ANALYTICAL EDITION
SEPTEMBER 15, 1939
503
light. The much faster fading in bright light is in conformity with the findings of other workers (1,d, 6,8). In other experiments carried out in tall Nessler tubes of 50-ml. capacity the amount of hydrogen peroxide was increased and the time during which the red color was permanent was greatly lengthened. The results are given in Table 11.
whether peroxide or permanganate is used, but, because of the fading of the ferric thiocyanate as soon as the permanganate is exhausted, precautions must be taken, when using this procedure, or the results will be low.
TABLE11. TIMEOF FADING OF FERRICTHIOCYANATE WITH HYDROGEN PER OX ID^ PRESENT
Hydrogen peroxide is a more satisfactory oxidant for iron than permanganate in the thiocyanate determination of iron. The red color can be made stable for several minutes, depending on the amount of peroxide used, and the faded color may be restored if necessary by the addition of more peroxide. Too much peroxide may cause a yellowish interfering color due to oxidation products of thiocyanate.
Iron
NaCNS, N
HzOZ, M
HCI, N
Time of Fading
P. p . m. 0.1 0.1
0.2 0.02
0.0032 0.0037
0.01
0.01
No fading in 15 minutes No fading in 2.5 hours
Further studies are being made by one of the authors on the rate of ferric thiocyanate fading under definite light conditions. The fading is believed by the authors to be due to the reduction of the iron and oxidation of the thiocyanate. If that is true, it should be possible to restore full color to a faded iron determination by adding peroxide. To test this hypothesis a solution 0.0028 M in peroxide, the same as used in making curve 2, Figure 1, was allowed to stand several hours. After 4 hours the fading had progressed from 35 scale units to 4. The addition of a second amount of peroxide, equal to the first, restored the original color. During the past 12 years many hundreds of colorimetric iron determinations have been made in this laboratory using peroxide alone as the oxidant; the results are the same
Summary
Literature Cited (1) Baudisoh, O., and Welo, L., J. Biol. Chem., 61, 261 (1924). (2) Bhattaoharyya, A. K.,and Dhar, N. R., J . Indian Chem.SOC., 6, 143, 197 (1929). (3) Hedenstrijm, A., and Kunau, E., 2. anal. Chem., 91, 17-26 (1932). (4) Offord, H.R., IND.ENQ.CHEM.,Anal. Ed., 7, 93 (1935). (5) Patten, C. G., and Smith, H. D., Trans. Roy. SOC.Can. (3), 22, 111, 221 (1928). (6) Pulsifer, H. B., J. Am. Chem. SOC.,26, 967 (1904). (7) Sharma, B.S., J. Chem. Soc., 1930,308. (8) Sharma. B. S., J. SOC.Chem. Ind., 48,336 (1929). (9) Urk, H, W.van, Pharm. Weekblad, 63, 1101-7 (1926). (IO) Winsor, H. H., IND.ENQ.CHEM., Anal. Ed., 9,446 (1937). FROM theses presented by Majel M. MacMasters and Chester L. French for the M.S. degree.
New Light Sources for Colorimetry F. L. MATTHEWS, R. H. CRIST, AND A. KNOLL Columbia University, New York, N. Y.
T
HE application of a colorimetric method of analysis requires that the system under consideration follow Beer’s law, which is theoretically valid only for monochromatic light. However, in practice, it is customary to use continuous or band emission light sources and filter combinations which produce incident illumination of varying degrees of monochromation. I n addition to the possibility of major deviations from Beer’s law there are other difficulties inherent in such procedures. When permanent standards, differing from the unknown, are used the difference in hue between unknown and standard is greater the broader the spectrum of the incident illumination. In addition, the practical difficulty of determining small changes of transmission in a restricted spectral region in the presence of strong accompanying transmission in other regions (background) becomes greater as the width of the spectrum of the incident light increases. Although approximations to monochromatic light have been obtained by the use of appropriate filters or combinations of filters, the necessary approximation can often be secured in this way only by the use of exceedingly intense sources (which in the case of tungsten filament lamps generate large amounts of heat) or by the sacrifice of intensity of illumination and therefore of speed and precision. The lamps described herein were developed in order to secure very high intensity sources of illumination which were relatively cool and whose emission was concentrated in some
restricted spectral region. These permitted the use of thick filters and the attainment of near monochromation without. too great a loss of intensity. These lamps were used with the standard Klett-Beaver visual colorimeter, but can be adapted readily to other instruments such as the photoelectric colorimeter, etc. For the colorimetric determination of sodium chloro hyllins (unpublished work) in aqueous solution use was made o f a neon source. The lamp was a spiral of 6-mm. tubing 45 mm. long and 35 mm. in external diameter. With this was used a Corning signal red filter (No. 243) and the standard was a copper sulfate solution (1). The precision attainable (in the determination of concentrations) in this way was approximately 5 parts per thousand for nine readings made in 2 minutes. For the colorimetric determination of a-naphthylamine (unpublished work) a green fluorescent lamp was used in conjunction with a Corning Sextant green (No. 401) filter with a pentammino cobaltic chloride solution ( I ) as the standard. This lam was in the sha e of a doubled “U” and was 11 cm. long and 3b)mm. in externaf diameter. The precision attainable in this case was the same as that given above. Both these lamps, when in use, were inserted in the place ordinarily occupied by the 25-watt tungsten lamp in the KlettBeaver colorimeter. Similar lamps, as well as a blue fluorescent lamp, can be obtained in a variety of designs from Claude Neon Lights, Inc., Long Island City, N. Y.
Literature Cited (1) Kasline, C. T., and Mellon, M. G., IND. ENQ.CHEM., Anal. Ed., 8,463 (1936).