This thesis presents the development of an electrochemical enzyme biosensor-on-a-chip for the measurement of glucose or eserine as an acetylcholinesterase (AChE) inhibitor. The design and characterisation of such a sensor requires an interdisciplinary approach. It is necessary to combine knowledge of electrochemistry, enzyme kinetics and microfluidics.
The main advantages of biosensors are that we can measure at the location and time of our choice. Device use and analysis procedures are simplified, so no specially trained personnel are required. Biosensors can also be used for preliminary analyses, prior to more complex and expensive analyses. Microfluidics has been added as an experimental technique that allows the upgrading of biosensor systems. The added value lies in the use of extremely small sample and reagent volumes, the possibility to perform several different experiments simultaneously and in a short time, the possibility to automate experiments and the possibility to quickly adjust the number of process units.
Biosensor operation can be described by biochemical and physical processes such as analyte transport in the microchannel, enzyme kinetics and chronoamperometry. I have combined these phenomena in a basic simulation model using the pyrroloquinoline quinone-glucose dehydrogenase B (PQQ-GdhB) biosensor as an example. The system of equations is solved numerically using the differential method for the approximative solving of partial differential equations. Model output is the response of the sensor to different glucose concentrations.
In my work, I report on two experimental designs of biosensors – a biosensor in a reservoir and a biosensor in a microfluidic system. In both cases the sensor is electrochemical. A typical three-electrode system was used. In the case of the biosensor in the reservoir, the working electrode (WE) was a thin-film gold electrode, the counter electrode (CE) was a platinum wire wrapped around a reference electrode (RE) which was a miniature Ag/AgCl in KCl. In the case of the biosensor in the microfluidic system, all three electrodes were thin film - gold WE and CE with Ag/AgCl RE. Switching between the two approaches can be done in a few minutes, as the fabricated chip is identical in both cases. Measurements with the biosensor in a reservoir are static, while the biosensor in a microfluidic system also allows for dynamic measurements in addition to static ones. The biosensor in the reservoir was used to optimise key steps in the biosensor development, such as assessing the purity of the thin-film gold electrodes, determining the appropriate deposition method for the mixed monolayer of self-assembled molecules (SAM), as well as determining the stability and shelf life of the biosensor.
In addition to the two configurations, I have also used so-called large and small electrodes. The large electrodes have a WE diameter of 3.0 mm and the small electrodes have a diameter of 0.8 mm. CE and RE are also shrunk by a similar amount. Additionally, in the case of the microfluidic biosensor, microchannels' shapes have been adapted to the electrode shapes. The length and height of the channel were 19.5 mm and 100 m, respectively, for both electrode sizes. For the larger electrodes the channels were 6.7 mm wide while the smaller ones had 1.2 mm wide channels.
The thin-film electrodes were fabricated using standard sputtering, photolithography and etching processes. In addition to the gold WE and CE, an Ag/AgCl RE was also produced by chemical and electrochemical processes. The most suitable method for further development of the biosensor turned out to be the FeCl3 drop-casting method, where the top layers of the silver electrode are converted into an AgCl layer. Afterfabrication of the electrodes, efficient cleaning of the electrodes is required for a reliable and a sensitive biosensor. We have cleaned the electrodes in three different ways: chemical cleaning with a solution of »piranha«, electrochemical cleaning with sulphuric acid in presence of current and cleaning with oxygen plasma. The latter method proved to be the most suitable for our biosensor, offering clean surface with the least degradation of the deposited Ag/AgCl layer. Indicators of electrode cleanliness and reversibility of the reaction were: the potential difference between the cathode and anode peak of the cyclic voltammogram after cleaning was 90 mV, the reduction and oxidation peaks of the cyclic voltammograms followed linearly the square root of the scan rate, and the ratio of the anode and cathode peak currents was on average equal to 0.98.
In addition to planar electrode construction, we have also tested the potential of nanostructuring the thin film WE to increase the effective surface area and thus the sensitivity. HAuCl4 and NH4Cl and pulsed amperometric techniques were used for nanostructuring. The fabricated nanostructures were observed by in-line electron microscopy. Needle-like aggregates of 10 mm size were observed, with individual needles of 500 nm. The effective surface area of the electrode, determined by cyclic voltammetry and the Randles-Ševčik equation, was increased by 41% by nanostructuring. However, all the further steps of biosensor construction were performed only on the planar electrodes.
