Wangyang Fu

Nernst Limit in Dual-Gated Si-Nanowire FET Sensors

Field effect transistors (FETs) are widely used for the label-free detection of analytes in chemical and biological experiments. Here we demonstrate that the apparent sensitivity of a dual-gated silicon nanowire FET to pH can go beyond the Nernst limit of 60 mV/pH at room temperature. This result can be explained by a simple capacitance model including all gates. The consistent and reproducible results build to a great extent on the hysteresis- and leakage-free operation. The dual-gate approach can be used to enhance small signals that are typical for bio- and chemical sensing at the nanoscale.

Graphene transistors are insensitive to pH changes in solution.

We observe very small gate-voltage shifts in the transfer characteristic of as-prepared graphene field-effect transistors (GFETs) when the pH of the buffer is changed. This observation is in strong contrast to Si-based ion-sensitive FETs. The low gate-shift of a GFET can be further reduced if the graphene surface is covered with a hydrophobic fluorobenzene layer. If a thin Al-oxide layer is applied instead, the opposite happens. This suggests that clean graphene does not sense the chemical potential of protons. A GFET can therefore be used as a reference electrode in an aqueous electrolyte. Our finding sheds light on the large variety of pH-induced gate shifts that have been published for GFETs in the recent literature.

Low-bias active control of terahertz waves by coupling large-area CVD graphene to a terahertz metamaterial

We propose an hybrid graphene/metamaterial device based on terahertz electronic split-ring resonators directly evaporated on top of a large-area single-layer CVD graphene. Room temperature time-domain spectroscopy measurements in the frequency range from 250 GHz to 2.75 THz show that the presence of the graphene strongly changes the THz metamaterial transmittance on the whole frequency range. The graphene gating allows active control of such interaction, showing a modulation depth of 11.5% with an applied bias of 10.6 V. Analytical modeling of the device provides a very good qualitative and quantitative agreement with the measured device behavior. The presented system shows potential as a THz modulator and can be relevant for strong light-matter coupling experiments.

Understanding the Electrolyte Background for Biochemical Sensing with Ion-Sensitive Field-Effect Transistors

Silicon nanowire field-effect transistors have attracted substantial interest for various biochemical sensing applications, yet there remains uncertainty concerning their response to changes in the supporting electrolyte concentration. In this study, we use silicon nanowires coated with highly pH-sensitive hafnium oxide (HfO(2)) and aluminum oxide (Al(2)O(3)) to determine their response to variations in KCl concentration at several constant pH values. We observe a nonlinear sensor response as a function of ionic strength, which is independent of the pH value. Our results suggest that the signal is caused by the adsorption of anions (Cl(-)) rather than cations (K(+)) on both oxide surfaces. By comparing the data to three well-established models, we have found that none of those can explain the present data set. Finally, we propose a new model which gives excellent quantitative agreement with the data.

True Reference Nanosensor Realized with Silicon Nanowires

Conventional gate oxide layers (e.g., SiO(2), Al(2)O(3), or HfO(2)) in silicon field-effect transistors (FETs) provide highly active surfaces, which can be exploited for electronic pH sensing. Recently, great progress has been achieved in pH sensing using compact integrateable nanowire FETs. However, it has turned out to be much harder to realize a true reference electrode, which--while sensing the electrostatic potential--does not respond to the proton concentration. In this work, we demonstrate a highly effective reference sensor, a so-called reference FET, whose proton sensitivity is suppressed by as much as 2 orders of magnitude. To do so, the Al(2)O(3) surface of a nanowire FET was passivated with a self-assembled monolayer of silanes with a long alkyl chain. We have found that a full passivation can be achieved only after an extended period of self-assembling lasting several days at 80 °C. We use this slow process to measure the number of active proton binding sites as a function of time by a quantitative comparison of the measured nonlinear pH-sensitivities to a theoretical model (site-binding model). Furthermore, we have found that a partially passivated surface can sense small changes in the number of active binding sites reaching a detection limit of δN(s) ≈ 170 μm(-2) Hz(-1/2) at 10 Hz and pH 3.

