Figure 1: Schematic showing main sectors of biosensor applications.

Moreover, biosensors have been extensively utilized in the field of medical sciences and clinical diagnostics . They have been employed in cancer diagnosis , cardiovascular studies , and diabetes monitoring . The application of biosensors in cancer diagnosis and therapy is very important due to the widespread frequency of the disease, high mortality rate, and recurrence after treatment. In addition, biosensors are applied to monitor blood glucose levels in diabetics, identify infections, and track cancer growth . The development of biosensor technologies for cancer screening is crucial and beneficial . Additionally, biosensors have been used to detect SARS-CoV-2, which causes COVID-19-related severe respiratory distress . For the accurate detection of COVID-19 RNA , proteins , and virus particles , various methods have been proposed, such as CRISPR systems , surface-enhanced Raman spectroscopy , microfluidic-coupled biochip , electrochemical , and field-effect transistor (FET)-based biosensors .

Biosensors offer several distinct benefits for virus recognition, including higher selectivity through improved target receptors and good sensitivity detection via a label-free procedure, real-time electrical signal in situ amplification, and cost-effective mass production, achieved through microelectronic manufacturing processes and a small size for portable point-of-care testing . Additionally, the application of biosensors for accurate detection of viruses , cancer , proteins , DNA, glucose , and nucleic acids has been strongly developed .

On the other hand, specific biomolecule classifications by microbiologists has led to the realization and development of different biosensors, significantly increasing their use in daily life . The first ion-sensitive field-effect transistor (IS FETs) biosensor combined the metal–oxide–semiconductor (MOS) structure with glass electrodes for measuring ion activities in electrochemical and biological environments . Subsequently, hydrogen-sensitive MOSFET technology rapidly emerged . However, all these FETs were highly bulky and required more space. Nakamoto et al. later devised a biosensor with a novel fabrication method aligned well with the complementary metal–oxide–semiconductor (CMOS) fabrication process.

Usually, biosensors convert biological characteristics of the target biomolecules into measurable and quantifiable electrical signals . In this sense, various types of biosensors have been designed using electrical , thermal , and optical signals . Among these, FET-based biosensors have garnered significant attention from researchers due to their desirable properties and advantages, including low cost, label-free operation, high sensitivity, robustness, low power consumption, and a straightforward fabrication process based on CMOS technology .

Figure 2: Schematic showing the organization of FET-based biosensors.

Figure 3: Schematic showing the general structure of DM FET-based biosensors with two nanogaps.

Figure 4: Adequate diameter for detecting different molecule types.

The development of advanced bioelectronic technologies involves exploration of novel material systems as well as emerging device architectures. However, an accurate assessment of different device architectures and their design criteria requires a systematic modeling of the electrical characteristics of the device.

In this context, due to diversities in the transduction and applications of FET-biosensors, the primary design improvisation involves novel device architectures for FET dedicated to biosensing applications. Therefore, an extensive and systematic review on emerging structures of FET-based biosensors is important for the design of future nanoscale FET-biosensors for different applications. There are a few reviews published regarding FET-based biosensors . Most of these works are focused on the materials-based performance optimization of FET-based sensors, and a few reviews of FET-based biosensors report on the analysis and development of nanostructured materials for those biosensors . However, to the best of our knowledge, none of the reported reviews have proposed a comprehensive overview on various structures for the design of FET-based biosensors. Therefore, this current article provides a comprehensive review of FET-based biosensor architectures. More specifically, apart from conventional MOSFET, the development of emerging non-CMOS devices including biosensors based on tunnel-field effect transistors (TFETs) and negative-capacitance field-effect transistors (NCFETs) has also been summarized.

This review is organized as follows: In the first section, we provide an introduction to FET-based biosensors, where we discuss biosensor technology and primary applications. The second section presents the latest emerging structures of FET-based biosensors. The third section presents a comparative analysis between the latest FET-based biosensors. The fourth section presents a summary and future research works. The final section concludes the manuscript.

Figure 5: Schematic showing the cross-sectional representation of a HG NT JL FET-based biosensor.

Figure 6: Schematic showing the cross-section (2D) of a VS NS FET-based biosensor.

Figure 7: Schematic showing the structure of the 3D SRG JL FET-based biosensor.

