A Tandem Van de Graaff accelerator has been modified for use in the direct measurement of natural abundances of 10Be and 14C. A description of the system is given and some 10Be results on oceanographic samples are discussed.

The Chalk River Tandem Accelerator Mass Spectrometry System has reached a state of reliable measurement of 14C using 2 to 5mg elemental carbon prepared by Mg reduction of CO2. For two comparisons of a near-modern unknown with the NBS oxalic acid standard we obtain a total error of ∼±4.5%, consisting of a random system error of about ±3.5% combined with the statistical counting error. Measurements have been made on 70 samples in 30 days of running time during the past year. Samples included deep rock carbonates, cosmogenic 14C in meteorites, charcoal from earthquake fault zones, collagen of bone artifacts and fossil beetle-fragments.

Our accelerator mass spectrometry (AMS) system shows a one-to-one relationship between sample 14C concentrations determined by AMS - and by β-counting. Measurements of unknown samples against a standard indicate that 14C concentration measurements to better than 2% can be made. For a 30-second data collection interval in a typical run of 100 intervals, the variability of the beam injected into the accelerator is ca 2%, that of the machine transmission is ca 4%, and counting statistics give 4.7% standard deviation for a sample of 80% of modern carbon.

Tests of performance of the tandem accelerator mass spectrometer at the NSF Regional Facility at the University of Arizona are discussed. Results of measurements on some tree rings and on some archaeologic samples are presented.

The Oxford 14C accelerator has operated with beam for some 400 hours. This report describes the progress made towards achieving dates from milligram samples with the required accuracy of better than 2%. In summary, it shows how 14C is relatively easily detected, but that the overall beam optical system is, at present, rather sensitive to effects which prevent reliable maintenance of the necessary isotope ratio stability. These effects can probably be eliminated by careful attention to details of the design rather than by major modifications.

In 1979 when accelerator mass spectrometry was started in the Netherlands, Gove (1978) and Litherland (1980) demonstrated that a tandem accelerator is eminently suited for this work. Isotope ratios below 10−15 could be determined and applications were stimulated on numerous fields of research. 14C dating in particular would become an important application because smaller samples could be employed and more complex dating problems could be attacked. However, the 5% accuracy reached at that time had to be improved towards the 0.5% obtained with the conventional method of counting decays. Purser and Hanley (1978) pointed out that this high accuracy could be reached in facilities especially dedicated to dating. Rapid switching between measurements on different isotopes turned out to be crucial (Suter et al, 1981). We decided to proceed in two phases. In the first phase, the existing EN-tandem facility would be used as before and modified for the detection of 14C ions. From measurements with 14C ions from different samples, information would be obtained about the limitations of the existing facility. In the second phase, this information would be used to develop a facility in which accurate measurements of isotope ratios can be performed for elements throughout the periodic system. Conventional 14C dating will be extended in collaboration with the Isotope Physics Laboratory in Groningen. Also, applications with other elements will be studied. The improved facility is scheduled for operation in the spring of 1983.

During the last three years the ETH-EN-tandem accelerator facility has been adapted for the quantitative determination of the rare isotopes 14C and 10Be in mg samples. The goal of this project is to routinely achieve a 1% accuracy when measuring 14C/12C ratios with a minimum expenditure of human resources and beam time. The concept is similar to that proposed by Purser and Henley (1978). The early evolution of this dating facility was described previously (Suter et al, 1981a,b). This paper is a brief report on the current status of the system and its development.

During the past year we have continued to work toward greater stability and flexibility in nearly all elements of our accelerator mass spectrometry (AMS) system, which is based upon an FN tandem Van de Graaff accelerator, and have carried out measurements of 14C/12C and 10Be/9Be isotopic abundance ratios in natural samples. The principal recent developments and improvements in the accelerator system and in our sample preparation techniques for carbon and beryllium are discussed, and the results of a study of 10Be cross-contamination of beryllium samples in the sputter ion source are presented.

Micro-ampere beams of C– have been produced with the Aarhus negative-ion source, operated on CO2 gas. The efficiency of the source and memory effects have been measured, using 13C-enriched CO2, in order to evaluate the applicability of the source to 14C dating by accelerator mass spectroscopy.

Chemical pretreatment procedures for the decontamination, extraction, and isolation of organic materials for 14C dating using the Oxford accelerator system are described. Specific details are given for the isolation and chromatographic purification of amino acids from bone and tooth collagen, of lipids from sediments, and of cellulose and glucose from wood, paper, and textiles. A description is also given of the apparatus used for the routine preparation of 1 to 5mg graphite samples on tantalum wire, for use in the accelerator ion source.

The high energy mass spectrometer (HEMS) approach to 14C dating allows the use of very small samples in the low milligram range. Sample pretreatment and decontamination procedures can be both more vigorous and more selective than those used by conventional dating laboratories. Specific chemical compounds can be isolated from archaeologic or geologic samples; such compounds may be characteristic of particular source materials and, hence, provide more detailed information than is generally possible using bulk organic samples. The Oxford Radiocarbon Unit has concentrated on three sample types that represent the kind of material we expect to work on initially: bone, lake sediment, and wood.

We present a survey of carbon beam yields from 20 simple carbon compounds using a caesium sputter source and the McMaster University tandem accelerator. The carbon yield was measured as a 35MeV 12C4+ beam. We found that the beam intensities could be related to a grouping of the carbides according to the chemical bonding of the compounds. The usefulness of the compounds for accelerator 14C dating was further related to their preparation chemistries. Strontium carbide was the equal of graphite in negative carbon ion beam production and aluminum carbide was found to be a good candidate for further tests because of its good sputter yield and preparation chemistry. Charcoal was also tested with varying amounts of silver added as a heat conduction aid.

A program of accelerator mass spectrometry has been started at the Rehovot 14UD Pelletron Accelerator Laboratory. Part of the initial emphasis has been directed to the detection of the 36Cl radioisotope. We report here on the present status of our work and describe our experimental system. Preliminary results are presented, showing that 36Cl/Cl concentrations ranging down to 1×10−14 could be measured with our system.

The international radiocarbon community met on June 25, 1982. Two reports were presented, and several measures related to radiocarbon dating discussed and adopted.