Documente Academic
Documente Profesional
Documente Cultură
Sohini Mondal Rajasree Hazra Shaswati Sardar Ratna Rajak Soham Maiti
B.TECH
IN Electronics & Communication Engineering
INSTITUTE OF SCIENCE & TECHNOLOGY (Approved by AICTE & Affiliated to West Bengal University of Technology)
UNDER THE GUIDANCE OF
ABSTRACT
With the advent of semiconductor technology in VLSI era, the channel length of a Metal oxide semiconductor has drastically gone down. Drain Induced Barrier Lowering is one of the short channel effect which degrades the performance of a MOSFET with its down scaling. To understand this effect the study of the nature of surface potential and energy is very important. In this project an analytical model for threshold voltage of shortchannel MOSFETs is presented. For such devices, the depletion regions due to source/drain junctions occupy a large portion of the channel, and hence are very important for accurate modeling. The proposed threshold voltage model is based on a realistic physically-based model for the depletion layer depth along the channel that takes into account its variation due to the source and drain junctions. With this, the unrealistic assumption of a constant depletion layer depth has been removed, resulting in an accurate prediction of the threshold voltage. The proposed model can predict the drain induced barrier lowering (DIBL) effect and hence, the threshold voltage roll-off characteristics quite accurately. The model predictions are verified against the simulator MATLAB and P-Spice.
CONTENTS
CHAPTER 1: INTRODUCTION
1.1 MOTIVATION OF THE THESIS 1.2 ANALYTICAL MODEL 1.3 ORGANISATION OF THE THESIS 1-6 2 3 6
16-18 16 19-27
4.1 MAT-LAB PROGRAM FOR VERIFICATION OF THRESHOLD 19 VOLTAGE WITH CHANNEL LENGTH 4.2 MAT-LAB PROGRAM FOR VERIFICATION OF THRESHOLD 23 VOLTAGE WITH CHANNEL LENGTH 4.3 MATLAB PROGRAM FOR VARIATION OF THRESHOLD 27 VOLTAGE REDUCTION WITH CHANNEL LENGTH
Chapter 1:
Introduction:
Demand for larger scale integration of MOS circuits on a single chip urged of miniaturization of MOS devices. As the channel length shrinks, many short channel effects were observed mainly reduction of threshold voltage, increased off-state leakage current and Drain-induced barrier lowering. Charge sharing model have been used to model the short-channel effects(SCEs). The charge sharing model assumptions of constant surface potential and no divergence of electric field lines in the gate oxide are invalid for high drain and substrate biases. On the other hand, two-dimensional analysis has accurately predicts the values of threshold voltage of short channel MOSFETs and breakdown voltage. We have derived the analytical relations for surface potential, threshold voltage and longitudinal field. Without any assumption, we have proposed a model for short channel factor which shows the dependence on the channel width and drain voltage. We have compared our results with various reported results. The effect of charge carrier density is to raise the threshold voltage of submicron devices operating at any drain voltage. It is observed from our analysis that short-channel effect is more dominant in the sub-micron devices with thinner gate oxide. Our study also predicts that DIBL effect in short channel device is more effective in presence of charge carrier density. It is also observed that the short channel factor shows a weak dependence on the substrate doping due to the inclusion of carrier charge density in the model. 1
1.1 Motivation:
The barrier lowering results in a shift of threshold voltage as a function of the drain voltage for short channel device because the depletion widths at source and drain junction becomes comparable to the channel length. The depletion region near the source and drain approach each other thus decreasing the potential barrier.With increase in the drain bias further result in lowering of barrier, which in turn causes substantial leakage current to flow. Detailed study of this variation of barrier potential is needed to understand the variation of surface potential along the channel for different drain biases. This effect is found by solving 2-D Poissons equation . The channel length shrinks, the lateral field due to the penetration of edge effect into the channel region becomes very important . Charge sharing model ,assuming a trapezoidal depleted channel region start penetrating from two ends. Ratnakumar and Meindl have solved threshold voltage model with boundary condition that assume constant surface potential . Study of threshold voltage reduction due to different parameter like drain to source voltage ,surface potential, channel length and accepter concentration.
