Content uploaded by Cedric Troadec
Author content
All content in this area was uploaded by Cedric Troadec on Oct 14, 2016
Content may be subject to copyright.
Nanogears mechanics: from a single molecule to
solid state nanogears on a surface
We-Hyo Soe1, Cedric Troadec1, Carlos Manzano1, Jie Deng1, Francisco
Ample1, Yang Jianshu2, and Christian Joachim1,2
1 IMRE, A*STAR (Agency for Science, Technology and Research), 3 Research Link, 117602
Singapore
2 CEMES/CNRS, PicoLab, 29 Rue Marvig, BP 94347, 31055 Toulouse Cedex, France
Abstract The first experimental demonstration of a controllable rotating molecule
gear is presented. A scanning tunneling microscope is used to construct, manipu-
late and observe the molecule gear rotation. The appropriate combination of mole-
cule design, molecule manipulation protocol and surface atomic structure selection
leads to the functioning of the molecule gear. Rotation of the molecule gear is
done step-by-step and totally under control. The fabrication of solid-state SiO2
nanogears with diameters ranging from 30 nm up to 1μm is also presented. They
can be manipulated using an atomic force microscope tip on a graphite surface.
Ranging in sizes from few tens of nanometers up to sub-micron diameters, they
are going to enable the transmission of mechanical motion from functional me-
chanical molecule machineries to larger sub-micron or micron sized devices
through series of solid state gears and mechanical components compatible with the
semiconductor and electronics industry technology.
Keywords molecule-gears - STM manipulation, nanolithography- solid state
nanogears. AFM manipulation
1 Introduction
Miniaturizing mechanical components is important not only to enable the crea-
tion of mechanical nanomachines, but also for data input on an atomic scale cir-
cuit, information encoding, energy harvesting, and also possibly production of
energy from the background. In this regard, as of 2007 solid-state gears (one of the
basic components of any mechanical machinery) have been to diameters down to
500 nm downsized making use of a top-down approach.[1] In 2009, a gear shaped
molecule, hexa-t-butyl-pyrimidopentaphenylbenzene (HB-NBP) i.e. C64N2H76, has
been used to construct a molecule gear by pinning it on an atom sized axle and
rotated successfully using a scanning tunneling microscope (ST M).[2]
2
This molecule gear having 1.2 nm diameter equipped with six teeth is the smallest
known working gear.
The construction of a rotating molecule gear device evidenced that gears with in-
termediate sizes in between 500 nm to 1.2 nm in diameter were not existing. Then,
we engaged the exploration of intermediate nanogear size nanofabrication be-
tween a molecule gear and the 500 nm solid state nanogear limit of 2007. In this
chapter, we develop this story into two parts. First, we present the intentionally
constructed molecule gear where an HB-NBP molecule is mounted and centered
on an atom axis uprising at a herringbone elbow on an Au(111) reconstructed sur-
face. Thereafter, we describe the nanofabrication process of solid-state nano-gears
down to 40 nm in outer diameter using electron beam nanolithography techniques,
where a hydrogen silsequiozane (HSQ) layer was selected as an ultrathin starting
material for this nanolithography leading to the fabrication of SiO2 solid-state
gears with a minimum thickness of 15 nm.
2 A single molecule gear
The HB-NBP molecule reported here has a gear-like structure and is chemical-
ly structured with six t-butyl outer legs uplifting a central aromatic core which is
composed of five phenyl rings and one pyrimidine ring all connected to a central
planar phenyl. The pyrimidine group works as an electronic tag for STM imaging
facilitating its identification and enabling to discriminate this gear tooth from the
other ones, since it appears as a bright protrusion during STM imaging (see figure
1).[3] Sub-monolayer amounts of this molecule were thermally sublimed on a
cleaned Au(111) single crystal substrate kept at room temperature. After this mo-
lecular sublimation, the sample was cooled down to cryogenic temperatures before
transferring it to the LT-UHV-STM chamber.
In general, the lateral manipulation of single atom and molecule using the STM
tip can be carried out in both constant-current and constant-height modes. In the
constant-current manipulation mode, the feedback loop enabling to keep the tun-
neling current constant is switched on during the whole manipulation process.
