Skip to main content
NCBI home page
ACS Omega logoLink to ACS Omega
. 2025 Sep 22;10(39):44989–44999. doi: 10.1021/acsomega.5c02855

Ge Pocket-Based Dual Gate Tunnel FET: A Label-Free Biosensor

M Salim Wani †, Anam Khan †, Abdullah G Alharbi ‡, Sultan S Aldkeelalah §, Sajad A Loan †,*
PMCID: PMC12508951  PMID: 41078750

Abstract

In this work, we designed and simulated a novel dielectrically modulated (DM) dual-metal-gate tunnel field effect transistor (DMG-TFET). The proposed TFET can efficiently facilitate the label-free detection of biomolecules and can work as a label-free biosensor. It employs a highly doped Germanium pocket with an optimized pocket length of 5 nm in the channel region near the source side. Besides, an enhanced on-current (I ON) is achieved in our proposed device owing to the introduction of a pocket and the use of a high dielectric constant k, as well as a dual-metal gate. The proposed TFET sensor features twin cavities within the gate oxides to sense biomolecules. The sensitivity of the biosensor is evaluated based on its ability to detect biomolecules using their inherent material parameters, such as dielectric constant (k) and charge density (ρ). The proposed DMG-TFET is compared to the conventional dielectrically modulated FET in terms of sensitivity for different values of dielectric constants (k). It has been found that the sensitivity of the proposed biosensor is 104 times greater than that of the conventional device at k = 10 and ρ = 0. Special cases of sensitivity have been analyzed for a single biomolecule of uricase (k = 1.54) placed at various positions and orientations. It has been observed that the biomolecule’s position in the cavity has a significant effect on the sensor’s performance. All results are deduced from simulations carried out on Silvaco Atlas. To observe the effect of noise on the biosensor’s sensitivity, noise-aware sensitivity estimation and detection analysis have been performed.

graphic file with name ao5c02855_0015.jpg

graphic file with name ao5c02855_0013.jpg

1. Introduction

Biomolecules are vital, complex organic molecules that are crucial for the sustenance of all living organisms. They are responsible for maintaining life in all organisms by specifically controlling metabolic processes like growth and development. Therefore, the need to accurately detect and categorize biomolecules is of prime importance. The classification of biomolecules into its subsets lays the foundation for investigation, examination, and treatment of various ailments. Hence, it highlights the need for disease surveillance and efficient handling of contagious entities present in the environment during infancy of the detection process. The design of biosensors is primarily dependent on their reception ability as well as their detection rate. FET-based biosensors find widespread applications in the field of bioelectronics as they provide label-free detection and sensing. In the context of biosensing, the term label-free refers to the detection of biomolecules without the need for any external markers or chemical labels, such as fluorescent dyes or enzymes. This approach allows for direct, real-time monitoring of biomolecular interactions in their native state. Label-free detection is advantageous because it simplifies the sensing process, preserves the integrity of the biomolecules, and reduces the preparation time and cost. In this work, the proposed DMG-GPTFET biosensor detects the presence of biomolecules based on their intrinsic propertiesspecifically, their dielectric constant and charge densitymaking it an effective and efficient label-free biosensor. FETs, being devices with high scalability and low power consumption, along with peculiar features like reduced size, economical production, and ease of fabrication, are prominent devices commonly used in biomolecule detection. − Biosensor characteristics like sensitivity and selectivity are of utter significance when choosing a biosensor for targeted application. Sensors with higher sensitivity are mostly preferred, as they can sense even a minor change in the target biomolecule concentration by equivalently changing the related electrical parameters, which is reflected by the change in the associated characteristics. − One of the most commonly used biosensors is the ion-sensitive FET (ISFET), which has been rigorously used for sensing since its inception decades ago. Even though it finds numerous applications in the field of biomedical sciences, it still suffers from certain limitations. One being its ability to only detect biomolecules with higher sensitivity. To overcome the challenges associated with the ISFET, new variants of FETs like dielectrically modulated FETs (DM-FET) have been explored and examined. DM-FETs have wide detection capabilities as they can sense both charged and neutral biomolecules, making them promising candidates in the field of biosensing applications. The elevated characteristics of DM-FETs are attributed to their architectural configuration. In this device, a nanogap is introduced within the dielectric material under the gate material near the source side. This nanogap serves as a cavity, which is responsible for capturing, sensing, and detecting a specific biomolecule by initiating the trapping mechanism. Once a biomolecule is trapped inside the nanogap cavity, the dielectric constant of the trapped entity causes a variation in the effective coupling between the channel and gate. This interrelated phenomena of dielectric variation and effective coupling lays the foundation of the working principle of a basic DMTFET. The biosensors based on dielectrically modulated FETs encompass FinFETs, FETs based on impact ionization, FETs using silicon nanowires, tunnel FETs, and the most widely used MOSFET-based biosensors. − MOSFET-based biosensors show increased sensitivity; however, they face severe short channel effects (SCEs), increased leakage, and reliability issues when scaled beyond a certain range, thereby arising the need to look for alternatives. In such a scenario, biosensors employing the tunnel field effect transistors are promising ones as they result in a sharp subthreshold slope and reduced leakage. − A biosensor with higher sensitivity can detect biomolecules even when their concentration is low. ,

A comparison is carried out between a simple DM-FET and our proposed DMTFET to investigate their operating subthreshold current range. The DM-FET has an operating range from 1 × 10–6 to 1 × 10–2 μA, while the DMTFET has an operating range from 1 × 10–9 to 1 × 10–7 μA. ,− The downgraded value of subthreshold current in the second case makes the biosensor more susceptible to noise. − Therefore, it highlights the importance of subthreshold current in precise sensing and accurate detection of trapped biomolecules, even when they are present in insignificant amounts. − ,− Also, the efficacy of TFETs can be enhanced by structural or material engineering strategies. − ,,

