key: cord-0725966-saqx72f9
authors: Biswas, Arpita; Rajan, Chithraja; Samajdar, Dip Prakash
title: Sensitivity Analysis of Physically Doped, Charge Plasma and Electrically Doped TFET Biosensors
date: 2021-10-21
journal: Silicon
DOI: 10.1007/s12633-021-01461-1
sha: 3a0580bac149b21b185449245809664626367a46
doc_id: 725966
cord_uid: saqx72f9

TFET based label-free biosensors are fast, sensitive and more power efficient as compared to CMOS biosensors, which are prone to short channel effects (SCEs). However, literature is flooded with various TFET biosensors that have become the reason of dilemma for researchers during pandemic situations like COVID-19. Therefore, in this work, a physically doped (PD), charge plasma (CP) and electrically doped (ED) dielectric modulated (DM) TFET based label-free biosensors are compared, which cover almost the entire range of doping and junctionless devices. Also, we found that the ED based TFET biosensors provide better current sensitivities of 5.10 × 10(7), 4.77 × 10(8) and 7.11 × 10(8) for biomolecules with K=12, positive charge= 1 × 10(13) C/cm(2) and negative charge= -1 × 10(13) C/cm(2) respectively. Hence, ED-DM-TFET based biosensors can act as promising candidates to provide better detection and identification quality.

The common human coronavirus (CoV), which infects millions of people every year are basically of two types: alpha (229E, NL63) and beta (OC43, HKU1, SARS-CoV, MERS-CoV). SARS (severe acute respiratory syndrome) CoV affected 8000 people in 2003 and MERS (Middle East respiratory syndrome) CoV infected more than 1700 people in 2012 [1, 2] . Now, the latest one, SARS-CoV-2, which is responsible for the rapid transmission of COVID (coronavirus disease) in 2019, turned into pandemic as it infected 230 million and killed 4.72 million people around the world as of September 2021. SARS-CoV-2 can spread directly from person to person through sneezing, coughing and talking within a six feet distance limit or through the indirect transmission from formites on which infection lasts for hours. Therefore, the rate of SARS-CoV-2 spread and death is much more as compared to the other CoVs as this virus damages respiratory system and fever, dry cough, shortness of breath are some of the symptoms which take 2-14 days to show its deadly effects [3] . Therefore, this is an alarming situation, where each individual has been instructed to wear a mask and maintain social distancing along with personal hygiene. Meanwhile, medical practitioners are searching for fast and accurate test kits, that could detect the infected persons and isolate them to stop transmission. However, the present reverse transcription polymerase chain reaction (RT-PCR) kits are slow and costly as they transform RNA to complementary DNA (C-DNA) and then measure the specific RNA amount by monitoring C-DNA amplification using PCR technique [4] . Unfortunately, the rapid test kits exhibit some limitations like sensitivity and wide deviation of test results (34 %-80 %), showing both false negative results and false positive results. Also, these rapid test kits can detect COVID only when someone is recently infected and could not provide any information about other diseases and their symptoms. Further, these kits rely on simultaneous capture and detection of the virus and so there is a possibility to miss the patients who have recovered from COVID-19. Therefore, the lack of accurate rapid test kits drove researchers to find suitable alternatives for the detection of SARS-CoV-2 and other such diseases in future.

