Enano newsletter issue22
1 like5,806 views
This document discusses recent developments in atomic scale and single molecule logic gate technologies as part of the European integrated project ATMoL. It highlights the project's objectives in molecular computing, joint academic collaborations, and the creation of nano-electronic devices, focusing on heat dissipation in nanometer-scale structures and advancements in atomic scale circuit fabrication. The document presents various articles and reports, emphasizing the innovations and challenges in the field.
![nanoresearch
Heat dissipation in nanometer-scale
ridges
Pierre-Olivier Chapuis(a)*, Andrey to electric and heat conduction, namely
Shchepetov(b), Mika Prunnila(b), Sampo particles free path and even wavelength in
Laakso(b), Jouni Ahopelto(b) and Clivia M. some cases. The case of the heat carriers
Sotomayor Torres(a)(c)(d)* is particularly of interest because it involves
(a) Institut Catala de Nanotecnologia (ICN), Centre phonons that have free paths estimated
d’Investigacio en Nanociencia e Nanotecnologia to be larger than the device characteristic
(CIN2-CSIC), Edificio CM3, Campus de la length in some cases, undergoing fewer
Universitat Autonoma de Barcelona,
collisions as a consequence. The so-called
08193 Bellaterra (Barcelona), Spain.
(b) VTT Technical Research Center of Finland, phonon rarefaction effect is then responsible
PO Box 1000, 02044 VTT, Espoo, Finland. for higher temperature in the heat source
(c) Department of Physics, Universitat Autonoma de
than what can be estimated with the usual
Barcelona, 08193 Bellaterra (Barcelona), Spain. Fourier-based heat conduction equation
(d) Institució Catalana de Recerca i Estudis
Avançats (ICREA), Passeig Lluís Companys, 23
[2]. Larger temperature gradients are then
08010 Barcelona, Spain. responsible for cracks as differential thermal
expansion at material interfaces is not
manageable.
Abstract > Heat management in today’s A key parameter in the improvement of the
electronic devices is critical to prevent device design is therefore the understanding
possible failures due particularly to cracks. of heat conduction in the devices. Here
Heat is carried in such devices by electrons we focus on one way to measure the
in electrical conductors and mainly by implications of the particle behaviour (mean
acoustic phonons in electrical insulators. free path) of acoustic phonons when they
We explain the first steps of our work are confined in tiny structures. We propose a
aiming to investigate experimentally the method to analyse the heat flux propagation
effect of lateral confinement of acoustic in these structures and show our first efforts
phonons in silicon ridges as a function of towards the efficient use of the designed
the temperature. Inspired by the electrical samples.
3ω method, we design a setup that can be
used as a mean to generate phonons in 2. Heat flux measurements in low-
~100 nm wide ridge nanostructures and dimensional samples
as a thermometer that allows tracking the
generated heat flux. A. The macroscopic 3ω method
The 3ω method has been developed since
1. Introduction the 1980s, in particular by Cahill [3], with the
Works performed over the last decades goal of studying the thermal conductivity
have shown that electronic devices with of planar materials. Thin films have been
nanometre-scale dimensions are subject investigated widely as well as the thermal
to larger temperature-driven stresses in boundary resistance between the films [4].
comparison to what had been estimated in It was extended for multilayer materials or
the past [1]. In particular, the size of different particle-based materials [5, 6]. The method
components of transistors and electronics is based on the Joule heating of a metallic
devices present in printed circuits are now wire of micrometric size deposited on top of
12 comparable to the key scales associated the substrate that “steals” part of the heat](https://image.slidesharecdn.com/enanonewsletter22versionfinal-111026044835-phpapp02/85/Enano-newsletter-issue22-12-320.jpg)
![nanoresearch
flux generated. As a consequence, the substrate thermal conductivity has to be
temperature of the wire gives an indication found. A 2D cross-section based model
on the ability of the sample to conduct heat. has been used extensively over the past
The use of a harmonic current to heat the 20 years. It is based on frequency sweeps.
wire allows the excitation of the temperature The slope of the temperature variation gives
higher-order harmonics. If the heating is not the thermal conductivity [3]. Some authors
too high, only the third harmonic is excited have underlined that better models can be
as will be seen in the following. One can used [7, 8]. Usually, the width of the wire is
write for a current I = I0 cos(ωt) the generated in the micrometric range, which is reached
power due to the Joule heating as with standard optical lithography in the
fabrication process.
P(t) = R I(t)2 = ½ R (1 + cos(2ωt))
and the total temperature reads including B. Implementation at the nanoscale
the heating TDC by DC current : The case of nanoelectronic devices is very
different to the micrometre-scale one. Even
T(t) = T0 + TDC + T2ω cos(2ωt+ 2ω). if the heating/sensing system that can be
The key point now is the dependence of used has the same principle, the sizes are
the wire resistivity to temperature, that is much smaller. We fabricated nanostructures
linear in first approximation for low heating: where the heater/sensor lies on top of silicon
R(T) = R0 (1+α ∆T). Finally, the voltage of the substrates as represented in Fig. 1. The top
wire can be written as of a ridge is a wire, either a metal or doped
silicon, which acts as a heater and as a
U = RI = R0I0 [1+α(TDC+T2ωcos(2ωt+ 2ω))] cos(ωt) thermometer at the same time. The doped-
= R0I0 [(1+αTDC) cos(ωt)+½αT2ωcos(ωt- 2ω) layer structure requires epitaxial growth of
+½αT2ωcos(3ωt+ doped silicon.
2ω)]
The use of a lock-in amplifier at the third
harmonic enables to measure the amplitude
and the phase of the third harmonic and thus
extract the local temperature. The amplitude
is U3ω = ½ αR0I0 T2ω. The frequency range
to be used here is generally between 10 and
5000 Hz. Care has to be taken with the wire
width and thickness that should be smaller Fig. 1 > Two types of resistive heater for the ridge
than the thermal diffusion wavelength in order experiments./
to prevent from a possible nonhomogeneity The substrates can be made of high-
of the heat generation in the wire. resistivity silicon. The submicrometer ridges
This experimental part of the work permits are fabricated with electron beam lithography
only to get a qualitative idea of the material and ICP dry etching (see Fig. 2, page 14).
thermal properties or to make an estimate This type of structure enables to generate
based on comparisons with reference phonons in the ridge and to measure the heat
materials the thermal conductivity of which is flux flowing to the substrate. An adaptation
known. This is not an easy task as a heating of the 3ω method is then used to heat the
device has to be deposited on top of each wire and measure the wire temperature. As
sample. it has been previously explained, a harmonic
If one wants to find the thermal conductivity electrical current generates the heat at 2ω
directly from the sample, a physical model due to Joule effect and lock-in detection
linking the measured temperature and the allows measuring the in-phase 3ω voltage 13](https://image.slidesharecdn.com/enanonewsletter22versionfinal-111026044835-phpapp02/85/Enano-newsletter-issue22-13-320.jpg)
![nanoresearch
measurements were reported in the 1960s reader to the mentioned references from
at lower temperature [9], and more recent the group of Goodson for the study of such
experiments using electrical methods [10] phenomena.
and time-domain thermoreflectance have
shown that part of the phonon mean free 4. Heat conduction in electric tracks and
path distribution should lie at lengths above ridges
500 nm. An alternative way to get insights In electronic devices with deposited metal
in the issue of mean free path is the use of lines or doped silicon tracks, the electronic
molecular dynamics simulation. Henry and path lies on top of planar substrates.
Chen [11] recently calculated a distribution Considering a phonon mean free path on
of the mean free paths for silicon, finding the order of 100 nm, we present in Fig. 3
indeed that around 30% of the thermal three types of possible devices that consist
conductivity was due to mean free paths of a ridge on a planar substrate of the same
larger than 1 micrometer. This is consistent or different material. For simplification, we
with the estimation [10] that the mean free start with only similar materials. The left
path should be around 300 nm for silicon. device can be treated with the usual Fourier
Here one should keep in mind that the heat conduction, the middle one is different
widespread evaluation of the mean free as even if the nanostructure on top is large
path vgτ from is delicate and is in a thermal equilibrium the thermal
in the sense that it counts all the optical constriction resistance to the cold bath has
modes in the specific heat cp, whereas they to be described by a subcontinuum heat
are not expected to play a key role in the conduction. The right device is even further
thermal conductivity due to the flatness of complicated as the size of the structure does
their dispersion relation (vg≈0). Here ρ is the not allow an equilibrium inside due to its small
material density as usual. This approximation size and the fact that phonons are not trapped
underestimates the effect of the phonon in the cavity but can also escape. The centre
rarefaction in small devices. Recent works figure is typical of the rarefaction effect [19],
performed with nanowires [12, 13, 14] and when the phonon statistics impinging the
with embedded nanoparticles [15, 16], constriction is different than the equilibrium
targeting thermal conductivity reduction one. The right one has been tackled in
in thermoelectric materials, have also a theoretical paper [20]. In principle, the
highlighted the effect of roughness [14, Boltzmann transport equation has to be used
15] in addition to the phonon-particle for calculating the heat flux in structures such
confinement effect. Here we do not discuss as the centre and right ones but approximate
the suspended wire issue as it is for the methods have been developed such as the
moment less relevant in nanoelectronics. ballistic-diffusive equation [21, 22].
We need also to underline the role of
the interaction of electrons and optical (a) (b) (c)
phonons with the acoustic ones [17, 18].
Even if optical phonons do not carry heat
significantly they interact with the acoustic
ones, therefore impacting the thermal
conductivity through the scattering mean
Fig. 3 > Three types of electrical conductors on a
free path. Note also that electron scattering
planar substrate. The substrate can be either an
with optical phonons is significant, and the electric conductor or an insulator. The Fourier
heat redistribution to acoustic phonons description of heat conduction is not adequate
takes place through optical/acoustic for the two last devices (b,c) if the phonon mean
phonons scattering interaction. We refer the free path is of the order of 100 nm or more./ 15](https://image.slidesharecdn.com/enanonewsletter22versionfinal-111026044835-phpapp02/85/Enano-newsletter-issue22-15-320.jpg)
![nanoresearch
Transient Heat Conduction Problems using
Ballistic-Diffusive Equations and Phonon
Boltzmann Equation”, Journal of Heat Transfer,
Vol. 127, pp.298-306 (2005).
[3] D. Cahill, “Thermal conductivity measurements
from 30 to 750 K: The 3ω method”, Review of
Scientific Instruments, Vol. 61, p802 (1990).
[4] S.-M. Lee and David G. Cahill, “Heat transport
in thin dielectric films,” Journal of Applied
Physics, Vol. 81, 2590 (1997).
[5] S.-M. Lee, David G. Cahill, and R.
Venkatasubramanian, “Thermal conductivity of
Si-Ge superlattices,” Applied Phyics. Letters,
Fig. 4 > Different regimes of heat conduction as a Vol. 70, 2957 (1997).
function of the shape of the body in contact with [6] D.-A. Borca-Tasciuc and G. Chen, “Thermal
the substrate. Adapted from Ref [20]./ Properties of Nanochanneled Alumina
Templates,” Journal of Applied Physics, Vol.
The purpose of our work is to observe 79, pp. 084303-1-9 (2005).
these kinds of subcontinuum effects [7] T. Borca-Tasciuc, R. Kumar, and G. Chen,
experimentally. We have already measured “Data Reduction in 3ω Method for Thin Film
[23] the expected strong reduction of the Thermal Conductivity Measurements,” Review
thermal conductance in the ballistic regime of Scientific Instruments, Vol. 72, o. 4, pp.
with respect to Fourier’s prediction. Our 2139-2147 (2001).
first results indicate in addition a different [8] T. Tong and A. Majumdar, “Reexamining
behavior than what can be calculated in the the 3-omega technique for thin film thermal
purely ballistic case, which is exactly what characterization”, Review of Scientific
is pointed out in the analysis developed in Instruments, Vol. 77, 104902-104902-9 (2006).
Figs. 3 and 4. [9] R. Gereth and K. Hubner, “Phonon MeAn
Free Path in Silicon Between 77 and 250°K”,
Acknowledgements Physical Review, Vol. 134, pp A235–A240
(1964).
We thank M. Tilli for providing high ohmic
[10] M. Asheghi, Y.K. Leung, S.S. Wong, and K.E.
8” Si wafers. M. Myronov and V. Shah are
Goodson., “Phonon-Boundary Scattering in
acknowledged for doing the n+ Si epitaxial
Thin Silicon Layers,” Applied Physics Letters,
growth.
Vol. 71, pp. 1798-1800 (1007).