Once a manipulation path is defined, the tip is moved horizontally along the input
pathway set while concurrently adjusting vertically the tip-substrate height to keep
the tunneling current constant. In a successful STM atom/molecule manipulation,
the most important parameter is the tunneling resistance, which can be adjusted by
changing the tip-molecule distance or by directly increasing or decreasing the tun-
neling current or the bias voltage.[4] The tunneling resistance is reduced to bring
the tip very close to the atom or molecule to be manipulated, so close that the tun-
neling junction is forming a potential energy trap strong enough for an atom or
molecule to stay moving along the tip tracking its lateral displacement. Therefore
this mode is ideal for manipulating over a long distance. On the other hand, the
3
constant-height mode is performed with feedback off, during the manipulation
process the tip is moved parallel to the substrate surface keeping the tip height
constant. Initial experiments show that when the gear molecule is manipulated its
movement involves translation and random rotation because it moves from ad-
sorption site to adsorption site on Au(111) looking for stable positions for its six
tertbutyl legs at the same time in registry with the underlying Au surface atoms
and adapting its conformation accordingly on the surface at every step through its
travel along the manipulation path. Because of the molecule conformation changes
resulting in its rotation, during manipulation using the constant height mode the
molecule might move off the given tip trajectory, then it would stop moving and
following the tip during manipulation. Therefore this mode is suitable for a short-
range manipulation accordingly it was used in the gear experiments described here
since our purpose is solely to rotate the gear molecule through very short manipu-
lation steps.
Before constructing the molecule gear, it was investigated how a free standing
HB-NBP molecule behaves during STM manipulations. When the molecule was
manipulated to move the shortest distance possible which is equal to the distance
between nearest neighboring surface atoms in an Au(111) plane i.e. 0.288 nm, the
molecule rotates randomly by approximately 30 degree, as presented in Figure 1.
Here a series of STM images show the orientation and position changes of a gear
molecule after each manipulation in between two herringbone ridges which are
clearly seen in the STM image background.
Fig. 1: Orientation and position changes of a HB -NBP molecule manipulated by
the shortest possible distance. Leftmost STM image shows the initial confor-
mation of molecule before giving movements i.e. angle and displacement are
equal to zero, and following images in order from left to right were taken after
each manipulation step; I = 5 pA, V = 100 mV. After each manipulation, the mol-
ecule was displaced around 0.3 nm and turned in multiple of 30°.
4
While moving along that path the molecule turns randomly in clock and anti-
clockwise directions while simultaneously being displaced by about 0.3 nm. It
intuitively indicates that the molecule needs a gear axis to be able to rotate it with-
out displacing it laterally.
To explore whether is possible to control the rotation of a HB-NBP molecule,
three types of atomic-scale pinning centers have been tested as well as an atom
quantum corral as presented in Figure 2. The explored pinning centers are: one
gold adatom, a pure herringbone elbow, an impurity natively bound to a herring-
bone elbow, and a circular potential moat created by superimposed surface states
standing waves scattering at surrounding gold atoms.
Fig. 2: Three candidates of atomic scale pinning center to be mounted molecule
gear; gold adatom (indicated by yellow colored arrows), clean herringbone elbow
(green arrows), and atomic sized impurity natively bound to the elbow (orange
arrows). Gold adatoms were created by gentle crashing STM tip into gold surface
outside the top-left corner of scanning area. Inset shows corral structure built by
gold adatoms using STM atom manipulation technique
Neither the bare elbow nor the Au adatom are suitable for a centered rotation.
Because the interaction between the aromatic core of HB-NBP molecule and the
herringbone elbow is not strong enough, small lateral displacements of the mole-
cule during its rotation happen easily. When the molecule is moved on the Au
adatom it is never able to be concentric to the atom, it always stabilizes with the
5
atom in the middle of two tertbutyl legs.[5] As an inevitable consequence, the
molecules are rotated pinned on an off-centered axis. The potential corral is also
able to confine the molecule but because it does not have a solid pinning center
like an adatom the molecule rotates while swaying inside the corral. Besides, in
this confining structure will complicate the construction of a gear train, even if it
is made by just two gears, because the surrounding adatoms will perturb the inter-
actions between the molecules. Finally, the HB-NBP molecule gear was con-
structed on an atomic-scale impurity bound to one herringbone elbow as shown in
Figure 3. Here the constant-current manipulation mode was applied to mount the
molecule on top of the impurity.
Fig. 3: STM images showing a single HB-NBP molecule (a) before and (b) after
manipulation. An atomic sized impurity bind to an elbow shown in (a) appears as
a small protrusion from center of molecule after being mounted on top of it shown
in (b). I = 10 pA, V = 100 mV.
The concentrically-mounted molecule gear on its atomic pinning center as pre-
sented in Figure 3b is able to rotate by pushing one of the molecule legs with the
STM tip, generally the one with the pyrimidine group. The reproducible step-by-
step rotations of the molecule gear were carried out in both clock and anticloc k-
wise directions by gently pushing the molecule's leg using the constant-height
manipulation mode. In Figure 4, a sequence of STM images showing a full rota-
tion of the molecule gear is presented. These images were taken after each manip-
ulation step to capture the new in-plane configuration of the molecule with respect
to the surface substrate, the pyrimidine tag was used as a reference to follow the
molecule step by step rotation and to calculate the molecule's rotation angle. The
rotation angle between each neighboring stable conformation is in average 30-40
degrees, except for two 63 degree jumps (0º - 63 º and 256 º - 318 º). These two
larger angular separations in between stable rotational positions are due to the
influence of the two ridges forming the underlying herringbone extending from
the elbow. The ridge's potential energy barrier rebounds the HB-NBP molecule by
interacting with its t-butyl-end groups.