Certainly, label-free dielectric modulated (DM) FET biosensors are one such alternative as they are capable to detect biomolecules with different dielectric constant values and are cost efficient as compared to the existing technologies [5] [6] [7] [8] [9] . Also, TiO 2 nanowires discussed in literature are good examples of glucose and vitamin detections using novel device techniques [10, 11] . SARS-CoV-2 structure is composed of multiple proteins such as spike (S), membrane (M), envelope (E) and hemagglutinin-esterase (HE) as shown in Fig. 1 . The S-protein has dielectric values ranging from 1 to 4 that could be easily detected from label-free DM-FET biosensors. Also, RNA transformed to C-DNA would have dielectric values ranging from 1 to 64. There is enough evidence available in literature that many FET based biosensors are effective for detection and diagnosis of COVID-19 [12, 13] . Most of the works are centred on graphene [12] and carbon nanotube [14] based FETs. In addition, many 2-D materials [15] and novel FETs [13] mechanisms are also discussed for COVID detection. However, the CMOS based DM-FET biosensors are not suitable for detection of wide variety of biomolecules as they have limitations like less sensitivity, low I ON /I OFF ratio, subthreshold swing (SS) greater than 60 mV/decade, short channel effects (SCEs) and random variations, which would reduce the test kit sensitivity and power efficiency. Alternatively, Tunnel FET (TFET) based DM biosensor provides better sensitivity due to tunneling phenomenon, SS lesser than 60 mV/ decade generates higher I ON /I OFF ratio at low supply voltage [16] [17] [18] . However, COVID-19 researchers are always in dilemma to select the appropriate DM-TFET biosensor as literature is flooded with such biosensors describing their applicability and hence, a decision based on proper guidance is required. Therefore, this manuscript is a small effort towards the comparative analysis of three basic structures which include physically doped (PD) [19] , charge plasma (CP) and electrically doped (ED) DM-TFET biosensors [20] [21] [22] [23] . Particularly, Silicon these three devices cover almost the entire research area of doping and doping-free (junctionless) TFETs and hence [24] , this manuscript provides a strong interpretation for the selection of proper device technology for designing rapid and accurate biosensors in future.

The cross-sectional view of PD-DM-TFET, CP-DM-TFET and ED-DM-TFET biosensors are presented in Fig. 1 . In PD-DM-TFET ( Fig. 2(a) ) N+ drain region and the P+ source region are created through doping where as CP-DM-TFET ( Fig. 2(b) ) and ED-DM-TFET ( Fig. 2(c) ) are junctionless devices as there are no physical boundaries between the regions through doping difference. But, in CP-DM-TFET and ED-DM-TFET external metal electrodes are placed above the source and drain regions and metal-semiconductor work function difference creates plasma of charge in CP-DM-TFET while opposite potentials applied on electrodes create source and drain in ED-DM-TFET [25] . TFET produces low ON current and high ambipolarity which affects biosensor sensing capability and power requirements. Therefore, hetero dielectric (HD) techniques are adopted by which half of the semiconductor from gate to source is covered with the high-K dielectric (HfO 2 ) which improves tunneling at the junction and hence [26, 27] , provides better ON current where as a low-K dielectric (SiO 2 ) in the other half reduces the ambipolar conduction at drain/channel junction.

Further, to introduce body fluid in the biosensor, a test cavity is formed inside HfO 2 below gate electrode towards source region where due to dielectric value difference, the band bending at gate/source junction changes. This results in the variation of drain current which when measured confirms the presence/absence of a particular biomolecule. All device physical parameters used in the device simulation are the optimum values obtained from literature and are tabulated in Table 1 .

The device fabrication process flow is as follows : Biosensors discussed in this manuscript include simple Silicon based TFET device that can be fabricated with the Silicon conventional MOSFET fabrication technologies. In PD biosensor, source and drain regions are created in Silicon substrate by selective window etching and doping by diffusion. Next, hetero dielectric layer can be deposited by liquid phase deposition method: first SiO 2 is completely deposited over Si substrate by dry oxidation process and then, half of the region is selectively etched by photolithography and then the Atomic Layer Deposition (ALD) deposits HfO 2 over the remaining portion. In CP and ED devices, additional metal layers are deposited over the oxide layers using metallization techniques. Finally, the test cavities can be formed by dry etching process and below the cavity, HfO 2 layer binds the biomolecules. The device simulations are carried out in Silvaco 2D-ATLAS TCAD tool. Schottky contact effect is incorporated using Universal Schottky tunneling model (UST), Shockley Read Hall (SRH) and Auger models are used for recombination of carriers. At the tunneling-junction, the carrier generation rate is calculated using non-local band-to-band tunneling (BTBT) and bandgap narrowing (BGN) models. The Fermi Dirac statistics and Klaassen's Unified Low Field Mobility models (KLA) are used to describe the lateral field mobility effect on the device characteristics. The models have been verified with the experimental result in [6] and we found that the simulated results are closely matched with the experimental data as depicted in Fig. 3 .

