Main New trends in nanotechnology and fractional calculus applications

New trends in nanotechnology and fractional calculus applications

, , , , , ,

In recent years fractional calculus has played an important role in various fields such as mechanics, electricity, chemistry, biology, economics, modeling, identification, control theory and signal processing. The scope of this book is to present the state of the art in the study of fractional systems and the application of fractional differentiation. Furthermore, the manufacture of nanowires is important for the design of nanosensors and the development of high-yield thin films is vital in procuring clean solar energy.

This wide range of applications is of interest to engineers, physicists and mathematicians.

Year: 2010
Edition: 1
Publisher: Springer Netherlands
Language: english
Pages: 531 / 544
ISBN 10: 9048132924
ISBN 13: 9789048132928
File: PDF, 9.39 MB
Download (pdf, 9.39 MB)
Preview
 
You can write a book review and share your experiences. Other readers will always be interested in your opinion of the books you've read. Whether you've loved the book or not, if you give your honest and detailed thoughts then people will find new books that are right for them.
New Trends in Nanotechnology
and Fractional Calculus Applications

New Trends in
Nanotechnology and
Fractional Calculus
Applications

Edited by

D. BALEANU
Çankaya University,
Balgat-Ankara, Turkey

Z.B. GÜVENÇ
Çankaya University,
Balgat-Ankara, Turkey
and

J.A. TENREIRO MACHADO
Institute of Engineering of Porto,
Porto, Portugal

123

Editors
Dumitru Baleanu
Çankaya University
Fac. Art and Sciences
Ogretmenler Cad. 14
06530 Ankara
Yüzüncü Yil, Balgat
Turkey
dumitru@cankaya.edu.tr

J.A. Tenreiro Machado
Institute of Engineering
of the Polytechnic Institute of Porto
Dept. Electrotechnical Engineering
Rua Dr. Antonio Bernardino de Almeida
4200-072 Postage
Portugal
jtm@isep.ipp.pt

Ziya B. Güvenç
Çankaya University
Fac. Engineering & Architecture
Ogretmenler Cad. 14
06530 Ankara
Yüzüncü Yil, Balgat
Turkey
guvenc@cankaya.edu.tr

ISBN 978-90-481-3292-8
e-ISBN 978-90-481-3293-5
DOI 10.1007/978-90-481-3293-5
Springer Dordrecht Heidelberg London New York
Library of Congress Control Number: 2009942132
c
Springer
Science+Business Media B.V. 2010
No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form or by
any means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without written
permission from the Publisher, with the exception of any material supplied specifically for the purpose
of being entered and executed on a computer system, for exclusive use by the purchaser of the work.
Cover design: eStudio Calamar S.L.
Printed on acid-free paper
Springer is part of Springer Science+Business Media (www.springer.com)

Preface

By the beginning of November 2008, the International Workshops on New Trends
in Science and Technology (NTST 08) and Fractional Differentiation and its
Applications (FDA08) were held at Çankaya University, Ankara, Turkey. These
events provided a place to exchange recent developments and progresses in several
emerging scientific areas, namely nanoscience, nonlinear science and complexity, symmetries and integrability, and application of fractional calculus in science,
engineering, economics and finance.
The organizing committees have invited presentations from experts representing the international community of scholars and welcomed contributions from the
growing number of researchers who are applying these tools to solve complex technical problems. Unlike the more established techniques of physics and engineering,
the new methods are still under development and modern work is proceeding by
both expanding the capabilities of these approaches and by widening their range of
applications. Hence, the interested reader will find papers here that focus on the underlying mathematics and physics that extend the ideas into new domains, and that
apply well established methods to experimental and to theoretical problems.
This book contains some of the contributions that were presented at NTST08
and FDA08 and, after being carefully selected and peer-reviewed, were expanded
and grouped into five main sections entitled “New Trends in Nanotechnology”,
“Techniques and Applications”, “Mathematical Tools”, “Fractional Modelling” and
“Fractional Control Systems”.
The selection of improved papers for publication in this book reflects the success
of the workshops, with the emergence of a variety of novel areas of applications.
Bearing these ideas in mind the guest editors would like to honor many distinguished
scientists that have promoted the development of nanoscience and fractional calculus and, in particular, Prof. George M. Zaslavsky that supported early this special
issue and passed away recently.
The organizing committees wish to express their thanks to Cem Ozdogan, Adnan
Bilgen, Ozlem Defterli, Burcin Tuna, Nazmi Battal as well as to our students for
their assistance.
The Editors would like to thank to Ozlem Defterli for helping in preparation of
this book.

v

vi

Preface

The organizing committees wish to thank the sponsors and supporters of NTST08
and FDA08, namely Çankaya University represented by the President of the Board
of Trustees Sıtkı Alp, the Rector Professor Ziya B. Güvenç, TUBITAK (The Scientific and Technological Research Council of Turkey), and the IFAC, for providing
the resources needed to hold this conference, the invited speakers for sharing their
expertise and knowledge, and the participants for their enthusiastic contributions to
the discussions and debates.
Ankara
March 31, 2009

Dumitru Baleanu
Ziya B. Güvenç
J.A. Tenreiro Machado

Contents

Part I New Trends in Nanotechnology
Novel Molecular Diodes Developed by Chemical Conjugation
of Carbon Nanotubes with Peptide Nucleic Acid : : : : : : : : : : : : : : : : : : : : : : : : : : : : :
Krishna V. Singh, Miroslav Penchev, Xiaoye Jing,
Alfredo A. Martinez–Morales, Cengiz S. Ozkan, and
Mihri Ozkan

3

Hybrid Single Walled Carbon Nanotube FETs for High
Fidelity DNA Detection : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 17
Xu Wang, Mihri Ozkan, Gurer Budak, Ziya B. Güvenç,
and Cengiz S. Ozkan
Towards Integrated Nanoelectronic and Photonic Devices: : : : : : : : : : : : : : : : : : : 25
Alexander Quandt, Maurizio Ferrari, and Giancarlo C. Righini
New Noninvasive Methods for ‘Reading’ of Random Sequences
and Their Applications in Nanotechnology : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 43
Raoul R. Nigmatullin
Quantum Confinement in Nanometric Structures: : : : : : : : : : : : : : : : : : : : : : : : : : : : 57
Magdalena L. Ciurea and Vladimir Iancu
Part II Techniques and Applications
Air-Fuel Ratio Control of an Internal Combustion Engine
Using CRONE Control Extended to LPV Systems : : : : : : : : : : : : : : : : : : : : : : : : : : : 71
Mathieu Moze, Jocelyn Sabatier, and Alain Oustaloup
Non Integer Order Operators Implementation via Switched
Capacitors Technology : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 87
Riccardo Caponetto, Giovanni Dongola, Luigi Fortuna,
and Antonio Gallo

vii

viii

Contents

Analysis of the Fractional Dynamics of an Ultracapacitor
and Its Application to a Buck-Boost Converter :: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 97
A. Parreño, P. Roncero-Sánchez, X. del Toro Garcı́a, V. Feliu,
and F. Castillo
Approximation of a Fractance by a Network of Four Identical
RC Cells Arranged in Gamma and a Purely Capacitive Cell : : : : : : : : : : : : : : : : 107
Xavier Moreau, Firas Khemane, Rachid Malti, and Pascal Serrier
Part III

Mathematical Tools

On Deterministic Fractional Models : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 123
Margarita Rivero, Juan J. Trujillo, and M. Pilar Velasco
A New Approach for Stability Analysis of Linear Discrete-Time
Fractional-Order Systems : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 151
Said Guermah, Said Djennoune, and Maamar Bettayeb
Stability of Fractional-Delay Systems: A Practical Approach : : : : : : : : : : : : : : : 163
Farshad Merrikh-Bayat
Comparing Numerical Methods for Solving Nonlinear
Fractional Order Differential Equations : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 171
Farhad Farokhi, Mohammad Haeri, and Mohammad Saleh
Tavazoei
Fractional-Order Backward-Difference Definition Formula
Analysis : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 181
Piotr Ostalczyk
Fractional Differential Equations on Algebroids and Fractional
Algebroids : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 193
Oana Chiş, Ioan Despi, and Dumitru Opriş
Generalized Hankel Transform and Fractional Integrals
on the Spaces of Generalized Functions : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 203
Kuldeep Singh Gehlot and Dinesh N. Vyas
Some Bounds on Maximum Number of Frequencies Existing
in Oscillations Produced by Linear Fractional Order Systems : : : : : : : : : : : : : : 213
Sadegh Bolouki, Mohammad Haeri, Mohammad Saleh Tavazoei,
and Milad Siami
Fractional Derivatives with Fuzzy Exponent : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 221
Witold Kosiński
Game Problems for Fractional-Order Systems : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 233
Arkadii Chikrii and Ivan Matychyn

Contents

ix

Synchronization Analysis of Two Networks:: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 243
Changpin Li and Weigang Sun
Part IV

Fractional Modelling

Modeling Ultracapacitors as Fractional-Order Systems : : : : : : : : : : : : : : : : : : : : : 257
Yang Wang, Tom T. Hartley, Carl F. Lorenzo, Jay L. Adams,
Joan E. Carletta, and Robert J. Veillette
IPMC Actuators Non Integer Order Models : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 263
Riccardo Caponetto, Giovanni Dongola, Luigi Fortuna,
Antonio Gallo, and Salvatore Graziani
On the Implementation of a Limited Frequency Band
Integrator and Application to Energetic Material Ignition
Prediction :: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 273
Jocelyn Sabatier, Mathieu Merveillaut, Alain Oustaloup,
Cyril Gruau, and Hervé Trumel
Fractional Order Model of Beam Heating Process
and Its Experimental Verification : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 287
Andrzej Dzieliński and Dominik Sierociuk
Analytical Design Method for Fractional Order Controller
Using Fractional Reference Model : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 295
Badreddine Boudjehem, Djalil Boudjehem, and Hicham Tebbikh
On Observability of Nonlinear Discrete-Time Fractional-Order
Control Systems : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 305
Dorota Mozyrska and Zbigniew Bartosiewicz
Chaotic Fractional Order Delayed Cellular Neural Network :: : : : : : : : : : : : : : : 313
Vedat Çelik and Yakup Demir
Fractional Wavelet Transform for the Quantitative Spectral
Analysis of Two-Component System : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 321
Murat Kanbur, Ibrahim Narin, Esra Özdemir, Erdal Dinç,
and Dumitru Baleanu
Fractional Wavelet Transform and Chemometric Calibrations
for the Simultaneous Determination of Amlodipine
and Valsartan in Their Complex Mixture : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 333
Mustafa Çelebier, Sacide Altınöz, and Erdal Dinç

x

Part V

Contents

Fractional Control Systems

Analytical Impulse Response of Third Generation CRONE
Control :: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 343
Rim Jallouli-Khlif, Pierre Melchior, F. Levron, Nabil Derbel,
and Alain Oustaloup
Stability Analysis of Fractional Order Universal Adaptive
Stabilization : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 357
Yan Li and YangQuan Chen
Position and Velocity Control of a Servo by Using GPC
of Arbitrary Real Order : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 369
Miguel Romero Hortelano, Inés Tejado Balsera, Blas Manuel
Vinagre Jara, and Ángel Pérez de Madrid y Pablo
Decentralized CRONE Control of mxn Multivariable System
with Time-Delay :: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 377
Dominique Nelson-Gruel, Patrick Lanusse, and Alain Oustaloup
Fractional Order Adaptive Control for Cogging Effect
Compensation : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 393
Ying Luo, YangQuan Chen, and Hyo-Sung Ahn
Generalized Predictive Control of Arbitrary Real Order : : : : : : : : : : : : : : : : : : : : 411
Miguel Romero Hortelano, Ángel Pérez de Madrid y Pablo,
Carolina Mañoso Hierro, and Roberto Hernández Berlinches
Frequency Response Based CACSD for Fractional Order
Systems : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 419
Robin De Keyser, Clara Ionescu, and Corneliu Lazar
Resonance and Stability Conditions for Fractional Transfer
Functions of the Second Kind : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 429
Rachid Malti, Xavier Moreau, and Firas Khemane
Synchronization of Fractional-Order Chaotic System
via Adaptive PID Controller :: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 445
Mohammad Mahmoudian, Reza Ghaderi, Abolfazl Ranjbar,
Jalil Sadati, Seyed Hassan Hosseinnia, and Shaher Momani
On Fractional Control Strategy for Four-Wheel-Steering
Vehicle : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 453
Ning Chen, Nan Chen, and Ye Chen
Fractional Order Sliding Mode Controller Design
for Fractional Order Dynamic Systems : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 463
Mehmet Önder Efe

