solving problems by isotope abundance measurements
SOLVING PROBLEMS BY ISOTOPE ABUNDANCE MEASUREMENTS
Etienne ROTH
103 rue BRANCAS, F- 92310, Sèvres France
e-mail:
Abstract
A variety of problems studied at Saclay with the use of deuterium, and isotopes of sulphur and rare earth, are explained in their historical context. They illustrate how techniques developed to solve applied problems were used to foster cognitive research; and conversely how analytical systems, set up for pure research purposes, provided data enabling to build plants that detritiate tons of heavy water. The role of instrumentation in authorising the discovery of disulphur monoxide, or to facilitate medical applications is underscored. In conclusion an expansion of applications of isotope abundance measurements is foreseen, and publication policies improving the possibility of comparisons of results from different laboratories are suggested.
Introduction
The discovery of stable isotopes, made in 1910 by J.J.Thompson, was not followed by applications before the 1930ies after the discovery of deuterium by H.C. Urey, and heavy water preparation. The latter was used first as tracer in human physiology. But it took less than six years for artificial radioactivity, discovered in 1934 by Irène and Frédéric Joliot-Curie, to be used in research.
Experimentation with isotopes was dependent, either on the possibility to procure spikes having a stable isotope composition different from that of the system to investigate, or of production of radioactive tracers by irradiation in accelerators , or later in nuclear reactors.
The gradual discovery of natural isotope fractionations, in nearly fifty elements having two or more isotopes, opened specific fields of applications to stable isotope measurements in addition to providing sources of natural spikes. This field is still expending because all ranges of natural isotope composition have not yet been tied to mechanisms helpful in problem solving. For practical purposes isotopes with half lives of the order of, or longer than, the estimated age of the universe, i.e. 15 Ga, are considered as stable. One may add two uranium isotopes to them because their half lives, though smaller, are long enough.
A largely unexplored field opened when non mass dependent isotope effects were observed in phenomena in the upper atmosphere. More recently similar occurrences
have been found in sulphur minerals. Considering that all elements of even atomic number, from oxygen up, possess a minimum of three stable isotopes, one may anticipate that studies of non mass dependent isotope effects will develop and may reveal new mechanisms of natural isotope fractionation.
We restrict this account of problems solved by the use of isotope abundance measurements to problems studied at Saclay. We describe first early work, at Saclay, which started with problems relative to deuterium concentration and sulphur isotope determinations.
We do not report isotope abundance measurements in dating nor in isotope dilution analysis, though it enabled for instance to date the Oklo phenomenon, or with deuterium, to measure a solubility as small as that of water in carbon tetrachloride.
Deuterium abundance
Earlier deuterium natural abundance measurements
Variations in deuterium content of natural waters
They have been found, even before World War two, by methods relying on density measurements, as delicate to master as the falling drop technique. Though mass spectrometry was the method of isotope analysis, in the fifties very few laboratories could make reliable measurements of deuterium by mass spectrometry. Either mass spectrometers could not not work in the very low mass range, or, if they could, the distance between mass two (H2) and mass three (HD) was so large that mass discrimination prevented ratios of peak heights to be representative of abundance ratios
In Saclay there was an apparatus dedicated to mass three to mass two determinations, that was built in H.G. Thode’s laboratory, from a prototype inherited from A. O. Nier. It consisted of a small tube equipped with two ion collectors, enclosed in an all glass envelop in which a small permanent magnet provided the deflecting field.
Hydrogen gas, not water could be admitted into the ion source. When samples are under water form, in early days deuterium was analysed by equilibrating water with hydrogen at a well controlled temperature and using a platinum oxide as catalyst in the water phase. The equilibrium constant was assumed from the literature. The preferred technique however soon became reducing water to hydrogen, initially over zinc, and later over uranium metal turnings, following a course advocated by
J. Bigeleisen.
When other molecules were investigated their hydrogen had to be converted to water. Several pitfalls had to be averted. To study deuterium in H2S exchange, H2S had to be burnt. An exchange with water before combustion would modify drastically the measured deuterium content, because this exchange is very fast and its equilibrium ratio, under working conditions, was always bigger than two. When analysing methane, errors due to incomplete catalytic conversion had to be avoided.
Work on D concentration in rivers
It started around 1955 with the purpose of choosing the best source of water for heavy water production. A location close to the natural gas wells of Lacq was considered because gas from Lacq contains about fifteen percent hydrogen sulphide. Advantage of this particularity was first taken to build a pilot plant to study [1]production of heavy water by the hydrogen sulphide water isotope exchange, saving the labour of preparing hydrogen sulphide in large quantities, and perhaps later a production plant might follow.
Lacq is located close to the Pyrénées. Investigations soon revealed that streams flowing down from hills had a deuterium content in small but characteristic excess over that of the main river originating from the mountains. An explanation was soon put forward: rains falling on the hills would deplete clouds from deuterium before they would deposit snow on the mountains where the river had its source.
In addition to survey of waters, however, we had, for the pilot plant operation, to be able to analyse deuterium in hydrogen sulphide. In the natural gas, collected at the exit of the pipes, its deuterium content was 86 atoms per million. As very small quantities of water are carried to the surface in the gas, the deuterium content of hydrogen sulphide corresponds to isotope equilibrium with water at the bottom of the well. Estimating temperature at the bottom of the well to be around 140°C assigns a deuterium content of this water at 152 atoms per million, a content close to that of local surface water. Reciprocally temperature at the bottom of the well can be estimated by the same reasoning when one considers that water in rocks forming the gas reservoir was in large quantity with respect to hydrogen sulphide, so that exchange did not modify its isotope composition supposed to be 150 ppm of deuterium[2].
