Instructions for various experimental protocols.
(Not yet written)
(Not yet written)
Recipe for obtaining equilibrium Δ47 by CO2-H2O isotope exchange: The way to obtain CO2 with equilibrium Δ47 value at low temperatures is through isotope exchange between CO2 and H2O at the desired temperature. This is commonly done in order to use CO2 δ18O as a measure for H2O δ18O. Clumped isotopes equilibrium is obtained by the same exchange reaction; only Δ47 does not depend on the δ18O of the water used for exchange (beyond linearity correction, of course).
In order to determine the reference frame for Δ47 as it varies among instruments, the community had decided to perform CO2-H2O equilibration at 10°C, 25°C and 50°C and use them as a basis for the Δ47 scale, through interlaboratory comparison.
There are many ways to perform these exchange experiments and a lab that is experience in doing that for δ18O, should use the same method. For labs that do not perform these experiments regularly, I describe below the way I do it. Again, any way to expose CO2 to water would work, as long as CO2 is then separated from the water before it has a chance to exchange at a different temperature.
Yale CO2-H2O exchange method for Δ47 analysis: In exchanging CO2 and H2O for Δ47 analysis, the value is not sensitive to the isotopic composition of the water used but is sensitive to temperature deviations, as long as there is trace of liquid water available to exchange. The method is therefore designed to quickly quench any liquid water, instead of focusing on preserving the original δ18O of the water. It is convenient to use water of a range of δ18O values in order to plot Δ47 vs δ47 in the equilibrated gases. For any given temperature this plot should be parallel to the heated gas line, thus reflecting the non-linearity of the mass spectrometer.
We typically use 3 different waters: our lab DIW, water that was enriched by evaporating this water, and mineral water that are 18O depleted. This yields a range of δ47 values that is similar to the range of heated gases we typically measure.
Equilibration is performed in a break seal (1/4” OD, ~20 cm long) that is fully immersed in water of the desired temperature (inside a temperature controlled bath), to guarantee that no part of the exchange reaction occurs at a different temperature. ~0.1ml of water is inserted into a break seal and the bottom is frozen in liquid N2. If the lab is humid this will potentially affect slightly the δ18O of the water, but not the Δ47 of the CO2. The break seal is then connected to a vacuum line and the air is pumped away while the water are still frozen (to minimize wetting the vacuum line). The valve to the break seal is closed. The CO2 aliquot is inserted into the line, frozen in a U trap and non-condensable gases are removed (if there are any). The CO2 is transferred into the break seal, which is then fire sealed. Typical CO2 amount: 100-150 μmol (the equivalent of 2-3 sample sizes), though smaller amount is OK as well.
Break seals are immersed in the desired temperature for approximately 3 days (this is likely to be much longer than necessary) to ensure CO2 reached isotopic equilibrium.
In order to extract the CO2, it has to be first quenched so that it does not continue exchanging with water at lab temperature during sample processing. This can be done by dipping the tip of the break seal in liquid N2, freezing both the CO2 and the water and preventing further exchange. The upper part of the break seal is scored, inserted carefully into the cracker and mounted on the vacuum line. Then, liquid N2 is replaced by a dry ice-ethanol mix to keep the water frozen but to release the CO2. CO2 is let thaw while headspace is pumped. The break seal is then cracked and processed like a CO2 sample, by passing through a dry ice-ethanol trap into a liquid N2 trap on the vacuum line.
CO2 that was isolated from water using dry ice-ethanol traps is collected in a dried break seal. Drying can be done by heating break seal with a flame while pumping on the vacuum line. After sealing and disconnecting the break seal the hot tip is dipped into the liquid N2 dewar to quickly cool it, avoiding exposure of CO2 to high temperature. After extraction, vacuum line is likely to contain some water and must be dried thoroughly (I do it with a low oxygen flame) before using it again; extraction should therefore be done at the end of the work day.
CO2 is not likely to be clean enough to measure as is, and should undergo cleaning, just like any other sample.
Heated gases are CO2 samples whose isotopes are randomly distributed among all isotopologues and therefore have Δ47 value of zero. This is obtained by heating the CO2 to 1000°C.
Make sure to test your quartz before using it to store heated gases. I tested 2 types of quartz tubes. ¼” tubes purchased from GM Associates were tested for storing heated gases and gas was found to be stable during 1 year. These tubes tend to produce signal in mass 48, that can be removed using the typical GC cleaning. Tubes from Hereaus do not produce signal in mass 48, however, heated gases can be used only if freshly made. After 2 weeks the Δ47 signal is partially lost and after a month the Δ47 value is closer to room temperature equilibrium than to heated gas. It is not recommended to use these tubes.
Use a range of δ13C and δ18O of the CO2 such that δ47 (which is about δ45+δ46) would bracket the range of your samples. This can be done either by having 2 types of CO2 and creating mixtures between them (remember that the mixing must be done before heating, as Δ47 does not mix linearly), or by recycling samples that are large enough (namely recollect the CO2 left in the bellows after the run and use it to make a heated gas). In both cases note that δ18O in heated gas tends to be a few ‰ more depleted than the original gas, so in order to bracket your enriched samples (most marine carbonates) you would need CO2 with relatively enriched δ18O (this can be done with enriched δ13C but it is difficult to fine a CO2 cylinder with enriched δ13C). Probably the easiest way is to use CO2 from a cylinder as the depleted end (but try not to use the common CO2 that has δ13C of -30 to -50‰ as the δ47 would be lighter than any likely sample). Then equilibrate a large aliquot (I do it in 500ml batches at atmospheric pressure) of that cylinder CO2 with water that is 18O enriched by evaporation.