Entwicklung eines amperometrischen Biosensors für Cyanid

The aim of this work was the development of an amperometric biosensor for cyanide by the use of a three-step reaction cascade involving two enzymes. In the first step of this cascade, the enzyme cyanidase (EC 3.5.5.1) hydrolyzes cyanide to formic acid and ammonia. In the following, formate, the corresponding base of formic acid, is oxidized to carbon dioxide by catalysis of formate dehydrogenase (EC 1.2.1.2), while NAD, serving as cosubstrate, is reduced to NADH. In the third step, the oxidized form of a redox mediator reacts with NADH, yielding NAD and the reduced form of the mediator. Amperometric detection is possible for formate on the first, NADH on the second and the reduced form of the mediator on the third step of the cascade. Technically, two different sensor setups were used. On the one hand, a flow-through system was employed. It consisted of an injection valve, enzyme cartridges and a flow cell with amperometric electrode chips. On the other hand, gas phase experiments were carried out by placing a mixture onto an electrode chip. The mixture contained the enzymes and reagents necessary for the reaction cascade. The electrode chip was put in the gas phase of a sealed vessel. An acid component resided at the bottom of this vessel. Addition of cyanide containing samples into the acid resulted in protonation of cyanide, leading to hydrogen cyanide. Hydrogen cyanide then entered the gas phase and reached the mixture covering the electrode chip. Detection of formate at the first step of the cascade was possible solely by platinum working electrodes. In a flow-through setup, cyanidase immobilized in a cartridge was not able to completely hydrolyze the cyanide in samples. The remaining cyanide interacted with the platinum electrode, interfering with measurements. A removal of non-hydrolyzed cyanide by formation of metal complexes was not successful. Detection of NADH was feasible by the use of a flow-through setup. The limit of detection and the limit of quantification were determined to be 0.7 and 2.4 µM, respectively. The linear range extended from the limit of detection up to 1 mM. Nitriles and sulfide did not interfere, while thiocyanate showed a moderate interference. However, formate is expected to cause disproportionate signals by design. Estimation of cyanide in real samples with complex matrices such as plant extracts were possible. The correctness of the determination of cyanide concentrations in extracts of leaves of cherry laurel were checked by the use of the DIN method 38405-13 as a reference. With an average deviation of -4,8%, the sensor results were in good conformance with the reference results. Moreover, a good storage stability was found. After more than three months, no decrease of sensitivity could be observed. In contrast, gas phase measurements proved to be problematic. Additions of cyanide-free liquids into the acid component induced signals similar to those of cyanide containing samples, supposedly caused by the influence of changes in water vapor pressure. Meldola Blue as a redox mediator together with graphite working electrodes showed to be the only combination able to differentiate between cyanide free and cyanide containing samples (millimolar concentration range). During these experiments, concentrations of hydrogen cyanide in gaseous phase were between 2 and 800 ppm. Other mediators such as ferricyanide or quinones were found to be unsuitable. Signal heights were hardly reproducible, even when using Meldola Blue. Furthermore, the rate of signal response changed in the course of experiments. Also, the orientation of the electrode chip influenced the signal trend. The mixture could not be fixated on the chip. Chemical and electrochemical modifications of the working electrode were dissatisfying. To successfully develop a practicable amperometric gas sensor, further investigations are necessary. Yet, the three-step reaction cascade basically seems to be suited for measurements in gaseous phase.