where is a constant of integration. The electric field is assumed to be zero in the neutral p region for x < -x, since the currents are zero in thermal equilibrium. As there are no surface charge densities within the p-n junction structure. the electric field is a continuous function. The constant of integration is determined by setting E = 0 at x = -x,. The electric field in the p region is then given by E = = In the n region, the electric field is determined from E = The constant C2 is determined by setting the E-field is assumed to be zero E = Again = . This equation states that the number of negative charges per unit area in the p region is equal to the number of positive charges per unit area in the n region. The potential in the junction is found by integrating the electric field. In the p region then, we have
or = where C1 is again a constant of integration. The potential difference through the p-n junction is the important parameter, rather than the absolute potential, so we may arbitrarily set the potential equal to zero at x = -Xp. The constant of integration is The potential in the p region can now be written as = The potential in these region is determined by integrating the electric field in the n region
Then = integration. =
where
is another constant of
Fig: 2 Electronic potential through space charge of uniformly doped p-n junction.
CHAPTER 2
The threshold voltage shift (DIBL) varied linearly with the drain voltage at drain terminal and the DIBL parameter varied with the channel length.
the same order of magnitude as the depletion-layer widths (XdD, XdS) of the source and drain junction. As the channel length L is reduced to increase both the operation speed and the number of components per chip, the so-called short-channel effects arise. The short-channel effects are attributed to two physical phenomena: i) the limitation imposed on electron drift characteristics in the channel, ii). the modification of the threshold voltage due to the shortening channel length. In particular five different short-channel effects can be distinguished: a. Drain-induced barrier lowering and punch through b. Surface scattering c. Velocity saturation d. Impact ionization e. Hot electrons
10
11
The electric flux lines generated by the charge on the MOS capacitor gate electrode terminate on the induced mobile carriers in the depletion region just under the gate. For short-channel MOSFETs, on the other hand, some of the field lines originating from the source and the drain electrodes terminate on charges in the channel region. Thus, less gate voltage is required to cause inversion. This implies that the fraction of the bulk depletion charge originating from the p-n junction depletion and hence requiring no gate voltage, must be subtracted from the V expression.
Fig : 4 Geometry of the gate-induced bulk depletion region The figure shows the simplified geometry of the gate-induced bulk depletion region and the p-n junction depletion regions in a short channel MOS transistor.
12
Note that the bulk depletion region is assumed to have and asymmetric trapezoidal shape, instead of a rectangular shape, to represent accurately the gate-induced charge. The drain depletion region is expected to be larger than the source depletion region because the positive drain-to-source voltage reversed-biases the drain substrate junction. We recognize that a significant portion of the total depletion region charge under the gate is actually due to the source and drain junction depletion, rather than the bulk depletion induced by the gate voltage. Since the bulk depletion charge in the short channel device is smaller than expected, the threshold voltage expression must be modified to account for this reduction: VTO(Short Channel) = VTO - VTO where VT0 is the zero-bias threshold voltage calculated using the conventional long-channel formula and VTO is the threshold voltage shift (reduction) due to the short-channel effect. The reduction term actually represents the amount of charge differential between a rectangular depletion region and a trapezoidal depletion region.
13
Let LS and LD represent the lateral extent of the depletion regions associated with the source junction and the drain junction, respectively. Then, the bulk depletion region charge contained within the trapezoidal region To calculate LS and LD, we will use the simplified geometry shown in the figure.
Here, XdS and XdD represent the depth of the pn-junction depletion regions associated with the source and the drain, respectively. The edges of the source and drain diffusion regions are represented by quarter-circular arcs, each with a radius equal to the junction depth, xj. The vertical extent of the bulk depletion region into the substrate is represented by xdm. The junction depletion region depths can be approximated by XdD = and XdS =
14
From figure, we find the following relationship between DLD and the depletion region depths.
