Research Article Potential and Quantum Threshold Voltage...
Transcript of Research Article Potential and Quantum Threshold Voltage...
Hindawi Publishing CorporationActive and Passive Electronic ComponentsVolume 2013 Article ID 153157 9 pageshttpdxdoiorg1011552013153157
Research ArticlePotential and Quantum Threshold Voltage Modeling ofGate-All-Around Nanowire MOSFETs
M Karthigai Pandian1 N B Balamurugan2 and A Pricilla3
1 Pandian Saraswathi Yadav Engineering College Sivagangai India2Thiagarajar Engineering College Madurai India3 St Michaelrsquos College of Engineering and Technology Sivagangai India
Correspondence should be addressed to M Karthigai Pandian karthickpandiangmailcom
Received 21 March 2013 Revised 15 August 2013 Accepted 15 August 2013
Academic Editor Gerard Ghibaudo
Copyright copy 2013 M Karthigai Pandian et al This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited
An improved physics-based compact model for a symmetrically biased gate-all-around (GAA) silicon nanowire transistor isproposed Short channel effects and quantummechanical effects caused by the ultrathin silicon devices are considered inmodellingthe threshold voltage Device geometrics play a very important role in multigate devices and hence their impact on the thresholdvoltage is also analyzed by varying the height and width of silicon channelThe inversion charge and electrical potential distributionalong the channel are expressed in their closed forms The proposed model shows excellent accuracy with TCAD simulations ofthe device in the weak inversion regime
1 Introduction
Semiconductor nanowires are attractive components forfuture nanoelectronics since they can exhibit a wide rangeof device function and at the same time serve as bridgingwires that connect larger scale metallization The nanoscaleFETs based on silicon nanowires have notable attentionfor their potential applications in electronics industry In acontinuous effort to increase current drive and better controlover SCEs silicon-on-insulator (SOI) MOS transistors haveevolved from classical planar and single-gate devices into3D devices with a multigate structure (double- triple- orgate-all-around devices)Thesemultigate nanowire FETs thatprevent the electric field lines from originating at the drainfrom terminating under the channel region are now widelyrecognized as one of the most auspicious solutions formeeting the roadmap requirements in the decananometerscale Multigate device structures of nanowire transistorspave the way for better electrostatic control and as a resultintrinsic channels get higher mobility and current [1]
CMOS devices can be scaled down up to a channel lengthof 10 nm when the number of gates in the device is increasedIn such transistors the short channel effects are controlled
by the device geometry and hence an undoped or lightlydoped ultrathin body is used to sustain the channel Variousdevice structures such as double gate fully depleted SOItrigate and all around gate structures have been extensivelyinvestigated to restrict SCEswithin a limitwhile achieving theprimary advantages of scaling that is higher performancelower power and ever increasing integration density [2] Thescaling theory and the analytical SCEs model for nanowiretransistors based on the concept of natural length are success-ful to a certain extent To address the issue of 2D effects in thegate insulator a more generalized concept of scale length hasbeen proposed recently [3 4]
Modeling of quantum confinement and transport in ananowire transistor has been addressed in the literature [5ndash7] The undoped cylindrical body GAA field effect tran-sistor that has a great control over the corner effects andchannel is considered to be promising candidate for sub-45 nm regime [8 9] An analytical threshold voltage modelfor GAA nanoscale MOSFETs (Figure 1) considering thehot carrier induced interface charges has been proposed byGhoggali et al in 2008 [10] Quantum confinement and itseffect on threshold voltage variations in short channel GAAdevices have been studied in 2009 [11] A compact analytical
2 Active and Passive Electronic Components
W
H
Vg
Vds
Source Drain
TOX
Figure 1 Schematic diagram of a gate-all-around SiNWMOSFET
threshold voltage model proposed by Te-Kuang deals withthe interface trapped charges in a nanowire channel [12]A physically based classical model for body potential of acylindrical GAA nanowire transistor has been proposed byRay andMahapatra in 2008 [13] and a quasianalytical modelfor predicting the potential of a nanowire FET has beenproposed by De Michielis et al in 2010 [14]
In this paper quantum threshold voltage modeling of alightly doped gate-all-around silicon nanowire transistor isproposed In modeling the threshold voltage the quantumeffects are also taken into account as the quantization ofelectron energy in ultrathin devices can never be ignoredOne important consequence of the quantum mechanicalcarrier distribution in accordance with the device behavioroccurs when the device geometrics and the silicon thicknessare varied so a reliable compact model for the nanowiretransistors must also take into account quantum effectsresulting out of these variations The proposed physicallybased closed form quantum threshold voltage model holdsgood for ultrathin and ultrashort channel gate all arounddevices and does not discuss any unphysical fitting parameterThe compact threshold voltage model is obtained by solvingthe 3DPoisson equation and 2D Schrodinger equations in theweak inversion region These equations are then consistentlysolved to obtain the potential distribution and inversioncharge density
2 Threshold Voltage Modeling
Here we consider a lightly doped nanowire MOSFET in theweak inversion region where both fixed and mobile chargedensities in the channel are negligible We have assumeda flat potential on the plane perpendicular to the source-drain direction Poisson-Schrodinger equations should besolved consistently to obtain the potential and inversioncharge density But in the weak inversion regime we haveapproximated the Poisson equation as Laplace equation withthe inversion charge density neglected and thus the twoequations are decoupledThemidgapmetals are used for gateintended to suppress the silicon gate poly depletion inducedparasitic capacitances [15]The 3D Poisson equation is solvedto obtain the threshold voltage in the weak inversion regionincluding the parabolic band approximation The potentialdistribution in insulator and silicon regions can be expressedas
1205752120595 (119909 119910 119911)
1205751199092+1205752120595 (119909 119910 119911)
1205751199102+1205752120595 (119909 119910 119911)
1205751199112= 0 (1)
The potential 120595 in terms of 119909 (width) 119910 (height) and119911 (length) is to be determined The boundary conditionsdefined by the physics of the device are given by
120595(119909minus119867
2minus 119879OX 119911) = 119881119892
120595 (minus119882
2minus 119879OX 119910 119911) = 119881119892
1015840
120575120595
120575119909
10038161003816100381610038161003816100381610038161003816119909=0= 0
120575120595
120575119910
10038161003816100381610038161003816100381610038161003816119910=0
= 0
120595 (119909 119910 0) = 120595bi
120595 (119909 119910 119871) = 120595bi + 119881ds
(2)
For a gate-all-around device we have to find the insulatorpotential on all sides of the channel under considerationSo the height and width of the channel are also taken intoaccount The insulator potential is now expressed as
120595 (119909 119910 119911) =119881119892minus 120595bi minus Φms
119879OX(119909 minus
119882
2) + 120595bi
for 1198822lt 119909 lt
119882
2+ 119879OX 0 lt 119910 lt
119867
2
120595 (119909 119910 119911) =119881119892minus 120595bi minus Φms
119879OX(119910 minus
119867
2) + 120595bi
for 0 lt 119909 lt 119882
2
119867
2lt 119910 lt
119867
2+ 119879OX
(3)
where 119881119892is the gate voltage 120595bi is the built-in potential
119871 is the channel length and 119881ds is the drain to sourcevoltage which is negligible for low 119881ds Φms is the workfunction difference By applying the superposition principlethe electrostatic potential can be now written as
120595 (119909 119910 119911) = 119880119871(119909 119910 119911) + 119880
119877(119909 119910 119911) + 119881
119892(119909 119910) (4)
Here 119881119892(119909 119910) is the 1D solution of the Poisson equation that
satisfies the gate boundary conditions119880119871satisfies the source
boundary condition but it is bound to have a null value onthe gate and drain boundaries Similarly119880
119877satisfies the drain
boundary condition and it is bound to have a null valueon the gate and source boundaries On further evaluation
Active and Passive Electronic Components 3
the term 119881119892+ 119880119871is found to satisfy the potential equation
when119880119877is on null value and in an exact repetition the term
119881119892+ 119880119877satisfies the potential equation when 119880
119871is on null
value From (1)
120595119909119909+ 120595119910119910+ 120595119911119911= 0 (5)
By solving the above equation using LDE method we obtainthe value of 120595 Then the limits are applied on the equationusing the boundary conditions Now the potentials 119880
119871is
given by
1198801198711= sum
119899
sum
119898
119862 times 119870119860119899119898
sin(Λ119899(119909 minus
119882
2minus 119879OX))
times cos (119872119898119910) sinh(sum
119899119898
(119871 minus 119911))
for 1198822lt 119909 lt
119882
2+ 119879OX 0 lt 119910 lt
119867
2
(6)
1198801198712= sum
119899
sum
119898
119862 times 119870119861119899119898
cos (Λ119899119909)
times sin(119872119898(119910 minus
119867
2minus 119879OX))
times sinh(sum119899119898
(119871 minus 119911)) for 0 lt 119909 lt 119882
2
119867
2lt 119910 lt
119867
2+ 119879OX
(7)
1198801198713= sum
119899
sum
119898
119862 times 119870119862119899119898
cos (Λ119899119909) cos (119872
119898119910)
times sinh(sum119899119898
(119871 minus 119911))
for 0 lt 119909 lt 119882
2 0 lt 119910 lt
119867
2
(8)
Similarly the values of potential 119880119877are also derived as
follows
1198801198771= sum
119899
sum
119898
119862 times 119870119875119899119898
sin(Λ119899(119909 minus
119882
2minus 119879OX))
times cos (119872119898119910)
times sinh(sum119899119898
119911) for 1198822lt 119909 lt
119882
2+ 119879OX
0 lt 119910 lt119867
2
1198801198772= sum
119899
sum
119898
119862 times 119870119876119899119898
cos (Λ119899119909)
times sin(119872119898(119910 minus
119867
2minus 119879OX)) sinh(sum
119899119898
119911)
for 0 lt 119909 lt 119882
2
119867
2lt 119910 lt
119867
2+ 119879OX
1198801198773= sum
119899
sum
119898
119862 times 119870119877119899119898
cos (Λ119899119909) cos (119872
119898119910)
times sinh(sum119899119898
119911) for 0 lt 119909 lt 119882
2
0 lt 119910 lt119867
2
(9)
where 119888 = 1sum119899119898119871 and 119870
119860119899119898 119870119861119899119898
119870119862119899119898
119870119875119899119898
119870119902119899119898
and 119870
119877119899119898are constants From (6) and (8) 119880
119871and (120597119880
119871)120597119909
are found to be continuous in the 119909 direction (119909 = 1198822)The first derivative function (120597119880
1198711)120597119909 itself has 120576OX120576si
times discontinuities at the silicon insulator interfaces Thusapplying continuity in both equations we proceed to equate(6) and (8) as follows
minus119870119860119899119898
sin (Λ119899119879OX) = 119870119862119899119898 cos(
Λ119899119882
2) (10)
Differentiating (6) with respect to 119909
1205971198801198711
120597119909= sum
119899
sum
119898
119862 times 119870119860119899119898
(Λ119899) cos(Λ
119899(119909 minus
119882
2minus 119879OX))
times cos (119872119898119910) sinh(sum
119899119898
(119871 minus 119911))
(11)
Differentiating (8) with respect to 119909
1205971198801198713
120597119909= sum
119899
sum
119898
119862 times 119870119888119899119898
(Λ119899) (minus sin (Λ
119899119909))
times cos (119872119898119910) times sinh(sum
119899119898
(119871 minus 119911))
(12)
Equating (11) and (12)
119870119860119899119898
120576OX cos (Λ 119899119879OX) = minus119870119862119899119898120576si sin(Λ119899
2) (13)
Dividing (10) by (13)
120576si tan (Λ 119899119879OX) minus 120576OXcot(Λ119899119882
2) = 0 (14)
Likewise from (7) and (8) 119880119871and (120597119880
119871120597119910) are found to
be continuous in the 119910 direction (119910 = 1198672) The function(1205971198801198712120597119910) itself has discontinuities at the silicon insulator
interfaces which are proportional to the dielectric constant120576OX120576si Thus applying continuity in both equations andequating (7) and (8) we get
minus119870119861119899119898
sin (119872119898119879OX) = 119870119862119899119898 cos(
119872119898119867
2) (15)
4 Active and Passive Electronic Components
Differentiating (7) with respect to 119910
1205971198801198712
120597119910= sum
119899
sum
119898
119862 times 119870119861119899119898
cos (Λ119899119909) (119872
119898)
times cos(119872119898(119910 minus
119867
2minus 119879OX))
times sinh(sum119899119898
(119871 minus 119911))
(16)
Differentiating (8) with respect to 119910
1205971198801198713
120597119910= sum
119899
sum
119898
119862 times 119870119862119899119898
cos (Λ119899119909) (119872
119898)
times (minus sin (119872119898119910)) sinh(sum
119899119898
(119871 minus 119911))
(17)
Equating (16) and (17)
119870119861119899119898
120576OX cos (119872119898119879OX) = minus119870119862119899119898120576si sin(119872119898119867
2) (18)
Dividing (15) by (18)
120576si tan (119872119898119879OX) minus 120576OXcot(119872119898119867
2) = 0 (19)
This natural length is an easy guide for choosing deviceparameters and has simple physical meaning that a smallnatural length corresponds to superb short channel effectimmunity [4] The value of Λ
119899and 119872
119898depends on device
parameters The potential 119880119871can be modified as
1198801198711198991198981
= 120572119899119898
times 119862 times sin(Λ119899(119909 minus
119882
2minus 119879OX))
times cos (119872119898119910) times sinh(sum
119899119898
(119871 minus 119911))
for 1198822lt 119909 lt
119882
2+ 119879OX 0 lt 119910 lt
119867
2
1198801198711198991198982
= 120573119899119898
times 119862 times cos (Λ119899119909)
times sin(119872119898(119910 minus
119867
2minus 119879OX)) sinh(sum
119899119898
(119871 minus 119911))
for 0 lt 119909 lt 119882
2
119867
2lt 119910 lt
119867
2+ 119879OX
1198801198711198991198983
= 120574119899119898
times 119862 times cos (Λ119899119909) cos (119872
119898119910)
times sinhsum119899119898
(119871 minus 119911) for 0 lt 119909 lt 119882
2 0 lt 119910 lt
119867
2
(20)
Similarly the potential 119880119877can be modified as
1198801198771198991198981
= 120572119899119898
times 119862 times sin(Λ119899(119909 minus
119882
2minus 119879OX))
times cos (119872119898119910)
times sinh(sum119899119898
119911) for 1198822lt 119909 lt
119882
2+ 119879OX
0 lt 119910 lt119867
2
1198801198771198991198982
= 120573119899119898
times 119862 times cos (Λ119899119909)
times sin(119872119898(119910 minus
119867
2minus 119879OX)) times sinh(sum
119899119898
119911)
for 0 lt 119909 lt 119882
2
119867
2lt 119910 lt
119867
2+ 119879OX
1198801198771198991198983
= 120574119899119898
times cos (Λ119899119909) cos (119872
119898119910)
times sinh(sum119899119898
119911) for 0 lt 119909 lt 119882
2 0 lt 119910 lt
119867
2
(21)
Now the 119892119899119898
can be obtained from the potential equations(20) by using different multipliers in different regions
1198921198991198981
=120576OX sin (Λ 119899 (119909 minus1198822 minus 119879OX)) cos (119872119898119910)
2120576si sin (Λ 119899119879OX) cos (119872119898 (1198672))
for 1198822lt 119909 lt
119882
2+ 119879OX 0 lt 119910 lt
119867
2
1198921198991198982
=120576OX cos (Λ 119899119909) sin (119872119898 (119910 minus 1198672 minus 119879OX))
2120576si cos (Λ 119899 (1198822)) sin (119872119898119879OX)
for 0 lt 119909 lt 119882
2
119867
2lt 119910 lt
119867
2+ 119879OX
1198921198991198983
= minuscos (Λ
119899119909) cos (119872
119898119910)
cos (Λ119899(1198822)) cos (119872
119898(1198672))
for 0 lt 119909 lt 119882
2 0 lt 119910 lt
119882
2
(22)
Subsequently the constants 120572119899119898 120573119899119898 and 120574
119899119898are evaluated
suitably
120572119899119898
= cos(Λ119899
119882
2) sin (119872
119898119879OX)
120573119899119898
= cos(119872119898
119867
2) sin (Λ
119899119879OX)
120574119899119898
= sin (119872119898119879OX) sin (Λ 119899119879OX)
(23)
From (20) the potential 119880119871119899119898
can be rewritten as
1198801198711198991198981
= minus cos(Λ119899
119882
2) sin (119872
119898119879OX)
times sin (Λ119899119879OX) cos(119872119898
119867
2)
Active and Passive Electronic Components 5
1198801198711198991198982
= minus cos(119872119898
119867
2) sin (Λ
119899119879OX)
times cos(Λ119899
119882
2) sin (119872
119898119879OX)
1198801198711198991198983
= minus sin (119872119898119879OX) sin (Λ 119899119879OX)
times cos(Λ119899
119882
2) cos(119872
119898
119867
2)
(24)
By multiplying with the corresponding orthogonal conjugatefunctions and integrating coefficients of119880
119871can be obtained
The coefficients of 119880119877are also obtained in a similar method
119894119899119898
= (int
119867+119879OX
0
int
119882+119879OX
0
(120595 (119909 119910 0))
minus119881 (119909 119910) 119892119899119898(119909 119910) 119889119909 119889119910)
times (int
119867+119879OX
0
int
119882+119879OX
0
119880119871119899119898
(119909 119910 0) 119892119899119898(119909 119910) 119889119909 119889119910)
minus1
119895119899119898
= (int
119867+119879OX
0
int
119882+119879OX
0
(120595 (119909 119910 119871))
minus119881 (119909 119910) 119892119899119898(119909 119910) 119889119909 119889119910)
times (int
119867+119879OX
0
int
119882+119879OX
0
119880119871119899119898
(119909 119910 119871) 119892119899119898(119909 119910) 119889119909 119889119910)
minus1
(25)
The above integrals (25) are evaluated to obtain explicitexpressions for 119894
119899119898and 119895119899119898
as follows
119894119899119898
= (119881119892minus 120595bi) 120596119899119898
119895119899119898
= (119881119892minus 120595bi minus 119881ds) 120596119899119898
120596119899119898
=Ω119899119898
120585119899119898
(26)
where
Ω119899119898
=120576OXΛ 119899 tan ((Λ 119899119882) 2) + 120576OX119872119898 tan (1198721198981198672)
2120576si119879OXΛ2
1198991198722119898
minus2
Λ119899119872119898
tan(Λ119899119882
