Equivalent Circuit Model Implementation of Dual Channel Vertical Strained Impact Ionization MOSFET (DC-VESIMOS) in Biosensor Applications

Ismail Saad, Andee H. S. Bacho, Chan B. Seng, Mohd. Z. Hamzah, N. Bolong

Nano Engineering & Material (NEMs) Research Group, Faculty of Engineering, Universiti Malaysia Sabah, 88999,

Kota Kinabalu, Sabah, Malaysia. ismail_s@ums.edu.my, abortme09@gmail.com

Abstract—A Dual Channel Vertical Strained Impact Ionization MOSFET (DC-VESIMOS) equivalent circuit model has been successfully designed and modelled in this paper. Miniaturization of biosensor using field effect transistor (FET) based sensors with superb performance and enhanced reliability plays a vital role in transducing component of a biosensor. The importance of subthreshold value (S) towards the sensitivity and selectivity of biosensor contributes to the significance of those parameter in designing and modelling equivalent circuitry model of DC-VESIMOS. A technique facilitating the optimization of impedance element value in the modelled DC-VESIMOS circuitry is explained in order to match the exact S value of the simulated device. DC-VESIMOS reported has a low S value of about 10.98 mV/dec which indicates fast switching behavior that leads to high sensitivity as a biosensor transducer element. A comparison of subthreshold slope between the TCAD simulated device and the PSpice circuitry modelled shows a 86.79% similarity which ensure that the DC-VESIMOS are able to work fine on circuit level implementation and in fact a virtual representation of the actual device characteristics.

Keywords-Equivalent Circuit Model, Biosensor, Impact Ionization, Dual Channel SiGe, ultra-sensitive

I. INTRODUCTION

The development of new analytical approaches in detection, recording and transmitting information in the diagnostics of physiological change or presence of biological and chemical additives in the environment is one of the major needs in modern medicine. The extensive usage of biosensor that combines a biological recognition mechanism with a physical transduction technique probe with higher throughput, smaller sample and set-up sizes, lower cost and easier disposal made sensing viruses and disease in the early stage with correct diagnosis conceivable which plays a vital role in modern medicine [1-2]. Integrated knowledge from different scientific fields that includes biochemistry, molecular biology, physiology, chemistry and physics is required to satisfy the need of modern society for low-cost, fast and reliable biosensing. Biosensors are widely used in biological and chemical analyses, clinical detection, health care, and environmental monitoring [3-4]. However, there is still a need for the development of cheap and ultrasensitive biosensing platform with the possibility of multiplex detection and reliable biosensor [5].

The two most significant characteristics of biosensor are its sensitivity and selectivity towards a target molecule or analyte. Selectivity depends on the attributes of the receptor element since it is where the analyte interacts with biosensor while sensitivity is determined by both biological compound and the biosensor transducer. An excellent recognition of analyte by the receptor element and a very efficient signal transduction to the output system is very essential for a high sensitivity biosensor. Miniaturization and simplification of biosensor are required for point of care diagnostics and lab on chip systems using field effect transistor (FET) based sensors where carbon nanostructures play the role of a transducing component in a biosensor [6-8]. High surface-to-volume ratio of the transducing element can increase the efficiency of the signal transfer. Transistor manufactured today almost twenty times faster and occupies less than one percent of the chip space of those built 20 years ago. However, there are several problems initiated by the semiconductor industry when the device further scaling beyond sub-100 nm. It includes low subthreshold voltage, high threshold voltage and high leakage current which then leads to high power consumption and heating problems [9-11].

The vertical impact ionization MOSFET (IMOS) has been introduced to overcome these issues [12]. However, it requires high supply voltage to obtain the desired better subthreshold slope swing. Therefore, vertical strained silicon germanium impact ionization MOSFET (VESIMOS) has been introduced to improve the device performance with moderate low threshold voltage [13-14]. Silicon Germanium (SiGe) has a lower band gap energy which improves electron mobility and reduce the threshold voltage of the device. To further improve the subthreshold swing and reduce the power consumption of VESIMOS device, a dual channel SiGe layers has been successfully designed and introduced in VESIMOS structure which is believed to be the best promising candidate for an ultra-sensitive and fast response biosensor due to their superb performance and high reliability [15]. In this paper, the equivalent circuit model of dual channel VESIMOS (DC-VESIMOS) was modelled using PSpice to investigate the feasibility of this device as a fantastic transducer element and ultra-sensitive biosensor detector in a circuit level implementation. The DC-VESIMOS was also simulated using Silvaco TCAD to ensure that the equivalent circuit model are mimicking the exact behavior and superb performance of DC-VESIMOS device.