We acknowledge the partial support [11] A. Henry and G. Chen, “Spectral Phonon
of the EU projects NANOPACK and Properties of Silicon Based Molecular Dynamics
NANOPOWER. P.O.C. acknowledges the and Lattice Dynamics Simulations,” Journal of
support of EU project nanoICT for the partial Computational and Theoretical Nanosciences,
funding of a stay at VTT. Vol. 5, pp. 141-152 (2008).
[12] D. Li, Y. Wu, P. Kim, L. Shi, P. Yang, A.
References Majumdar, “Thermal conductivity of individual
[1] D.G. Cahill, W. K. Ford, K.E. Goodson, G.D. silicon nanowires,” Applied Physics Letters, Vol.
Mahan, A. Majumdar, H.J. Maris, R. Merlin, and 83, pp. 2934-2936 (2003).
S.R. Phillpot, “Nanoscale thermal transport,” [13] A. I. Hochbaum, R. K. Chen, R. D. Delgado,
Applied Physics Reviews, Journal of Applied W. J. Liang, E. C. Garnett, M. Najarian, A.
Physics, Vol. 93, p793 (2003). Majumdar, P.D. Yang. “Enhanced thermoelectric
[2] R.G. Yang, G. Chen, M. Laroche, and Y. Taur, performance of rough silicon nanowires,”
16 “Simulation of Nanoscale Multidimensional Nature, Vol. 451, pp. 163-167 (2008).](https://image.slidesharecdn.com/enanonewsletter22versionfinal-111026044835-phpapp02/85/Enano-newsletter-issue22-16-320.jpg)
![nanoresearch
[14] I. Akram Boukai, Y. Bunimovich, J. Tahir-Kheli, Phonon Dispersion,” Journal of Applied
J.-K. Yu, W.A. Goddard III, and J.R. Heath, Physics, Vol. 96, no. 9, pp. 4998-5005 (2004).
“Silicon nanowires as efficient thermoelectric [19] G. Chen, G. Chen, “Nonlocal and
materials”, Nature, Vol. 451, pp168-171 (2008). Nonequilibrium Heat Conduction in the Vicinity
[15] W. Kim, J. Zide, A. Gossard, D. Klenov, S. of Nanoparticles,” ASME Journal of Heat
Stemmer, A. Shakouri, A. Majumdar, “Thermal Transfer, Vol. 118, pp. 539-545 (1006).
conductivity reduction and thermoelectric figure [20] S. Volz and P.-O. Chapuis, “Increase of thermal
of merit increase by embedding nanoparticles resistance between a nanostructure and a
surface due to phonon multireflections”, Journal
in crystalline semiconductors,” Physical Review
of Applied Physics, Vol. 103(3), p034306 (2008).
Letters, Vol. 96, p045901 (2006).
[21] G. Chen, “Ballistic-Diffusive Heat Conduction
[16] W. Kim, S. L. Singer, A. Majumdar, J. M. O.
Equations,” Physical Review Letters, Vol. 85,
Zide, D. Klenov, A. C. Gossard, S. Stemmer,
pp. 2297-2300 (2001).
“Reducing thermal conductivity of crystalline
[22] G. Chen, “Ballistic-Diffusive Equations for
solids at high temperature using embedded
Transient Heat Conduction from Nano to
nanostructures,” Nano Letters, Vol. 8, pp.
Macroscales”, Journal of Heat Transfer, Vol.
2097-2099 (2008). 124, pp. 320-328 (2002).
[17] S. Sinha, E. Pop, R.W. Dutton, K.E. Goodson, [23] P.O. Chapuis, M. Prunnila, A. Shchepetov, L.
“Non-Equilibrium Phonon Distributions in Sub- Schneider, S. Laakso, J. Ahopelto, and C.M.
100 nm Silicon Transistors,” ASME Journal of Sotomayor, “Effect of phonon confinement
Heat Transfer, Vol. 128, pp. 638-647 (2006). on heat dissipation in ridges”, Proceedings of
[18] E. Pop, B. Dutton, and K.E. Goodson, “Analytic the 16th International Worshop on THERMal
Band Monte Carlo Model for Electron Transport INvestigations of ICs and Systems (THERMINIC),
Modeling in Si Including Acoustic and Optical B. Courtois and M. Rencz (ed), (2010).
17](https://image.slidesharecdn.com/enanonewsletter22versionfinal-111026044835-phpapp02/85/Enano-newsletter-issue22-17-320.jpg)
![www.nanociencia.imdea.org
RESEARCH PROGRAMMES
• Molecular nanoscience
IMDEA-Nanociencia is a private Foundation created by joint initia-
tive of the Comunidad de Madrid and the Ministry of Education of
• Scanning probe microscopies the Government of Spain in February 2007 to manage a new
and surfaces
research Institute in Nanoscience and Nanotechnology (IMDEA-
Nanociencia). The Institute is located at the campus of the Univer-
• Nanomagnetism sidad Autónoma de Madrid in Cantoblanco.
The Institute aims at performing research of excellence in selected
• Nanobiosystems: biomachines and
manipulation of macromolecules areas and offers attractive opportunities to develop a career in sci-
ence at various levels from Ph.D. students to senior staff positions.
• Nanoelectronics and
superconductivity The Madrid Institute for Advanced Studies in Nanoscience also
develops an important program of technology transfer and creation
of spin-off companies.
• Semiconducting nanostructures
and nanophotonics
E-mail contacto.nanociencia@imdea.org
Phone 34 91 497 68 49 / 68 51
Fax 34 91 497 68 55
• Nanofabrication and advanced
instrumentation
[Nanociencia y Nanotecnología: lo pequeño es diferente small is different
Nanoscience and Nanotechnology:
]](https://image.slidesharecdn.com/enanonewsletter22versionfinal-111026044835-phpapp02/85/Enano-newsletter-issue22-20-320.jpg)
![nanoresearch
The raise up of UHV atomic scale
interconnection machines
J. S. Prauzner-Bechcicki1, D. Martrou2,
C. Troadec3, S. Gauthier2, M. Szymonski1 and
C. Joachim2,3
1Center for nanometer-Scale Science and Advanced
Materials (NANOSAM), Faculty of Physics,
Astronomy and Applied Computer Science
Jagiellonian University, Reymonta 4, Krakow, Poland.
2Centre d’Elaboration de Matériaux et d’Etudes Fig. 1 > A single five wings molecule-motor [1]
Structurales (CEMES-CNRS), 29, rue Jeanne Marvig, positioned between a 4 Au nano pads junction
BP 94347, 31055 Toulouse Cedex 4, France. constructed at the Si(100)-H surface. The 4 black
3Institute of Materials Research and Engineering,
wires getting out of the surface are indicative of the
A*STAR (Agency for Science, Technology and interconnections step 3 discussed in the text depending
Research), 3 Research Link, Singapore 117602. on the electronic gap of the supporting surface./
1. Introduction In section 2, the general principles of the
Single molecule mechanics [1], mono- few UHV atomic scale interconnection
molecular electronics [2] and surface machines under test to solve the problem are
atomic scale circuits [3] [4] are all requiring a described. Depending on the electronic gap
specific surface interconnection technology of the surface where the atomic scale devices
with an atomic precision and cleanness and machineries are supposed to work, two
[5]. In a planar configuration, this surface families of interconnections machines are
technology must be able to provide multiple being explored. Section 3 is providing one
access electronic channels to the atomic (or example of an atomic scale interconnection
molecular) scale machinery constructed on a machine designed for the surface of wide
surface (see for example Fig. 1). At the end gap semi-conductor and insulator materials.
of the 80’s, it was expected that the e-beam Section 4 is giving the example of two
nano-lithography technique would be able interconnection machines for moderate gap
to provide such a technology [6]. But with its semi-conductor surfaces. The design and
resist based approach, e-beam technique instrumentation works reported here are
will not face the challenge [7] because it is the consequence of the EU ICT integrated
not able to respect at the same time the project Pico-Inside in Krakow and Toulouse
atomic scale precision, the cleanness and together with the A*STAR VIP Atom Tech
the expected large number N of access Phase 2 project in Singapore. It is now further
channels to the atomic scale machinery developed in the new EU ICT integrated
[8]. Alternative nanolithography techniques project AtMol and in the Phase 3 of the
such as nano-imprint [9] or nano-stencil A*STAR VIP Atom Tech project in Singapore.
[10] are neither adapted to encompass
all the interconnection stages from the 2. Atomically precise electrical
macroscopic to the atomic scale nor clean interconnection machine
enough down to the atomic scale. At the turn An atomic scale precision, multiple access,
of the century, this problem triggers a new electrical interconnection instrument must
approach to planar electrical interconnects provide N conducting wires converging
starting from the bottom that is from the toward a very small surface area where
fundamentals of surface science. an active machinery (see Fig. 1 for a N=4 23](https://image.slidesharecdn.com/enanonewsletter22versionfinal-111026044835-phpapp02/85/Enano-newsletter-issue22-23-320.jpg)
![nanoresearch
To realize the 5 levels of interconnect 1. a flexural-hinge guided (XY) nano-
described in Fig. 2a in UHV, the deposition positioner stage (100 μm x 100 μm,
of molecules, their observation by NC-AFM repeatability 5 nm) with a closed loop
and the measurement of their electrical control based on capacitive sensors,
properties, the Toulouse group has designed 2. an evaporation system highly collimated
and constructed a dedicated UHV equipment on the cantilever to perform nano-stencil
called DUF (DiNaMo UHV Factory). This deposition,
equipment allows transferring samples under
UHV between five complementary UHV 3. a (XYZ) piezo driven table for positioning
chambers (see Fig. 3): the metallic micro-combs for the
electrical contacts,
(1) an MBE growth chamber dedicated
to nitride semiconductors growth, 4. an optical microscope to control the
metallic nano-pads growth and stencil positioning of the micro-combs.
evaporation for microelectrodes These modifications were introduced by
(2) a room temperature AFM/STM chamber the mechanical workshop of the Toulouse
for surface characterization by STM and laboratory. The main advantage of using
NC AFM a commercial UHV AFM/STM is to benefit
from the good characteristics for SPM
(3) an AFM/STM chamber modified for
imaging. But the piezo tube used to scan
nano-stenciling experiments and
has a range of a few μm only. The addition of
electrical measurements
a piezo table to move the sample offers the
(4) a preparation chamber for cleaning possibility to perform wide range scanning,
substrates, STM tips and AFM cantilevers up to 80 μm SPM images, while keeping the
(5) a mass spectrometer chamber possibility to realize atomic sale imaging with
transformed in a molecular ions source. the piezo tube.
For (3) a UHV Omicron Nanotechnology One of the disadvantages is the small
VT STM/AFM head has been modified to accessible space around the SPM head.
accommodate different tools, namely [11]: Indeed, it is not possible to place an optical
Fig. 3 > The DUF (DiNaMo UHV Factory) equipment allows to transfer samples between 5
complementary UHV chambers in order to realize the 5 levels of interconnect on wide band gap
semiconductors (GaN, AlN)./ 25](https://image.slidesharecdn.com/enanonewsletter22versionfinal-111026044835-phpapp02/85/Enano-newsletter-issue22-25-320.jpg)
![nanoresearch
microscope with normal incidence with Krakow and the other in Singapore. The
respect of the substrate, and an atomic Krakow’s system consists of three basic
source for the nano-stencil experiments with segments: multi-probe, low-temperature
normal incidence with respect to the AFM scanning probe microscope (LT-SPM)
cantilever. In our case, the image obtained and preparation chambers. Multi-probe
by the optical microscope comes from a segment is composed of 4-probe scanning
mirror with an angle of 30° with the substrate tunnelling microscope (STM) combined
plane. This gives distorted images, with with high resolution scanning electron
a loss of resolution: only 3 μm instead of microscope (HR-SEM) and hemispherical
1 μm in normal incidence. The effusion cell electron energy analyser (scanning Auger
is fixed on a port of the UHV chamber that microscope, SAM) (see Fig. 4). The Auger
makes an angle of 33° with the horizontal microscope part is the element not present
plane, and another angle of 28° between in Singapore’s setups. Composition of the
the two vertical planes passing through the multi-probe segment allows surface element
evaporation beam and the central axis of analysis, imaging and measurements of
the cantilever. This orientation of the atom nanostructures conductance with very
beam induces distortion, which should be high-resolution. In accord with the Fig. 2b
taken into account in the design of the nano- principle, HR-SEM may act as a navigation
pattern to be drilled into the pyramidal tip of to precisely position each of the 4 STM tips
the cantilever [11].