Contents

xi

A Fractional Order Adaptation Law for Integer Order Sliding
Mode Control of a 2DOF Robot : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 471
Mehmet Önder Efe
Synchronization of Chaotic Nonlinear Gyros Using Fractional
Order Controller : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 479
Hadi Delavari, Reza Ghaderi, Abolfazl Ranjbar,
and Shaher Momani
Nyquist Envelope of Fractional Order Transfer Functions
with Parametric Uncertainty : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 487
Nusret Tan, M. Mine Ozyetkin, and Celaleddin Yeroglu
Synchronization of Gyro Systems via Fractional-Order
Adaptive Controller : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 495
Seyed Hassan Hosseinnia, Reza Ghaderi, Abolfazl Ranjbar,
Jalil Sadati, and Shaher Momani
Controllability and Minimum Energy Control Problem
of Fractional Discrete-Time Systems : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 503
Jerzy Klamka
Control of Chaos via Fractional-Order State Feedback
Controller :: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 511
Seyed Hassan Hosseinnia, Reza Ghaderi, Abolfazl Ranjbar,
Farzad Abdous, and Shaher Momani
Index . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .521

Part I

New Trends in Nanotechnology

Novel Molecular Diodes Developed by Chemical
Conjugation of Carbon Nanotubes with Peptide
Nucleic Acid
Krishna V. Singh, Miroslav Penchev, Xiaoye Jing,
Alfredo A. Martinez–Morales, Cengiz S. Ozkan, and Mihri Ozkan

Abstract In this work single walled carbon nanotube (SWNT)-peptide nucleic acid
(PNA) conjugates are synthesized and their electrical properties are characterized.
Metal contacts to SWNT-PNA-SWNT conjugates, used for current–voltage (I –V )
measurements, are fabricated by two different methods: direct placement on prepatterned gold electrodes and metal deposition using focused ion beam (FIB). Backgated I –V measurements are used to determine the electronic properties of these
conjugates. Additionally, conductive atomic force microscopy (C-AFM) is used to
characterize the intrinsic charge transport characteristics of individual PNA clusters.
As electronic devices scale down, traditional lithography-based fabrication methods face unprecedented challenges more than ever before [1, 2]. The need for novel
bottom up techniques to get over the hurdle posed by downscaling is getting increasingly urgent [3–5]. Molecular electronics, based on the unique self-assembly
capabilities of molecules, exemplifies the idea of bottom-up fabrication approach
[6, 7]. Therefore, the study of the electrical properties of single molecular components, can serve as a starting point for the study and realization of molecular
electronics. Carbon nanotubes (CNTs) based bioconjugates are a suitable candidate
for molecular electronics as they incorporate the excellent electrical and structural
properties of CNTs [8,9] and the self assembly properties of bio-molecules [10–12].
In our previous work, we have synthesized single walled carbon nanotube (SWNT)peptide nucleic acid (PNA) conjugates [13]. The main aim behind this work is to
test these conjugates for their future use in molecular electronics applications.
The as-synthesized conjugates have the following structure: two SWNT ropes
joined by a PNA cluster, where PNA acts as a linker to bring two SWNT ropes

K.V. Singh
Department of Chemical and Environmental Engineering, University of California,
Riverside, CA 92521
M. Penchev, X. Jing, A.A. Martinez–Morales, and M. Ozkan ()
Department of Electrical Engineering, University of California, Riverside, CA 92521
e-mail: cozkan@engr.ucr.edu; mihri@ee.ucr.edu
C.S. Ozkan
Department of Mechanical Engineering, University of California, Riverside, CA 92521

D. Baleanu et al. (eds.), New Trends in Nanotechnology and Fractional
Calculus Applications, DOI 10.1007/978-90-481-3293-5 1,
c Springer Science+Business Media B.V. 2010


3

4

K.V. Singh et al.

together. Due to their unique structure these conjugates can serve a twofolded
purpose. On one hand, they can be used to develop CNT based molecular devices
as SWNTs are functionalized and conjugated with a molecule. On the other hand,
CNTs can act as electrodes to electrically characterize and test the functionality of
PNA. In fact, till date there is no report on electrical transport through PNA. Using
this approach of conjugating SWNTs with PNA, provides us with a tool to test for
such electrical characteristics. Hence this work also reports the use of single-walled
carbon nanotubes (SWNTs) as a wiring alternative for molecular-scale devices. The
appropriate nanometer dimensions, chemical and mechanical stability, and high carrier mobility make SWNTs an ideal candidate for the same [14]. Due to these
advantages provided by SWNTs as components for molecular devices, lots of advances have been made to incorporate them into molecular device platform [15–17].
These include the development of high quality nanotube syntheses and integrated
molecular-SWNT chemical and biological sensors [18]. The biggest challenge in
using SWNTs as wires for molecular circuits is to engineer synthesis techniques
of combining molecules with SWNTs in a way that it will not affect the intrinsic
electrical transport properties of SWNTs. This work also overcome this challenge
by optimizing the functionalization of SWNTs which result in predominant end oxidation and hence incorporation of PNA molecules at the tip of tubes [13].
The major challenge in electrically characterizing these conjugates was fabricating electrodes/contacts to measure their electrical transport. Two different
techniques: direct placement on pre-patterned gold electrodes and focused ion beam
(FIB) were utilized according to the available resources and technology to develop
these contacts. In addition, individual PNA clusters were also characterized by conductive atomic force microscopy (C-AFM). The electrical transport results present
very interesting phenomena for these conjugates. The conjugates have asymmetrical electrical transport, allowing current to flow only in one direction, at room
temperature which corresponds to diodic or rectifying behavior. In addition some
conjugates also show characteristics of negative differential resistance (NDR) [19].
In this work, back-gated measurements on conjugates were also performed, allowing us to determine the transconductance and mobility of the conjugates. Therefore,
this work presents electrical properties of novel SWNT-PNA-SWNT conjugates and
in addition also comments on the conductivity of PNA.
The synthesis route for producing these conjugates is given in detail in our previous report [13]. Differently to our previous work, here we have used highly pure
HiPCO SWNTs [20] to increase the reliability of electrical transport results as the
SWNTs conduct through their surface [21]. Due to decrease in the impurities in
SWNT structure, which contribute towards faster oxidation of SWNTs, we have to
modify oxidation conditions. The new optimized oxidation conditions for predominantly end functionalization (as required) [13] of SWNTs are 14 h of acid reflux
in 2.4 M of HNO3 . Increase in oxidation time and also the strength of acid used
in this work (previously 12 h and 1 M HNO3 / is a strong indicative of the purity of
SWNTs employed in the synthesis of these SWNT-PNA conjugates. After oxidation
and subsequent sonication of SWNTs, SWNT bearing NHS esters were prepared by
coupling with EDC and NHS [13]. Both end functionalization of PNA (AcLys–
GTGCTCATGGTG-Lys-NH2 ) led to formation of SWNT-PNA-SWNT conjugates

Novel Molecular Diodes Developed by Chemical Conjugation of Carbon Nanotubes

5

Fig. 1 SEM micrograph of a SWNT-PNA-SWNT conjugate

as an amide bond is formed between the amine of the amino-acid residue on the
PNA backbone and SWNT-bearing NHS esters [13]. A typical scanning electron
microscopy (SEM) image of a SWNT-PNA-SWNT conjugate is shown in Fig. 1.
In this work we also modified the amino acid residue on the PNA backbone to
Lysine to improve the solubility of PNA in water.
After synthesis of the SWNT-PNA-SWNT conjugates, electrical contacts were
fabricated at the ends of individual conjugates by the following methods. The first
method consists of a direct placement on pre-patterned gold electrodes. One block
of four gold electrodes was patterned on Si=SiO2 chips. The structure of one electrode consist of a large square pad (L  125 m) which is connected to a long metal
strip approximately 80 m long. In one block there were four such electrodes and
in the center of the block the separation between the metal strips is around 1 m.
On one single chip there were 289 such blocks. In the direct placement method the
conjugates are deposited by drop casting; bridging across the metal strips due to the
length of SWNTs. After locating the connected strips on a particular block, the electrical measurements are done by connecting the bigger pads of the corresponding
metal strips to external probes (tip diameter  1 m) in a probe station (Signatone).
Using an Agilent 4155 C semiconductor parameter analyzer the I –V characteristics
of these conjugates were obtained.
The major advantage of this method is the simplicity and less time consumption
in preparing the sample for electrical characterization. But the major drawback is
that this method works on “hit and trial” basis and locating a single conjugate connected across two metal strips is a time consuming step. In addition, the contact
between the conjugate and the electrode is not necessarily good (as the conjugate
is sitting on top of the electrode) and can create artifacts during the measurements.
Sometimes it is also possible that whole chip does not have the required connection
or electrodes are not connected by the right conjugates.
The second method used for fabricating the contacts employs the use of focused
ion beam (FIB). It consists of an electron beam (SEM) as well as an ion beam
(Gallium ions). This technique provides us the opportunity to visualize the conjugates (by SEM) and develop the contacts directly on the conjugates by metal
deposition assisted by the ion beam (Leo XB1540). The required conjugate is

6

K.V. Singh et al.

located on the pre-patterned electrode system (as discussed above) by SEM. The
measurements are made at the same time for the contacts. Then the deposition of
metal (i.e. Platinum) takes place by the following procedure. A gas containing metal
ions is introduced into the system and allowed to chemisorb onto the sample. By
scanning an area with the ion beam, the precursor gas is decomposed into volatile
and non-volatile components; the non-volatile component (platinum metal) remains
on the surface as a deposition while the volatile component is vaporized. One major
advantage of this system is that one can monitor the formation of contacts in real
time under SEM.
This technique overcomes the disadvantages of less control, lack of precision
and “hit and trial” approach of the previous technique discussed above. But this
technique has its own set of issues, which mainly include the destruction of sample
by ion beam and shifting (if the system is not calibrated precisely). In order to
avoid damaging the SWNT-PNA conjugates, the following parameters were chosen:
deposition current of 2 A and scanning frequency of 0.1 Hz, which worked well for
our conjugates. In addition, this technique can also be used to repair the damaged
electrodes after measurements and the same conjugate can be reused, which is not
possible by the other technique. Moreover, destructive ion milling can also be used
as means to isolate the conjugate from other materials. For this purpose currents
higher than 50 A were used.
In order to report the first electrical conductivity measurements of PNA
molecules, we prepared samples for C-AFM analysis (Fig. 3) by drop casting a
solution of PNA (100 M concentration) on an oxygen plasma cleaned n-type Si
substrate. Oxygen plasma cleaning ensured the removal of any carbonaceous impurities as they might interfere with the final results since PNA is also carbonaceous
in nature. During CAFM measurements a Pt/Ir coated AFM tip (20 nm radius
of curvature) was used as a top contact to measure the current with respect to an
applied bias voltage. The electrical measurements were taken by first performing
a morphology scan in contact mode and then driving the tip by a point and shoot
method to the top of a specific PNA cluster.
After contact fabrication the SWNT-PNA-SWNT conjugates were tested by different methods as described above. Most of the conjugates show asymmetrical
current–voltage (I –V ) characteristic. Most of which show a rectifying or diodic behavior. This behavior was independent of the method used to fabricate the contacts.
Typical diodic behavior is shown in Fig. 2a, c. In addition some conjugates also
show negative differential resistance, which is characteristic of resonance tunneling
diode (RTD). Figure 2b, d represent the NDR characteristic of few conjugates. Control devices based on SWNT-only samples were also fabricated and the results are
shown in Fig. 2e, f.
Additionally, the intrinsic charge transport characteristics of individual PNA
clusters (Fig. 3 inset) were also studied by C-AFM measurements. As shown in
Fig. 3, typical PNA current–voltage measurements at the nanoscale exhibit a rectifying behavior analogous to the I –V curves observed for the SWNT-PNA conjugates.
For the negative tip bias voltages, a steep and exponential increase of the tunneling current occurs beyond a threshold voltage of 6 V. The turn-on voltage