6
Figure 4: Full step-by-step molecule gear rotation. The image at 0° shows the
initial configuration imaged before start manipulations. The following images
taken after each manipulation in clockwise direction show the molecule gear stabi-
lized at different molecule angles. Repulsive barriers between the gear legs and
the ridges forming the herringbone elbow are located between 0°-63° and 256°-
318° respectively. There is also a rotation barrier between 123° and 166°.
An analysis of the rotation sequence shows that there are nine stable positions
for the molecule gear pinned on that herringbone elbow. These stable confor-
mations were identified using ASED+ molecular mechanics calculations per-
formed by taking into account the atomic scale surface structure under and around
the molecule gear and by calculating the STM image using the ESQC technique.
The detail atomic scale model of the surface including the elbow defined an effec-
tive Au(111) surface where the molecule gear conformation was optimized, the
constant current STM image calculated and compared to the experimental one.[2]
3 Solid state gears Nanofabrication and Manipulation
Although the diameter of a molecule gear can be increased by synthesizing a
molecule with a larger diameter and by adding more tertbutyl legs at the
periphery, ultimately there is a need for a nanosize scale solid state gear to
mechanically link a molecule gear to the existing micron size gears [6].
Furthermore, larger in diameter molecule gears will be too flexible to ensure a
good transmission of motion. Part of the energy required for rotating larger and
larger molecule will be distributed among the multiple degrees of freedom of such
7
large molecule and just a little portion of it will reach the collective rotation
degrees of freedom. It is therefore necessary to pass the lead to the solid state for
the intrinsic cohesion of solid state object be the condition for a robust rotation of
all the atoms composing a nanogear.
In our previous work, 6 teeth solid state nanogears made of hydrogen
silsequiozane (HSQ) were fabricated down to sizes of about 60 nm outer diameter
and 30 nm in thickness. A sacrificial layer of gold on top of the native oxide of the
silicon substrate was used to release the gears. This processed works fine, but for
molecule gears to be added on the same substrate, an inexpensive atomically
smooth substrate has to be used which can be easily found and processed. Highly
Oriented Pyrolitic Graphite (HOPG) fulfills the requirements and the following
results will be based solely on this substrate. Furthermore a 30 nm thickness is not
very compatible with the 1 nm van der Waals height of a molecule gear as
presented above together with the fact that 60 nm is still a gigantic diameter as
compared to the 1.2 nm of the molecule gear Figure 4.
Using an HOPG substrate, the fabrication process is greatly simplified: a min-
imum 15 nm HSQ layer (XR1541 from Dow Corning) was spun on a freshly
cleaved HOPG substrate, and then baked at 90°C for 5 mn. E-beam lithography
defines the nanogears, which are revealed after removing the surplus HSQ (the
detailed process will be reported somewhere else). The resulting sample can be
directly used for Scanning Electron Microscopy (SEM) and Atomic Force Micros-
copy (AFM) characterization and manipulations. Figure 5 shows the smallest solid
state gear fabricated to date, about 30 nm in diameter. The teeth are clearly de-
fined, but are at the limit of the fabrication process.
Fig. 5: The SEM image of the smallest solid state gear ever nanofabricated on a
graphite surface with a diameter below 30nm in diameter for a thickness of 15 nm.
Figure 6 shows an example of manipulation of such nanogear on the graphite
substrate. This was obtained in an AFM tapping mode by mounting the graphite
substrate on a movable piezo table. The AFM tip apex was fixed in x and y during
the manipulation sequence and the piezo table moved step by step toward the tar-
8
geted position on the graphite surface. A train of gears of different sizes can be
assembled this way and AFM imaged afterward (see Figure 6a). An SEM image
can be also recorded at the same location to appreciate the exact entanglement of
the teeth between the nanogears of the train since this cannot be determined using
the AFM images even in a taping mode (See Figure 6b). The step edges on the
graphite substrate can be clearly resolved by the AFM in taping mode of opera-
tion.
Fig. 6: a) AFM imaging of manipulated nanogears in an AFM tapping mode form-
ing a train of gears in interaction with a step edge. b) SEM image of the same
graphite surface area as in a).
Solid state gears of bigger size have also been achieved with the same process
up to a diameter of 10 μm. This is largely enough to offer a possibility to link with
the existing micron size gears [6]. However, manipulation of such gears was only
realized for diameter up to 1μm, as demonstrated in Figure 7. Two AFM images
9
of the same location show a 1μm diameter solid state HSQ gear before (a) and
after (b) the AFM tip located on the red cross was moved along the drawn path
resulting in a slight rotation and translation of the gear. This interplay between
rotation and translation has been discussed earlier [7] and depends on the friction
exerted on the gear by the substrate.
