Local modification and characterization of the electronic structure of carbon nanotubes

In december 1959, in his famous lecture "There's Plenty of Room at the
Bottom" given at Caltech, Richard Feynman imagined the possibility to
manufacture objects at the nanometer scale (1 nm = 10¡9 m) by maneu-
vering matter atom by atom. This revolutionary idea paved the way to
envision systems designed and engineered at the ultimate length scale rele-
vant to material science. Such systems have become a reality today and the
efforts to understand, build and use them encompass what is called nan-
otechnology. Today, nanoscience and nanotechnology constitute very active
and promising multidisciplinary research areas, bringing together engineers
and scientists from several ¯elds like physics, chemistry, materials science,
electronics, biology and medicine. A strong focus is given to the understand-
ing of the correlations between the structure of a material at the atomic
level and its optical, chemical and electronic properties. But nanoscience
and nanotechnology also aim at developing and improving techniques for
manufacturing nanomaterials for new applications.
When the dimensions of an object are shrunk down to a scale of the
order of the Fermi wavelength of the electrons, the said object will behave
according to the rules of quantum mechanics, and novel properties will
emerge that may be completely different from the bulk properties. One
well-known example for this is provided by graphite.
Graphite is a carbon allotrope whose structure consists of a stacking of
two-dimensional, sp2-bonded carbon layers interacting with each other by
van-der-Waals interactions. Because of these weak inter-layer interactions,
graphite is known as a mechanically soft material which is used in pencil
leads or, due to its high melting point and good electrical conductivity, in
the electrodes of arc lamps and arc furnaces. Now, if one imagines isolating a
small sheet of a graphite monolayer and roll it into a cylinder with nanome-
ter scale diameter, one will obtain a nano-object with amazing mechanical,
thermal and electronic properties. Such objects called carbon nanotubes
(CNT) have been discovered in 1991 1 and since then have led to an explo-
sion of research activities in many labs worldwide. In terms of mechanical
properties, carbon nanotubes are among the strongest and most resilient
materials known to exist in nature, with a Young's modulus approaching
1.2 TPa and a tensile strength 80 times higher than high strength steel.
Their electronic properties are unique in the sense that they are completely
determined by the tube geometry, resulting in semiconducting or metallic
character, with observed ballistic transport properties up to a few microm-
eters for the latter 2. During the last decade, CNT-based device prototypes
like single-electron transistors (SET) 3, ¯eld-effect transistor (FET) 4, logic
gates or memories 5;6 could be realized. Especially, it was demonstrated
that a CNTFET has superior performance over standard Si MOSFETs 7.
CNT based FETs are then very promising to be used as building blocks for
future large-scale integrated circuits as the actual silicon technology might
soon reach its limits in terms of miniaturization. Recent progress in separa-
tion techniques where individual chiralities can be isolated further sustain
this assumption 8. Also, growth techniques got improved with regard to
high purity of the raw material 9 and still constitute an active research
¯eld towards a highly desired chirality selective growth. Despite all these
advances, the development of a reliable and massively parallel integration
technology for CNT-based electronic devices that can compete in yield and
structure size with silicon technology is still missing and it is not yet clear
if such a technology can be established.
Up to now, especially in the worldwide race towards the realization of
the most competitive SWNT-based transistor, research focussed on mainly
defect free nanotubes. Nevertheless, researchers reported interesting results
from the investigation of intrinsic defects. For example, a classical p ¡ n
rectifying behavior has been reported for an intramolecular junction in a
single-walled carbon nanotube (SWNT), due to a special arrangement of
the C-C bonds at a tube kink 10. Furthermore, two separated intrinsic
defects in a metallic SWNT embodied in a source-drain-gate device showed
gate-dependent resonant backscattering properties 11. And more recently, a
high gate sensitivity at the position of individual defects of unknown nature
in similar devices has been observed 12.
These ¯ndings show that whereas a defect can be considered as a nui-
sance, on the other hand it can be regarded as an opportunity to tailor the
electronic properties of CNTs. Thus, the question arises to know how and
to which extent different types of defects can change the electronic proper-
ties of SWNTs. This is important in view to possibly modify and improve
the properties of existing CNT-based electronic devices such as CNTFETs,
or even further to de¯ne new kind of quantum devices with possibly new
properties entirely designed by a controlled creation of speci¯c defects.