2) tan(
119872119898119867
2)
120585119899119898
=sin (Λ
119899119879OX)
16 cos (Λ1198991198822)
(119882 +sin (Λ
119899119882)
Λ119899
) 1199031
+sin (119872
119898119879OX)
16 cos (1198721198981198672)
(119867 +sin (119872
119898119867)
119872119898
) 1199032
(27)
And the values of 1199031and 1199032are given by
1199031=120576OX119879OX cos (1198721198981198672)
120576si sin (119872119898119879OX)+119867 sin (119872
119898119879OX)
2 cos (1198721198981198672)
1199032=120576OX119879OX cos (Λ 1198991198822)
120576si sin (Λ 119899119879OX)+119882 sin (Λ
119899119879OX)
2 cos (Λ1198991198822)
(28)
The potential equation is now rewritten as
120595 (119909 119910 119911) = 119881119892+ (119881119892minus 120595bi)
timessum
119899
sum
119898
120588119899119898
sinh(sum119899119898
(119871 minus 119911))
+ (119881119892minus 120595bi minus 119881ds)sum
119899
sum
119898
sinh(sum119899119898
119911)
(29)
where
120588119899119898(119909 119910) = minus120596
119899119898120574119899119898
cos (Λ119899119909) cos (119872
119898119910)
sinh (sum119899119898119871)
(30)
Once the potential distribution at every point of the cross-section of the channel is known we calculate the inversioncharge density by using surface integral over the surface areaof the channel When the integrated charge at virtual sourcebecomes equal to critical charge the gate voltage of a lightlydoped body device is nearly equal to the threshold voltage ofthe device Hence the inversion charge can be expressed as
119876 = int
1198672
minus1198672
int
1198822
minus1198822
119902119899119894119890(120595119880119879)119889119909 119889119910 (31)
where 119902 is the elementary charge 119880119879is thermal voltage and
119899119894is the intrinsic carrier concentrationThe charge equation can now be approximated as
119876 asymp 119882119867119902119899119894119890((120595((311988214)(311986714)119911119888))119880119879) (32)
Here 119911119888is the virtual source position which is half of the
channel length for low 119881ds Using the inversion charge wecan obtain the classical threshold model as expressed in thefollowing
119881TC = (119880119879 ln119876
119882119867119902119899119894
+ 2119881bi12058811 (3119882
143119867
14)
times sinh(sum11119871
2))
times (1 + 212058811(3119882
143119867
14) sinh(
sum11119871
2))
minus1
(33)
3 Quantum Threshold Voltage Modeling
AsMOSFETdevices are further scaled into the deep nanome-ter regime it has become necessary to include quantummechanical effects while modeling their device behaviorIn this paper we approximate the actual potential well as
6 Active and Passive Electronic Components
E11
Ec
Ec
998779Ec
Ei119909i119910
Eco
Figure 2 Band diagram perpendicular to the gate-square wellpotential of a GAA silicon nanowire transistor
the square well potential since it is difficult to solve theSchrodinger equation to obtain the potential expressed in(29) The square well potential of a gate-all-around nanowiretransistor is shown in Figure 2 The quantum charge of thedevice is expressed as
119876 = sum
119894119909
sum
119894119910
119902int
infin
119864119894119909119894119910
1198731D119891 (119864) 119889119864 (34)
where1198731D is the 1D density-of-states and 119891(119864) is the Fermi-
Dirac distribution function 119864 is the energy of the electronwave The terms 119894
119909and 119894119910are positive natural numbers
In silicon six energy valleys are found to be present inits band structure (two lower energy valleys two middleenergy valleys and two higher energy valleys) If the thinfilm of device has equal height and width the two lowerenergy valleys and two middle energy valleys are combinedtogether to produce four lower energy valleys and the othertwo higher energy valleys remain in their own stateThus thecharge is given by
119876 = 119902sum
119894119909
sum
119894119910
radic(119898119911
2120587ℎ2)int
infin
119864119894119909119894119910
(119864 minus 119864119894119909119894119910)minus12
1 + 119890 ((119864 minus 119864119865) 119896119879)
119889119864 (35)
where 119898119911is the mass of the valley which is perpendicular to
the direction of quantizationThe Fermi energy level is muchlower than the conduction band energy in weak inversionregion Hence the charge equation can be approximated as
119876 = 119902radic(119898119911
2120587ℎ2)sum
119894119909
sum
119894119910
int
infin
119864119894119909119894119910
119890((119864119865minus119864)119896119879)
(119864 minus 119864119894119909119894119910)12119889119864 (36)
Using the Schrodinger equation the value of 119864119894119909119864119894119910is deter-
mined by the following formulation [5]
119864119894119909119894119910
= 119864co +ℎ21205872
2[1
119898119909
(119894119909
119868119909
)
2
+1
119898119910
(119894119910
119868119910
)
2
] (37)
where the conduction band energy is given as
119864co =119864119892
2minus 119902120595 (0 0 119911
119888) (38)
Using (36) and (37) the integrated charge can be obtained as
119876 = 119902radic(2119896119879119898
119905
ℎ2)sum
119894119909
sum
119894119910
exp(minus119864co + 1198641 (119894119909 119894119910)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2)sum
119894119909
sum
119894119910
exp(minus119864co + 1198642 (119894119909 119894119910)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2)sum
119894119909
sum
119894119910
exp(minus119864co + 1198643 (119894119909 119894119910)
119896119879)
(39)
where
1198641(119894119909 119894119910) =
ℎ21205872
2[1
1198981
(119894119909
119882)
2
+1
119898119905
(119894119910
119867)
2
]
1198642(119894119909 119894119910) =
ℎ21205872
2[1
119898119905
(119894119909
119882)
2
+1
1198981
(119894119910
119867)
2
]
1198643(119894119909 119894119910) =
ℎ21205872
2[1
119898119905
(119894119909
119882)
2
+1
119898119905
(119894119910
119867)
2
]
(40)
Here the 119898119905and 119898
1are the transverse and longitudinal
effective masses of the energy valleys of silicon The lengths119894119909and 119894119910carry distinct values contingent on the direction of
quantization Finally the quantum threshold voltage modelbecomes
119881TQ = (119864119892
2119902+ (
119896119879
119902) In(119876119879
120591)
+2120595bi12058811 (0 0) sinh((sum11119871)
2))
times (1 + 212058811(0 0) sinh(
(sum11119871)
2))
minus1
(41)
where
120591 = 119902radic(2119896119879119898
119905
ℎ2) exp(minus1198641 (1 1)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2) exp(minus1198642 (1 1)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2) exp(minus
1198643(1 1)
119896119879)
(42)
Active and Passive Electronic Components 7
10 20 30 40 50 6002
022
024
026
028
03
032
034
036
038
04
Channel length (nm)
Pote
ntia
l (V
)
Proposed modelTCAD simulation
Figure 3 Constant electrostatic potential obtained from the analyt-ical solution of a gate-all-around silicon nanowire transistor is 03 Vfor different channel length with a height of119867 = 9 nm and channelwidth of 119882 = 9 nm TCAD simulation shows that the potential isconstant at 0296V
The impacts on the threshold voltage due to quantum effectsare acquired by using the following equation
119881TC = 119881TQ + Δ119881119879 (43)
Here Δ119881119879is the difference between the quantum threshold
voltage and the classical threshold voltage
4 Results and Discussion
Figure 3 shows the electrostatic potential of the proposed gateall around transistor and it is found to be constant value at03 V Continuously varying the 119899 and 119898 terms in (4) hasno impact on the potential as it remains constant along theinsulator boundaries This is totally in contrast to the resultsobtained in [15] where the potential is found to be linearlyvarying in the insulator boundaries The constant potentialhas to be deduced as the resultant of the gate voltage appliedsymmetrically across the four sides of the transistor TheTCAD simulation of the device shows that the electrostaticpotential is constant at 0296V The simulation results arefound to in acceptance with the TCAD results
Figure 4 represents the variation of total quantum inte-grated charge with the gate voltage Equation (39) is usedto obtain the integrated charge with only one energy leveland one series term It clearly shows that the decrease inthe film thickness leads to the increase in the quantumthreshold voltage which is actually due to the increase inenergy quantization of the transistor With the height andlength of the device being constant the width of the device isvaried and henceforth the variation of charge in accordancewith the gate voltage is illustrated in Figure 4
The variation of quantum threshold voltage with widthand height of the film at a channel length of 20 nm is
0 01 02 03 04 05 06 070
05
1
15
2
25
3
35
4times10minus15
Gate voltage (V)
Char
ge (c
mminus1)
W = 9e minus 9
W = 5e minus 9
W = 3e minus 9
TCAD simulation for W = 9e minus 9
TCAD simulation for W = 5e minus 9
TCAD simulation for W = 3e minus 9
Figure 4 Variation of quantum integrated charge at virtual sourcewith gate voltage for different film widths where the height andlength are119867 = 9 nm 119871 = 20 nm
2
8
times10minus9
times10minus9
Film height
2
44
66
8
10
1003
0305
031
0315
032
Film width
Qua
ntum
thre
shol
d vo
ltage
(V)
ProposedTCAD
Figure 5 Variation of quantum threshold voltage with film heightand width for channel length (119871 = 20 nm)
shown in Figure 5 The short channel effects tend to decreasealong with the energy quantisation and this can be furtherexplained as a result of increase in the effective band gap ofsilicon due to quantum effects The effect of confinementexpressed as the difference in the threshold voltage and itsvariation with the channel length 119871 is illustrated in Figure 6
8 Active and Passive Electronic Components
2 3 4 5 6 7 8 9 10
times10minus9
0014
0016
0018
002
0022
0024
0026
0028
003
0032
Film height
Proposed modelTCAD simulation
ΔVT
(mV
)
Figure 6 Variation of threshold voltagewith filmwidth for differentheight at 119871 = 20 nm
3 4 5 6 7 8
times10minus9
02
022
024
026
028
03
032
034
036
038
04
Film height
Thre
shol
d vo
ltage
(V)
ClassicalQuantum
TCAD simulation for classicalTCAD simulation for quantum
9
Figure 7 Variation of quantum and classical threshold voltage withfilm height Here 119871 = 20 nm and119882 = 9 nm
Themost important thing about this gate all around nanowiretransistor is that any change in one of the dimensions canbe nullified by proper tuning of other dimensions as thetransistor is symmetric about its height and width
Figure 7 shows the variation of the classical thresholdvoltage and quantum threshold voltage with the film height ata constant width of 9 nmThe value of the classical thresholdvoltage ranges from 027V to 029V for the correspondingchanges in the film height Similarly the quantum thresholdvoltage ranges from 03V to 031 V It shows that the devicehas a highly improved control over the threshold voltageTheTCAD results justify the simulation results
5 Conclusion
In this paper a quantum threshold voltage model for a GAAsilicon nanowire transistor is proposed by solving the 3DPoisson and Schrodinger equations Analytical expressionsfor potential and the inversion charge are expressed intheir closed forms The results show that the integratedcharge and the threshold voltage calculated in accordancewith the quantum effects of this proposed model are highlyimprovedThe future considerations include deriving the I-Vcharacteristics of the gate all around nanowire transistors andstudying the impact of scaling on various device parametersFinally to conclude this model provides an analytical anduseful way for the threshold voltage evaluations in gate allaround nanowire devices with a unified formalism employedin both classical and quantum mechanical approaches
References
[1] J Wang E Polizzi and M Lundstrom ldquoA computationalstudy of ballistic silicon nanowire transistorsrdquo in Proceedings ofthe IEEE International Electron Devices Meeting pp 695ndash698December 2003
[2] J P Colinge FINFETS and Other Multi-Gate TransistorsSpringer New York NY USA 2007
[3] D J Frank Y Taur andH-S PWong ldquoGeneralized scale lengthfor two-dimensional effects in MOSFETrsquosrdquo IEEE ElectronDevice Letters vol 19 no 10 pp 385ndash387 1998
[4] B Yu L Wang Y Yuan P M Asbeck and Y Taur ldquoScaling ofnanowire transistorsrdquo IEEE Transactions on Electron Devicesvol 55 no 11 pp 2846ndash2858 2008
[5] G D Sanders C J Stanton and Y C Chang ldquoTheory of trans-port in silicon quantum wiresrdquo Physical Review B vol 48 no15 pp 11067ndash11076 1993
[6] M-Y Shen and S-L Zhang ldquoBand gap of a silicon quantumwirerdquo Physics Letters A vol 176 no 3-4 pp 254ndash258 1993
[7] J P Colinge X Baie V Bayot and E Grivei ldquoQuantum-wireeffects in thin and narrow SOI MOSFETsrdquo in Proceedings of theIEEE International SOI Conference pp 66ndash67 October 1995
[8] C P Auth and J D Plummer ldquoScaling theory for cylindricalfully-depleted surrounding-gate MOSFETrsquosrdquo IEEE ElectronDevice Letters vol 18 no 2 pp 74ndash76 1997
[9] J-T Park and J-P Colinge ldquoMultiple-gate SOI MOSFETsdevice design guidelinesrdquo IEEE Transactions on ElectronDevices vol 49 no 12 pp 2222ndash2229 2002
[10] Z Ghoggali F DjeffalM A Abdi D Arar N NLakhdar and TTBendib ldquoAn analytical threshold voltagemodel for nanoscalerdquoin Proceedings of the 3rd International Design and TestWorkshop(IDT rsquo08) pp 93ndash97 December 2008
[11] Y-S Wu and P Su ldquoQuantum confinement effect in short-channel gate-all-around MOSFETs and its impact on the sensi-tivity of threshold voltage to process variationsrdquo in Proceedingsof the IEEE International SOI Conference October 2009
[12] C Te-Kuang ldquoA compact analytical threshold-voltage modelfor surrounding-gateMOSFETswith interface trapped chargesrdquoIEEE Electron Device Letters vol 31 no 8 pp 788ndash790 2010
[13] B Ray and SMahapatra ldquoModeling and analysis of body poten-tial of cylindrical gate-all-around nanowire transistorrdquo IEEETransactions on Electron Devices vol 55 no 9 pp 2409ndash24162008
Active and Passive Electronic Components 9
[14] L De Michielis L Selmi and A M Ionescu ldquoA quasi-analytical model for nanowire FETs with arbitrary polygonalcross sectionrdquo Solid-State Electronics vol 54 no 9 pp 929ndash9342010
[15] P R Kumar and S Mahapatra ldquoQuantum threshold voltagemodeling of short channel quad gate silicon nanowire transis-torrdquo IEEE Transactions onNanotechnology vol 10 no 1 pp 121ndash128 2011
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of
2 Active and Passive Electronic Components
W
H
Vg
Vds
Source Drain
TOX
Figure 1 Schematic diagram of a gate-all-around SiNWMOSFET
threshold voltage model proposed by Te-Kuang deals withthe interface trapped charges in a nanowire channel [12]A physically based classical model for body potential of acylindrical GAA nanowire transistor has been proposed byRay andMahapatra in 2008 [13] and a quasianalytical modelfor predicting the potential of a nanowire FET has beenproposed by De Michielis et al in 2010 [14]
In this paper quantum threshold voltage modeling of alightly doped gate-all-around silicon nanowire transistor isproposed In modeling the threshold voltage the quantumeffects are also taken into account as the quantization ofelectron energy in ultrathin devices can never be ignoredOne important consequence of the quantum mechanicalcarrier distribution in accordance with the device behavioroccurs when the device geometrics and the silicon thicknessare varied so a reliable compact model for the nanowiretransistors must also take into account quantum effectsresulting out of these variations The proposed physicallybased closed form quantum threshold voltage model holdsgood for ultrathin and ultrashort channel gate all arounddevices and does not discuss any unphysical fitting parameterThe compact threshold voltage model is obtained by solvingthe 3DPoisson equation and 2D Schrodinger equations in theweak inversion region These equations are then consistentlysolved to obtain the potential distribution and inversioncharge density
2 Threshold Voltage Modeling
Here we consider a lightly doped nanowire MOSFET in theweak inversion region where both fixed and mobile chargedensities in the channel are negligible We have assumeda flat potential on the plane perpendicular to the source-drain direction Poisson-Schrodinger equations should besolved consistently to obtain the potential and inversioncharge density But in the weak inversion regime we haveapproximated the Poisson equation as Laplace equation withthe inversion charge density neglected and thus the twoequations are decoupledThemidgapmetals are used for gateintended to suppress the silicon gate poly depletion inducedparasitic capacitances [15]The 3D Poisson equation is solvedto obtain the threshold voltage in the weak inversion regionincluding the parabolic band approximation The potentialdistribution in insulator and silicon regions can be expressedas
1205752120595 (119909 119910 119911)
1205751199092+1205752120595 (119909 119910 119911)
1205751199102+1205752120595 (119909 119910 119911)
1205751199112= 0 (1)
The potential 120595 in terms of 119909 (width) 119910 (height) and119911 (length) is to be determined The boundary conditionsdefined by the physics of the device are given by
120595(119909minus119867
2minus 119879OX 119911) = 119881119892
120595 (minus119882
2minus 119879OX 119910 119911) = 119881119892
1015840
120575120595
120575119909
10038161003816100381610038161003816100381610038161003816119909=0= 0
120575120595
120575119910
10038161003816100381610038161003816100381610038161003816119910=0
= 0
120595 (119909 119910 0) = 120595bi
120595 (119909 119910 119871) = 120595bi + 119881ds
(2)
For a gate-all-around device we have to find the insulatorpotential on all sides of the channel under considerationSo the height and width of the channel are also taken intoaccount The insulator potential is now expressed as
120595 (119909 119910 119911) =119881119892minus 120595bi minus Φms
119879OX(119909 minus
119882
2) + 120595bi
for 1198822lt 119909 lt
119882
2+ 119879OX 0 lt 119910 lt
119867
2
120595 (119909 119910 119911) =119881119892minus 120595bi minus Φms
119879OX(119910 minus
119867
2) + 120595bi
for 0 lt 119909 lt 119882
2
119867
2lt 119910 lt
119867
2+ 119879OX
(3)
where 119881119892is the gate voltage 120595bi is the built-in potential
119871 is the channel length and 119881ds is the drain to sourcevoltage which is negligible for low 119881ds Φms is the workfunction difference By applying the superposition principlethe electrostatic potential can be now written as
