|
Abstract The most widely used method for pyrolysing samples for hydrogen or oxygen isotopic analysis involves heating them to greater than 1300°C in a helium stream passed through a glassy carbon tube in an alumina casing. There are a number of difficulties with this. Glassy carbon tubes are expensive and interaction between the carbon tube and the outer casing produces unwanted carbon monoxide by reduction of the alumina at high temperatures. The latter effect is overwhelming if temperatures of 1400°C or greater are used for pyrolysis. We experimented with lining alumina casings with pure molybdenum sheet. It is relatively cheap, conforms well to the interior of the reactor tube （to avoid carrier and sample bypassing of the carbon pack）, resists high temperatures and neither oxidises excessively nor absorbs the gases. The main disadvantages are that silver sample cups must be used and that the molybdenum degrades over time by formation of the carbide. We can maintain sharp peaks, high precision and good accuracy over more than 700 solid samples for both hydrogen and oxygen. The reactors last longer for water injections. The molybdenum in the columns does not contribute greatly to memory effects. The precision of analysis is dependent on other factors as well as the pyrolysis column, but for oxygen we typically achieve approximately <0.2‰ （sucrose）, <0.25‰ （water） and <0.25‰ （leaf）, sometimes using only a linear correction of drift, after dividing the run into 1 to 3 segments. Copyright ? 2008 John Wiley & Sons, Ltd.
|
