Contamination by recent carbon and the turnover of organic matter make dating of ancient soils difficult. In order to isolate the oldest organic fraction of sediments, two main extraction methods were previously proposed: 1) alkaline solubilization of humus that separates humins, humic acids, and fulvic acids, and 2) successive hydrolyses that solubilize increasingly resisting products. Both preparation methods were tested on the same actual or fossil soils of different pedologic types from five geologic profiles on which other chronologic data are available.

Analytic results show that 14C ages obtained from alkaline extraction products differ according to the duration of treatments and characteristics of soils: while hydrolysis should yield more homogeneous results and isolate oldest fractions. It seems likely that true ages of geologic formations were never obtained from their organic matter and that the oldest organic fraction, contemporaneous with the sediment formation, completely disappears. Thus, most ages from 14C dating of organic matter of soils must be too recent.

All background values measured in one of our 14C counters during 1977 were analyzed together with corresponding values for atmospheric pressure, outdoor temperature, and coincidence counting rate for investigation of the possibilities to separate the temperature dependency of the background from the better known atmospheric pressure dependency.

The background values were normalized to a common atmospheric pressure and plotted in a time diagram. A seasonal trend with higher background values in winter than in summer shown by this diagram was confirmed by regression analysis.

The regression line and the “Standard error of estimate” was calculated for the background values as a function of the atomspheric pressure, on the one hand and as a function of the coincidence counting rate, on the other hand. A comparison showed that the coincidence counting rate in this case was a better parameter than the atmospheric pressure for calculation of the actual background.

The approximate consistency of the temperature effects on the background was checked for the years 1976 to 1978.

A preliminary review is given of the evaluation of the International Cross-Calibration Radiocarbon Data. Consensus values and standard errors are D14C = 508.1 ± 2.0% (ANU Sucrose) and Δ = −2.5 ± 1.5% (1850 Wood). Reporting format follows recommendations of Stuiver and Polach (1977).

In November 1973 the British Museum Research Laboratory acquired a Hewlett Packard 2100A mini-computer for the storage, calculation and retrieval of scientific measurements made on museum objects. A part of the computer's work is the calculation of radiocarbon dates based on the liquid scintillation counting of 14C activities. A system of programs and files has been developed and has been in daily use since August 1974 (Hall and Hewson, 1977).

This paper describes changes and improvements to the system to make it more flexible so that it now provides the full range of facilities required by an active 14C laboratory. The reporting procedures in particular have been restructured in the light of experience. The paper will be of interest to all laboratories that have, or hope to have, access to similar mini- or micro-computers.

An interlaboratory comparison experiment for 14C measurements has been organized jointly by the Harwell and British Museum laboratories to include the working radiocarbon laboratories of the United Kingdom. The experiment has been run along the lines of that organized by the IAEA for tritium and has explored the problems of sample preparation, verification of equivalent levels, and presentation of results. Samples of benzene representing 5 age equivalent levels between twice modern and 20,000 years old have been prepared and distributed for measurement either by direct counting (liquid scintillation) or full process tests. Preliminary results received show excellent agreement both between laboratories and in comparison with the known relative activities of the prepared solutions. The possibility of extending the experiment to cover different sample types and a wider distribution of testing laboratories is briefly discussed.

Results of measurements in an underground laboratory, 70m of water equivalent below surface, are given. Background values of proportional gas counters with volumes between 16cc and 1.5L are lower by a factor of 2—4.5 compared to the values in the previously used laboratory (7m w.e.). High counting gas pressures, up to 5 at CH4 and up to 36 at P-10, enable the use of relatively small counters with correspondingly small background contributions from the walls. A separation of the residual background into different components is attempted, distinguishing a pressure dependent volume effect and surface correlated contributions. It can be shown that the selection of radioactively pure counter construction material is very important for a good low level counting system.

The “shielding” efficiency of the guard counters has been a main scope of the present investigation. Our special guard counters consist of closed shells (ca 3cm thick) filled with propane at 1.2 atmospheres pressure. These guard counters are nearly 100 percent efficient against charged particles, and 1 to 2 percent against gamma and neutrons. The efficiency has now been studied more in detail in an arrangement with four guard shells around a 14C counter. For each extra guard shell, the cosmic fraction of the counter background was reduced by ca 13 percent. The reduction does not involve penetrating high energy charged particles, but is related to γ ray showers penetrating the guards. A thicker old lead shield between 14C counter and the guard counters also reduces the background and serves the same purpose. In order to approach underground conditions for the 1.5 liter counter background (0.32 ± 0.01 c/min), most of the shielding material has to be put inside guard shells. An ordinary guard counter combined with an extra guard on top of the iron shield is very efficient. A background of 0.48 ± 0.01 c/min has already been obtained.