A TFET variant that has been dielectrically modulated (DM) is proposed and simulated in our work. The device proposed is a TFET incorporating a thin layer of germanium between the source and the channel (acting as a pocket), along with the drain side made up of a metal. In addition to this, the gates used are not composed of a single material but are made up of two different metals to further enhance device characteristics, hence named dual-metal-gate (DMG-GPTFET). The incorporation of a germanium pocket between the source region and the channel region is intended to elevate the on-current of the device. This substantial increase in on-current is because of the fact that germanium is a low-bandgap material with improved mobility of carriers. It thus reduces the barrier width by improving the alignment of the energy bands near the tunneling junction, i.e, the valence band of the source and the conduction band of the channel. Our device replaces conventional silicon on the drain side with hafnium (having WF = 3.9 eV). The primary advantage of using hafnium in our proposed dual-metal-gate tunnel field-effect transistor (DMG-TFET) lies in its contribution to reducing ambipolarity. Ambipolarity is an inherent property of TFETs. This occurs when a negative voltage is applied to a TFET, leading to the flow of current because of tunneling at the channel–drain junction. To remove ambipolarity, we use a metallic drain in place of conventional silicon. Furthermore, our device uses HfO2 (having k = 21) as a dielectric. This helps to enhance the TFET characteristics by creating a narrower window for electron tunneling. Another major material change incorporated is the use of a dual-metal gate: a gate having two unique work functions (WFs), one near the source side (G1) and the other near the drain side (G2). When we use a gate consisting of two distinctive materials, it helps to increase the on-current, thereby allowing the on-current to attain the highest possible value. Apart from increased I ON, the sensitivity of the device is also improved. A nanocavity is created to capture and detect biomolecules. To make this sensor work, we have built a tiny cavity (a nanocavity) right into the gate oxide layer, very close to where the “source” of the transistor is located. We created this cavity by etching away some of the material. This nanocavity is the key spot where the biomolecules we want to detect interact with the sensor. To help these biomolecules stick better and to prevent unwanted leakage of current from the gate, we coated the inside of the nanocavity with an extremely thin layer (just 0.5 nm) of HfO2. Our sensor works based on “band-to-band tunneling.” This means that when biomolecules enter the nanocavity, they change the electrical environment (specifically, the dielectric constant). This change affects how the gate couples to the channel of the transistor, which in turn alters the transistor’s electrical behavior. On top of that, the nanocavity’s small size and shape help to concentrate the electric field, making the sensor even more sensitive to the presence of those biomolecules. So, the size of this nanocavity, especially its thickness (t nc), becomes a crucial design ). Table shows the characteristics of various biomolecules used to be detected by the proposed biosensor.

1. Biomolecule Types with Their Dielectric Constants (k) and Nature of Charge (Positive/Negative/Neutral).

Type of Biomolecule “k” as mentioned in [refs − ] Charge associated (±/0) Size
APTES 3.57 0 0.5–0.58 nm
Biotin 2.63 0 3–5 nm
Streptavidin 2.1 0 5 nm
Uricase 1.54 0 L = 5 nm,1.2 nm thickness
DNA 1–64 - 2.5 nm
Open in a new tab

The behavior of our proposed device has been studied, and its efficiency has also been tested by taking into account parameters like surface potential and electric field. The transfer characteristics are thoroughly evaluated to correlate device parameters with its sensitivity. All the simulations carried out have been performed on Silvaco TCAD. The idea was presented in our earlier work; however, calibrated simulations using experimental data (already mentioned in the used reference number) have been performed. We have also performed different analyses, leading to substantial results regarding the biosensor’s sensing capability.

The paper content has been divided into four sections, namely: Section I, which gives an introduction to a biosensor and a preview of the simulated work; Section II includes the schematics and simulation setup involved for carrying out simulations. A proposed fabrication scheme for our designed device is also discussed in this section; Section III emphasizes the results obtained. Lastly, Section IV concludes the work and briefly discusses the key points, and, Section V includes the list of references used.

2. Device Design, Simulation Setup, and Fabrication Process

The schematics of the conventional DM-FET and the proposed DMG-GPTFET biosensor are illustrated in Figure . The germanium pocket introduced between the source and the channel region has a thickness (t p) of 5 nm and a high doping concentration of 1 × 1019 cm–3. The doping concentration of source region is 1 × 1020 cm–3 and that of the channel region is 1 × 1015 cm–3. A uniform doping profile is used for all regions of the device. The silicon body thickness (t Si) is considered fixed at 10 nm throughout all simulations; hence, the quantum effects are disregarded. All of the other parameters used in the simulations are mentioned in Table . The nanocavity depicted in our proposed device is created by the process of etching, and its dimensions are mentioned in Table . The inside of the nanocavity is coated with a 0.5 nm thick layer of HfO2, which acts as a binding material so that the biomolecules adhere to it. This layer also helps to diminish the gate leakage current and minimize degradation in sensitivity. Furthermore, the material’s intrinsic properties like charge density and dielectric constant, have been used to detect any biomolecule captured in the nanogap cavity, where the unity value of the dielectric constant signifies their absence, and another presumption is made that the biomolecules completely fill the nanogap cavity.

1.

1

Open in a new tab

Schematic layout of the (a) dielectrically modulated FET (conventional) and (b) proposed DM-based germanium pocket TFET (DMGP-TFET) biosensor.