This section deeply investigates the PD-DM-TFET, CP-DM-TFET and ED-DM-TFET biosensors and compare their sensing capability for a wide range of biomolecules present in human body, that can be captured through blood, swab or urine samples. These biomolucules having a dielectric value can be charged or neutral and hence, we considered the charge density (-1 × 10 13 C/cm 2 -1 × 10 13 C/cm 2 ) and K (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) values for the device sensitivity analysis.

In this section, we first focus on BTBT process variation at Source/Channel junction for various dielectric constants of biomolecules. Under ON state condition, the energy band diagram (EBD) of n-PD-DM-TFET, n-CP-DM-TFET, n and p ED-DM-TFET based biosensors are shown in Fig. 4 (a-d) Silicon with neutral charge and variable K values. In ON condition, the channel side conduction band (CB) and source side valence band (VB) aligns at source-channel interface. ON state EBD of all the three devices demonstrate that the bandbending increases and barrier width reduces with the increase of K values. It means that the electrons take less energy to tunnel through the barrier. Figure 5 (a-d) shows the variation in surface potential for three biosensors during ON and equilibrium conditions. It has been noticed that surface potential is lesser for K = 1 (vacant cavity) but in the presence of biomolecules (K>1), potential increases because the effective gate-capacitance (C eff ) of biosensor rises with the dielectric constant value. Figure 6 (a-d) shows the drain current (I d ) characteristics of n-PD-DM-TFET, n-CP-DM-TFET, n and p ED-DM-TFET based biosensors with different values of K. During ON-state, minimum I d is observed for air (K = 1). Also, drain current rises with the increment in dielectric constant (K>1) in all the three devices. The peak current values are listed in Table 2 .

For positively-charged biomolecules, the I d -V gs characteristics of n-PD-DM-TFET, n-CP-DM-TFET, n and p ED-DM-TFET based biosensors are shown in Fig. 7(a-d) with constant K = 5. In all three n-TFET based biosensors, the drain current (I d ) rises with the increment in positive charge density value of biomolecules but in p-ED-DM-TFET the drain current (I d ) decreases with the increment in positive charge density. This could be better explained through EBD of ED-DM-TFET in Fig. 8 . Figure 8 (a) shows that in n-ED-DM-TFET, bands steepen more at tunneling junction as positive charge increases whereas this effect is reverse in EBD of p-ED-DM-TFET ( Fig. 8(b) ) due to the dominance of opposite charges in both devices. For negatively-charged biomolecules, the I d -V gs characteristics of n-PD-DM-TFET, n-CP-DM-TFET, n and p ED-DM-TFET based biosensors are shown in Fig. 9 (a-d) with the same K value.

In all three n-TFET based biosensors, I d decreases with the increment in negative charge density value of biomolecules but in p-ED-DM-TFET based biosensor, I d rises with the increment in negative charge density. This effect is better understood though the EBD of ED-DM-TFET in Fig. 10 and the reason for the reverse trend in p-type and n-type TFET is similar to that explained 8(a) and (b). It can be clearly observed from Fig. 10(b) , that in p-ED-DM-TFET, bands steepen more at tunneling junction, whereas the opposite effect is noticed in EBD of n-ED-DM-TFET ( Fig. 10(a) ) due to the dominance of opposite charges in both devices. The peak Silicon current value of all the three devices for positive and negative charged biomolecules are illustrated in Table 3 .