Now, the amount of the threshold voltage reduction DVT0 due to shortchannel effects can be found as: . . .[(
-1) + -1]
The threshold voltage shift term is proportional to xj/L. As a result, this term becomes more prominent for MOS transistors with shorter channel lengths, and it approaches zero for long channel MOSFETs where L >> xj.
15
The depletion region charge density at VSB=0 is found as follows: ln = 0.026.ln{(1.45*1010)/(2*1015)} = -0.3077 =
The oxide-interface charge is: Qox= q Nox= 1.6*10-19*4*1010 = 6.4 10-9 C/cm2 The gate oxide capacitance per unit area is calculated using the dielectric constant of the silicon dioxide and the oxide thickness tox: F/ we can combine all components and calculate the threshold voltage
We find the long-channel zero-bias threshold voltage for the process described above as + = 0.40 +{(1.6*1010)* 1011} = 0.855 V
Next, the amount of the threshold voltage reduction due to shortchannel effects must be calculated. The source and drain junction built-in voltage is 0.026 ln = 0.76 v
The depths of source and drain junction depletion regions is found as XdS = = 0 .026.ln{(2*1020)/(1016)}/(1.45*1010)2 =0.9147
17
XdS = Xds = =
= 0.314 mm =
.[(
-1) +
-1]
=1/(7.03*10-8)* (2(1.6*10-19)*11.7*8.85*10-14*2*10-15*2*0.3077)*1/10-6[1+(2*8.03*10-5 (0+VDS)-1] + {1+(2*0.76*10-6)/106 } = 0.2873+0.5875+1+16.06*10-9*(VDS+0.91)-1 = ( 0.343/ L[m] ) * (-0.724 + ( 1 + 2 xdD) The threshold voltage of this short-channel MOS transistor is calculated as VT0 = 0.855V - VTO
18
Chapter 4 The Mat lab simulating program, Results, Command: 4.1 Mat lab Program for verification of threshold voltage with Channel Length
close all, clear all; % arrays for the different values of Vth @ Vds=1V,Vds=3V, Vds=5V. vector_vt1=[ ]; vector_vt3=[ ]; vector_vt5=[ ]; % arrays for the different values of L @ Vds=1V,Vds=3V, Vds=5V. vector_l1=[ ]; vector_l3=[ ]; vector_l5=[ ]; % for Vds=1 V vds=1; for l=0.5:0.1:6 rad2=sqrt(0.13*(0.76+vds)); rad1=sqrt(1+(2*rad2)); deltavt=(0.343/l)*(rad1-0.724); vt=0.855-deltavt; vector_l1=[vector_l1,l]; vector_vt1=[vector_vt1,vt]; end %for Vds=3 V vds=3; for l=0.5:0.1:6 rad2=sqrt(0.13*(0.76+vds)); rad1=sqrt(1+(2*rad2));
19
deltavt=(0.343/l)*(rad1-0.724); vt=0.855-deltavt; vector_l3=[vector_l3,l]; vector_vt3=[vector_vt3,vt]; end %for Vds=5 V vds=5; for l=0.5:0.1:6 rad2=sqrt(0.13*(0.76+vds)); rad1=sqrt(1+(2*rad2)); deltavt=(0.343/l)*(rad1-0.724); vt=0.855-deltavt; vector_l5=[vector_l5,l]; vector_vt5=[vector_vt5,vt]; end % graps of Vth vs. L @ Vds=1V,Vds=3V, Vds=5V. plot(vector_l1,vector_vt1,':',vector_l3,vector_vt3,'.',vector_l5,vector_vt5,'.'), %comments on the plot xlabel('L: Channel length [um]'), ylabel('Vth: Threshold voltage [V]'), title('(Vth vs. L) @ Vds=1V [-----] Vds=3V [_._._] Vds=5V [.....]');
20
21
