Validated methods for sampling and headspace analysis of carbon monoxide in seawater

Abstract

Abstract A headspace analysis system with well demonstrated precision and accuracy for measuring carbon monoxide (CO) in natural waters and for CO incubation experiments is described. High water/gas volume ratios are accurately set by injecting known volumes of CO-free air into known volumes of water in glass syringes. CO in equilibrated headspace gas is separated chromatographically and quantified by a mercuric oxide reduction detector. A water/gas ratio of ∼7 is sensitive and precise enough for determining low-level CO; sensitivity can be increased by raising the water/gas ratio. At a water/gas ratio of 7 (40 ml total), the analytical blank, precision, and accuracy are 0.02 nM (nanomolar), ±0.018 nM±2%, and better than ±10%, respectively. Recovery of CO from the water phase is ∼88%. The system is efficient, simple, convenient, rapid and robust; it responds linearly up to ∼12 nM, and can process ∼8–12 samples/h. Several applications are illustrated: studies elucidating subtle CO-contamination artifacts, microbial oxidation incubations, and an oceanic profile. Validated low-contamination sampling methods are presented, and contamination control measures are recommended. A detailed 0–200-m profile at BATS in summer shows less “deep” CO than previously reported, but there is CO well below the seasonal mixed layer (ML) and even at the 1% light level.