Short communicationSensitivity and detection limit analysis of silicon nanowire bio(chemical) sensors
Introduction
With the rising popularity of silicon nanowire (Si-NW) field-effect (FET) biosensors, understanding their physical behavior is important for systematically improving their sensitivity and limit of detection (LOD). Si-NW FET sensors are different than their planar predecessors [1] due to their nanoscale three-dimensional geometry, which is important for chemical and biochemical sensors. The three-dimensional geometry provides a multi-gate interface with a large surface area to volume ratio, which can render them extremely sensitive to the binding of small quantities of target species on the sensor surface [2], [3], [4]. Additionally, the nanoscale multi-gate sensing surface requires a smaller number of molecules to generate the same gate potential change compared to the macroscale planar sensors.
A typical Si-NW biosensor system is comprised of a multi-gate Si-NW FET biosensor, reference electrode and power supply (Vfg) for front-gate control (fg), power supply (Vbg) for back-gate (bg) control, power supply (Vds) for the NW bias current control, and a current measurement instrument (I), as shown in Fig. 1(a). Multi-gate Si-NW FET biosensors are usually configured in a dual gate configuration, indicated as fg and bg in Fig. 1(a). The bg controls the back gating of the sensor body, and the fg controls the front gating of the upper sensor surface in contact with the electrolyte solution. Surface reactions at the gate-oxide surface, due to electrolyte ions, or the binding of charged target molecules, result in a change in the surface charge density Δσ and subsequent surface potential change Δψo, which induce a change in the Si-NW current ΔI. The majority of Si-NW FET biosensors use a p-type doped body and the operating region of the depletion-mode devices depends on the body doping and the bias voltages Vfg and Vbg.
Fig. 1(b) shows typical I–V curves from depletion-mode Si-NW sensors with triangular cross-section (length 7 μm, height 100 nm), developed by our research group, and measured in a 0.1 M NaCl supporting electrolyte with a 10 mM universal buffer mixture (10 mM citric acid, 10 mM phosphoric acid, and 20 mM boric acid, pH 7.0) using a reference electrode (REF200, Ag/AgCl, Radiometer Analytical) for liquid gating. In these measurements, the sensor current I is measured with a lock-in amplifier instrument (SR830, Stanford Research Systems), however, a DC power source for Vds is used for the rest of the noise measurements unless mentioned specifically. Four different operation regions are indicated in the measured I–V curves. For VT ≤ Vfg ≤ Voff, where Voff is the pinch-off voltage and VT is the threshold voltage (Fig. 1(b), inset), the devices operate in the weak current region, sometimes referred to as a subthreshold region [5]. For Vfb ≤ Vfg ≤ VT, where Vfb is the applied gate voltage necessary to induce a flat energy band at the silicon surface, the current decreases from the non-depleted flat-band condition (Vfg = Vfb). For Vfg < Vfb, the total device current is due to a combination of the majority carriers in the device body and the accumulation surface layer.
Section snippets
Methodology
We previously reported an analytical model that describes the current of a Si-NW in the depletion and accumulation regions of operation [1]. The total current is the sum of the current of the device body and the accumulated majority carriers at the upper sensor surfaces I = Id + Ia, which depends on the region of operation. The depletion current can be approximated with Id ≈ qμbVdsNaL−1Ac(fd, a), where q is the electronic charge, μb is the majority carrier mobility in the NW body, Na is the NW body
Results and discussion
Since the LOD of the sensor system is determined by the transconductance of the sensor and the noise of the measurement system, we now determine both characteristics of the Si-NW sensor system. The transconductance is a function of the fg bias gm(Vfg) and can be directly calculated from the sensor I–V characteristics, as shown in Fig. 2(a). The noise characteristics of ion-sensitive field effect transistor (ISFET) sensors have been previously studied, and it was determined that the ISFET noise
Conclusions
In conclusion, we have presented an analysis of the sensitivity and LOD of Si-NW sensor systems. Using the simple analytical model, the LOD can be estimated with a few fitting parameters, or by directly using I–V and noise measurements. Sensor operation in the high transconductance region is preferred for a low LOD, and equivalently high SNR. The LOD can be improved by increasing the surface to square root of volume ratio.
Acknowledgements
This work was supported by NanoNextNL, a nanotechnology consortium with 130 research partners funded by the Government of the Netherlands.
Songyue Chen received her M.Sc. in biomedical engineering at Zhejiang University, China, in 2006 and at University of Twente, the Netherlands, in 2007, and her Ph.D. at University of Twente in 2011. She is currently a postdoc in MESA+ Institute for Nanotechnology, and her main research interests are silicon nanowire biosensors.
References (23)
- et al.
Detection limits of chemical sensors: applications and misapplications
Sens. Actuators B: Chem.
(2012) - et al.
Al2O3/silicon nanoISFET with near ideal Nernstian response
Nano Lett.
(2011) - et al.
Direct ultrasensitive electrical detection of DNA and DNA sequence variations using nanowire nanosensors
Nano Lett.
(2004) - et al.
Silicon nanowire arrays for label-free detection of DNA
Anal. Chem.