2016 First International Conference on Micro and Nano Technologies, Modelling and Simulation

978-1-5090-2406-3/16 $31.00 © 2016 IEEE

DOI 10.1109/MNTMSim.2016.20

59

2016 First International Conference on Micro and Nano Technologies, Modelling and Simulation

978-1-5090-2406-3/16 $31.00 © 2016 IEEE

DOI 10.1109/MNTMSim.2016.20

59

II. DEVICE STRUCTURE

Figure 1 shows detailed cross-sections of the DC-

VESIMOS simulated using the Silvaco software package for performance analysis [27]. The DC-VESIMOS structure is similar to the vertical IMOS structure [13] with the only difference is the introduction of dual SiGe layers into the Abelein et al., structure. This structure comprises n+ doped source and drain regions, an intrinsic channel containing a highly-doped delta P+ (δp+) layer, dual SiGe layer and two sided gates. The high doping of the delta layer (NA = 4x1019

cm-3) creates a large potential barrier separating source and drain, which are doped the same. It also allows the device to achieve high electric fields near the drain side intrinsic zone without the need for a high supply voltage to initiate the device’s ON state [16]. Since high S values result from a delta layer that is too wide, an optimum value of delta layer thickness and doping concentration was chosen to achieve the desired S value and to ensure that the device works in impact ionization (II) mode.

From the device dimensions shown in figure 1, it can be seen that the thicknesses of the intrinsic Silicon (i-Si) regions are fixed. The i-Si regions reduce the lateral electric field near the source and drain [29] so an optimum thickness for them has been taken to effectively reduce the lateral electric fields. The strained SiGe layer thickness was 20nm with Ge=30%. The dual SiGe layers capped by a very thin intrinsic layer are sandwich between source, δp+ layer and drain region. Due to the presence of i-Si layers between the highly-doped source and drain regions, impurity scattering is also reduced. The source and drain region are n+ doped with doping concentration of 2x1018 cm-3 and thickness of 200 nm. In this paper, the DC-VESIMOS is simulated at VDS = 1.75V where the best subthreshold slope was obtained in this operating voltage which the device is in the Impact Ionization mode.

Fig. 1: DC-VESIMOS Device Structure

III. EQUIVALENT CIRCUIT MODEL

At present, most of the diagnostic estimation are based on

optical measurement. Nevertheless, complex instrumentation, time-consuming sample pre-processing, expansive and compulsory labeling causes optical methods non-feasible for biosensing. Utilization of nanomaterials and FET into the biosensor transducer can significantly mend the optical common detection limitation techniques. A FET contains source and drain electrodes, a semiconducting channel and a gate electrode. Electrical transport through the semiconductor channel is modulated by an applied gate voltage (VGS). The adsorption of additional charges due to biorecognition reactions on the surface of semiconductor will lead to changes in transport characteristics of the device. These changes can be detected by measuring source-drain current of the transistor or subthreshold slope of the device which indicates the sensitivity of the biosensor. DC-VESIMOS structure is more alike of Bipolar Junction Transistor (BJT) with NPN configuration where N+ source, δp+ layer and N+ drain acts as a collector, base and emitter of a BJT.

The DC-VESIMOS equivalent circuit is modelled using ORCAD PSPICE software with a level 3 MOSFET model parameter due to the high sensitivity of DC-VESIMOS [21]. The equivalent circuit model parameter must follow the exact parameter value extracted from the simulated TCAD model to ensure that the excellent performance of DC-VESIMOS are not interrupted and mimicking the exact behavior of simulated TCAD DC-VESIMOS device. The circuit concept is adopted from the Electrostatic Discharge (ESD) protection device circuit [17]. This protection device circuit preventing high current from occurring which may damage the circuitry system due to the uncontrollable supply of drain voltage [18]. The continuous supply voltage causes another breakdown called ‘Near Avalanche and Snapback Breakdown’ which is the continuity of avalanche breakdown to the second order effects [19].

