Reliability Investigation and Application of High Performance Ion-Sensitive Field Effect Transistors

• Main Authors: Kuo-Yi Chao, 趙高毅
• Other Authors: Kow-Ming Chang
• Format: Others
• Language: en_US
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/80029062390039682802

博士 === 國立交通大學 === 電子工程系所 === 96 === This dissertation proposes to use zirconium oxide as a membrane of ion-sensitive field effective transistor (ISFET), and to control the effects of temperature and process deviation. When a stable sensitivity and intrinsic drift of SiO2 gate ISFET can be found, the thickness of hydration layer that is introduced by the drift effect will be found, too. After we understand the drift effect, we also propose two methods to solve the unwanted effect. In Chapter 2, theories and models of ISFET are described, including site-dissociation model, Gouy-Chapman theory, Gouy-Chapman-Stern theory, and drift model. Chapter 3 describes the application of the zirconium oxide (ZrO2) membrane as a pH-sensitive layer for ISFETs. It exhibited an excellent response range of 56.7~58.3 mV/pH from the fixed current measurement using HP4156A. The ZrO2 membrane prepared by direct current (DC) sputtering was used as a pH-sensitive film that showed good surface adsorption with oxide and silicon. The pH sensitivities slightly decreased in 1 M NaCl solution; however, the device showed a perfect linear response of 52.5 mV/pH. In Chapter 4, a reference field effective transistor (REFET) is used to control the effects of temperature and process deviation. After the calibration of REFET, a very stable sensitivity and intrinsic drift of SiO2 gate ISFET can be obtained. It can be used to define the thickness of hydration layer that is introduced by the drift effect. Results of this study will show that the thickness of hydration is about 50 nm in SiO2 membrane ISFET. It exhibits a stable response of 28~32 mV/pH from the fixed current measurement by HP4156A. This method is a really simple way to find the thickness of hydration layer, and it will be useful in the study of the real mechanism in drift effect. When we understand the phenomenon of drift, we try to design two methods to reduce this effect. One is in Chapter 5. A simple and cheap way to solve the drift problem is presented which describes the relation of drift and gate voltage. Constant various gate voltages are biased in sensing layers with reference electrode. It obviously shows a strong relation of gate drifts and gate stress voltages. When the gate voltage is controlled as 0.5 V, the drift voltage of SiO2 gate ISFET will decrease from 56.12 to 2.94 mV in ten hours measurement. The improvement of drift voltage reaches 94.8%. When the gate voltage is controlled as -1 V, the drift voltage of ZrO2 gate ISFET will also decrease from -57.94 to 0.76 mV. The improvement of drift voltage reaches 98.7%. Another one in Chapter 6 of the thesis is using REFET to reduce the drift effect. A simple CMOS compatible REFET for pH detection by post NH3 plasma surface treatment of a ZrO2 membrane ISFET has been developed. It is a novel study that has latent capacity to integrate the ISFET devices into a chemical micro system for in vivo analysis or become a part of lab-on-a-chip. With the fixed current measurement by HP4156, we can get not only the individual sensitivities of ISFET and REFET, but also the differential sensitivities of ISFET/REFET pair. The ZrO2 membrane ISFET exhibits an excellent response of 56.7~58.3mV/pH with deviation of 3% and the REFET shows a small response of 27.6~29 mV/pH with a deviation of 5%. Using this ISFET/REFET differential pair, we can get a very stable differential sensitivities of 29.1~29.3 mV/pH with a small deviation of 0.7%. This result indicates that the research not only makes the ISFET integrate into a micro system in a simple way possible, but also increases the stability of sensitivity.