120595 (119909 119910 119911) = 119880119871(119909 119910 119911) + 119880
119877(119909 119910 119911) + 119881
119892(119909 119910) (4)
Here 119881119892(119909 119910) is the 1D solution of the Poisson equation that
satisfies the gate boundary conditions119880119871satisfies the source
boundary condition but it is bound to have a null value onthe gate and drain boundaries Similarly119880
119877satisfies the drain
boundary condition and it is bound to have a null valueon the gate and source boundaries On further evaluation
Active and Passive Electronic Components 3
the term 119881119892+ 119880119871is found to satisfy the potential equation
when119880119877is on null value and in an exact repetition the term
119881119892+ 119880119877satisfies the potential equation when 119880
119871is on null
value From (1)
120595119909119909+ 120595119910119910+ 120595119911119911= 0 (5)
By solving the above equation using LDE method we obtainthe value of 120595 Then the limits are applied on the equationusing the boundary conditions Now the potentials 119880
119871is
given by
1198801198711= sum
119899
sum
119898
119862 times 119870119860119899119898
sin(Λ119899(119909 minus
119882
2minus 119879OX))
times cos (119872119898119910) sinh(sum
119899119898
(119871 minus 119911))
for 1198822lt 119909 lt
119882
2+ 119879OX 0 lt 119910 lt
119867
2
(6)
1198801198712= sum
119899
sum
119898
119862 times 119870119861119899119898
cos (Λ119899119909)
times sin(119872119898(119910 minus
119867
2minus 119879OX))
times sinh(sum119899119898
(119871 minus 119911)) for 0 lt 119909 lt 119882
2
119867
2lt 119910 lt
119867
2+ 119879OX
(7)
1198801198713= sum
119899
sum
119898
119862 times 119870119862119899119898
cos (Λ119899119909) cos (119872
119898119910)
times sinh(sum119899119898
(119871 minus 119911))
for 0 lt 119909 lt 119882
2 0 lt 119910 lt
119867
2
(8)
Similarly the values of potential 119880119877are also derived as
follows
1198801198771= sum
119899
sum
119898
119862 times 119870119875119899119898
sin(Λ119899(119909 minus
119882
2minus 119879OX))
times cos (119872119898119910)
times sinh(sum119899119898
119911) for 1198822lt 119909 lt
119882
2+ 119879OX
0 lt 119910 lt119867
2
1198801198772= sum
119899
sum
119898
119862 times 119870119876119899119898
cos (Λ119899119909)
times sin(119872119898(119910 minus
119867
2minus 119879OX)) sinh(sum
119899119898
119911)
for 0 lt 119909 lt 119882
2
119867
2lt 119910 lt
119867
2+ 119879OX
1198801198773= sum
119899
sum
119898
119862 times 119870119877119899119898
cos (Λ119899119909) cos (119872
119898119910)
times sinh(sum119899119898
119911) for 0 lt 119909 lt 119882
2
0 lt 119910 lt119867
2
(9)
where 119888 = 1sum119899119898119871 and 119870
119860119899119898 119870119861119899119898
119870119862119899119898
119870119875119899119898
119870119902119899119898
and 119870
119877119899119898are constants From (6) and (8) 119880
119871and (120597119880
119871)120597119909
are found to be continuous in the 119909 direction (119909 = 1198822)The first derivative function (120597119880
1198711)120597119909 itself has 120576OX120576si
times discontinuities at the silicon insulator interfaces Thusapplying continuity in both equations we proceed to equate(6) and (8) as follows
minus119870119860119899119898
sin (Λ119899119879OX) = 119870119862119899119898 cos(
Λ119899119882
2) (10)
Differentiating (6) with respect to 119909
1205971198801198711
120597119909= sum
119899
sum
119898
119862 times 119870119860119899119898
(Λ119899) cos(Λ
119899(119909 minus
119882
2minus 119879OX))
times cos (119872119898119910) sinh(sum
119899119898
(119871 minus 119911))
(11)
Differentiating (8) with respect to 119909
1205971198801198713
120597119909= sum
119899
sum
119898
119862 times 119870119888119899119898
(Λ119899) (minus sin (Λ
119899119909))
times cos (119872119898119910) times sinh(sum
119899119898
(119871 minus 119911))
(12)
Equating (11) and (12)
119870119860119899119898
120576OX cos (Λ 119899119879OX) = minus119870119862119899119898120576si sin(Λ119899
2) (13)
Dividing (10) by (13)
120576si tan (Λ 119899119879OX) minus 120576OXcot(Λ119899119882
2) = 0 (14)
Likewise from (7) and (8) 119880119871and (120597119880
119871120597119910) are found to
be continuous in the 119910 direction (119910 = 1198672) The function(1205971198801198712120597119910) itself has discontinuities at the silicon insulator
interfaces which are proportional to the dielectric constant120576OX120576si Thus applying continuity in both equations andequating (7) and (8) we get
minus119870119861119899119898
sin (119872119898119879OX) = 119870119862119899119898 cos(
119872119898119867
2) (15)
4 Active and Passive Electronic Components
Differentiating (7) with respect to 119910
1205971198801198712
120597119910= sum
119899
sum
119898
119862 times 119870119861119899119898
cos (Λ119899119909) (119872
119898)
times cos(119872119898(119910 minus
119867
2minus 119879OX))
times sinh(sum119899119898
(119871 minus 119911))
(16)
Differentiating (8) with respect to 119910
1205971198801198713
120597119910= sum
119899
sum
119898
119862 times 119870119862119899119898
cos (Λ119899119909) (119872
119898)
times (minus sin (119872119898119910)) sinh(sum
119899119898
(119871 minus 119911))
(17)
Equating (16) and (17)
119870119861119899119898
120576OX cos (119872119898119879OX) = minus119870119862119899119898120576si sin(119872119898119867
2) (18)
Dividing (15) by (18)
120576si tan (119872119898119879OX) minus 120576OXcot(119872119898119867
2) = 0 (19)
This natural length is an easy guide for choosing deviceparameters and has simple physical meaning that a smallnatural length corresponds to superb short channel effectimmunity [4] The value of Λ
119899and 119872
119898depends on device
parameters The potential 119880119871can be modified as
1198801198711198991198981
= 120572119899119898
times 119862 times sin(Λ119899(119909 minus
119882
2minus 119879OX))
times cos (119872119898119910) times sinh(sum
119899119898
(119871 minus 119911))
for 1198822lt 119909 lt
119882
2+ 119879OX 0 lt 119910 lt
119867
2
1198801198711198991198982
= 120573119899119898
times 119862 times cos (Λ119899119909)
times sin(119872119898(119910 minus
119867
2minus 119879OX)) sinh(sum
119899119898
(119871 minus 119911))
for 0 lt 119909 lt 119882
2
119867
2lt 119910 lt
119867
2+ 119879OX
1198801198711198991198983
= 120574119899119898
times 119862 times cos (Λ119899119909) cos (119872
119898119910)
times sinhsum119899119898
(119871 minus 119911) for 0 lt 119909 lt 119882
2 0 lt 119910 lt
119867
2
(20)
Similarly the potential 119880119877can be modified as
1198801198771198991198981
= 120572119899119898
times 119862 times sin(Λ119899(119909 minus
119882
2minus 119879OX))
times cos (119872119898119910)
times sinh(sum119899119898
119911) for 1198822lt 119909 lt
119882
2+ 119879OX
0 lt 119910 lt119867
2
1198801198771198991198982
= 120573119899119898
times 119862 times cos (Λ119899119909)
times sin(119872119898(119910 minus
119867
2minus 119879OX)) times sinh(sum
119899119898
119911)
for 0 lt 119909 lt 119882
2
119867
2lt 119910 lt
119867
2+ 119879OX
1198801198771198991198983
= 120574119899119898
times cos (Λ119899119909) cos (119872
119898119910)
times sinh(sum119899119898
119911) for 0 lt 119909 lt 119882
2 0 lt 119910 lt
119867
2
(21)
Now the 119892119899119898
can be obtained from the potential equations(20) by using different multipliers in different regions
1198921198991198981
=120576OX sin (Λ 119899 (119909 minus1198822 minus 119879OX)) cos (119872119898119910)
2120576si sin (Λ 119899119879OX) cos (119872119898 (1198672))
for 1198822lt 119909 lt
119882
2+ 119879OX 0 lt 119910 lt
119867
2
1198921198991198982
=120576OX cos (Λ 119899119909) sin (119872119898 (119910 minus 1198672 minus 119879OX))
2120576si cos (Λ 119899 (1198822)) sin (119872119898119879OX)
for 0 lt 119909 lt 119882
2
119867
2lt 119910 lt
119867
2+ 119879OX
1198921198991198983
= minuscos (Λ
119899119909) cos (119872
119898119910)
cos (Λ119899(1198822)) cos (119872
119898(1198672))
for 0 lt 119909 lt 119882
2 0 lt 119910 lt
119882
2
(22)
Subsequently the constants 120572119899119898 120573119899119898 and 120574
119899119898are evaluated
suitably
120572119899119898
= cos(Λ119899
119882
2) sin (119872
119898119879OX)
120573119899119898
= cos(119872119898
119867
2) sin (Λ
119899119879OX)
120574119899119898
= sin (119872119898119879OX) sin (Λ 119899119879OX)
(23)
From (20) the potential 119880119871119899119898
can be rewritten as
1198801198711198991198981
= minus cos(Λ119899
119882
2) sin (119872
119898119879OX)
times sin (Λ119899119879OX) cos(119872119898
119867
2)
Active and Passive Electronic Components 5
1198801198711198991198982
= minus cos(119872119898
119867
2) sin (Λ
119899119879OX)
times cos(Λ119899
119882
2) sin (119872
119898119879OX)
1198801198711198991198983
= minus sin (119872119898119879OX) sin (Λ 119899119879OX)
times cos(Λ119899
119882
2) cos(119872
119898
119867
2)
(24)
By multiplying with the corresponding orthogonal conjugatefunctions and integrating coefficients of119880
119871can be obtained
The coefficients of 119880119877are also obtained in a similar method
119894119899119898
= (int
119867+119879OX
0
int
119882+119879OX
0
(120595 (119909 119910 0))
minus119881 (119909 119910) 119892119899119898(119909 119910) 119889119909 119889119910)
times (int
119867+119879OX
0
int
119882+119879OX
0
119880119871119899119898
(119909 119910 0) 119892119899119898(119909 119910) 119889119909 119889119910)
minus1
119895119899119898
= (int
119867+119879OX
0
int
119882+119879OX
0
(120595 (119909 119910 119871))
minus119881 (119909 119910) 119892119899119898(119909 119910) 119889119909 119889119910)
times (int
119867+119879OX
0
int
119882+119879OX
0
119880119871119899119898
(119909 119910 119871) 119892119899119898(119909 119910) 119889119909 119889119910)
minus1
(25)
The above integrals (25) are evaluated to obtain explicitexpressions for 119894
119899119898and 119895119899119898
as follows
119894119899119898
= (119881119892minus 120595bi) 120596119899119898
119895119899119898
= (119881119892minus 120595bi minus 119881ds) 120596119899119898
120596119899119898
=Ω119899119898
120585119899119898
(26)
where
Ω119899119898
=120576OXΛ 119899 tan ((Λ 119899119882) 2) + 120576OX119872119898 tan (1198721198981198672)
2120576si119879OXΛ2
1198991198722119898
minus2
Λ119899119872119898
tan(Λ119899119882
2) tan(
119872119898119867
2)
120585119899119898
=sin (Λ
119899119879OX)
16 cos (Λ1198991198822)
(119882 +sin (Λ
119899119882)
Λ119899
) 1199031
+sin (119872
119898119879OX)
16 cos (1198721198981198672)
(119867 +sin (119872
119898119867)
119872119898
) 1199032
(27)
And the values of 1199031and 1199032are given by
1199031=120576OX119879OX cos (1198721198981198672)
120576si sin (119872119898119879OX)+119867 sin (119872
119898119879OX)
2 cos (1198721198981198672)
1199032=120576OX119879OX cos (Λ 1198991198822)
120576si sin (Λ 119899119879OX)+119882 sin (Λ
119899119879OX)
2 cos (Λ1198991198822)
(28)
The potential equation is now rewritten as
120595 (119909 119910 119911) = 119881119892+ (119881119892minus 120595bi)
timessum
119899
sum
119898
120588119899119898
sinh(sum119899119898
(119871 minus 119911))
+ (119881119892minus 120595bi minus 119881ds)sum
119899
sum
119898
sinh(sum119899119898
119911)
(29)
where
120588119899119898(119909 119910) = minus120596
119899119898120574119899119898
cos (Λ119899119909) cos (119872
119898119910)
sinh (sum119899119898119871)
(30)
Once the potential distribution at every point of the cross-section of the channel is known we calculate the inversioncharge density by using surface integral over the surface areaof the channel When the integrated charge at virtual sourcebecomes equal to critical charge the gate voltage of a lightlydoped body device is nearly equal to the threshold voltage ofthe device Hence the inversion charge can be expressed as
119876 = int
1198672
minus1198672
int
1198822
minus1198822
119902119899119894119890(120595119880119879)119889119909 119889119910 (31)
where 119902 is the elementary charge 119880119879is thermal voltage and
119899119894is the intrinsic carrier concentrationThe charge equation can now be approximated as
119876 asymp 119882119867119902119899119894119890((120595((311988214)(311986714)119911119888))119880119879) (32)
Here 119911119888is the virtual source position which is half of the
channel length for low 119881ds Using the inversion charge wecan obtain the classical threshold model as expressed in thefollowing
119881TC = (119880119879 ln119876
119882119867119902119899119894
+ 2119881bi12058811 (3119882
143119867
14)
times sinh(sum11119871
2))
times (1 + 212058811(3119882
143119867
14) sinh(
sum11119871
2))
minus1
(33)
3 Quantum Threshold Voltage Modeling
AsMOSFETdevices are further scaled into the deep nanome-ter regime it has become necessary to include quantummechanical effects while modeling their device behaviorIn this paper we approximate the actual potential well as
6 Active and Passive Electronic Components
E11
Ec
Ec
998779Ec
Ei119909i119910
Eco
Figure 2 Band diagram perpendicular to the gate-square wellpotential of a GAA silicon nanowire transistor
the square well potential since it is difficult to solve theSchrodinger equation to obtain the potential expressed in(29) The square well potential of a gate-all-around nanowiretransistor is shown in Figure 2 The quantum charge of thedevice is expressed as
119876 = sum
119894119909
sum
119894119910
119902int
infin
119864119894119909119894119910
1198731D119891 (119864) 119889119864 (34)
where1198731D is the 1D density-of-states and 119891(119864) is the Fermi-
Dirac distribution function 119864 is the energy of the electronwave The terms 119894
119909and 119894119910are positive natural numbers
In silicon six energy valleys are found to be present inits band structure (two lower energy valleys two middleenergy valleys and two higher energy valleys) If the thinfilm of device has equal height and width the two lowerenergy valleys and two middle energy valleys are combinedtogether to produce four lower energy valleys and the othertwo higher energy valleys remain in their own stateThus thecharge is given by
119876 = 119902sum
119894119909
sum
119894119910
radic(119898119911
2120587ℎ2)int
infin
119864119894119909119894119910
(119864 minus 119864119894119909119894119910)minus12
1 + 119890 ((119864 minus 119864119865) 119896119879)
119889119864 (35)
where 119898119911is the mass of the valley which is perpendicular to
the direction of quantizationThe Fermi energy level is muchlower than the conduction band energy in weak inversionregion Hence the charge equation can be approximated as
119876 = 119902radic(119898119911
2120587ℎ2)sum
119894119909
sum
119894119910
int
infin
119864119894119909119894119910
119890((119864119865minus119864)119896119879)
(119864 minus 119864119894119909119894119910)12119889119864 (36)
Using the Schrodinger equation the value of 119864119894119909119864119894119910is deter-
mined by the following formulation [5]
119864119894119909119894119910
= 119864co +ℎ21205872
2[1
119898119909
(119894119909
119868119909
)
2
+1
119898119910
(119894119910
119868119910
)
2
] (37)
where the conduction band energy is given as
119864co =119864119892
2minus 119902120595 (0 0 119911
119888) (38)
Using (36) and (37) the integrated charge can be obtained as
119876 = 119902radic(2119896119879119898
119905
ℎ2)sum
119894119909
sum
119894119910
exp(minus119864co + 1198641 (119894119909 119894119910)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2)sum
119894119909
sum
119894119910
exp(minus119864co + 1198642 (119894119909 119894119910)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2)sum
119894119909
sum
119894119910
exp(minus119864co + 1198643 (119894119909 119894119910)
119896119879)
(39)
where
1198641(119894119909 119894119910) =
ℎ21205872
2[1
1198981
(119894119909
119882)
2
+1
119898119905
(119894119910
119867)
2
]
1198642(119894119909 119894119910) =
ℎ21205872
2[1
119898119905
(119894119909
119882)
2
+1
1198981
(119894119910
119867)
2
]
1198643(119894119909 119894119910) =
ℎ21205872
2[1
119898119905
(119894119909
119882)
2
+1
119898119905
(119894119910
119867)
2
]
(40)
Here the 119898119905and 119898
1are the transverse and longitudinal
effective masses of the energy valleys of silicon The lengths119894119909and 119894119910carry distinct values contingent on the direction of
quantization Finally the quantum threshold voltage modelbecomes
119881TQ = (119864119892
2119902+ (
119896119879
119902) In(119876119879
120591)
+2120595bi12058811 (0 0) sinh((sum11119871)
2))
times (1 + 212058811(0 0) sinh(
(sum11119871)
2))
minus1
(41)
where
120591 = 119902radic(2119896119879119898
119905
ℎ2) exp(minus1198641 (1 1)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2) exp(minus1198642 (1 1)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2) exp(minus
1198643(1 1)
119896119879)
(42)
Active and Passive Electronic Components 7
10 20 30 40 50 6002
022
024
026
028
03
032
034
036
038
04
Channel length (nm)
Pote
ntia
l (V
)
Proposed modelTCAD simulation
Figure 3 Constant electrostatic potential obtained from the analyt-ical solution of a gate-all-around silicon nanowire transistor is 03 Vfor different channel length with a height of119867 = 9 nm and channelwidth of 119882 = 9 nm TCAD simulation shows that the potential isconstant at 0296V
The impacts on the threshold voltage due to quantum effectsare acquired by using the following equation
119881TC = 119881TQ + Δ119881119879 (43)
Here Δ119881119879is the difference between the quantum threshold
voltage and the classical threshold voltage
4 Results and Discussion
Figure 3 shows the electrostatic potential of the proposed gateall around transistor and it is found to be constant value at03 V Continuously varying the 119899 and 119898 terms in (4) hasno impact on the potential as it remains constant along theinsulator boundaries This is totally in contrast to the resultsobtained in [15] where the potential is found to be linearlyvarying in the insulator boundaries The constant potentialhas to be deduced as the resultant of the gate voltage appliedsymmetrically across the four sides of the transistor TheTCAD simulation of the device shows that the electrostaticpotential is constant at 0296V The simulation results arefound to in acceptance with the TCAD results