The recognition element of the biosensor is PQQ-GdhB in the case of glucose detection and AChE in the case of eserine detection. PQQ-GdhB is immobilised on the gold electrode by 6-mercapto-1-hexanol acid (6-MCH) and 11-mercaptoundecanoic acid (11-MUA), and 1-ethyl-3-carbodiimide (EDC) and sulfo-N-hydroxysulphosuccinimide (S-NHS). AChE is immobilised on the gold electrode by cysteamine and glutaraldehyde. I compared two methods to immobilise the enzyme on the electrode. The simpler way is to immerse the whole electrode chip in the solutions. The process is simple and does not require additional components, but it does not offer selective deposition and it is slow. A more advanced method is micro-contact printing, which is performed with a polydimethylsiloxane (PDMS) stamp. Here, the deposition of the molecules is local, only on the WE. In addition, the deposition time is reduced to few minutes. Local deposition also has the advantage of allowing in situ immobilisation, i.e. immobilisation of the enzyme within the microchannel where the biosensor is positioned, in the later steps of the microfluidic biosensor fabrication. Enzyme can be immobilized immediately before the biosensor is used. In this way, the biological activity of the enzyme is also preserved as much as possible. The performance of the improved selective deposition method was verified electrochemically, very high coverage of the WE with minimal RE and CE contamination was observed.
The microfluidic part of the biosensor is fabricated by the standard photolithography, etching and soft lithography processes using PDMS elastomer for the channel walls.
PQQ-GdhB biosensor is based on the oxidation of glucose, whereby the glucose is converted to gluconolactone in the presence of GdhB. Reaction generates two electrons that are observable products, which are then transferred to the PQQ cofactor and from there to the mediator. The latter carries the electrons to the electrode where they are detected chronoamperometrically or by cyclic voltammetry. Meanwhile, the mechanism of the AChE biosensor is based on the hydrolysis of acetylthiocholine chloride (ATCl). The product of the reaction is thiocholine chloride (TChCl), which gives an electron to the mediator. Electrons are then transfered to the electrode. However, since we are interested in the concentration of the inhibitor that inhibits the enzyme, we determine the inhibition by the measured signal in the absence and presence of eserine.
Experiments show that the microfluidic electrochemical PQQ-GdhB biosensor has a measurement range of up to 10 mM glucose concentration, with a linear range of up to 200 M. The lower limit of detection is 30 M. The biosensor response follows Michaelis-Menten kinetics. The Michaelis constant is defined as 3.0 mM and 1.5 mM in the case of large and small electrodes. The biosensor sensitivity in the linear part is up to 0.79 nA/M/mm2 in the case of flow of 20 L/min. The stability of the developed biosensor was also evaluated – after 11 days the signal decreased about 45%. Preliminary measurements of the AChE biosensor, performed in the reservoir only, show that the chronoamperometric signal in response to different glucose concentrations follows Michaelis-Menten kinetics, with a Michaelis constant of 2.6 mM. Here, the measurement range is also limited to a 10 mM ATCl concentration. After exposing the biosensor to 25 M of eserine for 10 min, 70% inhibition of the enzyme was observed. The reactivation of the inihibited AChE was determined as 0.016 min-1. The measured parameters of both biosensors are comparable to the literature.
The developed approach of in situ immobilisation of the enzyme, together with the overall development of the microfluidic biosensor presented in this thesis, can be widely applied, e.g. with other biological recognition elements. Additionally, asthe glucose biosensor was primarily designed for glucose monitoring in pharmaceutical processes, further development offers a possibility of continuous and time-independent sensing. On the other hand, the AChE biosensor is developed with biomedical applications in mind. By replacing or adding other biosensing recognition elements the microfluidic system can be customised for other applications and fields. The device could be also used in the fields of medical diagnostics, pharmaceuticals, food industry, ecology, security and beyond.
|