Data sheet forVth =1 v Channel length (m) 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.7 1.8 1.9 2.0 2.1 2.2 2.6 2.8 3.0 3.3 3.6 3.9 4.2 4.6 5.0 5.4 5.7 6.0
For Vds =1v 0.2181 0.3242 0.400 0.4589 0.5011 0.5365 0.5655 0.5896 0.61 0.6275 0.6427 0.6677 0.6781 0.6874 0.6958 0.7033 0.7102 0.7325 0.7413 0.7488 0.7585 0.7665 0.7733 0.7792 0.7858 0.7913 0.796 0.7991 0.80
For Vds =3v 0.2893 0.3836 0.4509 0.5014 0.5407 0.5721 0.5979 0.6193 0.6374 0.653 0.6664 0.6886 0.6979 0.7061 0.7136 0.7203 0.7264 0.7462 0.754 0.7607 0.7693 0.7764 0.7825 0.7877 0.7935 0.7984 0.8026 0.8054 0.8079
For Vds =5v 0.3921 0.4692 0.5243 0.5657 0.5978 0.6235 0.6446 0.6621 0.677 0.6897 0.7007 0.17188 0.7264 0.7332 0.7393 0.7448 0.7498 0.766 0.7723 0.7778 0.7849 0.7907 0.7957 0.7999 0.8047 0.8087 0.8121 0.8144 0.8164
22
4.2 Mat-lab Program for verification of threshold voltage with Channel Length
close all, clear all; % arrays for the different values of Vth @ Vds=1V,Vds=3V, Vds=5V. vector_vt1=[ ]; vector_vt3=[ ]; vector_vt5=[ ]; % arrays for the different values of L @ Vds=1V,Vds=3V, Vds=5V. vector_l1=[ ]; vector_l3=[ ]; vector_l5=[ ]; % for Vds=1 V vds=1; for l=0.5:0.1:6 rad2=sqrt(0.65*(0.91+vds)); rad1=sqrt(1+(2*rad2)); deltavt=(0.1437/l)*(rad1-0.4125); vt=0.2235-deltavt; vector_l1=[vector_l1,l]; vector_vt1=[vector_vt1,vt]; end %for Vds=3 V vds=3; for l=0.5:0.1:6 rad2=sqrt(0.65*(0.91+vds)); rad1=sqrt(1+(2*rad2)); deltavt=(0.1437/l)*(rad1-0.4125); vt=0.2235-deltavt; vector_l3=[vector_l3,l]; vector_vt3=[vector_vt3,vt]; end
23
%for Vds=5 V vds=5; for l=0.5:0.1:6 rad2=sqrt(0.65*(0.91+vds)); rad1=sqrt(1+(2*rad2)); deltavt=(0.1437/l)*(rad1-0.4125); vt=0.2235-deltavt; vector_l5=[vector_l5,l]; vector_vt5=[vector_vt5,vt]; end % graps of Vth vs. L @ Vds=1V,Vds=3V, Vds=5V. plot(vector_l1,vector_vt1,':',vector_l3,vector_vt3,'.',vector_l5,vector_vt5,'.'), grid %comments on the plot xlabel('L: Channel length [um]'), ylabel('Vth: Threshold voltage [V]'), title('(Vth vs. L) @ Vds=1V [-----] Vds=3V [_._._] Vds=5V [.....]');
24
Fig : 6 Variation of threshold voltage with Channel Length Appling drain voltage applied to the device ,the barrier height is lowered even more, resulting further decrease of threshold voltage. It explains increasing of substrate current with drain voltage in short channel MOSFET. It explain substrate characteristics of long and short channel devices at different drain bias voltage. For long channel devices , the substrate current is independent of drain voltage .For short channel devices however there is a potential shift of the curve to a lower threshold voltage for high drain bias condition. At even shorter channel length the substrate slope starts to degrade as the surface potential is more controlled by the gate. Eventually the device reaches the punchthrough condition when the gate totally loses control of the channel and high drain current persists independent of gate voltage.