Novel Molecular Diodes Developed by Chemical Conjugation of Carbon Nanotubes

7

Fig. 2 Two terminal electrical characterization of SWNT-PNA-SWNT conjugates. (a) and (c)
Diodic behavior is observed for both direct placement and focused ion beam (FIB) method. (b)
and (d) Similarly, negative differential resistance behavior was observed in few conjugates for
both methods. (e) and (f) SWNTs-only samples show symmetric behavior with high conductivity
irrespective of method

observed in the PNA cluster is in good agreement with the measurements made
on the SWNT-PNA conjugates (Fig. 2a). It is also interesting to point out that PNA
shows extremely good current-blocking behavior under positive tip bias voltage of
up to 10 V.

8

K.V. Singh et al.

Fig. 3 Charge transport characterization by C-AFM. The I –V curve shows a characteristic diodic
behavior for a single PNA cluster. Inset: AFM topography image of a single PNA cluster

Field effect transistors (FETs) were fabricated on single conjugates by using
a Si=SiO2 substrate as the back-gate, as the back gate and insulator respectively,
during the electrical measurements. A representative I –V curve for these gated
studies is represented in Fig. 4a showing that the SWNT-PNA-based FETs behave
as ‘p’ type conjugates. Few conjugates did not show any change in conductivity
on applying a gate voltage (Fig. 4b). To further test the electrical properties of
our device structure control devices based on SWNT ropes alone (Fig. 4c, d) were
also fabricated. The ropes which were semiconducting were found to be ‘p’ type
while metallic ropes do not show any semiconducting behavior. The back-gated
measurements were used to determine transconductance and mobility of the SWNTPNA-SWNT conjugate FET device (Fig. 4e, f).
The diodic behavior observed in the SWNT-PNA-SWNT conjugates is not a
new phenomenon in molecular electronics. In 1974 Aviram and Ratner proposed
a molecule based rectifying behavior [22]. That work was one of the pioneers in the
field of molecular electronics. Since then there have been numerous efforts to develop AR theory based rectifiers. In the literature, there are several molecules which
have shown this rectifying behavior [23–25] but this kind of observation is made
here for the first time for PNA. The mechanism for this rectifying behavior is explained in detail elsewhere [23–25]. In short, for an ideal AR molecular diode, the
rectifying molecule has a D-¢-A structure, where D is a good electron donor, ¢ is
the insulating bridge and A is the good electron acceptor. The rectifying behavior of
the molecule is observed when this molecule is connected to the conductors (Conductor (C1)-Molecule (M)-Conductor (C2)) on both ends. The mechanism involves
two molecular orbitals, the highest occupied molecular orbital (HOMO), mainly
localized on D, which would be filled, and the lowest unoccupied molecular orbital (LUMO), mainly localized on A. Electrons transfer from one contact to the
other contact by tunneling through the D-¢-A molecule which forms the preferentially excited electronic state. DC -¢-A . Inelastic “downhill” tunneling within the
molecule (involving either phonon emission or photon emission) then would reset

Novel Molecular Diodes Developed by Chemical Conjugation of Carbon Nanotubes

9

Fig. 4 FET characterization. (a) and (b) Gated study of SWNT-PNA-SWNT conjugates. The
electrical behavior of the conjugates is modulated by the type of SWNTs connecting the conjugate. (c) and (d) SWNT ropes are shown to behave either as ‘p’ type semiconductors or metallic,
respectively. (e) Transconductance of SWNT-PNA-SWNT conjugates. (f) Mobility of SWNTPNA-SWNT conjugates

10

K.V. Singh et al.

the ground state D-¢-A, but an electron would have been moved from metal electrode C1 to metal C2; hence the rectifying effect [11]. The proposed molecule was
never synthesized, but helped in developing the theory behind the rectifying behavior of molecules in molecular electronics. The observance of diodic behavior is also
due to the chemical structure of the molecule. When the relevant molecular energy is
in resonance with the Fermi level of the metal electrode, there is a dramatic increase
in the current through the molecule, and a dramatic selectivity of electron transport
through the C1-M-C2 sandwich [26]. Hence when the molecular orbitals of PNA
come to resonance with the molecular orbitals of SWNTs attached to them, there is
an observed increase in the current. But this phenomenon is not reversible and the
current can only be conducted in one direction only. Therefore, both the structure
and contact of PNA with conductors (i.e. SWNTs) on both ends are responsible for
the diodic behavior observed in the conjugates (Fig. 2a, c). Furthermore, the observance of diodic behavior of the PNA clusters in conductive AFM (Fig. 3) is also due
to this C1-M-C2 sandwich. In this case the conductors are the AFM tip (metal) on
the top and Silicon (semiconductor) on the bottom. This observation further supports the fact that the observed diodic behavior of SWNT-PNA-SWNT conjugates
is not because of SWNTs contact with PNA but rather due to the PNA itself.
The exact mechanism of transfer between SWNT-PNA-SWNT will require extensive modeling based on molecular dynamics. Only then we can locate the various
molecular orbitals in the conjugates and their behavior under an external electric
field. But the mechanism explained above gives us a right start in this direction.
As far as NDR effect is concerned, there are many reports on observation of
NDR in molecular electronics [27, 28]. Many mechanisms have been proposed for
the same but there is no consensus in the literature. In fact, we have previously observed a similar NDR behavior in our earlier work related to SWNT-DNA-SWNT
conjugates [29]. Since the bonding between SWNT and DNA is analogous to the
one between SWNT and PNA, we propose the following qualitative explanation
for the observance of NDR effect in SWNT-PNA-SWNT conjugates (Fig. 5) [29].
At zero bias voltage, chains of SWNT-PNA-SWNT conjugates have uniform Fermi

Fig. 5 Schematic illustration of electrons transferring through energy barriers of PNA molecules

Novel Molecular Diodes Developed by Chemical Conjugation of Carbon Nanotubes

11

energy levels. When applied voltage increases, energy levels tilt and electrons start
tunneling from the voltage source through the energy barriers of PNA molecules.
Correspondently, current increases until the localized energy band inside quantum
well shifts to below Fermi energy from the voltage source, leaving no corresponding energy levels for after-tunneling electrons to stay. As a result, current starts to
decrease. As the applied voltage continues to increase, the higher unoccupied energy levels in PNA shift down to the energy level which are in alignment with the
Fermi energy from the energy source and current starts to increase again. Since our
conjugates consists of SWNT ropes formed by many intertwined tubes and in consequence of numerous PNA molecules, alignment and misalignment do not happen
at the same voltage, it is reasonable that we get multiple current peaks for different
SWNT-PNA conjugates.
In addition, Lake et al. [30] have postulated SWNT-pseudo peptide-SWNT
nanostructure could exhibit RTD I –V response via computations based on the density functional theory (DFT) and non-equilibrium Green function (NEGF) approach.
Our results are in accordance with these theoretical and experimental analyses.
Control measurements were done on SWNT ropes alone with a two-fold purpose.
Firstly, to differentiate the electrical characteristics obtained for the conjugates versus the electrical properties of SWNT ropes. Secondly, to indirectly prove that PNA
is indeed joining two different SWNT ropes. Representative I –V curves of SWNT
ropes (Fig. 2e, f) show a symmetrical nature and higher conductivity for the ropes.
The electrical characteristics clearly show that the ropes are fundamentally a different system from that of the conjugates.
The gated study presented in this work is the first of its kind for PNA based
carbon nanotube conjugates. It was found that the conjugates were semiconducting as well as metallic (Fig. 4a, b). A control study was also performed on SWNT
ropes alone (Fig. 4c, d). Few of the ropes were found to be metallic and some to
be semiconducting, as expected. But the difference in the nature of SWNT-PNASWNT conjugates can also be explained on the basis of SWNTs. Since PNA is very
small compared to SWNTs and also much less conductive than SWNTs (as shown
in I –V characteristics); the influence on total gated behavior will be modulated by
the SWNTs of the conjugate. If PNA is attached to semiconducting SWNTs on both
the ends, the conjugate will behave as semiconductor but if either or both of the
SWNTs are metallic the conjugate will then behave as a metallic component.
Overall, the gated study confirms that PNA behaves as a hole conducting
molecule. This study also confirms the theoretical model explained elsewhere
for CNT-Peptide-CNT system [31]. Lake et al. modeled the peptide molecule and
found out that peptide linker acts as a good bridge for hole transmission in the CNT
valence band and strongly suppresses electron transmission in the CNT conduction
band [31].
During the electrical characterization of these conjugates, the biggest challenge
was to understand the difference in behavior observed among different conjugates.
The reason for this variation could be result of the following three main factors:
variation in number of SWNTs, variation of number of PNAs, and variation of the
type of SWNTs in the conjugates.

12

K.V. Singh et al.

The SWNTs used here are ropes and these ropes attach themselves to PNA
molecules by covalent coupling as described above. But all the ropes are not of
same diameter and hence do not contain the same number of tubes. Therefore, this
variation in number of tubes will always be observed from one conjugate to another.
This variation in number of tubes on both sides of a PNA cluster will also change
the number of PNAs from one conjugate to another. But the number of PNAs can
be estimated by the following methodology.
The number of PNA attached in one conjugate can be calculated from the diameter of the SWNT rope in that sample. Haddon et al. reported that the efficiency
of the oxidation process for carbon nanotubes (tubes @ Rice are approximated for
HiPCO tubes) is around 2% [32]. In this work we have preferentially oxidized the
tips of SWNTs. Therefore, we can estimate the number of oxidized carbon atoms at
the tip by this formula:
; D Number of oxidized carbon atoms in on tube


  dtube
D
 .Efficiency of oxidation process/
Length of C–C bond in SWNT .nm/
where, dt ube W Diameter of single tube (nm)
It is estimated that on an average in a SWNT rope of 20 nm diameter there are
around 500 tubes [33]. To get the number of oxidized sites in a rope (), we can
multiply ; with rope correction factor ‰ (which gives the number of SWNTs in one
rope of diameter davg .)

where ' D

davg
20.nm/


 500

Therefore,
˝ D Number of oxidized carbon atoms in one rope D ;  '
The efficiency of esterification (formation of SWNT-NHS esters) is nearly 100%
as the intermediates are in excess. As per the chemistry, we also keep the amines
(in our case PNA) in excess. Therefore, all the oxidation sites on the SWNT ropes
will be utilized by PNA molecules. Since for one site we can only have one PNA
molecule attached the number of PNA molecules attached will be equal to .
A major challenge of using SWNTs in bulk or in solution is that it contains both
metallic and semiconducting tubes/ropes. There is no easy way to separate them and
utilize them separately. Our conjugates also suffer from this inherent disadvantage.
In the conjugates, three types of configuration are possible; metallic (M)-PNA-M,
semiconducting (SC)-PNA-SC and SC-PNA-M; will occur. In fact, this variation
is clearly verified by the gated study of these conjugates. This configuration will
affect the shape, position of NDR peaks and nature of the current–voltage response
for SWNT-PNA-SWNT conjugates since the resonance of energy levels between
SWNTs and PNA is responsible for the rectifying nature of these conjugates.

