Since their first discovery and fabrication in 1991,
CNTs have received considerable attention because of
the prospect of new fundamental science
and many potential applications.
Stretching And confined deformation
Strain of less than 1% results in
the CNT changing from metal to
Twisting and bending
The encapsulated fullerenes can rotate
freely in the space of a (10, 10) tube at
room temperature, and the rotation of
fullerenes will affect C60@(10, 10)
peapod electronic properties significantly;
generally, orientational disorderwill
remove the sharp features of the average
density of states (DOS). However, the
rotation of fullerenes cannot induce a
metal–insulator transition. Unlike the
multicarrier metallic C60@(10, 10)
peapod, the C60@(17, 0) peapod is a
semiconductor, and the effects of the
encapsulated fullerenes on tube valence
bands and conduction bands are
The distances between the centres of the
fullerenes are 0.984 and 1.278 nm for the
C60@(10, 10) peapod and C60@(17, 0)
J. Chen, and J. Dong, J. Phys. Condens. Matter, 16, 1401 (2
Semiconducting behavior in nanotubes was first reported by Tans et al. in 1998.
Fig. 5 shows a measurement of the
conductance of a semiconducting SWNT
as the gate voltage applied to the
conducting substrate is varied. The tube
conducts at negative Vg and turns off with
a positive Vg. The resistance change
between the on and off state is many orders
of magnitude. This device behavior is
analogous to a p-type metal–oxide–
semiconductor field-effect transistor
(MOSFET), with the nanotube replacing
Si as the semiconductor. At large positive
gate voltages, n-type conductance is
sometimes observed, especially in
McEuen et al., IEEE Trans. Nanotechn., 1, 78 (2002)
It is shown that, by appropriate work function engineering of the source, drain and gate contacts to the
device, the following desirable properties should be realizable: a sub-threshold slope close to the
thermionic limit; a conductance close to the interfacial limit; an ON/OFF ratio of around 1000; ON current
and transconductance close to the low-quantum-capacitance limit.
Semiconducting nanotubes are typically p-type at Vg=0 because of the contacts and also because chemical species,
particularly oxygen, adsorb on the tube and act as weak p-type dopants. Experiments have shown that changing a
tube’s chemical environment can change this doping level—shifting the voltage at which the device turns on by a
significant amount. This has spurred interest in nanotubes as chemical sensors.
Adsorbate doping can be a problem for reproducible device behavior, however.
Controlled chemical doping of tubes, both p- and n-type, has been accomplished in a number of ways. N-type doping
was first done using alkali metals that donate electrons to the tube. This has been used to create n-type transistors, p-n
junctions, and p-n-p devices. Alkali metals are not air-stable, however, so other techniques are under development,
such as using polymers for charge-transfer doping
Scattering sites in nanotubes:
I–V characteristics at different Vgs for a p-
type SWNT FET utilizing an electrolyte gate
in order to improve gate efficiency.
Maximum transconductance dI/dVg=20uA/V at Vg=-0.9V. Implying a mean-free path of approx. 700 nm.
Normalizing this to the device width of ~2nm: 10mS/um. McEuen et al., IEEE Trans. Nanotechn., 1, 78 (2002)
Bottom - gated CNT FET
Calculated conductance vs gate voltage at room temperature, varying
(a) the work function of the metal electrode, and (b) doping of the NT.
In (a) the work function of the metal electrode is changed by -0.2 eV (red
dashed), -0.1 eV (orange dashed), 0 eV (green), +0.1 eV (light blue),
and +0.2 eV (blue), from left to right, respectively. In (b) the doping
atomic fraction is n-type 0.001 (red), 0.0005 (orange), and 0.0001
(green), and p-type 0.0001 (blue dashed), from left to right, respectively.
Thus the gate field induces switching by modulating the contact resistance (the
junction barriers). Oxygen adsorption at the junctions modifies the barriers
(i.e. the local band-bending of the CNT) and affects the injection of carriers
(holes or electrons).
The inverse subthreshold slope, which is a measure of the
efficiency of the gate field in turning on the device,
decreases with a decrease in gate oxide thickness. This
behavior cannot be explained by conventional field-effect
transistor models, and has in fact been shown to be a result
of the presence of Schottky barriers at the metal/nanotube
interface at the source and drain.
There is a clear difference in the inverse subthreshold slope for the case of sweeping all gate segments together (S=400
mV/dec) versus sweeping only the inner segments (S=180 mV/dec). We attribute the observed change in S to a change from
Schottky barrier modulation to bulk switching. (b) shows linear plots of the subthreshold portion (where the current is
dominated by carrier density) of the transfer characteristics when the inner gate segments are swept together or separately.