The use of materials having high levels of 14C activity (up to 113 times the activity of modern carbon) enabled a quantitative analysis of the magnitude and sites of memory occurring in the routine synthesis of benzene, via lithium carbide, for radiocarbon dating. Memory may be expressed as the percentage or fractional contribution of carbon from sources other than the original sample in this synthesis. Although tritium and radon contamination have also been found, the major site of memory was the inner surface of the stainless steel reaction vessel used for lithium carbide production. Up to 1 percent memory has been found there under extreme conditions in the routine dating system at Harwell. Values of half to one-third this size were more usual, but even etching and scouring the inner surface of the reaction vessel reduced the memory only by a factor of four. This lower limit is believed to exist because of the carburization and decarburization of steel which occurs at the temperature required for the production of lithium carbide. The levels of memory found are of the same order as the levels of significance associated with present radiocarbon techniques. With the accuracy and extended chronologies expected from direct 14G measurement by accelerator techniques, these levels of memory become increasingly important in the preparation of acetylene and pyrolitic graphite (via lithium carbide) used as target materials. This effect, however, can be limited by lowering the temperature of the carbide reaction stage or by lining or impregnating the lithium carbide reaction vessel with some carbon-impermeable alloy or material.

Radiocarbon dates calculated from the ratio of modern carbon-14 activity and sample activity and the half-life of carbon-14 need to be calibrated to compensate for temporal variations in the concentration of carbon-14 in the atmosphere. Development of a suitable calibration scheme has been an ongoing process of the last twenty years, ever since the discovery of variations in historical times of the atmospheric radiocarbon content which parallel climatic and solar phenomena (de Vries, 1958; 1959) and the recent depletion due to industrial effects (Suess, 1955).

In 1957 the National Bureau of Standards (NBS) agreed, at the request of James R Arnold, to store and distribute an oxalic acid contemporary carbon-14 standard. In 1978 stocks of this standard were practically gone. We approached Chas Pfizer and Company, Inc, which had provided the original 1000 pounds of oxalic acid to Dr Arnold to see if they could provide another one-batch lot of 1000 pounds of oxalic acid to replace the old standard. This they did generously, at no cost to NBS. The oxalic acid was prepared by fermentation of French beet molasses from the 1977 spring, summer, and autumn harvest using Aspergillus niger var. The acid was separated from the broth as the insoluble calcium salt and reconverted to free acid using sulfuric acid. The free acid was then crystallized, redissolved, filtered, recrystallized, and dried.

We here report on two technical research projects of the Quaternary Isotope Laboratory (QL) vis, (1) the use of thermal diffusion isotopic enrichment to extend the technical range of 14C dating, (2) the preparation of samples for ion counting using a Van de Graaff tandem accelerator. The second project is carried out in cooperation with, and partly at, the Nuclear Physics Laboratory.

A gain in dating range of 3 to 4 half-lives can routinely be obtained with the QL and the Groningen enrichment systems. The same gain in age range can be obtained for ion counting with a simplified system that requires only 0.5 to 2g of carbon and 3 to 7 days enrichment time.

A method to convert CO2 quantitatively via CO into carbon is described. For short intervals the carbon deposit yields good 12C— beams. We also give a different procedure to make graphite-like carbon samples. The preparation of beryllium metal samples is given last.

There is much controversy about the significance to radiocarbon dating of isotopic fractionation during the oxidation to CO2 of NBS oxalic acid, the carbon-14 dating standard. To check the effect of fractionation, 30 routine potassium permanganate oxidations of oxalic acid were performed. The isotopic composition of resulting CO2 was determined mass-spectrometrically. δ13C results ranged between —17.7 and —21.2, with an average of —18.9‰ wrt PDB. The effect of not applying an isotopic fractionation correction to NBS oxalic acid when performing age calculations is evaluated.