Another junctionless (JL) surrounding-gate stack FET-based biosensor concept has been analyzed and proposed by Chakraborty et al. . This structure uses two surrounding nanogap cavities separated by HfO2 as a high-k dielectric material and SiO2 as an interface layer. It has been reported that this structure offers higher sensitivity compared to that of the dual-material JL MOSFET-biosensor proposed by Ahangari et al. and the double-gate JL MOSFET biosensor proposed by Narang et al. .

Figure 8: Schematic showing the cutting-plane view of a FinFET-based biosensor.

Various circuit architectures combining both sensing and signal readout functions have been investigated by Rigante et al. . In this context, the FinFET device was implemented in both the metal gate and the sensor device. Moreover, the FinFET structure with a hybrid partially gated design is reported as an exceptional device allowing digital gates with biosensing integration. Also, a hardware description language (Verilog-A) has been used for the modeling of each device structure to be incorporated into various system designs via electronic design automation simulations (EDAS) . Furthermore, an amplifier with common-source architecture has been incorporated into a simple analog circuit for the amplification of the threshold-voltage change, leading to the measurement of pH change at the level of the biosensor surface. Thus, two different circuits for signal conditioning and biosensing have been implemented, such as the pseudo-differential amplifier and ring oscillator circuits. It has been demonstrated that the FinFET-based sensing amplifier exhibits a readout sensitivity that is at least ten times higher than that of individually addressable single sensors, where the sensitivity parameter is limited by the fundamental Nernst equation. It has also been reported that the signal improvement results from the designed circuit, and the sensitivity limit has been addressed by means of the back gate architecture , or the structure of double-side gates .

Figure 9: Schematic showing the cross-section of an NW TFET-based biosensor with an additional electrode (ASE).

Furthermore, the fabrication of a silicon NW TFET-based biosensor has been reported by Gao et al. , targeting high sensitivity, versatility, and reliability for point-of-care diagnostic systems. This device was realized using a top-to-down approach with an anisotropic and cost-effective self-stop etching method . A novel CMOS anisotropic technique was implemented for the etching process, combining classical optical and electron beam lithography with anisotropic wet etching via the tetramethylammonium hydroxide method. This structure offers anti-interference and strong capability, demonstrating inherent ambipolarity through CYFRA21-1 and pH sensing. The device proposed by Gao et al. provides enhanced operational conditions compared to classical FET-based biosensors, with a low limit of quantification and detection.

A real-time simulation of a highly sensitive specific antigen biosensor was also performed by Gao et al. using silicon NW FET-based CMOS technology. In this work, both P- and N-type NW arrays were designed and incorporated into one chip using CMOS technology, combined with optical lithography and an anisotropic self-stop etching method . The incorporated P- and N-type NWs showed complementary electrical responses upon prostate-specific antigen binding, providing a unique means of internal command for biosensing signal verification. It was demonstrated that this biosensor structure is reliable, accurate, and effective at concentrations as low as 1 fg·mL−1, making it suitable for prostate cancer and clinical diagnosis applications. Additionally, this structure is low cost and exhibits a good ability for the detection of biomolecule species.

Wenga et al. designed a step-gate polysilicon FET-based biosensor for the detection of DNA. An emerging step-gate polycrystalline silicon NW FET-based biosensor was fabricated to enable highly sensitive electrical biosensing for DNA hybridization detection with a low-cost and simple fabrication process. The polysilicon NW FET (PSi-NW) is based on a silicon NW transistor with CMOS technology, implemented in highly sensitive biosensors for the detection of biological species. The PSi-NW is synthesized using the sidewall spacer formation technique and designed with a five-mask process with a maximum temperature of 600 °C, featuring N-type devices with different parallel polysilicon channels using the side-wall spacer technique. It has been reported that the developed step-gate PSi-NW FET-based biosensor offers a high surface-to-volume ratio and low-cost devices, allowing better and direct sensitive recognition of various biological species.

Another silicon NW-based biosensor concept was reported by Buitrago et al. . This NW structure uses a vertically stacked architecture and full-depleted body channels for better array concentration. The classical top-down clean room method was employed for the fabrication of the NW structure . Moreover, this device is gated by a back gate and one or two platinum side gates via a liquid which has been characterized for possible implementation in reliable sensing applications. This biosensor uses HfO2 as a high-k gate dielectric material and silicon-on-insulator substrates with low-doped device layers and small nanowire diameters, achieving a fully depleted mechanism and allowing better surface-to-volume ratios and higher sensitivity applications.

