An informal collaborative group of radiocarbon dating laboratories, the Low-Level Club, has been established to measure the gamma radiation flux and to test the efficiency of the anticoincidence counting system in laboratories with a NaI detector unit. The detector will record gamma radiation from cosmogenic nuclides, muons and secondary γ radiation formed in the passive shield by charged cosmic-ray particles. We present here the first phase of this work.

In the “sublimation technique,” carbon dioxide entrapped in ice is recovered by sublimation, converted to graphite and ratio of 14C/13C in the CO2 determined by AMS measurements. We describe here several experiments performed to check the validity of such measurements and to study the effect of cosmogenically produced in-situ14C on the measurements.

We have performed a new determination of the half-life of 41Ca by measuring the specific activity of an enriched Ca material with known 41Ca abundance. We measured the activity via the 3.3-keV X-rays emitted in the electron capture decay of 41Ca, and the 41Ca abundance was measured by low-energy mass spectrometry. The result, t1/2 = (1.01 ± 0.10) × 105 yr, agrees with the recent ‘geological’ half-life of Klein et al., (1991), t1/2 = (1.03 ± 0.07) × 105 yr, and with the corrected value of Mabuchi et al. (1974), t1/2 = (1.13 ± 0.12) × 105 yr. We recommend the weighted mean of these three measurements, t1/2 = (1.04 ± 0.05) × 105 yr, as the most probable half-life of 41Ca. We also discuss the situation of the radioisotopes, 32Si, 44Ti, 79Se and 126Sn, whose half-lives, though still uncertain, are potentially interesting for future AMS studies and other applications.

A new facility for accelerator mass spectrometry has been established at Purdue University. First results have been obtained for 10Be and 36C1, and several internal research projects have been initiated. Plans are to become a national AMS facility to serve the Earth and planetary science communities for the full range of cosmogenic radionuclides.

A 14C measurement capability has been developed on the 14UD accelerator at the Australian National University. At present, this system operates on a medium-precision, low-throughput basis with slow cycling between isotopes. We describe unusual features of the system, and review preliminary experience with this mode of operation, in sample preparation, and with a recently installed injection system.

I present design details of a tandem accelerator mass spectrometer, which has been installed at the National Ocean Sciences AMS Facility at Woods Hole, Massachusetts, to provide precision 14C/13C/12C isotopic ratios for sub-milligram-size samples of graphite with throughputs of >4000 samples per year. A unique feature is the capability for simultaneous measurement of all three isotopes after acceleration, to avoid differential transmission effects and to allow on-line fractionation corrections and diagnosis of instrument health. Using filamentous graphite fabricated from a recent sample, we have established the counting rate of 14C ions at between 60–120 s–1.

Three years ago, funds were raised to equip the 3 MV Pelletron accelerator at the Department of Physics, Lund University for accelerator mass spectroscopy (AMS). We have modified the accelerator for mass spectroscopy by relocating focusing devices on both the low- and high-energy side of the accelerator and installing a Wien velocity filter and detectors for measuring the particle energy (E) and energy loss (ΔE). We have been working exclusively with 14C during the initial period. About 40 samples of elemental carbon have been produced, using Fe or Co as catalyst, during the last two years. The 12C− current from these samples is about 3–5 μA using an ANIS sputtering source. We are now planning 14C analyses in the fields of archaeology, Quaternary geology and radioecology.

We report on the present status of the Lawrence Livermore AMS spectrometer, including sample throughput and progress towards routine 1% measurement capability for 14C, first results on other isotopes and experience with a multisample high-intensity ion source.

Start-up performance and first results of the new Woods Hole Accelerator Mass Spectrometer are discussed. Special attention is given to the hemispherical ionizer sputter source and the triple-isotope injector design.

A computer program for convenient calibration of radiocarbon dates has been developed. The program has a simple user interface, which includes pull-down menus, windows and mouse support. All important information, such as calibration curves, probability density function and results, in text form, are displayed on the screen and easily can be rearranged by the user. Two versions of CalibETH, one for an IBM-PC and one for the Macintosh, are available. CalibETH runs under the graphics interface, GEM, from Digital Research, on an IBM PC.

The Groningen Radiocarbon Data base has been operating on an Apple IIe PC until recently. Both software and hardware of this data base made exchange with other systems very difficult. Therefore, the data base has been transferred to an IBM AT compatible computer. A new program has been written in Turbo-Pascal, using the Turbo-Pascal Database Toolbox. The program features versatile search routines. The output can be directed to various date-list formats, such as the International Radiocarbon Data Base (IRDB). The program is written in such a way that new date lists can be added with minor adjustments to the source code.

A new GPIB/IEEE-488 based data acquisition system has been built for the Groningen proportional counter setup, consisting of 11 counters. The IEEE bus is connected to an XT-compatible host PC. A versatile computer program controls the data entry; the same program can be used offline for final calculations.

As a follow-up to the meeting of experts convened at the International Atomic Energy Agency (IAEA) in February 1989, and the International 14C Workshop held in Glasgow in September 1989, the 14C Quality Assurance Program was formulated. In a joint effort of several radiocarbon teams and IAEA staff, we have prepared a set of five new intercomparison materials. These are natural materials frequently used by radiocarbon laboratories. The materials were distributed to 137 laboratories in May 1990. In February 1991, a meeting of experts was convened in Vienna to evaluate the results, to determine the radiocarbon activity of the five samples expressed in % Modern (pMC) terms and to define the 13C/12C ratio, and to make recommendations on further use of these materials. We present here the results of the exercise and the agreed consensus values for each of the five materials and discuss the different analyses that were undertaken.

