Comparative performance analysis of JL DG- MOSFET with Underlap JL DG-MOSFET

Comparative performance analysis of JL DG- MOSFET with Underlap JL DG-MOSFET

Roopayan Patnaik1, Rohan Kumar Majhi 2, Prabin Kumar Roul3, Biswajit Baral4

Department of Electronics and Telecommunication Engineering Silicon Institute of Technology, Bhubaneswar

Abstract-The CMOS technology has seen immense development in the last decade with number of analog and RF applications. An important aspect of this technology now is system on chip applications i.e. SOC applications. Numerous investigatory studies have been done on Junctionless devices to examine their aptness for various applications. Though the absence of junctions and presence of double gates enhanced the device performance, it was sighted to be susceptible to SCEs (Short Channel Effects). The replacement of this architecture by an Underlap device showed reduced SCEs. The gate underlap also reduces gate edge direct tunneling leakage [9] and gate sidewall fringe capacitance, resulting in an improved circuit performance. This paper investigates the D.C, analog and linear performance of the aforesaid devices. Further comparative investigations are done using TCAD SILVACO device simulator to establish the superiority of either device for specific applications.

Keywords Junction less transistor, Transconductance(gm), Transconductance Generation Factor(gm/id),Parasitic capacitances

CMOS technologies, which is low cost. The inter modulation distortion & higher order harmonics at the output need to be mitigated. Various figure of merits (FOM) such as Transconductance(gm), Gate to source and Gate to drain capacitances(CGS and CGD), Transconductance Generation Factor(TGF) and VIP2. Further, different parameters like work function, substrate doping, oxide layer of both the devices are also studied in order to obtain overall device performance. The result obtained from both models are verified and simulated for obtaining a good result, validating the models.

This paper is organized as follows: Section II illustrates the device structure and the simulation setup considered in this study. Section III provides the variation of analog performance parameter as a function gate-length downscaling. In section IV gives the idea about the impact of gate length downscaling on RF performance. Section V provides the Linearity performance. Finally, in section VI, the conclusions is drawn.

1. INTRODUCTION

   The remarkable performance of CMOS based devices has made the CMOS technology a formidable force in the digital market. With the need for increase in packaging density on IC and low power consumption the downscaling of devices has become an obvious need [1]. However downscaling leads to more pronounced SCEs like DIBL(Drain Induced Barrier Lowering),high off state leakage current, Threshold voltage roll off, bringing down the device performance [3-6]. In order to overcome DIBL and SCEs the Double Gate (DG) structure has been proposed. The absence of junctions (Junction-less) offers simplicity of design as well as overcomes the problem of leakage current. A plethora of investigative studies have been done on JL DG-MOSFET establishing their superior DC, Analog, RF and Linear performance. The substrate of the device proposed is Silicon (Si) and oxide layer is SiO2. However, aggressive downscaling in sub 45nm region has led to effect of SCEs even in the double gate device. Consideration of fabrication defects and theoretical studies have brought another device with Underlap Gate structure into domain of research. The basic D.C investigation is necessary to establish the efficiency of device under biasing conditions. Analog investigation and study of parasitic capacitances of device are necessary to model the device. Linearity property of any device is highly important to predict their performance in mobile communication, computing & multimedia application has resulted interest in SOC application based on

2. DEVICE STRUCTURE AND SIMULATION

   Fig. 1: 2D schematic view of a DG JL FET

   The structure under consideration have uniform doping type and concentration 1019 cm-3 throughout the structures. L is the channel length with an oxide thickness tox=0.8nm,thickness of

   Si body tsi=10nm. A p-type polysilicon gate with work function m=5.5eV is used. The MOS model is selected using Newton Maxtrap method. while AuGe/Ni/Au is used for the formation of ohmic contacts in different terminals. The two dimensional device performances are simulated for JL-DG- MOSFET and Underlap JL-DG-MOSFET using SILVACO- ATLAS [10]. In this simulation Fermi Dirac carrier statistics

   along with conventional Drift Diffusion (DD) model has been adapted to model carrier transport. Shockley-Read-hall (SRH) recombination model combined with Auger Recombination model has been employed in order to model recombination

   characteristics. Newton and Gummels numerical iteration method are used to solve coupled differential equation such as continuity equation of electron and hole and current equation of electron and hole in ATLAS. The simulation software was used to generate layouts for both normal and underlap device.

   Fig. 2: Device Layout of Underlap JL DG-MOSFET using SILVACO TCAD device simulator for 10nm channel length with Si substrate

   Fig. 3: Variation of Drain Current ID as a function of Gate-to-source voltage VGS for channel length L=10nm with device parameter values VDS=0.2V, tSi=10nm and tOX=0.8nm

3. ANALOG PERFORMANCE INVESTIGATION

   This Section shows evaluation of the analog performance parameter of Junction-less DG MOSFET and Underlap Junction-less DG MOSFET.