Novel Molecular Diodes Developed by Chemical Conjugation of Carbon Nanotubes

13

As discussed above, in addition to developing SWNT based devices, this
structure could also served as a way of utilizing SWNTs as electrodes for the
characterization of molecular structures. The most common method of testing
molecules electrically is by Langmuir-Blodgett (LB) thin films [26, 34, 35]. In
contrast to the LB thin film technique there are several advantages in using the
architecture presented in this work for characterization of molecules. First, the
electrical transport is confined to one dimension along the molecules, whereas in
the thin film approach, conduction also takes place along the latitudinal direction
as well. Second, the number of molecules attached is restricted by the coupling
sites available on the SWNTs, permitting high accuracy in calculating the number of molecules attached. The number of functionalized sites in a rope/tube can
be estimated (as explained above). Therefore, from the number of these sites the
number of attached molecules can also be calculated. Third, CNTs themselves have
exceptional electronic properties and also have excellent mechanical and chemical properties as well that could be useful for the characterization of the intrinsic
properties of molecules.
In summary, we have synthesized single walled carbon nanotube (SWNT)peptide nucleic acid (PNA) conjugates, which are characterized by several
different techniques to determine their electrical properties. Our results demonstrate that the conjugates exhibit rectifying and negative differential resistance I –V
characteristics, making them ideal candidates for future electronic applications [36]
as molecular diodes. Furthermore, the excellent structural and electrical properties
of SWNTs enable us to use them as test electrodes in order to study the electrical
and electronic properties of PNA cluster.
Acknowledgements We gratefully acknowledge financial support from the Nanomanufacturing
Program of the National Science Foundation (NSF) (grant no: 0800680), the FCRP Center on
Functional Engineered Nano Architectonics funded by the SRC and DARPA, and the Center for
Hierarchical Manufacturing (CHM) funded by the NSF.

References
1. Lin BJ (2006) The ending of optical lithography and the prospects of its successors. Microelectronic Engineering 83(4–9):604–613
2. Lin BJ (2006) Optical lithography—present and future challenges. Comptes Rendus Physique
7(8):858–874
3. Huang Y, Duan XF, Cui Y, Lauhon LJ, Kim KH, Lieber CM (2001) Logic gates and computation from assembled nanowire building blocks. Science 294(5545): 1313–1317
4. Cui Y, Lieber CM (2001) Functional nanoscale electronic devices assembled using silicon
nanowire building blocks. Science 291(5505):851–853
5. Lu W, Lieber CM (2007) Nanoelectronics from the bottom up. Nature Materials 6(11):841–850
6. Fu L, Cao L, Liu Y, Zhu D (2004) Molecular and nanoscale materials and devices in electronics.
Advances in Colloid and Interface Science 111(3):133–157
7. Balzani V, Credi A, Venturi, M (2003) Molecular logic circuits. Chemphyschem 4(1):49–59
8. Baughman RH, Zakhidov AA, de Heer WA (2002) Carbon nanotubes – the route toward
applications. Science 297(5582):787–792

14

K.V. Singh et al.

9. Wildoer JWG, Venema LC, Rinzler AG, Smalley RE, Dekker C (1998) Electronic structure of
atomically resolved carbon nanotubes. Nature 391(6662): 59–62
10. Bashir R (2001) DNA-mediated artificial nanobiostructures: state of the art and future directions. Superlattices and Microstructures 29(1):1–16
11. Chakrabarti R, Klibanov AM (2003) Nanocrystals modified with peptide nucleic acids (PNAs)
for selective self-assembly and DNA detection. Journal of the American Chemical Society
125(41):12531–12540
12. Vernille JP, Kovell LC, Schneider JW (2004) Peptide nucleic acid (PNA) amphiphiles: Synthesis, self-assembly, and duplex stability. Bioconjugate Chemistry 15(6):1314–1321
13. Singh KV, Pandey RR, Wang X Lake R, Ozkan CS, Wang K, Ozkan M (2006) Covalent functionalization of single walled carbon nanotubes with peptide nucleic acid: Nanocomponents
for molecular level electronics. Carbon 44(9):1730–1739
14. Star A, Han TR, Gabriel JCP, Bradley K, Gruner G (2003) Interaction of aromatic compounds
with carbon nanotubes: Correlation to the Hammett parameter of the substituent and measured
carbon nanotube FET response. Nano Letters 3(10):1421–1423
15. Kong J, Chapline MG, Dai HJ (2001) Functionalized carbon nanotubes for molecular hydrogen
sensors. Advanced Materials 13(18):1384–1386
16. Lin Y, Taylor S, Li HP, Fernando KAS, Qu LW, Wang W, Gu LR, Zhou B, Sun YP (2004).
Advances toward bioapplications of carbon nanotubes. J Mat Chem 14(4):527–541
17. Tasis D, Tagmatarchis N, Bianco A, Prato M (2006) Chemistry of carbon nanotubes. Chemical
Reviews 106(3):1105–1136
18. Kong J, Soh HT, Cassell AM, Quate CF, Dai HJ (1998) Synthesis of individual single-walled
carbon nanotubes on patterned silicon wafers. Nature 395(6705):878–881
19. Sze SM, Ng KK (2006) Tunnel Devices. Physics of Semiconductor Devices. 3rd edn. Wiley,
Hoboken p 417
20. Nikolaev P (2004) Gas-phase production of single-walled carbon nanotubes from carbon
monoxide: A review of the HiPCO process. Journal of Nanoscience and Nanotechnology.
4(4):307–316
21. Odom TW, Huang JL, Kim P, Lieber CM (1998) Atomic structure and electronic properties of
single-walled carbon nanotubes. Nature 391(6662):62–64
22. Aviram A, Ratner MA (1974) Molecular Rectifiers. Chem Phys Lett 29(2):277–283
23. Metzger RM (1999) Electrical rectification by a molecule: The advent of unimolecular electronic devices. Acc Chem Res. 32(11):950–957
24. Metzger RM (2003) Unimolecular electrical rectifiers. Chem Rev. 103(9):3803–3834
25. Metzger RM (2004) Electrical rectification by monolayers of three molecules. Macromolecular
Symposia 212(1):63–72
26. Metzger RM (2006) Unimolecular rectifiers: Methods and challenges. Analytica Chimica Acta
568(1–2):146–155
27. Chen J, Reed MA, Rawlett AM, Tour JM (1999) Large on-off ratios and negative differential
resistance in a molecular electronic device. Science 286(5444):1550–1552
28. James DK, Tour JM (2006) Organic synthesis and device testing for molecular electronics.
Aldrichimica Acta 39(2):47–56
29. Wang X, Liu F, Andavan GTS, Jing XY, Singh K, Yazdanpanah VR, Bruque N, Pandey RR,
Lake R, Ozkan M, Wang KL, Ozkan CS (2006) Carbon Nanotube-DNA nanoarchitectures and
electronic functionality. Small 2(11):1356–1365
30. Pandey RR, Bruque N, Alam K, Lake RK (2006) Carbon nanotube – molecular resonant tunneling diode. Physica Status Solidia – Applications and Materials Science 203(2):R5–R7
31. Bruque N, Pandey RR, Lake RK, Wang H, Lewis JP (2005) Electronic transport through a
CNT-Pseudopeptide-CNT hybrid material. Molecular Simulation 31(12):859–864
32. Hu H, Bhowmik P, Zhao B, Hamon MA, Itkis ME, Haddon RC (2001) End-group and defect
analysis of soluble single-walled carbon nanotubes. Chem Phys Lett 345(1–2):25–28
33. Thess A, Lee R, Nikolaev P, Dai H, Petit P, Robert J, Xu C, Lee YH, Kim SG, Rinzler AG,
Colbert DT, Scuseria GE, Tomanek D, Fischer JE, Smalley RE (1996) Crystalline ropes of
metallic carbon nanotubes. Science 273(5274):483–487

Novel Molecular Diodes Developed by Chemical Conjugation of Carbon Nanotubes

15

34. Metzger RM, Panetta CA (1989) Langmuir-blodgett films of potential donor sigma acceptor
organic rectifiers. J Mol Elect 5(1):1–17
35. Okazaki N, Sambles JR, Jory MJ, Ashwell GJ (2002) Molecular rectification at 8 K in an
Au/C(16)H(33)Q-3CNQ LB film/Au structure. Applied Physics Letters 81(12): 2300–2302
36. Kumar MJ (2007) Molecular Diodes and Applications. Recent Patents on Nanotechnology
1:51–57

Hybrid Single Walled Carbon Nanotube FETs
for High Fidelity DNA Detection
Xu Wang, Mihri Ozkan, Gurer Budak, Ziya B. Güvenç, and Cengiz S. Ozkan

Abstract A novel application for detecting specific biomolecules using SWNTssDNA nanohybrid is described. SWNT-ssDNA hybrid is formed by conjugating
amino-ended single strand of DNA (ssDNA) with carboxylic group modified
SWNTs through a straightforward EDC coupling reaction. ssDNA functionalized SWNT hybrids could be used as high fidelity sensors for biomolecules. The
sensing capability is demonstrated by the change in the electronic properties of
SWNT. Employing DNA functionalized SWNT FETs could lead to dramatically
increased sensitivity in biochemical sensing and medical diagnostics applications.

