Amperometric Measurement of Hydrogen Evolution in Chlamydomonas

Abstract

We report details of an amperometric method for measurement of hydrogen (H2) in the aqueous phase and its application to hydrogen evolution by an alga. Although amperometric measurement of hydrogen is not common , we wondered whether the conventional Clark electrode for measurement of oxygen could be applied to measurement of hydrogen simply by proper choice of voltage. Initial tests were encouraging. Further effort was necessary to obtain stable response. The resulting procedure has merits of high sensitivity at S FIG. 1. The cuvet assembly. Two ports of the water-jacketed cuvet received the electrode, E, and the stopper, S. Liquid level in the cuvet was maintained slightly above the point of seal between glass ball and bottom of the ground glass taper. low concentrations of hydrogen, rapid response, and use of conventional and available components. The apparatus (Fig. 1) used was a OX 700 Clark-type electrode (equivalent to YSI 5331, platinum electrode about 0.5 mm) and OX 705 water-jacketed cuvet obtained from Gilson Medical Electronics (Middleton, Wis., 53562). The cuvet was stirred by a glass-enclosed steel bar driven by a magnet on the shaft of a 450 rpm synchronous motor. electrodes were covered by a film of half-saturated KCl enclosed within a 19 u polypropylene membrane. The 0-ring holding the membrane also served as a seal for the electrode at its insertion into the side port of the cuvet. Temperature was held at 25 ±4-0.03 C by circulated water. The platinum electrode was polarized +0.60 v versus Ag/AgCl, and the current measured by a Keithley 150A microvolt-ammeter, was displayed on a Brown recorder. The same equipment has been used routinely for measurements of oxygen concentration. For measurement of hydrogen, two technical problems required attention. The cuvet proved to be more leaky for hydrogen than for oxygen. The top port of the cuvet is normally sealed by a ground-glass stopper with capillary open to air. Except at oxygen concentrations far removed from that at equilibrium with air, diffusion leakage through the capillary is negligible. However, even at very low concentrations of hydrogen, leakage was appreciable as evidenced by drift toward lower concentrations with time. The glass stopper was redesigned (Fig. 1). A 6 mm inner diameter glass tube was sealed to a standard taper joint 5 mm inner diameter at the lower opening. A spherical glass ball at the end of a 2 mm glass rod was hand-ground to seal against the inside bottom of the tapered joint. Liquid level in the cuvet was maintained slightly above the sealing point. Raising the ball to an upper part of the taper gave sufficient clearance for insertion cf a 20 gauge needle for aeration or for injection of added microliter quantities of solutions. With this modification the only leakage path remaining was that through the membrane and electrode assembly. A second technical difficulty arose in stability of electrode sensitivity, which drifted with time after imposition of a chosen polarizing potential. Polarograms of current signal versus potential showed a marked hysteresis when traversed first toward higher and then toward lower potentials. For example, the current at 0.6 v after a period at 0.7 v was almost twice that observed after a period at 0.5 v. Following procedures suggested by Gilman (2), we conditioned the electrode surface by a treatment consisting of about 10 min of timed 50/min alternations of polarizing potential between +0.2 and +0.8 volts. After such treatment, the current signal in response to a small concentration of hydrogen decayed slowly, about 10%/-in the 1st hr and about 25%" in 30 hr. In practice a new membrane preparation and electrode conditioning were applied once each day. Figure 2 demonstrates linearity and typical sensitivity. Time response (not shown) is estimated as follows. After a step change by injection of hydrogen-saturated water, the initial current response was rapid (<1 sec) but approach to final current was much slower (50% in 4 sec, 90%e in 50 sec). We consider the current signal specific for hydrogen. The electrochemical reactant must be (a) a nonionized small molecule in order to penetrate the membrane, (b) oxidizable at 0.6 v versus Ag/AgCl, and (c) re-108