The current nearly identical, despite the fact that the effective gate lengths differ by a factor of 1.6 . This is in contrast to the
expected behavior of diffusive transport, where the current varies inversely with the gate length.
Calculated output characteristics of the symmetric
(dashed lines) and the asymmetric (solid lines) CNFET.
We have introduced nanotemplate to control selective growth,
length and diameter of CNT. Ohmic contact of the CNT/metal
interface was formed by rapid thermal annealing (RTA).
Diameter control and surface modification of CNT open the
possibility to energy band gap modulation.
Diode-like rectifying behavior was observed in a CNx /C multiwalled
nanotube due to its being one half doped with nitrogen.
FETs based on an individual CNx /C nanotube were fabricated by focused
ion-beam technology. The nanotube transistors exhibited n-type
semiconductor characteristics, and the conductance of nanotube FETs can
be modulated more than four orders of magnitude at room temperature.
The electron mobility of a CNx /C NT FET estimated from its
transconductance was as high as 3840 cm2/Vs. The n-type gate
modulation could be explained as due the effect of bending of the valence
band in the Schottky-barrier junction.
CNTs doped with fullerenes inside nanotubes (so-called peapods) are interesting materials for novel CNT FET
channels. Transport properties of various peapods such as C60-, Gd@C82-, and Ti2@C92-peapods have been
studied by measuring FET I-V characteristics. Metallofulleren peapod FETs exhibited ambipolar behavior both
p- and n-type characteristics by changing the gate voltage, whereas C60-peapod FETs showed unipolar p-type
characteristics similar to the FETs of intact single-walled nanotubes. This difference can be explained in terms
of a bandgap narrowing of the single-walled nanotube due to the incorporation of metallofullerenes. The
bandgap narrowing was large in the peapods of metallofullerene, where more electrons are transferred from
encapsulated metal atoms to the fullerene cages.
The entrapped fullerene molecules are capable of
modifying the electronic structure of the host tube. It is,
therefore, anticipated that the encapsulation of fullerene
molecules can play a role in band gap engineering in
nanotubes and hence that peapods may generate
conceptually novel molecular devices.
Schematic illustration of elastic strain distributed around the site
of metallofullerenes in a small-diameter nanotube peapod and the
corresponding changes in conduction and valence band edges.
Charge transport in a partially filled peapod FET in
“metal-on-top” setup. (a) Transfer characteristics at
various temperatures. Data were taken at Vds = 0.3
Current vs. voltage characteristics of an all-carbon transistor
with semiconducting nanotube as channel, with different voltag
at the carbon gate. The back gate is kept at 0 V. The measurem
were carried out at 4 K.
Ambipolar conduction leads to a large
leakage current that exponentially
increases with the power supply voltage,
especially when the tube diameter is
large. An asymmetric gate oxide SB
CNTFET has been proposed as a means
of suppressing ambipolar conduction. SB
CNTFETs of any type, however, will likely
suffer from the need to place the gate
electrode close to the source (which
increases parasitic capacitance) and
metal-induced gap states, which
increase source to drain tunneling and
limit the minimum channel length.
The band profile of the SB CNTFET at the minimal leakage bias (VG=0V) for VD=0.6V. The band profile of
the MOS CNTFET when the source-drain current is low. (VD=0.6V and VG=-0.3V). The channel is a (13,0)
Id vs. Vd characteristics at VG = 0.4V for the MOS Id vs. Vg characteristics at Vd = 0.4V for the zero
CNTFET (the solid line) and the SB CNTFETs (the barrier SBFET and the MOS CNTFET. The gated
dashed lines). The off-current of all transistors (defined channel of both transistors is a 5nm-long, intrinsic
at Vd=0.4V and Vg=0) was set at 0.01µA by adjusting (13, 0) CNT.
the flat band voltage for each transistor. For the SB
CNTFETs, three barrier heights we simulated. The
channel is a (13,0) nanotube, which results in a diameter
of d≈ 1 nm, and a bandgap of Eg≈ 0.83 eV.
By eliminating the Schottky barrier between the source and channel, the transistor will be capable
of delivering more on-current. The leakage current of such devices will be controlled by the full
bandgap of CNTs (instead of half of the bandgap for SB CNTFETs) and band-to-band tunneling.
These factors will depend on the diameter of nanotubes and the power supply voltage.