Figure 4 represents the variation of total quantum inte-grated charge with the gate voltage Equation (39) is usedto obtain the integrated charge with only one energy leveland one series term It clearly shows that the decrease inthe film thickness leads to the increase in the quantumthreshold voltage which is actually due to the increase inenergy quantization of the transistor With the height andlength of the device being constant the width of the device isvaried and henceforth the variation of charge in accordancewith the gate voltage is illustrated in Figure 4
The variation of quantum threshold voltage with widthand height of the film at a channel length of 20 nm is
0 01 02 03 04 05 06 070
05
1
15
2
25
3
35
4times10minus15
Gate voltage (V)
Char
ge (c
mminus1)
W = 9e minus 9
W = 5e minus 9
W = 3e minus 9
TCAD simulation for W = 9e minus 9
TCAD simulation for W = 5e minus 9
TCAD simulation for W = 3e minus 9
Figure 4 Variation of quantum integrated charge at virtual sourcewith gate voltage for different film widths where the height andlength are119867 = 9 nm 119871 = 20 nm
2
8
times10minus9
times10minus9
Film height
2
44
66
8
10
1003
0305
031
0315
032
Film width
Qua
ntum
thre
shol
d vo
ltage
(V)
ProposedTCAD
Figure 5 Variation of quantum threshold voltage with film heightand width for channel length (119871 = 20 nm)
shown in Figure 5 The short channel effects tend to decreasealong with the energy quantisation and this can be furtherexplained as a result of increase in the effective band gap ofsilicon due to quantum effects The effect of confinementexpressed as the difference in the threshold voltage and itsvariation with the channel length 119871 is illustrated in Figure 6
8 Active and Passive Electronic Components
2 3 4 5 6 7 8 9 10
times10minus9
0014
0016
0018
002
0022
0024
0026
0028
003
0032
Film height
Proposed modelTCAD simulation
ΔVT
(mV
)
Figure 6 Variation of threshold voltagewith filmwidth for differentheight at 119871 = 20 nm
3 4 5 6 7 8
times10minus9
02
022
024
026
028
03
032
034
036
038
04
Film height
Thre
shol
d vo
ltage
(V)
ClassicalQuantum
TCAD simulation for classicalTCAD simulation for quantum
9
Figure 7 Variation of quantum and classical threshold voltage withfilm height Here 119871 = 20 nm and119882 = 9 nm
Themost important thing about this gate all around nanowiretransistor is that any change in one of the dimensions canbe nullified by proper tuning of other dimensions as thetransistor is symmetric about its height and width
Figure 7 shows the variation of the classical thresholdvoltage and quantum threshold voltage with the film height ata constant width of 9 nmThe value of the classical thresholdvoltage ranges from 027V to 029V for the correspondingchanges in the film height Similarly the quantum thresholdvoltage ranges from 03V to 031 V It shows that the devicehas a highly improved control over the threshold voltageTheTCAD results justify the simulation results
5 Conclusion
In this paper a quantum threshold voltage model for a GAAsilicon nanowire transistor is proposed by solving the 3DPoisson and Schrodinger equations Analytical expressionsfor potential and the inversion charge are expressed intheir closed forms The results show that the integratedcharge and the threshold voltage calculated in accordancewith the quantum effects of this proposed model are highlyimprovedThe future considerations include deriving the I-Vcharacteristics of the gate all around nanowire transistors andstudying the impact of scaling on various device parametersFinally to conclude this model provides an analytical anduseful way for the threshold voltage evaluations in gate allaround nanowire devices with a unified formalism employedin both classical and quantum mechanical approaches
References
[1] J Wang E Polizzi and M Lundstrom ldquoA computationalstudy of ballistic silicon nanowire transistorsrdquo in Proceedings ofthe IEEE International Electron Devices Meeting pp 695ndash698December 2003
[2] J P Colinge FINFETS and Other Multi-Gate TransistorsSpringer New York NY USA 2007
[3] D J Frank Y Taur andH-S PWong ldquoGeneralized scale lengthfor two-dimensional effects in MOSFETrsquosrdquo IEEE ElectronDevice Letters vol 19 no 10 pp 385ndash387 1998
[4] B Yu L Wang Y Yuan P M Asbeck and Y Taur ldquoScaling ofnanowire transistorsrdquo IEEE Transactions on Electron Devicesvol 55 no 11 pp 2846ndash2858 2008
[5] G D Sanders C J Stanton and Y C Chang ldquoTheory of trans-port in silicon quantum wiresrdquo Physical Review B vol 48 no15 pp 11067ndash11076 1993
[6] M-Y Shen and S-L Zhang ldquoBand gap of a silicon quantumwirerdquo Physics Letters A vol 176 no 3-4 pp 254ndash258 1993
[7] J P Colinge X Baie V Bayot and E Grivei ldquoQuantum-wireeffects in thin and narrow SOI MOSFETsrdquo in Proceedings of theIEEE International SOI Conference pp 66ndash67 October 1995
[8] C P Auth and J D Plummer ldquoScaling theory for cylindricalfully-depleted surrounding-gate MOSFETrsquosrdquo IEEE ElectronDevice Letters vol 18 no 2 pp 74ndash76 1997
[9] J-T Park and J-P Colinge ldquoMultiple-gate SOI MOSFETsdevice design guidelinesrdquo IEEE Transactions on ElectronDevices vol 49 no 12 pp 2222ndash2229 2002
[10] Z Ghoggali F DjeffalM A Abdi D Arar N NLakhdar and TTBendib ldquoAn analytical threshold voltagemodel for nanoscalerdquoin Proceedings of the 3rd International Design and TestWorkshop(IDT rsquo08) pp 93ndash97 December 2008
[11] Y-S Wu and P Su ldquoQuantum confinement effect in short-channel gate-all-around MOSFETs and its impact on the sensi-tivity of threshold voltage to process variationsrdquo in Proceedingsof the IEEE International SOI Conference October 2009
[12] C Te-Kuang ldquoA compact analytical threshold-voltage modelfor surrounding-gateMOSFETswith interface trapped chargesrdquoIEEE Electron Device Letters vol 31 no 8 pp 788ndash790 2010
[13] B Ray and SMahapatra ldquoModeling and analysis of body poten-tial of cylindrical gate-all-around nanowire transistorrdquo IEEETransactions on Electron Devices vol 55 no 9 pp 2409ndash24162008
Active and Passive Electronic Components 9
[14] L De Michielis L Selmi and A M Ionescu ldquoA quasi-analytical model for nanowire FETs with arbitrary polygonalcross sectionrdquo Solid-State Electronics vol 54 no 9 pp 929ndash9342010
[15] P R Kumar and S Mahapatra ldquoQuantum threshold voltagemodeling of short channel quad gate silicon nanowire transis-torrdquo IEEE Transactions onNanotechnology vol 10 no 1 pp 121ndash128 2011
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of
Active and Passive Electronic Components 3
the term 119881119892+ 119880119871is found to satisfy the potential equation
when119880119877is on null value and in an exact repetition the term
119881119892+ 119880119877satisfies the potential equation when 119880
119871is on null
value From (1)
120595119909119909+ 120595119910119910+ 120595119911119911= 0 (5)
By solving the above equation using LDE method we obtainthe value of 120595 Then the limits are applied on the equationusing the boundary conditions Now the potentials 119880
119871is
given by
1198801198711= sum
119899
sum
119898
119862 times 119870119860119899119898
sin(Λ119899(119909 minus
119882
2minus 119879OX))
times cos (119872119898119910) sinh(sum
119899119898
(119871 minus 119911))
for 1198822lt 119909 lt
119882
2+ 119879OX 0 lt 119910 lt
119867
2
(6)
1198801198712= sum
119899
sum
119898
119862 times 119870119861119899119898
cos (Λ119899119909)
times sin(119872119898(119910 minus
119867
2minus 119879OX))
times sinh(sum119899119898
(119871 minus 119911)) for 0 lt 119909 lt 119882
2
119867
2lt 119910 lt
119867
2+ 119879OX
(7)
1198801198713= sum
119899
sum
119898
119862 times 119870119862119899119898
cos (Λ119899119909) cos (119872
119898119910)
times sinh(sum119899119898
(119871 minus 119911))
for 0 lt 119909 lt 119882
2 0 lt 119910 lt
119867
2
(8)
Similarly the values of potential 119880119877are also derived as
follows
1198801198771= sum
119899
sum
119898
119862 times 119870119875119899119898
sin(Λ119899(119909 minus
119882
2minus 119879OX))
times cos (119872119898119910)
times sinh(sum119899119898
119911) for 1198822lt 119909 lt
119882
2+ 119879OX
0 lt 119910 lt119867
2
1198801198772= sum
119899
sum
119898
119862 times 119870119876119899119898
cos (Λ119899119909)
times sin(119872119898(119910 minus
119867
2minus 119879OX)) sinh(sum
119899119898
119911)
for 0 lt 119909 lt 119882
2
119867
2lt 119910 lt
119867
2+ 119879OX
1198801198773= sum
119899
sum
119898
119862 times 119870119877119899119898
cos (Λ119899119909) cos (119872
119898119910)
times sinh(sum119899119898
119911) for 0 lt 119909 lt 119882
2
0 lt 119910 lt119867
2
(9)
where 119888 = 1sum119899119898119871 and 119870
119860119899119898 119870119861119899119898
119870119862119899119898
119870119875119899119898
119870119902119899119898
and 119870
119877119899119898are constants From (6) and (8) 119880
119871and (120597119880
119871)120597119909
are found to be continuous in the 119909 direction (119909 = 1198822)The first derivative function (120597119880
1198711)120597119909 itself has 120576OX120576si
times discontinuities at the silicon insulator interfaces Thusapplying continuity in both equations we proceed to equate(6) and (8) as follows
minus119870119860119899119898
sin (Λ119899119879OX) = 119870119862119899119898 cos(
Λ119899119882
2) (10)
Differentiating (6) with respect to 119909
1205971198801198711
120597119909= sum
119899
sum
119898
119862 times 119870119860119899119898
(Λ119899) cos(Λ
119899(119909 minus
119882
2minus 119879OX))
times cos (119872119898119910) sinh(sum
119899119898
(119871 minus 119911))
(11)
Differentiating (8) with respect to 119909
1205971198801198713
120597119909= sum
119899
sum
119898
119862 times 119870119888119899119898
(Λ119899) (minus sin (Λ
119899119909))
times cos (119872119898119910) times sinh(sum
119899119898
(119871 minus 119911))
(12)
Equating (11) and (12)
119870119860119899119898
120576OX cos (Λ 119899119879OX) = minus119870119862119899119898120576si sin(Λ119899
2) (13)
Dividing (10) by (13)
120576si tan (Λ 119899119879OX) minus 120576OXcot(Λ119899119882
2) = 0 (14)
Likewise from (7) and (8) 119880119871and (120597119880
119871120597119910) are found to
be continuous in the 119910 direction (119910 = 1198672) The function(1205971198801198712120597119910) itself has discontinuities at the silicon insulator
interfaces which are proportional to the dielectric constant120576OX120576si Thus applying continuity in both equations andequating (7) and (8) we get
minus119870119861119899119898
sin (119872119898119879OX) = 119870119862119899119898 cos(
119872119898119867
2) (15)
4 Active and Passive Electronic Components
Differentiating (7) with respect to 119910
1205971198801198712
120597119910= sum
119899
sum
119898
119862 times 119870119861119899119898
cos (Λ119899119909) (119872
119898)
times cos(119872119898(119910 minus
119867
2minus 119879OX))
times sinh(sum119899119898
(119871 minus 119911))
(16)
Differentiating (8) with respect to 119910
1205971198801198713
120597119910= sum
119899
sum
119898
119862 times 119870119862119899119898
cos (Λ119899119909) (119872
119898)
times (minus sin (119872119898119910)) sinh(sum
119899119898
(119871 minus 119911))
(17)
Equating (16) and (17)
119870119861119899119898
120576OX cos (119872119898119879OX) = minus119870119862119899119898120576si sin(119872119898119867
2) (18)
Dividing (15) by (18)
120576si tan (119872119898119879OX) minus 120576OXcot(119872119898119867
2) = 0 (19)
This natural length is an easy guide for choosing deviceparameters and has simple physical meaning that a smallnatural length corresponds to superb short channel effectimmunity [4] The value of Λ
119899and 119872
119898depends on device
parameters The potential 119880119871can be modified as
1198801198711198991198981
= 120572119899119898
times 119862 times sin(Λ119899(119909 minus
119882
2minus 119879OX))
times cos (119872119898119910) times sinh(sum
119899119898
(119871 minus 119911))
for 1198822lt 119909 lt
119882
2+ 119879OX 0 lt 119910 lt
119867
2
1198801198711198991198982
= 120573119899119898
times 119862 times cos (Λ119899119909)
times sin(119872119898(119910 minus
119867
2minus 119879OX)) sinh(sum
119899119898
(119871 minus 119911))
for 0 lt 119909 lt 119882
2
119867
2lt 119910 lt
119867
2+ 119879OX
1198801198711198991198983
= 120574119899119898
times 119862 times cos (Λ119899119909) cos (119872
119898119910)
times sinhsum119899119898
(119871 minus 119911) for 0 lt 119909 lt 119882
2 0 lt 119910 lt
119867
2
(20)
Similarly the potential 119880119877can be modified as
1198801198771198991198981
= 120572119899119898
times 119862 times sin(Λ119899(119909 minus
119882
2minus 119879OX))
times cos (119872119898119910)
times sinh(sum119899119898
119911) for 1198822lt 119909 lt
119882
2+ 119879OX
0 lt 119910 lt119867
2
1198801198771198991198982
= 120573119899119898
times 119862 times cos (Λ119899119909)
times sin(119872119898(119910 minus
119867
2minus 119879OX)) times sinh(sum
119899119898
119911)
for 0 lt 119909 lt 119882
2
119867
2lt 119910 lt
119867
2+ 119879OX
1198801198771198991198983
= 120574119899119898
times cos (Λ119899119909) cos (119872
119898119910)
times sinh(sum119899119898
119911) for 0 lt 119909 lt 119882
2 0 lt 119910 lt
119867
2
(21)
Now the 119892119899119898
can be obtained from the potential equations(20) by using different multipliers in different regions
1198921198991198981
=120576OX sin (Λ 119899 (119909 minus1198822 minus 119879OX)) cos (119872119898119910)
2120576si sin (Λ 119899119879OX) cos (119872119898 (1198672))
for 1198822lt 119909 lt
119882
2+ 119879OX 0 lt 119910 lt
119867
2
1198921198991198982
=120576OX cos (Λ 119899119909) sin (119872119898 (119910 minus 1198672 minus 119879OX))
2120576si cos (Λ 119899 (1198822)) sin (119872119898119879OX)
for 0 lt 119909 lt 119882
2
119867
2lt 119910 lt
119867
2+ 119879OX
1198921198991198983
= minuscos (Λ
119899119909) cos (119872
119898119910)
cos (Λ119899(1198822)) cos (119872
119898(1198672))
for 0 lt 119909 lt 119882
2 0 lt 119910 lt
119882
2
(22)
Subsequently the constants 120572119899119898 120573119899119898 and 120574
119899119898are evaluated
suitably
120572119899119898
= cos(Λ119899
119882
2) sin (119872
119898119879OX)
120573119899119898
= cos(119872119898
119867
2) sin (Λ
119899119879OX)
120574119899119898
= sin (119872119898119879OX) sin (Λ 119899119879OX)
(23)
From (20) the potential 119880119871119899119898
can be rewritten as
1198801198711198991198981
= minus cos(Λ119899
119882
2) sin (119872
119898119879OX)
times sin (Λ119899119879OX) cos(119872119898
119867
2)
Active and Passive Electronic Components 5
1198801198711198991198982
= minus cos(119872119898
119867
2) sin (Λ
119899119879OX)
times cos(Λ119899
119882
2) sin (119872
119898119879OX)
1198801198711198991198983
= minus sin (119872119898119879OX) sin (Λ 119899119879OX)
times cos(Λ119899
119882
2) cos(119872
119898
119867
2)
(24)
By multiplying with the corresponding orthogonal conjugatefunctions and integrating coefficients of119880
119871can be obtained
The coefficients of 119880119877are also obtained in a similar method
119894119899119898
= (int
119867+119879OX
0
int
119882+119879OX
0
(120595 (119909 119910 0))
minus119881 (119909 119910) 119892119899119898(119909 119910) 119889119909 119889119910)
times (int
119867+119879OX
0
int
119882+119879OX
0
119880119871119899119898
(119909 119910 0) 119892119899119898(119909 119910) 119889119909 119889119910)
minus1
119895119899119898
= (int
119867+119879OX
0
int
119882+119879OX
0
(120595 (119909 119910 119871))
minus119881 (119909 119910) 119892119899119898(119909 119910) 119889119909 119889119910)
times (int
119867+119879OX
0
int
119882+119879OX
0
119880119871119899119898
(119909 119910 119871) 119892119899119898(119909 119910) 119889119909 119889119910)
minus1
(25)
The above integrals (25) are evaluated to obtain explicitexpressions for 119894
119899119898and 119895119899119898
as follows
119894119899119898
= (119881119892minus 120595bi) 120596119899119898
119895119899119898
= (119881119892minus 120595bi minus 119881ds) 120596119899119898
120596119899119898
=Ω119899119898
120585119899119898
(26)
where
Ω119899119898
=120576OXΛ 119899 tan ((Λ 119899119882) 2) + 120576OX119872119898 tan (1198721198981198672)
2120576si119879OXΛ2
1198991198722119898
minus2
Λ119899119872119898
tan(Λ119899119882
2) tan(
119872119898119867
2)
120585119899119898
=sin (Λ
119899119879OX)
16 cos (Λ1198991198822)
(119882 +sin (Λ
119899119882)
Λ119899
) 1199031
+sin (119872
119898119879OX)
16 cos (1198721198981198672)
(119867 +sin (119872
119898119867)
119872119898
) 1199032
(27)
And the values of 1199031and 1199032are given by
1199031=120576OX119879OX cos (1198721198981198672)
120576si sin (119872119898119879OX)+119867 sin (119872
119898119879OX)
2 cos (1198721198981198672)
1199032=120576OX119879OX cos (Λ 1198991198822)
120576si sin (Λ 119899119879OX)+119882 sin (Λ
119899119879OX)
2 cos (Λ1198991198822)
(28)
The potential equation is now rewritten as
120595 (119909 119910 119911) = 119881119892+ (119881119892minus 120595bi)
timessum
119899
sum
119898
120588119899119898
sinh(sum119899119898
(119871 minus 119911))
+ (119881119892minus 120595bi minus 119881ds)sum
119899
sum
119898
sinh(sum119899119898
119911)
(29)
where
120588119899119898(119909 119910) = minus120596
119899119898120574119899119898
cos (Λ119899119909) cos (119872
119898119910)
sinh (sum119899119898119871)
(30)
Once the potential distribution at every point of the cross-section of the channel is known we calculate the inversioncharge density by using surface integral over the surface areaof the channel When the integrated charge at virtual sourcebecomes equal to critical charge the gate voltage of a lightlydoped body device is nearly equal to the threshold voltage ofthe device Hence the inversion charge can be expressed as
119876 = int
1198672
minus1198672
int
1198822
minus1198822
119902119899119894119890(120595119880119879)119889119909 119889119910 (31)
where 119902 is the elementary charge 119880119879is thermal voltage and
119899119894is the intrinsic carrier concentrationThe charge equation can now be approximated as
119876 asymp 119882119867119902119899119894119890((120595((311988214)(311986714)119911119888))119880119879) (32)
Here 119911119888is the virtual source position which is half of the
channel length for low 119881ds Using the inversion charge wecan obtain the classical threshold model as expressed in thefollowing
119881TC = (119880119879 ln119876
119882119867119902119899119894
+ 2119881bi12058811 (3119882
143119867
14)
times sinh(sum11119871
2))
times (1 + 212058811(3119882
143119867
14) sinh(
sum11119871
2))
minus1
(33)
