The ARAMIS system

All JRAS sets have been prepared using the ARAMIS system at BGC-Jena. ARAMIS (‘Acid Reaction and Air Mixing System’) was built to enable the production of well defined SRMs (Standard Reference Materials). It is designed to minimize variability between different preparations/batches and to ensure a consistent production of reference CO₂ mixed into CO₂ free air. (for details see http:\\...).

JRAS preparations/batches

Each digestion of calcite results in three flasks of mixed gas. These three flasks are filled simultaneously from the mixing volume and represent a preparation or batch. The CO₂ and N₂O concentrations in each flask are measured by the BGC Gas-Lab prior to isotope analysis in BGC Iso-Lab. Isotope analysis of δ¹³C and δ¹⁸O are repeated three times for each flask providing a total of nine measurements per batch. Below are the results of the batches prepared within the IMECC project (which funds this activity). Each data point represents the average of the nine measurements with the error bars showing one standard deviation.

Figure 1. Batch wise averages (n=9) from the MAR-J1 preparations made since the beginning of the IMECC project. The overall average for δ¹³C is 1.956 ‰ with a standard deviation of 0.012 ‰ . The δ¹⁸O average is -2.58 ‰ ,  σ = 0.035 ‰. The time scale refers to the batch production dates.

 The average δ¹³C values for both MAR-J1 and OMC-J1 show a variability that is comparable or better than the average flask to flask variability obtainable by the BGC IsoLab (±0.02 ‰ as opposed to tank measurements which has a long term precision of ~0.013 ‰). The oxygen values are as expected not quite as precise, although the overall variability is also less than the average precision for flask measurements (±0.04‰). Based on these observations we can conclude that the preparation process is not adding any detectable variability when comparing the batch averages. Also, even though only a limited number of batches have been prepared so far, no systematic trend in time is visible as more preparations are added, indicating that the working reference air-based analysis is consistent and stable.

Figure 2. Batch wise averages (n=9) from the OMC-J1 preparations made since the beginning of the IMECC project. The overall average for δ13C
is -4.371‰ with a standard deviation of 0.017 ‰. The δ18O average is -8.927 ‰,  σ =0.027 ‰.

	

While this is all good news for JRAS, it does not leave much room for improvements. On the other hand, if we take a closer look at the two isotopic values for each calcite together we still find a correlation between δ¹³C and δ¹⁸O for both calcites (Figure 3). Although the ranges are too small to allow for a quantification of the correlation with any certainty, the sign of the slope and to some degree the magnitude suggest that this is a consequence of mass dependent isotope fractionation. Where this fractionation could take place is not clear. Any of the expansion steps when the CO₂ formed during digestion is moved within the ARAMIS system is a possible culprit.

This inherent “shortcoming” of the ARAMIS system does not have any direct consequence for the JRAS SRMs as they are assigned a value that is measured after production. Nevertheless, the understanding and potential elimination of this systematic feature could possibly limit the isotopic variability of the preparations and, if successfully addressed, allow to improve our results even further. We are therefore investigating whether the digestion protocol or the ARAMIS system itself could be further improved. Also, as we are making more preparations, the preparation record will be updated regularly on this web-site.

Figure 3. Batch wise average values of δ¹³C and δ¹⁸O for each calcite, error bars are one standard deviation of the  averages (n=9)