high in cerium group content, mere found to yield low results b y the polarographic method. I n the simultaneous determination of europium and ytterbium i t is necessary to correct for the effect of decreasing drop time with increasing negative potential.
Literature Cited Bruckl, 8 . ,and Noddack, W., Angew. Chem., 49,533 (1936). Heyrovsky, J., and Ilkovic, D., Collection Czechoslov. Chem. Commun., 7, 198 (1935). Holleck, L., 2. anal. Chem., 116, 161 (1939). Holleck, L., 2. Elektrochem., 46, 69 (1940). Ilkovic, D., Collection Czechoslov. Chem. Commun., 6, 498 (1934). Jantsch, G., Grubitsch, H., and Lischke, E., 2. Elektrochern., 43, 293 (1937). Lingane, J. J., J . Am. Chem. SOC.,61, 2099 (1939). Lingane, J. J., and Kolthoff, I. M., Ibid., 61, 825 (1939). Lingane, J. J., and Laitinen, H. A., IND.EXG.CHEX,ANAL.ED., 11, 504 (1939). McCarty, C. N., Scribner, L. R., Lawrena, M., and Hopkins, B. S., Ibid., 10, 184 (1938). McCoy, H. N., J . Am. Chem. Soc., 58, 1578 (1936). Noddack, W., and Bruckl, A., Angew. Chem., 50, 362 (1937). Tomes, J., Collection Czechoslov. Chem. Commun., 9,12 (1937).
Determination of Lead Content of Commercial
Ciders and Vinegars by Spectrographic Methods CHARLES W. SCHROEDER AND HERItIANN C. LYTHGOE Massachusetts Department of Public Health, Boston, Riass.
B
ECAUSE of the necessary spraying of vegetation subsequently to be used for food there results a certain amount of residual lead which often cannot be removed, particularly in the by-products of the apple crop. Excessive spray residue is often removed from apples which are sold for consumption as such, but not necessarily from apples intended for use in the manufacture of cider and vinegar. This investigation was undertaken to devise rapid and accurate spectrographic methods for the determination of lead in cider and vinegar, as well as to determine the range of lead content of these products as they appear on the market. All apparatus used was freed from lead by rinsing with hot nitric acid and lead-free water redistilled in a Pyrex still. The bismuth chloride was freed from lead by precipitating bismuth oxychloride with lead-free water, decanting the supernatant liquid, dissolving the precipitate in double-distilled hydrochloric acid, and repeating the process until the salt was spectroscopically free from lead. The calcium acetate was freed from lead by precipitating with hydrogen sulfide, using copper as a coprecipitmt. ANALYSISOF CIDER. Twenty-five cubic centimeters of cider were placed in a platinum dish, to which were added 0.5 mg. of bismuth as bismuth chloride and 0.04 gram of calcium acetate. The addition of calcium was found to enhance both the lead and the bismuth lines. The mixture was evaporated nearly to dryness on the steam bath and charred over a small flame, being allowed to take fire when charring was almost complete. The char was ground to a fine powder and three approximately equal por-
tions of such size, experimentally determined, as would give lines of convenient length on the plate were placed in cupped graphite electrodes-about one sixth of the total char was placed in each electrode. It was found unnecessary to weigh these portions, as considerable variation in the amount used had no appreciable effect on the analysis. These samples were then burned in an arc of 220 volts and 9 amperes, the lead being determined by the internal standard method, using a rotating logarithmic sector in front of the slit. The lengths of the bismuth line a t 2898 A. and the lead line a t 2833 A. were compared. The lengths of the lines from the divided samples were averaged before plotting. TABLE I. LEADCONTENT OF CIDERAND
No. of samples
Cider 13'3 7
Cider Vinegar 49
OF
VINEGAR
Distilled or Spirit Malt Vinegar Vinegar 5
1
Parts per million 0.18 0.00 0.37 0.51 0.55b 0:iS 0:63
Malt and Spirit Vinegar 1 ?