(a)
4. UHV interconnection machine for
moderate semi-conductor surface gap
For a moderated valence band-conduction
band electronic surface gap (around a few
eV), it is not possible to use very long surface
metallic circuitry due to the possible lateral
surface leakage current between the surface
electrodes. In this case, one solution is to
use ultra sharp STM like tips positioned from
the top on the surface as microelectrodes
(Fig. 2b). In this case, the core of the tips will (b)
not be in contact with the supporting surface
and one can go continuously from a tip apex
radius of curvature of a few nanometer up
to a 100 microns or more section for the
tip body. In this case, navigation on the
surface can be performed using an UHV-
SEM (Fig. 2b) by grounding the sample
during the SEM imaging to avoid the surface
charging effect. This is the basics of the
UHV interconnection machines described
here. A low temperature approach is
compulsory with those systems because of Fig. 4 > (a) View on sample stage of 4-probe
the low electronic gap at the surface of the microscope; in upper part one can see SEM
column and next to it an entrance to hemispherical
supporting material.
electron energy analyser; below SEM column
There are two apparatuses that realize the there are three of four STM probes. (b) SEM image
26 above described design, one is housed in of four STM probes./](https://image.slidesharecdn.com/enanonewsletter22versionfinal-111026044835-phpapp02/85/Enano-newsletter-issue22-26-320.jpg)
![nanoresearch
that will be used as microelectrodes. First
measurements of conductance of gold
nanostructures on Ge(001) surface are in
progress.
Next, LT-SPM segment consist of scanning
probe microscope that may work both as
STM and NC-AFM in a range of temperatures
Fig. 5 > InSb surface imaged with q-sensor NC-
from 4K up to room temperature. Thanks AFM in temperature 4K./
to use of scanner and sample holder
embedded in a cryostat, the LT-UHV STM
SAM images of metallic nano-mesa grown
allows for a very high resolution imaging, as
on semiconductor substrate are shown.
well as, stable spectroscopic measurements
The Ag/Ge(111) is an example of a system
and atomic scale manipulations.
for which depending on the deposition
Furthermore, NC-AFM mode is based on conditions, on the successive thermal
q-sensor device (tuning fork) that enables annealing and on the amount of deposited
imaging of conducting, semiconducting and material the resulting overlayer morphology
insulating samples at low temperatures (see can be switched from an atomically smooth
Fig. 5) and, if required, also simultaneous to a columnar-like [12]. Sample is prepared in
measurements of tunnelling current. This the following way: silver in amount of nominal
option makes the Krakow’s system a very 5 ML is deposited on the germanium surface
powerful tool. Last but not least, is the kept at low temperature. Low energy electron
preparation segment that consist of typical diffraction studies performed immediately
preparation equipment such as a XYZ after deposition reveal that compact silver
manipulator with electric contact allowing for film is crated. Such a conclusion follows from
resistive heating up to 1000K (using a direct the fact that reflections characteristic for
heating mode 1200K may be achieved), the unreconstructed Ag(111) overlayer are
furthermore the manipulator allows for observed exclusively on LEED image. One
cooling the sample down to 100K with may assume it is a clear and direct indication
nitrogen vapour, an ion gun, a low energy that Ge substrate is completely buried.
electron diffraction system for quick sample In the next step the sample is annealed to
quality tests and several ports allowing room temperature. After annealing a massive
for incorporation of additional elements
(for instance effusion cells or quartz
microbalance thickness monitor) into the
chamber. All segments are composed of the
highest quality elements all of them being
compatible to work in UHV environment
(less than 3×10-10mbar), and thus allowing
Fig. 6 > STM, HR-SEM and SAM images of the Ag/
for conducting very complex experiments Ge(111) sample. Image size: 220nm × 270nm. Left
in a single set-up in a very controlled panel: STM image; Middle panel: HR-SEM image;
way, starting from sophisticated sample Right panel: SAM image. STM image reveals bright
preparation and ending with extensive and nanostructures 2.5nm high. The same regions
complete characterisation. are marked red in HR-SEM image. In SAM image
those structures are black. SAM image was taken
As the Auger microscopic part of the
for Ge line (E=1144 eV), thus exposing as bright
Krakow’s nano-probe instrument is not regions containing Ge. Therefore, it is possible to
present in the other setups (see above and identify the black structures in SAM image (bright
below) its potential is briefly described in the and red regions on STM and HR-SEM images,
following. In Fig. 6, the STM, HR-SEM and respectively) as silver islands./ 27](https://image.slidesharecdn.com/enanonewsletter22versionfinal-111026044835-phpapp02/85/Enano-newsletter-issue22-27-320.jpg)
![nanoresearch
atomic scale interconnection machines
housed in Toulouse, Krakow and Singapore
possess potential to meet requirements
and demands arising upon realization
of such a connection [13]. Furthermore,
experience and knowledge gained by
Krakow’s and Toulouse’s groups through
EU ICT integrated project Pico-Inside and
by Singapore’s group in parallel through A
*STAR VIP Atom Tech Phase 2 set a very
Fig. 8 > SEM image of the tip apex of 4 chemically
etched tungsten tips converging toward 4 Au
fortunate starting point for that quest which
nano-islands which have been manipulated one by is now in full development in the new ICT
one to enable a 4 points like surface conductance integrated project AtMol (www.atmol.eu).
measurement. This image has been recorded on a
multi-probe system from Zyvex./ References
[1] Rapenne G, Launay J-P and Joachim C 2006
Under the UHV-SEM, each nano-island of a J. Phys.: Condens. Matter 18 S1797.
contacting nanostructure can be electrically [2] Joachim C, Gimzewski J K and Aviram A 2000
contacted from the top using one ultra- Nature 408 541.
sharp tip per nano-island. Each tip apex is [3] Wada Y 1996 Microelectronic Engineering 30
positioned using the UHV-SEM for a precise 375.
navigation. The soft contact between a nano- [4] Ample F, Ami S, Joachim C, Thieman F and
island and its tip apex is controlled by the Rapenne G 2007 Chem. Phys. Lett. 434, 280.
feedback loop system of an STM meaning [5] Joachim C 2002 Nanotechnology 13 R1.
that there is one STM control electronic per [6] Itoua S, Joachim C, Rousset B and Fabre N
tip. A SEM image of the contact preparation 1992 Nanotechnology 3 10.
of 4 nano-islands manipulated on purpose in [7] Saifullah M S M, Ondarcuhu T, Koltsov
a row on a MoS2 surface is presented in Fig. D F, Joachim C and Welland M 2002
8 together where the 4 tip apex approaching Nanotechnology 13, 659.
for the contact are also imaged. Electrical [8] Cacciollati O, Joachim C, Martinez J-P and
measurements are now in progress to first Carsenac F 2004 Int. Journ. Nanosci. 3 233.
record the “nano-pad – MoS2 surface – [9] Chou S Y, Krauss P R, Renstrom P J 1996
nano-pad” I-V characteristics for an inter Science 272 85.
nano-pad distance lower than 10 nm. Then, [10] Thet N T, Lwin M H, Kim H H, Chandrasekhar
metallic nano-islands will be transfer printed N and Joachim C 2007 Nanotechnology 18,
on Si(100)H surface for surface atomic wire 335301.
conductance measurements. [11] Guo H, Martrou D, Zambelli T, Dujardin E and
Gauthier S 2008 Rev. Sci. Instrum. 79 103904.
5. Conclusion [12] Krok F., Buatier de Mongeot F., Goryl M.,
Connecting the atomic (or molecular) scale Kolodziej J.J., and Szymonski M., Phys. Rev. B
machinery to the outer world respecting the 81, 235414 (2010).
atomic scale precision of the construted [13] C. Joachim, D. Martrou, M. Rezeq, C. Troadec,
machinery is a very challenging and Deng Jie, N. Chandrasekhar and S. Gauthier J.
30 complex task. Each of the three described Phys. CM, 22, 084025 (2010).](https://image.slidesharecdn.com/enanonewsletter22versionfinal-111026044835-phpapp02/85/Enano-newsletter-issue22-30-320.jpg)
Enano newsletter issue22
- 1. No. 22 /// August 2011 www.phantomsnet.net Atomic Scale and Single Molecule Logic Gate Technologies (AtMol) Heat dissipation in nanometer-scale ridges The raise up of UHV atomic scale interconnection machines
- 3. dear readers, This edition contains two articles providing new insights in a relevant field for future nano- electronics applications, i.e. molecular computing at the atomic scale. Since 1974, Molecular electronics had been always associated to the possible future of computers. A new European Integrated Project AtMol was proposed and accepted by the European Commission after the FP7 ICT Call 6 to create this new technology (www.atmol.eu). The nanoICT “Mono- Molecular Electronics” Working Group was also set-up in 2008 to reach this objective. In 2010, the nanoICT project launched its first call for exchange visits for PhD students with the following main objectives: 1. To perform joint work or to be trained in the leading European industrial and academic research institutions; 2. To enhance long-term collaborations within the ERA; 3. To generate high-skilled personnel and to facilitate technology transfer; The first outcome report (“Heat dissipation in nanometer-scale ridges”) of this call investigates experimentally the effect of lateral confinement of acoustic phonons in silicon ridges as a function of the temperature. We would like to thank all the authors who contributed to this issue as well as the European Commission for the financial support (projects nanoICT No. 216165 and AtMol No. 270028). > Dr. Antonio Correia Editor - Phantoms Foundation contents 05 > nanoresearch. Atomic Scale and Single Molecule Logic Gate Technologies (AtMol) /// C. Joachim 12 > nanoresearch. Heat dissipation in nanometer-scale ridges /// P.-O. Chapuis, A. Shchepetov, M. Prunnila, S. Laakso, J. Ahopelto and C. M. Sotomayor Torres 18 > nanojobs 21 > nanoICT Conf Report 23 > nanoresearch. The raise up of UHV atomic scale interconnection machines /// J. S. Prauzner-Bechcicki, D. Martrou, C. Troadec, S. Gauthier, M. Szymonski and C. Joachim editorial information No 22. August 2011. Published by Phantoms Foundation (Spain) editor > Dr. Antonio Correia > antonio@phantomsnet.net assistant editors > José Luis Roldán, Maite Fernández, Conchi Narros, Carmen Chacón and Viviana Estêvão. 1500 copies of this issue have been printed. Full color newsletter available at: www.phantomsnet.net/Foundation/newsletter.php For any question please contact the editor at: antonio@phantomsnet.net editorial board > Adriana Gil (Nanotec S.l., Spain), Christian Joachim (CEMES-CNRS, France), Ron Reifengerger (Purdue University, USA), Stephan Roche (ICN-CIN2, Spain), Juan José Saenz (UAM, Spain), Pedro A. Serena (ICMM-CSIC, Spain), Didier Tonneau (CNRS-CINaM Université de la Méditerranée, France) and Rainer Waser (Research Center Julich, Germany). deadline for manuscript submission depósito legal printing Issue No 24: October 31, 2011. legal deposit Gráficas Issue No 25: December 30, 2011. BI-2194/2011 Valdés, S.L. 03
- 5. nanoresearch Atomic Scale and Single Molecule Logic Gate Technologies (AtMol) FET ICT Integrated Project (2011-2014) C. Joachim CEMES-CNRS, France. on very specific fabrication and measurement AtMol Coordinator know-hows like the break junction or LT-UHV STM techniques. According to the ITRS roadmap, the transistor The new European Integrated Project i.e. the basic switching element of any AtMol was proposed and accepted by the processor Arithmetic and Logic Unit (ALU) will European Commission after the FP7 ICT reach its bottom source – drain distance limit Call 6 to create this new technology for (between 5 nm to 10 nm) in the years 2020’s. molecular computing at the atomic scale. This will be one limit more after the switching The members of the AtMol consortium energy problem which started to stop the started from the observation that for single increase of the computer performance in molecule electronics to diffuse towards years 2005-2006. Our global need for more applications and at the same time be used and more computing power in portable and as epistemological devices to explore the limit sustainable forms is now pushing the semi- of calculating machines, it is required to start conductor industry to explore new avenues for from the best of surface science. We need to producing machines to transmit and process create the ultra clean technology required to information. On the other side of the road, this construct atomic scale calculating circuits, to extreme miniaturisation of the ALU and its interconnect them to the external world and associated memory bring again on the table at the same time to invent a proper packaging the problem of the limits of the machine in term technology to protect the constructed atomic of size and power consumption whatever the scale circuit when ready to be extracted from machine: a mechanical machine, a calculator, its native UHV environment. an emitter… The technical and fundamental limits of machines from the gears to the The AtMol molecular chip concept heat engine, from the relay to the transistor For AtMol, a molecular chip is a fully have always been a fantastic playground for packaged and interconnected planar physicist and chemists to make progresses in atomic scale complex logic circuit where the our understanding of the laws of physics. ALU is constructed with a set of complex Since 1974, Molecular electronics had been molecule logic gates which may be one always associated to the possible future day embedded in a single molecule. These of computers. It is now one of the options are interconnected by surface atomic or among others like quantum computers on the molecular wires constructed with atomic way to bring the next computing technology precision on a substrate with a large enough after our fantastic transistor era. Step by step, surface electronic band gap. The central Molecular electronics evolves and gives rise and entirely new concept to be explore to different branches like organic electronics, within AtMol is the separation in space of single molecular devices and molecular the atomic scale structures of the ALU from logics. Organic electronics had developed the nano/mesoscopic scale interconnects. its own dedicated technologies like printed This new interconnect concept is motivated electronics. This had not been the case yet for by the fact that, whatever the architecture of the others branches which remain dependent the planar atomic scale complex logic circuit 05