2. Parameters Used in Simulations for the DMG-GPTFET.

Parameter Symbol Value
Source length L S 100 nm
Channel length L C 50 nm
Gate length L G 50 nm
Drain metal WF ϕD 3.9 eV
WF of gate G1 ϕ1 4.1 eV
Wf of gate G2 ϕ2 4.8 eV
Length of G1 L G1 20 nm
Length of G2 L G2 30 nm
Thickness of the nanocavity t np 5.5 nm
Length of the nanocavity L nc 20 nm
Open in a new tab

Our proposed device has been simulated using Atlas Silvaco (5.26.1.R.) Figure a compares the transfer characteristics between our simulated results and the reported work in order to verify the validity of our simulation scheme. As shown in Figure , there is a clear agreement between the two, which validates our simulation scheme. The architecture and size of the device remained intact throughout the simulation to match the results reported in ref . The device characteristics improved by using a fine meshing network at the junction of the source and channel in comparison to the rest of the device. A dense meshing of 0.0001 has been used in the tunneling area. This is done to achieve higher levels of accuracy in predicting the sensitivity levels. A number of models are used while carrying out the simulations like the BTBT model to account for nonlocal band-to-band tunneling, the FLDMOB model for parallel electric-field dependence, the CONMOB model for concentration-dependent mobility, the UST model for Universal Schottky Tunneling, the BGN model for bandgap narrowing in heavily doped regions, and lastly, the Auger model and SRH model for recombination. The Wentzel–Kramers–Brillouin (WKB) method is used for determining numerical tunneling probability.

2.

2

Open in a new tab

Calibration with the results reported in ref .

A process flow for the fabrication of our proposed device is shown in Figure . It involves various steps, as shown in Figure . On an intrinsic semiconductor wafer, a p-type region can be created by selective diffusion of p-type impurities (or p-type ion implantation). The wafer can then be annealed after ion implantation to repair lattice damage and place atoms in substitutional positions. To create a germanium pocket, a window needs to be created in the wafer at an appropriate place. A dry etching approach can be used to create this window for pocket realization near the source by plasma etching. The germanium material can then be directly deposited into the window near the source employing molecular beam epitaxy (MBE), which controls the rate of deposition accurately and precisely. A thin HfO2 layer is conformally deposited above the intrinsic channel using the atomic layer deposition (ALD) technique, followed by annealing in a rapid thermal annealing (RTA) system to enhance the film quality. Next, the gate is realized by directly depositing the metal on the oxide layer in an evaporation vacuum chamber. The creation of nanocavities in the oxide is an important step in the proposed process flow. Using the selective etching mechanism, a nanocavity is created near the source side of the gate dielectric. The chosen silicon area can be protected by an oxide layer created by a dry oxidation mechanism. There is also an alternative approach that employs a neutral-beam-assisted atomic ALE. This technique is advantageous as it provides damage-free etching and preserves immunity against process variations for such biosensors. The final step involves direct metal deposition to create a metallic drain. In this way, a germanium pocket-based double-gate TFET biosensor can be realized. The proposed structure can be flipped to create its flipped alternative, which can be bonded with the alternative structure to realize the proposed biosensor. The proposed process flow just gives the direction for realizing the proposed structure. However, a lot of challenges will occur in actually fabricating the structure, which can be overcome in due course of time as fabrication techniques are improving day by day with the availability of advanced fabrication facilities.

3.

3

Open in a new tab

Proposed Fabrication Scheme for the DMG-GPTFET.

3. Results and Discussion

A study has been performed on the device performance when different biomolecules are used, and the impact on device characteristics has been carefully examined. This analysis considers dielectric values (k > 1) and charge density values ranging from ρ = −1011 C cm–2 to −1012 C cm–2), which have been employed in the nanogap cavity. The on-state of our proposed device is achieved when V GS = 1 V and V DS = 0.5 V, and the device is considered to be in the off-state when V GS = 0 V and V DS = 0.5 V. The energy band diagram of our proposed device in the on-state and off-state is depicted in Figure . In the off-state, the barrier between the energy bands of the source and channel is wide, thereby reducing the chances of band-to-band tunneling (BTBT), as higher energy is required for electron tunneling from the valence band of the source region to the conduction band of the channel region. However, the behavior of the device in the on-state is completely changed, because the energy required for tunneling from the valence band to the conduction band is much lower, as evident from Figure a. The BTBT rate is elevated in the on-state as the tunneling width at the junction is reduced, thereby providing a smaller path for electron transition. Thus, the on-current of the device is enhanced.

4.

4

Open in a new tab

(a) Energy distribution curve (band diagram) of our proposed device in the ON/OFF state. (b) Band diagram of the proposed device when the dielectric constant is varied. (c) Band diagram of the proposed device when charge density is varied.

The energy band diagram of the DMG-GPTFET in the on-state for different values of k (ρ = 0) is shown in Figure b. It is clear from the energy band representation that the band bending enhances as the value of k is increased. More the band bending, the lesser the tunneling width, making it easier for carriers to tunnel through. Figure c depicts the impact on the energy band diagram because of the variation in charge density (ρ). The carrier width is reduced owing to the formation of holes in the channel when negative charge density increases (−ρ). The change in the energy band diagram by changing ρ or k is better understood by considering the visual representation of the electric field along the device length. Figure a illustrates the electric field (EF) strength throughout the device when ρ = 0 and k is varied from 1 to 10. When ρ is fixed at 0 and the value of k is increased, the ability of the gate to control the channel is substantially improved. This reduces the tunneling barrier width, because of which the EF increases. The maximum value of EF obtained at the tunneling junction is 3.65 × 106 V/cm, which is achieved when k = 10. A similar graph for EF through the device is shown in Figure b, which is obtained when k = 5 and ρ is varied. The EF strength diminishes for an increasing value of −ρ. The highest value of the electric field is 3.54 × 106 V/cm, which is obtained at ρ = 0.

5.

5

Open in a new tab

Electric field of the DMG-GPTFET along the device length for (a) neutral and (b) charged biomolecules.

Figure a,b illustrates the variation in the surface potential (SP) of the DMG-GPTFET within the L G1 region in response to various biomolecule values. For k = 1, that is when air is present in the nanocavity, the SP exhibits a downgraded value. With an increasing value of the dielectric constant (k > 1), a subsequent change is observed in the SP. The increased coupling between the channel and the gate forces the SP to increase as the values of k are increased. The equation for the flat-band voltage is given by eq :

Flatbandvoltage=qρ/Ceff 1

6.