25
Data sheet for Channel length (m) 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.7 1.8 1.9 2.0 2.1 2.2 2.6 2.8 3.0 3.3 3.6 3.9 4.2 4.6 5.0 5.4 5.7 6.0
Vth =0.2 v
Threshold voltage For Vds =1v For Vds =3v For Vds =5v -0.2954 -0.2089 -0.1472 -0.1008 -0.06479 -0.03596 -0.01238 0.00728 0.02391 0.03817 0.05052 0.07087 0.07935 0.08694 0.09377 0.09995 0.1056 0.1237 0.1308 0.137 0.1449 0.1514 0.157 0.1617 0.1671 0.1716 0.1755 0.178 0.1803 -0.2461 -0.1679 -0.1119 -0.07002 -0.03741 -0.01131 0.02782 0.04287 0.05578 0.06696 0.07674 0.08537 0.09305 0.09991 0.1061 0.1117 0.1168 0.1332 0.1396 0.1452 0.1523 0.1583 0.1633 0.1676 0.1725 0.1765 0.18 0.1823 0.1844 -0.17430 -0.108 -0.06067 -0.02515 0.002475 0.02458 0.04266 0.05773 0.07048 0.08141 0.09089 0.09917 0.1065 0.113 0.118 0.124 0.1288 0.1331 0.147 0.1525 0.1572 0.1632 0.1682 0.1725 0.1761 0.1803 0.1837 0.1867 0.1903
26
4.3 Mat lab Program for variation of threshold voltage reduction with Channel Length
%variation of threshold volt with channel length clc; clear all; close all; L=0:0.1:3; eox=3.9*8.85*10^-14;% dielectric constant of sio2 esi=11.7*8.85*10^-14; tox=450*10^-8; cox=eox/tox; e=1.6*10^-19;% electronic charge vt=0.0259; % kT/q Na1=3*10^16; Na2=10^15; Na3=10^16; ni=1.5*10^10; rj=0.5*10^-6;% diffusion junction depth fp1=vt*log(Na1/ni);% Fermi potential fp2=vt*log(Na2/ni); fp3=vt*log(Na3/ni); xdt1=sqrt((4*esi*fp1)/(e*Na1));% Junction depth xdt2=sqrt((4*esi*fp2)/(e*Na2)); xdt3=sqrt((4*esi*fp3)/(e*Na3)); dvt1=-(e*Na1*xdt1/cox)*(rj./L)*(sqrt(1+(2*xdt1/rj))-1);% voltage shift dvt2=-(e*Na2*xdt2/cox)*(rj./L)*(sqrt(1+(2*xdt2/rj))-1); dvt3=-(e*Na3*xdt3/cox)*(rj./L)*(sqrt(1+(2*xdt3/rj))-1); subplot(311) plot(L,dvt1); title('L vs. dvt1...') grid; subplot(312) plot(L,dvt2); title('L vs. dvt2...') grid;
27
threshold
ylabel('Thresold volt reduction --->'); subplot(313) plot(L,dvt3); title('L vs. dvt3...') grid; %title('Plotting of channel length vs. reduction of thresold volt...') xlabel('channel length (in micron)--------->'); %variation of thresold volt with channel length
Data sheet For Channel length (in micron) Vs Threshold voltage reduction: Chanel length 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 2 2.2 2.4 2.6 2.8 3.0 For 1 For 2 For 3
-4.252*10-5 -2.126*10-5 -1.417*10-5 -1.063*10-5 -8.503*10-6 -7.086*10-6 -6.074*10-6 -5.314*10-6 -4.724*10-6 -4.252*10-6 -3.865*10-6 -3.543*10-6 -3.271*10-6 -3.037*10-6 -2.834*10-6 -2.657*10-6 -2.501*10-6 -2.126*10-6 -1.933*10-6 -1.771*10-6 -1.635*10-6 -1.518*10-6 -1.417*10-6
-1.585*10-5 -7.924*10-6 -5.282*10-6 -3.962*10-6 -3.169*10-6 -2.641*10-6 -2.264*10-6 -1.981*10-6 -1.761*10-6 -1.585*10-6 -1.441*10-6 -1.321*10-6 -1.291*10-6 -1.132*10-6 -1.056*10-6 -9.904*10-7 -9.322*10-7 -8.341*10-7 -7.546*10-7 -6.89*10-7 -6.095*10-7 -5.66*10-7 -5.282*10-7
