Abstract. Water molecules in vapor can exchange with gaseous water molecules sticking to surfaces of sampling tubing, and exchange rates are unique for each water isotopologue and tubing material. Therefore, water molecules on tubing walls take some time to reach isotopic equilibrium with a new vapor isotopic signal. This creates a memory effect that is observed as attenuation time for signal propagation in continuous stable water vapor isotope measurement systems. Tubing memory effects in δD and δ18O measurements can limit the ability to observe fast changes, and because δD and δ18O memory are not identical, this introduces transient deuterium excess (D-excess, defined as δD-8×δ18O) artifacts in time-varying observations. To our knowledge, a comprehensive performance comparison of commonly used tubing material water exchange properties in laser-based measurement systems has not been published. We compared how a large isotopic step change propagated through five commonly used tubing materials for water isotopic studies – perfluoroalkoxy (PFA), fluorinated ethylene propylene (FEP), polytetrafluoroethylene (PTFE), high-density polyethylene (HDPE), and copper – at two different temperatures and an airflow rate of 0.635 L min−1 through approximately 100 ft (30.5 m) of 1/4 in. (6.4 mm) outer diameter (o.d.) tubing. All commonly used tubing materials performed similarly to each other in terms of attenuation times, reaching δ18O-location-adjusted δD and δ18O 95 % completion in less than 45 s, with slight variations based on temperature. PFA does appear to perform slightly better than the other materials, although memory metric differences are small. A tubing material commonly used in the early 2000s but reported to have memory effects on δD, Dekabon, was also tested at ambient temperature and changing humidities. The Dekabon isotopic equilibrium was not reached until nearly an hour after source transition, much later than H2O mixing ratios equilibrated. Bev-A-Line XX (used in some soil O2 and CO2 gas studies) was also tested at ambient temperature, but it did not approach isotopic equilibrium until after nearly 6 h of testing. Therefore, we cannot recommend the use of Bev-A-Line XX or Dekabon in water vapor isotope applications. Source transition from heavy to light or from light to heavy affected isotopic transition speed only in experiments where H2O ppmv was changing. While a shorter tubing lengths and smaller inner diameters shorten the delay of signal propagation through the tubing, they did not greatly change the attenuation curves under these conditions for the current commonly used tubing materials tested. However, in Dekabon, attenuation curves were greatly extended with increased tubing length. Our results show that the commonly used plastic tubing materials tested were not inferior to copper in terms of isotopic memory under these conditions, and they are easier to work with and are less expensive than copper.
Read full abstract