- 6. nanoresearch to be prepared on the surface, there is an incompatibility to deal at the same time and on the same surface with all the interconnection scales from the atomic to the mesoscopic scales (and beyond). All the known nano-scale fabrication techniques, including e-beam nano- lithography, nano-stencil, nano-imprint, and Focused Ion Beam (FIB), are not atomically clean techniques. For example, in e-beam Fig. 1 > An LT UHV STM image (5.8 nm x 6.7 nm) lithography a resist is used which is very difficult of a Ge(100)H surface carefully prepared by the to remove entirely after the nanofabrication Krakow AtMol partner. Large terraces of Ge(100)H step and thus precludes the use of atomically in Krakow, Si(100)H in Dresden, Nottingham and clean surfaces and the associated state-of- Singapore, MoS2 in Singapore and AlN in Toulouse the-art surface science characterisation and AtMol partners are now prepared to become the atomic-scale manipulation tools. Similarly, supporting surface of the AtMol atomic scale atoms can diffuse laterally and in a random interconnects./ manner in the nano-stencil technique and when engraving using a FIB. On the other hand, while LT-UHV STM or UHV NC-AFM microscopes are capable of manipulating single atoms, these instruments are not capable of constructing interconnects from the atomic scale (0.1 nm) to the mesoscopic scale (100 nm). A spatial separation of the interconnects between the two faces of a same wafer is the solution proposed by the Fig. 2 > The detail configuration of the targeted AtMol consortium. It has the great advantage AtMol molecular chip structure with its back that the top surface of the wafer is reserved for interconnects, its vias through the surface stopped the planar atomic scale circuit constructions. just before disturbing it and its top packaging chip. In this drawing, the top packaging chip is laterally The AtMol process flow cross cut to help in locating the atomic scale circuit supported by an Si(100)H surface in this case. The AtMol is proposing a comprehensive process insert is presenting this circuit with its 16 Au nano- flow, spanning the atomic to mesoscopic islands interconnection pads and the location of the scale for processing and fabricating a active circuit. The question mark is an indication that molecular chip. the exact optimal architecture of this circuit is not yet determined in AtMol (see the main text for the 1) The Atomic scale logic gates and atomic different possible choices). The surface size of the scale circuits are going to be constructed atomic chip is 54 x 54 surface SiH dimers. This is an on the front side of the wafer whose atomic indication of the simulation target of the Toulouse, scale surface is going to be prepared Singapore and Barcelona AtMol partners to succeed with care (Fig. 1). Then, the nanoscale to in associating semi-empirical (N-ESQC) and DFT mesoscopic (and, indeed, macroscopic) (TranSiesta) surface transport calculations to predict interconnects will be fabricated on the back the best surface atomic scale surface architecture./ side of the same wafer (See Fig. 2). From atomic scale structures of the top surface wafer from the back to the front side. Of to mesoscopic connections of the back, course, the piercing is going to be stopped solid and rigid nano-vias will be fabricated just before the perturbation of the top surface 06 by piercing, with nanoscale precision, the atomic order.
- 7. 07
- 8. nanoresearch system under a scanning electron microscope (or an optical for insulating surface) navigation system. Uniquely, AtMol will use and develop further the only N-probe UHV interconnection machines which are currently existing in the world i.e. in Singapore, Krakow and Toulouse (Fig. 4). These machines are the ultimate UHV compatible multi-probe testers reaching the nanoscale precision. Before encapsulating the front side under UHV conditions, electrical characterisation will be carried out in parallel with the electrical testing of the back Fig. 3 > A detail atomic scale representation of interconnects. a top surface contact LT-UHV experiment using multiple STM tips, each one contacting one Au nano-pads. The molecule logic gate represented is here a starphene molecule interconnected in a classical way to 3 quite large surface dangling bond atomic wires, each reaching an Au nano- pad. The 3 AtMol UHV interconnection machines able to perform such multi-probes experiments are discussed in Fig. 4. The insert is presenting the LT-UHV STM image of the starphene electronic ground state obtained by the Singapore and Toulouse AtMol LT-UHV STM. This starphene molecule was synthesized by the Tarragona AtMol partner. Long molecular wires, new molecule logic gates and latching molecules are going to be synthesized by the Berlin, Tarragona and Toulouse AtMol Chemists./ 2) The back-side mesoscopic interconnection circuitry and the nano-via through the wafer indicated in Fig. 2 are prepared and UHV cleaned for example before the atomic scale construction step. In this case, the front side is going to be encapsulated using a wafer bonding technique without modifying the back mesoscopic face of the wafer. Fig. 4 > The 3 AtMol UHV atomic scale interconnection machines in construction. The first one is now being 3) The top surface planar atomic scale logic tested in Singapore, the second one in Krakow circuits will be tested using N probes (see and the third one in Toulouse. Each one has is the testing principle presented in Fig. 3) own specific characteristics. The Singapore one is located within a UHV-compatible atomic fully LT and equipped with a UHV transfer printer. scale interconnection machine (Fig. 4). The Krakow one is equipped by a hemispherical These interconnection instruments are electron energy analyser (Auger microscope). Both integrated within one large UHV system which are equipped by the required high resolution (4 nm) incorporates a surface preparation chamber, UHV-SEM. The Toulouse one is more dedicated a UHV-transfer printing device, an FIM atomic towards large electronic gap surfaces which scale tip preparation device, an LT-UHV-STM explained this peculiar multi-probes contacting 08 (or NC-AFM) microscope, and a N-nanoprobe approach based on metallic cantilevers./
- 9. nanoresearch 4) New atomic scale construction and Exploring the different molecule logic fabrication techniques will be fully developed architectures without fear of seeing the atomically precise At the atomic scale, the main advantage of circuits being destroyed by any subsequent the AtMol atomic scale chip technology is that mesoscopic scale interconnection fabrication it offer a definitive working bench to determine step. The atomic scale fabrication techniques the optimal atomic scale architecture for are under development before being constructing a complex logic gate able, for incorporated into back interconnected example, to add two binary numbers. The fantastic advantage of the front-back side wafers. For example, AtMol will exploit this interconnection AtMol innovation described capability to develop a unique UHV atomic above is that any circuit architecture can be scale transfer printing technique able to constructed and tested in full planar and UHV integrate nano-scale contacting metallic technology with an atomic scale technology pads, long molecular wires, active molecule and with the possibility to determine the logic gates, and latch molecules on the front exact atomic scale structure of the ALU being (atomically clean) surface (Fig. 5). This will be constructed. It can be expanded to single supported by the objective of improving the molecule mechanics and transmission of construction of long dangling bond atomic mechanical motion. wires on a semiconductor surfaces using local For “single molecule” molecular logic, the forces instead of inelastic electronic effects of standard solution coming from the 70’s is the STM. an hybrid molecular electronics architecture where each molecule in the circuit acts as a switch (or, better, as a transistor). Of course, there are significant problems with this conventional scheme. In particular, it requires a command (for example an electric field or a mechanical push) to be applied on each molecule-switch in the circuit. As a consequence, the distance between each molecule-switch has to be larger than the electron mean free path of the interconnection materials for the electric field to be well defined on each molecule of the circuit. This also requires bringing the command electrodes on many points of the atomic scale circuit. Fig. 5 > A cartoon indicating that the Berlin and For AtMol, this is not the way to go. But this Singapore AtMol partners are exploring UHV type of design can be well tested on the AtMol transfer printing to avoid any nano-lithography wafer top. steps. Initially developed in Singapore to transfer metallic nano-pads on the Si(100)H surface, the The first AtMol objective is to test semi- UHV transfer printing technique is now generalized classical electronic circuit laws at the atomic in AtMol to molecular wires and molecule logic scale. Distinct from the well-known Kirchhoff gates. This cartoon is also presented here to circuit laws, these laws were demonstrated illustrate how AtMol via its Madrid partner is pro- theoretically at the end of the 90s but active in diffusing information to a wide public remain to be experimentally tested. Be it by about AtMol and about the development of Atom synthesising specific long molecules having Tech in general. This cartoon belongs to a series the shape of an electronic circuit or by of 30 drawn by the famous cartoon drawer G. constructing atom by atom such a circuit at Cousseau invited for the AtMol Kick-off meeting./ the surface of a passivated semiconductor, 09
- 10. nanoresearch atomic and molecular manipulation and very precise dI/dV spectroscopy using scanning probes (LT-UHV-STM/AFM) are going to provide the definitive experimental testing of those laws. Having verified atomic scale electronic circuit laws, AtMol designer and chemists can generate complete designs of fully integrated ALU atomic scale circuits. More complex semi-classical circuits will be designed theoretically and tested up to the point where the maximum reachable complexity will be attained due, for example, to the output current intensity falling below a minimum detectable threshold. Fig. 6 > The detail atomic scale representation of a LT-UHV STM experiment performed by the The second AtMol objective is to determine Singapore AtMol Partner to demonstrate how the how the very new Quantum Hamiltonian Boolean truth table of a QHC molecule NOR logic Computing (QHC) concept introduced in gate can be measured. Each Au individual atom the European Pico-Inside project can reach is STM manipulated to interact (or not) with one a larger Boolean logic function complexity branch of the conjugated starphene molecule. as compared with semi-classical atomic The positioning of an Au atom nearby the scale circuits. In QHC, logic operations are molecular branch is a classical input converted in quantum information by the molecule itself. The carried out not via charge processing but starphene ground state position is determined through quantum information manipulation by performing an STS spectrum on the output inside the atomic scale circuit or inside the branch of the molecule. The Berlin AtMol partner molecule. In QHC, the electron transfer will perform single molecule pulling experiments rate between a drive electrode and an to determine the conductance of a single output electrode is controlled by locally molecular wire (or intramolecular circuit). The changing the Hamiltonian of the molecule Singapore, Barcelona and Toulouse molecule (or of the surface atom circuit). In QHC, the (or surface atomic circuit) AtMol designers are decoherence phenomenon is used to build now in full interactions with the Berlin, Tarragona and Toulouse AtMol chemists to create new up a measurable output current averaging molecule logic gates and with the Dresden, all the quantum fluctuations. One major Nottingham, Krakow and Singapore AtMol architectural innovation is that the inputs are surface scientists to create the first surface atomic basically classical but locally converted in scale fully interconnected logic gate circuits./ quantum information (Fig. 6). This conversion is performed by the energy and phase surface supporting the logic gates. A changes of some of the electronic states of detail theory of the surface atomic scale the molecule (or of the surface atomic scale structure, its relaxation while contructing circuit). Two strategies will be considered: surface atomic wire or adsorbing and either the molecule(s) can do everything or interconnecting molecule logic gate and a surface atomic scale circuit completed molecular wire is going to be developed by molecule latches to handle the inputs together with the calculation of the running is preferable to reach large output current current (and its associated electronic (perhaps up to the microampere range). effects) through the atomic scale surface This design effort will be supported by an circuit. New quantum design rules may intense theory of surface science efforts to emerged from those developments which take into account for the first time the full will be immediate feed backed to the 10 electronic contribution of the underneath AtMol chemists.