3 Quantum Threshold Voltage Modeling
AsMOSFETdevices are further scaled into the deep nanome-ter regime it has become necessary to include quantummechanical effects while modeling their device behaviorIn this paper we approximate the actual potential well as
6 Active and Passive Electronic Components
E11
Ec
Ec
998779Ec
Ei119909i119910
Eco
Figure 2 Band diagram perpendicular to the gate-square wellpotential of a GAA silicon nanowire transistor
the square well potential since it is difficult to solve theSchrodinger equation to obtain the potential expressed in(29) The square well potential of a gate-all-around nanowiretransistor is shown in Figure 2 The quantum charge of thedevice is expressed as
119876 = sum
119894119909
sum
119894119910
119902int
infin
119864119894119909119894119910
1198731D119891 (119864) 119889119864 (34)
where1198731D is the 1D density-of-states and 119891(119864) is the Fermi-
Dirac distribution function 119864 is the energy of the electronwave The terms 119894
119909and 119894119910are positive natural numbers
In silicon six energy valleys are found to be present inits band structure (two lower energy valleys two middleenergy valleys and two higher energy valleys) If the thinfilm of device has equal height and width the two lowerenergy valleys and two middle energy valleys are combinedtogether to produce four lower energy valleys and the othertwo higher energy valleys remain in their own stateThus thecharge is given by
119876 = 119902sum
119894119909
sum
119894119910
radic(119898119911
2120587ℎ2)int
infin
119864119894119909119894119910
(119864 minus 119864119894119909119894119910)minus12
1 + 119890 ((119864 minus 119864119865) 119896119879)
119889119864 (35)
where 119898119911is the mass of the valley which is perpendicular to
the direction of quantizationThe Fermi energy level is muchlower than the conduction band energy in weak inversionregion Hence the charge equation can be approximated as
119876 = 119902radic(119898119911
2120587ℎ2)sum
119894119909
sum
119894119910
int
infin
119864119894119909119894119910
119890((119864119865minus119864)119896119879)
(119864 minus 119864119894119909119894119910)12119889119864 (36)
Using the Schrodinger equation the value of 119864119894119909119864119894119910is deter-
mined by the following formulation [5]
119864119894119909119894119910
= 119864co +ℎ21205872
2[1
119898119909
(119894119909
119868119909
)
2
+1
119898119910
(119894119910
119868119910
)
2
] (37)
where the conduction band energy is given as
119864co =119864119892
2minus 119902120595 (0 0 119911
119888) (38)
Using (36) and (37) the integrated charge can be obtained as
119876 = 119902radic(2119896119879119898
119905
ℎ2)sum
119894119909
sum
119894119910
exp(minus119864co + 1198641 (119894119909 119894119910)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2)sum
119894119909
sum
119894119910
exp(minus119864co + 1198642 (119894119909 119894119910)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2)sum
119894119909
sum
119894119910
exp(minus119864co + 1198643 (119894119909 119894119910)
119896119879)
(39)
where
1198641(119894119909 119894119910) =
ℎ21205872
2[1
1198981
(119894119909
119882)
2
+1
119898119905
(119894119910
119867)
2
]
1198642(119894119909 119894119910) =
ℎ21205872
2[1
119898119905
(119894119909
119882)
2
+1
1198981
(119894119910
119867)
2
]
1198643(119894119909 119894119910) =
ℎ21205872
2[1
119898119905
(119894119909
119882)
2
+1
119898119905
(119894119910
119867)
2
]
(40)
Here the 119898119905and 119898
1are the transverse and longitudinal
effective masses of the energy valleys of silicon The lengths119894119909and 119894119910carry distinct values contingent on the direction of
quantization Finally the quantum threshold voltage modelbecomes
119881TQ = (119864119892
2119902+ (
119896119879
119902) In(119876119879
120591)
+2120595bi12058811 (0 0) sinh((sum11119871)
2))
times (1 + 212058811(0 0) sinh(
(sum11119871)
2))
minus1
(41)
where
120591 = 119902radic(2119896119879119898
119905
ℎ2) exp(minus1198641 (1 1)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2) exp(minus1198642 (1 1)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2) exp(minus
1198643(1 1)
119896119879)
(42)
Active and Passive Electronic Components 7
10 20 30 40 50 6002
022
024
026
028
03
032
034
036
038
04
Channel length (nm)
Pote
ntia
l (V
)
Proposed modelTCAD simulation
Figure 3 Constant electrostatic potential obtained from the analyt-ical solution of a gate-all-around silicon nanowire transistor is 03 Vfor different channel length with a height of119867 = 9 nm and channelwidth of 119882 = 9 nm TCAD simulation shows that the potential isconstant at 0296V
The impacts on the threshold voltage due to quantum effectsare acquired by using the following equation
119881TC = 119881TQ + Δ119881119879 (43)
Here Δ119881119879is the difference between the quantum threshold
voltage and the classical threshold voltage
4 Results and Discussion
Figure 3 shows the electrostatic potential of the proposed gateall around transistor and it is found to be constant value at03 V Continuously varying the 119899 and 119898 terms in (4) hasno impact on the potential as it remains constant along theinsulator boundaries This is totally in contrast to the resultsobtained in [15] where the potential is found to be linearlyvarying in the insulator boundaries The constant potentialhas to be deduced as the resultant of the gate voltage appliedsymmetrically across the four sides of the transistor TheTCAD simulation of the device shows that the electrostaticpotential is constant at 0296V The simulation results arefound to in acceptance with the TCAD results
Figure 4 represents the variation of total quantum inte-grated charge with the gate voltage Equation (39) is usedto obtain the integrated charge with only one energy leveland one series term It clearly shows that the decrease inthe film thickness leads to the increase in the quantumthreshold voltage which is actually due to the increase inenergy quantization of the transistor With the height andlength of the device being constant the width of the device isvaried and henceforth the variation of charge in accordancewith the gate voltage is illustrated in Figure 4
The variation of quantum threshold voltage with widthand height of the film at a channel length of 20 nm is
0 01 02 03 04 05 06 070
05
1
15
2
25
3
35
4times10minus15
Gate voltage (V)
Char
ge (c
mminus1)
W = 9e minus 9
W = 5e minus 9
W = 3e minus 9
TCAD simulation for W = 9e minus 9
TCAD simulation for W = 5e minus 9
TCAD simulation for W = 3e minus 9
Figure 4 Variation of quantum integrated charge at virtual sourcewith gate voltage for different film widths where the height andlength are119867 = 9 nm 119871 = 20 nm
2
8
times10minus9
times10minus9
Film height
2
44
66
8
10
1003
0305
031
0315
032
Film width
Qua
ntum
thre
shol
d vo
ltage
(V)
ProposedTCAD
Figure 5 Variation of quantum threshold voltage with film heightand width for channel length (119871 = 20 nm)
shown in Figure 5 The short channel effects tend to decreasealong with the energy quantisation and this can be furtherexplained as a result of increase in the effective band gap ofsilicon due to quantum effects The effect of confinementexpressed as the difference in the threshold voltage and itsvariation with the channel length 119871 is illustrated in Figure 6
8 Active and Passive Electronic Components
2 3 4 5 6 7 8 9 10
times10minus9
0014
0016
0018
002
0022
0024
0026
0028
003
0032
Film height
Proposed modelTCAD simulation
ΔVT
(mV
)
Figure 6 Variation of threshold voltagewith filmwidth for differentheight at 119871 = 20 nm
3 4 5 6 7 8
times10minus9
02
022
024
026
028
03
032
034
036
038
04
Film height
Thre
shol
d vo
ltage
(V)
ClassicalQuantum
TCAD simulation for classicalTCAD simulation for quantum
9
Figure 7 Variation of quantum and classical threshold voltage withfilm height Here 119871 = 20 nm and119882 = 9 nm
Themost important thing about this gate all around nanowiretransistor is that any change in one of the dimensions canbe nullified by proper tuning of other dimensions as thetransistor is symmetric about its height and width
Figure 7 shows the variation of the classical thresholdvoltage and quantum threshold voltage with the film height ata constant width of 9 nmThe value of the classical thresholdvoltage ranges from 027V to 029V for the correspondingchanges in the film height Similarly the quantum thresholdvoltage ranges from 03V to 031 V It shows that the devicehas a highly improved control over the threshold voltageTheTCAD results justify the simulation results
5 Conclusion
In this paper a quantum threshold voltage model for a GAAsilicon nanowire transistor is proposed by solving the 3DPoisson and Schrodinger equations Analytical expressionsfor potential and the inversion charge are expressed intheir closed forms The results show that the integratedcharge and the threshold voltage calculated in accordancewith the quantum effects of this proposed model are highlyimprovedThe future considerations include deriving the I-Vcharacteristics of the gate all around nanowire transistors andstudying the impact of scaling on various device parametersFinally to conclude this model provides an analytical anduseful way for the threshold voltage evaluations in gate allaround nanowire devices with a unified formalism employedin both classical and quantum mechanical approaches
References
[1] J Wang E Polizzi and M Lundstrom ldquoA computationalstudy of ballistic silicon nanowire transistorsrdquo in Proceedings ofthe IEEE International Electron Devices Meeting pp 695ndash698December 2003
[2] J P Colinge FINFETS and Other Multi-Gate TransistorsSpringer New York NY USA 2007
[3] D J Frank Y Taur andH-S PWong ldquoGeneralized scale lengthfor two-dimensional effects in MOSFETrsquosrdquo IEEE ElectronDevice Letters vol 19 no 10 pp 385ndash387 1998
[4] B Yu L Wang Y Yuan P M Asbeck and Y Taur ldquoScaling ofnanowire transistorsrdquo IEEE Transactions on Electron Devicesvol 55 no 11 pp 2846ndash2858 2008
[5] G D Sanders C J Stanton and Y C Chang ldquoTheory of trans-port in silicon quantum wiresrdquo Physical Review B vol 48 no15 pp 11067ndash11076 1993
[6] M-Y Shen and S-L Zhang ldquoBand gap of a silicon quantumwirerdquo Physics Letters A vol 176 no 3-4 pp 254ndash258 1993
[7] J P Colinge X Baie V Bayot and E Grivei ldquoQuantum-wireeffects in thin and narrow SOI MOSFETsrdquo in Proceedings of theIEEE International SOI Conference pp 66ndash67 October 1995
[8] C P Auth and J D Plummer ldquoScaling theory for cylindricalfully-depleted surrounding-gate MOSFETrsquosrdquo IEEE ElectronDevice Letters vol 18 no 2 pp 74ndash76 1997
[9] J-T Park and J-P Colinge ldquoMultiple-gate SOI MOSFETsdevice design guidelinesrdquo IEEE Transactions on ElectronDevices vol 49 no 12 pp 2222ndash2229 2002
[10] Z Ghoggali F DjeffalM A Abdi D Arar N NLakhdar and TTBendib ldquoAn analytical threshold voltagemodel for nanoscalerdquoin Proceedings of the 3rd International Design and TestWorkshop(IDT rsquo08) pp 93ndash97 December 2008
[11] Y-S Wu and P Su ldquoQuantum confinement effect in short-channel gate-all-around MOSFETs and its impact on the sensi-tivity of threshold voltage to process variationsrdquo in Proceedingsof the IEEE International SOI Conference October 2009
[12] C Te-Kuang ldquoA compact analytical threshold-voltage modelfor surrounding-gateMOSFETswith interface trapped chargesrdquoIEEE Electron Device Letters vol 31 no 8 pp 788ndash790 2010
[13] B Ray and SMahapatra ldquoModeling and analysis of body poten-tial of cylindrical gate-all-around nanowire transistorrdquo IEEETransactions on Electron Devices vol 55 no 9 pp 2409ndash24162008
Active and Passive Electronic Components 9
[14] L De Michielis L Selmi and A M Ionescu ldquoA quasi-analytical model for nanowire FETs with arbitrary polygonalcross sectionrdquo Solid-State Electronics vol 54 no 9 pp 929ndash9342010
[15] P R Kumar and S Mahapatra ldquoQuantum threshold voltagemodeling of short channel quad gate silicon nanowire transis-torrdquo IEEE Transactions onNanotechnology vol 10 no 1 pp 121ndash128 2011
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of
4 Active and Passive Electronic Components
Differentiating (7) with respect to 119910
1205971198801198712
120597119910= sum
119899
sum
119898
119862 times 119870119861119899119898
cos (Λ119899119909) (119872
119898)
times cos(119872119898(119910 minus
119867
2minus 119879OX))
times sinh(sum119899119898
(119871 minus 119911))
(16)
Differentiating (8) with respect to 119910
1205971198801198713
120597119910= sum
119899
sum
119898
119862 times 119870119862119899119898
cos (Λ119899119909) (119872
119898)
times (minus sin (119872119898119910)) sinh(sum
119899119898
(119871 minus 119911))
(17)
Equating (16) and (17)
119870119861119899119898
120576OX cos (119872119898119879OX) = minus119870119862119899119898120576si sin(119872119898119867
2) (18)
Dividing (15) by (18)
120576si tan (119872119898119879OX) minus 120576OXcot(119872119898119867
2) = 0 (19)
This natural length is an easy guide for choosing deviceparameters and has simple physical meaning that a smallnatural length corresponds to superb short channel effectimmunity [4] The value of Λ
119899and 119872
119898depends on device
parameters The potential 119880119871can be modified as
1198801198711198991198981
= 120572119899119898
times 119862 times sin(Λ119899(119909 minus
119882
2minus 119879OX))
times cos (119872119898119910) times sinh(sum
119899119898
(119871 minus 119911))
for 1198822lt 119909 lt
119882
2+ 119879OX 0 lt 119910 lt
119867
2
1198801198711198991198982
= 120573119899119898
times 119862 times cos (Λ119899119909)
times sin(119872119898(119910 minus
119867
2minus 119879OX)) sinh(sum
119899119898
(119871 minus 119911))
for 0 lt 119909 lt 119882
2
119867
2lt 119910 lt
119867
2+ 119879OX
1198801198711198991198983
= 120574119899119898
times 119862 times cos (Λ119899119909) cos (119872
119898119910)
times sinhsum119899119898
(119871 minus 119911) for 0 lt 119909 lt 119882
2 0 lt 119910 lt
119867
2
(20)
Similarly the potential 119880119877can be modified as
1198801198771198991198981
= 120572119899119898
times 119862 times sin(Λ119899(119909 minus
119882
2minus 119879OX))
times cos (119872119898119910)
times sinh(sum119899119898
119911) for 1198822lt 119909 lt
119882
2+ 119879OX
0 lt 119910 lt119867
2
1198801198771198991198982
= 120573119899119898
times 119862 times cos (Λ119899119909)
times sin(119872119898(119910 minus
119867
2minus 119879OX)) times sinh(sum
119899119898
119911)
for 0 lt 119909 lt 119882
2
119867
2lt 119910 lt
119867
2+ 119879OX
1198801198771198991198983
= 120574119899119898
times cos (Λ119899119909) cos (119872
119898119910)
times sinh(sum119899119898
119911) for 0 lt 119909 lt 119882
2 0 lt 119910 lt
119867
2
(21)
Now the 119892119899119898
can be obtained from the potential equations(20) by using different multipliers in different regions
1198921198991198981
=120576OX sin (Λ 119899 (119909 minus1198822 minus 119879OX)) cos (119872119898119910)
2120576si sin (Λ 119899119879OX) cos (119872119898 (1198672))
for 1198822lt 119909 lt
119882
2+ 119879OX 0 lt 119910 lt
119867
2
1198921198991198982
=120576OX cos (Λ 119899119909) sin (119872119898 (119910 minus 1198672 minus 119879OX))
2120576si cos (Λ 119899 (1198822)) sin (119872119898119879OX)
for 0 lt 119909 lt 119882
2
119867
2lt 119910 lt
119867
2+ 119879OX
1198921198991198983
= minuscos (Λ
119899119909) cos (119872
119898119910)
cos (Λ119899(1198822)) cos (119872
119898(1198672))
for 0 lt 119909 lt 119882
2 0 lt 119910 lt
119882
2
(22)
Subsequently the constants 120572119899119898 120573119899119898 and 120574
119899119898are evaluated
suitably
120572119899119898
= cos(Λ119899
119882
2) sin (119872
119898119879OX)
120573119899119898
= cos(119872119898
119867
2) sin (Λ
119899119879OX)
120574119899119898
= sin (119872119898119879OX) sin (Λ 119899119879OX)
(23)
From (20) the potential 119880119871119899119898
can be rewritten as
1198801198711198991198981
= minus cos(Λ119899
119882
2) sin (119872
119898119879OX)
times sin (Λ119899119879OX) cos(119872119898
119867
2)
Active and Passive Electronic Components 5
1198801198711198991198982
= minus cos(119872119898
119867
2) sin (Λ
119899119879OX)
times cos(Λ119899
119882
2) sin (119872
119898119879OX)
1198801198711198991198983
= minus sin (119872119898119879OX) sin (Λ 119899119879OX)
times cos(Λ119899
119882
2) cos(119872
119898
119867
2)
(24)
By multiplying with the corresponding orthogonal conjugatefunctions and integrating coefficients of119880
119871can be obtained
The coefficients of 119880119877are also obtained in a similar method
119894119899119898
= (int
119867+119879OX
0
int
119882+119879OX
0
(120595 (119909 119910 0))
minus119881 (119909 119910) 119892119899119898(119909 119910) 119889119909 119889119910)
times (int
119867+119879OX
0
int
119882+119879OX
0
119880119871119899119898
(119909 119910 0) 119892119899119898(119909 119910) 119889119909 119889119910)
minus1
119895119899119898
= (int
119867+119879OX
0
int
119882+119879OX
0
(120595 (119909 119910 119871))
minus119881 (119909 119910) 119892119899119898(119909 119910) 119889119909 119889119910)
times (int
119867+119879OX
0
int
119882+119879OX
0
119880119871119899119898
(119909 119910 119871) 119892119899119898(119909 119910) 119889119909 119889119910)
minus1
(25)
The above integrals (25) are evaluated to obtain explicitexpressions for 119894
119899119898and 119895119899119898
as follows
119894119899119898
= (119881119892minus 120595bi) 120596119899119898
119895119899119898
= (119881119892minus 120595bi minus 119881ds) 120596119899119898
120596119899119898
=Ω119899119898
120585119899119898
(26)
where
Ω119899119898
=120576OXΛ 119899 tan ((Λ 119899119882) 2) + 120576OX119872119898 tan (1198721198981198672)
2120576si119879OXΛ2
1198991198722119898
minus2
Λ119899119872119898
tan(Λ119899119882
2) tan(
119872119898119867
2)
120585119899119898
=sin (Λ
119899119879OX)
16 cos (Λ1198991198822)
(119882 +sin (Λ
119899119882)
Λ119899
) 1199031
+sin (119872
119898119879OX)
16 cos (1198721198981198672)
(119867 +sin (119872
119898119867)
119872119898
) 1199032
(27)
And the values of 1199031and 1199032are given by
1199031=120576OX119879OX cos (1198721198981198672)
120576si sin (119872119898119879OX)+119867 sin (119872
119898119879OX)
2 cos (1198721198981198672)
1199032=120576OX119879OX cos (Λ 1198991198822)
120576si sin (Λ 119899119879OX)+119882 sin (Λ
119899119879OX)
2 cos (Λ1198991198822)
(28)
The potential equation is now rewritten as
