The problems we were presented with were:

  • The sample gas was in high vacuum
  • The mass spectrometer measurement required a pure CO2 gas in high vacuum
  • Even the smallest amount of lost CO2 could drastically affect the customers isotopic measurement (isotopic fractionation), so they required the CO2 to be purified but none of the CO2 to be lost.
  • The system had to be simple and robust enough for staff and visitors with no GC training to use on a day-to-day basis, thus all valve control, baking and timing had to be simple and automated as much as possible.

We had a series of technical meetings with SUERC, especially to discuss the peculiarities of the vacuum to Helium system at the sample inlet, and Helium to vacuum interface at the mass spec, and we came up with a workable design, which ordered.

The system consisted of:

  • Agilent 6890 GC with associated regulators for high purity carrier and TCD detector
  • Two sequential Hayesep-Q GC columns, one short one for backflush to remove the worst contaminants and a longer one as the main column to separate the most similar compounds from the peak of interest, CO2
  • Three computer controlled electrically actuated Valco valves, one 4-port for the vacuum to Helium interface, one 4-port for the Helium to vacuum interface, and one 6-port to control the backflush column
  • Two “cryoloop” systems connected to the 4-port Valco valves at the inlet and vent ends of the GC, allowing introduction of Helium at the inlet and cryogenic separation of the CO2 from the helium using liquid nitrogen.
  • DataApex Clarity software with interface hardware to sequence two of the Valco valves, working in conjunction with the 6890 GC’s own sequencing. Clarity was configured to trigger along with the GC, making the whole system start with one button
  • Hardware to interface the GC inlet to our vacuum-to-Helium system and to interface the final cryoloop to the inlet of the mass spectrometer

We installed and commissioned the system, trained the client fully on the use of the system and assisted them in setting up the timing and temperature profiles required to separate their sample gases by using a variety of test gases.

It is now almost 3 years since the GC system was installed, and apart from very minor timing changes to the sequence it is unchanged from what was provided by Speck and Burke.  It is in day-to-day usage, and produces incredibly clean CO2 samples, typically 99.9999%+ purity from gases which are sometimes 30% CO2 and 70% contaminant at the gas inlet.  The measurements from these gases are superb, typically with accuracy and precision of around 0.010permil (parts per thousand), which translates to an error of <2 degrees Celsius of the temperature measurement we are trying to obtain.  SUERC have now completed two major projects between mid-2017 and late-2018, and our ongoing work is to integrate the GC system into a fully automated gas extraction system over the early part of 2019.

In the field of isotope geochemistry, there are some techniques which cannot be performed without having the correct equipment and gas handling apparatus, and in the case of our client the amount of effort and attention to detail provided by Speck and Burke in providing the GC system has taken them from a non-functioning method to one which is world class.

Further Technical Information

In mid-2015 SUERC started work on commissioning a clumped isotope analytical system. Clumped isotopes are a relatively new isotopic analytical technique, first used in around 2005, and mainly used as a thermometric technique, measuring the temperature of formation of carbonates in everything from the skeletons of sea animals, terrestrial rocks, extra-terrestrial rocks like meteorites and esoteric samples like dinosaur teeth, by dissolving the carbonate in acid to produce CO2, and measuring it in a stable isotope mass spectrometer. We started with a vacuum chemistry line built in-house and a Thermo Fisher MAT253 mass spectrometer configured for measuring clumped isotopes in CO2.  Our early measurements proved problematic.  Most labs doing clumped isotopes work with very clean, pure carbonates, however our first samples were early solar-system meteorites which are typically only 2-10% carbonate, with the rest of the meteorite consisting of organic material, sulphur and various other chemicals.  The entire measurement scale for clumped isotopes is one part per thousand of the total component, i.e. the “coldest” measurement is a mere 1/1000th bigger than the “hottest”, so the tiniest amount of contamination can hugely affect the measurement.  To obtain high precision (typically in the order of 10ppm) we measure the gas for a long time, but to obtain the equivalent accuracy we require a gas which is completely free of interfering contaminants.  We had tried using a simple molecular sieve inline trap at this point, which proved only partially successful in keeping our gas clean.  Early in 2016 we approached Speck and Burke with the purpose of discussing a full GC system to purify our CO2.