- 11. nanoresearch AtMol in 2014 January 2012 and the next one on “UHV The AtMol consortium members are very surface chemistry and UHV transfer printing” committed by the fact that at the same in Berlin 6 months after. Output from these time chemistry, surface science, single workshops (lecture notes, etc.) concerning atom manipulation, new UHV multi-channel atomic scale technology will be regrouped interconnection machines, new UHV printing in a new series of books entitled “Advances in Atom and Single Molecule Machine” and packaging techniques need to be published by Springer. developed to construct the first ever single molecule chip. After 2014, AtMol is expecting The AtMol consortium (www.atmol.eu) that its “concept Chip” will be the pivot of the development of real molecular chips and at CEMES-CNRS (Toulouse) the same time will point forward the material LETI-CEA (Grenoble) and technological down limit of a calculator. The Phantoms Foundation (Madrid) To accompany its efforts and to associate ICIQ Institute (Tarragona) more academic and industrial groups to the CIN2 Institute (Barcelona) creation of its new atomic scale technology, AtMol is organizing each 6 months a F. Haber Institute (Berlin) workshop on topics of high concerns. Humbolt University (Berlin) The 1st AtMol workshop on “Atomic scale TU Dresden (Dresden) interconnection machines” ran in Singapore Nottingham University (Nottingham) from the 28th to 29th of June 2011. The 2nd workshop on “Molecular logic architecture at Jagiellonian University (Krakow) the atomic scale” will be running in Barcelona, IMRE A*STAR (Singapore) Impact Factor 7.333 2010 Journal Citation Reports® (Thomson Reuters, 2011) provides the very best forum for experimental For subscription details please and theoretical studies contact Wiley Customer Service: of fundamental and >> cs-journals@wiley.com applied interdisciplinary (Americas, Europe, Middle East and Africa, Asia Pacific) research at the micro- >> service@wiley-vch.de and nanoscales (Germany/Austria/Switzerland) 2011. Volume 7, 24 issues. >> cs-japan@wiley.com (Japan) Print ISSN 1613-6810 / Online ISSN 1613-6829 For more information please visit www.small-journal.com or contact us at small@wiley-vch.de 11
- 12. nanoresearch Heat dissipation in nanometer-scale ridges Pierre-Olivier Chapuis(a)*, Andrey to electric and heat conduction, namely Shchepetov(b), Mika Prunnila(b), Sampo particles free path and even wavelength in Laakso(b), Jouni Ahopelto(b) and Clivia M. some cases. The case of the heat carriers Sotomayor Torres(a)(c)(d)* is particularly of interest because it involves (a) Institut Catala de Nanotecnologia (ICN), Centre phonons that have free paths estimated d’Investigacio en Nanociencia e Nanotecnologia to be larger than the device characteristic (CIN2-CSIC), Edificio CM3, Campus de la length in some cases, undergoing fewer Universitat Autonoma de Barcelona, collisions as a consequence. The so-called 08193 Bellaterra (Barcelona), Spain. (b) VTT Technical Research Center of Finland, phonon rarefaction effect is then responsible PO Box 1000, 02044 VTT, Espoo, Finland. for higher temperature in the heat source (c) Department of Physics, Universitat Autonoma de than what can be estimated with the usual Barcelona, 08193 Bellaterra (Barcelona), Spain. Fourier-based heat conduction equation (d) Institució Catalana de Recerca i Estudis Avançats (ICREA), Passeig Lluís Companys, 23 [2]. Larger temperature gradients are then 08010 Barcelona, Spain. responsible for cracks as differential thermal expansion at material interfaces is not manageable. Abstract > Heat management in today’s A key parameter in the improvement of the electronic devices is critical to prevent device design is therefore the understanding possible failures due particularly to cracks. of heat conduction in the devices. Here Heat is carried in such devices by electrons we focus on one way to measure the in electrical conductors and mainly by implications of the particle behaviour (mean acoustic phonons in electrical insulators. free path) of acoustic phonons when they We explain the first steps of our work are confined in tiny structures. We propose a aiming to investigate experimentally the method to analyse the heat flux propagation effect of lateral confinement of acoustic in these structures and show our first efforts phonons in silicon ridges as a function of towards the efficient use of the designed the temperature. Inspired by the electrical samples. 3ω method, we design a setup that can be used as a mean to generate phonons in 2. Heat flux measurements in low- ~100 nm wide ridge nanostructures and dimensional samples as a thermometer that allows tracking the generated heat flux. A. The macroscopic 3ω method The 3ω method has been developed since 1. Introduction the 1980s, in particular by Cahill [3], with the Works performed over the last decades goal of studying the thermal conductivity have shown that electronic devices with of planar materials. Thin films have been nanometre-scale dimensions are subject investigated widely as well as the thermal to larger temperature-driven stresses in boundary resistance between the films [4]. comparison to what had been estimated in It was extended for multilayer materials or the past [1]. In particular, the size of different particle-based materials [5, 6]. The method components of transistors and electronics is based on the Joule heating of a metallic devices present in printed circuits are now wire of micrometric size deposited on top of 12 comparable to the key scales associated the substrate that “steals” part of the heat
- 13. nanoresearch flux generated. As a consequence, the substrate thermal conductivity has to be temperature of the wire gives an indication found. A 2D cross-section based model on the ability of the sample to conduct heat. has been used extensively over the past The use of a harmonic current to heat the 20 years. It is based on frequency sweeps. wire allows the excitation of the temperature The slope of the temperature variation gives higher-order harmonics. If the heating is not the thermal conductivity [3]. Some authors too high, only the third harmonic is excited have underlined that better models can be as will be seen in the following. One can used [7, 8]. Usually, the width of the wire is write for a current I = I0 cos(ωt) the generated in the micrometric range, which is reached power due to the Joule heating as with standard optical lithography in the fabrication process. P(t) = R I(t)2 = ½ R (1 + cos(2ωt)) and the total temperature reads including B. Implementation at the nanoscale the heating TDC by DC current : The case of nanoelectronic devices is very different to the micrometre-scale one. Even T(t) = T0 + TDC + T2ω cos(2ωt+ 2ω). if the heating/sensing system that can be The key point now is the dependence of used has the same principle, the sizes are the wire resistivity to temperature, that is much smaller. We fabricated nanostructures linear in first approximation for low heating: where the heater/sensor lies on top of silicon R(T) = R0 (1+α ∆T). Finally, the voltage of the substrates as represented in Fig. 1. The top wire can be written as of a ridge is a wire, either a metal or doped silicon, which acts as a heater and as a U = RI = R0I0 [1+α(TDC+T2ωcos(2ωt+ 2ω))] cos(ωt) thermometer at the same time. The doped- = R0I0 [(1+αTDC) cos(ωt)+½αT2ωcos(ωt- 2ω) layer structure requires epitaxial growth of +½αT2ωcos(3ωt+ doped silicon. 2ω)] The use of a lock-in amplifier at the third harmonic enables to measure the amplitude and the phase of the third harmonic and thus extract the local temperature. The amplitude is U3ω = ½ αR0I0 T2ω. The frequency range to be used here is generally between 10 and 5000 Hz. Care has to be taken with the wire width and thickness that should be smaller Fig. 1 > Two types of resistive heater for the ridge than the thermal diffusion wavelength in order experiments./ to prevent from a possible nonhomogeneity The substrates can be made of high- of the heat generation in the wire. resistivity silicon. The submicrometer ridges This experimental part of the work permits are fabricated with electron beam lithography only to get a qualitative idea of the material and ICP dry etching (see Fig. 2, page 14). thermal properties or to make an estimate This type of structure enables to generate based on comparisons with reference phonons in the ridge and to measure the heat materials the thermal conductivity of which is flux flowing to the substrate. An adaptation known. This is not an easy task as a heating of the 3ω method is then used to heat the device has to be deposited on top of each wire and measure the wire temperature. As sample. it has been previously explained, a harmonic If one wants to find the thermal conductivity electrical current generates the heat at 2ω directly from the sample, a physical model due to Joule effect and lock-in detection linking the measured temperature and the allows measuring the in-phase 3ω voltage 13
- 14. nanoresearch (a) (b) in certain cases. The standard wire method deals in general with a few Ohms. Lithography mask 3. Phonons in silicon and some size effects A. Silicon properties Silicon is a semiconductor where electrons (c) (d) are the charge carriers but most of the heat is carried through phonons. The thermal conductivity of pure silicon is at room temperature, which is high in comparison to amorphous materials such as SiO2 with thermal conductivities two orders of magnitude below. Note that gold, one of the best metallic heat conductors, is only conducting heat two times better than silicon. Despite this high thermal conductivity, it can be shown with the Wiedemann-Franz law Fig. 2 > Examples of the fabricated structures: (a) Overall view of one layout (b) Zoom on a ridge that the electronic contribution to thermal with a metal wire before mask removal (c) Zoom conductivity is negligible even at moderate on a ridge with a doped silicon layer showing a doping levels. The phononic thermal nonrectangular shape after reactive ion etching conductivity of a crystal can be written (d) Connection between the measured wire and electrical access./ component proportional to the wire 2ω temperature. Note that one needs generally where the integration spans over the to filter the spurious signal generated by frequency ω and the discrete sum over the the source at 3ω. A 4 points measurement three different polarizations. is the reduced is better usually, but 2 points can be also Planck constant; f is the Bose-Einstein used in some cases. The difference with distribution; T the temperature; g is the the macroscopic method is that a different phononic density of states; vg the phonon physical model has to be used to link the mode group velocity; τ is the phonon wire temperature to the heat flux transmitted relaxation time and (vg τ) is the phonon to the substrate. mode mean free path. The high thermal The first test is to measure the electrical conductivity is therefore due to either high resistance of the device as a function of velocities, large density of states or large the temperature, as varying this parameter phonons mean free path. allows determining the value of the temperature coefficient α needed for the B. Acoustic phonons, mean free paths measurements. It can be found that α In general, mean free paths are parameters is positive or negative depending on the that are not very-well known as they temperature and the type of heater/sensor. (1) depend strongly on the frequency The major experimental difference with the whereas they are generally calculated macroscopic method is the value of the as an average and (2) are very difficult to 14 resistances that can be as high as 20 kΩ measure at room temperature. Some early
- 15. nanoresearch measurements were reported in the 1960s reader to the mentioned references from at lower temperature [9], and more recent the group of Goodson for the study of such experiments using electrical methods [10] phenomena. and time-domain thermoreflectance have shown that part of the phonon mean free 4. Heat conduction in electric tracks and path distribution should lie at lengths above ridges 500 nm. An alternative way to get insights In electronic devices with deposited metal in the issue of mean free path is the use of lines or doped silicon tracks, the electronic molecular dynamics simulation. Henry and path lies on top of planar substrates. Chen [11] recently calculated a distribution Considering a phonon mean free path on of the mean free paths for silicon, finding the order of 100 nm, we present in Fig. 3 indeed that around 30% of the thermal three types of possible devices that consist conductivity was due to mean free paths of a ridge on a planar substrate of the same larger than 1 micrometer. This is consistent or different material. For simplification, we with the estimation [10] that the mean free start with only similar materials. The left path should be around 300 nm for silicon. device can be treated with the usual Fourier Here one should keep in mind that the heat conduction, the middle one is different widespread evaluation of the mean free as even if the nanostructure on top is large path vgτ from is delicate and is in a thermal equilibrium the thermal in the sense that it counts all the optical constriction resistance to the cold bath has modes in the specific heat cp, whereas they to be described by a subcontinuum heat are not expected to play a key role in the conduction. The right device is even further thermal conductivity due to the flatness of complicated as the size of the structure does their dispersion relation (vg≈0). Here ρ is the not allow an equilibrium inside due to its small material density as usual. This approximation size and the fact that phonons are not trapped underestimates the effect of the phonon in the cavity but can also escape. The centre rarefaction in small devices. Recent works figure is typical of the rarefaction effect [19], performed with nanowires [12, 13, 14] and when the phonon statistics impinging the with embedded nanoparticles [15, 16], constriction is different than the equilibrium targeting thermal conductivity reduction one. The right one has been tackled in in thermoelectric materials, have also a theoretical paper [20]. In principle, the highlighted the effect of roughness [14, Boltzmann transport equation has to be used 15] in addition to the phonon-particle for calculating the heat flux in structures such confinement effect. Here we do not discuss as the centre and right ones but approximate the suspended wire issue as it is for the methods have been developed such as the moment less relevant in nanoelectronics. ballistic-diffusive equation [21, 22]. We need also to underline the role of the interaction of electrons and optical (a) (b) (c) phonons with the acoustic ones [17, 18]. Even if optical phonons do not carry heat significantly they interact with the acoustic ones, therefore impacting the thermal conductivity through the scattering mean Fig. 3 > Three types of electrical conductors on a free path. Note also that electron scattering planar substrate. The substrate can be either an with optical phonons is significant, and the electric conductor or an insulator. The Fourier heat redistribution to acoustic phonons description of heat conduction is not adequate takes place through optical/acoustic for the two last devices (b,c) if the phonon mean phonons scattering interaction. We refer the free path is of the order of 100 nm or more./ 15
