< Back to previous page

Project

Low-dimensional electrical biosensors for labelfree multiplex biosensing

The amalgamation of technology with biology has resulted in a significant improvement both in the quality and the extent of human lifespan. This is more than evident in the current COVID‑19 situation when hospitals and pharmaceutical industry are able to rapidly test millions of people, provide intensive care to those in critical condition and accelerate research to screen for vaccines, thanks to the support provided by technology. However, there are still major challenges to further improve this integration. To be able to diagnose better and receive specifically designed treatments that match patients’ personal physiological profile, a move toward personalized medicine is needed. To enable this, we not only need sensor technology that can provide fast response time, high throughput, low limit of detection for bulk concentration sensing, but also improvements and innovation for single-molecule analysis such as DNA sequencing or proteomics.

Among several potential technologies, field-effect transistor (FET) based sensors stand out for such applications. FET-based biosensors (or bioFETs) have been claimed to show very short response times ranging from a few seconds to a few minutes and extremely low limit of detection (LOD) down to attomolar (10^-18 M) bulk biomolecular concentration. Additionally, as FET-based sensors can be fabricated using CMOS technology, they can be integrated to form a dense multiplexed array for high throughput applications. The origin of this high bulk concentration sensitivity is controversial and has mostly been attributed to high surface-to-volume ratio of the nanowire/nano-FETs. This has led to scaling down of the width and the height of FET-based sensors to achieve a higher surface-to-volume ratio. However, this argument is only qualitative and does not support the orders of magnitude reduction in LOD claimed in the nanowire/nano-FET papers. Additionally, some contradictory reports have also been published raising questions about high FET sensitivity and short response times. This points towards several inconsistencies in the literature and a lack of understanding for FET-based sensing mechanism.

This thesis aims to understand the origins of the high sensitivities and the fast sensor response observed for FET-based sensors for bulk biomolecular sensing. The intrinsic transducer sensitivity is often confused with total sensor sensitivity for bulk concentration sensing, which not only includes the transducer response but also the biomolecular association kinetics and signal post-processing and amplification aspects. In this work we isolate the transducer performance from the biomolecular association kinetics and investigate separately the FET transduction performance and the contribution of biomolecular association kinetics, to understand the claims made in the literature.

We first start by examining biomolecular association kinetics. When the binding regions of interacting biomolecules are oppositely charged, long-range electrostatic interactions can increase association rate constants resulting in an enhancement in capture kinetics with a reduction in ionic strength. We investigate to what extent the unintentional side-effect of using low ionic strength buffers for bioFETs measurement plays a role in their exceptional performance i.e., high sensitivity and enhanced capture kinetics. We observe an enhancement in association kinetics when using low ionic strength buffer, however, the magnitude of enhancement observed was not of the order (10² to 10³ enhancement with 10x drop in ionic strength) that was expected based on the nanowire/FET based sensing results. This could be a consequence of parasitics such as denaturing of the selected set of biomolecules at low ionic strength or falling into mass transport limit resulting in slowing down of interaction kinetics.

On the transduction aspect, we investigate the claim of high surface-to-volume ratio as the reason for high FET sensitivity motivating researchers to scale down FET size for higher sensitivities. We experimentally investigate pH sensing using FETs for two different oxide surfaces and for various FET dimensions and observe that the pH sensitivity is independent of FET width (down to 10 nm) as well as FET length (down to 145 nm). We also experimentally investigated the FET as a transducer for bulk biomolecular sensing using PNA-DNA hybridization system for multiple FET dimensions and find that FET voltage transduction sensitivity to be independent of FET width (down to 60 nm) as well as FET length (down to 450 nm). Furthermore, we also discuss the parasitics associated with FET-based sensing resulting in size-dependent sensitivities as obtained in the literature.

FET transduction is significantly affected by the screening of analyte charge from the bulk electrolyte ions and is typically discussed in terms of the linear Debye-Huckel screening model. However, much less attention has been given to the phenomenon of increased non-linear electrolytic screening when an electric double layer (EDL) is present at the sensor surface, e.g., in response to surface charges, or even simply when the generated bio-signal exceeds a few kT/q (= 25 mV). We investigate this phenomenon of localized excess screening of biomolecular charge originating from fixed surface charge and from pH-sensitive surface states. We find that maximum sensitivity for bulk biomolecular sensing is obtained when the initial sensor surface is charged approximately half and opposite in polarity of the expected biomolecular charge, such that the EDL potential swings roughly symmetrically around zero. In case of a small biomolecular density or even single-molecule sensing, however, the analyte itself does not significantly influence the EDL, and the sensor surface potential should be close to zero mV (i.e., an uncharged surface), allowing the ion concentration in the EDL to be the same as the bulk ionic strength.

For bulk biomolecular sensing applications, since larger transistors exhibit the same voltage transduction sensitivity as smaller transistors for the same surface charge density, using large transistors will help in reducing the fabrication costs, device noise and process complexities associated with nanoFET fabrication. For single-molecule analysis applications (e.g., DNA sequencing), going to extremely small dimensions may be advantageous, as the effective biomolecular coverage for the given sensor area increases for small sensors, leading to a gain in the signal. However, to enable single-molecule sensing using FETs, one should also consider other aspects, e.g., optimizing sensor surface charge to improve signal transduction, minimizing parasitics associated with measurements, improving surface chemistry for selective biomolecular functionalization, and other such aspects. The combined improvement could result in highly sensitive FET-based sensors, optimum for single molecular analysis to further advance the study of the genome and proteome for personalized medicine.

Date:1 Feb 2016 →  11 Jan 2021
Keywords:biosensor
Disciplines:Condensed matter physics and nanophysics
Project type:PhD project