Extended-gate biosensors achieve fluid stability with no loss in charge sensitivity

The detection of toxic chemicals/biomolecules is of paramount importance for medical applications, environmental monitoring, food and pharmaceutical industries. Among the sensors available, FET based chemical/biosensors promise highly-sensitive, label-free detection for point-of-care applications. Further, compatibility with CMOS technology reduces costs and allows functional integration. Unfortunately, very poor reliability/stability of these sensors in the fluidic environment has been a key roadblock to the commercialization of technology. Fig. 1(a) shows the schematic of a conventional ISFET [1]: The gate oxide of the transistor is directly exposed to the ionic solution. Ions from the solution can penetrate into the gate oxide, causing voltage-dependent hysteresis [2] in the measured IV characteristics. Since, the sensing mechanism relies on the induced change in conductance/threshold voltage; this hysteresis can lead to false positives. Extended-gate FETs (Figs. 1(b) and 1(c)) promise to improve reliability by isolating the sensor (A_sensor) from the transducer (A_ox), connecting the two by an interconnect (A_int) [3][4]. The theory of pH-sensitivity (A_ISFET) for a classical ISFET is known from 1970s, however, a theoretical understanding of how the decoupling of sensor-transducer changes the sensitivity of an EGFET (S_EGFET) remains unknown. In this paper, we use detailed numerical simulations, compact modeling, and experimental results to conclude that regardless of the interconnect penalty, S_EGFET → A_ISFET with A_sensor ≫ A_ox.