The fundamental questions put above constitute the main motivation
of this Ph.D thesis. Here, we propose a study of the interplay between ar-
ti¯cially created defects and the electronic structure of SWNTs by means
of low temperature scanning tunneling microscopy and spectroscopy (LT-
STM/STS). LT-STM/STS constitutes the perfect investigation tool to achieve
this goal since it enables us to directly probe the electronic structure of solids
with atomic resolution. The defects we investigate here are created by ion
bombardment. In order to study the effect of well-de¯ned individual de-
fects, a low invasive method for their creation must be privileged. To serve
this purpose, we have chosen an ECR plasma source producing low energy
ions of the order of a few eV.
In a ¯rst phase, we investigated hydrogen ion-induced defects, motivated
by a earlier work on graphite where it was shown that such defects act as
scattering centers leading to large momentum scattering. In a second phase,
we studied the effects of an exposition of the tubes to nitrogen plasma. This
is motivated by the fact that the intrinsic p-type doping of CNTFETs draws
up a demand on techniques allowing n-type doping in view to de¯ne p-n
junctions, which are indispensable key-blocks towards a future SWNT-based
technology. Different approaches like deposition of K donor atoms 13 or
attempts to create substitutional N sites during the tube growth 14 present
inhomogeneity problems. A controlled substitution technique is thus highly
desirable.
In the last phase, we were interested to create harsher defects like va-
cancies and double vacancies, which have shown to largely increase the
resistivity of SWNTs 15. This could be achieved by medium energy argon
ions of 200 eV and 1.5 keV produced by an ion gun.
Outline
The present Ph.D thesis is organized as follows:
Chapter 1 gives a brief introduction on the geometrical structure, synthe-
sis and electronic structure of SWNTs. The actual knowledge on the
effects of structural defects on the electronic and transport properties
of SWNT will be given in the form of an overview of the literature.
Chapter 2 describes the experimental methods used in this work. As the
principal investigation tool, the basic theoretical principles of STM
and STS will be described, followed by a technical description or our
LT-STM/STS setup. A short introduction to tmAFM will also be
given. And ¯nally, the basic principles of cold plasma physics will be
described, with emphasis on ECR plamsa and DC glow discharge.
Chapter 3 describes in details the different steps involved in the sample
preparation, from the SWNT suspension to the defect free SWNT
sample characterization by means of tmAFM and LT-STM/STS.
Chapter 4 and 5 present topography and spectroscopy investigations on
ECR H- and N-plasma-induced defects. In both cases, new defect-
induced gap states in semiconducting SWNTs could be observed.
For the ECR-H plasma treatment, STM/STS investigations combined
with ¯rst principle ab initio calculations demonstrated that a corre-
lated chemisorption of H-adatoms on the SWNT wall gives rise to
symmetric paired gap states.
Chapter 6 presents topography and spectroscopy investigations on 200
eV and 1.5 keV Ar+ion-induced defects. From ¯rst principle ab initio
calculations combined with our experimental results, we concluded
on the formation of two main defect types: vacancies and C-adatoms
giving rise to new states in the semiconducting gap. An increase of
the complexity of the defect con¯guration has been observed for 1.5
keV treatement, compared to 200 eV.
Chapter 7 gives an extended discussion on the often observed NDR be-
havior in the I ¡ V curves recorded at defect sites. Within a simple
tunneling model, we could explain this phenomenon by a voltage de-
pendence of the tunneling barrier height.
Chapter 8 describes electronic con¯nement effects observed between con-
secutive defects in metallic SWNTs. The capability of our method to
create su±ciently strong scattering centers shows the possibility to de-
¯ne room temperature active intra-tube quantum dots. The discussion
of the experimental observations mainly based on a Fourier analysis of
the spatial variation of the differential conductance is made on the ba-
sis of a scattering matrix formalism as well as a geometrical approach
within the Fourier projection-slice theorem.
Details on the ab initio calculations for different defect structures and ad-
ditional calculation results are given in Appendices A and B, respectively.