- 16. nanoresearch Transient Heat Conduction Problems using Ballistic-Diffusive Equations and Phonon Boltzmann Equation”, Journal of Heat Transfer, Vol. 127, pp.298-306 (2005). [3] D. Cahill, “Thermal conductivity measurements from 30 to 750 K: The 3ω method”, Review of Scientific Instruments, Vol. 61, p802 (1990). [4] S.-M. Lee and David G. Cahill, “Heat transport in thin dielectric films,” Journal of Applied Physics, Vol. 81, 2590 (1997). [5] S.-M. Lee, David G. Cahill, and R. Venkatasubramanian, “Thermal conductivity of Si-Ge superlattices,” Applied Phyics. Letters, Fig. 4 > Different regimes of heat conduction as a Vol. 70, 2957 (1997). function of the shape of the body in contact with [6] D.-A. Borca-Tasciuc and G. Chen, “Thermal the substrate. Adapted from Ref [20]./ Properties of Nanochanneled Alumina Templates,” Journal of Applied Physics, Vol. The purpose of our work is to observe 79, pp. 084303-1-9 (2005). these kinds of subcontinuum effects [7] T. Borca-Tasciuc, R. Kumar, and G. Chen, experimentally. We have already measured “Data Reduction in 3ω Method for Thin Film [23] the expected strong reduction of the Thermal Conductivity Measurements,” Review thermal conductance in the ballistic regime of Scientific Instruments, Vol. 72, o. 4, pp. with respect to Fourier’s prediction. Our 2139-2147 (2001). first results indicate in addition a different [8] T. Tong and A. Majumdar, “Reexamining behavior than what can be calculated in the the 3-omega technique for thin film thermal purely ballistic case, which is exactly what characterization”, Review of Scientific is pointed out in the analysis developed in Instruments, Vol. 77, 104902-104902-9 (2006). Figs. 3 and 4. [9] R. Gereth and K. Hubner, “Phonon MeAn Free Path in Silicon Between 77 and 250°K”, Acknowledgements Physical Review, Vol. 134, pp A235–A240 (1964). We thank M. Tilli for providing high ohmic [10] M. Asheghi, Y.K. Leung, S.S. Wong, and K.E. 8” Si wafers. M. Myronov and V. Shah are Goodson., “Phonon-Boundary Scattering in acknowledged for doing the n+ Si epitaxial Thin Silicon Layers,” Applied Physics Letters, growth. Vol. 71, pp. 1798-1800 (1007). We acknowledge the partial support [11] A. Henry and G. Chen, “Spectral Phonon of the EU projects NANOPACK and Properties of Silicon Based Molecular Dynamics NANOPOWER. P.O.C. acknowledges the and Lattice Dynamics Simulations,” Journal of support of EU project nanoICT for the partial Computational and Theoretical Nanosciences, funding of a stay at VTT. Vol. 5, pp. 141-152 (2008). [12] D. Li, Y. Wu, P. Kim, L. Shi, P. Yang, A. References Majumdar, “Thermal conductivity of individual [1] D.G. Cahill, W. K. Ford, K.E. Goodson, G.D. silicon nanowires,” Applied Physics Letters, Vol. Mahan, A. Majumdar, H.J. Maris, R. Merlin, and 83, pp. 2934-2936 (2003). S.R. Phillpot, “Nanoscale thermal transport,” [13] A. I. Hochbaum, R. K. Chen, R. D. Delgado, Applied Physics Reviews, Journal of Applied W. J. Liang, E. C. Garnett, M. Najarian, A. Physics, Vol. 93, p793 (2003). Majumdar, P.D. Yang. “Enhanced thermoelectric [2] R.G. Yang, G. Chen, M. Laroche, and Y. Taur, performance of rough silicon nanowires,” 16 “Simulation of Nanoscale Multidimensional Nature, Vol. 451, pp. 163-167 (2008).
- 17. nanoresearch [14] I. Akram Boukai, Y. Bunimovich, J. Tahir-Kheli, Phonon Dispersion,” Journal of Applied J.-K. Yu, W.A. Goddard III, and J.R. Heath, Physics, Vol. 96, no. 9, pp. 4998-5005 (2004). “Silicon nanowires as efficient thermoelectric [19] G. Chen, G. Chen, “Nonlocal and materials”, Nature, Vol. 451, pp168-171 (2008). Nonequilibrium Heat Conduction in the Vicinity [15] W. Kim, J. Zide, A. Gossard, D. Klenov, S. of Nanoparticles,” ASME Journal of Heat Stemmer, A. Shakouri, A. Majumdar, “Thermal Transfer, Vol. 118, pp. 539-545 (1006). conductivity reduction and thermoelectric figure [20] S. Volz and P.-O. Chapuis, “Increase of thermal of merit increase by embedding nanoparticles resistance between a nanostructure and a surface due to phonon multireflections”, Journal in crystalline semiconductors,” Physical Review of Applied Physics, Vol. 103(3), p034306 (2008). Letters, Vol. 96, p045901 (2006). [21] G. Chen, “Ballistic-Diffusive Heat Conduction [16] W. Kim, S. L. Singer, A. Majumdar, J. M. O. Equations,” Physical Review Letters, Vol. 85, Zide, D. Klenov, A. C. Gossard, S. Stemmer, pp. 2297-2300 (2001). “Reducing thermal conductivity of crystalline [22] G. Chen, “Ballistic-Diffusive Equations for solids at high temperature using embedded Transient Heat Conduction from Nano to nanostructures,” Nano Letters, Vol. 8, pp. Macroscales”, Journal of Heat Transfer, Vol. 2097-2099 (2008). 124, pp. 320-328 (2002). [17] S. Sinha, E. Pop, R.W. Dutton, K.E. Goodson, [23] P.O. Chapuis, M. Prunnila, A. Shchepetov, L. “Non-Equilibrium Phonon Distributions in Sub- Schneider, S. Laakso, J. Ahopelto, and C.M. 100 nm Silicon Transistors,” ASME Journal of Sotomayor, “Effect of phonon confinement Heat Transfer, Vol. 128, pp. 638-647 (2006). on heat dissipation in ridges”, Proceedings of [18] E. Pop, B. Dutton, and K.E. Goodson, “Analytic the 16th International Worshop on THERMal Band Monte Carlo Model for Electron Transport INvestigations of ICs and Systems (THERMINIC), Modeling in Si Including Acoustic and Optical B. Courtois and M. Rencz (ed), (2010). 17
- 18. nanojobs • PostDoctoral Position (CEA-Léti, We are looking for talented chemists France): ”Theory and modeling of NEMS motivated to pursue a PhD in the area of based digital functions” Materials Science. The students will be CMOS technology scaling has enabled able to join to a pioneer, dynamic and significantly reduced energy per operation active group from the Department of in integrated circuits. However, these NanoScience and Organic Materials (www. improvements are not in line with the icmab.es/nmmo). This research group expected performances of future approaches some of the most exciting and challenging fields that a chemist and autonomous systems. Autonomous a materials scientist can explore nowadays systems that use energy harvesting are — the study of advanced organic functional attractive for many applications (medical materials and nanoscopic systems with implants, micro-sensors, internet of things useful electronic (superconductors, metallic devices…). Today technologies used in conductors, semiconductors), magnetic autonomous systems are not efficient (ferromagnets, superparamagnets, single enough for ultra low power applications. The molecule magnets, nanoporous magnets, transistor threshold voltage has been scaled etc), biological and/or optical properties. to optimally balance leakage and dynamic We also involved on the development of power but optimized performances are materials processing techniques, molecular below autonomous systems specifications. self-assembly and on the preparation of The deadline for submitting applications functional nanostructured materials. is October 13, 2011 The deadline for submitting applications For further information about the position, is October 15, 2011 please contact: For further information about the position, Hervé Fanet (herve.fanet@cea.fr) please contact: Jaume Veciana (vecianaj@icmab.es) • PostDoctoral Position (CEA-Léti, France): ”Characterization of a flexible array • PhD Position (IMM - CSIC, Spain): of tactile sensors” ”New paradigms and New Devices based We aim to develop a flexible tactile sensor on Nanomechanics” based on MEMS technology developed The aim of this PhD project is the at CEA Léti. Three-axis force sensors development of new NEMS devices and developed at Léti and already tested for new sensing paradigms to achieve the texture measurements will be integrated in ultimate limits in biological detection based an array. The work proposed will include on nanomechanics. Silicon nanowires both the integration and characterization of together carbon nanotubes represent the flexible sensor array. the ultimate limit in the minituarization of The deadline for submitting applications nanomechanical resonators. It is expected is October 13, 2011 that these devices can be applied for ultrasensitive mass sensing at the sub- For further information about the position, zeptogram level and for mass spectroscopy please contact: of single biomolecules. However, the Caroline Coutier achievement of the optimal performance (caroline.coutier@cea.fr) of these devices requires a detailed understanding of the nanomechanical • PhD Position (ICMAB - CSIC, Spain): response and a major development of the “Functionalisation of surfaces with optical instrumentation for the detection of functional organic molecules for electronic the picometer scale vibrations. In this PhD 18 or biological applications” project advanced optical instrumentation
- 19. nanojobs and modeling of the nanomechanical and Group and ETSF Scientific Development optical response of the silicon nanowires Centre in Spain and the Theory group of will be developed. Finite element simulations the Fritz-Haber-Institut in Berlin. The aim and analytical models will be developed in of the research project is to develop new order to describe how the static and dynamic concepts for understanding, identifying, response of nanomechanical systems with and quantifying the different contributions different geometries behaves when subject to energy harvesting and storage as well as to biological adsorption. The final aim will be describing transport mechanisms in natural to establish the potential for weighing single light harvesting complexes, photovoltaic biomolecules and measuring molecular materials, fluorescent proteins and artificial recognition at the level of few events. (nanostructured) devices by means of The deadline for submitting applications theories of open quantum systems, non- is October 18, 2011 equilibrium processes and electronic For further information about the position, structure. please contact: The deadline for submitting applications Montserrat Calleja is October 31, 2011 (mcalleja@imm.cnm.csic.es) For further information about the position, please contact: • PostDoctoral Position (ICFO, Spain): ”Optics and Photonics” Angel Rubio (Angel.Rubio@ehu.es) ICFO – The Institute of Photonic Sciences is • PostDoctoral Position (CEA-Léti, a center based in Castelldefels (Barcelona), France): ”CMOS electro-optical bridge for Spain, devoted to the research and network-on-chip and optical network” education of the optical and photonic sciences, at the highest international level. The forecasted developments of high- No restrictions of citizenship apply to the performance computing (HPC) and ICFO post-doctoral contracts. Candidates Cloud computing induce new needs for must hold an internationally-recognized computation density and data mining. The PhD-equivalent degree in a field of science architectural model is built from a large and engineering related to optics and number of processors sharing a huge photonics. Suitable backgrounds include memory (eventually several terabytes). optics, physics, mathematics, electronics Besides, while networks-on-chip (NoC) are and telecommunications engineering. becoming the dominant interconnection The deadline for submitting applications paradigm within chips, the connections to is October 20, 2011 large memories are still point-to-point. The For further information about the position, gap between the theoretical computing please contact: power and the effective or real computing Ariadna García power is hence widening because of (ariadnag@heuristica.org) bandwidth limitations to shared memory and increasing communication latency. • Postdoctoral and PhD positions Emerging high-bandwidth connection (University of the Basque Country UPV/ standards (DDR3, WideIO…) remain EHU, Spain): ”Dynamical processes in incremental solutions and do not allow Open Quantum Systems” concurrent accesses to a large number of Applications are invited for postdoctoral and memory banks. PhD positions link to a five year project on The deadline for submitting applications the topic of Dynamical processes in open is October 31, 2011 quantum systems as part of an European Research Council Advanced grant (DYNamo For further information about the position, project). The project will be conducted please contact: between the NanoBio Spectroscopy Yvain Thonnart (yvain.thonnart@cea.fr) 19
