Advanced Field Effect Transistor (FET) Devices: Lecture 12b
Advanced Field Effect Transistor (FET) Devices: Lecture 12b
Advanced Field Effect Transistor (FET) Devices: Lecture 12b
Reading:
(Cont’d) Notes and Anderson2 Chapter 8.11
E2
E1
The channel is a 2D electron gas contained in the triangular quantum well created by the ∆Ec of the AlGaAs-
GaAs heterojunction. Also known as a MODFET, HFET, TEGFET, SDHT.
Advantages:
• The mobility is not degraded by surfaces or interfaces like in a MOSFET. The AlGaAs-GaAs
interface is lattice matched and thus has ZERO interface states.
• In a 2D electron gas (channel) the electron scattering is reduced by a factor of approximately x2/3
making the mobility higher than in bulk. Some of this increase is due to separation of the dopants
from the channel (modulation doping) and a small additional enhancement is due to the quantum
2D nature of the channel.
• Example: Bulk GaN µ~1000-1200 cm2/V-Sec but in an AlN/GaN channel, µ~2000 cm2/V-Sec .
• The device is very fast because the channel thickness can be very precisely controlled (minimized)
making the channel easy to deplete with a small gate voltage.
Georgia Tech ECE 3080 - Dr. Alan Doolittle
Advanced Devices: FET Devices
Pseudomorphic High Electron Mobility Transistor (PHEMT)
InGaAs InGaAs
GaAs GaAs
or InP or InP
The channel is a a strained lattice 2D electron gas contained in the triangular quantum well created by the
∆Ec the AlGaAs-InGaAs heterojunction.
Advantages:
• Higher channel charge (higher conductivity) due to a higher ∆Ec for InGaAs vs GaAs Channels
• Higher saturation velocity of InGaAs results in higher frequency operation
• Lower Noise.
• Higher mobility of InGaAs results in smaller parasitic resistances and higher low field electron
velocity. InGaAs µ~12,000 cm2/V-Sec. In is typically 15-20% and Channels are typically 11-15
nm.
• The device is very fast because the channel thickness can be very precisely controlled (minimized)
making the channel easy to deplete with a small gate voltage.
Cross section of a MESFET under small VDS bias and the corresponding energy band diagrams at the source
end and drain end of the gate.
The electron potential energy, (EC) along the channel of a MESFET for VDS < VDSsat. The channel
thickness decreases with increasing distance along the channel.
The same MESFET VDS > VDSsat. At the source, the diagram is the same as before. At the drain,
however, the two depletion regions overlap pinching off the channel.
Georgia Tech ECE 3080 - Dr. Alan Doolittle
Advanced Devices: FET Devices
Junction FET (JFET)
The energy band diagram, (EC), for VDS = 0
p- substrate
JFET: The yellow shaded areas represent the depletion regions. When VDS > 0 (but not yet in saturation)
the depletion region at the drain end increases, narrowing the channel and increasing the channel
resistance.
Georgia Tech ECE 3080 - Dr. Alan Doolittle
Advanced Devices: FET Devices
Junction FET (JFET)
p- substrate
p- substrate
JFET: For higher Gate voltages, the channel can be pinched off. In some JFETs, there is a symmetric
gate-channel region even creating a ring and cylindrical channel.
N+ N+
Since large electric fields (high voltages) must be supported, thick low doped regions must be used. Thus, most power mosfets use a
substantial portion of the substrate as a “drift region” (Green N- region) separate from the “channel region” that controls current flow.
Georgia Tech ECE 3080 - Dr. Alan Doolittle
Advanced Devices: FET Devices
Insulated Gate Bipolar Junction Transistor
The IGBT is very similar to a n-channel vertical power MOSFET except the n+ drain is replaced
with a p+ collector, creating a vertical PNP bipolar junction transistor.
The IGBT is used in primarily applications above 600 V blocking rating whereas power
MOSFETs are used below 600V.
IGBTs can handle thousands of volts (off state) with minimal leakage currents and conduct 1000s
of amps (on state) with typically a few volts to 10s of volts forward voltage drop (on state).
Georgia Tech ECE 3080 - Dr. Alan Doolittle
Advanced Devices: FET Devices
Insulated Gate Bipolar Junction Transistor
The additional p+ region creates a cascaded connection of a PNP bipolar junction transistor with the surface n-channel
MOSFET.
This connection results in a significantly lower forward voltage drop compared to a conventional MOSFET. By injecting
minority carriers (holes) from the collector p+ region into the n- drift region during forward conduction, the resistance of
the n- drift region is considerably reduced. This minority carrier injection in power devices is called conductivity
modulation. When in the off state, this conductivity modulation does not occur allowing the low doped drift region to
support very high voltages in the off state.
Since this is a minority carrier device, this results in longer switching time (slower speed) and hence higher switching
losses compared to a power MOSFET.
Georgia Tech ECE 3080 - Dr. Alan Doolittle
Advanced Devices: FET Devices
Insulated Gate Bipolar Junction Transistor
Transistor /
Two Transistor Thyrister
Model Model
Georgia Tech Figures after Dimitrijev, Understanding Semiconductor Devices ECE 3080 - Dr. Alan Doolittle