It is not easy to determine the value of resistor and capacitance [20]. To find the value this formula was used.

(1)

where ρsub is the substrate resistivity, and that Csub is in- dependent of ρsub because of rsub ρsub2. The equivalent circuit model incorporates three additional resistors Rg, Rs, and Rd which are used to account losses at the gate, source and drain respectively. It is known that Rs and Rd exhibits gate bias dependence in MOSFETs [22]. The device electrical performance and reliability, saturated drain current (Ion), trans-conductance (gm), noise figure, cut-off frequency and hot carrier degradation effects depends more on Rs rather than Rd [23]. The parasitic overlap capacitance is also considered for the vertical IMOS which is expressed as gate-source capacitance, Cgs and gate-drain capacitance, Cgd. Besides that, the voltage supplies applied to the circuit is used for the capacitor whereas, the extrinsic resistances (Rg,

6060

Rs, and Rd) are used to enhance the fitting quality to the high frequencies. The Vdd is a supply voltage that plays a role to switch on the circuit.

The intrinsic layer which is an oxide layer is represented by locating the capacitor, Ci1 between the drain and body, while the second capacitor, Ci2 represents the second layer is located between the body and the source. The capacitance and resistor are set parallel to each other as the equivalent with the strained SiGe layer. The capacitance, C2 and resistor, R2 are connected from the source to the body as it is for the second layer of strained SiGe layer and for the first layer of capacitance, C1 and resistor, R1 is connected from the body to the drain of the MOSFET. The SiGe layer is believed to increase the mobility carrier of the device or the current flow [15].

Ohm’s Law is used to determine the value of the resistor. To increase the current, the resistance should be lowered down while the applied voltage is constant. In DC-VESIMOS, carrier mobility is higher at near drain region. Thus, the value of resistor should be lower than the extrinsic resistance in order for high current to flow through the drain region. Strain engineering has been applied in an attempt to reduce the supply voltage significantly without compromising excellent performance exhibited in IMOS device. Besides that, it also helps to increase the electron mobility of the device. In DC-VESIMOS, the mobility in the second SiGe layer is not higher compared to the first SiGe layer [15]. Therefore, the second SiGe layer resistor value is set to be higher than the extrinsic resistance to limit the current flow through the source region while allows more current to flow through the drain region. The capacitance of DC-VESIMOS is assumed as an overlap capacitance which is the same value with extrinsic capacitance.

IV. RESULTS AND DISCUSSIONS

The equivalent circuit model consists of n-MOSFET and NPN BJT device has been successfully designed and modelled in PSpice simulator. Figure 2 shows the equivalent circuit modeling for the DC-VESIMOS. The equivalent circuit model parameter used are exactly the same with the simulated TCAD DC-VESIMOS device to conform the modelled circuitry are in fact a virtual representative of the actual device characteristic and simulation works.

Fig. 2: Equivalent Circuit Model of DC-VESIMOS

0 0.5 1 1.5 2 2.5 3 3.5 4

10-15

10-10

10-5

Vds=1.25VVds=1.50VVds=1.75VVds=2.00VVds=2.25V

Fig. 3: Transfer Characteristics of IDS vs VGS of Equivalent Circuit Model

DR

AIN

CU

RR

EN

T, I

DS (

µA

/µm

)

GATE VOLTAGE, VGS

VDS increase

6161

VESIMOS-DP has three distinguished modes of operation depending on the supply voltage being applied: conventional MOSFET, impact ionization (II) and bipolar (BJT) mode. In conventional MOSFET mode, low drain voltage (VDS) and insufficient electric field limits the impact ionization rate which contributed to the high subthreshold value (S). As VDS rises, the electric field in the drain side intrinsic region increases until impact ionization (II) occurs in the drain side intrinsic zone. A significantly lower S value is obtained due to the extremely fast rising current in this II mode. Further increasing the VDS increases the II rate exponentially until the δp+ layer region cannot contain the influx of holes which initialized BJT mode. Although a very low S value were obtained in this mode, the device suffers from Parasitic Bipolar Transistor (PBT) effect which alter the reliability of the device [24]. Figure 3 shows transfer characteristics of the DC-VESIMOS equivalent circuit model.