The biomolucule detection quality of a biosensor is obtained though sensitivity analysis. Hence, higher the sensitivity, higher would be the detection probability. The biosensor sensitivity is calculated from Eq. (1).

where I d Bio indicates drain current of the device in the presence of biomolecules and I d Air indicates drain current of the device for vacant cavity.

Greater the difference between the I d Bio and I d Air , higher would be the sensitivity. Therefore, increasing the difference between the drain current for K=1 and K > 1, increases I d -V gs sensitivity as shown in Fig. 11(a-d) . Table 4 illustrates the I d -V gs sensitivity value of all three devices for several neutral For positively-charged biomolecules, the I d -V gs sensitivity of all three devices are shown in Fig. 12 (a-d) with the constant K = 5 value. Depicted from the difference in drain current, the I d -V gs sensitivity rises with the increment in positive charge density for n-TFETs but in p-ED-DM-TFET the I d -V gs sensitivity sensitivity decreases with the increment in positive charge density value of biomolecules. For negative-charged biomolecules, the I d -V gs sensitivity of all three devices are shown in the Fig. 13 (a-d) for same K value. Here, also due to the drain current difference, I d -V gs sensitivity decreases with the increment in negative charge density in all three nTFET based biosensor but in p-ED-DM-TFET, I d -V gs sensitivity rises with the increment in negative charge density. The Table 5 . Results observed in this paper show excellent agreement with the data presented in [7, [16] [17] [18] , which signifies that the comparative analysis performed in this article is authentic for biomedical applications. 5.31×10 −7 6.06×10 −8 1.37×10 −7 4.47×10 −8 5×10 12 6.46×10 −7 6.98×10 −8 1.67×10 −7 4.01×10 −8 1×10 13 9.76×10 −7 1.09×10 −7 3.16×10 −7 3.33×10 −8 -1×10 13 1.10×10 −7 3.83×10 −8 7.45×10 −8 1.57×10 −7 -5×10 12 2.13×10 −7 4.65×10 −8 8.78×10 −8 9.09×10 −8 -3×10 12 2.72×10 −7 4.88×10 −8 9.53×10 −8 7.36×10 −8 -1×10 12 3.43×10 −7 5.11×10 −8 1.04×10 −7 6.08×10 −8 -5×10 11 3.63×10 −7 5.18×10 −8 1.07×10 −7 5.08×10 −8 

A detailed literature survey has been performed with the Si based biosensors and it is found that these works has closely matched sensitivity values as shown in Table 6 .

The temperate dependent sensitivity of the biosensor is also computed for the in-depth analysis of the biosensor. Figure 14 (a) shows that as the temperature increases, OFF current increased recombination process in ED-DM-TFET. However, there is no appreciable variation in drain current for higher V gs value. Therefore, the biosensor sensitivity depends on both V gs and temperature. Hence, Fig. 14(b) shows the peak value of biosensor sensitivity in PD-DM-TFET, CP-DM-TFET and ED-DM-TFET and it is clear that the sensitivity reduces for higher temperature and the variation of sensitivity between the three types of biosensors also reduces with temperature.

A comparative analysis of PD-DM-TFET, CP-DM-TFET, n-ED-DM-TFET and p-ED-DM-TFET based biosensors are done in this paper. Both charged and neutral biomolecules are identified and detected by these three devices. Some characteristics like energy band diagram, the surface potential, I d -V gs (transfer) characteristics and I d -V gs sensitivity are investigated to find better quality biosensor among these three. For neutral biomolecules with K=12 the I d -V gs sensitivity value of 5.10 × 10 7 is obtained for p-ED-DM-TFET biosensor which is the maximum among the three nTFET based biosensors. For positively-charged biomolecules with charge density of 1 × 10 13 C/cm 2 , the I d -V gs sensitivity value of 4.77 × 10 8 is obtained for n-ED-DM-TFET biosensor, which is maximum among the three TFET based biosensors. For negatively charged biomolecules with charge density of -5 × 10 11 C/cm 2 , the I d -V gs sensitivity value of 1.76 × 10 6 for p-ED-DM-TFET biosensor is the maximum. So, it could be concluded that ED-DM-TFET based biosensor exhibits better detection and identification capability among the three different configurations of TFET biosensors.