1 Introduction
Carbon nanotubes (CNT) have been utilized widely in nanoelectronic devices such
as field effect transistors (FET) [1], single-electron transistors [2], rectifying diodes
[3] and logic circuits [4] due to its unique mechanical, thermal and electrical properties. They are chemically inert and it is difficult to conduct synthetic chemical
treatment on them because they are resistant to wetting and indissolvable in water
and organic solvents. In order to expand their potential applications in biomedical
and optoelectronic devices, surface functionalization strategies have been explored
by many research groups within recent years. The attachment of chemical functional
X. Wang
Department of Chemical Engineering, University of California, Riverside, CA 92521
M. Ozkan
Department of Electrical Engineering, University of California, Riverside, CA 92521
e-mail: mihri@ee.ucr.edu
G. Budak
Nanomedicine Research Laboratory, Gazi University, Besevler, Ankara, Turkey 06510
Z.B. Güvenç
Electronic and Communication Engineering, Cankaya University, Ankara, Turkey 06530
e-mail: guven@cankaya.edu.tr
C.S. Ozkan ()
Department of Mechanical Engineering, University of California, Riverside, CA 92521
e-mail: cozkan@engr.ucr.edu
D. Baleanu et al. (eds.), New Trends in Nanotechnology and Fractional
Calculus Applications, DOI 10.1007/978-90-481-3293-5 2,
c Springer Science+Business Media B.V. 2010


17

18

X. Wang et al.

groups represents a strategy for overcoming the disadvantages of CNTs and has
become attractive for synthetic chemists and materials scientists. Functionalization can improve CNTs solubility and processibility, and will allow combination
of unique properties of CNTs with those of other types of materials. The functionalization of CNTs can be divided into covalent and noncovalent types. Covalent
functionalization is based on covalent linkage of functional entities onto CNTs ends
and/or sidewall. Non-covalent functionalization is mainly based on the adsorption
forces between functional entities and CNTs, such as van der waals and -stacking
interaction. With the successful surface functionalization of CNTs, various strategies of forming CNT hybrids with chemicals, polymers, and biological species have
been developed, including fluorination of nanotubes [5], cholorination of nanotubes
[6], formation of carbon nanotube-acyl amides [7], and carbon nanotube-esters [8].
The integration of biomaterials, such as proteins, enzymes, antigens, antibodies, and
nucleic acids with CNTs would combine the conductive or semiconductive properties of CNTs with recognition or catalytic properties of biomaterials. A number of
researchers focus on DNA assemblies with CNTs because of the molecular recognition capability and high aspect ratio nanostructures. DNA has been utilized as
scaffolding materials or fabrics with applications in electronics; such constructs include DNA lattices [9], grids [10], tiles [10], ribbons [10], tubes [10], and origami
[11] for organizing components of electronics.
CNT-DNA complexes have been assembled via different methods. DNA’s interaction with CNT through the physical binding has been explored. DNA’s nonspecific binding to CNT wall has been visualized by high resolution transmission
electron microscopy [12].
DNA transport through a single MWNT cavity has been directly observed
by fluorescence microscopy [13]. During the process, both Van der Waals and
hydrophobic forces are found to be important, with the former playing a more dominant role on CNT-DNA interactions [14]. DNA interaction with CNT through
chemical covalent binding has also been described [15]. The amide linkage is
formed by the reaction of carboxylic groups on CNT with the amine groups of ssDNA in a solution. Such heterostructures indicated a negative differential resistance
(NDR) effect indicating a biomimetic route to forming resonant tunneling diodes
(RTD). CNT-DNA assemblies have been applied into detection of biomaterials and
chemical species. Label free detection of DNA hybridization using carbon nanotube
network field-effect transistors [16] has been demonstrated. DNA functionalized
single wall carbon nanotubes for electrochemical detection has been reported [17].
Most of this prior work presents the sensing capability of CNT networks or CNT
film structures. In this work, the detection of specific sequences of DNA using a
single SWNT field effect transistor is described. SWNTs are purified and dispersed
in o-dichlorobenzene (ortho-dichlorobenzene) solvent before functionalized by ssDNA. The functionalization is completed by forming an amide linkage between
carboxylic groups of SWNT and amine groups of ssDNA via the EDC coupling
method. Modified SWNT based biosensor in the configuration of a field effect transistor (FET) is fabricated using electron beam lithography (EBL). When specific
sequences of ssDNA which are complementary to the ssDNA covalently bound on

Hybrid Single Walled Carbon Nanotube FETs for High Fidelity DNA Detection

19

the SWNT surface are exposed to the device, modulation of the current-voltage
characteristics demonstrate the capability of SWNT-ssDNA nanohybrids for applications in high fidelity biosensing.

2 Experimental Section
2.1 SWNT Purification and Dispersion
SWNTs with carboxylic functional groups in 2.73 wt% were purchased from Cheap
Tubes, Inc. They were first purified and dispersed following a previously defined
procedure [18] as follows: SWNT-COOH (1 mg) was added in o-dichlorobenzene
(o-DCB) solvent (10 mL), followed by sonication in an ice bath for 10 min. Sonication usually generates a lot of heat, therefore, an ice bath is used for protecting
the SWNTs from physical damage. After sonication, the mixture solution was centrifuged for 90 min at 13,000 rpm. The supernatant was then further centrifuged at
55,000 rpm for 2 h. The resulting supernatant solution is almost transparent, and the
resulting functionalized SWNTs are shown in Fig. 1.

a

b

c

d

Fig. 1 SWNT purification and dispersion process. (a) SEM image of commercial carboxylic group
functionalized SWNTs. (b) SWNTs sonicated in ODCB for 5 min. (c) Supernatant of SWNT solution collected after centrifugation at 13,000 rpm for 90 min. (d) Supernatant of SWNT solution
collected after centrifugation at 55,000 rpm for 2 h

20

X. Wang et al.

2.2 Device Fabrication
A drop of purified SWNT dispersion solution was deposited on a marked heavily
doped p C Si=SiO2 (300 nm) substrate. After the solution was dried at room temperature, discrete SWNTs and groups were left on the surface of the substrate. Metal
electrode contacts were deposited at the ends of a single SWNT by using electron beam lithography and lift-off patterning (Fig. 2). Initial electrical testing was
carried out by sweeping the back-gate voltage from 10 to C10 V under a fixed
source-drain voltage at 1 V using an Agilent 4155C semiconductor parametric analyzer. Current–voltage (I –V ) measurements indicated that the SWNT was of p-type
(Fig. 3).

Fig. 2 (a) SEM image of SWNT field effect transistor fabricated with electron beam lithography.
(b) AFM image of another SWNT FET device

Current (uA)

3.0

2.5

2.0

1.5

Fig. 3 I –Vg measurements
of the SWNT FET for
Vds D 1 V with a gate oxide
thickness of 500 nm

−10 −8 −6 −4 −2

0

2

Voltage (v)

4

6

8 10

Hybrid Single Walled Carbon Nanotube FETs for High Fidelity DNA Detection

21

2.3 Synthesis of SWNT-ssDNA Conjugations and Detection
of Specific DNA Sequences
SWNT-ssDNA conjugations were formed by reacting the amine group at the end
of a single strand DNA with the carboxylic group on the surface of SWNTs via
the EDC coupling reagent. Since SWNTs were fixed by the metal electrodes on
the substrate, the substrate was immersed into the EDC solution for 30 minutes.
Amine functional group modified ssDNA (sequence: 50 -CTCTCTCTC-NH2 30 ,
from Sigma-Gynosis) and NHS-sulfo reagent were added to the solution. After incubating for 12 h, the sample was dried at room temperature. During the incubation
process, ssDNA molecules bound to the SWNT surfaces via amide linkage. After
obtaining an initial I –V measurement of the SWNT-ssDNA FET structure, it was
then immersed into a complementary strand DNA (cDNA) solution where fragments
with the complementary sequence of 50 -GAGAGAGAG-30 were hybridized to the
ssDNA at 42 C for 4 h. I –V measurements were conducted and the modulation of
the conductivity was recorded.

3 Results and Discussion
Commercial SWNTs were dispersed in dionized water, and a drop of dispersion
solution was dried on a silicon substrate and imaged as reference (Fig. 1a). A lot
of impurities, such as carbonaceous graphite particles, sonopolymers that were
involved during SWNT fabrication and acid oxidization are observed. Most of
SWNTs bundle together due to van der waals interactions between SWNTs. After sonication in o-DCB, a drop of sample was taken for SEM imaging (Fig. 1b),
indicating the dispersion of SWNTs becoming much better although impurities still
existed. According to our experience, o-DCB exhibits stronger -orbital interaction
with the sidewalls of SWNTs. During a sonication process, o-DCB molecules penetrate SWNT bundles by overcoming the van der waals interaction [18]. Therefore,
sonication of SWNTs in o-DCB is critical to obtain well dispersed SWNTs. In order
to remove the impurities, centrifuging with different speeds conducted. Centrifuging under low speed was performed first, followed by ultra-centrifugation under high
speed. Larger impurities settled down and were excluded after the first centrifugation step (Fig. 1c). With the centrifugation speed increasing, a decreasing number of
SWNTs with an increase in quality (much less impurities) as shown in Fig. 1d.
Purified SWNTs were deposited on a pC doped silicon substrate capped with
500 nm SiO2 . SWNT field effect transistors were fabricated via electron beam
lithography. Figure 2a shows the configuration of the device. A single SWNT was
fixed at both ends by metal electrode contacts patterned by electron beam evaporation. The contacts made in this way are reliable for a long time and can withstand
immersion in water bath [19]. Another sample is presented by AFM imaging in
Fig. 2b. Most of SWNTs after dispersion have a diameter of 15–20 nm, and are

22

X. Wang et al.

isolated from each other in a well dispersed manner. SWNT FET characterization
was carried out by measuring the current between source and drain electrodes under
gate voltage sweeping. I –V curve in Fig. 3 shows that the current is decreasing with
applying a positive voltage, which demonstrates that the SWNT in the FET is of a
p-type semiconductor.
Due to the carboxylic groups of SWNT, amino ended ssDNA readily binds to
SWNT under EDC coupling and NHS-sulfo reagents acting in the solution. After
ssDNA attach to the carboxylic group sites on the surface of SWNT, the functionalized SWNT was immersed into a target DNA (cDNA) solution. SWNT serves as the
semiconductor, and ssDNA bound along the surface of SWNT serves as the receptors for the target DNA fragments. I –V measurements of SWNT, SWNT-ssDNA
hybrids and SWNT-ssDNA-cDNA hybrids were recorded respectively. From the
I –V curves (Fig. 4), after ssDNA fragments covalently bind to the SWNT, the conductivity of the SWNT is reduced (Fig. 4, red) compared to that of before binding
(Fig. 4, black). We suggest that upon SWNT-ssDNA binding, geometric deformations occurs, leading to charge carrier scattering sites in the SWNT, hence the
reduced conductivity [20]. With the target DNA hybridizing with ssDNA, the conductivity increases (Fig. 4, green). The increase in conductivity is due to an increase
in the density of negative charges at the SWNT surface associated with the binding
of cDNA. In the sensor device, ssDNA serves not only as receptors for targets, but
also as the gate dielectric. When cDNA is added, ssDNA hybridizes with cDNA instead of binding to SWNT directly. cDNA molecules bear negative charges on their
backbone. Even though cDNA is dried during the measurements, residual water
molecules from the buffer solution are still adsorbed on DNA’s hydrophilic phosphoric acid backbone by forming hydrogen bonds [21], together with the cations
counterbalancing the negative charge of DNA [22]. Also, the effect of measurement
environment after DNA molecules dryed could not be ignored [23, 24]. Under a
high humidity level, water molecules would accumulate at the phosphate backbone
of DNA [24]. The electrical measurements in this paper are conducted under an ambient humidity level of 40%. Therefore, cDNA molecules bear negative charges with

SWNT

6

SWNT-ssDNA

Fig. 4 I –V curves of SWNT
before and after ssDNA
covalent binding (black and
red). I –V measurements of
ssDNA-SWNT nanohybrids
detecting the target DNA
(cDNA) is shown in green

Current (uA)

4

SWNT-ssDNA-cDNA

2
0
−2
−4
−6
−8

−5

−4

−3

−2

−1

0

1

Voltage (V)

2

3

4

5

Hybrid Single Walled Carbon Nanotube FETs for High Fidelity DNA Detection

23

water molecules surrounding them. cDNA hybridization with ssDNA is consistent
with applying a negative gate voltage on SWNT FET. Thus, the conductivity of
p-type SWNT increases when cDNA fragments hybridize to the ssDNA receptors.

4 Conclusion
SWNT-ssDNA based hybrid biosensor for the detection specific sequences of DNA
has been developed. SWNT is purified and well dispersed before conjugating with
ssDNA. SWNT FET measurements indicate a p-type semiconductor behavior. After
functionalized by amino-ended ssDNA, the SWNT FET is used for detecting target DNA molecules. Adding target DNA molecules, which hybridize with ssDNA
molecules on the surface of SWNT results in a significant modulation of SWNT
conductivity. The bio-sensing process is analogous to applying a negative bias voltage on the gate of SWNT FET. Therefore, the conductivity of SWNT increases. Our
results illustrate the promise of hybrid SWNT FETs for detecting a broad range of
biological and chemical species.
Acknowledgement The authors gratefully acknowledge financial support of this work by the
Center for Nanotechnology for the Treatment, Understanding and Monitoring of Cancer (NanoTumor) funded by the National Cancer Institute, and the Center for Hierarchical Manufacturing
(CHM) funded by the National Science Foundation.