Ultra-sensitive and fast electrical biosensors require ultra-low power, a low S value, and a high breakdown voltage. In general, a device’s transfer characteristics are examined by biasing its VDS and ramping its VGS at defined bias steps. A change in the gate electrode potential leads to a change in the electrostatic potential drop over the surface of the semiconductor. The adsorption of additional charges due to biorecognition reactions on the surface of semiconductor will lead to a change in transport characteristics of the device. These changes can be detected by measuring source-drain current of the device. Figure 3 revealed that the S value of modelled DC-VESIMOS decrease from 70.02mV/dec at VDS=1.25 to 9.53mV/dec at VDS=1.75V. Subthreshold voltage and leakage currents are in direct proportion and therefore a lower subthreshold voltage ensures low leakage currents [25].

The significantly low S value (S=9. 53 mV/dec) for VDS above 1.75V gives DC-VESIMOS the fastest switching behavior and enhance its electrical performance and response time which are crucial characteristics for bio-based sensors. The only problem detected in the DC-VESIMOS device were when the analyte detected has potential voltage below 1.75V. A viruses or biomolecule below this voltage will not be able be detected abruptly by the transistor due to the high S value below the respective voltage. Figure 4 shows the transfer characteristic comparison of TCAD simulated device and the PSpice circuitry modelled. It is revealed that the subtreshold slope is synchronous with the simulated data extracted and closely approximates of 86.79%. The comparison confirms that the DC-VESIMOS are able to work fine on circuit level implementation and in fact a virtual representation of the actual device characteristics. In the OFF-state operation mode, the device shows a leakage current which is independent to the VGS, but increases with the increment of VDS.

0 0.5 1 1.5 2 2.5 3 3.5 410

-18

10-16

10-14

10-12

10-10

10-8

10-6

10-4

10-2

TCAD Simulation

PSPICE Simulation

Fig. 4: Comparisons TCAD and PSPICE Simulation of Transfer Characteristics of IDS vs VGS at VDS = 1.75V

A very low off-state leakage current (IOFF=10-16 µA/µm)

and good drive current (ION=10-3 µA/μm) at VDS=1.75V is explicitly shown for DC-VESIMOS device in TCAD simulation.While, PSPICE simulation results observed that the IOFF=10-17µA/µm and the drive current, ION=10-4µA/µm respectively. With regard to output characteristics, a very good drain current at different gate voltages with increasing drain voltage was observed and is shown in figure 5. This is due to the existence of the strained SiGe in the channel region which enhanced carrier transport in the VESIMOS-DP channel. Figure 5 shows the excellent performance of the output characteristic for the DC-VESIMOS equivalent circuit model.

0 0.5 1 1.5 2 2.5 310

-15

10-10

10-5

100

Vgs=1.25V

Vgs=1.50V

Vgs=1.75V

Fig. 5: Output Characteristic of IDS vs VDS of Equivalent Circuit Model

DR

AIN

CU

RR

EN

T, I

DS (

µA

/µm

)

DRAIN VOLTAGE, VDS

DR

AIN

CU

RR

EN

T, I

DS (

µA

/µm

)

GATE VOLTAGE, VGS

6262

0 0.5 1 1.5 2 2.5 310

-8

10-6

10-4

10-2

100

PSPICE Simulation

TCAD Simulation

Fig. 6: Comparisons of TCAD and PSPICE Simulation of the Output

Characteristic of IDS vs VDS at VGS=1.75V.

The figure revealed that the equivalent circuit model has a very good drain current at different gate voltage with the increment of VDS. The drain current rises steadily before going into breakdown state as VDS increases. The device undergoes a breakdown state above VDS=2.4V, which is considered exceptional as most operating voltage for biosensor application is below 1.5V. A comparison of output characteristic between TCAD simulated device and the PSpice equivalent circuit model is shown in figure 6. Figure 6 revealed that the breakdown voltage of the TCAD simulated device is 2.6V at VGS=1.75V. The difference in breakdown voltage was the circuit model was assumed at ideal condition without any external disturbance that can alter the performance of the device.

Fig. 7: Equivalent Circuit Model illustrated as a circuit biosensor

0 0.5 1 1.5 2 2.5 3-1

0

1

2

3

4

5

6

Vin=0V

Vin=0.1V

Vin=0.2V

Vin=0.3V

Vin=0.4V

Vin=0.5V

Fig. 8: Current-Voltage characteristics of difference in potential of the input.