   Fig. 4. Plot of variation of transconductance (gm) as a function of Gate-to- source voltage Vgs for different Channel Length L=10nm for normal and underlap device with device parameter values VDS=0.2v, tsi=10nm and tox=0.8nm.

   t is clear that in Fig. 4 that transconductance increase with Gate to source Voltage of both the devices are comparable.

   I

   Fig.5. Variation of Transconductance Generation Factor (TGF) as a function of Drain current Id for Channel Length L=10nm with device

   parameter values Vds=0.2V, tsi=10nm and tox=0.8nm.

   Fig.6. Variation of Transconductance Generation Factor (TGF) as a function of Drain current Id for different Channel Length L=10nm with device

   parameter values Vds=0.2V, tsi=10nm and tox=0.8nm for Underlap device

   Fig. 4&5, this figure of merit shows that the measure of efficiency to convert the current into transconductance. It indicates that lower TGF reduces input device performance and higher power dissipation. It is clear from Fig. 4&5 that lower TGF is obtained with a device of shorter gate length.

4. RF PERFORMANCE INVESTIGATION

   This Section shows evaluation of the analog performance parameter of Junction-less DG MOSFET .The cutoff frequency fT and the maximum frequency of oscillation fmax are two important parameters for evaluating the device potentials for RF applications. The cutoff frequency fT is the frequency when the current gain is unity ,whereas fmax is the frequency when the power gain is unity [11].

   Standard figure of merits (FOMs) require to examine the RF performance are,

   1. Cut-off frequency (fT)

   2. Maximum frequency of oscillation (fmax)

   3. Gain Bandwidth Product (GBW)

   The standard analytical expression for evaluating fT, fmax, GBW are given by

   Fig. 7 Variation of Gate-to-Drain Capacitance Cgd as a function of Vgs for Channel Length L=10nm with device parameter values Vds=0.2V, tsi=10nm and tox=0.8nm

   .

   Fig. 8: Variation of Gate-to-Source Capaciance Cgs as a function of Vgs for different Channel Length L=10nm. with device parameter values Vds=0.2V, tsi=10nm and tox=0.8nm for Underlap device

   Fig. 7 shows the plot of Cgs as a function of Vgs for 10nm JL DG-MOSFET. Fig. 8 shows the plot of Cgs as a function of Vgs for 10nm underlap JL DG-MOSFET. As we see the values of Cgs are substantially lower compared to normal JL DG MOSFET.

   gm

   gm

   gm

   gm

   fT

   gm

   2Cgs

   1 2 Cgd

   C

   2 Cgd Cgs

   2Cgg

   gs (1)

   fmax

   gm (2)

   c

   2Cgs

   4R R R g g

   s i g ds m c

   GBW gm

   2 10 Cgd

   gd

   gs (3)

   Fig. 9: Variation of Cut-off Frequency (fT ) as a function of Gate to Source Voltage Vgs for different Channel Length L=10nm with device parameter

   where, Cgs , Cgg , Cgd represents the Gate to Source, Gate, Drain capacitance and total gate capacitance including the fringing and overlap parasitic capacitances. .

   values VDS=0.2V,tSi=10nm and tOX=0.8nm.

   At the sub-threshold region fT is proportional to a 1/L2 as gm proportional to1/L and Cgd/Cgs proportional to L fT attains a

   lower value and increases with ID until it reaches a maximum value at a specific gate bias. At maximum transconductance fT is at maximum point and gate to source/drain capacitance is minimum.

5. LINEARITY TESTING

   Linearity is an essential requirement in all RF system in order to ensure minimal inter-modulation and higher order harmonics at the output. Intermodulation Distortion (IMD) due to non-linearity, generate unwanted signal with different frequencies [12]-[15]. These unwanted (noise) signals may interfere, change or even corrupt the desired output components. So there is a need to estimate the linearity distortion analysis for JL DG MOSFETs as well as Underlap device. A transistor-level linearization is more appropriate for power amplifiers in portable systems, which requires an analysis of the linearity behavior at the device level.

   In this section we investigate the RF performance using standard figure of merits

   4.gm1

   concluded that Underlap JL DG MOSFET is a competitive contender for next generation SOC applications with an improved linearity and analog performance. The major drawback of the device is its high power dissipation and low TGF. However, it may be a contender for select RF/Analog applications. Our work also opens up probable substrate variations, change in number of gates and their material, use of higher order dielectrics and other investigations.

   REFERENCES

   1. N. Mohankumar,Binit Syamal, Chandan Kumar Sarkar N. Mohankumar, "Influence of Channel and Gate Engineering on the Analog and RF Performance of DG MOSFETs," IEEE TRANSACTIONS ON ELECTRON DEVICES, vol. 57, no. 4, pp. 820-826, April 2010.

