The subproject, development of micro-sensing device, will be an important one for the integration in the main project, molecular imprinting micro-sensing chip. There is a critical problem in sensors, namely, suffering from poor selectivity. Although a bio-chemical sensor provides high selectivity, it is very sensitive to environmental conditions such as pH and temperature. Molecular imprinting technique is a promising technique to produce molecules that possess the ability of molecular recognition at low interference level. MIP is an approach of using the molecules or the molecular analogs or molecules which are complementary templates to the target molecules as the template. The general principle of MIP is a process for synthetic polymers where functional and cross-linking monomers are co-polymerized in the presence of the target analyte. Obviously, MIP is an art of artificial antibody. During the polymerization, the templates are mixed with the functional monomers. After the formation of the polymer, the template molecules are then extracted or purged out and hence specific caves can be formed. When the target molecules (usually are the same as the template molecules being removed) are presented to such polymers, the molecules can be selectively entrapped into the caves of these functional polymers^{(4 -7)}. By this way, the detection of the target molecules can be performed. The templates are attached to the polymer via either covalent or non-covalent bonds. In early researches, it is covalent-type MIPs which were investigated frequently. Comparing with non-covalent-type of MIPs, covalent-type MIPs can result in more homogeneous binding sites and higher yields of binding sites while non-covalent-type MIPs can be more flexible from their weaker binding forces. The templates of non-covalent type MIPs are very easily removed by solvents such as water. Nevertheless, non-covalent-type MIPs get more attention recently. Furthermore, a mixed covalent-noncovalent type of MIPs is also developed so that the polymers can contain the templates by covalent bonds while, at the same time, the templates can be washed out by non-covalent methods.

The development of templates for MIPs were from small molecules such as amino acids, chiral drugs, steroids, glucose, and aromatic compounds, medium-sized molecules such as peptides and enzymes to large molecules such as proteins, cells, and antibodies (also one kind of proteins). Smaller molecules can be more easily used to be the templates since their structures are not complex at all. Large molecules such as proteins not only have complex 3D conformations but also express by these conformations. As a result, using such large molecules for the templates can be very difficult and a lot of MIP related researches are focused on such molecules. The difficulties for protein templates may come from loss of expression during polymerization process and recognition mistakes with the other similar molecules. Nevertheless, the strategy to be established is finding the best recipe of monomers, cross-linkers and polymerization conditions for a particular combination of analyte and application. The template molecules chosen have a similar basic structure of a ring. In addition, with the ability for hydrogen bonding, the non-covalent type of MIPs can be possibly formed. This can be realized from the investigation on using amino acids in sequence as templates. Hydrogen bonding is a very important characteristic bonding force which can stabilize the amino acid sequence. In this technique, a template molecule (for example, CRP, cholesterol, bilirubin, and morphine) associates with one or more monomers (for example, MAA, HEMA, cyclodextrin) to form a complex. The complex is then polymerized with a cross-linker (for example, EGDMA) to produce a molecularly imprinted polymer (MIP). The MIPs produce an artificial antibody, which makes an absolute selectivity in sensing property. In principle, the molecular imprinting polymers are better than any biological molecules because the molecular imprinting film can characterize the protein and can endure environmental conditions. In this subproject, microfabrication technique will be utilized and MIFET is one of the transducer for this device.

When the MIPs carry with the analytes, they gain weights though the weight change can be very small. Hence, the mass sensitive transducers may be combined with MIP technique for the measurements of the analytes. However, there are a lot of transducers can also be chosen for the acquisition of the detection signals. These methods include SAW (surface-acoustic wave), QCM (quartz crystal microbalance), and ISFET (ion-sensitive field transistor). The first two methods are both very sensitive to mass change. However, the mass change is correlated to either wave shift or frequency shift. There have been plenty of successful QCM applications on sensors. However, only recently multi-channel QCM can also be miniaturized on a single chip. ISFET was proposed in 1970 by Bergveld. The concentrations of chemical quantities can be measured with a solid-state device. It requires a layer of ion-sensitive materials such as Si₃N₄, Al₂O₃, Ta₂O₅. It has advantages over the other methods such as small size, low cost, and robustness. Based on ISFET, a miniaturized urea sensor has been developed. The sensor for the measurement of living cells based on ISFET system is also reported. In this research, we prefer to use MISFET (molecular ion-sensitive field transistor) or QCM to be the transducer for MIP-analyte measurement. The easily formed noise at high frequencies is also one disadvantage from using QCM and should be solved. The signals from sensors carrying MIP will also be detected by MISFET approach. Molecular imprinting field effect transistor (MIFET) approach with microfabrication technique would certainly be the first choice in making a specific and reliable microsensor.

TOP

The methodology of the subproject, development of micro-sensing device, is shown in Figure 1. The methodology is divided into three parts. In the first part, MIP & ISFET materials will be prepared. The polymerization and deposition conditions will be optimized. The FET will be integrated with MIP using microfabrication techniques. The preparation and deposition of the MIPs as films over the microfabricated sensor sites will be discussed. In the second part, the characterization of MIPs will be carried out by analytical techniques such as AFM, SEM, NMR, in-situ IR, Raman, etc. The electrochemical and electrical characteristics will be evaluated by cyclic voltammetry, AC-impedance, spectroscopy, and I-V measurements. From these characterizations, fundamental mechanisms such as formation mechanism, incorporation and extraction mechanism, non-specific or specific binding mechanism, sensing mechanism, and aging mechanism will all be investigated. By the understanding of these mechanisms, the sensing performance such as sensitivity, selectivity, power consumption, etc., of the micro-sensing device will be improved.

The strategic plan for the micro-sensing development is shown in Figure 2. As shown in this figure, there are three items. The first item deals with MIPs. The organic and inorganic MIPs will be synthesized by Prof. T. Y. Lin, from KSUT and Prof. T. R. Ling, from ISU, respectively. The first item also addresses the mechanism of MIPs, for example, incorporation and extraction mechanisms. This work will be carried out by Prof. W. J. Chen, from CUT. The second item deals with integration of MIP with FET. The compatibility studies of MIP with FET will be performed by Prof. J. M. Ting, from NCKU. The last item includes the development of micro-sensors. The performance of several types of micro-sensing devices will be studied by Prof. T. C. Chou, Prof. K. C. Ho, Prof. M. C. Yang, and Prof. M. J. Syu. A possible configuration for a micro sensing system is shown in Figure 3. Finally, the developed micro-sensing device will be integrated with MIMSC in addition to micro-battery and signal/data processing system as well as telemetry.

TOP