Author Contributions All the authors have contributed to the design, investigation, conceptualization and formal analysis. and design. Author Arpita Biswas prepared the first draft of the manuscript after perfroming the simulation study and validation. Author Chithraja Rajan edited the manuscript after data analysis and validation. D P Samajdar commented on the manuscript and supervised the entire work. All the authors read and approved the final version of the manuscript. Code Availability Not Applicable.

Consent to Participate All the authors contributed voluntarily to this work.

In accordance with the copyright transfer or open access rules.

Identification of a novel coronavirus in patients with severe acute respiratory syndrome

Isolation of a novel coronavirus from a man with pneumonia in Saudi Arabia

A familial cluster of pneumonia associated with the 2019 novel coronavirus indicating person-to-person transmission: a study of a family cluster

Molecular diagnosis of COVID-19: an update and review

Reconfigurable FET biosensor for a wide detection range and electrostatically tunable sensing response

Vertical Si-nanowire n-type tunneling FETs with low subthreshold swing (≤= 50 mV/ decade) at room temperature

Sensitivity Analysis of a Novel Negative Capacitance FinFET for Label-Free Biosensing IEEE Transactions on Electron Devices

Design principles for a novel lightweight configurable PUF using a reconfigurable FET

Nanowire reconfigurable FET as biosensor: Based on dielectric modulation approach. Solid-State Electron 161

A novel Ag nanoparticles/TiO 2 nanowires-based photodetector and glucose concentration detection

Detection technique for vitamin D 3 using Er-doped TiO 2 nanowire-based UV photodetector

Selective and electronic detection of COVID-19 (Coronavirus) using carbon nanotube field effect transistor-based biosensor: A proof-of-concept study

A dielectric modulated biosensor for SARS-CoV-2

Rapid detection of COVID-19 causative virus (SARS-CoV-2) in human nasopharyngeal swab specimens using field-effect transistor-based biosensor

Two-dimensional-materialbasedfield-effect transistor biosensor for detecting COVID-19 virus (SARS-CoV-2)

Performance assessment of a novel vertical dielectrically modulated TFETbased biosensor

Performance analysis of gate engineered dielectrically modulated TFET biosensors

Sensitivity and sensing speed analysis of extended nanocavity and source over electrode in Si/SiGe based TFET biosensor

Design of NW TFET biosensor for enhanced sensitivity and sensing speed by using cavity extension and additional source electrode

Design and analysis of dual-metal-gatedouble-cavitycharge-plasma-TFET as a label free biosensor

A charge-plasma-baseddielectric-modulated junctionless TFET for biosensor label-free detection

A new structure of electrically doped TFET for improving electronic characteristics

Comparative analysis of full-gate and short-gate dielectric modulated electrically doped Tunnel-FET based biosensors10.1016/ j

A new design approach of dopingless tunnel FET for enhancement of device characteristics

A fair comparison of the performance of charge plasma and electrostatic tunnel FETs for low-power high-frequency applications

Double-gate tunnel FET with highgate dielectric

Dielectric engineered tunnel field-effect transistor

Junctionless based dielectric modulated electrically doped tunnel FET based biosensor for label-free detection

Label free detection of biomolecules using charge-plasma-based gate underlap dielectric modulated junctionless TFET

Comparative analysis of circular gate TFET and Heterojunction TFET for dielectric-modulated label-free biosensing

Application of nanocavity embedded dual metal double gate TFET in biomolecule detection

SiGe source charge plasma TFET for biosensing applications

Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations

The authors declare that they have no conflict of interest.Ethics Approval Not Applicable.