NOV., 1959
FRACTIONATION OF OXYGEN ISOTOPES BETWEEN WATER AND SULFUR
incompletely filled d bands. On this basis an associative mechanism is suggested rather than a dissociative mechanism of exchange.
DIOXIDE
1885
Acknowledgment.-The authors wish to thank R. P. Eischens, L. C. Roess and Irving Wender for their interest and suggestions.
THE FRACTIONATION OF OXYGEN ISOTOPES BETWEEN WATER AND SULFUR DIOXIDE' BY L. L. BROWNAND J. S. DRURY Chemistry Division, Oak Ridge National Laboratory, Oak Ridge, Tenlbessee Received N a y 8 , 1060
The isotopic fractionation of oxygen between water and sulfur dioxide was measured at 21". The experimental value (1.016) agreed well with a theoretical value calculated from Urey's data (1.014). Oxygen-18 concentrated in the gas phase. a vacuum stopcock and a standard taper joint to a manifold for evacuation and filling. Diametrically opposite the filling port was a second vacuum stopcock which shut off a 20-ml. bulb from the main reaction volume. The entire container was evacuated.and 0.998 g. of HzOIB(about 1.10 X normal) was distilled into the small bulb where it was isolated. Then 2.443 g. of SO2 was added to the large flask. Exact weights of reagents were determined by weighing the tared From Urey's data2 the factor may be calculated to flask after each addition of reagent. The HzOwas transferred to the larger container by turning the flask on its side in be 1.012 a t 25". Halperin and Taube3showed that an evaporating dish of liquid nitrogen and opening the conthe exchange of oxygen between liquid water and necting stopcock to the small bulb. This action froze both SO2 was fast. Hoering and Kennedy4 also ob- the SO2 and HzOon the wall of the large flask and voided the served that the exchange was rapid and called small bulb which was isolated by turning the stopcock to its attention to the unsuitability of SO2 for Oi8 analyses closed position. The flask was put in a room thermostated t 21' for 48 hours, during which interval it was occasiona!ly in mass spectrometers. A modification of the ashaken to break up the liquid phase. At the end of the equihmethod of Gregg and Lauder5 for reduction of SO2 bration time, the gas and liquid phases were separated by to water has provided a practical way to assay SO2 draining the water into the small bulb and closing the stopfor Oi8, and made feasible the determination of cock. With the phases thus separated, the SO2 was subjected to reduction as described below. Later the HzO was disthe single stage fractionation factor (cy) for oxygen tilled out of the small bulb and portions were pipetted into isotopes between SO2and liquid H20. small bulbs for analysis by equilibration with COZ. Equilibration B.-The reaction vessel for the second equiExperimental libration was a 500-ml. round flask to which a three-way Materials.-The Matheson Company purified H2S and stopcock was attached. The two external arms of the stopanhydrous SO2 were used after being purified further by two cock were fitted with standard taper joints; one connected vacuum distillations in which only the center fractions were the apparatus to the vacuum manifold, the other accepted retained. the product water receiver. The flask was evacuated and Laboratory-distilled water was vacuum distilled into the 0.466 g. of normal water and 1.245 g. of sulfur dioxide were reaction vessels following several cycles of freezing and pump- added directly to the reaction flask by cooling with liquid ing to remove dissolved gases. nitrogen. The exact amounts of reagents were obtained by Water, containing 1.5% 0 1 8 , was obtained from Stuart weighing after each addition. The equilibration was conOxygen Company. tinued with occasional shoaking of the flask for 44 hours in a Methods.-The sulfur dioxide and water were equilibra- room thermostated at 21 Sampling of the liquid was done ted by two different procedures in order to minimize syste- by collecting the water above the stopcock and diverting I t matic errors. In one equilibration, oxygen-18 enriched into an evacuated bulb of 5-ml. capacity. The water was water was used so that a relatively large isotopic change removed by pipet into another bulb for the Cot equilibrawould occur in the SO2. I n the other equilibration, the iso- tion. The SO2 was retained for further processing as detopic composition of both reactants was normal. The pres- scribed below. sure of SO2 in each case was approximately one atmosphere. Analyses of SO2 for O18.--Gregg and Lauder6 reduced SO2 The isotopic composition of all water samples was deter- to HzO by using H2S with catalytic amounts of water. mined indirectly from the mass analysis of COZwith which Their isotopic measurements were done by density deterthe water samples were equilibrated previously. In the first minations of the product water. I n the present instance, experiment average peak heights of masses 44 and 46 were greater precision and convenience were obtained by analyzobtained with a modified General Electric 60' sector, 6 in. ing the water samples by means of a mass spectrometer. radius spectrometer. In the second experiment measure- This was done indirectly by equilibrating the water with ments of the 46/44 mass ratio were obtained directly with a COZand calculating the isotopic composition of the water dual collector, ratio spectrometer constructed by Nuclide from that of COZ. The latter method requires a weighab!e Analysis Associates. The SO2 samples from each experi- amount of water and necessitates reducing more SO2 than IS ment were reduced to HzO with HzS and then were treated required for the ordinary mass spectrometer sample. The as described above. reduction reaction is Equilibration A.-The first equilibration was made in a 1 SO2 2H2S ---f 2H20 35 liter, round-bottom flask. The flask was connected through The stoichiometry requires a 2/1 mole ratio of HzS/S08 and (1) This paper is based on work performed for the U. S. Atomic it was found that a 2.2/1 ratio gave 1 0 0 ~conversion. o The Energy Commission by Union Carbide Corporation. reduction would not go if both reagents were anhydrous (2) H. C. Urey, J. Chem. Soc., 562 (1947). even with prolonged heating at 500". MatthewsB studied (3) J. Halperin and H . Taube, J . A m . Chem. SOC.,74, 375 (1952). various inorganic and organic solvent catalysts which pro(4) T. C. Hoering and J. W. Kennedy, ibid., 79, 57 (1957).
