690
ANALYTICAL CHEMISTRY
However, any cation forming an insoluble oxinate a t pEI values of 4 or above will interfere, and the procedure is less clean-cut than that for gallium. ACKNOW LEDG,MENT
The authors wish to express sincere appreciation to the Office of Naval Research for the support which rendered this investigation possible. LITERATURE CITED ( 1 ) Chrktien, A . , and Longi, Y.. Bull. S O C . c h i n . , 11, 245 (1944). (2) Feigl, F., Nature. 161,436 (1948). (3) Feigl, F., and Baurnfeld, L., Anal. Chim. Acta, 3, 83 (1949). (4) Geilmann, W.,and Wrigge, F. W., Z. anorg. allgem. Chem., 209, 129 (1932). (6) Gentry, C. H. R., and Sherrington, L. G., A n a l y s t , 71, 432 (1946).
(6) Lacroix, S., A d . Chi'nt. .4ctrr, 1, 260 (1947).
(7) Ibid.,2, 167 (1948). (8) Lacroix, S., dissertation, rniversity of Paris, 1948. (9) Moeller, T., IND.END.C H E S i . , . i s k i , . ED..15, 270 (1943). (10) Ibid., p. 346. ( 1 1 ) Moeller, T., and Cohen, A. .J.. Birnonthl,f/ Report 6, NR 052060, N60ri-71, T. 0. XVII, Ciiiversity of Illinois, . i p d 1 t o (12) (13) (14) (15) (16) (17)
June 1,1948. Parks, T. D., and Lykkeii, L., Ax,ii,. CHEM., 20, 1102 (1948). Rey, A. B., Ion, 7 , 3 8 9 (1947). Royer, G . L., IND.ENG.C m a i . . VAL. ED.,12, 439 (1940). Sandell, E. B., . ~ N A L .(:HEM.. 19, 68 (1947). Sandell, E. B., IND.E s c . C H E x AN.AL.ED.,13, 544 (1941 1 Spacu, G., and Pop, A . . Z . ami. Chem.. 120, 322 (1940).
RECEIVED Aiiril 12, 1949. . \ h t r a r t e d f r o m papers presented a t t h e I l 4 t i l iMeeting of t h e AMERICAN CHLVICAL S O C I E T Y , Portland, Ore., and a t t h e 115th Meeting of t h e . 4 v E ! a r + \ C I i E \ i r c h L SOCIETY, San Francisco. Calif.
Infrared Spectrometric Determination of Deuterium Oxide in Water VERNON TIIORNTON AND F. E. CONDON Phillips Petroleum Company, Bartlesville, Okla. The deuterium oxide content of deuterium oxide-water mixtures containing 3 to 1 0 0 q ~deuterium oxide is determined with a precision of 1 part in 100 from the absorbance of the 3.98-micron deuterium oxide band in the infrared absorption spectrum of a sample diluted with water s o that the deuterium oxide concentration is in a measurable range. The per cent deuterium oxide is read from a calibration curve. The time required for each determination is approximately 10 minutes.
I
N .4 continuation of a study of spectra and structure (2), acetone-& was prepared from acetone by repented deuteriumhydrogen cxchange with 99.8% deuterium oside in the presence of potassium carbonate as catalyst ( 1 ) . Progress of the cxchange reaction in its intermediate stages was followed by determining the dcuteriuni oxide content of the water remaining after reriioval of the acetone-d, by fractional distillation. This determiii:~tioir was accomplished by measuring the intensity of the 3.98-mi(*rori deuterium oxide band in the infrared absorption spectrum of ii sample diluted with water so that the deuterium oside content was about 3%. The deuterium oxide content of the diluted swnple was read from a calibration curve prepared by determining the absorbances of synthetic blends of water and 99.8% deuterium oxide. The deuterium oxide content of the sample before dilution was then simply calculated from the weights of the sample before and after dilution and the measured deuterium oside content of the diluted sample. Samples for calibration :tilt1 for :inalysis were prepared in a 10ml. glass-stoppered graduatcd rylirider or mixing bottle. About 0 . 1 to 0 . 5 gram of 99.Syo deuterium oxide or of unknown was transferred to the tared cylinder i)y means of a medicine dropper, care being taken to avoid prolonged exposure of the material to atmospheric moisture, and iveighed to the nearest 0.1 mg. Enough distilled water was addrri to bring the deuterium oxide content of the mixture to O.to 5% (preferably shout 3% for unknowns) and its &ght was obtained. The contents of the cylinder were then mixed, first by rotating the cylinder i n an inclined position so as to avoid contact with the ground-glass surfaces, and finally by inversion of the cylinder a number of times. About 0.5 gram of diluted sample is required for use in filling the infrared absorption cell. The absorbance of the diluted sample in a calcium fluoride cell, 0.043 mm. thick, was determined with a Perkin-Elmer hlodel 12B infrared spectrometer with a sodium chloride prism, GAT amplifier, and Speedomax G recorder. With the wave-length drive set at 3.98 microns and the recorder in motion, a record was made by
the "shutter in-shutter out" technique, from nliich the xt)sorbance was computed in the customai~y nia~iner. Tlie :it)sorharice of water in the same cell was redcterminctl at the time of each analysis as a check of the calibration, and, if neces%:iry,B correction was applied to the absorbance of the unkiioivn. The time required for each determination \vas approsiin:~tt~ly 10 minutes. Figure 1 shows the c:ilihration curve obtained 1"rorii these data, it w:is cstiniatcd that 0 to 3%
iii
tiiis work.
dcsutoi,iuiii
osiilca
Figure 1. Calibration Curve for Infrared Spectrometric Determination of Deuterium Oxide in Water
V O L U M E 2 2 , NO. 5, M A Y 1 9 5 0
691
in water can be determined in this way with a precision of 0.03 percentage unit, and that larger concentrations can be determined with a precision of 0.03 to 1 percentage unit, depending on the dilution necessary to bring the deuterium oxide content to within the measurable range (about 3% for greatest Precision Of measurement of the absorbance in a 0.043-mm. cell).
