Sensitivity Limits and Scaling of Bioelectronic Graphene Transducers

Abstract

Semiconducting nanomaterials are being intensively studied as active elements in bioelectronic devices, with the aim of improving spatial resolution. Yet, the consequences of size-reduction on fundamental noise limits, or minimum resolvable signals, and their impact on device design considerations have not been defined. Here, we address these key issues by quantifying the size-dependent performance and limiting factors of graphene (Gra) transducers under physiological conditions. We show that suspended Gra devices represent the optimal configuration for cardiac extracellular electrophysiology in terms of both transducer sensitivity, systematically ~5× higher than substrate-supported devices, and forming tight bioelectronic interfaces. Significantly, noise measurements on free-standing Gra together with theoretical calculations yield a direct relationship between low-frequency 1/f noise and water dipole-induced disorders, which sets fundamental sensitivity limits for Gra devices in physiological media. As a consequence, a square-root-of-area scaling of Gra transducer sensitivity was experimentally revealed to provide a critical design rule for their implementation in bioelectronics.

Determination of the Thermal Noise Limit of Graphene Biotransistors

“…As expected, we observe that Z re (and therefore S V,th ) scales inversely with A. Previous measurements of noise in GFETs, 5,6,16 and related systems such as metal electrodes, 8 show the same trend. The relationship S V ∝ A -1 summarizes the trade-off between voltage resolution and spatial resolution.…”

Charge Noise in Organic Electrochemical Transistors

“…Since S V g ¼ S q =C 2 gate , the scaling of S V g with geometry depends on the corresponding scaling of the charge noise S q and the gate capacitance C gate . Our experimental data could be explained by assuming that both C gate and S q scale only with the area leading to S V g ∝ 1=wl, as is commonly accepted for liquid-gated CNTs and graphene transistors [19,[25][26][27][28]. However, the absence of a clear interface between the polymer channel and the electrolyte might lead to a different geometry scaling of S V g .…”

“…The charge-noise model was successfully applied to describe 1=f noise measured for liquid-gated SWCNTs [27,28] and single-bilayer graphene [19,25,26] where the channel material is in direct contact with the electrolyte. The observed 1=f noise is associated with charge fluctuations caused either by trap states in the substrate for substrate-bound SWCNTs [27,28] and graphene devices [19,25] or by the Brownian motion of ions of the electrolyte for suspended carbon nanotubes (CNTs) [28].…”

“…In biosensing, since the binding kinetics of many species of interest (e.g., proteins) typically require time scales up to a few minutes [21], the 1=f noise becomes a key parameter limiting the performance of the sensor. For this reason, lowfrequency noise has been studied in depth for several transistor-based biosensors, such as silicon nanowires (SiNWs) operated as ion-sensitive field-effect transistors [20,22,23], diamond solution-gated field-effect transistors [24], liquid-gated graphene [19,25,26], and single-walled carbon-nanotube (SWCNT) transistors [27,28]. While some studies on OFETs exist [29,30], to the best of our knowledge, no measurements have been reported so far for OECTs.…”

Nano-Bioelectronics

“…Moreover, their work showed that the fluctuations of water ultimately set the fundamental sensitivity limit of graphene-based bioelectronics. 546 …”