Determination of Deuterium in Water A M a s s Spectrometric Method R O B E R T B. F I S C H E R ' , R I C H A R D 4. P O T T E R * , AND ROGER J. \OSKLYL'
T h e Argonne .Vational Laporatory, Chicago, I l l . A procedure is given for the determination of deuterium in water of any isotopic composition. The procedure is simple and rapid when the equipment is available. The method does not, in general, require any removal of dissolsed impurities as would be the case for a specific gravity method. The water sample may be recovered practically unchanged after the analysis is completed.
SI"':":
its discovery in 1932 by Lrey, Brickwedde, and Vurphy ( I 2 ) ,deuterium, the mass two isotope of hydrogen, has found much use in scient,ific work in the form of heavy vater. Biological tracer studies using deut,erium have revealed useful information. Several organic reaction mechanisms have been worked out using heavy hydrogen ( 6 ) . lfore recently heavy wat,er has come into prominence in the field of atomic energy as a moderator-i.e., as a slowing down agent for neutrons ( I O ) . Any physical or chemical property of light and heavy hydrogen or vater may be used as the basis of a method for the analytical determination of deuterium, provided that the difference in that property between the tivo isotopes and the experimental ability of measurement are sufficient to provide t,he desired degree of precision and accuracy. Specific gravity methods involving both float and pycnometer have been used Tyith good results. Refractive indices ( 2 ) , at,omic spectra (12), thermoconductivity of hydrogen gas ( 3 ) , thermoconductivity of water vapor ( I ) , infrared absorption ( 6 ) , freezing point ( 7 ) , TTiscosity ( 8 ) , and potentials of light' and heavy hydrogen electrodes (11) have all been used 15-ith varying degrees of success as the bases of methods for the determination of deuterium in water. 3Iass spectrometer methods cannot be performed directly upon water molecules because water vapor is readily adsorbed t'o an appreciable extent on the inner surfaces of the apparatus. Preparation of hydrogen gas from water makes mass spectrometer analysis satisfact,ory. Decomposition of ivater by electrolysis and by chemical reduction have proved unsatisfactory, however. A successful method of preparat,ion of a hydrogen gas corresponding to a given water sample involves the equilibration reaction between the hydrogen isotopes in the water specimen and in a hydrogen gas. The first equilibration was performed by Farkas ( 4 ) in 1936. The general met,hod was developed and used for the analysis of water specimens very near to normal deuterium abundance and very near to pure deuterium oxide early in the Manhat,tan Project. This paper ext,ends the method to one of general applicability for the determination of deuterium in water specimens of any deuterium-protium ratio.
steps: evaporation of the wtter, each type of liquid n-ater molecule (H20, HDO, and D 2 0 ) being in dynamic equilibrium rvith the corresponding Ivater vapor molecule, as well as in equilibrium n-ith each other; and dynamic equilibrium reactions b e t m e n the water vapor molecules and the hydrogen gas molecules (H2, HD, and D2). L4cat,alyst is required for the latter step, and platinum oxide, Adams' catalyst, has been found satisfactory. The derivat,ion must, be based upon all the equilibria involved
Any other possible combination between gas and/or water molecules is mathematically equivalent to some combination of these equilibria. Multiply Kz by K5, divide by K6, and transpose:
Multiply K3 by K6, divide by K7, and transpose:
A system of units for the liquid conceztrations must be established. Any consistent units are permissible; the units will be defined as mole per cent of the total-that is,
THEORETICAL
This method of determining deuterium in water involves an equilibration of the water sample with hydrogen gas and subsequent analysis on the mass spectrometer of the equilibrated gas. Inasmuch as the result of a water analysis must be calculated from an observed gas composition, i t is necessary to establish some mathematical basis upon which the water isotopic composition may be calculated from the experimental data. The equilibration reaction must proceed through two general 1 Present
[H*0(01
+ [D?O(l)] + [HDO(I)]
=
100
(10)
Substitute Equations 8 and 9 into 10; and solve for [HDO(I)]:
From Equations 8 and 9 viith 11,solve for [H?O(Z)]and [DzO(Z)] in terms of gas concentration:
address, Department of Chemistry, VniT ersity of Ill., Urbana,
Ill 2 Present address, Department of Chemistry, Rollins College, K l n t e r Park, Fla. 8 Present address, T h e a t o n College, Wheaton, Ill.
571
ANALYTICAL CHEMISTRY
572
Because the concentrations are already defined in mole per cent units, the mole % ’ D in the water may be expressed as: Mole % D in water = [D20(1)1
+
l/?