The major findings of the Intercomparison Study (ICS) have already been published (Scott, Long & Kra 1990), but a number of questions remain unresolved. We address here some issues of user and technical relevance, which include: 1) further investigation of the quoted errors and their relation to the perceived precision and accuracy, which is of interest to users of 14C dates; 2) the analysis of the known-age wood samples provided in Stages 2 and 3 of the ICS; 3) an investigation of the corresponding δ13C data base, of more technical relevance to laboratories.

Following recommendations of the Glasgow International Workshop on Intercomparison of Radiocarbon Laboratories (Scott, Long & Kra 1990), a further international intercomparison is planned. This new intercomparison is complementary to the existing IAEA intercalibration, and will make use of natural samples whose ages will be unknown to the participants. The study has been funded by the UK Research Councils (SERC and NERC), and samples will be free to all participants. We anticipate that this intercomparison will be ongoing, with distribution of samples in 1992, and presentation of the results at a later meeting. We present here details of the samples available and the time scale of the study. Briefly, we envisage that the new study will be more focused than the ICS (Scott et al. 1986), and will include natural samples in both pretreated and unpretreated forms.

Some examples are given to show that the depth distribution curves of natural 14C and 13C of thin-layer sampled soil profiles can be used for inferring changes in soil organic matter and climate changes. By using a simple exchange model, we can determine whether decomposition products are fixed by clay or transported downward toward the groundwater table. We can also estimate the amount of the Greenhouse gases, CO2 and CH4, produced by the decomposition of the organic matter in terrestrial and paddy soils and emitted from the soil. A change from C3 to C4 plants, which might occur during a predicted temperature rise in some areas, thereby influencing the carbon balance, can be clearly detected by the δ13C depth profiles. A change in organic matter input can also be calculated under certain circumstances.

Soil organic matter sequesters close to three times the carbon existing totally in the living biomass and nearly the same for the total carbon in the atmosphere. Models, such as Jenkinson's or Parton's Century model, help to define soil organic matter fractions of different functions, based on residence time/14C age. Rejuvenation of soil carbon was felt to be the principal impediment to absolute soil dating, in addition to the ambiguity of the initiation point of soil formation and soil age. Recent studies, for example, of Becker-Heidmann (1989), indicate that a soil 14C age of >1000 yr cannot have >0.1% rejuvenation in the total soil organic matter compartments/fractions to be possible and sustainable. Always problematic in earlier observations were age vs. depth increases, in 14C profile curves showing an inflection of reduced age in the deepest samples, i.e., from the rim of the organic matter containing epipedon. We attribute this phenomenon, in mollic horizons, to earthworm casts in the terminal part of the escape tube. Becker-Heidmann (1989) has shown, in thin layer soil profile dating, a highly significant correlation between the highest 14C ages and the highest clay content. Thus, optimization of soil dating is, to a lesser degree, related to the applied extracting solvent system than to soil texture fractions. Such observations allow us to mitigate error ranges inherent in dating dynamic soil systems.

We present 14C AMS measurements and discuss the extraction procedure used on pollen extracted from peat samples. Microscopic examination of the extracts shows that the procedure is sufficient to remove most non-pollen materials and results in an extract that is composed predominantly of pollen. The 14C dates that we obtained for pollen extracts from peat samples associated with the Mazama Ash layer are consistent with the range of bulk-sample dates obtained by others in previous studies. The limited measurement time and resulting precision (± 100 yr) of these initial measurements restrict our ability to draw firm conclusions from a comparison of the pollen extract dates with previous bulk-sample dates. We intend to adjust our procedure to improve the rejection of non-pollen materials and to increase the precision of our 14C measurements on pollen extracts from peat samples in future studies.

We have developed and tested a practical device for manually separating pollen from pollen concentrates in sufficient quantity for AMS 14C dating. It is a combination of standard, commercially available equipment handled in a clean room by an individual trained to recognize pollen. A typical example requires about 15–20 h of hand-picking under the microscope. We show the usefulness of this procedure with results on a mid-Holocene segment from a core from Mono Lake. Sediments from this hardwater lake contain pollen and finely disseminated organic matter, but no macrofossils. The pollen dated ca. 1000 yr younger than the bulk sediment. The sediment “date” is most likely affected by incorporation of limestone-derived carbon, and is erroneously old.

We have measured 14C, 210Pb and 137Cs profiles in two representative cores from Manasbal Lake, Kashmir, India. The sedimentation rate derived from 210Pb and 137Cs in the upper part of the core is in the range of 3.4 to 5.5 mm yr−1. In contrast, 14C ages show an inversion at depths >20 cm. These results are attributed to the erosion of the ubiquitous 10–20-m-thick loess mantle, based on the similarity of 14C ages of the inversion layer in the sediments and the paleosols present in the catchment area. Frequency-dependent mineral magnetic susceptibility (χfd), carbon to nitrogen ratios and pigment concentrations in the profile show a significant amount of allochthonous component in the lake deposits and support the conclusion that the 14C dates do not reflect the chronology of the in-situ lake sedimentation but episodic deposition of the surrounding loess. Thus, 14C serves as a useful tracer to understand source components of the sediments.