Any type of molecules or analyte detected is determined by how the sensor is working. In terms of the electrical signal, any types of sensor will detect potential difference of the samples. This occurs because of the materials used can convert the biological response to the electrical signal. In this research, the researchers try to illustrate the circuit model [26] as a sensor by varying the input voltage and the output observed is measured as shown in figure 7.

Figure 8 shows the corresponding output when the input voltage is varied. Small potential difference values are used, to see how the corresponding output. Besides that, the researchers also wanted to illustrate any samples with small potential difference can be detected by the sensor. This means that if the sensor can detect it, the device has a probability of high sensitivity because of the sensing capabilities which sense small changes or small potential difference.

0 0.5 1 1.5 2 2.5 3-1

0

1

2

3

4

5

6

7

PSPICE Simulation at Vin=0.5V

F. Ramli at pH=3

Fig. 9: Comparisons of output voltage of pH sensor [27] with the PSPICE output voltage act as circuit sensor

DR

AIN

CU

RR

EN

T, I

DS (

µA

/µm

)

DRAIN VOLTAGE, VDS

CU

RR

EN

T (

mA

)

OUTPUT VOLTAGE

CU

RR

EN

T (

mA

)

OUTPUT VOLTAGE

6363

The researchers also compared the circuit model with experimental works by sensing pH value [27] as shown in figure 9. The experimental works with pH 3 shows them having almost the same pattern of the graph when the input voltage of 0.5V. This is to prove that the DC-VESIMOS device in terms of circuit model is applicable to detect the changes with small difference in potential. Not apart from that, any value of potential difference is reliable for the DC-VESIMOS device to measure or detect it and show at the output signal. Hence, the equivalent circuit model of the DC-VESIMOS is viable for future biosensor applications.

V. CONCLUSIONS

The equivalent circuit model of DC-VESIMOS is invented by using the concept of ‘Near Avalance and Snapback Breakdown’. The TCAD and PSPICE simulation of DC-VESIMOS is successfully simulated. By comparing the analysis in PSPICE and TCAD simulation of DC-VESIMOS, it is shown that there is 86.79% similarity in terms of subthreshold slope which is important in the sensitivity of the device. However, the current ratio of ON and OFF is found to be almost the same. The equivalent circuit model of DC-VESIMOS is simulated to be used as based biosensor applications. The results showed that the circuit model is applicable to become circuit biosensor for biological recognition in the future.

ACKNOWLEDGMENT

The authors would like to acknowledge the financial support from FRGS (FRGS0302-TK02). The authors also express their deep gratitude to the Universiti Malaysia Sabah (UMS) for providing an excellent research environment to complete this work.

REFERENCES [1] Cornell B. A., et al. "A biosensor that uses ion-channel

switches." NATURE-LONDON-(1997): 580-582.

[2] Eggins, Brian R. Biosensors: an introduction. Springer-Verlag (2013). p. 2.

[3] Baby, Tessy Theres, et al. "Metal decorated graphene nanosheets as immobilization matrix for amperometric glucose biosensor." Sensors and Actuators B: Chemical 145.1 (2010): 71-77.

[4] Eggins, Brian R. Biosensors: an introduction. Springer-Verlag (2013). pp.10-11.

[5] Chen, Kuan-I., et al. "Silicon nanowire field-effect transistor-based biosensors for biomedical diagnosis and cellular recording investigation." Nano Today 6.2 (2011): pp.131-154.

[6] Xia, Lin, et al. "Conducting polymer nanostructures and their application in biosensors." Journal of colloid and interface science 341.1 (2010): pp.1-11.

[7] Arya, Sunil K., et al. "Recent advances in ZnO nanostructures and thin films for biosensor applications: review." Analytica chimica acta 737 (2012): pp.1-21.

[8] Li, Feng, et al. "Development of an electrochemical DNA biosensor with a high sensitivity of fM by dendritic gold nanostructure modified electrode." Biosensors and Bioelectronics 26.5 (2011): pp.2619-2625.

[9] Gopalakrishnan Kailash, et al. "I-MOS: A novel semiconductor device with a subthreshold slope lower than kT/q." Electron Devices Meeting, 2002. IEDM'02. International. IEEE, 2002.