References
1. Jimenez D, Cartoixa X, Miranda E et al (2007) A simple drain current model for Schottkybarrier carbon nanotube field effect transistors. Nanotechnology 18(2):Article No. 025201
2. Li H, Zhang Q, Li JQ (2006) Carbon-nanotube-based single-electron/hole transistors. Appl
Phys Lett 88(1):Article No. 013508
3. Kim BK, Kim JJ, So HM et al (2006) Carbon nanotube diode fabricated by contact engineering
with self-assembled molecules. Appl Phys Lett 89(24):Article No. 243115
4. Raychowdhury A, Roy K (2007) Carbon nanotube electronics: Design of high-performance
and low-power digital circuits. IEEE Trans Circ Syst I–Reg Pap 54:2391–2401
5. Kelly KF, Chiang IW, Mickelson ET et al (1999) Insight into the mechanism of sidewall functionalization of single-walled nanotubes: an STM study. Chem Phys Lett 313(3–4):445–450
6. Fagan SB, da Silva AJR, Mota R et al (2003) Functionalization of carbon nanotubes through
the chemical binding of atoms and molecules. Phys Rev B 67(3):Article No. 033405, 4 pages
7. Liu J, Rinzler AG, Dai HJ et al (1998) Fullerene pipes. Science 280(5367):1253–1256
8. Qu LW, Martin RB, Huang WJ et al (2002) Interactions of functionalized carbon nanotubes
with tethered pyrenes in solution. Jf Chem Phys 117(17):8089–8094
9. Dwyer C, Johri V, Cheung M et al (2004) Design tools for a DNA-guided self-assembling
carbon nanotube technology. Nanotechnology 15(9):1240–1245
10. Zhang JP, Liu Y, Ke YG et al (2006) Periodic square-like gold nanoparticle arrays templated
by self-assembled 2D DNA nanogrids on a surface. Nano Lett 6(2):248–251
11. Rothemund PWK (2006) Folding DNA to create nanoscale shapes and patterns. Nature
440(7082):297–302

24

X. Wang et al.

12. Guo ZJ, Sadler PJ, Tsang SC (1998) Immobilization and visualization of DNA and proteins on
carbon nanotubes. Adv Mat 10(9):701–703
13. Ito T, Sun L, Crooks RM (2003) Observation of DNA transport through a single carbon nanotube channel using fluorescence microscopy. Chem Commun (13):1482–1483
14. Gao HJ, Kong Y, Cui DX et al (2003) Spontaneous insertion of DNA oligonucleotides into
carbon nanotubes. Nano Lett 3(4):471–473
15. X. Wang, Liu F, Andavan GTS et al (2006) Carbon Nanotube-DNA nanoarchitectures and
electronic functionality. Small 2:1356–1365
16. Star A, Tu E, Niemann J et al (2006) Label-free detection of DNA hybridization using carbon
nanotube network field-effect transistors. Proceedings of the National Academy of Sciences of
the United States of America 103(4):921–926
17. Hu CG, Zhang YY, Bao G et al (2005) DNA functionalized single-walled carbon nanotubes
for electrochemical detection. J Phys Chem B 109(43):20072–20076
18. Kim DS, Nepal D, Geckeler KE (2005) Individualization of single-walled carbon nanotubes:
is the solvent important? Small 1(11):1117–1124
19. Cengiz Ozkan, Xu Wang (2008) Multisegment Nanowire Sensors for the Detection of DNA
Molecules. Nano Lett 8(2):398–404
20. Star A, Gabriel JCP, Bradley K et al (2003) Electronic detection of specific protein binding
using nanotube FET devices. Nano Lett 3(4):459–463
21. Lee Otsuka Y, Gu H-Y, Lee J.-H, Yoo J-O, Tanaka K-H, Kawai H, Tabata T (2002) Jpn J Appl
Phys 41:891; Ha DH, Nham H, Yoo KH et al (2002) Humidity effects on the conductance of
the assembly of DNA molecules. Chem Phys Lett 355(5–6):405–409
22. Otto P, Clementi E, Ladik J (1983) The electronic-structure of DNA related periodic polymers.
J Chem Phys 78(7):4547–4551; Lewis JP, Ordejon P, Sankey OF (1997) Electronic-structurebased molecular-dynamics method for large biological systems: application to the 10 basepair
poly(dG)center dot poly(dC) DNA double helix. Phys Rev B 55(11):6880–6887; York DM,
Lee TS, Yang WT (1998) Quantum mechanical treatment of biological macromolecules in
solution using linear-scaling electronic structure methods. Phys Rev Lett 80(22):5011–5014;
Ye YJ, Jiang Y (2000) Electronic structures and long-range electron transfer through DNA
molecules. Int J Quant Chem 78(2):112–130
23. Jo YS, Lee Y, Roh Y (2003) Effects of humidity on the electrical conduction of lambda-DNA
trapped on a nano-gap Au electrode. J Korean Phys Soc 43:909–913
24. Kleine-Ostmann T, Jordens C, Baaske K et al (2006) Conductivity of single-stranded and
double-stranded deoxyribose nucleic acid under ambient conditions: the dominance of water.
Appl Phys Lett 88(10):Article No. 102102

Towards Integrated Nanoelectronic
and Photonic Devices
Alexander Quandt, Maurizio Ferrari, and Giancarlo C. Righini

Abstract State of the art nanotechnology appears like a confusing patchwork of
rather diverse approaches to manipulate matter at the nanometer scale. However,
there are strong economic and technological driving forces behind those developments. One key technology consists of a rather dramatic shrinking of integrated
electronic devices towards the very size limits of nanotechnology, just to satisfy
the growing demand for commonly available computing power. Furthermore, the
corresponding step from microelectronics to nanoelectronics pushes another important technological sector, which aims at the development of novel optical devices,
that ought to furnish the bandwidth and speed to ship the plethora of accumulating
processing bits. In the following, we point out some of the basic technological challenges involved, and present a selection of experimental and numerical approaches
that aim at the development of novel types of optoelectronic nanodevices.

1 Introduction
Nanotechnology has become a common buzzword for a general technological
development, that promises to make our lives easier and longer. It is based on
our unique abilities to manipulate matter at the atomic scale. But even the biggest
enthusiast of nanotechnology might become rather thoughtful, after putting away
Drexler’s Engines of Creation [11], and browsing in Hero of Alexandria’s Pneumatics [43], which stems from the first century AD. How could it be, that it took
almost 1,700 years until the steam engine finally initiated the industrial revolution,
A. Quandt ()
Institut für Physik, Universität Greifswald, Felix-Hausdorff-Str. 6, 17489 Greifswald, Germany
e-mail: alexander.quandt@physik.uni-greifswald.de
M. Ferrari
CSMFO Lab., CNR-IFN Trento, Via alla Cascata 56/C, 38100 Povo, Italy
e-mail: mferrari@science.unitn.it
G.C. Righini
CNR-Nello Carrara Institute of Applied Physics, MDF Lab, 50019 Sesto Fiorentino, Italy
e-mail: giancarlo.righini@cnr.it

D. Baleanu et al. (eds.), New Trends in Nanotechnology and Fractional
Calculus Applications, DOI 10.1007/978-90-481-3293-5 3,
c Springer Science+Business Media B.V. 2010


25

26

A. Quandt et al.

although the basics of steam power had already been known at the time of Hero?
The rather sobering answer might be that at the time of Hero, and for a long period
of time thereafter, the huge amounts of power provided by the steam engine were
simply not needed. Because manpower was abundant, due to slavery and serfdom.
In the dawning age of nanotechnology, it might be rather bewildering to see the
same major driving forces at work, which actually led to slavery and serfdom at
the time of Hero: comfort, health, business and entertainment. With the distinction
that nanotechnology should make those things available to everyone. In fact, there
is hardly any part of human life that would not sooner or later go online, and the
corresponding need for bandwidth and higher bit rates is growing enormously. As a
measure for the technological evolution of optical networks, one usually considers
the product of the length L times the maximum bit rate B0 of the communication
link. It turns out that LB0 is approximately increasing by a factor of ten every four
years (optical Moore’s law [39]).
Recently, a group of researchers at Nippon Telegraph and Telephone Corporation
(NTT) carried out a study of cutting edge optical fiber communication technologies
[17]. They reported a bandwidth of 14 TBits/s transmitted over 160 km of optical
fiber, which involved wavelength and polarization multiplexing techniques [39],
using 140 channels within a window of 1,450–1,650 nm wavelengths of the optical
carrier waves. This amounts to a bit rate of 111 Gbit/s per channel, which is more
than double the maximum line rate that is commercially available now. Those results
are most likely the prelude to a novel 100 Gigabit Ethernet standard, which is badly
needed, in order to satisfy the growing need for broadband-access lines. And to
provide the necessary flexibility in handling novel types of internet based services
like file swapping or video sharing, which are extremely bandwidth intensive, and
rather unpredictable [17].
Attached to the nodes of a rapidly growing global communication internet are
novel types of computers with rapidly shrinking processor units. This massive integration process more or less follows Moore’s law [23], which states that the number
of transistors per square centimeter doubles every 12 months. It is quite obvious
that such a development will sooner or later hit the limits of nanotechnology itself,
which are located in the Å domain (1 Å D 1010 m), being the typical size range
of single atoms.
By now the most recent processor generations are already based on transistor
technologies with gate lengths in the range of several dozen nanometers (1 nm D
109 m). To maintain the reliability of established microfabrication techniques
at such a tiny length scale represents a formidable technological challenge [35].
Furthermore the miniaturization of transistors implies a dramatic increase in switching speed, such that signal propagation delays in the interconnects between transistors become a real issue, and optical interconnects ought to come to the rescue of
chip design [1].
The proper integration of optical interconnects will pose a serious problem for
future integrated nanocircuits. Following Moore’s law, the expected gate lengths
of the electronic elements might rapidly drop below the 10 nm range [15], whereas
optical interconnects might need to stay compatible with the standards of long-range

Towards Integrated Nanoelectronic and Photonic Devices

27

optical interconnects, and use standard wavelengths in the infrared domain around
1,550 nm. Optical devices that ought to interact with standard optical carrier waves
must be of nearly the same size range, which will be orders of magnitudes larger
than the electronic components of future integrated optoelectronic chips.
In the following, we will present a selection of experimental and numerical approaches to develop and characterize basic electronic and optical devices, which
might be most valuable in the design of future integrated optoelectronic chips. In
Sect. 2 we will discuss the practical limits of MOSFET design within the nanodomain, and present alternative design approaches based on carbon nanotubes [2]
and graphene [12]. In this context, we will also illustrate the rather important role
of numerical simulation methods [34]. In Sect. 3 we will study some of the key
elements for the integration of optical devices like VCSEL photonic sources and
photonic crystal waveguides, and present some experimental and numerical approaches [18, 39] to optimize their basic functionalities.
Finally, we will summarize our findings in Sect. 4. Note that our selection of
topics is neither intended to be exhaustive nor representative. Nor might it be completely unbiased. Our main goal here is to point out some of the most promising
starting points for one’s own experimental or numerical access to the development
of novel electronic and optical devices for the nanotechnology era.