[HDO(l)]
(11)
Substitute Equations 11 and 13 into 14: Mole
70D in water
Kz = 3.62 K I = 3.11 K s = Pnzo Kt = P H D O K I = PD?O and using the relationship PD’o ~
PHDO
% D in water
=
(;Z)’’’ =
+
1.067
Exptl. hlole % D 48.4 26.1 77.3
37.7
36.0
+
Equation 16 may be used to calculate the isotopic composition o f the water from the experimentally determined composition of the equilibrated gas. The equation holds good for the reaction a t 25” C. for any composition of water. The gas molecule concentrations may, of course, be expressed in any consistent set of units; mole per cent of each gas species is probably the most convenient set of units to use. This equation could be simplified to eliminate one of the three types of gas molecules, from the calculation, yet it has becn found advisable to keep it as given above. EXPERIMENTAL
55.6
librium. Analyses were performed on two separate portions of each water standard, using tank hydrogen and tank deuterium, respectively, as the equilibrating gases. In this manner it was possible to approach the state of equilibrium from both directions for each water specimen. After the reaction had proceeded to equilibrium, the gas was analyzed on a mass spectrometer of the Xier type (9). In several cases the gas was reanalyzed after standing for a considerably longer period of time with the water specimen and catalyst to provide a check on whether or not the reaction had been a t equilibrium the first time. Corrections were applied to all gas analyses to eliminate error from dilution or concentration of the water by the gas during the reaction. The results of the analyses are given in Table I. LITERATURE CITED
IO55[Dz] 159.0 [HD] 10.55[D*] 0.8231 [Hz] (16) 3.179[HD]
+
Water Analyses
Calcd. Mole % D 49.4 24.9 78.9 57.5
=
Equation 15 is the equation for water isotopic composition in ’ terms of the equilibrated gas composition and the equilibria constants involved in the reaction. Insert the values for the constants into Equation 15 (all these values are correct at 25 C.) :
Mole
Table I. ’
,
A series of water standards of known isotopic composition was prepared by weight dilution methods, using normal distilled water and nearly pure deuterium oxide as the ingredients. For each analysis about 5 ml. of the water specimen, a few milligrams of platinum oxide catalyst, and an atmosphere of equilibrating gas were introduced into a small flask, about 30 ml. in volume. The flask and its contents were allowed to stand for about 1 hour, during vihich time the equilibration reaction proceeded to equi-
( 1 ) Clemo, G. R., and Swan, G. A , , J . Chem. Soc.,1942, 370.
(2) Crist, R. H., Jfurphy, G. XI., and Urey, H. C., J . Chem. Phys., 2, 112 (1934). (3) Farkas, A., “Light and Heavy Hydrogen,” pp. 134-6, London, Cambridge University Press, 1935. (4) Farkas, A., T r a n s . Faraday SOC.,32,413 (1936). ( 5 ) J . Amlied Phws.. 13. 526-69 (1942). (6) Kistiakowsky,-G. B:, and Tichenor, R. I., J. Am. Chem. SOC., 64, 2302 (1942). (7) Laliler, V. K., Eichelberger, TI-. C., and Urey, H. C., I b i d . , 56, 248 (1934). (8) Lewis, G . N., and MacDonald, R. T., Ibid., 55, 4730 (1933). (9) Sier, A. 0. C., Rev. Sci. Instruments, 11, 212 (1940). (10) Smyth, H. D., “Atomic Energy for Military Purposes,” U. S. War Department, 1945. (11) Topley, B., and Wynne-Jones, W. P. K., Xature, 134, 574 (1934). (12) Urey, H. C., Brickwedde, F. G., and Murphy, G. XI., P h y s . Ren., 39, 164, 864 (1932); 40, 1 (1932). KECEII-EI)September IO, 1947. Based on work performed under Contract No. W-31-109-eng-38 for the Atomic Energy P-oject a t the drgonne National Laboratory, and the information contained therein will appear in Division I11 of the National Nuclear Energy Series (Rfanhattan Project Technical Section). Although this work was completed in 1945, the results were n o t released for publication until Bugust 29, 1947. A t t h a t time, the intermediate numerical d a t a involved in t h e analyses were not available t o the authors.
Determination of Major Constituents of Cedar Oil Vapor in Cedar Chests F. W. H A Y W A R D AND R . B. S E Y M O U R Industrial Research Znstitute, University of Chattanooga, Chattanooga, Tenn. 8
A procedure is given for rapid colorimetric determination of cedrene and cedrol in the air of cedar chests. The method is based on a red-violet color formation resulting from the reaction of cedrene with vanillin in the presence of hydrochloric acid. The cedrol is dehydrated in situ by phosphoric acid and determined as cedrene.
c
EDAR chests are made from the heartwood of the eastern red cedar (Juniperus airginiana). The aroma in these chests is due t o the volatile oil which is present in cedar wood t o the extent of about 2%. Rabak (3) found the principal constituents of this oil t o be sesquiterpenes, a hydrocarbon, cedrene (C1~H~4), and an alcohol, cedrol (C15H260). I n this investigation a method has been developed for the determination of these volatile constitutents in cedar chests.
Semmler ( 6 ) , Ruzicka (j),and Treibs ( 7 ) determined in principle the structural formulas of these compounds. Cedrol was found to have a tertiary alcohol group and could be dehydrated with formic acid t o produce cedrene, which had a double bond. Rosenthaler (4)discovered that turpentine reacted with vanillin in hydrochloric acid t o give a green coloration. Bogatskif (1) applied this reaction t o a quantitative colorimetric method for turpentine and found that compounds containing active