-3.172*10-5 -1.563*10-6 -1.042*10-6 -7.817*10-6 -6.253*10-6 -5.211*10-6 -4.467*10-6 -3.908*10-6 -3.474*10-6 -3.127*10-6 -2.842*10-6 -2.606*10-6 -2.233*10-6 -2.084*10-6 -1.954*10-6 -1.839*10-6 -1.737*10-6 -1.646*10-6 -1.563*10-6 -1.421*10-6 -1.203*10-6 -1.117*10-6 -1.042*10-6
29
It is seen that, for short channels and large Vsb , the dependence of the effective threshold on Vsb diminishes. This correspondences to the fact that the lower base of the trapezoidal diminishes in length. The bottom of the trapezoidal is then practically cutoff from the rest of the substrate , and thus the control of the substrate on the charge inside the trapezoidal is small. Decreasing L tends to increases the DIBL, decreasing tox tends to decrease it. This is because then the gate is closer to the channel and is thus better able to keep control of the depletion region charge.
30
CHAPTER 5
5.1Conclusions
The analytical threshold voltage model, presented in this project, is based on a model for depletion layer depth along the channel that takes into account its variations due to the source and drain junctions. By doing so, the sharing of the gate control on VT by the source/drain has properly been incorporated. In fact, the origin of threshold voltage reduction with reduced channel length, reduced substrate doping, and/or increased drain bias is primarily due to the non-negligible influence of the same on the overall channel depletion layer depth under the gate; particularly for short-channel MOSFETs. From the comparisons of the results with MATLAB, it can be concluded that the proposed analytical approach has the accuracy of near 2D numerical results.
5.2 RESEARCH AND FUTURE WORKSIn future another programs can be done in MATLAB or spice simulation can be done to get the Spice Simulations for ID VGS.In near future many research works regarding these topics can be done.
31
APPENDIX-A
List of Abbreviation:
2-D : Two Dimensional DIBL : Drain Induced Barrier Lowering VLSI : Very Large Scale Integrated Circuit CMOS : Complementary Metal Oxide Semiconductor MOSFET : Metal Oxide Semiconductor Field Effect Transistor
32
5.4 References
AJAY KUMAR SINGH An Analytical Model of Short Channel Effects in Sub-Micron MOS Devices Ritesh Gupta, Mridula Gupta, R.S. Gupta Generalized guide for MOSFET miniaturization . K.N Ratnakumar and J.D MeindlShort channel MOSFET threshold voltage model. J.D.Marshall Performance limits of silicon enhancement/depletion MOSFET integrated circuit, Stanford University, CA,Tech pp 46-47. Xing Zhou, Khee Yong Lim - A general approach to compact threshold Voltage formulation based on 2-D numerical simulation and experimental correlation for development Vivek K. De and James D. Meindl An analytical Threshold voltage and sub-threshold current model for short channel MOSFET . Kai Chen and Chenming Hu - Performance and Vdd scaling in deep sub micrometer CMOS Resve Saleh, Michael Benoit and Pete McCrorie - Power distribution planning Sung-Mo, Yusuf CMOS digital integrated circuits deep.sub-micron VLSI technology
33
34