....
.. *.
.. .. *.
....
Lowest 0.10 Lower quartile 0.18 hfedian 0.27 Average 0.43 0:OS Geometric mean 0.32 0.50b Upper qdartile 0.54 0.78 Highest 1.50 11.8OC 1:20 .. Not including 5 samples containing less than 1 part lead per million. b Exclusive of highest sample. Including highest sample: average = 0.78 geometric mean = 0.53. 6 Duplicate determination by dithizone extraction method gave 8.0 pnrts of lead per million.
.. ..
830
Vol. 13, No. 11
INDUSTRIAL AND ENGINEERING CHEMISTRY
I n a series of seven analyses in the range of from 0.14 to 3.4 parts per million of lead the average error was 1.5 per cent and the maximum error was 6.0 per cent, using cider containing known added amounts of lead. I n this case six separate differences in line lengths were averaged to give each of the seven points on the working curve. Eighteen samples of commercial cider were examined, and all contained minute amounts of lead (Table I). Five samples which were examined only to an accuracy of less than 1 p. p. m . of lead are not included in the table. ANALYSISOF VINEGAR.Twenty-five cubic centimeters of vinegar were evaporated to dryness in a platinum dish, and the residue was dissolved in 0.5 cc. of double-distilled hydrochloric acid containing 1 mg. of bismuth per cc. The well-mixed solution of the vinegar residue was divided equally among three cupped graphite electrodes which had previously been treated with one drop of lead-free kerosene to prevent the solution from soaking into the electrodes. The determination was then carried out as in the cider method.
I n one series of twelve analyses, using vinegar containing known added amounts of lead in the range of from 0.4 t o 1.2 p. p. m., the average error was 7 per cent and the maximum error was 20 per cent. I n another series of six analyses in the range of from 0.08 to 0.4 p. p. m. of lead the average error was 9 per cent and the maximum error Ras 25 per cent.
Fifty-six samples of commercial vinegar were examined, and all but two, both distilled and spirit vinegar, contained lead (Table I). The geometric mean is nearer the median than is the arithmetic mean, indicating the geometric nature of the series. The lead tolerance adopted by the United States Department of Agriculture under the old food and drug lam was 0.025 grain per pound, which is equivalent t o 3.5 p. p. m. Only one of these samples exceeded that figure. The next lower result was 1.8 p. p. m. or approximately one half of the tolerance. The lead content of the cider was on the average less than that of the cider vinegar. There are two possible explanations. The cider was made from Kew England apples which do not require so much spraying as do western apples, but the vinegar was not exclusively a Massachusetts product. It is customary in the manufacture of sweet cider prior to placing it on the market to filter the apple juice through sand or some other type of filter which may remove some of the lead, but in the manufacture of vinegar this filtration is not always carried out. Furthermore, vinegar is a good solvent for lead in paint, metals, etc., with which it may come in contact during processing. P R E ~ E N TinEpart D before the Division of Agricultural and Food Chemiitry a t the 100th Meeting of the American Chemical Society, Detroit, Mich.
A Method for the Identification of Nitriles HAROLD B. CUTTER
AND
MICHAEL TARAS, Wayne University, Detroit, Mich.
S
EVERAL methods have been proposed for the identification of nitriles (1, 2, 8). The authors have found that adaptation of the well-known method of reduction of nitriles to primary amines by sodium and absolute alcohol gives a practical method for the identification of aliphatic nitriles. Aromatic nitriles do not give such good results, but this class can usually be readily determined by hydrolysis to the corresponding amide or acid.