2 Integrated Electronic Devices
Metal oxide semiconductor field-effect transistors (MOSFETs) as depicted in
Fig. 1(a) are today’s standard workhorses of integrated electronics. The layer-bylayer layout of complex networks containing billions of transistors on the area of a
single chip requires a permanent refinement of Very Large Scale Integration (VLSI)
techniques [26]. As VLSI is based on the optical projection of photomasks, state of
the art VLSI already involves projection techniques using electron-beam lithography or illumination wavelengths within the deep UV range [16]. The latter may be

Fig. 1 Towards nanoelectronics. (a) Basic structure of a MOSFET transistor. (b) FET transistor,
which involves semiconducting carbon nanotube channels

28

A. Quandt et al.

combined with immersion and double-projection techniques to create feature sizes
of the order of a few dozen nanometers, which are well below the wavelengths of
visible light, and close to the Rayleigh-limit of optical projection methods [41].
What about physical limits on nanometer sized MOSFETs? Thanks to Moore’s
law, there has always been a bulk of literature on the projected scaling of such devices [25], and about the technological problems involved in downsizing the basic
MOSFET design [15, 27]. We will discuss some of those issues in Sect. 2.1. Then
in Sect. 2.2 we will illustrate some of the improved device characteristics for FETs
with nanotubular components. Furthermore, alternative semiconducting substrates
like graphene [12,31] might actually allow for a further extension of Moore’s law towards the very size limits of nanotechnology, using a cluster based nano-patterning
approach described in Sect. 2.3.

2.1 Scaling of MOSFET Devices
The basic layout of a MOSFET transistor is sketched in Fig. 1(a). It consists of
source and drain contacts, which are doped, and conducting diffusion layers isolated
by a semiconducting substrate. The third contact called the gate is also conducting,
and it is separated from the other components by a thin insulating layer. As long
as the gate voltage is low, there is no current flowing from the source to the drain,
due to the semiconducting properties of the substrate. But once the gate voltage
overcomes a certain threshold voltage, there will be current flowing, as soon as we
apply an appropriate electric field between the source and the drain.
In order to run such a transistor with technologically appealing device characteristics, one has to dope it in a systematic fashion, such that the doping of the substrate
shows an opposite polarity to the source and drain. This effectively creates two backto-back junction diodes. A suitable voltage Vgs applied through the gate will pull
mobile carriers (electrons or holes) to the underside of the metal oxide layer, thus
opening a conducting channel through the substrate. Once the voltage is turned off,
the surface under the gate will be depleted of carriers, and no current will be able to
flow any more.
The gate/oxide/substrate sandwich of gate length L, width W and thickness D
may be pictured as a capacitor with dielectric constant " and capacitance Cg D
"W L=D. Thus there is a charge Qg D Cg Vgs accumulating in this conducting
channel. Once a voltage Vds is applied between the source and the drain, there is
a current Ids flowing that experiences a resistance R D LD="W Vgs . An elementary derivation of these results can be found in [14]. A rough estimate of the
device speed is related to the time constant  of a model RC circuit with the same
characteristics:
 D RCg D

L2
Vgs

(1)

Towards Integrated Nanoelectronic and Photonic Devices

29

Thus, shorter discharge times  may be obtained by shrinking the gate length L.
However, this implies a shorter distance between the source and the drain, which
might lead to difficulties in switching off an operating device. This could still be
avoided using massive doping. Another possibility to increase device speed would
be through an increase of the mobility  for the carriers that travel from the source to
the drain, for example by straining the substrate [15]. A third possibility would be to
increase the gate voltage Vgs . But the metal oxide layer is already close to its physical limit (around 1 nm), and increasing the gate voltage will lead to leakage currents.
Furthermore the power consumption will sharply increase, as the switching power
is proportional to the operating frequency f , and to the dynamic switching energy.
The latter may be estimated from:
ED

1
2
.Cg C Cw /Vgs
2

(2)

where Cw is the effective capacitance of the wiring, which is a rather complex metaldielectric interconnect structure.
Let us consider a single metal line with contact capacitance Cw , and of length L
and diameter A. Then we note that the corresponding resistance R  L=A will
obviously increase with shrinking line diameter A, leading to longer signal delay
times related to  D RCw . This signal delay will not be an issue for similarly
shrinking local interconnects with small lengths L, but it will become a serious
problem for the much longer global interconnects, which ought to join important
parts of a processor [35].
Here we close our short discussion of basic design problems for MOSFET
transistors within the nanodomain. Alternative design concepts like the FinFET
transistor are shortly described in [23], and a more detailed description of ultimate
device limits may be found in [27].

2.2 Nanotube Transistors and Interconnects
An alternative road to the design of nanoelectronic devices is the employment of
semiconducting carbon nanotube (CNT) channels as integral part of a working FET,
which is indicated in Fig. 1(b). Carbon nanotubes may be pictured as rolled up versions of rectangular strips cut out of a single layer of graphite called graphene [10].
A single graphene layer consists of carbon atoms located on the vertices of a honeycomb lattice. Depending on the direction of the cut, the resulting carbon nanotubes
exhibit different chiralities, which influence their basic electronic properties quite
strongly (i.e. metallic vs. semiconducting [10]). Unfortunately, it is hard to control
this chirality during the synthesis of CNTs.
The major advantages of implementing semiconducting carbon nanotubes as
FET channels have recently been pointed out in [2]: the nanotube channel is quite
small (1–2 nm) and atomically smooth, the carrier mobilities are very high at low

30

A. Quandt et al.

gate voltage, and the capacitance of CNTs is rather low. The gap size of semiconducting CNTs is inversely proportional to their diameters, which allows for a rather
flexible use of CNTs as basic nanoelectronic components [10]. Furthermore, CNTs
dispose of a rather favorable optical properties, to be discussed in Sect. 3.
Therefore the integration of CNT components might allow for novel high-speed,
low power and nanometer sized FET and optoelectronic devices discussed in [2].
However, major technological challenges are represented by a controlled layout, a
method to separate metallic/semiconducting CNTs during synthesis, and a systematic control of contact barriers between CNTs and the source/drain of the MOSFET
device shown in Fig. 1(b). Note that the contacts of a CNT based FET are usually
made of metals.
Partial technological solutions for some of these problems are discussed in [2].
But progress may also be made through the employment of boron nanotubes (BNTs)
[36]. Those materials are the brainchild of extensive numerical simulations on small
boron clusters [5], which suggested [6] the existence of stable boron sheets (i.e. the
boron analogue of graphene) and boron nanotubes (i.e. the boron analogue of carbon
nanotubes shown in Fig. 2(a)).
Note that numerical simulations on unknown boron nanomaterials are far from
trivial. Those materials were outside the horizon of standard textbook wisdom [33],
and therefore the tedious identification of stable ground state configurations of planar and tubular boron clusters required the usage of ab initio simulation methods
at the highest level of numerical accuracy (for a survey of such methods see [34]).
Nevertheless, these earlier results not only stimulated the successful synthesis of
BNTs [9], alongside a plethora of novel types of semiconducting boron nanowires
(see [36]). But they were also the basis of recent refinements of the atomic structures
of BNTs, based on a remarkable hole-doping scenario [40].
There is now some general consensus about a number of very favorable properties for nanotechnological implementations of boron nanotubes, as pointed out

Fig. 2 Tubular carbon–boron interconnects. (a) Model armchair (top) and zigzag (bottom) boron
nanotube (BNT). (b) Strong dependence of elastic properties on the chirality of various BNTs [22].
(c) Stable boron–carbon heterojunction (CNT at the top, BNT at the bottom) [20]

Towards Integrated Nanoelectronic and Photonic Devices

31

in [36]: first of all, BNTs should always be metallic, independent of their chirality, which would make them perfect conducting nanowires. On the other hand, the
elastic properties of BNTs are strongly dependent on their chirality, as shown in
Fig. 2(b) and pointed out in [22]. This is most obvious from the constricted nature of
the zigzag BNT indicated at the bottom of Fig. 2(a), as compared to the stable round
structure of the armchair BNT shown at the top of Fig. 2(a) (for details see [21]). In
contrast to CNTs, the mechanical properties of BNTs might actually be controlled
during synthesis, thus leading to some control over their chiralities [22].
Furthermore, similar bond lengths and the electron deficient nature of boron
should make boron based nanomaterials largely compatible to carbon nanomaterials, to silicon substrates, and to all sorts of metallic wirings. The basic compatibility
between carbon and boron nanomaterials has already been demonstrated in numerical simulations of nanotubular carbon–boron heterojunctions [20]. One exemplary
metallic carbon–boron junction is shown in Fig. 2(c), and it consists of a CNT on
top, and a BNT at the bottom. Another interesting feature of such junctions is the fact
that they might easily be formed by excessive doping of CNTs with boron atoms:
simulations revealed that boron atoms have a strong tendency to migrate towards
the open ends of CNTs [13], where they grow BNT type of extensions.
Note that the formation of stable heterojunctions between BNTs and CNTs could
actually induce a similar structure control over the CNT components, simply by controlling the BNT segments attached to them (see [22]). Therefore BNT–CNT based
networks could become vital components of future nanotube based FET design,
where the metallic boron component might be responsible for structure control, as
well as for stable interconnects with the outside world.

2.3 Ultimate Integrated Devices Based on Graphene
The scaling of important device properties for integrated nanoelectronic circuits
described in Sect. 2.1 points towards thinner and thinner MOSFET devices. One
ultimate technological limit would be the controlled layout of integrated circuits on
a 2D semiconducting substrate. Lucky enough, a suitable substrate material has already been identified in terms of graphene [31]. This term denotes a whole family
of nanomaterials, which consist of (irregularly shaped) flakes of carbon monolayers,
cut from a basic carbon honeycomb sheet sketched in Fig. 3(a).
Small amounts of graphene may be produced in a disarmingly simple fashion,
using adhesive tape to gradually cleave small flakes of graphite into thinner and
thinner fragments (for a Do It Yourself description of this process see [12]). The
electronic properties of graphene flakes depend on the nature of their borders [30],
but a safe bet is to either obtain semiconducting flakes from scratch, or otherwise
turn a given flake into a semiconducting one by manipulating its borders. Note that
the mobility of conducting electrons within graphene is very high. Furthermore, the
conducting electrons seem to move ballistically, i.e. without being scattered by the
carbon atoms of the underlying honeycomb lattice [12].

32

A. Quandt et al.

Fig. 3 Graphene based nanoelectronics. (a) Honeycomb lattice of single graphene layer. (b) Chain
of B7 -clusters embedded into semiconducting armchair graphene nanoribbon (top). Note that this
system is supposed to be periodic in the y-direction, such that the boron clusters are not directly
connected, but separated by a full carbon honeycomb. This functionalization nevertheless induces
conducting channels inside the gap of the undoped graphene substrate (bottom). (c) FET type of
wiring and basic functionalization of a semiconducting graphene substrate, based on unconnected
chains of embedded boron clusters

Recent numerical simulations [38] uncovered a way to functionalize semiconducting graphene sheets, based on conducting nanowires only a few atoms thick.
The corresponding model system is shown at the top of Fig. 3(b). It consists of
a small hexagonal B7 -clusters being embedded into a semiconducting rectangular graphene nanoribbon with armchair borders. Note that the structure shown in
Fig. 3(b) is actually periodic in the y-direction, such that neighboring B7 -clusters
are not directly connected, but separated by a full carbon honeycomb. Nevertheless,
we notice the appearance of conducting channels in the gap of the original semiconducting substrate, as shown at the bottom of Fig. 3(b).
This points towards a very robust way of laying out a basic wiring within a
semiconducting graphene substrate, like the basic FET type of blueprint shown in
Fig. 3(c). The fact that the boron clusters were not directly connected in the model
system of Fig. 3(b) also suggests that even a rather irregular and heterogeneous embedding of boron islands might work in practice. Therefore the cluster embedding
could in principle be carried out with state of the art technological equipment, as
described in [37].

