Sample treatment AMS

A.T. Aerts-Bijma, D. van Zonneveld, F.N. Ghebru,  HA.J. Meijer, J. van der Plicht, T.F. Dijkstra, J.J. Spriensma (carbonate)

H.A. Been, E.N.H. Kuperus (until 5/00) (technical support)

J.C. Roeloffzen (software support)

 

The AMS sample handling laboratory has, at present, a capacity of about 2000 samples annually. For more information numbers and various materials, we refer to section RA-03 of this report.

 

Analogous to the conventional ¹⁴C laboratory we realised a separate preparation laboratory. The organic samples are chemically pretreated to remove contamination, before they are combusted to CO₂. The form and the intensity of the pretreatment depends on the type, quality and quantity of the sample. The usual method is Acid-Alkali-Acid (AAA) for the peat  and other organic deposits, charcoal and wood, Longin for bone and phosphoric acid for carbonates.

 

For inorganic samples bi-carbonate from water is converted to CO₂. CO₂ from air samples is trapped  cryogenically . Breakseals are broken on one of the laboratory’s mass spectrometers, enabling direct measurements of stable isotopic ratios (¹³C, ¹⁸O), or they are directly broken on the graphitization setup.

 

The organic samples are combusted in an automatic Elemental Analyser Carlo Erba 1500 which is coupled to a Micromass Optima IRMS (see SI-01). It consists of a Cr₂O₃ flash combustion tube, a silvered cobaltous cobaltic oxide purification furnace, a Cu reduction tube,  a water trap and a gas chromatographic column to separate N₂ and CO₂. The Elemental Analyser is coupled on-line to a continuous flow stable isotope mass spectrometer enabling high-precision d ¹³ C and d ¹⁵ N determinations. The CO₂ is trapped cryogenically for graphitization later on.

 

The (purified) CO₂ as produced by the Elemental Analyser, breakseals, the water laboratory or the conventional ¹⁴C laboratory (test measurements, standards and conventional samples which appear too small) is cryotrapped in small glass vessels. These are transferred to the graphitization room. For graphite production, we employ the method of reduction under an excess of hydrogen gas, using iron powder as a catalyst: CO₂ + 2H₂ ® (Fe) ® 2H₂O + C. The H₂O is trapped in a cold finger, which is cooled to –18^oC by a Peltier element. During the reduction reaction, the pressure is monitored on computer screens. In general the reaction takes 100 min. We employ one 10-fold and one 15-fold graphitization system, with volumes ranging from 3 to 8 ml. With this setup we can handle samples from 150 m g C and up.

 

In original, commercially available EA systems, a Gas Chromatographic column (GC) is used to separate the CO₂ and N₂. At first, the CO₂ was trapped cryogenically by hand, and transferred to the graphitization system in a separate room. Recently, the EA/IRMS system has been expanded with an automatic Cryogenic Trapping device (CT). This system consists of 40 separate freeze fingers in a vacuum system. The CT system is described in more detail in (van der Plicht et al., 2000). Each sample, combusted by the EA will deliver its CO₂ to one of the traps. Each trap has a pneumatic valve; a small cart/elevator with a dewar filled with Liquid Nitrogen can move from trap to trap.

 

A photograph of the combined EA/IRMS/CT system is shown in fig. 1. The system is fully automated using LabView, under communication with the commercial IRMS software. The total system can combust samples and collect CO₂ unattended for up to 40 samples. It is designed to run overnight. The trapped CO₂ are then collected and transferred to the graphitization setup.

 

In a conventional EA system, the sample is combusted in a hot furnace, yielding CO₂ and N₂ gases after purification and water vapour removal. The complete system is flushed with He gas. The GC column separates the CO₂ from the N₂. After the GC column a TCD detector measures the amounts of CO₂ and N₂ in the He flow. A small fraction of the EA output (typically  1%) is led to the IRMS in order to measure d ¹³ . Most of the He flow is led through  liquid N₂, where the CO₂ is cryogenically collected. This EA system (combustion and GC separation) is shown schematically in fig. 2 (top). Such EA/GC/IRMS systems are designed for stable isotope measurements. We call here such systems “conventional”, as opposed to “modified” (described below).

 

Radiocarbon analyses, in particular for old samples, are vulnerable for contamination effects. In a conventional EA system as described above, contamination with rest material from previous samples (memory effect) can occur, both in the combustion furnace, and in the GC column. The purpose of the column is the separation of CO₂ and N₂. The commercial EA/IRMS combinations, designed for stable isotope analysis, deliver typically a few m g C to the IRMS. They are not intended for CO₂ trapping, which is needed for subsequent ¹⁴C analysis. Such EA systems, however, are used by many AMS laboratories. Since CO₂ trapping is done cryogenically, the GC column is in fact superfluous since the Liquid Nitrogen trap separates the CO₂ in the He-flow from other gases such as N₂, and possibly traces of O₂. Therefore, we modified the EA/IRMS design as shown in fig. 2 (bottom). The gas flow through the system (CO₂ after combustion and He carrier gas) is 120 ml/min with 1 mg C. The flow is split in 2 pathways: (1) 119.5 ml/min He (‘waste’) with ca. 2 mg C to the CT system, and (2) 0.5 ml/min He with ca. 8 m g C to a new GC column (for column types see fig. 2) followed by a splitter valve. The latter reduces the flow to the IRMS to the desired 0.1 ml/min (2 m g C), whereas its waste line contains the remaining waste . The bulk of the CO₂ (corresponding to ca. 2 mg C) now flows directly from the combustion/purification ovens into the cryogenic trapping system, without being affected by the GC columns. Only the small fraction needed for the IRMS (≤ 10 m g C), is GC-separated by a small column.