6

Open in a new tab

(a) Surface potential (SP) of the DMG-GPTFET along the X-position for varying k (ρ = 0) and (b) varying ρ (k = 5). (c) Nonlocal BTBT e– tunneling rate of the DMG-GPTFET for varying k (ρ = 0) at V GS = 1 V and V DS = 0.5 V.

Where the charge density is represented as ρ, and the capacitance between the gate and the channel is defined as effective capacitance and is represented as C eff. The presence of negatively charged biomolecules tends to increase the value of the flat band voltage. The overall effect is observed on the surface potential, which is developed in the channel region near the source side; hence, a diminished value of SP is observed within the L G1 region. Figure c vividly explains that whenever a biomolecule is present with k > 1, the tunneling barrier width reduces appreciably. This reduction in tunneling distance accelerates the BTBT rate of carriers from the valence band to the conduction band. This, in turn, leads to a rise in the value of drain current (I DS), which means that the performance of the device in the on-state is enhanced.

Next, we carried out simulations to observe the effect of changes in parameters related to device architecture, like the thickness and length of the nanocavity, which are represented by t nc and L nc, respectively. A graphical representation of transfer characteristics with varying t nc and L nc for fixed values of k = 5 and ρ = 0 is depicted in Figure a. In our simulations, t nc was varied from 5.5 to 10 nm, and we observed that as the thickness of the nanocavity in our proposed device is increased, there is a proportional increase in the tunneling width, which, in itself, means a drop in the value of the on-current. However, when simulations were carried out by varying the length of the nanocavity, no significant change was observed. Figure b demonstrates that when L nc is varied from 20 to 30 nm, there is no measurable impact on the value of I DS. The reason behind this lies in the fact that the working mechanism of our proposed device is based on BTBT rather than bulk conduction, which is the case when using conventional FETs. A change in I ON is observed with a change in L nc when the nanocavity is filled with air (k = 1). This has been demonstrated and reported by Narang et al. However, when biomolecules are to be sensed for k > 1, I ON does not vary with varying L nc.

7.

7

Open in a new tab

Impact on drain current (ID) when (a) the cavity thickness (t nc) and (b) the cavity length (L nc) is varied.

Another set of simulations has been carried out to observe the impact of dielectric constant variation on I D V G characteristics of the DMG-GPTFET. This effect is explained using the graph shown in Figure a, which also provides a comparison of DMG-GPTFET characteristics with those of a conventional DM-FET. The minimum value of I DS (when air is present in the nanocavity) is in the order of 10–10 (A/μm) with V GS = 1 V and V DS = 0.5 V. As soon as the biomolecule gets trapped inside the nanocavity (k > 1), the I DS value is enhanced. The maximum value of I DS attained in this case is in the order of 10–6 (A/μm) for k = 10 at V GS = 1 V and V DS = 0.5 V. Three more plots are obtained with varying values of k (at k = 5, k = 7, k = 10) and also for ρ = 0.5 × 10–11, 1 × 10–11, and 5 × 10–11. Figure b–d shows the corresponding related plots. The main reason behind this behavior is that when the negative charge densities increase, the SP decreases, which in turn decreases the drain current. Figure a,b represents the current sensitivity with varying gate bias of the DM-FET and our proposed device for different values of the dielectric constant in the first case and varying both ρ and k in the second case. Sensitivity analysis on our proposed device has also been carried out using the I DS sensitivity formula, which is determined by eq :

Idss=(IB−IAIA) 2

8.

8

Open in a new tab

Transfer characteristics of the DMG-GPTFET for different values of (a) dielectric constants (k), (ρ = 0), (b) negative charge densities (ρ) (k = 5), (c) negative charge densities (ρ) (k = 7), and (d) negative charge densities (ρ) (k = 10).

9.

9

Open in a new tab

(a) Comparison of current sensitivity in the DM-FET and DMG-GPTFET biosensors with varying dielectric constants. (b) Sensitivity analysis of the DMG-GPTFET with varying ρ and k.

Where I B is the value of the drain current when the biomolecules (different k and ρ) are present in the nanocavity, and I A is the value of the drain current when air (k = 1) is present in the nanocavity. When ρ = 0, we observe that there is a change in device sensitivity. The sensitivity increases when k is increased, and a maximum value is achieved at k = 10. A significant change in sensitivity with gate bias implies that our proposed device has a better ability to detect the changes in its nanocavity but it is observed that the current sensitivity reduces when the negative charge density increases, as illustrated in Figure b. The suggested biosensor can detect a single biomolecule trapped in the nanocavity, indicating its detection limit.

To accomplish this, a case study of a single molecule of uricase with dimensions of 1.2 and 5 nm in the cavities was examined. It is seen that the sensitivity also depends on the position and orientation of a biomolecule present in the nanocavity. We placed a single biomolecule of uricase vertically and horizontally at different positions in the nanocavity, as shown in Figure , cases A–D. The I DS sensitivity for cases A–D is depicted in Figure . In cases A and B, I DS sensitivities (A > B) are higher compared to cases C and D, as the biomolecule is present above the tunneling junction, leading to higher gate capacitance and effective gate voltage. However, in cases C and D, when the biomolecule is present away from the tunneling junction, I DS sensitivities (C > D) are reduced due to the lower gate capacitance and, consequently, the lower effective gate voltage within the area of the tunneling junction. The values of drain current sensitivity (I DS sensitivity) in cases A–D are 0.173, 0.114, 0.0265, and 0.0103, respectively. It is evident from Figure A–D that a biomolecule placed at the same position with different orientations results in different values of I DS sensitivity. In Table , the proposed model’s sensitivity (k = 12) is compared with earlier works. ,,, Table compares the sensitivity of the biomolecule at different positions and variations.

10.

10

Open in a new tab

Schematic of the DMG-GPTFET’s nanocavity filled with a single biomolecule of uricase- Case A: placed vertically above the tunneling junction; Case B: placed horizontally above the tunneling junction; Case C: placed vertically away from the tunneling junction; Case D: placed horizontally away from the tunneling junction.