The purpose of the workshop was to clarify any possible confusion related to the reporting 14C data. Two areas of concern, the use of the new NBS oxalic acid standard (Cavallo and Mann, 1980), and the reporting of marine sample ages, were singled out for detailed discussions.

Amino acid composition data and stable isotope ratios (14N, D, and13C) are being evaluated as sources of information to indicate the presence of non-indigenous organics in bone samples intended for 14C analyses. The study is being conducted in the context of the planned measurement of Pleistocene bone samples by a high energy mass spectrometric 14C detection system.

A commercial liquid scintillation counter (LSC) was tested for its ability of low level counting (LLC) in two underground laboratories (70m of water equivalent cover). Reduction of total background and background components compared to the normal counting laboratory (5m w.e. cover) up to a factor of 9 depending on the energy range and the shielding of counting cell and laboratory walls were obtained. First experiments with a large volume cell (50cc) led to a modern 14C net effect of 425cpm and a corresponding background of 1.6cpm.

To improve reporting of radiocarbon dates, Stuiver and Polach (1977) recommend that reported standard errors should include the error in the applied δ13C value, and suggest estimated mean values for δ13C to be applied when not measured. Based on δ13C data for ca 250 samples measured during 1975-1979, mean values for different materials dated by the Trondheim radiocarbon laboratory have been compiled. All material is from Norway and Svalbard (marine bone collagen). For peat, gyttja, and terrestrial bone material, δ13C should be measured to obtain optimal precision in the dates. For shell, wood, charcoal, and marine bones, the standard error in an estimated δ13C value will only increase uncertainty of a date from ±50 years to ca ±55 years.

Two Pinus aristata samples submitted by C W Ferguson were separated in different fractions, as was done earlier with Pinus silvestris L from Sweden, to yield different fractions for studies of the pretreatment. One sample in this new series consisted of heartwood and the second of sapwood.

The treatment performed in the radiocarbon laboratory involved an acid treatment by boiling, washings, an hydroxide treatment at 80°C, washings and, finally, another acid treatment before being dried before the combustion. The sodium-hydroxide treatment was repeated to yield at least two soluble and two insoluble fractions.

The treatment performed by the wood chemists involved extractions with ethanol-benzene and water. The remaining wood was dated but was also used for the production of holocellulose. The extractives were partitioned between ethyl ether and water and that from the older wood was used for the isolation of neutrals, acids, and phenols. In all, 19 fractions of these two wood samples were dated.

It is confirmed that a treatment for 30 min in sodium hydroxide at 80°C is not sufficient to remove the extractives from the heartwood or the sapwood. A treatment at 80°C overnight with 1 percent NaOH yielded a sample from the older wood with a 14C content in good agreement with the results predicted from the curve presented by Stuiver (1978). The final statistical uncertainty in the present investigation was ca 4‰. The younger wood yielded results indicating a lower activity than that given by Stuiver.

Wood from regular timbering of a shallow seated mine in Tuscany gave 14C ages of 5730 ± 100 years BP, much too old to be attributed to one of the known civilizations of Italy. This mine is located in a region of declining volcanic activity, noticeable especially through numerous emanations (natural or incidentally induced by drillings). It was suspected that the analyzed wood had grown in an environment where the normal atmosphere had been diluted by volcanic emanations. To check this hypothesis, living plants (trees, bushes and reeds) and volcanic emanations have been sampled and their 14C content measured. All present day plants are depleted in 14C, giving a fictitious age different from 0 (1805, 1820, 2540, 4350 years BP). Of the gaseous emanations sampled, two have a high pressure and show virtually no 14C (>41,000 years BP). Two others have a pressure close to atmospheric, and small amounts of 14C were introduced by atmospheric contamination (22,570 and 30,580 years BP).

Conclusion: plants grown in the vicinity of volcanic emanations have 14C activities that are too low because of a natural 14C dilution and yield anomalously high ages. The δ13C values obtained for two of these plants (—27.4‰ and —23.7‰) are close to the average for plants in general (—25‰), whereas the CO2 of mofettes is heavier than atmospheric CO2.

Although initial studies in the 1960's established the feasibility of applying the radiocarbon technique directly to ceramics, subsequent analyses have revealed this category of materials to be more complex than previously believed. Consideration of the points at which various types of organic carbon are introduced into ceramic materials by natural or human agencies allows more effective screening of potential sherd samples. It also suggest ways to overcome some of the limiting factors involved in dating ceramics.