   2. P. H. Woerlee, M. J. Knitel, R. van Langevelde, D. B. M. Klaassen, L.

      F. Tiemeijer, A. J. Scholten, and A. T. A. Zegers-van Duijnhoven, RFCMOS performance trends, IEEE Trans. Electron Devices, vol. 48, no. 8, pp. 17761782, Aug. 2001.

   3. V. Kilchytska, A. Nève, L. Vancaillie, D. Levacq, S. Adriaensen,H. van Meer, K. De Meyer, C. Raynaud, M. Dehan, J.-P. Raskin, and D. Flandre, Influence of device engineering on the analog and RF

      performance of SOI MOSFETs, IEEE Trans. Electron Devices, vol. 50,

      VIP2

      gm2

      const. Vds

      no. 3, pp. 577588, Mar. 2003.

      , (4) [4] D. J. Frank, R. H. Dennard, E. Nowak, P. M. Solomon, Y. Taur, and H.-

      S. P. Wong, Device scaling limits of Si MOSFETs and their application

      dependencies, Proc. IEEE, vol. 89, no. 3, pp. 259288, Mar. 2001.

      1. S. S. Suryagandh, M. Garg, and J. C. S. Woo, A device design

         methodology for sub-100-nm SOC applications using bulk and SOI

         MOSFETs, IEEE Trans. Electron Devices, vol. 51, no. 7, pp. 1122 1128, Jul. 2004.

      2. J. Liang, H. Xiao, R. Huang, P.Wang, and Y.Wang, Design

         optimization of structural parameters in double gate MOSFETs for RF

         applications, Semicond. Sci. Technol., vol. 23, no. 5, pp. 18, May 2008.

      3. J.P. Colinge, C.W. Lee, A. Afzalian, N. Dehdashti Akhavan, R. Yan, I. Ferain, P. Razavi, B. ONeill, A. Blake, M. White, A.M. Kelleher, B. McCarthy, R. Murphy, " Nanowire transistors without junctions," Nature Nanotechnology, vol. 5, no. 3, pp. 225-229,2010.

      4. Yogesh Pratap, Subhasis Haldar, R.S Gupta and Mridula Gupta,

         Linearity Performance Investigation of high-k Spacer based Junctionless

         Nanowire Transistor (JLNWT) for RFIC Design, ISDRS,2013.

      5. W. Ma, S. Kaya, A. Asenov, Study of RF linearity in sub- 50nm MOSFETs Using Simulations, Journal of Computational Electronics,vol. 2, no. 2-4, pp. 347-352,2000.

         Fig. 10: Variation of VIP2 as a function of Gate to Source Voltage Vgs for Channel Length L=10nm for simple device compared with underlap with device parameter values VDS=0.2v,tSi=10nm and tOX=0.8nm.

         The comparison plot shows distinctively low VIP2 values for the Underlap device compared to the simple JL DG MOSFET. It can be a prospective linear device.

6. CONCLUSION

In this paper, selected FOMs(Figure Of Merits) of DC, Analog, RF and linearity performances of JL DG MOSFET and Underlap JL DG MOSFET were compared in terms of transconductance Generation Factor(TGF) gm/Ids, cutoff frequency fT, gain band width product GBW and VIP2.Comparison has been performed for different channel length and it was observed that device shows excellent RF performance for shorter channel length but the analog performance of the Device was poor. Hence, this paper

1. ATLAS Device Simulation Software, Silvaco Int., Version 5.14.0.R

2. J. Liang, H. Xiao, R. Huang, P.Wang, and Y.Wang, Design

   optimization of structural parameters in double gate MOSFETs for RF

   applications, Semicond. Sci. Technol., vol. 23, no. 5, pp. 18, May 2008.

3. S. Kaya, W. Ma, and A. Asenov, (2003) Design of DG-MOSFETs for High Linearity PerformanceIEEE International SOI Conference, pp. 68-69.

4. H. Zhang, E. Sánchez-sinencio, and L. Fellow, Linearization techniques for CMOS low noise amplifiers: A tutorial, IEEE Trans. Syst. Circuits, vol. 58, no. 1, pp. 2236, Jan. 2011.

5. Y. Ding and R. Harjani, A CMOS high efficiency +22 dBm linear power amplifier, in Proc. IEEE Custom Integr. Circuit Conf., Oct. 2004,pp. 557560.

6. B. Razavi, RF Microelectronics, Prentice-Hall, 1998

7. Sarkar A, Das AK, De S, Sarkar CK. Effect of gate engineering in double-gate MOSFETs for analog/RF applications, Microelectron. J 2012; 43:873-882

8. Baishya S, Mallik A, Sarkar CK. A subthreshold surface potential model for short-channel MOSFET taking into account the varying depth of channel depletion layer due to source and drain Junctions, IEEE Transactions on Electron Devices 2006; 53:507-14.