11.

11

Open in a new tab

Sensitivity of uricase placed at different positions in the nanocavity.

3. Comparative Analysis of the Proposed Sensor with State-of-the-Art Sensors.

Reference Sensitivity
2.59 × 107
5.45 × 109
4.756 × 105
1.04 × 106
1.73 × 1010
This work 1.26 × 106
Open in a new tab

4. Sensitivity of Uricase with Respect to Its Positions and Variations.

Uricase Position Relative Sensitivity Specific Sensitivity Value (I DS Sensitivity)
Vertically above the tunneling junction Highest 0.173
Horizontally above the tunneling junction High 0.114
Vertically away from the tunneling junction Low 0.0265
Horizontally away from the tunneling junction Lowest 0.0103
Open in a new tab

3.1. Noise-Aware Sensitivity Estimation and Detectability Analysis

The sensors’ ability to sense and detect biomolecules can be affected by noise generated internally. Although the simulations carried out in this work demonstrate noticeable variations in drain current (I DS) due to the presence, position, and orientation of a single uricase biomolecule in the nanocavity, we acknowledge that a rigorous claim of single biomolecule detection demands a more quantitative approach involving signal-to-noise analysis. The minimum I DS when the cavity is filled with air (k = 1) is observed to be in the range of 10–10 A/μm, whereas the I DS increases to the order of 10–6 A/μm for a biomolecule with a dielectric constant k = 10, as shown in Figure a. This indicates a signal margin (ΔI DS) of approximately 4 orders of magnitude. However, it is important to consider that nanoscale devices are influenced by various noise sources, such as flicker noise, thermal noise, and random telegraph noise (RTN), which can contribute to background current fluctuations in the range from 10–12 to 10–9 A, depending on biasing and operating conditions. Therefore, the results presented here indicate the potential for single biomolecule detection under idealized conditions.

To support the claim of single biomolecule detection, a quantitative evaluation of the signal-to-noise characteristics is essential. Following the approach outlined by Bhattacharyya et al., we performed a noise-aware sensitivity estimation by comparing the change in drain current (ΔI DS) induced by biomolecule insertion against expected background noise levels in nanoscale TFET devices. As shown in Figure a, when the nanocavity is empty (k = 1), the drain current (I DS) is approximately 10–10 A/μm, whereas for a dielectric constant k = 10, representing the presence of a biomolecule, I DS rises to ∼10–6 A/μm. This corresponds to a signal margin of:

ΔIDS=IDS(k=10)−IDS(k=1)≈10−6−10−10=9.99×10−7A/μm 3

In realistic biosensing scenarios, the dominant noise components in TFETs include thermal noise, flicker noise, and random transmission noise (RTN). These contribute to background fluctuations in the range from 10–12 to 10–9 A, depending on the gate bias, channel material, and ambient conditions. To assess detectability, we calculated the signal-to-noise ratio (SNR):

SNR=ΔIDSInoise 4

Using I noise = 10–9 A (worst-case), we get:

SNR=10−610−9=103 5

Even under the most conservative noise assumptions, the resulting SNR indicates that the sensor is theoretically capable of detecting a single biomolecule with a distinguishable signal margin. This is further supported by Bhattacharyya et al., who demonstrate similar current levels and emphasize on the importance of noise-aware modeling for accurate sensitivity estimation in Bio-TFETs.

In Figure , the signal current (ΔI DS ≈ 1 μA) corresponds to the change in drain current induced by the presence of a single molecule (from k = 1 to k = 10). The noise current is varied from 1 pA to 1 nA, covering the expected range for nanoscale TFETs. The blue line at an SNR of 1 represents the minimum threshold for detectability. The plot confirms that the biosensor maintains a detectable signal even under worst-case noise conditions, supporting its potential for single biomolecule detection in ideal environments.

12.

12

Open in a new tab

Signal-to-noise ratio (SNR) versus background noise current for the proposed DMG-GPTFET biosensor.

4. Conclusion

A novel dual-gate TFET has been proposed and simulated in this work, with the objective of using this device effectively as a biosensor. Our proposed device incorporates architectural as well as material changes in a basic DM-FET, like the use of a dual metal gate, the introduction of a highly doped germanium pocket at the source-channel junction, the use of high dielectric constant k, and the use of a metallic drain. For this device to be used as a biosensor, nanocavities are introduced underneath the gate (near the source side), which serve as trapping sites for neutral as well as charged biomolecules. Biomolecules have distinctive dielectric constants and charge densities, wherein these two parameters strongly affect the coupling created between the gate electrode and the channel region. The sensed biomolecule, whether neutral or charged, has been used to study properties such as the electric field, surface potential, energy band diagram, and transfer characteristics. For enhanced comprehension of sensor’s sensitivity, a special case of uricase biomolecules has been analyzed. Our simulations have prominently revealed that the value of drain current sensitivity is strongly affected by the location and orientation of the biomolecule inside the nanocavity. Further, we have established that the dimensions of the nanocavity play a pivotal role in enhancing the drain current value. A study has also been carried out to study the effect of noise on the device’s sensing ability. In conclusion, we can say that better sensitivity can be achieved even with minuscule specimens available for trials when a suitable selection of cavity dimensions and proper placement of probes inside the cavity is made.

Acknowledgments

Princess Nourah bint Abdulrahman University Researchers Supporting Project Number (PNURSP2025R919), Princess Nourah bint Abdulrahman University, Riyadh Saudi Arabia.

○.

Department of Electrical Engineering, College of Engineering, Islamic University of Madinah, Madinah, 42351, Saudi Arabia

The authors declare no competing financial interest.

