Bisegmented flow system for determination of low concentrations of gaseous constituents in gaseous samples

Abstract

The monosegmented flow approach has been modified to produce a bisegmented flow system aimed at the determination of gaseous analytes present in low concentrations in gaseous samples. The concentration ranges can vary from < 1% (v/v) to nl l⁻¹. A flow pattern constituted of b#1-ls#1-b#2-ls#2-b#3, where b indicate a gas bubble and ls a liquid segment, is produced by a single movement of the injection device. The system is computer controlled and two procedures can be used for the determination of the gaseous analytes. In the first the b#2 segment constitutes the sample and a suitable reagent is added in ls#1 after the whole pattern is present inside the glass flow manifold. The gas sample is forced to flow over the liquid reagent layer left behind by ls#1 and a detectable specimen is formed. The ls#2 collects the product and carries it to the detection point. In the second procedure ls#1 contains a suitable absorbing reagent and forms an absorbing layer on the tube wall. The carrier flow is stopped when b#2 reaches an entrance arranged perpendicular to the glass reactor tube. Through this entrance a large volume of a gaseous sample is impelled. The gaseous analyte is absorbed and concentrated in the liquid layer. About 85% of the analyte can be retained by the absorbing layer. When this operation is finished the carrier flow is re-established and ls#2 removes the absorbed analyte from the liquid layer. This segment may or may not contain a suitable reagent. The system has been applied for the determination of O₂ (0–1% v/v) in the atmosphere of food packages by reacting the analyte with pyrogallate and of NO₂ in synthetic air by absorbing the analyte in triethanolamine and reacting it with Saltzman reagent in ls#2. The system is capable of determining NO₂ in air samples in the range 25–250 nl l⁻¹.

Introduction

The use of flow systems for determination of gaseous analytes in gaseous samples is very seldom described in the literature, despite the necessity for procedures that are rapid, robust, less reagent demanding, less prone to interference and with lower cost per analysis. Gaseous constituents can range, on a volume basis, from 50% or more to a few nl l⁻¹ for those analytes of interest for pollution studies, such as nitrogen dioxide and ozone.

The flow injection (FI) approach [1]has been employed in the past aiming for the determination of SO₂, Cl₂, Br₂ and CO₂2, 3, 4, 5. In some cases the flow system was employed only for sample transportation [2]while in others an in line reaction was promoted between the gaseous analyte and a reagent stream [3]. Low sensitivity and lack of versatility were, perhaps, the most unfavourable characteristics of such proposals. Recently an interesting flow cell has been developed for the spectrophotometric determination of low contents of NO₂ in gaseous samples employing permeation through a PTFE membrane and a static reagent acceptor containing Saltzman reagent [6]. The sensitivity attained was very good and demonstrates that the chemical approach to gas determination still has many aspects to be exploited. However, the analyte transportation to the reagent solution is restricted by the membrane and, at suitable flow rates, only about 2% of the analyte reaches the reagent solution.

Even more recently the monosegmented approach to flow analysis [7]has been used for gas analysis. The monosegmented system has intrinsically the characteristic that gas and liquid segments are simultaneously generated in the system by a single movement of an injection device [8]. Aiming at the determination of gaseous analytes, one of the gaseous segments is constituted by the sample and various strategies can be carried out in order to provide an analytical response dependent on the analyte concentration. The first approach employs the volume contraction of the gas segment, occurring after selective absorption of a given analyte, as the analytical parameter. O₂ and CO₂ have been determined using this approach in samples containing 2–80% (v/v) and 2–50% (v/v) of the gaseous analytes, respectively [9]. Further development of the monosegmented system allowed the determination of CO₂ in atmospheric samples containing a low concentration of this substance (100–800 μl l⁻¹) using conductimetric detection of the ions formed after CO₂ dissolution in the deionized water layer of the reactor tube[10]. These ions were collected in a liquid monosegment, also constituted of deionized water. Therefore, no reagent is necessary to perform CO₂ determination.

On the other hand, many chemical procedures for spectrophotometric determination of gaseous analytes are based on the absorption of the analyte in a suitable reagent that can produce a detectable product or simply retain the analyte in a solid or liquid substrate. Liquid absorbing reagents have the advantage of being renewed from one determination to other and are attractive for use within flow systems. Unfortunately, the monosegmented system, as originally conceived, is not capable of realising the two basic operations necessary for gas determination, namely, the gas absorption/reaction and product transportation to the detector. Furthermore, the determination of low concentration samples (for instance, in the range of nl l⁻¹) would require a pre-concentration of the analyte/product which can not be supplied by the simpler monosegmented systems described so far.

In view of the drawbacks indicated above, this paper proposes, a new flow system that produces a bisegmented flow pattern which can ensure in full the requirements for determination of low content gaseous samples. As the system produces a pattern containing two sequentially arranged liquid segments bordered by three gas bubbles it was named as `bisegmented flow system'. The use and performance of this system for determination of O₂ in the atmosphere of food packages and of NO₂ in synthetic atmospheric air are described below.