8.5. CHARGE CONTROL MODEL FOR THE MODFET 387
assumeΔnpar=
∫ 0
−dn(AlGaAs)dx→^0. From equation 8.5.9, we havens(VG)=(
Nd−Nd^0)
dδ−/e[VG−(φb−ΔEc/e)]
D(8.5.22)
From this expression, we can solve forΔΔVnGs:
Δns
ΔVG=
1
D
[
−
ΔN^0 d
ΔVG·dδ+e]
(8.5.23)
Inserting this into equation 8.5.11 gives us the following expression for the modulation effi-
ciency:
ME(VG)=1−e
ΔNd^0
ΔVG·dδ=1−Cp,ef f
Cp(8.5.24)
where
Cp,ef f=eΔNd^0
ΔVGCp=dδAgain, if the change in charge density in the AlGaAsΔNd^0 is negligible, thenCp,ef f→ 0 and
ME→ 1. Finding an expression forCp,ef frequires solving for the conduction band occupancy
in the AlGaAs as a function ofVG. This must be done numerically and is left as a problem for
the reader.
Example 8.1Consider ann-type GaAs/Al 0. 3 Ga 0. 7 As MODFET at 300 K with the
following parameters:Schottky barrier height, φb =0.9V
Barrier doping, Nd =10^18 cm−^3
Conduction band discontinuity, ΔEc =0.24 eV
Dielectric constant of the barrier, b =12. 2
Spacer layer thickness, ds =30A ̊
Barrier thickness, d = 350A ̊Calculate the 2DEG concentration atVG=0andVG=−0.5V.
The parameterVp 2 of this structure is given byVp 2 =eNd
b(d−ds)^2 =(
1. 6 × 10 −^19 C
)(
1018 cm−^3)(
320 × 10 −^8 cm) 2
12 .2(8. 85 × 10 −^14 F/cm)
=1.52 VThe threshold voltageVof fis given byVof f=0. 9 − 0. 24 − 1 .52 =− 0 .86 V