Selective sodium sensing with gold-coated silicon nanowire field-effect transistors in a differential setup.

Ion-sensitive field-effect transistors based on silicon nanowires with high dielectric constant gate oxide layers (e.g., Al2O3 or HfO2) display hydroxyl groups which are known to be sensitive to pH variations but also to other ions present in the electrolyte at high concentration. This intrinsically nonselective sensitivity of the oxide surface greatly complicates the selective sensing of ionic species other than protons. Here, we modify individual nanowires with thin gold films as a novel approach to surface functionalization for the detection of specific analytes. We demonstrate sodium ion (Na(+)) sensing by a self-assembled monolayer (SAM) of thiol-modified crown ethers in a differential measurement setup. A selective Na(+) response of ≈-44 mV per decade in a NaCl solution is achieved and tested in the presence of protons (H(+)), potassium (K(+)), and chloride (Cl(-)) ions, by measuring the difference between a nanowire with a gold surface functionalized by the SAM (active) and a nanowire with a bare gold surface (control). We find that the functional SAM does not affect the unspecific response of gold to pH and background ionic species. This represents a clear advantage of gold compared to oxide surfaces and makes it an ideal candidate for differential measurements.

Signal-to-noise ratio in dual-gated silicon nanoribbon field-effect sensors

Recent studies on nanoscale field-effect sensors reveal the crucial importance of the low frequency noise for determining the ultimate detection limit. In this letter, the 1/f-type noise of Si nanoribbon field-effect sensors is investigated. We demonstrate that the signal-to-noise ratio can be increased by almost two orders of magnitude if the nanoribbon is operated in an optimal gate voltage range. In this case, the additional noise contribution from the contact regions is minimized, and an accuracy of 0.5% of a pH shift in one Hz bandwidth can be reached.

High-yield fabrication of nm-size gaps in monolayer CVD graphene

Herein we demonstrate the controlled and reproducible fabrication of sub-5 nm wide gaps in single-layer graphene electrodes. The process is implemented for graphene grown via chemical vapor deposition using an electroburning process at room temperature and in vacuum. A yield of over 95% for the gap formation is obtained. This approach allows producing single-layer graphene electrodes for molecular electronics at a large scale. Additionally, from Raman spectroscopy and electroburning carried out simultaneously, we can follow the heating process and infer the temperature at which the gap formation happens.

High mobility graphene ion-sensitive field-effect transistors by noncovalent functionalization

Noncovalent functionalization is a well-known nondestructive process for property engineering of carbon nanostructures, including carbon nanotubes and graphene. However, it is not clear to what extend the extraordinary electrical properties of these carbon materials can be preserved during the process. Here, we demonstrated that noncovalent functionalization can indeed delivery graphene field-effect transistors (FET) with fully preserved mobility. In addition, these high-mobility graphene transistors can serve as a promising platform for biochemical sensing applications.

Large-scale fabrication of BN tunnel barriers for graphene spintronics

We have fabricated graphene spin-valve devices utilizing scalable materials made from chemical vapor deposition (CVD). Both the spin-transporting graphene and the tunnel barrier material are CVD-grown. The tunnel barrier is realized by Hexagonal boron nitride, used either as a monolayer or bilayer and placed over the graphene. Spin transport experiments were performed using ferromagnetic contacts deposited onto the barrier. We find that spin injection is still greatly suppressed in devices with a monolayer tunneling barrier due to resistance mismatch. This is, however, not the case for devices with bilayer barriers. For those devices, a spin relaxation time of ∼260 ps intrinsic to the CVD graphene material is deduced. This time scale is comparable to those reported for exfoliated graphene, suggesting that this CVD approach is promising for spintronic applications which require scalable materials.