transistors, for which SS values near the the-
oretical limit of 0.06 V/dec have been reported
( 3 , 4 , 23 – 25 ). Although the suspended nano-
tube transistors fabricated inside the TEM are
ideal for in situ manipulation, characteriza-
tion, and measurements, such configuration is
not a realistic way to produce CNT field-effect
transistors for large-scale applications.
We next compared the transfer character-
istics of the transistors with channel lengths
of ~19.8, ~2.8, and ~0.55 nm (Fig. 2G and fig. S7).
Although theIONcurrent was similar (~1mA)
for the three devices with aVSDof 0.5 V,IOFF
was <0.01 nA for the 19.8-nm-channel device
and ~ 0.2 nA for the 2.8-nm-channel transistor
(Fig. 2H). The device with the channel length
of 0.55 nm could not be switched off byVG, an
effect that could have been caused by quantum
tunneling at such a short distance. We observed
resonant conductance in the gap region for the
2.8-nm-channel transistor, indicating the tran-
sition from a semiconductor with continuous
energy bands to molecules with discrete energy
states. Thus, as the size of the nanotube tran-
sistors was reduced, quantum effect and tun-
neling behavior became more pronounced.
The nanometer length of the chirality-
transformed CNT channel creates a quan-
tum mechanically confined region in the axial
direction in addition to the circumferential
direction. In Fig. 3A, a TEM image and sche-
matic show a SWCNT segment with the length
of ~8.1 nm, with small kinks marked by arrows.
Oscillations of conductance at the ON state
were observed in theVG-dependent measure-
ments (Fig. 3B). The oscillation peaks grad-
ually smeared out asVSDwas increased from
0.1 to 1.0 V, an effect likely caused by a screen-
ing effect of the largerISDand charge density.
WhenVSDwas again decreased, the oscillation
peaks recovered at the same positions. Such a
phenomenon can be understood as Fabry-Pérot
interference ( 26 , 27 ).
The diameter,d, of the narrow section was
0.86 nm, which would result in a bandgap of
0.81 eV (EG= 0.7 eV/d). With aVGgap of 12.5 V,
the gate efficiency is 0.065ða¼EVGGÞ. With a
period of conductanceVPof 3.8 V, the energy
spacing is calculated as 0.25 eV (DE=aVP).
The length of the oscillator is calculated to
be 9.1 nm (L¼ 2 hvDFE, wherehis Planck’s con-
stant andvFis the Fermi velocity), near the
measured length (~8.1 nm). We calculated
the propagation process of an electron wave
in a nanotube junction using wave packet
(WP) dynamics. Snapshots of WP propaga-
tion through a (22,4)/(9,4)/(22,4) junction
(Fig. 3C and movie S1) revealed the transmit-
tance, reflection, and interference of the elec-
tron waves. Spatial and spectral distributions of
the wave functions showed the localized states
and energy-dependent channels (fig. S8). The
observation of quantum interference at room
temperature resulted from the large energy
spacing from the short chirality-altered seg-
ment and from the reduced electron scattering
at the covalently bonded nanotube junctions.
The detailed chiral transition processes in a
triple-wall CNT (TWCNT) were analyzed by
nanobeam electron diffraction (NBED), with
intermediate states shown in Fig. 4, A to C. We
analyzed the spacings of the layer lines in the
NBED patterns to calculate the chiral angles of
each wall at each step (Fig. 4D) and found a
trend for the chiral angles to increase to the
larger angle region near the armchair type of
chirality. For a shell with an initial chiral angle
of ~9.2°, the chiral angle gradually increased to
~23.2°. For a shell with an initial chiral angle
of ~24.4°, the chiral angle increased to 30° and
then fluctuated in the large angle region (~27°
to ~30°).
Generally, plastic deformation and chirality
transformations of CNTs have been attributed
to dislocation activity ( 14 ). In theory, two mech-
anisms have been discussed for the nucleation
of dislocations in CNTs, that is, bond flip driven
by stress and thermally activated carbon dimer
evaporation ( 15 ). The chirality changes due to
the gliding of dislocations nucleated from bond
flip and associated Stone-Thrower-Wales de-
fects ( 28 , 29 ) have been studied in detail to
predict a decreasing trend for the chiral angles
to near-zigzag-type chirality owing to the ener-
getically favorable (0,1) dislocations ( 14 , 15 ),
which is the opposite of the experimentally
observed increasing trend for the chiral angles
in this work. As mentioned, a near-transition
condition was used in our experiments so that
theCNTswereexposedtoahightemperature
from Joule heating and accompanied by a slow
elongation in a quasi-static process. Therefore,
we considered the orientation-dependent for-
mation energy of the dislocations generated
by the thermally activated evaporation of
carbon dimers (C 2 ) and associated formation
of 5|8|5 defects (Fig. 4E). The 5|8|5 defects
can yield 5|7 dislocation cores responsible for
the chirality transitions (Fig. 4F). The calcu-
lated formation energy of 5|8|5 defects (fig. S9)
indicated that at a lower chiral angle, the (1,0)
dislocation was energetically favored (fig. S10).
Therefore, nanotubes with initially small chiral
angles tend to increase the chiral angle during
the transitions. In a higher-angle region, the
formation energies of 5|8|5 defects associated
with (1,0) and (0,1) dislocations became close,
which explains the fluctuation behavior of
the chiral angles in the high-angle region. The
ratio of probabilities of forming (1,0) and (0,1)
dislocations was calculated to predict the chi-
rality dynamics (Fig. 4G), showing a converg-
ing trend toward larger angles, consistent with
the experimental observations.REFERENCES AND NOTES- R. Saito, M. Fujita, G. Dresselhaus, M. S. Dresselhaus,
Appl. Phys. Lett. 60 , 2204–2206 (1992). - A. D. Franklin, Z. Chen,Nat. Nanotechnol. 5 , 858–862 (2010).