References

  1. Barbaro M., Bonfiglio A. R., Raffo L.. A charge-modulated FET for detection of biomolecular processes: conception, modeling, and simulation. IEEE Trans. Electron Dev. 2006;53(1):158–166. doi: 10.1109/TED.2005.860659. [DOI] [Google Scholar]
  2. Kim C.-H., Jung C.. et al. Novel dielectric modulated field-effect transistor for label-free DNA detection. BioChip J. 2008;2(2):127–134. [Google Scholar]
  3. Seabaugh A. C., Zhang Q.. Low-voltage tunnel transistors for beyond CMOS logic. Proc. IEEE. 2010;98(12):2095–2110. doi: 10.1109/JPROC.2010.2070470. [DOI] [Google Scholar]
  4. Choi W. Y.. et al. Tunneling field-effect transistors (TFETs) with subthreshold swing (SS) less than 60 mV/dec. IEEE Electron Device Lett. 2007;28(8):743–745. doi: 10.1109/LED.2007.901273. [DOI] [Google Scholar]
  5. Wang P. F.. et al. Complementary tunneling transistor for low power application. Solid-State Electron. 2004;48(12):2281–2286. doi: 10.1016/j.sse.2004.04.006. [DOI] [Google Scholar]
  6. Ionescu A. M., Riel H.. Tunnel field-effect transistors as energy efficient electronic switches. Nature. 2011;479(7373):329–337. doi: 10.1038/nature10679. [DOI] [PubMed] [Google Scholar]
  7. Gao X. P. A., Zheng G., Lieber C. M.. Subthreshold regime has theoptimal sensitivity for nanowire FET biosensors. Nano Lett. 2010;10(2):547–552. doi: 10.1021/nl9034219. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Ahn J.-H., Choi S.-J., J.-W.Han T. J. P., Lee S. Y., Lee, andy.-K. Choi S. Y., Choi Y.-K.. Investigation of size dependence on sensitivity fornanowire FET biosensors. IEEE Trans. Nanotechnol. 2011;10(6):1405–1411. doi: 10.1109/TNANO.2011.2157519. [DOI] [Google Scholar]
  9. Rajan N. K., Routenberg D. A., Reed M. A.. Optimal signal-tonoiseratio for silicon nanowire biochemical sensors. Appl. Phys. Lett. 2011;98(26):264107. doi: 10.1063/1.3608155. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Bergveld P.. Development, Operation, and Application of the Ion-Sensitive Field-Effect Transistor as a Tool for Electrophysiology. IEEE Trans. Biomed. Eng. 1972;19(5):342–351. doi: 10.1109/TBME.1972.324137. [DOI] [PubMed] [Google Scholar]
  11. Im H., Huang X.-J., Gu B., Choi Y.-K.. A dielectric-modulatedfield-effect transistor for biosensing. Nat. Nanotechnol. 2007;2(7):430–434. doi: 10.1038/nnano.2007.180. [DOI] [PubMed] [Google Scholar]
  12. Rajani G.. et al. Numerical model of Gate-All-Around MOSFET with vacuum gate dielectric for biomolecule detection. IEEE Electron Device Lett. 2012;33(12):1756–1758. doi: 10.1109/LED.2012.2216247. [DOI] [Google Scholar]
  13. Kannan N., Kumar M. J.. Dielectric-modulated impact-ionization MOS transistor as a label-free biosensor. IEEE Electron Device Lett. 2013;34(12):1575–1577. doi: 10.1109/LED.2013.2283858. [DOI] [Google Scholar]
  14. Kanungo S., Chattopadhyay S., Gupta P. S., Rahaman H.. Comparative performance analysis of the dielectrically modulated full-gate and short-gate tunnel FET-based biosensors. IEEE Trans. Electron Dev. 2015;62(3):994–1001. doi: 10.1109/TED.2015.2390774. [DOI] [Google Scholar]
  15. Kanungo S.. et al. Study and analysis of the effects of SiGe source and pocket-doped channel on sensing performance of dielectrically modulated tunnel FET-based biosensors. IEEE Trans. Electron. Dev. 2016;63(6):2589–2596. doi: 10.1109/TED.2016.2556081. [DOI] [Google Scholar]
  16. Narang R.. et al. A dielectric-modulated tunnel-FET-based biosensor for label-free detection: Analytical modeling study andsensitivity analysis. IEEE Trans. Electron Dev. 2012;59(10):2809–2817. doi: 10.1109/TED.2012.2208115. [DOI] [Google Scholar]
  17. Narang R., Saxena M., Gupta M.. Comparative analysis of dielectric-modulated FET and TFET-based biosensor. IEEE Trans. Nanotechnol. 2015;14(3):427–435. doi: 10.1109/TNANO.2015.2396899. [DOI] [Google Scholar]
  18. Lam K.-T.. et al. A simulation study of graphene-nanoribbon tunneling FET with heterojunction channel. IEEE Electron Dev. Lett. 2010;31(6):555–557. doi: 10.1109/LED.2010.2045339. [DOI] [Google Scholar]
  19. Vishnoi R., Kumar M. J.. Compact analytical drain current model of gate-all-around nanowire tunneling FET. IEEE Trans. Electron. Dev. 2014;61(7):2599–2603. doi: 10.1109/TED.2014.2322762. [DOI] [Google Scholar]
  20. Dwivedi P., Kranti A.. Applicability of transconductance-tocurrentratio (gm/Ids) as a sensing metric for tunnel FETbiosensors. IEEE Sensors J. 2017;17(4):1030–1036. doi: 10.1109/JSEN.2016.2640192. [DOI] [Google Scholar]
  21. Grieshaber D., Mackenzie R., Vörös J., Reimhult E.. ElectrochemicalbiosensorsSensor principles and architectures. Sensors. 2008;8(3):1400–1458. doi: 10.3390/s80314000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kim C.-H., Ahn J.-H., K.-B.Lee C. J., Park H. G., Choi Y.-K., Choi Y.-K.. A new sensing metric to reduce data fluctuations ina nanogap-embedded field-effect transistor biosensor. IEEE Trans. Electron Dev. 2012;59(10):2825–2831. doi: 10.1109/TED.2012.2209650. [DOI] [Google Scholar]