- 20. www.nanociencia.imdea.org RESEARCH PROGRAMMES • Molecular nanoscience IMDEA-Nanociencia is a private Foundation created by joint initia- tive of the Comunidad de Madrid and the Ministry of Education of • Scanning probe microscopies the Government of Spain in February 2007 to manage a new and surfaces research Institute in Nanoscience and Nanotechnology (IMDEA- Nanociencia). The Institute is located at the campus of the Univer- • Nanomagnetism sidad Autónoma de Madrid in Cantoblanco. The Institute aims at performing research of excellence in selected • Nanobiosystems: biomachines and manipulation of macromolecules areas and offers attractive opportunities to develop a career in sci- ence at various levels from Ph.D. students to senior staff positions. • Nanoelectronics and superconductivity The Madrid Institute for Advanced Studies in Nanoscience also develops an important program of technology transfer and creation of spin-off companies. • Semiconducting nanostructures and nanophotonics E-mail contacto.nanociencia@imdea.org Phone 34 91 497 68 49 / 68 51 Fax 34 91 497 68 55 • Nanofabrication and advanced instrumentation [Nanociencia y Nanotecnología: lo pequeño es diferente small is different Nanoscience and Nanotechnology: ]
- 21. nanoICT Conf Report 7th International Thin Film Transistor Conference-ITC 2011 3-4 March 2011, Clare College, Cambridge Organisers were presentations on CNTs and nanowires W.I.Milne > Engineering Dept, University of for use in TFTs. In terms of distribution, an Cambridge, UK. equal balance in presentations was achieved Arokia Nathan > Electrical & Electronic Engineering University College, London, UK. between materials and applications, fulfilling www-g.eng.cam.ac.uk/edm/itc2011/ the primary theme of ITC2011. Much of the meeting concentrated on the Sponsored by: production, characterisation and application of metal oxide based semiconductors although there were also several reports on the use of organic based material systems for TFTs. Metal Oxide transistors are becoming increasingly important as their mobility is much higher than those of amorphous The aim of this meeting was to highlight silicon based TFTs and, as their stability the on-going work on Thin Film Transistors is improved, their use in practical systems (TFTs), including a-Si:H and related materials including flat panel displays, sensors and systems such as nano, micro and poly LEDs cannot be far away. crystalline silicon. Sessions however also included work on metal oxides, organics, There were 7 oral sessions and a poster semiconducting nanowires, carbon session on both afternoons. The first session nanotubes (CNTs) and naturally the new was based on Materials & Processing “material of choice” graphene. Thin-film and the invited papers in this session were Transistors (TFTs) have become increasingly presented by Hiroshi Tanabe from NEC important since amorphous silicon (a-Si:H) and Richard Wilson from CDT. Dr Tanabe’s TFTs were first incorporated in the backplanes presentation was on TFT technologies for in AMLCD TVs. They of course are now being Flexible Displays based on the production considered for a variety of other applications of metal oxide TFTs at low temperatures including RFID tags, sensors, smart tags, using an excimer laser annealing technique. etc. and increasingly in flexible electronics. Richard Wilson’s talk concentrated on However the electronic properties of a-Si:H solution processing of organic TFTs with field limit its possible applications and a variety effect mobilities in excess of 1 cm2V-1 cm-2. of different material systems are now being The optimisation of the solvent selection investigated as alternatives. from which the material is deposited is key to This year the conference theme was on enhancing and controlling crystalline domain Novel Materials, Processing and Device- formation. Circuit Integration. There were 110 abstracts Session 2 and 8 concentrated on Thin Film submitted and 150 attendees many from the Transistors themselves and the invited talks Far East. here were given by Kenji Nomura from Tokyo There were 17 invited speakers and 29 Tech and Elvira Fortunato from FCT-UNL, contributed papers who presented their work Portugal. In his presentation Prof Nomura on a variety of thin film material systems. There described the work they have been doing to 21
- 22. nanoICTConfReport improve the stability of a-In-Ga-Zn-O TFTs graphene by Markuu Rouval from the Nokia and Prof Fortunato’s talk concentrated on Research Centre. transparent electronics with emphasis on 74 posters were presented in the two the production of both p-type and n-type sessions and the banquet was held in TFTs. Gilles Horowitz from the Université Clare College which is the second oldest Denis-Diderot covered the modelling Cambridge College, having been founded of organic TFTs and Simon Ogier from in 1326. PeTEC presented their work on backplane technologies for flexible displays. All the sessions were exceedingly well attended despite a tight two-day program Novel devices and their applications were described in Session 3. Sigurd Wagner from with back-to-back talks and posters. Princeton reviewed their work on self aligned Excellent feedback was received from the amorphous silicon transistors and Yue Kuo attendees on the technical quality of the from Texas A&M described his work on program and the general organization. non-volatile memory based devices based ITC 2012 will be held in Lisbon, hosted by on floating gate amorphous silicon TFTs. Uninova, in January 2012. This was followed by Mutsuku Hatano from Tokyo Tech who gave her vision of the future- nanoICT Coordination Action (nanoICT) integration of wireless-communication www.nanoict.org functions on Display Panels using TFT technology. Sessions 4 and 6 looked at TFT circuits and System Integration and involved 4 further invited talks. Prof Takao Someya from Tokyo The nanoICT plan to strengthen scientific University gave an excellent presentation and technological excellence will go beyond on his work on foldable and stretchable the organisation of conferences, workshops, electronics using organic based transistors exchange of personnel, WEB site, etc. and memories and this was followed by Prof developing the following activities: Jin Jang from Kyung Hee University in Korea who presented their research on the stability 1. Consolidation and visibility of the and flexibility of a-IGZO Transistors on plastic research community in ICT nanoscale and their application to circuits. devices The second session (Session 5) on Materials 2. Mapping and benchmarking of research and Processing was held on the morning of at European level, and its comparison the second day and mostly concentrated on with other continents metal oxide materials and devices. Andrew 3. Identification of drivers and measures Flewitt from Cambridge University and to assess research in ICT nanoscale Thomas Anthopoulos from I.C. were the devices, and to assess the potential invited speakers and covered respectively of results to be taken up in industrial insulators and semiconducting materials research deposited at low temperature using a novel sputtering method and spray pyrolysis 4. Coordination of research agendas and processed ZnO for use in TFT manufacture. development of research roadmaps Session 7 was sponsored by nanoICT EU 5. The coordination of national or regional project and the invited talks were on CNTs for research programmes or activities, with TFTs by Prof Didier Pribat of Sungkyunkwan the aim to involve funding authorities in 22 University in Korea and Circuits based on building the ERA around this topic.
- 23. nanoresearch The raise up of UHV atomic scale interconnection machines J. S. Prauzner-Bechcicki1, D. Martrou2, C. Troadec3, S. Gauthier2, M. Szymonski1 and C. Joachim2,3 1Center for nanometer-Scale Science and Advanced Materials (NANOSAM), Faculty of Physics, Astronomy and Applied Computer Science Jagiellonian University, Reymonta 4, Krakow, Poland. 2Centre d’Elaboration de Matériaux et d’Etudes Fig. 1 > A single five wings molecule-motor [1] Structurales (CEMES-CNRS), 29, rue Jeanne Marvig, positioned between a 4 Au nano pads junction BP 94347, 31055 Toulouse Cedex 4, France. constructed at the Si(100)-H surface. The 4 black 3Institute of Materials Research and Engineering, wires getting out of the surface are indicative of the A*STAR (Agency for Science, Technology and interconnections step 3 discussed in the text depending Research), 3 Research Link, Singapore 117602. on the electronic gap of the supporting surface./ 1. Introduction In section 2, the general principles of the Single molecule mechanics [1], mono- few UHV atomic scale interconnection molecular electronics [2] and surface machines under test to solve the problem are atomic scale circuits [3] [4] are all requiring a described. Depending on the electronic gap specific surface interconnection technology of the surface where the atomic scale devices with an atomic precision and cleanness and machineries are supposed to work, two [5]. In a planar configuration, this surface families of interconnections machines are technology must be able to provide multiple being explored. Section 3 is providing one access electronic channels to the atomic (or example of an atomic scale interconnection molecular) scale machinery constructed on a machine designed for the surface of wide surface (see for example Fig. 1). At the end gap semi-conductor and insulator materials. of the 80’s, it was expected that the e-beam Section 4 is giving the example of two nano-lithography technique would be able interconnection machines for moderate gap to provide such a technology [6]. But with its semi-conductor surfaces. The design and resist based approach, e-beam technique instrumentation works reported here are will not face the challenge [7] because it is the consequence of the EU ICT integrated not able to respect at the same time the project Pico-Inside in Krakow and Toulouse atomic scale precision, the cleanness and together with the A*STAR VIP Atom Tech the expected large number N of access Phase 2 project in Singapore. It is now further channels to the atomic scale machinery developed in the new EU ICT integrated [8]. Alternative nanolithography techniques project AtMol and in the Phase 3 of the such as nano-imprint [9] or nano-stencil A*STAR VIP Atom Tech project in Singapore. [10] are neither adapted to encompass all the interconnection stages from the 2. Atomically precise electrical macroscopic to the atomic scale nor clean interconnection machine enough down to the atomic scale. At the turn An atomic scale precision, multiple access, of the century, this problem triggers a new electrical interconnection instrument must approach to planar electrical interconnects provide N conducting wires converging starting from the bottom that is from the toward a very small surface area where fundamentals of surface science. an active machinery (see Fig. 1 for a N=4 23
- 24. example) has been constructed with an atomic scale precision. Those N interconnects are positioned somewhere on a large wafer surface. As a consequence, a very efficient navigation system must be designed to locate this very small active area from a macroscopic perspective while keeping the local atomic precision of the interconnection. The solution to this navigation requirement is to combine two types of microscopy: a far field one (optical, scanning electron microscope (SEM)) for large scale navigation and a near field one (Scanning Tunneling Microscope (STM), Atomic Force Microscope (AFM)) for the atomic scale part with a full overlay between those 2 types of microscopy. An UHV atomic scale interconnection machine is Fig. 2 > Scheme of the atomic scale interconnection designed to follow a dedicated interconnection machines for (a) wide and (b) moderate surface protocol. On an atomically clean well-prepared band gap substrates. A: Atomic scale circuitry, surface, an atomic scale circuitry is fabricated B: Contacting metallic nanopads, C1: Ultrasharp (A). To reach a large number N of interconnects metallic tips, C2: Nanowires, D: Microelectrodes, E: and to be able to interconnect each atomic Metallic microcantilevers./ wire to the external world, there is a necessary lateral extension of this circuit to reach N 3. UHV interconnection machine for large contacting metallic nanopads (B) that are surface gap positioned around the atomic scale circuit. In the example of Fig. 2, a molecule is connected For a large valence-conduction band to these nanopads by atomic metallic wires. electronic surface gap (more than a few eV Depending on the electronic surface gap of up to 8 eV for standard insulators), SEM the supporting material, the nanopads (B) is difficult to use as a navigation far field have to be contacted from the top by a series microscope because its electron beam of N atomically sharp metallic tips (C1) or by will charge the surface. In this case, an a series of N nano-scale wires (C2) up to the optical microscope is natural candidate for point where mesoscopic metallic wiring or coarse-grained positioning. It determines microelectrodes (D) can be surface fabricated the minimum length of metallic surface and contacted by a series of N micro-scale wiring which must be fabricated starting metallic cantilevers (E) also from the top of from the nano-pads (B) in Fig. 2a toward the wafer. During the process, the sequence the next contact stage based on metallic of those different steps depends on the micro-cantilever. Fortunately enough, with a machine and on the supporting material. What large surface gap, the surface area of those is triggering the choice of the interconnection interconnects can be expanded horizontally technology between C1 and C2 (and after the without too much lateral leakage current need for the D and E interconnection steps between the different electrodes. This is the in Fig. 2a) is the electronic gap of the surface basis of the UHV interconnection machine that in turn will determine the kind of far field described in this section where a low microscopy to be used for navigation over the temperature approach is not compulsory 24 wafer surface. but preferable.