3 Photonic Devices and Interconnects
Optical interconnects are a well-established high-bandwidth technology for longdistance communication. And the dreaded internet bottlenecks are not related to
occasional cable accidents on the high seas, but rather to data traffic jams on the
notorious last mile, where optical fiber technologies are only gradually replacing
electrical or wireless carriers. On the other hand, due to Moore’s law and modern
multi-core processor architectures, the requirements for on-board IO bandwidth will

Towards Integrated Nanoelectronic and Photonic Devices

33

rapidly approach the range of 1 Tbit/s (and beyond), thus shifting the bottlenecks
to the last (centi-)meter. As pointed out in [1], commercially available processors
are now featuring data buses that operate at 1.3–1.4 GHz, and at bandwidths around
17 Gbit/s. Experimental electrical transmission lines slightly surpass this value, but
for achieving bandwidths in the range of Tbit/s over the area of a microchip, electrical transmission looks rather insufficient.
In order to get more insight into this problem, we again pick up our theoretical
considerations about the scaling of electrical interconnects from Sect. 2.1. But at
this point, we would like to be a little bit more precise: let us once again consider
a wire of length L and area A. Again we find that the resistance should scale as
R  L=A. But for the capacitance Cw of such a wire, we will now assume a simple,
but much more realistic linear dependance Cw  L, which presupposes a chosen
standard geometry for such a wire. This leads to a time constant  of
 D RCw 

L2
A

(3)

A conservative estimation [28] for the corresponding bandwidth B, which ought to
be inversely proportional to the time constant , amounts to
B  1016

A
bit/s
L2

(4)

This result is rather interesting, for various reasons: if we play around with some
realistic figures, where a global interconnect might have a cross-section of 1 m2
and a length of 1 mm or less, we will see that we are talking about bandwidths
in the range of MBit/s or Gbit/s, rather than the targeted Tbit/s. Even worse, the
geometrical factor A=L2 in Eq. 4 tells us that, as long as we have to stick
p to a
certain standard wire geometry characterized by a fixed aspect ratio L= A, we
will never be able to increase the corresponding data bandwidth by only an overall
miniaturization of such a wire.
According to [1], a conventional integrated photon source (see Sect. 3.1) will
easily deliver a bandwidth of 10 Gbit/s and more, and that bandwidth will be largely
independent of the length of the transmission line, at least on the length scale of
on-chip optical interconnects. By combining whole arrays of such devices, it should
be possible to achieve Tbit/s bandwidths with device technologies that might be
implemented into the standard VLSI type of chip production. Other advantages of
optical interconnects are the reduction or absence of inter-channel crosstalk and
electromagnetic interference [1, 28]. It allows for much denser packings of integrated optical waveguides, thus leading to a dramatic increase in the available
bandwidth. Furthermore, optical technologies may in principle be implemented with
lower power expenditures. For a detailed discussion of the technological aspects of
optical interconnects, we refer the interested reader to [28].
Nevertheless, it should not be concealed here that the basic technology for
integrated optical interconnects is still in its infancy. And that the candidate materials, devices and fabrication processes still have to be explored and gradually

34

A. Quandt et al.

improved. Therefore we will only be able to sketch some blueprints for integrated
optical interconnects in Sect. 3.1. In Sects. 3.2 and 3.3 we will present some basic
photonic components, which could become key elements of future optical circuits.
Section 3.2 will be devoted to photonic sources, where we will also present some
experimental and numerical results about Bragg reflectors and microcavities, which
are vital parts of such devices. Furthermore, in Sect. 3.3 we will illustrate optical
waveguiding based on photonic crystals.
Finally we would like to point out that other promising technological solutions
to achieve nanometer sized integrated optoelectronic circuits may be provided by
plasmonics [3, 24, 32]. The latter denotes a very fascinating branch of photonics,
that is dealing with processes that take place at the interface between a dielectric
medium and a metallic medium. It basically turns out that light, which is hitting such
a metal/dielectric interface, may stimulate coherent electron oscillations known as
surface-plasmon polaritons (SPPs). It is interesting to note that even at the time of
Hero, artists were already make use of plasmonics to achieve the most vibrant colors
for their glass ware. A brief survey over the most recent technological applications
of metal nanoparticles embedded into a surrounding dielectric medium is given by
[24].
Interesting enough, the wavelengths of these SPPs are actually smaller than the
wavelengths of the incident light, which offers the possibility to shrink optical technologies towards the size ranges of integrated electronic devices [3, 24]. The SPPs
rapidly decay into both media, and unfortunately they will only be able to propagate for a limited range along the interface. Nevertheless, a lot of progress has been
made to extend the range of SPPs towards technologically interesting distances, and
the first working plasmonic devices even give reason to dream of future integrated
all-photonic (or better: all-plasmonic) chip technologies [32].

3.1 Optical Interconnects
The application of optical technologies in the framework of nanoelectronics should
mainly provide sufficient bandwidth for global data interconnects on a given chip.
Like any conventional optical transmission system, on-board optical interconnects
will be composed of three major components: photonic sources, photonic links and
photonic detectors [39]. Photonic sources, like light emitting diodes (LEDs) or
lasers (see Sect. 3.2), as well as the corresponding photo-detectors ought to be directly coupled to the electronic circuitry, with acceptable signal-to-noise ratios. The
specific technological requirements for these components are summarized in [28].
Some all-silicon based solutions would certainly have the advantage to be easily integrable into conventional chip production processes, and some progress has
already been made in that direction [1]. Nevertheless, the standard technology for
photonic sources or detectors involves III–V semiconductors [1], and due to the lattice mismatches between those materials and silicon, the corresponding integrated
optoelectronic chips have to be produced outside the silicon fabrication process

Towards Integrated Nanoelectronic and Photonic Devices

35

itself. Later on, the completed optoelectronic chips must be mounted on the electronic chips, in a process that is called flip–chip bonding [39].
Channels for optical data transmission might be conventional fibers, free-space
transmission, or optical waveguide links directly built onto or into the substrate of
the processor board. Optical fibers are most suitable for straight linear interconnects,
but for microelectronic applications they have to be modified in order to sustain elevated temperatures for a long period of time [1]. For free space routing, there
are interesting solutions, which involve external reflection holograms [39]. For onboard waveguiding, one faces the usual technical problems of confining an optical
mode, which strongly depends on the refractive index of the available materials.
Furthermore one might need to guide modes around sharp bends, which may actually be solved using photonic crystals, as described in Sect. 3.3.

3.2 Photonic sources and Bragg Reflectors
Light emission by solid-state semiconductor devices is based on electron-hole
recombination. The perfect setup is a junction with a p-doped semiconductor on
one side, and an n-doped semiconductor on the other side (p-n-junction). Operating
such a device in a forward bias mode (i.e. positive voltage at the p-side, and negative voltage at the n-side), one can achieve a sufficiently high recombination rate to
see more than just a feeble glow, which is called injection electroluminescence.
The standard light-emitting diode (LED ) is operating on this principle of injection electroluminescence. The good news is that such devices may be miniaturized
to a degree that will allow for their combination with microelectronic circuits, where
electronic data streams will directly modulate the LED to allow for optical transmissions (electronic-to-optical transducers [39]).
Note that the radiation from a standard LED will be incoherent, as long as the
recombination will only be caused by spontaneous emission. In order to achieve optical amplification or laser action, one has to accomplish some population inversion.
The latter might be achieved by raising the forward voltage beyond a certain level,
or by some optical pumping, for example when using rare-earth dopants like Er 3C
as active components. The pumping will excite electrons from the valence band to
the conduction band of the active component, where these electrons will undergo a
fast non-radiative transition to a lower, and long-lived excited level. A photon of just
the right energy will then be able to induce a recombination of the corresponding
electron-hole pair, and thus clone a second photon of the same phase and energy.
This process is called stimulated emission, and it leads to coherent radiation.
In order to achieve constant laser action, one has to provide a certain feedback, where photons are kept long enough inside the active zone to cause whole
avalanches of photon clones. This may be achieved by using Bragg reflectors. The
latter are stacks of dielectric layers i with effective thickness di , and with varying refractive index ni , as shown in Fig. 4. When light of a certain wavelength 

36

A. Quandt et al.

Fig. 4 Photonic devices. (a) Two Bragg reflectors and a larger spacer layer, which form a photonic
microcavity. (b) Principal layout of a vertical-cavity surface-emitting laser (VCSEL). The whole
system is mounted on a substrate, and the reflectance of the laser mode by the Bragg reflectors is
rather high. Laser modes are generated through electron-hole recombinations inside the pumped
spacer layer, and the resulting coherent light will escape vertically, as indicated

impacts on such a structure, partial reflectance and constructive interference may
occur whenever  D 4ni di (“quarter wave reflection grating”). This might add up
to total reflectance [4], the so-called stop bands. The aim of a related optical microcavity shown in Fig. 4(a) is to trap standing waves inside a spacer layer between two
Bragg reflectors.
Now we dispose of all the basic photonic components to devise a suitable
coherent light source for nanotechnological purposes. It is called vertical-cavity
surface-emitting laser (VCSEL) , and the basic blueprint of such a device is shown
in Fig. 4(b). A VCSEL is a sandwich of an active lasing area between two Bragg
reflectors mounted on a suitable substrate, where one reflector is n-doped, and the
other reflector is p-doped. Recombination takes place inside the active area, which
may be pumped electrically. The microcavity provides the necessary feedback for
laser action, and the coherent radiation escapes vertically, as indicated in Fig. 4(b).
Such VCSEL devices have the distinct advantage that they may be fabricated in a
layer-by-layer fashion, with standard layout techniques known from microelectronic
VLSI, and in a size range that will allow for their implementation as active parts of
on-board optical interconnects [39]. Nevertheless, the optimization of the various
photonic components of a VCSEL for microelectronic purposes is still the subject of
ongoing research. In Fig. 5(a) we show a picture of a photonic microcavity made of
S iO2 and T iO2 , which disposes of an Er 3C doped spacer layer [7]. Such a device
could act as an optical filter and amplifier. The corresponding normal transmittance
is shown in Fig. 5(a), where we recognize a broad stop band with a cavity mode at
1,544 nm.
Note that active VCSEL components may be analyzed in a similar fashion.
Further improvement in the design of such devices might actually come from numerical simulations [39]. An example is shown in Fig. 5(b), and it refers to the
model microcavity of Fig. 4(a). We just assumed that the index of refraction for the
two media is n D 1 and n D 2, and that the thickness of the plates is d D 2=3
and d D 1=3 in arbitrary units. The spacer layer is characterized by n D 2 and

Towards Integrated Nanoelectronic and Photonic Devices

a

b

1.5

Banddiagram

1.5

1

1

0.5

0.5

Transmission

ωa/2πc

Transmittance [%]

60

37

50
40
30

1544 nm

20
10
0
1000 1200 1400 1600 1800 2000 2200
Wavelength [nm]

0
0

0.5
ka / π

1

0
0

0.5
T

1

Fig. 5 Microcavity. (a) Bragg reflectors made from SiO2 (black) and T iO2 (grey), where the
spacer layer is doped with Er 3C -ions. The corresponding normal transmittance shows an ample
stop band around the cavity mode [7]. (b) A supercell simulation (left) of an infinite model Bragg
reflector shows band gaps and cavity modes (horizontal line). The transmittance (right) for a finite
and air terminated Bragg reflector is marked by a (near) perfect reflectance of the stop bands, and
the appearance of a sharp cavity mode

d D 2=3. If we extend this model periodically in the framework of a supercell
model, the propagating electromagnetic modes are B