  23. Choi J.-M., Han J.-W., Choi S.-J., Choi Y.-K.. Analytical modelingof a nanogap-embedded FET for application as a biosensor. IEEE Trans. Electron Dev. 2010;57(12):3477–3484. doi: 10.1109/TED.2010.2076152. [DOI] [Google Scholar]
  24. Pandey R., Rajamohanan B., Liu H., Narayanan V., Datta S.. Electrical noise in heterojunction interband tunnel FETs. IEEE Trans. Electron Dev. 2014;61(2):552–560. doi: 10.1109/TED.2013.2293497. [DOI] [Google Scholar]
  25. Fan M.-L., Hu V. P.-H., Chen Y.-N., Su P., Chuang C.-T., Chuang C.-T.. Analysisof single-trap-induced random telegraph noise and its interaction withwork function variation for tunnel FET. IEEE Trans. Electron Dev. 2013;60(6):2038–2044. doi: 10.1109/TED.2013.2258157. [DOI] [Google Scholar]
  26. Cervenka J., Kosina H., Selberherr S., Zhang J., Hrauda N., Stangl J., Bauer G., Vastola G., Marzegalli A., Montalenti F., Miglio L.. TunnelingFETs on SOI: Suppression of ambipolar leakage, low-frequencynoise behavior, and modeling. Solid-State Electron. 2011;65(66):226–233. doi: 10.1016/j.sse.2011.06.041. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Narang R., Saxena M., Gupta R. S., Gupta M.. Dielectric modulatedtunnel field-effect transistorA biomolecule sensor. IEEE Electron Device Lett. 2012;33(2):266–268. doi: 10.1109/LED.2011.2174024. [DOI] [Google Scholar]
  28. Khatami Y., Banerjee K.. “Steep subthreshold slope n- and p-typetunnel-FET devices for low-power and energy-efficient digital circuits,”IEEE Trans . Electron Dev. 2009;56(11):2752–2761. doi: 10.1109/TED.2009.2030831. [DOI] [Google Scholar]
  29. Asra R., Shrivastava M., Murali K. V. R. M., Pandey R. K., Gossner H., Rao V. R.. A tunnel FET for VDD scaling below0.6 V with a CMOS-comparable performance. IEEE Trans. Electron Dev. 2011;58(7):1855–1863. doi: 10.1109/TED.2011.2140322. [DOI] [Google Scholar]
  30. Bhuwalka K. K., Schulze J., Eisele I.. “Scaling the vertical tunnelFET with tunnel bandgap modulation and gate workfunction engineering,”IEEE Trans . Electron Dev. 2005;52(5):909–917. doi: 10.1109/TED.2005.846318. [DOI] [Google Scholar]
  31. Narang R., Saxena M., Gupta R. S., Gupta M.. Assessmentof ambipolar behavior of a tunnel FET and influence of structuralmodifications. J. Semicond. Technol. Sci. 2012;12(4):482–49. doi: 10.5573/JSTS.2012.12.4.482. [DOI] [Google Scholar]
  32. Kao K.-H., Verhulst A. S., Vandenberghe W. G., Sorée B., Groeseneken G., De Meyer K.. Direct and indirect band-to-bandtunneling in germanium-based TFETs. IEEE Trans. Electron Dev. 2012;59(2):292–301. doi: 10.1109/TED.2011.2175228. [DOI] [Google Scholar]
  33. Agarwal S., Klimeck G., Luisier M.. Leakage-reduction designconcepts for low-power vertical tunneling field-effect transistors. IEEE Electron Device Lett. 2010;31(6):621–623. doi: 10.1109/LED.2010.2046011. [DOI] [Google Scholar]
  34. Singh G., Amin S. I., Anand S., Sarin R. K.. Design of Si 0.5 Ge 0.5 based tunnel field effect transistor and its performance evaluation. ”superlatticesmicrostruct. 2016;92:143–156. [Google Scholar]
  35. Vandooren A.. et al. Analysis of trap-assisted tunneling in verticalSi homo-junction and SiGe hetero-junction tunnel-FETs. Solid-State Electron. 2013;83:50–55. doi: 10.1016/j.sse.2013.01.026. [DOI] [Google Scholar]
  36. Nayfeh O. M., Hoyt J. L., Antoniadis D. A.. Strained-Si1–x Gex/Siband-to-band tunneling transistors: Impact of tunnel-junction germaniumcomposition and doping concentration on switching behavior. IEEE Trans. Electron Devices. 2009;56(10):2264–2269. doi: 10.1109/TED.2009.2028055. [DOI] [Google Scholar]
  37. Zhao Q.-T., Richter S., Schulte-Braucks C., Knoll L., Blaeser S., Luong G. V., Trellenkamp S., Schafer A., Tiedemann A., Hartmann J.-M., Bourdelle K.. Strained Si and SiGe nanowire tunnel FETs for logicand analog applications. IEEE J. Electron Devices Soc. 2015;3(3):103–114. doi: 10.1109/JEDS.2015.2400371. [DOI] [Google Scholar]
  38. Khan, A. ; Loan, S. A. . Double gate TFET with germaniumpocket and metal drain using dual oxide International Conference On Microelectronics (ICM) ; IEEE; 2021. 170–173 10.1109/ICM52667.2021.9664949 [DOI] [Google Scholar]
  39. ATLAS Device Simulation Software; Silvaco: Santa Clara, CA, USA, 2016. [Google Scholar]
  40. Boucart K., Ionescu A. M.. Double gate tunnel FETwith high-k gate dielectric. IEEE Trans. Electron Dev. 2007;54(7):1725–1733. doi: 10.1109/TED.2007.899389. [DOI] [Google Scholar]