 

We have performed contamination tests by ¹⁴C analysis of ANU-sucrose and anthracite, by means of alternating combustion in our EA/IRMS/CT system. This was done for both the conventional EA system, and for the modified EA system. The radiocarbon activity of ANU-sucrose is 150.6%  whereas anthracite is the commonly used material for background determination. The sample size was 2 mg C in all cases. After combustion and trapping, the CO₂ was transported to the graphite system. Next, the graphite was pressed into sample holders for the AMS ion source  and finally the ¹⁴C content was measured by the AMS.

 

The results of the contamination tests are shown in fig. 3. The conventional and modified EA systems were used in parallel, using the same IRMS and the same CT system. Fig. 3 (top) shows the AMS-¹⁴C results for alternating combustion of, at first, ANU-sucrose, followed by 2 combustions of anthracite. The anthracites for the conventional system (squares) show significant memory effects. For our modified system (circles), no memory effects are visible. From the comparison of the two, it is clear that in the conventional system, even the second anthracite sample in row still suffers from significant traces of the sucrose sample, present in the GC column.

 

Combustion of blank samples (empty tin cups commonly used for the EA) in between all samples, for reasons of cleaning the combustion furnace, does not lead to improvement for the anthracite results using the conventional EA setup (fig. 3, bottom).

 

Apparently, the traces of CO₂ from the previous sample in the GC column cannot be washed out by the flow; they can only be diluted by a new burst of CO₂. The modified system does not show further improvement by the blank combustion either, which indicates that memory effects due to remains in the combustion oven are negligible.

 

Comparison of the behaviour of the conventional and modified EA setup (fig.4) shows that in the conventional setup,  about  0.4% of the previous sample is admixed to the next one. This explains the average value of 0.57% for anthracite samples in the conventional setup, as well as its relatively large spread of 0.17%.

 

In the modified system, the average value for anthracite is 0.24 ± 0.05%. This value overlaps with the Rommenhöller gas background value for the graphitization system (see fig.2), from which we conclude that the memory effect of our modified system is now negligible. Especially the large reduction of the spread in the background values from conventional to modified system has improved the accuracy of dating older samples in our laboratory.

 

More detailed reports are presented at various conferences (ref. 1- )

 

References

 

1.       Aerts-Bijma, A.T., Meijer, H.A.J. and van der Plicht, J., 1997. AMS sample handling in Groningen. Proceedings AMS-7, A.J.T. Jull, J.W. Beck, G.S. Burr (eds.). Tucson, AZ, USA May 20-24, 1996. Nuclear Instruments and Methods B 123, 221-225.

2.       Aerts-Bijma, A.T., van der Plicht, J., and Meijer, H.A.J., Automatic AMS sample combustion and CO₂ collection. Radiocarbon (submitted). Proceedings of the 17^th International Radiocarbon Conference, Jerusalem, eds…

3.       van der Plicht, J., Aerts-Bijma, A.T., Wijma, S., Zondervan A., 1995. First results from the Groningen AMS facility. Proceedings of the 15^th International ¹⁴C Conference, Cook, G.T, Harkness, D.D., Miller, B.F. and Scott, M. (eds.). Radiocarbon 37(2), 657-661.

4.       Van der Plicht, J., Wijma, S., Aerts-Bijma, A.T., Pertuisot, M.H. and Meijer, H.A.J. 2000. Status report: the Groningen AMS facility. Proceedings AMS-8, Kutschera, W., Golser, R., Priller, A. and Strohmaier, B. (eds.). Vienna, Austria, 6-10 September 1999. Nuclear Instruments and Methods B 172, 58-65.

 

Fig. 1.     Photograph of the EA/IRMS/CT (from left to right) system, employed for combustion of organic samples for the Groningen AMS facility.

 

Fig. 2.     Schematic setup of the Elemental Analyser system.

                   top: conventional combustion and GC separation

                   bottom: modified: combustion without GC separation.

 

Fig. 3.     Results (¹⁴C/¹²C ratio) for alternate combustions of ANU-sucrose and Anthracite showing contamination effects for both the conventional and the modified EA setup

                   top: alternating combusion of sucrose and then 2 anthracite samples

                   bottom: alternating combustion of sucrose, blank and anthracite samples.