This paper studies the effect of atomic layer deposition (ALD) temperature on the performance of top-down ZnO nanowire transistors. Electrical characteristics are presented for 10-μm ZnO nanowire field-effect transistors (FETs) and for deposition temperatures in the range 120°C to 210°C. Well-behaved transistor output characteristics are obtained for all deposition temperatures. It is shown that the maximum field-effect mobility occurs for an ALD temperature of 190°C. This maximum field-effect mobility corresponds with a maximum Hall effect bulk mobility and with a ZnO film that is stoichiometric. The optimized transistors have a field-effect mobility of 10 cm2/V.s, which is approximately ten times higher than can typically be achieved in thin-film amorphous silicon transistors. Furthermore, simulations indicate that the drain current and field-effect mobility extraction are limited by the contact resistance. When the effects of contact resistance are de-embedded, a field-effect mobility of 129 cm2/V.s is obtained. This excellent result demonstrates the promise of top-down ZnO nanowire technology for a wide variety of applications such as high-performance thin-film electronics, flexible electronics, and biosensing.

Abstract. In this work, we investigate how the sensitivity of a nanowire or nanoribbon sensor is influenced by the subthreshold slope of the sensing transistor. Polysilicon nanoribbon sensors are fabricated with a wide range of subthreshold slopes and the sensitivity is characterized using pH measurements. It is shown that there is a strong relationship between the sensitivity and the device subthreshold slope. The sensitivity is characterized using the current sensitivity per pH, which is shown to increase from 1.2%/pH to 33.6%/pH as the subthreshold slope improves from 6.2 V/dec to 0.23 V/dec respectively. We propose a model that relates current sensitivity per pH to the subthreshold slope of the sensing transistor. The model shows that sensitivity is determined only on the subthreshold slope of the sensing transistor and the choice of gate insulator. The model fully explains the values of current sensitivity per pH for the broad range of subthreshold slopes obtained in our fabricated nanoribbon devices. It is also able to explain values of sensitivity reported in the literature, which range from 2.5%/pH to 650%/pH for a variety of nanoribbon and nanowire sensors. Furthermore, it shows that aggressive device scaling is not the key to high sensitivity. For the first time, a figure-of-merit is proposed to compare the performance of nanoscale field effect transistor sensors fabricated using different materials and technologies.

Abstract-Biosensors are commonly produced using an SOI CMOS process and advanced lithography to define nanowires. In this work, a simpler and cheaper junctionless 3-mask process is investigated, which uses thin film technology to avoid the use of SOI wafers, in-situ doping to avoid the need for ion implantation and direct contact to a low doped polysilicon film to eliminate the requirement for heavily doped source/drain contacts. Furthermore, TiN is used to contact the biosensor source/drain because it is a hard, resilient material that allows the biosensor chip to be directly connected to a printed circuit board without wire bonding. pH sensing experiments, combined with device modelling, are used to investigate the effects of contact and series resistance on the biosensor performance, as this is a key issue when contacting directly to low doped silicon. It is shown that in-situ phosphorus doping concentrations in the range 4×10

Introduction: ZnO nanowire field effect transistors (NWFETs) can be fabricated using bottom-up or top-down approaches. Even though bottom-up devices may have better electrical characteristics; their orientation, dimensions and addressability are difficult to achieve. Therefore, top-down fabrication method is favorable for producing controlled nanowire dimensions and location. In a back gate FET configuration, the nanowire channel is exposed to the atmospheric environment. The surface charges between insulator, air and nanowire channel can deplete the channel surface to lower the output drain current and affect the electrical performances such as threshold shift -making the device to behave like enhancement mode [2]. This has also been observed in silicon-based nanowire FET [3]. It has been shown that passivation of ZnO TFT and bottom-up ZnO nanowire can improve electrical performance such as transconductance and mobility [4]. However, there is still little work done in the electrical performance of top-down spacer fabrication method ZnO nanowire under unpassivated and passivated conditions. In this work, we will investigate the effect of surface passivation of ZnO nanowire channel with Al2O3 using atomic layer deposition method and the electrical characteristics comparison to an unpassivated FET device.Fabrication: The ZnO NWFET is fabricated using our top-down spacer method [2], [3], which uses standard photolithography; remote plasma atomic layer deposition (RPALD) and anisotropic dry etch. RPALD process is conducted at 900 cycles, with each cycle consisting of 50 ms diethyl zinc (DEZ) precursor dose time, a 4 s Ar purge , a 2.65 s oxygen plasma, and a final 4 s Ar purge. The deposition temperature is done at 190 o C and the RF power and pressure were set at 100 W and 15 mtorr respectively. The deposited ZnO layer thickness is 76 nm and was measured using an ellipsometer. The natural n-type carrier concentration is 1 x 10 18 cm -3 and measured using a Hall Effect system. We use anisotropic inductively coupled plasma (ICP) etcher and CHF3 gas chemistry to remove the ZnO layer. Figure 1 (a) shows the top view microscope image of the ZnO NWFET with the source drain connection; and Figure 1(b) shows the cross section scanning electron micrograph of the ZnO nanowire. Then thermal ALD is used to deposit a layer of 18 nm thick Al2O3 over the ZnO nanowires at temperature of 200 o C. The aluminum layer is formed using trimethyl-aluminum (TMA) precursor and distilled ionized water is used for the oxidation process. Finally, aluminum lift-off process is used to form ohmic contact at the source and drain region. The fabricated NWFETs have channel length of 8.6 µm, thickness of 80 nm and width of 20 nm. The electrical I-V characterization of the ZnO NWFETs is done using Agilent B1500A semiconductor parametric analyzer.Results: Figure 2(a) shows the IdVd characteristics for un-passivated device with output drain current Id = 0.18 µA; whereas the passivated device has Id = 0.4 µA when mesured at Vd = 1 V and Vg = 20 V. Figu...

Abstract. We demonstrate the advantages of dual-gate polysilicon nanoribbon biosensors with a comprehensive evaluation of different measurement schemes for pH and protein sensing. In particular, we compare the detection of voltage and current changes when top-and bottom-gate bias is applied. Measurements of pH show that a large voltage shift of 491 mV/pH is obtained in the subthreshold region when the top-gate is kept at a fixed potential and the bottomgate is varied (voltage sweep). This is an improvement of 16 times over the 30 mV/pH measured using a top-gate sweep with the bottom-gate at a fixed potential. A similar large voltage shift of 175 mV is obtained when the protein avidin is sensed using a bottom-gate sweep. This is an improvement of 20 times compared with the 8.8 mV achieved from a topgate sweep. Current measurements using bottom-gate sweeps do not deliver the same signal amplification as when using bottom-gate sweeps to measure voltage shifts. Thus, for detecting a small signal change on protein binding, it is advantageous to employ a double-gate transistor and to measure a voltage shift using a bottom-gate sweep. For top-gate sweeps, the use of a dual-gate transistor enables the current sensitivity to be enhanced by applying a negative bias to the bottom-gate to reduce the carrier concentration in the nanoribbon. For pH measurements, the current sensitivity increases from 65% to 149% and for avidin sensing it increases from 1.4% to 2.5%.