It is worthwhile mentioning that to study isotope exchange reactions, absolute isotope abundance must be measured. Relative deuterium contents of stream and rivers could have been established using relative values, expressed by deltas, but temperature determinations call for absolute figures.
The same can be said of the requirements for any thermometer, e.g. that of the well known 18O one developed by Epstein and Urey.
Deuterium measurements at higher concentrations
Heavy water analysis
Mass spectrometric analysis of water concentrated in deuterium is inevitably affected by contamination by background, whichever procedure is used. It introduces uncertainties that check endeavours to use mass spectroscopy.
Fortunately in the near infrared region of the water spectrum, there is an interval where absorption is due to HDO[3]. Water is analysed in the liquid form, and contamination from absorption cells that can be dried is not observed. This enables to measure D concentrations of heavy water safely and fast up to around 99.9 percent. Infrared analysis of heavy water was performed first with this technique
when concentration was checked on every drum during filling the tank of the first nuclear reactor at Saclay, late in 1952.
Medical application
This easy analytical procedure was attractive to workers outside Saclay.. In particular medical doctors interested in different body water pools. The most sophisticated application was to measure extravascular lung water[4]. When injecting a few cc. of D2O in a vein , they mix with lung water by diffusion within the lungs. The transit time to an artery of an induced deuterium signal in blood depends on the water volume with which this spike mixes. This volume is calculated by comparison with the transit time of the non diffusible indocyanide green. A special infrared machine was designed that enabled to work “on line” on a patient’ bedside. A by product of the technique is to give a value of the blood flow to the heart.
Sulphur isotope abundance measurements
Work on sulphur samples from volcanic origin
This work was carried out in parallel with the early deuterium work. A problem was whether they were anomalies in sulphurs collected from volcano eruptions. The ground to start this research, in 1949, was to check an hypothesis put forward by a volcanologist, Dr. Noetzlin. In March 1939, two months only after the discovery of fission, he had made the supposition that energy liberated by fission might be the driving force of volcanic eruptions[5]. During the intervening years he wrote several papers developing this hypothesis, and made calculations resting on what became known, little by little, of the theory of chain reactions. He thus decided to look for isotopes that might have captured neutrons, and wanted to investigate sulphur isotopes because sulphur samples could be obtained from many volcanic sites and would lend themselves easily to generate SO2 for mass spectrometric analysis. This, of course was not a choice resting on optimum nuclear reaction considerations! However Saclay could measure sulphur isotope ratios because there was at Saclay a mass spectrometer of the same type as the one with which H.G. Thode had measured natural variations of sulphur isotope abundance as soon as 1947
Half a dozen volcanic samples were analysed for 34S/32S ratios[6]. Little or no variation between samples from volcanoes of different parts of the world was found, and no difference either with native sulphur from non volcanic origin. A comprehensive review of isotope abundances of sulphur is found in Coplen’s report (ref.1). One may compare these findings with the evidence given in ref.2 that 34S abundance in Lacq sulphur was significantly different, though by a small amount, from what it was in neighbouring sulphuretted water sources.
S2O, an interesting side result of this research.
Peaks in the background spectrum of the mass spectrometer were found around mass 8O. A first interpretation was to assign them to a small amount of SO3. But peaks at mass 82 were much too big to be due to 34SO3+, but they could originate from 34S in S2O+ ions. One was to find whether such ions were due to reactions within the mass spectrometer, or belonged to bona fide molecules of a sulphur oxide. The only oxide described in the literature that could generate those ions would have been the dubious (SO)n.
A series of experiments showed that an S2O molecule could be prepared. Its stability region and thermodynamic properties were established[7]. We gave it the name “Sulphur hemioxide”.It is surprising that this molecule was not discovered sooner because it is formed when burning sulphur under a low oxygen pressure, or alternatively under action of an electric discharge. When condensed in a dry ice cold trap, it gives a red precipitate instead of the white one of sulphur dioxide. Ignoring preceding publications, the same molecule has been described under the name of disulphur monoxide from its raman spectrum in 1975[8]. It may be pointed out that working with instruments that are equipped with programs that limit exploration of masses to the “most interesting” ones would have prevented making this discovery.
Later deuterium work at natural concentration, at Saclay,
Such work has mainly been done on samples fractionated during transition from condensed to vapour phase, or the reverse, and been compared to, or associated with, 18O and tritium measurements
Exploration of ice caps
When it became known that Saclay could routinely analyse deuterium in water at its natural concentrations, it started requests for measurements from several parts. Under impulse from a Swiss teacher, Dr. Renaud, the french explorer Paul Emile Victor proposed to bring samples from ice caps to make isotope determinations. This would be parallel to work done in Chicago by Sam Epstein, to estimate annual precipitation, relying on seasonal variations of 18O in ice. The aim was essentially of geographical interest. In Europe W. Dansgaard studied Greenland. and equipped his laboratory for 18O abundance measurements.
Saclay could not look routinely at 18O, and it was decided to analyse deuterium.
Seasonal variations showed up nicely in ice cores; fear that diffusion would rapidly erase seasonal variations of D contents were alleviated. It was nevertheless necessary to prove that determinations of deuterium and of 18O, reputedly less diffusible in ice, gave the same results. For this purpose, we analysed deuterium in samples from an iceberg, in collaboration with Dansgaard, who did the 18O determinations. The linear relationship found between D and 18O values. proved the two techniques to be equivalent[9].