(2007) - et al.
A silicon nanowire ion-sensitive field-effect transistor with elementary charge sensitivity
Appl. Phys. Lett.
(2011) - et al.
All-(1 1 1) surface silicon nanowires: selective functionalization for biosensing applications
Appl. Mater. Interfaces
(2010) - et al.
Electrochemical biosensors: recommended definitions and classification
Pure Appl. Chem.
(1999)
Electronic detection of DNA by its intrinsic molecular charge
Proc. Natl. Acad. Sci. U.S.A.
Cited by (23)
Amyloid-beta peptide detection via aptamer-functionalized nanowire sensors exploiting single-trap phenomena
2020, Biosensors and BioelectronicsCitation Excerpt :In these measurements, the drain current was recorded by sweeping liquid-gate voltage, while the drain-source voltage was kept constant at −20 mV to provide the linear working regime of the transistor structure. The device demonstrated transfer characteristics behavior typical of p+-p-p+ FETs (Bedner et al., 2014; Chen et al., 2015; Pud et al., 2014). It should be noted that the registered leakage current IG ≤ 0.1 nA was at least three orders of magnitude lower than the drain current within the entire range of the liquid-gate voltages applied.
Performance analysis of solid-state nanopore chemical sensor
2019, Sensors and Actuators, B: ChemicalCitation Excerpt :The current power spectral density (PSD) can be expressed with Hooge’s empirical equation SI ≈ Io2(α/N)(1/f)β [20,28], where α is the dimensionless constant (an empirically determined proportionality constant that depends on the type and concentration of charge carriers), β is the frequency index, and N is the total number of charge carriers in the channel. The rms current noise amplitude în ≈ (SIΔf)1/2, where for the remained analysis Δf = 1 Hz is chosen [29]. Therefore, the LOD of surface charge density change can be expressed with σLOD = 3SI1/2/Vo gm.
Covalent functionalization of polycrystalline silicon nanoribbons applied to Pb(II) electrical detection
2018, Sensors and Actuators, B: ChemicalCitation Excerpt :One of its deposition technique is Low-Pressure Chemical Vapor Deposition (LPCVD) which can provide good homogeneity of the silicon in terms of thickness, doping concentration and stability. Most electronic sensors are based on field effect transistors (FETs) which have proven high sensitivity and electrical performances [11–13]. However, they require several photolithography processes, plasma etching and deposition steps that can provide reproducibility drawbacks.
Nanotube-on-graphene heterostructures for three-dimensional nano/bio-interface
2018, Sensors and Actuators, B: ChemicalCitation Excerpt :Semiconductor nanomaterials with controlled structures, such as nanowires (NWs), nanotubes (NTs), and graphene have been developed to overcome such challenges and record the properties of biological species with a higher resolution [4–7]. One or several dimensions of the nanomaterials match the length-scale of target biological species and could provide information about electrical signal propagation in biological systems at a far higher spatial resolution than previously demonstrated [8–11]. Furthermore, the microfabricated device arrays can be readily fabricated on flexible polymer substrates, which enable minimally invasive and gentle contact of device chips with non-planar biological systems [12,13].
Design and numerical analysis: Effect of core and cladding area on hybrid hexagonal microstructure optical fiber in environment pollution sensing applications
2017, Karbala International Journal of Modern ScienceCitation Excerpt :But the sensor has a detection limit. The detection limit (LOD) of sensor depends on cross-sectional geometry and size [8]. Besides sensitivity, many different uniform properties like large effective area [9], nonlinearity [10,11], high birefringence [2,6], absolutely single polarization [12], and endlessly single mode [13] are found from PCF.
Songyue Chen received her M.Sc. in biomedical engineering at Zhejiang University, China, in 2006 and at University of Twente, the Netherlands, in 2007, and her Ph.D. at University of Twente in 2011. She is currently a postdoc in MESA+ Institute for Nanotechnology, and her main research interests are silicon nanowire biosensors.
Albert van den Berg received his M.Sc. in applied physics in 1983, and his Ph.D. in 1988 both at the University of Twente, the Netherlands. He is currently a full professor on Miniaturized Systems for (Bio)Chemical Analysis in the Faculty of Electrical Engineering. He is member of the Dutch National Health Council and became a board member of the Royal Dutch Academy of Sciences (KNAW) in 2011. His current research interests focus on microanalysis systems and nanosensors, nanofluidics and single cells, tissues and organs on chips, especially with applications in personalized health care, drug development and development of sustainable (nano)technologies.
Edwin T. Carlen received the Ph.D. from the University of Michigan, USA in 2001. In 2006, he was a visiting scholar at the Institute of Physics, Leiden University, the Netherlands and later that year joined the scientific staff of the BIOS-Lab on a Chip Group and MESA+ Institute for Nanotechnology, University of Twente, the Netherlands, as a 3TU assistant professor. He was promoted to the rank of associate professor in 2011. Currently, he is an associate professor in the Graduate School of Pure and Applied Sciences at University of Tsukuba, Japan, and leads a research group developing plasmonic devices and nanosensors.