Microfluidic biosensing systems Part I. Development and optimisation of enzymatic chemiluminescent µ-biosensors based on silicon microchips

Abstract

Chemiluminescent (CL) enzyme-based flow-through microchip biosensors (µ-biosensors) for detection of glucose and ethanol were developed for the purpose of monitoring real-time production and release of glucose and ethanol from microchip immobilised yeast cells. Part I of this study focuses on the development and optimisation of the µ-biosensors in a microfluidic sequential injection analysis (µSIA) system. Glucose oxidase (GOX) or alcohol oxidase (AOX) was co-immobilised with horseradish peroxidase (HRP) on porous silicon flow through microchips. The hydrogen peroxide produced from oxidation of the corresponding analyte (glucose or ethanol) took part in the chemiluminescent (CL) oxidation of luminol catalysed by HRP enhanced by addition of p-iodophenol (PIP). All steps in the µSIA system, including control of syringe pump, multiposition valve (MPV) and data readout, were computer controlled. The influence of flow rate and luminol- and PIP concentration were investigated using a 2³-factor experiment using the GOX-HRP sensor. It was found that all estimated single factors and the highest order of interaction were significant. The optimum was found at 250 µM luminol and 150 µM PIP at a flow rate of 18 µl min⁻¹, the latter as a compromise between signal intensity and analysis time. Using the optimised system settings one sample was processed within 5 min. Two different immobilisation chemistries were investigated for both µ-biosensors based on 3-aminopropyltriethoxsilane (APTS)- or polyethylenimine (PEI) functionalisation followed by glutaraldehyde (GA) activation. GOX-HRP µ-biosensors responded linear in a log–log format within the range 10–1000 µM glucose. Both had an operational stability of at least 8 days, but the PEI-GOX-HRP sensor was more sensitive. The AOX-HRP µ-biosensors responded linear (log–log) in the range between 1 and 10 mM ethanol, but the PEI-AOX-HRP sensor was in general more sensitive. Both sensors had an operational stability of at least 8 h, but with a half-life of 2–3 days.