3. C. Qiuet al.,Science 355 , 271–276 (2017).
4. Q. Cao, J. Tersoff, D. B. Farmer, Y. Zhu, S.-J. Han,Science 356 ,
1369 – 1372 (2017).
5. M. M. Shulakeret al.,Nature 501 , 526–530 (2013).
6. G. Hillset al.,Nature 572 , 595–602 (2019).
7. M. S. Arnold, A. A. Green, J. F. Hulvat, S. I. Stupp,
M. C. Hersam,Nat. Nanotechnol. 1 , 60–65 (2006).
8. J. R. Sanchez-Valenciaet al.,Nature 512 , 61–64 (2014).
9. F. Yanget al.,Nature 510 , 522–524 (2014).
10. S. Zhanget al.,Nature 543 , 234–238 (2017).
11. L. Chico, V. H. Crespi, L. X. Benedict, S. G. Louie, M. L. Cohen,
Phys. Rev. Lett. 76 , 971–974 (1996).
12. Z. Yao, H. W. C. Postma, L. Balents, C. Dekker,Nature 402 ,
273 – 276 (1999).
13. M. Ouyang, J.-L. Huang, C. L. Cheung, C. M. Lieber,
Science 291 , 97–100 (2001).
14. B. I. Yakobson,Appl. Phys. Lett. 72 , 918–920 (1998).
15. F. Ding, K. Jiao, M. Wu, B. I. Yakobson,Phys. Rev. Lett. 98 ,
075503 (2007).
16. J. Y. Huanget al.,Nature 439 , 281–281 (2006).
17. D.-M. Tanget al.,Ultramicroscopy 194 , 108–116 (2018).
18. E. S. Penev, V. I. Artyukhov, B. I. Yakobson,ACS Nano 8 ,
1899 – 1906 (2014).
19. A. A. Odintsov,Phys. Rev. Lett. 85 , 150–153 (2000).
20. M. Muothet al.,Nat. Nanotechnol. 5 , 589–592 (2010).
21. A. D. Franklinet al.,Nano Lett. 12 , 758–762 (2012).
22. S. Heinzeet al.,Phys. Rev. Lett. 89 , 106801 (2002).
23. A. Javeyet al.,Nat. Mater. 1 , 241–246 (2002).
24. Y. Luet al.,J. Am. Chem. Soc. 128 , 3518–3519 (2006).
25. A. Javeyet al.,Nano Lett. 4 , 447–450 (2004).
26. W. Lianget al.,Nature 411 , 665–669 (2001).
27. A. Javey, J. Guo, Q. Wang, M. Lundstrom, H. Dai,Nature 424 ,
654 – 657 (2003).
28. P. A. Thrower,Chem. Phys. Carbon 5 , 217–320 (1969).
29. A. J. Stone, D. J. Wales,Chem. Phys. Lett. 128 , 501–503 (1986).
ACKNOWLEDGMENTS
The authors thank P. Thrower (Penn State) for the constructive
advice. V.A.D., D.G.K., and P.B.S. are grateful to P. Vancsó (KFKI) for
fruitful discussions. The authors are grateful to a supercomputer
cluster at NUST“MISIS”provided by the Materials Modeling and
Development Laboratory and to the Joint Supercomputer Center of the
Russian Academy of Sciences.Funding:This work was supported by
JSPS KAKENHI (grants JP20K05281, JP25820336, JP18H05329,
JP19H02543, JP20H00220, and JP20KK0114); JST, CREST grant
JPMJCR20B5, Japan; the Ministry of Science and Technology of China
(grant 2016YFA0200101); and the National Natural Science Foundation
of China (NSFC) (grants 52188101, 51625203, and 52130209).
S.V.E. and P.B.S. acknowledge financial support from the Ministry of
Education and Science of the Russian Federation in the framework of
the Increase Competitiveness Program of NUST MISIS (K2-2020-
023). D.G.K. and V.A.D. acknowledge support from the Ministry of
Science and Higher Education of the Russian Federation (project
Nº01201253304). D.G. is grateful to the Australian Research Council
(ARC) for granting Laureate Fellowship FL160100089.Author
contributions:D.-M.T. conceived of and conducted the in situ TEM
experiments. S.V.E. and P.B.S. performed theoretical calculations
on the chirality dynamics. V.A.D. and D.G.K. performed theoretical
calculations on the quantum transport. O.C. made the LabVIEW
program. S.J., L.Z., P.-X.H., C.L., and H.-M.C. performed the growth of
SWCNTs through floating catalyst CVD. G.C. and D.N.F. performed
the growth of SWCNTs through super-growth CVD. Y.Z., R.X., and S.M.
performed the growth of SWCNTs through ethanol CVD. X.Z. and
F.-C.H. made probes. N.K., M.M., Y.N., F.U., and M.T. assisted with
TEM characterizations. C.L., H.-M.C., P.B.S., Y.B. and D.G. oversaw the
project. All authors discussed the results and revised the manuscript.
Competing interests:None declared.Data and materials
availability:All data needed to evaluate the conclusions in the paper
are present in the paper or the supplementary materials. Requests
for carbon nanotubes under a materials transfer agreement should
be made to D.-M.T.SUPPLEMENTARY MATERIALS
science.org/doi/10.1126/science.abi8884
Materials and Methods
Figs. S1 to S10
Table S1
References ( 30 – 46 )
Movie S1
7 April 2021; accepted 2 November 2021
10.1126/science.abi88841620 24 DECEMBER 2021•VOL 374 ISSUE 6575 science.orgSCIENCE
RESEARCH | REPORTS