Introduction The literature has no experimental values of the fractionation factor for oxygen isotopes for the reaction SO1601fl(g)+ HzO'S(1) I_ S0'6O1a(g) + HzO'B(1)
.
I
+
+
(6)E. M. C. Gregg and I. Lauder, Trans. Faraday Soo., 46, 1039 (1960).
(6) E. Matthews, J. Chem. Soc., 2270 (1926).
L. L. BROWN AND J. S. DRURY
1886
ro
VPCUUU
MPNiFOLO
v
Fig. 1.-Sample
preparation manifold.
mote (or prevent) the reaction; from his data, water was selected as the least likely to yield spurious side products. The manifold for preparing samples is shown in Fig. 1. The total volume confined inside the stopcocks was approximately 90 ml. The SOz sample pressure was determined first, then the gas was condensed in a reduction bulb of 100ml. size. SO2 at a pressure of about 50 cm. yielded a satisfactory amount of water. Hydrogen sulfide now was measured into the evacuated manifold in two increments, each of which was about 1.1 times the original SO2 pressure. The catalytic water reservoir was kept in a bath a t 15'. An indefinite increment of water vapor was added to the reaction vessel by a rapid 180' turn of the confining stopcock. At most 1 mg. (I00 ml., 13 mm.) of water could be added, and likely much less, since the water reservoir had only 75 ml. of gas instantaneously available above 25 ml. of water. The catalytic water was normal in 018, and no corrections were made for its presence in the sample. The amount of water added was small and nearly equal for every sample. Its presence would not affect the ratio of isotopic ratios in a pair of samples. The loaded reduction tube was sealed off with the contents at liquid nitrogen temperature. It then was allowed to warm to room temperature. As it warmed most of the SO2 was reduced, leaving a coat of sulfur on the wall of the bulb. At this point water equivalent to 75% conversion could be recovered; the remainder of the oxygen probably was present in the sulfur as SO2 or HzO. A 100% recovery of water was obtained by heating the flask overmght at 350'. The excess HsS was removed by repeatedly freezing the sample in a Dry Ice-acetone bath and pumping off the gas abovq; the solid. The purified water was distilled into a tared receiver, then a known amount of COZwas added and the weight of water was determined. After equilibrating the mixture of water and COafor three days at 25O, the isotopic composition of SO? was calculated from analysis of the COZ using the equation'
where
Neb. Re
%
= atom fraction
0 ' 8 in sample C01SO*6/C0216 for the sample ( R o = reference COZ) = COZ/H20 mole ratio in sample equilibration = 2.076 at 25'
=
Results The isotopic data from both equilibrations are summarized in Table I. Data from equilibration A were obtained in the usual manner from measure(7) I. Dostrovsky and F. 8. Klein, Anal. Cheham., 24, 414 (1952).
Vol. 63
ments of mass 44 and 46 peak heights. Data from equilibration B were arrived at differently since the results of the mass analysis were of the form R = (46/44)sample/(46/44)standard. The 0l8/O16 for the standard was taken as 0.00204. A calculation showed that a 4.1% increase in this ratio would cause the separation factor to appear only 0.0004 greater. Since this quantity is less than the confidence interval of the assays, it may be seen that the experimental a is not strongly dependent on the concentration chosen for 01* in the standard COz. The average value of the single stage separation is 1.016 factor [a = (01S/01~)S02/(01~/016)H20] and is probably correct to f 0.003. The uncertainty in the analyses of the samples from the second equilibration is known to be less than 0.003, and the agreement between the results of the two experiments makes that value a reasonable limit. The difference in the average experimental value, and the value calculated from Urey's2 data extrapolated t o 21" is less than the uncertainties in the experimental measurements. For both equilibrations the experimental a is greater than the calculated. The heavy isotope of oxygen enriches in the gas phase. TABLE I MASSASSAYDATAAND SEPARATION FACTORS Equilibration
A
Rn HzO-1 223.7 Hz0-2 223.1
Sample
HzOb (g.)
0"/01'
ac
0.0751 0.002159 .ON2 (av.) 1.0157
B
SO2 COZ HzO
219.9 .lo72 238.1 (ref. value) 1.0203 .2762 i 0.0003 (95% C.I.)
,0020059
SO2
1.0365 i 0.0001 (95% C.I.)
.5050
.0020377
.002193
1.0159
cos .00204d For equilibration A, R = COz1~/CO16018. For equilibration B, R = (46/44) sample/(46/44) standard. Weight of = water equilibrated with 1.228 X lo-* mole COZ. ( O18/0l6)SO~/( 018/016)H20. d Assumed value for standard COa * a
A chemical exchange reaction having a separation factor as large as that observed in this work is of interest as a possible method of separating isotopes. If the SO2-H20 reaction were used to separate oxygen-18, the critical reflux would be the reduction of SO2 to HzO since the desired isotope is concentrated in this species. SO2 reduction might be accomplished by the reaction already described for isotopic sample preparation. Similar work has been done on a large scale by NishimorP to remove HzSfrom a gas stream. (8) T. Nishimori, Japan Patent 2362 (1955); C. A . , 51, 12453 (1957).