120595 (119909 119910 119911) = 119881119892+ (119881119892minus 120595bi)
timessum
119899
sum
119898
120588119899119898
sinh(sum119899119898
(119871 minus 119911))
+ (119881119892minus 120595bi minus 119881ds)sum
119899
sum
119898
sinh(sum119899119898
119911)
(29)
where
120588119899119898(119909 119910) = minus120596
119899119898120574119899119898
cos (Λ119899119909) cos (119872
119898119910)
sinh (sum119899119898119871)
(30)
Once the potential distribution at every point of the cross-section of the channel is known we calculate the inversioncharge density by using surface integral over the surface areaof the channel When the integrated charge at virtual sourcebecomes equal to critical charge the gate voltage of a lightlydoped body device is nearly equal to the threshold voltage ofthe device Hence the inversion charge can be expressed as
119876 = int
1198672
minus1198672
int
1198822
minus1198822
119902119899119894119890(120595119880119879)119889119909 119889119910 (31)
where 119902 is the elementary charge 119880119879is thermal voltage and
119899119894is the intrinsic carrier concentrationThe charge equation can now be approximated as
119876 asymp 119882119867119902119899119894119890((120595((311988214)(311986714)119911119888))119880119879) (32)
Here 119911119888is the virtual source position which is half of the
channel length for low 119881ds Using the inversion charge wecan obtain the classical threshold model as expressed in thefollowing
119881TC = (119880119879 ln119876
119882119867119902119899119894
+ 2119881bi12058811 (3119882
143119867
14)
times sinh(sum11119871
2))
times (1 + 212058811(3119882
143119867
14) sinh(
sum11119871
2))
minus1
(33)
3 Quantum Threshold Voltage Modeling
AsMOSFETdevices are further scaled into the deep nanome-ter regime it has become necessary to include quantummechanical effects while modeling their device behaviorIn this paper we approximate the actual potential well as
6 Active and Passive Electronic Components
E11
Ec
Ec
998779Ec
Ei119909i119910
Eco
Figure 2 Band diagram perpendicular to the gate-square wellpotential of a GAA silicon nanowire transistor
the square well potential since it is difficult to solve theSchrodinger equation to obtain the potential expressed in(29) The square well potential of a gate-all-around nanowiretransistor is shown in Figure 2 The quantum charge of thedevice is expressed as
119876 = sum
119894119909
sum
119894119910
119902int
infin
119864119894119909119894119910
1198731D119891 (119864) 119889119864 (34)
where1198731D is the 1D density-of-states and 119891(119864) is the Fermi-
Dirac distribution function 119864 is the energy of the electronwave The terms 119894
119909and 119894119910are positive natural numbers
In silicon six energy valleys are found to be present inits band structure (two lower energy valleys two middleenergy valleys and two higher energy valleys) If the thinfilm of device has equal height and width the two lowerenergy valleys and two middle energy valleys are combinedtogether to produce four lower energy valleys and the othertwo higher energy valleys remain in their own stateThus thecharge is given by
119876 = 119902sum
119894119909
sum
119894119910
radic(119898119911
2120587ℎ2)int
infin
119864119894119909119894119910
(119864 minus 119864119894119909119894119910)minus12
1 + 119890 ((119864 minus 119864119865) 119896119879)
119889119864 (35)
where 119898119911is the mass of the valley which is perpendicular to
the direction of quantizationThe Fermi energy level is muchlower than the conduction band energy in weak inversionregion Hence the charge equation can be approximated as
119876 = 119902radic(119898119911
2120587ℎ2)sum
119894119909
sum
119894119910
int
infin
119864119894119909119894119910
119890((119864119865minus119864)119896119879)
(119864 minus 119864119894119909119894119910)12119889119864 (36)
Using the Schrodinger equation the value of 119864119894119909119864119894119910is deter-
mined by the following formulation [5]
119864119894119909119894119910
= 119864co +ℎ21205872
2[1
119898119909
(119894119909
119868119909
)
2
+1
119898119910
(119894119910
119868119910
)
2
] (37)
where the conduction band energy is given as
119864co =119864119892
2minus 119902120595 (0 0 119911
119888) (38)
Using (36) and (37) the integrated charge can be obtained as
119876 = 119902radic(2119896119879119898
119905
ℎ2)sum
119894119909
sum
119894119910
exp(minus119864co + 1198641 (119894119909 119894119910)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2)sum
119894119909
sum
119894119910
exp(minus119864co + 1198642 (119894119909 119894119910)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2)sum
119894119909
sum
119894119910
exp(minus119864co + 1198643 (119894119909 119894119910)
119896119879)
(39)
where
1198641(119894119909 119894119910) =
ℎ21205872
2[1
1198981
(119894119909
119882)
2
+1
119898119905
(119894119910
119867)
2
]
1198642(119894119909 119894119910) =
ℎ21205872
2[1
119898119905
(119894119909
119882)
2
+1
1198981
(119894119910
119867)
2
]
1198643(119894119909 119894119910) =
ℎ21205872
2[1
119898119905
(119894119909
119882)
2
+1
119898119905
(119894119910
119867)
2
]
(40)
Here the 119898119905and 119898
1are the transverse and longitudinal
effective masses of the energy valleys of silicon The lengths119894119909and 119894119910carry distinct values contingent on the direction of
quantization Finally the quantum threshold voltage modelbecomes
119881TQ = (119864119892
2119902+ (
119896119879
119902) In(119876119879
120591)
+2120595bi12058811 (0 0) sinh((sum11119871)
2))
times (1 + 212058811(0 0) sinh(
(sum11119871)
2))
minus1
(41)
where
120591 = 119902radic(2119896119879119898
119905
ℎ2) exp(minus1198641 (1 1)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2) exp(minus1198642 (1 1)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2) exp(minus
1198643(1 1)
119896119879)
(42)
Active and Passive Electronic Components 7
10 20 30 40 50 6002
022
024
026
028
03
032
034
036
038
04
Channel length (nm)
Pote
ntia
l (V
)
Proposed modelTCAD simulation
Figure 3 Constant electrostatic potential obtained from the analyt-ical solution of a gate-all-around silicon nanowire transistor is 03 Vfor different channel length with a height of119867 = 9 nm and channelwidth of 119882 = 9 nm TCAD simulation shows that the potential isconstant at 0296V
The impacts on the threshold voltage due to quantum effectsare acquired by using the following equation
119881TC = 119881TQ + Δ119881119879 (43)
Here Δ119881119879is the difference between the quantum threshold
voltage and the classical threshold voltage
4 Results and Discussion
Figure 3 shows the electrostatic potential of the proposed gateall around transistor and it is found to be constant value at03 V Continuously varying the 119899 and 119898 terms in (4) hasno impact on the potential as it remains constant along theinsulator boundaries This is totally in contrast to the resultsobtained in [15] where the potential is found to be linearlyvarying in the insulator boundaries The constant potentialhas to be deduced as the resultant of the gate voltage appliedsymmetrically across the four sides of the transistor TheTCAD simulation of the device shows that the electrostaticpotential is constant at 0296V The simulation results arefound to in acceptance with the TCAD results
Figure 4 represents the variation of total quantum inte-grated charge with the gate voltage Equation (39) is usedto obtain the integrated charge with only one energy leveland one series term It clearly shows that the decrease inthe film thickness leads to the increase in the quantumthreshold voltage which is actually due to the increase inenergy quantization of the transistor With the height andlength of the device being constant the width of the device isvaried and henceforth the variation of charge in accordancewith the gate voltage is illustrated in Figure 4
The variation of quantum threshold voltage with widthand height of the film at a channel length of 20 nm is
0 01 02 03 04 05 06 070
05
1
15
2
25
3
35
4times10minus15
Gate voltage (V)
Char
ge (c
mminus1)
W = 9e minus 9
W = 5e minus 9
W = 3e minus 9
TCAD simulation for W = 9e minus 9
TCAD simulation for W = 5e minus 9
TCAD simulation for W = 3e minus 9
Figure 4 Variation of quantum integrated charge at virtual sourcewith gate voltage for different film widths where the height andlength are119867 = 9 nm 119871 = 20 nm
2
8
times10minus9
times10minus9
Film height
2
44
66
8
10
1003
0305
031
0315
032
Film width
Qua
ntum
thre
shol
d vo
ltage
(V)
ProposedTCAD
Figure 5 Variation of quantum threshold voltage with film heightand width for channel length (119871 = 20 nm)
shown in Figure 5 The short channel effects tend to decreasealong with the energy quantisation and this can be furtherexplained as a result of increase in the effective band gap ofsilicon due to quantum effects The effect of confinementexpressed as the difference in the threshold voltage and itsvariation with the channel length 119871 is illustrated in Figure 6
8 Active and Passive Electronic Components
2 3 4 5 6 7 8 9 10
times10minus9
0014
0016
0018
002
0022
0024
0026
0028
003
0032
Film height
Proposed modelTCAD simulation
ΔVT
(mV
)
Figure 6 Variation of threshold voltagewith filmwidth for differentheight at 119871 = 20 nm
3 4 5 6 7 8
times10minus9
02
022
024
026
028
03
032
034
036
038
04
Film height
Thre
shol
d vo
ltage
(V)
ClassicalQuantum
TCAD simulation for classicalTCAD simulation for quantum
9
Figure 7 Variation of quantum and classical threshold voltage withfilm height Here 119871 = 20 nm and119882 = 9 nm
Themost important thing about this gate all around nanowiretransistor is that any change in one of the dimensions canbe nullified by proper tuning of other dimensions as thetransistor is symmetric about its height and width
Figure 7 shows the variation of the classical thresholdvoltage and quantum threshold voltage with the film height ata constant width of 9 nmThe value of the classical thresholdvoltage ranges from 027V to 029V for the correspondingchanges in the film height Similarly the quantum thresholdvoltage ranges from 03V to 031 V It shows that the devicehas a highly improved control over the threshold voltageTheTCAD results justify the simulation results
5 Conclusion
In this paper a quantum threshold voltage model for a GAAsilicon nanowire transistor is proposed by solving the 3DPoisson and Schrodinger equations Analytical expressionsfor potential and the inversion charge are expressed intheir closed forms The results show that the integratedcharge and the threshold voltage calculated in accordancewith the quantum effects of this proposed model are highlyimprovedThe future considerations include deriving the I-Vcharacteristics of the gate all around nanowire transistors andstudying the impact of scaling on various device parametersFinally to conclude this model provides an analytical anduseful way for the threshold voltage evaluations in gate allaround nanowire devices with a unified formalism employedin both classical and quantum mechanical approaches
References
[1] J Wang E Polizzi and M Lundstrom ldquoA computationalstudy of ballistic silicon nanowire transistorsrdquo in Proceedings ofthe IEEE International Electron Devices Meeting pp 695ndash698December 2003
[2] J P Colinge FINFETS and Other Multi-Gate TransistorsSpringer New York NY USA 2007
[3] D J Frank Y Taur andH-S PWong ldquoGeneralized scale lengthfor two-dimensional effects in MOSFETrsquosrdquo IEEE ElectronDevice Letters vol 19 no 10 pp 385ndash387 1998
[4] B Yu L Wang Y Yuan P M Asbeck and Y Taur ldquoScaling ofnanowire transistorsrdquo IEEE Transactions on Electron Devicesvol 55 no 11 pp 2846ndash2858 2008
[5] G D Sanders C J Stanton and Y C Chang ldquoTheory of trans-port in silicon quantum wiresrdquo Physical Review B vol 48 no15 pp 11067ndash11076 1993
[6] M-Y Shen and S-L Zhang ldquoBand gap of a silicon quantumwirerdquo Physics Letters A vol 176 no 3-4 pp 254ndash258 1993
[7] J P Colinge X Baie V Bayot and E Grivei ldquoQuantum-wireeffects in thin and narrow SOI MOSFETsrdquo in Proceedings of theIEEE International SOI Conference pp 66ndash67 October 1995
[8] C P Auth and J D Plummer ldquoScaling theory for cylindricalfully-depleted surrounding-gate MOSFETrsquosrdquo IEEE ElectronDevice Letters vol 18 no 2 pp 74ndash76 1997
[9] J-T Park and J-P Colinge ldquoMultiple-gate SOI MOSFETsdevice design guidelinesrdquo IEEE Transactions on ElectronDevices vol 49 no 12 pp 2222ndash2229 2002
[10] Z Ghoggali F DjeffalM A Abdi D Arar N NLakhdar and TTBendib ldquoAn analytical threshold voltagemodel for nanoscalerdquoin Proceedings of the 3rd International Design and TestWorkshop(IDT rsquo08) pp 93ndash97 December 2008
[11] Y-S Wu and P Su ldquoQuantum confinement effect in short-channel gate-all-around MOSFETs and its impact on the sensi-tivity of threshold voltage to process variationsrdquo in Proceedingsof the IEEE International SOI Conference October 2009
[12] C Te-Kuang ldquoA compact analytical threshold-voltage modelfor surrounding-gateMOSFETswith interface trapped chargesrdquoIEEE Electron Device Letters vol 31 no 8 pp 788ndash790 2010
[13] B Ray and SMahapatra ldquoModeling and analysis of body poten-tial of cylindrical gate-all-around nanowire transistorrdquo IEEETransactions on Electron Devices vol 55 no 9 pp 2409ndash24162008
Active and Passive Electronic Components 9
[14] L De Michielis L Selmi and A M Ionescu ldquoA quasi-analytical model for nanowire FETs with arbitrary polygonalcross sectionrdquo Solid-State Electronics vol 54 no 9 pp 929ndash9342010
[15] P R Kumar and S Mahapatra ldquoQuantum threshold voltagemodeling of short channel quad gate silicon nanowire transis-torrdquo IEEE Transactions onNanotechnology vol 10 no 1 pp 121ndash128 2011
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of
Active and Passive Electronic Components 5
1198801198711198991198982
= minus cos(119872119898
119867
2) sin (Λ
119899119879OX)
times cos(Λ119899
119882
2) sin (119872
119898119879OX)
1198801198711198991198983
= minus sin (119872119898119879OX) sin (Λ 119899119879OX)
times cos(Λ119899
119882
2) cos(119872
119898
119867
2)
(24)
By multiplying with the corresponding orthogonal conjugatefunctions and integrating coefficients of119880
119871can be obtained
The coefficients of 119880119877are also obtained in a similar method
119894119899119898
= (int
119867+119879OX
0
int
119882+119879OX
0
(120595 (119909 119910 0))
minus119881 (119909 119910) 119892119899119898(119909 119910) 119889119909 119889119910)
times (int
119867+119879OX
0
int
119882+119879OX
0
119880119871119899119898
(119909 119910 0) 119892119899119898(119909 119910) 119889119909 119889119910)
minus1
119895119899119898
= (int
119867+119879OX
0
int
119882+119879OX
0
(120595 (119909 119910 119871))
minus119881 (119909 119910) 119892119899119898(119909 119910) 119889119909 119889119910)
times (int
119867+119879OX
0
int
119882+119879OX
0
119880119871119899119898
(119909 119910 119871) 119892119899119898(119909 119910) 119889119909 119889119910)
minus1
(25)
The above integrals (25) are evaluated to obtain explicitexpressions for 119894
119899119898and 119895119899119898
as follows
119894119899119898
= (119881119892minus 120595bi) 120596119899119898
119895119899119898
= (119881119892minus 120595bi minus 119881ds) 120596119899119898
120596119899119898
=Ω119899119898
120585119899119898
(26)
where
Ω119899119898
=120576OXΛ 119899 tan ((Λ 119899119882) 2) + 120576OX119872119898 tan (1198721198981198672)
2120576si119879OXΛ2
1198991198722119898
minus2
Λ119899119872119898
tan(Λ119899119882
2) tan(
119872119898119867
2)
120585119899119898
=sin (Λ
119899119879OX)
16 cos (Λ1198991198822)
(119882 +sin (Λ
119899119882)
Λ119899
) 1199031
+sin (119872
119898119879OX)
16 cos (1198721198981198672)
(119867 +sin (119872
119898119867)
119872119898
) 1199032
(27)
And the values of 1199031and 1199032are given by
1199031=120576OX119879OX cos (1198721198981198672)
120576si sin (119872119898119879OX)+119867 sin (119872
119898119879OX)
2 cos (1198721198981198672)
1199032=120576OX119879OX cos (Λ 1198991198822)
120576si sin (Λ 119899119879OX)+119882 sin (Λ
119899119879OX)
2 cos (Λ1198991198822)
(28)
The potential equation is now rewritten as
120595 (119909 119910 119911) = 119881119892+ (119881119892minus 120595bi)
timessum
119899
sum
119898
120588119899119898
sinh(sum119899119898
(119871 minus 119911))
+ (119881119892minus 120595bi minus 119881ds)sum
119899
sum
119898
sinh(sum119899119898
119911)
(29)
where
120588119899119898(119909 119910) = minus120596
119899119898120574119899119898
cos (Λ119899119909) cos (119872
119898119910)
sinh (sum119899119898119871)
(30)
Once the potential distribution at every point of the cross-section of the channel is known we calculate the inversioncharge density by using surface integral over the surface areaof the channel When the integrated charge at virtual sourcebecomes equal to critical charge the gate voltage of a lightlydoped body device is nearly equal to the threshold voltage ofthe device Hence the inversion charge can be expressed as
119876 = int
1198672
minus1198672
int
1198822
minus1198822
119902119899119894119890(120595119880119879)119889119909 119889119910 (31)
where 119902 is the elementary charge 119880119879is thermal voltage and
119899119894is the intrinsic carrier concentrationThe charge equation can now be approximated as
119876 asymp 119882119867119902119899119894119890((120595((311988214)(311986714)119911119888))119880119879) (32)
Here 119911119888is the virtual source position which is half of the
channel length for low 119881ds Using the inversion charge wecan obtain the classical threshold model as expressed in thefollowing
119881TC = (119880119879 ln119876
119882119867119902119899119894
+ 2119881bi12058811 (3119882
143119867
14)
times sinh(sum11119871
2))
times (1 + 212058811(3119882
143119867
14) sinh(
sum11119871
2))
minus1
(33)
3 Quantum Threshold Voltage Modeling
AsMOSFETdevices are further scaled into the deep nanome-ter regime it has become necessary to include quantummechanical effects while modeling their device behaviorIn this paper we approximate the actual potential well as
6 Active and Passive Electronic Components
E11
Ec
Ec
998779Ec
Ei119909i119910
Eco
Figure 2 Band diagram perpendicular to the gate-square wellpotential of a GAA silicon nanowire transistor
the square well potential since it is difficult to solve theSchrodinger equation to obtain the potential expressed in(29) The square well potential of a gate-all-around nanowiretransistor is shown in Figure 2 The quantum charge of thedevice is expressed as