Figure 10: Schematic showing the view of an SE SB FET-based biosensor.

Figure 11: Schematic showing the structure of TMD FET-based biosensors without a gate above the nanocavity (a) and with a gate above the nanocavity (b).

2.2.3 Split-gate junctionless FET-based biosensors. A dielectric-modulated split-gate junctionless (SPG JL) FET-based biosensor was introduced by Singh et al. for label-free detection of various biomolecule types (neutral and charged).

Figure 12: Schematic showing the structure of an SPG JL FET-based biosensor.

Figure 13: Schematic showing the structure of a DC DG JL MOSFET-based biosensor.

2.2.5 Charge plasma four-gated MOSFET-based biosensors. Chanda et al. designed a charge plasma four-gated (CP FG) MOSFET-based biosensor structure to enhance the detection of various types of biomolecules. The concept of charge plasma has been developed and implemented using appropriate metal work-function electrodes . This structure is akin to the classical double-gate device, with the exception that the mid-region between the gates is etched out to capture biomolecules. The device reported by Chanda et al. can be realized by following the techniques described by Ahm et al. . Additionally, the proposed structure has a gate length of 300 nm, an underlap length between the gates of 200 nm, a channel thickness of 25 nm, a thickness of the front and back gate oxide of 10 nm, an oxide thickness layer at the mid-region of 1 nm, and a channel doping concentration of 1016 cm−3. Moreover, significant improvements on the on-current state and on the threshold voltage of the device are observed when charged biomolecules are immobilized. It has also been indicated that the CP FG MOSFET-based biosensor provides high sensitivity and a low thermal budgeting scheme. Therefore, this structure could be a promising candidate to replace classical FET-based biosensors due to its low thermal budgeting and compatibility with the fabrication processes of CMOS technology .

Figure 14: Schematic showing the structure of a UII MOS-based biosensor.

2.2.7 Inverter and ring oscillator-based biosensors. Kanungo et al. developed an inverter-based strategy for dielectric-modulated biosensing applications . In this case, a bioring oscillator was incorporated for the detection of biomolecules. Thus, the biomolecule species are detected through the oscillation frequency amplification with conjugation. Therefore, a DM fringing FET-based transducer was implemented as the pull-up and pull-down elements for this bioinverter. Also, different bioinverter structures and configurations have been reported for adequate detection of uncharged and charged biomolecule species. It has been indicated that the sensor/sensor and “P”- type FET-sensor configurations are the most suitable candidates for the detection of uncharged and charged biomolecule species. However, for the realization of bioring oscillators, the sensor/sensor configuration is suitable for the detection of uncharged biomolecules, whereas the P-type FET-sensor configuration is adequate for the detection of positively charged and the sensor N-type configuration is suitable for negatively charged biomolecules.

2.2.8 MoS2 FET-based biosensors. Nam et al. identified two different physical principles for the accurate operation of MoS2 FET-based biosensors based on the positions of antibody functionalization. When antibodies are immobilized at the level of the insulated layer covering the MoS2 device, antibody–antigen binding events mostly change the threshold voltage of the device, which can be described by the classical capacitor-model approach. Moreover, if the antibody particles are grafted at the level of the MoS2 body channel, the binding events mainly modulate the on-state transconductance of the device, credited to the antigen-induced disordered potential in the MoS2 body channel. This physics study of the biosensor device simplifies the biosensor architecture for accurate femtomolar detection and quantification of various biomolecule species.

2.2.9 Vertical-strained impact-ionization MOSFET-based biosensors. Saad et al. reported an equivalent electrical circuit model dedicated to vertical-strained impact-ionization (VSR II) MOSFET, which can be used for biosensing applications. The proposed structure is designed to reduce the supply voltage .

When the strained layer (SiGe) is implemented in the framework level, the threshold and supply voltage are considerably reduced, as well as the leakage current and subthreshold. Therefore, the VSR II MOSFET devices are promising candidates for high-sensitive biosensor applications.