  41. Mahalaxmi, Acharya B., Mishra G. P.. Designand Analysis of Dual-Metal-Gate Double-Cavity Charge-Plasma-TFET as a Label Free Biosensor. IEEE Sens. J. 2020;20:13969–13975. doi: 10.1109/JSEN.2020.2979016. [DOI] [Google Scholar]
  42. Goswami R.. et al. Comparative analysis of circular gate TFET and Heterojunction TFET for dielectric-modulated label-free biosensing. IEEE Sensors J. 2019;19(21):9600–9609. doi: 10.1109/JSEN.2019.2928182. [DOI] [Google Scholar]
  43. Ajay, Saxena M., Gupta M.. Model of GaSb-InAs p-i-n Gate All Around BioTunnel FET. IEEE Sensors J. 2019;19(7):2605–2612. doi: 10.1109/JSEN.2018.2887277. [DOI] [Google Scholar]
  44. Anand S., Singh A., Amin A. S., Thool A. S.. Design and Performance Analysis of Dielectrically Modulated Doping-Less Tunnel FET-Based Label Free Biosensor. IEEE Sensors J. 2019;19(12):4369–4374. doi: 10.1109/JSEN.2019.2900092. [DOI] [Google Scholar]
  45. Kruskopf, Kruskopf M., Rigosi A. F., Panna A. R., Patel D. K., Jin H., Marzano M., Berilla M., Newell D. B., Elmquist R. E.. et al. Two-Terminal and Multi-Terminal Designs for Next-Generation Quantized Hall Resistance Standards: Contact Material and Geometry. IEEE Trans. Electron Dev. 2019;66(4):1905–1910. doi: 10.1109/TED.2019.2926684. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Ahn J.-H., Choi S.-J., J.-W.Han T. J. P., Lee S. Y., Lee, Lee S. Y., Choi Y.-K.. Investigation of size dependence on sensitivity fornanowire FET biosensors. IEEE Trans. Nanotechnol. 2011;10(6):1405–1411. doi: 10.1109/TNANO.2011.2157519. [DOI] [Google Scholar]
  47. Wadhwa G., Raj B.. Label Free Detection of Biomolecules Using Charge-Plasma-Based Gate Underlap Dielectric Modulated Junctionless TFET. J. Electron. Mater. 2018;47(8):4683–4693. doi: 10.1007/s11664-018-6343-1. [DOI] [Google Scholar]
  48. Wangkheirakpam V. D., Bhowmick B., Pukhrambam P. D.. N+ Pocket Doped Vertical TFET Based Dielectric-ModulatedBiosensor Considering Non-Ideal Hybridization Issue: ASimulationStudy. IEEE Trans. Nanotechnol. 2020;19:156–162. doi: 10.1109/TNANO.2020.2969206. [DOI] [Google Scholar]
  49. Patil M., Gedam A., Mishra G. P.. Performance Assessment of a Cavity on Source ChargePlasmaTFET-Based Biosensor. IEEE Sens. J. 2021;21(3):2526–2532. doi: 10.1109/JSEN.2020.3027031. [DOI] [Google Scholar]
  50. Salim Wani, M. ; Loan, S. A. . Design and Simulation of Dual-Metal-Gate Tunnel Field Effect Transistor with Biomolecule Sensing Applications; Khalifa University Abu Dhabi, 2023. [Google Scholar]
  51. Anam K., Loan S. A.. Metal drain double-gate tunnel field effect transistor with underlap: Design and simulation. Silicon. 2021;13(5):1421–1431. doi: 10.1007/s12633-020-00528-9. [DOI] [Google Scholar]
  52. Anam K., Alharbi A. G., Loan S. A.. Germanium pocke based tunnel FET with underlap: design and simulation. Analog Integr Circ Sig Process. 2023;114(2):207–215. doi: 10.1007/s10470-022-02109-9. [DOI] [Google Scholar]
  53. Munguía-Cortés L.. et al. APTES-Functionalization of SBA-15 Using Ethanol or Toluene: Textural Characterization and Sorption Performance of Carbon Dioxide. J. Mex Chem. Soc. 2017;61(4):273–281. [Google Scholar]
  54. Elmlund L., Suriyanarayanan S., Wiklander J. G., Aastrup T., Nicholls I. A.. et al. Biotin selective polymer nano-films, Elmlund et al. J. Nanobiotechnol. 2014;12(1):8. doi: 10.1186/1477-3155-12-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Kuzuya, A. , et al. Single molecule accommodation of Streptavidin in nanometer scale wells formed in DNA Nanostructures. Nucleic Acids Symp. Ser. 2008. 52, 681–682. 10.1093/nass/nrn344. [DOI] [PubMed] [Google Scholar]
  56. Paliwal A., Tomar M., Gupta V.. et al. Complex dielectric constant of various biomolecules as a function of wavelength using surface plasmon resonance, AIP Publishing. J. Appl. Phys. 2014;116(2):023109–023109–7. [Google Scholar]
  57. National Nanotechnology Initiative. https://www.nano.gov/nanotech-101/what/nano-size#:.
  58. Fan M.-L., Hu V. P.-H., Chen Y.-N., Su P., Chuang C.-T.. Analysis of single-trap-induced random telegraph noise and its interaction with work function variation for tunnel FET”IEEE. Transactions On Electron Dev. 2013;60(6):2038–2044. doi: 10.1109/TED.2013.2259284. [DOI] [Google Scholar]
  59. Bhattacharyya A., De D., Chanda M.. Sensitivity Measurement for Bio-TFET Considering Repulsive Steric Effects With Better Accuracy. IEEE Trans. Nanotechnol. 2022;21:100–109. doi: 10.1109/TNANO.2022.3148922. [DOI] [Google Scholar]
  60. Dharmender, Nigam K. K., Yadav P., Tikkiwal V. A.. Vinay Anand Tikkiwal, Performance analysis of dual material control gate cavity on source electrically doped TFET biosensor for biomedical applications. Micro. Nanostruct. 2024;191:207844. doi: 10.1016/j.micrna.2024.207844. [DOI] [Google Scholar]

Articles from ACS Omega are provided here courtesy of American Chemical Society

RESOURCES