- 25. nanoresearch To realize the 5 levels of interconnect 1. a flexural-hinge guided (XY) nano- described in Fig. 2a in UHV, the deposition positioner stage (100 μm x 100 μm, of molecules, their observation by NC-AFM repeatability 5 nm) with a closed loop and the measurement of their electrical control based on capacitive sensors, properties, the Toulouse group has designed 2. an evaporation system highly collimated and constructed a dedicated UHV equipment on the cantilever to perform nano-stencil called DUF (DiNaMo UHV Factory). This deposition, equipment allows transferring samples under UHV between five complementary UHV 3. a (XYZ) piezo driven table for positioning chambers (see Fig. 3): the metallic micro-combs for the electrical contacts, (1) an MBE growth chamber dedicated to nitride semiconductors growth, 4. an optical microscope to control the metallic nano-pads growth and stencil positioning of the micro-combs. evaporation for microelectrodes These modifications were introduced by (2) a room temperature AFM/STM chamber the mechanical workshop of the Toulouse for surface characterization by STM and laboratory. The main advantage of using NC AFM a commercial UHV AFM/STM is to benefit from the good characteristics for SPM (3) an AFM/STM chamber modified for imaging. But the piezo tube used to scan nano-stenciling experiments and has a range of a few μm only. The addition of electrical measurements a piezo table to move the sample offers the (4) a preparation chamber for cleaning possibility to perform wide range scanning, substrates, STM tips and AFM cantilevers up to 80 μm SPM images, while keeping the (5) a mass spectrometer chamber possibility to realize atomic sale imaging with transformed in a molecular ions source. the piezo tube. For (3) a UHV Omicron Nanotechnology One of the disadvantages is the small VT STM/AFM head has been modified to accessible space around the SPM head. accommodate different tools, namely [11]: Indeed, it is not possible to place an optical Fig. 3 > The DUF (DiNaMo UHV Factory) equipment allows to transfer samples between 5 complementary UHV chambers in order to realize the 5 levels of interconnect on wide band gap semiconductors (GaN, AlN)./ 25
- 26. nanoresearch microscope with normal incidence with Krakow and the other in Singapore. The respect of the substrate, and an atomic Krakow’s system consists of three basic source for the nano-stencil experiments with segments: multi-probe, low-temperature normal incidence with respect to the AFM scanning probe microscope (LT-SPM) cantilever. In our case, the image obtained and preparation chambers. Multi-probe by the optical microscope comes from a segment is composed of 4-probe scanning mirror with an angle of 30° with the substrate tunnelling microscope (STM) combined plane. This gives distorted images, with with high resolution scanning electron a loss of resolution: only 3 μm instead of microscope (HR-SEM) and hemispherical 1 μm in normal incidence. The effusion cell electron energy analyser (scanning Auger is fixed on a port of the UHV chamber that microscope, SAM) (see Fig. 4). The Auger makes an angle of 33° with the horizontal microscope part is the element not present plane, and another angle of 28° between in Singapore’s setups. Composition of the the two vertical planes passing through the multi-probe segment allows surface element evaporation beam and the central axis of analysis, imaging and measurements of the cantilever. This orientation of the atom nanostructures conductance with very beam induces distortion, which should be high-resolution. In accord with the Fig. 2b taken into account in the design of the nano- principle, HR-SEM may act as a navigation pattern to be drilled into the pyramidal tip of to precisely position each of the 4 STM tips the cantilever [11]. (a) 4. UHV interconnection machine for moderate semi-conductor surface gap For a moderated valence band-conduction band electronic surface gap (around a few eV), it is not possible to use very long surface metallic circuitry due to the possible lateral surface leakage current between the surface electrodes. In this case, one solution is to use ultra sharp STM like tips positioned from the top on the surface as microelectrodes (Fig. 2b). In this case, the core of the tips will (b) not be in contact with the supporting surface and one can go continuously from a tip apex radius of curvature of a few nanometer up to a 100 microns or more section for the tip body. In this case, navigation on the surface can be performed using an UHV- SEM (Fig. 2b) by grounding the sample during the SEM imaging to avoid the surface charging effect. This is the basics of the UHV interconnection machines described here. A low temperature approach is compulsory with those systems because of Fig. 4 > (a) View on sample stage of 4-probe the low electronic gap at the surface of the microscope; in upper part one can see SEM column and next to it an entrance to hemispherical supporting material. electron energy analyser; below SEM column There are two apparatuses that realize the there are three of four STM probes. (b) SEM image 26 above described design, one is housed in of four STM probes./
- 27. nanoresearch that will be used as microelectrodes. First measurements of conductance of gold nanostructures on Ge(001) surface are in progress. Next, LT-SPM segment consist of scanning probe microscope that may work both as STM and NC-AFM in a range of temperatures Fig. 5 > InSb surface imaged with q-sensor NC- from 4K up to room temperature. Thanks AFM in temperature 4K./ to use of scanner and sample holder embedded in a cryostat, the LT-UHV STM SAM images of metallic nano-mesa grown allows for a very high resolution imaging, as on semiconductor substrate are shown. well as, stable spectroscopic measurements The Ag/Ge(111) is an example of a system and atomic scale manipulations. for which depending on the deposition Furthermore, NC-AFM mode is based on conditions, on the successive thermal q-sensor device (tuning fork) that enables annealing and on the amount of deposited imaging of conducting, semiconducting and material the resulting overlayer morphology insulating samples at low temperatures (see can be switched from an atomically smooth Fig. 5) and, if required, also simultaneous to a columnar-like [12]. Sample is prepared in measurements of tunnelling current. This the following way: silver in amount of nominal option makes the Krakow’s system a very 5 ML is deposited on the germanium surface powerful tool. Last but not least, is the kept at low temperature. Low energy electron preparation segment that consist of typical diffraction studies performed immediately preparation equipment such as a XYZ after deposition reveal that compact silver manipulator with electric contact allowing for film is crated. Such a conclusion follows from resistive heating up to 1000K (using a direct the fact that reflections characteristic for heating mode 1200K may be achieved), the unreconstructed Ag(111) overlayer are furthermore the manipulator allows for observed exclusively on LEED image. One cooling the sample down to 100K with may assume it is a clear and direct indication nitrogen vapour, an ion gun, a low energy that Ge substrate is completely buried. electron diffraction system for quick sample In the next step the sample is annealed to quality tests and several ports allowing room temperature. After annealing a massive for incorporation of additional elements (for instance effusion cells or quartz microbalance thickness monitor) into the chamber. All segments are composed of the highest quality elements all of them being compatible to work in UHV environment (less than 3×10-10mbar), and thus allowing Fig. 6 > STM, HR-SEM and SAM images of the Ag/ for conducting very complex experiments Ge(111) sample. Image size: 220nm × 270nm. Left in a single set-up in a very controlled panel: STM image; Middle panel: HR-SEM image; way, starting from sophisticated sample Right panel: SAM image. STM image reveals bright preparation and ending with extensive and nanostructures 2.5nm high. The same regions complete characterisation. are marked red in HR-SEM image. In SAM image those structures are black. SAM image was taken As the Auger microscopic part of the for Ge line (E=1144 eV), thus exposing as bright Krakow’s nano-probe instrument is not regions containing Ge. Therefore, it is possible to present in the other setups (see above and identify the black structures in SAM image (bright below) its potential is briefly described in the and red regions on STM and HR-SEM images, following. In Fig. 6, the STM, HR-SEM and respectively) as silver islands./ 27
- 28. nanoresearch morphological reorganization of the Ag film is (E = 352 eV). A strong signal contrast in observed that leads to formation of metallic the Ge-SAM image (Fig. 6, right panel) is islands. Left panel in Fig. 6 contains STM observed. It clearly indicates the metallic image of the Ag/Ge(111) sample annealed composition of the nanomesa against the to room temperature. As clearly seen, the semiconductor substrate. Bright regions Ag continuous film has been converted into on right panel in Fig. 6, reveal a germanium a network of “nano-mesa” of 10 ML height content, whereas black regions prove lack with atomically flat tops and extended pits of the germanium. Those black regions exposing the underlying wetting layer (bright nicely correspond to red structures seen and dark structures on left panel in Fig. 6, in HR-SEM image (Fig. 6, middle panel) respectively). The atomic steps bounding and to bright regions in SEM image (Fig. the edges are developing into considerably 6, left panel). Thus analysis with Auger elongated ridges, and appear to be spectroscopy strengthens the conclusion predominantly oriented along the threefold that observed nanostructures are truly silver symmetry of the (111) substrate as proved nanomesa. Furthermore, in the image, the by the HR-SEM analysis. Ag ridges of the width as low as 20 nm can be resolved providing the capability of high To further confirm metallic composition lateral resolution of our Auger microscope of the observed topographic structures based on the Gemini electron column. Here Auger spectroscopy in scanning mode is it is worth noting, that in the Ag-SAM image performed. The corresponding Scanning (not shown), together with the high lateral Auger Microscopy images were taken resolution a lower contrast of the Auger at the Ge line (E = 1144 eV) and Ag line signal is observed. The reason for that is 28
- 29. nanoresearch the composition of the wetting layer which is actually the intermixed layer of Ag and Ge. Following the Fig. 2b principle, the Singapore system consists of five UHV chambers interconnected in a star configuration. A first chamber is dedicated to the surface wafer preparation. A second one is dedicated to metallic nano-islands deposition and transfer (when necessary). A third chamber is equipped with a LT-UHV STM for fabricating the atomic wires, manipulating the molecule logic gate and positioning the metallic nano-islands, one on each termination Fig. 7 > The final IMRE Singapore interconnection machine designed in Singapore and constructed of an atomic wire. The fourth chamber is by Omicron in 2010. Five main UHV chambers: equipped with a very efficient UHV-SEM with (A) The preparation chamber with the FIM tip a resolution better than 5 nm to be able to preparation chamber, (B) the LT-UHV-STM from image the contacting metallic nano-islands Omicron Nanotechnology and (C) the nano- and to position a tip apex on each of these probe system from Omicron Nanotechnology nano-islands. Finally, a lateral chamber is with its UHV-SEM (3 nm resolution) and the 4 dedicated to the UHV tip preparation, integrated STM. (D) is the delicate LT transfer sharpening and cleaning. The details of the from the LT-UHV-STM to the UHV-SEM. The IMRE Singapore interconnection machine UHV transfer print chamber is behind the LT- are presented in Fig. 7. UHV-STM chamber B./ 29
- 30. nanoresearch atomic scale interconnection machines housed in Toulouse, Krakow and Singapore possess potential to meet requirements and demands arising upon realization of such a connection [13]. Furthermore, experience and knowledge gained by Krakow’s and Toulouse’s groups through EU ICT integrated project Pico-Inside and by Singapore’s group in parallel through A *STAR VIP Atom Tech Phase 2 set a very Fig. 8 > SEM image of the tip apex of 4 chemically etched tungsten tips converging toward 4 Au fortunate starting point for that quest which nano-islands which have been manipulated one by is now in full development in the new ICT one to enable a 4 points like surface conductance integrated project AtMol (www.atmol.eu). measurement. This image has been recorded on a multi-probe system from Zyvex./ References [1] Rapenne G, Launay J-P and Joachim C 2006 Under the UHV-SEM, each nano-island of a J. Phys.: Condens. Matter 18 S1797. contacting nanostructure can be electrically [2] Joachim C, Gimzewski J K and Aviram A 2000 contacted from the top using one ultra- Nature 408 541. sharp tip per nano-island. Each tip apex is [3] Wada Y 1996 Microelectronic Engineering 30 positioned using the UHV-SEM for a precise 375. navigation. The soft contact between a nano- [4] Ample F, Ami S, Joachim C, Thieman F and island and its tip apex is controlled by the Rapenne G 2007 Chem. Phys. Lett. 434, 280. feedback loop system of an STM meaning [5] Joachim C 2002 Nanotechnology 13 R1. that there is one STM control electronic per [6] Itoua S, Joachim C, Rousset B and Fabre N tip. A SEM image of the contact preparation 1992 Nanotechnology 3 10. of 4 nano-islands manipulated on purpose in [7] Saifullah M S M, Ondarcuhu T, Koltsov a row on a MoS2 surface is presented in Fig. D F, Joachim C and Welland M 2002 8 together where the 4 tip apex approaching Nanotechnology 13, 659. for the contact are also imaged. Electrical [8] Cacciollati O, Joachim C, Martinez J-P and measurements are now in progress to first Carsenac F 2004 Int. Journ. Nanosci. 3 233. record the “nano-pad – MoS2 surface – [9] Chou S Y, Krauss P R, Renstrom P J 1996 nano-pad” I-V characteristics for an inter Science 272 85. nano-pad distance lower than 10 nm. Then, [10] Thet N T, Lwin M H, Kim H H, Chandrasekhar metallic nano-islands will be transfer printed N and Joachim C 2007 Nanotechnology 18, on Si(100)H surface for surface atomic wire 335301. conductance measurements. [11] Guo H, Martrou D, Zambelli T, Dujardin E and Gauthier S 2008 Rev. Sci. Instrum. 79 103904. 5. Conclusion [12] Krok F., Buatier de Mongeot F., Goryl M., Connecting the atomic (or molecular) scale Kolodziej J.J., and Szymonski M., Phys. Rev. B machinery to the outer world respecting the 81, 235414 (2010). atomic scale precision of the construted [13] C. Joachim, D. Martrou, M. Rezeq, C. Troadec, machinery is a very challenging and Deng Jie, N. Chandrasekhar and S. Gauthier J. 30 complex task. Each of the three described Phys. CM, 22, 084025 (2010).