Procedure A solution of 1 cc. (0.8 t o 1.0 gram) of the nitrile in 20 cc. of absolute alcohol is placed in a 200-cc. round-bottomed flask fitted with a reflux condenser. (It is essential that the alcohol be absolute, otherwise considerably less derivative is obtained.) The flask is immersed t o the neck in a water bath heated to 50" to 60" C. Fresh, finely cut sodium (1.5 grams) is added gradually through the top of the condenser as rapidly as possible without allowin the reaction t o become too vigorous. When all the sodiumaas reacted (10 to 15 minutes) the reduction is complete. The mixture is cooled to 20°,and 10 cc. of concentrated hydrochloric acid are added in small portions through the top of the condenser. Care is necessary on account of the s attering which takes place when the acid strikes the strongly aPkaline mixture. The reflux condenser is disconnected, the system is set up for ordinary distillation, and 20 cc. of alcohol are distilled into a graduated cylinder. The residue in the flask is cooled to 20" and a solution of 6 grams of sodium hydroxide in 6 cc. of water is cautiously added. The reaction at this point is violent, and care is necessary to avoid loss of amine by volatilization. The flask is swirled to ensure mixture of the ingredients and then rapidly reconnected to the condenser. Using a smoky flame the flask is heated until the contents are nearly dry, catching the distillate in a 50-cc. Erlenmeyer flask containing 3 cc. of water. The condenser should be fitted with an adapter dipping beneath the surface of the water in the flask. If the original substance was a nitrile, the distillate.wil1 be alkaline at this point. Phenylisothiocyanate (0.5 to 1.0 cc.) is then added to the distillate, and the mixture vigorously shaken for 2 or 3 minutes. If no derivative forms on shaking, scratching the walls of the Erlenmeyer and cooling under a t a or in an ice bath will brin down the precipitate. Aliphatic grivatives as a rule responj t o shaking; aromatic compounds require cooling in an ice bath.
The crude derivative is filtered, washed with 50 per cent alcoho and recrystallized from dilute alcohol in the usual manner. Because reduction in the case of aromatic nitriles is less smooth, an initial sample of 2 cc. or 2 grams is recommended. I n Table I are given the results obtained with ten aliphatic and four aromatic nitriles. The product was in most casea recrystallized from dilute alcohol, two recrystallizations usually being sufficient to yield a pure product. The method is applicable only to those aliphatic nitriles which form a volatile amine upon reduction. I n the aromatic series the method works less well, probably because of the lack of volatility of the amine and the fact that reduction in the aromatic series is accompanied by side reactions. I
TABLE I. IDENTIFICATION OF KITRILES Weight of Weight No. of M. P. of CrYS- Phenyl Derivaof Nitrogen Nitrile tallica- Thio- tive Ohtained Calculated Found urea tions Used O
c.
Gram
70
... ...
0.8 2 106 0.8 Acetonitrile 0.8 2 63 0.8 Propionitrile 0.8 2 65 0.7 n-Butyronitrile 0.8 2 82 0.6 Isobutyronitrile 0.8 2 69 0.7 n-Valeronitrile 0.8 2 102 0.6 Isovaleronitrile 0.8 2 77 0.3 n-Capronitrile 0.8 2 112 0.25 li:85 Isocapronitrilea 0.8 2 148 0.9 15.04 Glutaronitrilea 0.7 2 168 0.9 15.63 Succinonitrile" 2.0 2 147 0.3 11.56 Benzonitrile" 1.0 3 144 0.10 10.93 p-Tolunitrilen 3.0 4 179 0.3 10.93 o-Tolunitrile' 5 140 0.33 9.58 8-Naphthonitrile" 3.0 , a Derivatives not reviously described. No satisfactory resulta obtained with a-napgthonitrile or with m-tolunitrile.
... ... ... ...
%
... ...
... ... ... ... li:97 15.00 15.64 11.62 10.84 11.35 9.69 aould be
Literature Cited (1) Condo, Hinkel, Fassero, and Shriner, J. Am. Chem. SOC.,59, 230 (1937). (2) Howells and Little, Ibid., 54, 2451 (1932). (3) Shriner and Turner, Ibid., 52, 1267 (1930).