2.2.10 Extended-gate FET-based biosensors. Guan et al. designed an off-chip extended-gate (EG) FET-based biosensor with better sensitivity and non-enzymatic targeting detection of uric acid in serum and urine. In this sense, the authors developed an enzymatic potentiometric technique that can be employed for the accurate determination of uric acid density in human urine and serum. In this context, the detection of uric acid is achieved by the measurements of interfacial potential by means of the extended-gate structure combined with ferrocenyl–alkanethiol-modified gold electrodes . This structure uses reusable back-end detection parts and disposable front-end biosensing chip parts. A quasi-reference electrode is made using Ag/AgCl. The front-end biosensing parts are made using gold electrode material and manufactured by a lithography process, liftoff technique, and metal evaporation process. It has been indicated that the proposed EG FET-based biosensors exhibit super selectivity and high sensitivity for a reliable detection of uric acid in human urine and serum.

Figure 15: Schematic showing the structure of a BE TC OTFT-based biosensor.

2.2.12 Pocket-doped TFET-based biosensors. An emerging pocket-doped tunnel (PKD) FET-based biosensor has been designed for biosensing applications by Rashid et al. .

Figure 16: Schematic showing the structure of a PKD TFET-based biosensor.

Figure 17: Schematic showing the structure of a GE H TFET-based biosensor.

2.2.14 Staggered heterojunction vertical TFET-based biosensors. A GaSb/Si staggered heterojunction vertical tunnel FET (SHV TFET)-based biosensor has been analytically investigated and proposed for biological applications by Khan et al. . The proposed SHV TFET biosensor, with the designed nanocavity segment, offers better sensitivity (≈7.7 × 109) compared to that of classical DM PNPN TFET and full-gate DM TFET . The SHV TFET structure exhibits very good resistance to short-channel effects (SCEs) and leakage currents, making this biosensor structure promising and an excellent candidate for various highly sensitive biomedical applications.

Figure 18: Schematic showing the structure of an NC DG TFET-based biosensor.

It has been demonstrated that the NC DG TFET-based biosensor provides high on-state current by applying lower gate voltage values, creating a steep subthreshold and therefore high current sensitivity. This type of biosensor offers high current sensitivity (9.08 × 1012) compared to that of classical Ge/GaAs TFET (4.04 × 107) and full-gate TFET biosensors (2.5 × 104) .

2.2.16 Pocket-doped vertical TFET-based biosensors. A vertical tunnel field-effect transistor (TFET)-based biosensor with N+-pocket doping (PKD V TFET) has been proposed by Devi et al. to enhance sensitivity in FET biosensors. This structure utilizes a 30% concentration of germanium in the Si/Ge N+ pocket and an N-type doping concentration of 1019 cm−3.

In this case, two different gate metals with distinct work functions were designed to achieve maximum on-state current and an improved Ion/Ioff ratio. The fabrication process flow for the proposed biosensor structure is detailed in . Sensitivity parameters were assessed by measuring the drain-current shift in relation to changes in the dielectric constant. It has been demonstrated that this type of biosensor structure exhibits a significant deviation in the drain current, making the on-state current a suitable sensing parameter. The PKD V TFET-based biosensor has been reported to demonstrate superior sensitivity compared to that of MOSFET biosensors. The proposed PKD V TFET-based biosensor exhibits enhanced sensitivity (approximately 106 for a dielectric constant equal to 12) compared to that of MOSFET biosensors.

Figure 19: Schematic showing the structure of a CG HJ TFET-based biosensor.

2.2.18 Electrically doped junctionless TFET-based biosensors. Chandan et al. investigated electrically doped junctionless tunnel FET (ED JL TFET)-based biosensors. This structure employs a single type of doping concentration, an appropriate work function, and a polarity-gate bias. Additionally, an N+ heavily doped silicon layer has been implemented with two separated gates to develop P+ and intrinsic source regions under the control-gate and the polarity-gate electrode, each with suitable work functions and polarity bias over the silicon region. A nanogap cavity has been embedded within the control-gate dielectric and implemented by etching the control-gate dielectric region towards the polarity-gate side for sensing and biomolecule detection purposes.

It has been reported that ED JL TFET-based biosensors offer a simpler fabrication process, highe