LITERATURE CITED
(1) Halford, J. o., Anderson, L. C., Bates, J. R., and Swisher, R. D.. J . Am. Chem. Soc., 57, 1663 (1935). (2) McMurry, H. L., Thornton, V., and Condon, F. E., J . Chern. P h y s . , 17, 918 (1949). R~~~~~~~
November 18, 1949.
Report 774-49R, Phillips
Petroleum
Company.
Ferrate Oxidimet ry Oxidation of Arsenite with Potassium Ferrate(VI) J. M. SCHREYER, G. W. THOMPSON1, A N D L. T. OCKERMAN University of K e n t u c k y , Lexington, K y . Methods of analyzing potassium ferrate(V1) are described. The methods are based upon the oxidizing property of the ferrate(V1) ion and the determination of the total iron present in the compound.
T
H E availability of potassium ferrate(V1) in a state of high purity ( 4 ) and its potentiality as a new and powerful oxidizing agent indicated the necessity of developing methods of analysis for the compound. The impurities usually present in samples of potassium ferrate (VI) are potassium chloride and hydrous ferric oxide. For the analysis of impure as well as highly purified samples, use of the oxidizing property of the ferrate(VI) ion seemed the most logical approach for the development of suitable methods. De Mollins (3) reported the quantitative reduction of the ferrate(V1) ion by means of the iodide ion in acid solution. 2FeOa--
The alkalinities of the solutions were varied between approximately neutral and 10 molar in sodium hydroxide. As shown in Figure 1, higher percentages of potassium ferrate(V1) were found in those determinations conducted in the more alkaline arsenite solutions. One of the difficulties encountered in the use of potassium bromate for the standardization of highly alkaline arsenite solutions was the fact that the concentration of hydrochloric acid a t the end point was found to be critical. Reproducible results in standardization of alkaline arsenite solutions were obtained only if the concentration of hydrochloric acid was 1.5 to 2.0 N.
+ 81- + 16H+--+-2Fe++ + 8H20 + 412
In view of the expected reduction of any ferric ions present as
an impurity and the instability of the ferrate(VI) ion in acid solution, it appeared impossible that this reaction would be a quantitative measure of the ferrate(V1) ion present in a sample. The method developed in this laboratory makes use of the increased stability of the ferrate(V1) ion in strongly alkaline solution and is based upon the reduction of the ferrate(V1) ion to ferric ion in alkaline arsenite solution. A weighed sample of potassium ferrate(V1) is added to a standard alkaline arsenite solution containing a quantity of arsenite in excess of that required for the reduction of the ferrate(V1) ion. The excess arsenite is back-titrated with standard bromate or standard cerate solution. The following equation represents the chemical reaction upon which the method is based: 2Fe04--
+ 3AsOa--- + 11H20----t 2Fe(OH)a(H20)3
+ 3kSO4--- + 4 0 H -
For confirmation by an independent method, the samples under investigation are analyzed to determine the total iron present in the compound. The hydrous ferric oxide impurity is removed by solution of the sample in sodium hydroxide solution and subsequent filtration. After the filtrate is acidified, the ferric ions are reduced to the ferrous state and titrated with a standard cerate solution. DEVELOPMENT OF METHODS
Preliminary studies showed increased stability of the ferrate (VI) ion in alkaline solutions. Investigations were undertaken to determine the optimum alkalinities of arsenite solutions used in the determination of potassium ferrate(VI) in a sample. 1
Present address, 2632 Dominguez St., Long Beach, Calif.
2650
I
'
I O
'
2 0
' 30
4D
5 0
60
70
BO
90
100
I I1
N Q 0 H ( M oI e s/ LIt er )
Figure 1. Average Per Cent Potassium Ferrate us. Molarity of Sodium Hydroxide
Smith (5) found that Gyory's method ( I ) for the determination of arsenite with bromate need not be conducted in heated solutions. He reported that if methyl orange was used as an indicator in the titration conducted a t room temperature, the response to the first excess drop of oxidant required a t least 30 seconds. Gyory's original method seemed preferable because the solutions were already hot as a result of the neutralization of the strongly alkaline arsenite solutions that were employed. No difficulty was encountered in standardizing strongly alkaline arsenite solutions with standard cerate solutions. In the investigation of the method based on the determination of total iron present in a sample, results were obtained which were not in agreement with those obtained by the arsenite methods. The error was found to be caused by the presence of certain reducible substances in the caustic solution used to dissolve the potassium ferrate(V1). Heinemann and Rohn (2) have reported