119876 = sum
119894119909
sum
119894119910
119902int
infin
119864119894119909119894119910
1198731D119891 (119864) 119889119864 (34)
where1198731D is the 1D density-of-states and 119891(119864) is the Fermi-
Dirac distribution function 119864 is the energy of the electronwave The terms 119894
119909and 119894119910are positive natural numbers
In silicon six energy valleys are found to be present inits band structure (two lower energy valleys two middleenergy valleys and two higher energy valleys) If the thinfilm of device has equal height and width the two lowerenergy valleys and two middle energy valleys are combinedtogether to produce four lower energy valleys and the othertwo higher energy valleys remain in their own stateThus thecharge is given by
119876 = 119902sum
119894119909
sum
119894119910
radic(119898119911
2120587ℎ2)int
infin
119864119894119909119894119910
(119864 minus 119864119894119909119894119910)minus12
1 + 119890 ((119864 minus 119864119865) 119896119879)
119889119864 (35)
where 119898119911is the mass of the valley which is perpendicular to
the direction of quantizationThe Fermi energy level is muchlower than the conduction band energy in weak inversionregion Hence the charge equation can be approximated as
119876 = 119902radic(119898119911
2120587ℎ2)sum
119894119909
sum
119894119910
int
infin
119864119894119909119894119910
119890((119864119865minus119864)119896119879)
(119864 minus 119864119894119909119894119910)12119889119864 (36)
Using the Schrodinger equation the value of 119864119894119909119864119894119910is deter-
mined by the following formulation [5]
119864119894119909119894119910
= 119864co +ℎ21205872
2[1
119898119909
(119894119909
119868119909
)
2
+1
119898119910
(119894119910
119868119910
)
2
] (37)
where the conduction band energy is given as
119864co =119864119892
2minus 119902120595 (0 0 119911
119888) (38)
Using (36) and (37) the integrated charge can be obtained as
119876 = 119902radic(2119896119879119898
119905
ℎ2)sum
119894119909
sum
119894119910
exp(minus119864co + 1198641 (119894119909 119894119910)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2)sum
119894119909
sum
119894119910
exp(minus119864co + 1198642 (119894119909 119894119910)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2)sum
119894119909
sum
119894119910
exp(minus119864co + 1198643 (119894119909 119894119910)
119896119879)
(39)
where
1198641(119894119909 119894119910) =
ℎ21205872
2[1
1198981
(119894119909
119882)
2
+1
119898119905
(119894119910
119867)
2
]
1198642(119894119909 119894119910) =
ℎ21205872
2[1
119898119905
(119894119909
119882)
2
+1
1198981
(119894119910
119867)
2
]
1198643(119894119909 119894119910) =
ℎ21205872
2[1
119898119905
(119894119909
119882)
2
+1
119898119905
(119894119910
119867)
2
]
(40)
Here the 119898119905and 119898
1are the transverse and longitudinal
effective masses of the energy valleys of silicon The lengths119894119909and 119894119910carry distinct values contingent on the direction of
quantization Finally the quantum threshold voltage modelbecomes
119881TQ = (119864119892
2119902+ (
119896119879
119902) In(119876119879
120591)
+2120595bi12058811 (0 0) sinh((sum11119871)
2))
times (1 + 212058811(0 0) sinh(
(sum11119871)
2))
minus1
(41)
where
120591 = 119902radic(2119896119879119898
119905
ℎ2) exp(minus1198641 (1 1)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2) exp(minus1198642 (1 1)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2) exp(minus
1198643(1 1)
119896119879)
(42)
Active and Passive Electronic Components 7
10 20 30 40 50 6002
022
024
026
028
03
032
034
036
038
04
Channel length (nm)
Pote
ntia
l (V
)
Proposed modelTCAD simulation
Figure 3 Constant electrostatic potential obtained from the analyt-ical solution of a gate-all-around silicon nanowire transistor is 03 Vfor different channel length with a height of119867 = 9 nm and channelwidth of 119882 = 9 nm TCAD simulation shows that the potential isconstant at 0296V
The impacts on the threshold voltage due to quantum effectsare acquired by using the following equation
119881TC = 119881TQ + Δ119881119879 (43)
Here Δ119881119879is the difference between the quantum threshold
voltage and the classical threshold voltage
4 Results and Discussion
Figure 3 shows the electrostatic potential of the proposed gateall around transistor and it is found to be constant value at03 V Continuously varying the 119899 and 119898 terms in (4) hasno impact on the potential as it remains constant along theinsulator boundaries This is totally in contrast to the resultsobtained in [15] where the potential is found to be linearlyvarying in the insulator boundaries The constant potentialhas to be deduced as the resultant of the gate voltage appliedsymmetrically across the four sides of the transistor TheTCAD simulation of the device shows that the electrostaticpotential is constant at 0296V The simulation results arefound to in acceptance with the TCAD results
Figure 4 represents the variation of total quantum inte-grated charge with the gate voltage Equation (39) is usedto obtain the integrated charge with only one energy leveland one series term It clearly shows that the decrease inthe film thickness leads to the increase in the quantumthreshold voltage which is actually due to the increase inenergy quantization of the transistor With the height andlength of the device being constant the width of the device isvaried and henceforth the variation of charge in accordancewith the gate voltage is illustrated in Figure 4
The variation of quantum threshold voltage with widthand height of the film at a channel length of 20 nm is
0 01 02 03 04 05 06 070
05
1
15
2
25
3
35
4times10minus15
Gate voltage (V)
Char
ge (c
mminus1)
W = 9e minus 9
W = 5e minus 9
W = 3e minus 9
TCAD simulation for W = 9e minus 9
TCAD simulation for W = 5e minus 9
TCAD simulation for W = 3e minus 9
Figure 4 Variation of quantum integrated charge at virtual sourcewith gate voltage for different film widths where the height andlength are119867 = 9 nm 119871 = 20 nm
2
8
times10minus9
times10minus9
Film height
2
44
66
8
10
1003
0305
031
0315
032
Film width
Qua
ntum
thre
shol
d vo
ltage
(V)
ProposedTCAD
Figure 5 Variation of quantum threshold voltage with film heightand width for channel length (119871 = 20 nm)
shown in Figure 5 The short channel effects tend to decreasealong with the energy quantisation and this can be furtherexplained as a result of increase in the effective band gap ofsilicon due to quantum effects The effect of confinementexpressed as the difference in the threshold voltage and itsvariation with the channel length 119871 is illustrated in Figure 6
8 Active and Passive Electronic Components
2 3 4 5 6 7 8 9 10
times10minus9
0014
0016
0018
002
0022
0024
0026
0028
003
0032
Film height
Proposed modelTCAD simulation
ΔVT
(mV
)
Figure 6 Variation of threshold voltagewith filmwidth for differentheight at 119871 = 20 nm
3 4 5 6 7 8
times10minus9
02
022
024
026
028
03
032
034
036
038
04
Film height
Thre
shol
d vo
ltage
(V)
ClassicalQuantum
TCAD simulation for classicalTCAD simulation for quantum
9
Figure 7 Variation of quantum and classical threshold voltage withfilm height Here 119871 = 20 nm and119882 = 9 nm
Themost important thing about this gate all around nanowiretransistor is that any change in one of the dimensions canbe nullified by proper tuning of other dimensions as thetransistor is symmetric about its height and width
Figure 7 shows the variation of the classical thresholdvoltage and quantum threshold voltage with the film height ata constant width of 9 nmThe value of the classical thresholdvoltage ranges from 027V to 029V for the correspondingchanges in the film height Similarly the quantum thresholdvoltage ranges from 03V to 031 V It shows that the devicehas a highly improved control over the threshold voltageTheTCAD results justify the simulation results
5 Conclusion
In this paper a quantum threshold voltage model for a GAAsilicon nanowire transistor is proposed by solving the 3DPoisson and Schrodinger equations Analytical expressionsfor potential and the inversion charge are expressed intheir closed forms The results show that the integratedcharge and the threshold voltage calculated in accordancewith the quantum effects of this proposed model are highlyimprovedThe future considerations include deriving the I-Vcharacteristics of the gate all around nanowire transistors andstudying the impact of scaling on various device parametersFinally to conclude this model provides an analytical anduseful way for the threshold voltage evaluations in gate allaround nanowire devices with a unified formalism employedin both classical and quantum mechanical approaches
References
[1] J Wang E Polizzi and M Lundstrom ldquoA computationalstudy of ballistic silicon nanowire transistorsrdquo in Proceedings ofthe IEEE International Electron Devices Meeting pp 695ndash698December 2003
[2] J P Colinge FINFETS and Other Multi-Gate TransistorsSpringer New York NY USA 2007
[3] D J Frank Y Taur andH-S PWong ldquoGeneralized scale lengthfor two-dimensional effects in MOSFETrsquosrdquo IEEE ElectronDevice Letters vol 19 no 10 pp 385ndash387 1998
[4] B Yu L Wang Y Yuan P M Asbeck and Y Taur ldquoScaling ofnanowire transistorsrdquo IEEE Transactions on Electron Devicesvol 55 no 11 pp 2846ndash2858 2008
[5] G D Sanders C J Stanton and Y C Chang ldquoTheory of trans-port in silicon quantum wiresrdquo Physical Review B vol 48 no15 pp 11067ndash11076 1993
[6] M-Y Shen and S-L Zhang ldquoBand gap of a silicon quantumwirerdquo Physics Letters A vol 176 no 3-4 pp 254ndash258 1993
[7] J P Colinge X Baie V Bayot and E Grivei ldquoQuantum-wireeffects in thin and narrow SOI MOSFETsrdquo in Proceedings of theIEEE International SOI Conference pp 66ndash67 October 1995
[8] C P Auth and J D Plummer ldquoScaling theory for cylindricalfully-depleted surrounding-gate MOSFETrsquosrdquo IEEE ElectronDevice Letters vol 18 no 2 pp 74ndash76 1997
[9] J-T Park and J-P Colinge ldquoMultiple-gate SOI MOSFETsdevice design guidelinesrdquo IEEE Transactions on ElectronDevices vol 49 no 12 pp 2222ndash2229 2002
[10] Z Ghoggali F DjeffalM A Abdi D Arar N NLakhdar and TTBendib ldquoAn analytical threshold voltagemodel for nanoscalerdquoin Proceedings of the 3rd International Design and TestWorkshop(IDT rsquo08) pp 93ndash97 December 2008
[11] Y-S Wu and P Su ldquoQuantum confinement effect in short-channel gate-all-around MOSFETs and its impact on the sensi-tivity of threshold voltage to process variationsrdquo in Proceedingsof the IEEE International SOI Conference October 2009
[12] C Te-Kuang ldquoA compact analytical threshold-voltage modelfor surrounding-gateMOSFETswith interface trapped chargesrdquoIEEE Electron Device Letters vol 31 no 8 pp 788ndash790 2010
[13] B Ray and SMahapatra ldquoModeling and analysis of body poten-tial of cylindrical gate-all-around nanowire transistorrdquo IEEETransactions on Electron Devices vol 55 no 9 pp 2409ndash24162008
Active and Passive Electronic Components 9
[14] L De Michielis L Selmi and A M Ionescu ldquoA quasi-analytical model for nanowire FETs with arbitrary polygonalcross sectionrdquo Solid-State Electronics vol 54 no 9 pp 929ndash9342010
[15] P R Kumar and S Mahapatra ldquoQuantum threshold voltagemodeling of short channel quad gate silicon nanowire transis-torrdquo IEEE Transactions onNanotechnology vol 10 no 1 pp 121ndash128 2011
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of
6 Active and Passive Electronic Components
E11
Ec
Ec
998779Ec
Ei119909i119910
Eco
Figure 2 Band diagram perpendicular to the gate-square wellpotential of a GAA silicon nanowire transistor
the square well potential since it is difficult to solve theSchrodinger equation to obtain the potential expressed in(29) The square well potential of a gate-all-around nanowiretransistor is shown in Figure 2 The quantum charge of thedevice is expressed as
119876 = sum
119894119909
sum
119894119910
119902int
infin
119864119894119909119894119910
1198731D119891 (119864) 119889119864 (34)
where1198731D is the 1D density-of-states and 119891(119864) is the Fermi-
Dirac distribution function 119864 is the energy of the electronwave The terms 119894
119909and 119894119910are positive natural numbers
In silicon six energy valleys are found to be present inits band structure (two lower energy valleys two middleenergy valleys and two higher energy valleys) If the thinfilm of device has equal height and width the two lowerenergy valleys and two middle energy valleys are combinedtogether to produce four lower energy valleys and the othertwo higher energy valleys remain in their own stateThus thecharge is given by
119876 = 119902sum
119894119909
sum
119894119910
radic(119898119911
2120587ℎ2)int
infin
119864119894119909119894119910
(119864 minus 119864119894119909119894119910)minus12
1 + 119890 ((119864 minus 119864119865) 119896119879)
119889119864 (35)
where 119898119911is the mass of the valley which is perpendicular to
the direction of quantizationThe Fermi energy level is muchlower than the conduction band energy in weak inversionregion Hence the charge equation can be approximated as
119876 = 119902radic(119898119911
2120587ℎ2)sum
119894119909
sum
119894119910
int
infin
119864119894119909119894119910
119890((119864119865minus119864)119896119879)
(119864 minus 119864119894119909119894119910)12119889119864 (36)
Using the Schrodinger equation the value of 119864119894119909119864119894119910is deter-
mined by the following formulation [5]
119864119894119909119894119910
= 119864co +ℎ21205872
2[1
119898119909
(119894119909
119868119909
)
2
+1
119898119910
(119894119910
119868119910
)
2
] (37)
where the conduction band energy is given as
119864co =119864119892
2minus 119902120595 (0 0 119911
119888) (38)
Using (36) and (37) the integrated charge can be obtained as
119876 = 119902radic(2119896119879119898
119905
ℎ2)sum
119894119909
sum
119894119910
exp(minus119864co + 1198641 (119894119909 119894119910)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2)sum
119894119909
sum
119894119910
exp(minus119864co + 1198642 (119894119909 119894119910)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2)sum
119894119909
sum
119894119910
exp(minus119864co + 1198643 (119894119909 119894119910)
119896119879)
(39)
where
1198641(119894119909 119894119910) =
ℎ21205872
2[1
1198981
(119894119909
119882)
2
+1
119898119905
(119894119910
119867)
2
]
1198642(119894119909 119894119910) =
ℎ21205872
2[1
119898119905
(119894119909
119882)
2
+1
1198981
(119894119910
119867)
2
]
1198643(119894119909 119894119910) =
ℎ21205872
2[1
119898119905
(119894119909
119882)
2
+1
119898119905
(119894119910
119867)
2
]
(40)
Here the 119898119905and 119898
1are the transverse and longitudinal
effective masses of the energy valleys of silicon The lengths119894119909and 119894119910carry distinct values contingent on the direction of
quantization Finally the quantum threshold voltage modelbecomes
119881TQ = (119864119892
2119902+ (
119896119879
119902) In(119876119879
120591)
+2120595bi12058811 (0 0) sinh((sum11119871)
2))
times (1 + 212058811(0 0) sinh(
(sum11119871)
2))
minus1
(41)
where
120591 = 119902radic(2119896119879119898
119905
ℎ2) exp(minus1198641 (1 1)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2) exp(minus1198642 (1 1)
119896119879)
+ 119902radic(2119896119879119898
119905
ℎ2) exp(minus
1198643(1 1)
119896119879)
(42)
Active and Passive Electronic Components 7
10 20 30 40 50 6002
022
024
026
028
03
032
034
036
038
04
Channel length (nm)
Pote
ntia
l (V
)
Proposed modelTCAD simulation
Figure 3 Constant electrostatic potential obtained from the analyt-ical solution of a gate-all-around silicon nanowire transistor is 03 Vfor different channel length with a height of119867 = 9 nm and channelwidth of 119882 = 9 nm TCAD simulation shows that the potential isconstant at 0296V
The impacts on the threshold voltage due to quantum effectsare acquired by using the following equation
119881TC = 119881TQ + Δ119881119879 (43)
Here Δ119881119879is the difference between the quantum threshold
voltage and the classical threshold voltage
4 Results and Discussion
Figure 3 shows the electrostatic potential of the proposed gateall around transistor and it is found to be constant value at03 V Continuously varying the 119899 and 119898 terms in (4) hasno impact on the potential as it remains constant along theinsulator boundaries This is totally in contrast to the resultsobtained in [15] where the potential is found to be linearlyvarying in the insulator boundaries The constant potentialhas to be deduced as the resultant of the gate voltage appliedsymmetrically across the four sides of the transistor TheTCAD simulation of the device shows that the electrostaticpotential is constant at 0296V The simulation results arefound to in acceptance with the TCAD results
Figure 4 represents the variation of total quantum inte-grated charge with the gate voltage Equation (39) is usedto obtain the integrated charge with only one energy leveland one series term It clearly shows that the decrease inthe film thickness leads to the increase in the quantumthreshold voltage which is actually due to the increase inenergy quantization of the transistor With the height andlength of the device being constant the width of the device isvaried and henceforth the variation of charge in accordancewith the gate voltage is illustrated in Figure 4
The variation of quantum threshold voltage with widthand height of the film at a channel length of 20 nm is
0 01 02 03 04 05 06 070
05
1
15
2
25
3
35
4times10minus15
Gate voltage (V)
Char
ge (c
mminus1)
W = 9e minus 9
W = 5e minus 9
W = 3e minus 9
TCAD simulation for W = 9e minus 9
TCAD simulation for W = 5e minus 9
TCAD simulation for W = 3e minus 9
Figure 4 Variation of quantum integrated charge at virtual sourcewith gate voltage for different film widths where the height andlength are119867 = 9 nm 119871 = 20 nm
2
8
times10minus9
times10minus9
Film height
2
44
66
8
10
1003
0305
031
0315
032
Film width
Qua
ntum
thre
shol
d vo
ltage
(V)
ProposedTCAD
Figure 5 Variation of quantum threshold voltage with film heightand width for channel length (119871 = 20 nm)
shown in Figure 5 The short channel effects tend to decreasealong with the energy quantisation and this can be furtherexplained as a result of increase in the effective band gap ofsilicon due to quantum effects The effect of confinementexpressed as the difference in the threshold voltage and itsvariation with the channel length 119871 is illustrated in Figure 6