[10] Gopalakrishnan, Kailash, et al. "Impact ionization MOS (I-MOS)-Part I: device and circuit simulations." Electron Devices, IEEE Transactions on 52.1 (2005): pp. 69-76.

[11] Gopalakrishnan, Kailash, et al. "Impact Ionization MOS (I-MOS)-part II: experimental results." Electron Devices, IEEE Transactions on 52.1 (2005): pp. 77-84.

[12] Abelein, U., et al. "A novel vertical impact ionisation MOSFET (IMOS) concept." Microelectronics, 2006 25th International Conference on. IEEE, 2006.

[13] Toh, Eng-Huat, et al. "A complementary-I-MOS technology featuring SiGe channel and i-region for enhancement of impact-ionization, breakdown voltage, and performance." Solid State Device Research Conference, 2007.

[14] Saad, Ismail, et al. "The vertical strained impact ionization MOSFET (VESIMOS) for ultra-sensitive biosensor application." Regional Micro and Nanoelectronics (RSM), (2015): pp. 1-4.

[15] Saad, Ismail, et al. "Single and dual strained channel analysis of vertical strained—SiGe impact ionization MOSFET (VESIMOS)." Regional Micro and Nanoelectronics (RSM), 2013.

[16] Abelein, Ulrich, et al. "Improved reliability by reduction of hot-electron damage in the vertical impact-ionization MOSFET (I-MOS)." Electron Device Letters, IEEE 28.1 (2007): pp. 65-67.

[17] Toteva, Ina, and Anna Andonova. "Modeling NMOS snapback characteristic using PSpice." Electronics Technology (ISSE), 2012 35th International Spring Seminar on. IEEE, 2012.

[18] Tam, Wing-Shan, et al. "Snapback breakdown ESD device based on zener diodes on silicon-on-insulator technology." Microelectronics Reliability 54.6 (2014): 1163-1168.

[19] Donald A. Neamen. "Semiconductor Physics and Devices". Fourth Edition. New York: McGraw-Hill. (2012). pp 465-468.

[20] Pfost, Martin, Hans-Martin Rein, and Thomas Holzwarth. "Modeling substrate effects in the design of high-speed Si-bipolar ICs." Solid-State Circuits, IEEE Journal of 31.10 (1996): pp 1493-1501.

[21] Vladimirescu, Andrei, and Sally Liu. The simulation of MOS integrated circuits using SPICE2. No. UCB/ERL-M80/7. CALIFORNIA UNIV BERKELEY ELECTRONICS RESEARCH LAB, 1980.

[22] Hu, Genda J., Chi Chang, and Yu-Tai Chia. "Gate-voltage-dependent effective channel length and series resistance of LDD MOSFET's." Electron Devices, IEEE Transactions on 34.12 (1987): pp 2469-2475.

[23] Chen, Jone F., et al. "Performance and reliability comparison between asymmetric and symmetric LDD devices and logic gates." Solid-State Circuits, IEEE Journal of 34.3 (1999): pp 367-371.

[24] Saad, Ismail, et al. "Investigation of incorporating dielectric pocket (DP) on Vertical Strained-SiGe Impact Ionization MOSFET (VESIMOS-DP). (ICSE) (2012): pp. 249-253.

[25] Roy, K., et al. "Leakage current mechanisms and leakage reduction techniques in deep-submicrometer CMOS circuits," in Proc. IEEE, vol. 91, no. 2, pp. 305-327, 2003.

[26] Rahman, Md Mahbubur, et al. "A comprehensive review of glucose biosensors based on nanostructured metal-oxides." Sensors 10.5 (2010): pp.4855-4886.

[27] Ramli, NF, “FABRICATION OF POROUS SILICON FOR pH-BASED BIOSENSOR”. Faculty of Electrical (UTM). 2014

[28] Atlas and Athena user Manual Device and Process Simulation Software, Silvaco International, 2005.

[29] A. C. Lamb, L. S. Riley, S. Hall, V. D. Kunz, C. H. de Groot and P. Ashburn, “A 50 nm vertical MOSFET concept incorporating a retrograde channel and a dielectric pocket. 31st European Solid-State Device Research Conference (ESSDERC 2001), 2001, Pg. 347–350. DOI:10.1109/ESSDERC.2001.195272

6464