In recent years, as biological information analysis technology rapidly develops in hematology, biochemistry and microbiology areas, demand for portable measurement systems become more and more important. This study makes efforts in developing disposable hydrogen and potassium ion sensor and microsystem for analysis application. The measured ion concentration data by this analysis microsystem provide people a judgement on their health condition, and furthermore an important reference for medical treatment for patients. There are several advantages in using IC or MEMS technology to manufacture portable measurement system, the advantages are down-scaling, short reaction time, trace chemical analysis, low power dissipation, and low cost. So the thesis uses extended gate field effect transistor, in order to measure multiple ions at the same time, multiple transistors are manufactured on the same chip with an ion selective membrane on top of the gate sensitive layer. This allows the measurement result of the multiple ion be shown at the same time. The main processing steps of the ion sensor developed in this study involve at least four photolithographic and three thin-film deposition processes. Based on the measurement result, the hydrogen ion sensor¡¦s sensitivity is 30.7 mV/decade for a sensing range pH1 ~ pH13. The sensitivity of the potassium ion sensor is 11.5 mV/decade for a sensing range 10-1M to 10-3M.

Human blood reflects the health of internal organs and tissues, the constituent can be affected on account of abnormal health status. Moreover, the ion concentration of chloride and ammonium in human blood relates to the functionality of our kidney and liver, which is one of the most important health indicators. For real-time monitoring and detecting applications, this study uses micro-electromechanical systems (MEMS) surface micromachining process technology to develop an extended-gate field-effect transistors (EGFET), and by combining two different ion-selective membranes (ISM), a disposable silicon-based chloride and ammonium micro ion sensor was fabricated. The main process steps include four photolithography process and two thin film deposition. In order to probe into the influences on modulating three designing parameters: (i) channel width to length ratio; (ii) channel shape; (iii) mixing ratio of the sensing film additives, the size of the sensing area is fixed to 1¡Ñ1 mm2 while the device is fixed to 6 mm3. In this thesis, the optimization of the development chloride and ammonium micro ion sensors, according to the measuring results from the commercial semiconductor analyzer (Agilent B1500A), when measuring range of 10-1~10-5 mol/L, the sensitivity and linearity for the chlorine ion sensor is 47.5 mV/pCl and 99.13%, as for the ammonium ion sensor are 41.2 mV/pNH4 and 99.28%, the interfering ion selectivity coefficients (log KCl,OH, log KNH ,Na) are -4.71 and 0.53 respectively.

Water resource is one of the most important natural resources on earth. In recent years, due to the discharges of large industrial and domestic waste-water into the nature, water pollution problem is getting more and more serious and how to monitor the quality of water in real time has become a very important research issue. The dissolved oxygen is one of the critical indexes for evaluating the quality of water. Although the conventional dissolved oxygen detectors presented a high sensitivity and high accuracy, the high cost, large dimension, low capability of batch fabrication and real-time monitoring will limit their applications. In this thesis, an extended-gate field-effect transistor (EGFET) based dissolved oxygen microsensor is developed utilizing micro-electromechanical system (MEMS) technology. The gate voltages of EGFET under different concentrations of dissolved oxygen can be detected by the Cr/Au sensing electrode. To further enhance the sensitivity of the proposed microsensor, a polystyrene layer with very high permeation rate of the dissolved oxygen gas is adopted and coated on the surface of Cr/Au layer. The main processing steps of the presented microsensor involve four photolithographic and four thin-film deposition processes. The influence of the channel¡¦s width/length ratio, source/drain geometry and polystyrene additional layer on the sensitivity of the EGFET based dissolved oxygen microsensor are investigated in this study. The chip size of the implemented dissolved oxygen microsensor is 11 mm¡Ñ13 mm¡Ñ 0.5 mm and the sensing area is 1 mm¡Ñ1 mm. As the dissolved oxygen concentration varies from 2 ppm to 6 ppm, a very high sensitivity (35.36 mV/ppm) and sensing linearity (98.83%) of the proposed EGFET microsensor can be demonstrated. In addition, the response time of the presented dissolved oxygen microsensor is only about III 180~200 seconds, hence it is very suitable for developing a real-time monitoring microsystem.

The large carbon dioxides produced by highly developed industries not only result in serious air pollution and health problems, but also cause ocean acidification and decrease the survival rate of fry in aquaculture. Therefore, to develop a system for real-time detection of the concentration of carbon dioxide in aquaculture has become a very important research issue. Optical analysis and gas-chromatography are the two main methods adopted in conventional gas detection. Although the conventional carbon dioxide detectors presented high sensitivity and accuracy, the high fabrication cost, large dimension, low capability of batch fabrication and without real-time monitoring function will limit their applications. This thesis utilizes MEMS technology to implement an extended-gate field-effect transistor (EGFET) with an integrated gas permeable membrane for development of a high-sensitivity, small size and low cost carbon dioxide microsensor. The main material of the carbon dioxide gas permeable membrane adopted in this research is dioctyl sebacate. The main processing steps of the proposed microsensor include four photolithography and four thin-film deposition processes. In addition, the influences of the channel width/length ratio of EGFET and the coating of gas permeable membrane on the sensing performances of presented microsensor are also investigated in this study. The chip size of the implemented carbon dioxide microsensor is 11 mm¡Ñ13 mm¡Ñ 0.5 mm and the sensing area is 1 mm¡Ñ1 mm. As the carbon dioxide concentration varies from 0.25 mM to 50 mM, a very high sensitivity (42.3 mV/ppm) and sensing linearity (99.2%) of the proposed EGFET microsensor can be demonstrated. In addition, the response time of the presented carbon dioxide microsensor is only about 100 seconds, hence it is very suitable for developing a real-time monitoring microsystem.

As working time increases for most people, dining out and staying up late is inevitable, resulting in bad health conditions. The concentrations of calcium and sodium ion in human blood not only respond directly to health conditions, but can also obtain symptoms of different diseases by observing it. This shows that the concentrations of calcium and sodium ion in human blood are an important index of health. In order to manufacture disposable ion sensor and make it easy to measure, this study uses extended gate field effect transistor (EGFET) with an ion selective membrane(ISM) on top of the gate sensitive layer to replace traditional ion sensitive field effect transistor (ISFET). The ISM adsorbs the appointed ion by means of ion selective medicament which is covered by a macromolecule. The main processing steps of the extended gate field effect transistor developed in this study involve at least four photolithographic and two thin-film deposition processes. The influence of the channel¡¦s width to length ratio, the design of channel, the area of the gate sensitive layer, the energy and dose of ion implantation used for the transistor and ion sensor were investigated. Based on the measurement results of the ion sensor, a sensitivity of 40mV/decade with linearity of 98.589 % is measured for calcium ion concentration in human blood ranging from 5 ¡Ñ 10-3 mol/L to 5 ¡Ñ 10-4 mol/L. On the other hand, a sensitivity of 56 mV/decade with linearity of 98.589 % is measured for sodium ion concentration in human blood ranging from 1 mol/L to 10-1 mol/L.