8 Active and Passive Electronic Components
2 3 4 5 6 7 8 9 10
times10minus9
0014
0016
0018
002
0022
0024
0026
0028
003
0032
Film height
Proposed modelTCAD simulation
ΔVT
(mV
)
Figure 6 Variation of threshold voltagewith filmwidth for differentheight at 119871 = 20 nm
3 4 5 6 7 8
times10minus9
02
022
024
026
028
03
032
034
036
038
04
Film height
Thre
shol
d vo
ltage
(V)
ClassicalQuantum
TCAD simulation for classicalTCAD simulation for quantum
9
Figure 7 Variation of quantum and classical threshold voltage withfilm height Here 119871 = 20 nm and119882 = 9 nm
Themost important thing about this gate all around nanowiretransistor is that any change in one of the dimensions canbe nullified by proper tuning of other dimensions as thetransistor is symmetric about its height and width
Figure 7 shows the variation of the classical thresholdvoltage and quantum threshold voltage with the film height ata constant width of 9 nmThe value of the classical thresholdvoltage ranges from 027V to 029V for the correspondingchanges in the film height Similarly the quantum thresholdvoltage ranges from 03V to 031 V It shows that the devicehas a highly improved control over the threshold voltageTheTCAD results justify the simulation results
5 Conclusion
In this paper a quantum threshold voltage model for a GAAsilicon nanowire transistor is proposed by solving the 3DPoisson and Schrodinger equations Analytical expressionsfor potential and the inversion charge are expressed intheir closed forms The results show that the integratedcharge and the threshold voltage calculated in accordancewith the quantum effects of this proposed model are highlyimprovedThe future considerations include deriving the I-Vcharacteristics of the gate all around nanowire transistors andstudying the impact of scaling on various device parametersFinally to conclude this model provides an analytical anduseful way for the threshold voltage evaluations in gate allaround nanowire devices with a unified formalism employedin both classical and quantum mechanical approaches
References
[1] J Wang E Polizzi and M Lundstrom ldquoA computationalstudy of ballistic silicon nanowire transistorsrdquo in Proceedings ofthe IEEE International Electron Devices Meeting pp 695ndash698December 2003
[2] J P Colinge FINFETS and Other Multi-Gate TransistorsSpringer New York NY USA 2007
[3] D J Frank Y Taur andH-S PWong ldquoGeneralized scale lengthfor two-dimensional effects in MOSFETrsquosrdquo IEEE ElectronDevice Letters vol 19 no 10 pp 385ndash387 1998
[4] B Yu L Wang Y Yuan P M Asbeck and Y Taur ldquoScaling ofnanowire transistorsrdquo IEEE Transactions on Electron Devicesvol 55 no 11 pp 2846ndash2858 2008
[5] G D Sanders C J Stanton and Y C Chang ldquoTheory of trans-port in silicon quantum wiresrdquo Physical Review B vol 48 no15 pp 11067ndash11076 1993
[6] M-Y Shen and S-L Zhang ldquoBand gap of a silicon quantumwirerdquo Physics Letters A vol 176 no 3-4 pp 254ndash258 1993
[7] J P Colinge X Baie V Bayot and E Grivei ldquoQuantum-wireeffects in thin and narrow SOI MOSFETsrdquo in Proceedings of theIEEE International SOI Conference pp 66ndash67 October 1995
[8] C P Auth and J D Plummer ldquoScaling theory for cylindricalfully-depleted surrounding-gate MOSFETrsquosrdquo IEEE ElectronDevice Letters vol 18 no 2 pp 74ndash76 1997
[9] J-T Park and J-P Colinge ldquoMultiple-gate SOI MOSFETsdevice design guidelinesrdquo IEEE Transactions on ElectronDevices vol 49 no 12 pp 2222ndash2229 2002
[10] Z Ghoggali F DjeffalM A Abdi D Arar N NLakhdar and TTBendib ldquoAn analytical threshold voltagemodel for nanoscalerdquoin Proceedings of the 3rd International Design and TestWorkshop(IDT rsquo08) pp 93ndash97 December 2008
[11] Y-S Wu and P Su ldquoQuantum confinement effect in short-channel gate-all-around MOSFETs and its impact on the sensi-tivity of threshold voltage to process variationsrdquo in Proceedingsof the IEEE International SOI Conference October 2009
[12] C Te-Kuang ldquoA compact analytical threshold-voltage modelfor surrounding-gateMOSFETswith interface trapped chargesrdquoIEEE Electron Device Letters vol 31 no 8 pp 788ndash790 2010
[13] B Ray and SMahapatra ldquoModeling and analysis of body poten-tial of cylindrical gate-all-around nanowire transistorrdquo IEEETransactions on Electron Devices vol 55 no 9 pp 2409ndash24162008
Active and Passive Electronic Components 9
[14] L De Michielis L Selmi and A M Ionescu ldquoA quasi-analytical model for nanowire FETs with arbitrary polygonalcross sectionrdquo Solid-State Electronics vol 54 no 9 pp 929ndash9342010
[15] P R Kumar and S Mahapatra ldquoQuantum threshold voltagemodeling of short channel quad gate silicon nanowire transis-torrdquo IEEE Transactions onNanotechnology vol 10 no 1 pp 121ndash128 2011
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of
Active and Passive Electronic Components 7
10 20 30 40 50 6002
022
024
026
028
03
032
034
036
038
04
Channel length (nm)
Pote
ntia
l (V
)
Proposed modelTCAD simulation
Figure 3 Constant electrostatic potential obtained from the analyt-ical solution of a gate-all-around silicon nanowire transistor is 03 Vfor different channel length with a height of119867 = 9 nm and channelwidth of 119882 = 9 nm TCAD simulation shows that the potential isconstant at 0296V
The impacts on the threshold voltage due to quantum effectsare acquired by using the following equation
119881TC = 119881TQ + Δ119881119879 (43)
Here Δ119881119879is the difference between the quantum threshold
voltage and the classical threshold voltage
4 Results and Discussion
Figure 3 shows the electrostatic potential of the proposed gateall around transistor and it is found to be constant value at03 V Continuously varying the 119899 and 119898 terms in (4) hasno impact on the potential as it remains constant along theinsulator boundaries This is totally in contrast to the resultsobtained in [15] where the potential is found to be linearlyvarying in the insulator boundaries The constant potentialhas to be deduced as the resultant of the gate voltage appliedsymmetrically across the four sides of the transistor TheTCAD simulation of the device shows that the electrostaticpotential is constant at 0296V The simulation results arefound to in acceptance with the TCAD results
Figure 4 represents the variation of total quantum inte-grated charge with the gate voltage Equation (39) is usedto obtain the integrated charge with only one energy leveland one series term It clearly shows that the decrease inthe film thickness leads to the increase in the quantumthreshold voltage which is actually due to the increase inenergy quantization of the transistor With the height andlength of the device being constant the width of the device isvaried and henceforth the variation of charge in accordancewith the gate voltage is illustrated in Figure 4
The variation of quantum threshold voltage with widthand height of the film at a channel length of 20 nm is
0 01 02 03 04 05 06 070
05
1
15
2
25
3
35
4times10minus15
Gate voltage (V)
Char
ge (c
mminus1)
W = 9e minus 9
W = 5e minus 9
W = 3e minus 9
TCAD simulation for W = 9e minus 9
TCAD simulation for W = 5e minus 9
TCAD simulation for W = 3e minus 9
Figure 4 Variation of quantum integrated charge at virtual sourcewith gate voltage for different film widths where the height andlength are119867 = 9 nm 119871 = 20 nm
2
8
times10minus9
times10minus9
Film height
2
44
66
8
10
1003
0305
031
0315
032
Film width
Qua
ntum
thre
shol
d vo
ltage
(V)
ProposedTCAD
Figure 5 Variation of quantum threshold voltage with film heightand width for channel length (119871 = 20 nm)
shown in Figure 5 The short channel effects tend to decreasealong with the energy quantisation and this can be furtherexplained as a result of increase in the effective band gap ofsilicon due to quantum effects The effect of confinementexpressed as the difference in the threshold voltage and itsvariation with the channel length 119871 is illustrated in Figure 6
8 Active and Passive Electronic Components
2 3 4 5 6 7 8 9 10
times10minus9
0014
0016
0018
002
0022
0024
0026
0028
003
0032
Film height
Proposed modelTCAD simulation
ΔVT
(mV
)
Figure 6 Variation of threshold voltagewith filmwidth for differentheight at 119871 = 20 nm
3 4 5 6 7 8
times10minus9
02
022
024
026
028
03
032
034
036
038
04
Film height
Thre
shol
d vo
ltage
(V)
ClassicalQuantum
TCAD simulation for classicalTCAD simulation for quantum
9
Figure 7 Variation of quantum and classical threshold voltage withfilm height Here 119871 = 20 nm and119882 = 9 nm
Themost important thing about this gate all around nanowiretransistor is that any change in one of the dimensions canbe nullified by proper tuning of other dimensions as thetransistor is symmetric about its height and width
Figure 7 shows the variation of the classical thresholdvoltage and quantum threshold voltage with the film height ata constant width of 9 nmThe value of the classical thresholdvoltage ranges from 027V to 029V for the correspondingchanges in the film height Similarly the quantum thresholdvoltage ranges from 03V to 031 V It shows that the devicehas a highly improved control over the threshold voltageTheTCAD results justify the simulation results
5 Conclusion
In this paper a quantum threshold voltage model for a GAAsilicon nanowire transistor is proposed by solving the 3DPoisson and Schrodinger equations Analytical expressionsfor potential and the inversion charge are expressed intheir closed forms The results show that the integratedcharge and the threshold voltage calculated in accordancewith the quantum effects of this proposed model are highlyimprovedThe future considerations include deriving the I-Vcharacteristics of the gate all around nanowire transistors andstudying the impact of scaling on various device parametersFinally to conclude this model provides an analytical anduseful way for the threshold voltage evaluations in gate allaround nanowire devices with a unified formalism employedin both classical and quantum mechanical approaches
References
[1] J Wang E Polizzi and M Lundstrom ldquoA computationalstudy of ballistic silicon nanowire transistorsrdquo in Proceedings ofthe IEEE International Electron Devices Meeting pp 695ndash698December 2003
[2] J P Colinge FINFETS and Other Multi-Gate TransistorsSpringer New York NY USA 2007
[3] D J Frank Y Taur andH-S PWong ldquoGeneralized scale lengthfor two-dimensional effects in MOSFETrsquosrdquo IEEE ElectronDevice Letters vol 19 no 10 pp 385ndash387 1998
[4] B Yu L Wang Y Yuan P M Asbeck and Y Taur ldquoScaling ofnanowire transistorsrdquo IEEE Transactions on Electron Devicesvol 55 no 11 pp 2846ndash2858 2008
[5] G D Sanders C J Stanton and Y C Chang ldquoTheory of trans-port in silicon quantum wiresrdquo Physical Review B vol 48 no15 pp 11067ndash11076 1993
[6] M-Y Shen and S-L Zhang ldquoBand gap of a silicon quantumwirerdquo Physics Letters A vol 176 no 3-4 pp 254ndash258 1993
[7] J P Colinge X Baie V Bayot and E Grivei ldquoQuantum-wireeffects in thin and narrow SOI MOSFETsrdquo in Proceedings of theIEEE International SOI Conference pp 66ndash67 October 1995
[8] C P Auth and J D Plummer ldquoScaling theory for cylindricalfully-depleted surrounding-gate MOSFETrsquosrdquo IEEE ElectronDevice Letters vol 18 no 2 pp 74ndash76 1997
[9] J-T Park and J-P Colinge ldquoMultiple-gate SOI MOSFETsdevice design guidelinesrdquo IEEE Transactions on ElectronDevices vol 49 no 12 pp 2222ndash2229 2002
[10] Z Ghoggali F DjeffalM A Abdi D Arar N NLakhdar and TTBendib ldquoAn analytical threshold voltagemodel for nanoscalerdquoin Proceedings of the 3rd International Design and TestWorkshop(IDT rsquo08) pp 93ndash97 December 2008
[11] Y-S Wu and P Su ldquoQuantum confinement effect in short-channel gate-all-around MOSFETs and its impact on the sensi-tivity of threshold voltage to process variationsrdquo in Proceedingsof the IEEE International SOI Conference October 2009
[12] C Te-Kuang ldquoA compact analytical threshold-voltage modelfor surrounding-gateMOSFETswith interface trapped chargesrdquoIEEE Electron Device Letters vol 31 no 8 pp 788ndash790 2010
[13] B Ray and SMahapatra ldquoModeling and analysis of body poten-tial of cylindrical gate-all-around nanowire transistorrdquo IEEETransactions on Electron Devices vol 55 no 9 pp 2409ndash24162008
Active and Passive Electronic Components 9
[14] L De Michielis L Selmi and A M Ionescu ldquoA quasi-analytical model for nanowire FETs with arbitrary polygonalcross sectionrdquo Solid-State Electronics vol 54 no 9 pp 929ndash9342010
[15] P R Kumar and S Mahapatra ldquoQuantum threshold voltagemodeling of short channel quad gate silicon nanowire transis-torrdquo IEEE Transactions onNanotechnology vol 10 no 1 pp 121ndash128 2011
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of
8 Active and Passive Electronic Components
2 3 4 5 6 7 8 9 10
times10minus9
0014
0016
0018
002
0022
0024
0026
0028
003
0032
Film height
Proposed modelTCAD simulation
ΔVT
(mV
)
Figure 6 Variation of threshold voltagewith filmwidth for differentheight at 119871 = 20 nm
3 4 5 6 7 8
times10minus9
02
022
024
026
028
03
032
034
036
038
04
Film height
Thre
shol
d vo
ltage
(V)
ClassicalQuantum
TCAD simulation for classicalTCAD simulation for quantum
9
Figure 7 Variation of quantum and classical threshold voltage withfilm height Here 119871 = 20 nm and119882 = 9 nm
Themost important thing about this gate all around nanowiretransistor is that any change in one of the dimensions canbe nullified by proper tuning of other dimensions as thetransistor is symmetric about its height and width
Figure 7 shows the variation of the classical thresholdvoltage and quantum threshold voltage with the film height ata constant width of 9 nmThe value of the classical thresholdvoltage ranges from 027V to 029V for the correspondingchanges in the film height Similarly the quantum thresholdvoltage ranges from 03V to 031 V It shows that the devicehas a highly improved control over the threshold voltageTheTCAD results justify the simulation results
5 Conclusion
In this paper a quantum threshold voltage model for a GAAsilicon nanowire transistor is proposed by solving the 3DPoisson and Schrodinger equations Analytical expressionsfor potential and the inversion charge are expressed intheir closed forms The results show that the integratedcharge and the threshold voltage calculated in accordancewith the quantum effects of this proposed model are highlyimprovedThe future considerations include deriving the I-Vcharacteristics of the gate all around nanowire transistors andstudying the impact of scaling on various device parametersFinally to conclude this model provides an analytical anduseful way for the threshold voltage evaluations in gate allaround nanowire devices with a unified formalism employedin both classical and quantum mechanical approaches
References
[1] J Wang E Polizzi and M Lundstrom ldquoA computationalstudy of ballistic silicon nanowire transistorsrdquo in Proceedings ofthe IEEE International Electron Devices Meeting pp 695ndash698December 2003
[2] J P Colinge FINFETS and Other Multi-Gate TransistorsSpringer New York NY USA 2007
[3] D J Frank Y Taur andH-S PWong ldquoGeneralized scale lengthfor two-dimensional effects in MOSFETrsquosrdquo IEEE ElectronDevice Letters vol 19 no 10 pp 385ndash387 1998
[4] B Yu L Wang Y Yuan P M Asbeck and Y Taur ldquoScaling ofnanowire transistorsrdquo IEEE Transactions on Electron Devicesvol 55 no 11 pp 2846ndash2858 2008
[5] G D Sanders C J Stanton and Y C Chang ldquoTheory of trans-port in silicon quantum wiresrdquo Physical Review B vol 48 no15 pp 11067ndash11076 1993
[6] M-Y Shen and S-L Zhang ldquoBand gap of a silicon quantumwirerdquo Physics Letters A vol 176 no 3-4 pp 254ndash258 1993
[7] J P Colinge X Baie V Bayot and E Grivei ldquoQuantum-wireeffects in thin and narrow SOI MOSFETsrdquo in Proceedings of theIEEE International SOI Conference pp 66ndash67 October 1995
[8] C P Auth and J D Plummer ldquoScaling theory for cylindricalfully-depleted surrounding-gate MOSFETrsquosrdquo IEEE ElectronDevice Letters vol 18 no 2 pp 74ndash76 1997
[9] J-T Park and J-P Colinge ldquoMultiple-gate SOI MOSFETsdevice design guidelinesrdquo IEEE Transactions on ElectronDevices vol 49 no 12 pp 2222ndash2229 2002
[10] Z Ghoggali F DjeffalM A Abdi D Arar N NLakhdar and TTBendib ldquoAn analytical threshold voltagemodel for nanoscalerdquoin Proceedings of the 3rd International Design and TestWorkshop(IDT rsquo08) pp 93ndash97 December 2008
[11] Y-S Wu and P Su ldquoQuantum confinement effect in short-channel gate-all-around MOSFETs and its impact on the sensi-tivity of threshold voltage to process variationsrdquo in Proceedingsof the IEEE International SOI Conference October 2009
[12] C Te-Kuang ldquoA compact analytical threshold-voltage modelfor surrounding-gateMOSFETswith interface trapped chargesrdquoIEEE Electron Device Letters vol 31 no 8 pp 788ndash790 2010
[13] B Ray and SMahapatra ldquoModeling and analysis of body poten-tial of cylindrical gate-all-around nanowire transistorrdquo IEEETransactions on Electron Devices vol 55 no 9 pp 2409ndash24162008
Active and Passive Electronic Components 9
[14] L De Michielis L Selmi and A M Ionescu ldquoA quasi-analytical model for nanowire FETs with arbitrary polygonalcross sectionrdquo Solid-State Electronics vol 54 no 9 pp 929ndash9342010
[15] P R Kumar and S Mahapatra ldquoQuantum threshold voltagemodeling of short channel quad gate silicon nanowire transis-torrdquo IEEE Transactions onNanotechnology vol 10 no 1 pp 121ndash128 2011
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of
Active and Passive Electronic Components 9
[14] L De Michielis L Selmi and A M Ionescu ldquoA quasi-analytical model for nanowire FETs with arbitrary polygonalcross sectionrdquo Solid-State Electronics vol 54 no 9 pp 929ndash9342010
[15] P R Kumar and S Mahapatra ldquoQuantum threshold voltagemodeling of short channel quad gate silicon nanowire transis-torrdquo IEEE Transactions onNanotechnology vol 10 no 1 pp 121ndash128 2011
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporation httpwwwhindawicom
Journal ofEngineeringVolume 2014
Submit your manuscripts athttpwwwhindawicom
VLSI Design
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation httpwwwhindawicom
Volume 2014
The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
DistributedSensor Networks
International Journal of