Fig.7: AFM images of a 1μm HSQ solid state gear nanofabricated on a graphite
surface. a) before manipulation by the AFM tip following the path depicted by the
red arrow. B) after manipulation, the nanogear clearly rotated and translated t o-
wards the interaction with the AFM tip.
The new process using graphite as a substrate is very promising as it will ena-
ble to integrate the solid state gears with the molecule gears, but two challenges
remain in order to fulfill the gap between molecular scale and nanoscale gears.
The first challenge is the coupling between a molecule gear and a solid state
gear. The teeth defined by e-beam lithography are still too rough, uneven and big
to directly couple with a molecule. Van der Walls coupling could be achieved, but
might be too small to pass a torque along. A mechanical coupling would still be a
better choice, and decorating a solid state disk with the same chemical group
which terminates the molecule legs/teeth would be a valid option. Another option
would be to use the emerging 2 dimensional materials available (graphene, h-BN,
MoS2 etc…) to nanofabricate an intermediary nanogear, but that would require
cutting the outlay of the gear with atomic precision, which is not yet achieved.
The second challenge that linger concerns the axle. It is of prime importance to
tackle this issue for practical nanomechanical mechanisms to be created. Using the
tips of a multiprobe system [8] in place of axle is very useful to study and under-
10
stand the mechanics in such nanogears. But is not a long term and scalable option.
Focused Ion Beam-assisted chemical vapor deposition offers an alternative which
is scalable. A small nanometer dot deposited using this technique could potentially
be used as an axle. The deposition could be done at different stages, with each its
own challenge: the dot can be patterned prior to the deposition of the HSQ layer
so the e-beam lithography step to define the gear will have to be done in a very
precise manner, but the gear will be pinned from the start. If high resolution in the
post patterned gear cannot be achieved, the FIB assisted dot could be made after
the gear making process. Thus the gear will have to be either mounted on the axle
using AFM or multiprobe manipulation or the gear will be put in the desired pos i-
tion before the dot is deposited. This method would be advantageous as it would
be more flexible in making complex machinery.
4 Conclusion
The precise manipulation of a molecule gear by a Scanning Tunneling Micro-
scope tip has been achieved on an atomically flat gold surface. This itself is an
achievement, but in order to harness the possibilities opened by these tiny gears,
we also presented the smallest solid state gears we could achieve by nanopattern-
ing a 15 nm thick hydrogen silsequiozane layer by e-beam lithography to fabri-
cate gears from micrometer size down to 30nm outer diameter. These gears can be
manipulated by an Atomic Force Microscope tip to construct various assemblies
which could potentially be used to be interconnected with planar single molecule
mechanical machines. However, we are still facing a few important challenges
like the lack of solution to easily create an axle and the problem of coupling mole-
cule and solid state gears.
Acknowledgments
This work was supported by A*STAR (the Agency for Science, Technology
and Research) funding under project no 1021100072.
References:
[1] Yun Y. J., Ah C. S., Kim S., Yun W. S., Park B. C., Ha D. H.: Manipulation of freestanding
Au nanogears using an atomic force microscope. Nanotechnology 18: 505304. (2007)
[2] Manzano C., Soe W.-H., Wong H. S., Ample F., Chandrasekhar N., Joachim C.: Step-by-step
rotation of a molecule-gear mounted on an atomic-scale axis. Nature Materials. 8: 576.
(2009)
[3] De Sarkar A., Manzano C., Soe W.-H., Chandrasekhar N., Joachim C.: Conformation de-
pendence of tag induced intramolecular STM contrast in hexaphenylbenzen molecules.Surf.
Sci. 603: L57. (2009)
[4] Hla S. W., Braun K. F., Rieder K. H.: Single-atom manipulation mechanisms during a quan-
tum corral construction. Phys. Rev. B67: 201402(R). (2003)
[5] Gross L., Rieder K. H., Moresco F., Stojkovic S. M., Gourdon A., Joachim C.: Trapping and
moving metal atoms with a six-leg molecule. Nature Materials 4: 892. (2005)
11
[6] Rogers M.S., Sniegowski S.S., Miller S., LaVigne G.F.: Designing and Operating Electrostat-
ically Driven Microengines. Proceedings of the 44th International Instrumentation Symposi-
um, Reno, NV, May 3-7: 56-65. (1998)
[7] Deng J., Troadec C., Ample F., Joachim C.: Fabrication and manipulation of solid-state SiO2
nano-gears on a gold surface, Nanotechnology 22: 275307. (2011)
[8] Joachim C., Martrou D., Rezeq M., Troadec C., Deng J., Chandrasekhar N. and Gauthier S.
J. Phys.: Condens. Matter 22: 084025. (2010)