The Single-Atom Transistor: perspectives for quantum electronics on the atomic-scale
Ch. Obermair 1 2
F.-Q. Xie 1
Th. Schimmel 0 1
0 Institute of Nanotechnology, Forschungszentrum Karlsruhe, Karlsruhe Institute of Technology (KIT) , 76021 Karlsruhe , Germany
1 Institute of Applied Physics and Center for Functional Nanostructures (CFN), University of Karlsruhe, Karlsruhe Institute of Technology (KIT) , 76131 Karlsruhe , Germany
2 DOI: 10.1051/epn/2010403
Controlling the electronic conductivity on the quantum level will impact the development of future nanoscale electronic circuits with ultralow power consumption. Here we report about the invention of the single-atom transistor, a device which allows one to open and close an electronic circuit by the controlled and reproducible repositioning of one single atom. It opens intriguing perspectives for the emerging fields of quantum electronics and logics on the atomic scale.
PERSPECTiVES FOR QuAN TuM ElE CTRONiCS ON THE ATOMiC-SCAlE
ascinating physical properties and technological In our new approach, a three-terminal, gate-controlled
perspectives have motivated investigation of ato- atomic quantum switch was fabricated by
electrochemimic-scale metallic point contacts in recent years cal deposition of silver between two nanoscale gold
]. e quantum nature of the electron is electrodes (see Fig. 1) [
]. A comparison of the
expedirectly observable in a size range where the width of the rimental data with theoretical calculations indicates
contacts is comparable to the Fermi wavelength of the perfect atomic order within the contact area without
electrons, and conductance is quantized in multiples of volume or surface defects [
2e2/h for ballistic transport through ideal junctions [
In metallic point contacts, which have been fabricated by Switching an atom
mechanically controlled deformation of thin metallic We control individual atoms in the quantum point
] and electrochemical fabrication techniques contact by a voltage applied to an independent gate
], the conductance depends on the chemical valence electrode, which allows a reproducible switching of the
]. Two-terminal conductance-switching devices based contact between a quantized conducting “on-state” and
on quantum point contacts were developed both with an an insulating “off-state” without any mechanical
moveSTM-like setup [
] and with electrochemical methods [
]. ment of an electrode (see Fig. 2).
Article available http://www.europhysicsnews.org or http://dx.doi.org/10.1051/epn/2010403
Schematic of the
atomic switch is
ent third gate
allowing to open and
close a metallic
the source and
relocation of one
single silver atom.
Schematic of the
A metal atom (see
arrow) is switched
graph) and an
To fabricate the initial atomic-scale contact we deposit As soon as the conductance exceeds a preset “target”
silver within a narrow gap between two macroscopic value, the deposition is stopped and the voltage is
revergold electrodes (gap width: typically 50 nm) by applying sed to dissolve the junction again. Aer the
an electrochemical potential of 10-40 mV to a gate elec- conductance drops below a preset value, the
]. e gold electrodes are covered with an tion/dissolution cycle is repeated automatically by the
insulating polymer coating except for the immediate computer-controlled setup. During the first cycles, the
contact area, and serve as electrochemical working elec- conductance at the end of the deposition step varies
trodes. ey correspond to the “source” and “drain” strongly from cycle to cycle [
]. Aer repeated cycling,
electrodes of the atomic-scale transistor. Two silver an abrupt change is observed from this irregular
variawires serve as counter and quasi-reference electrodes. tion to a bistable switching between zero and a finite,
e potentials of the working electrodes with respect quantized conductance value at an integer multiple
to the quasi-reference and counter electrodes are set by of G0 (= 2e2/h).
a computer-controlled bipotentiostat (see Fig.3). All
experiments are performed at room temperature, the Controlling the junction
electrolyte being kept in ambient air. For conductance at the single-atom level
measurements, an additional voltage in the millivolt Figure 4 shows a sequence of reproducible switching
range is applied between the two gold electrodes. To events between an insulation “off-state” and a
quantifabricate the atomic transistor, silver is deposited on zed conducting “on-state” (at 1 G0), where the
each of the two working electrodes, until finally two sil- quantum conductance (red curves) of the switch is
ver crystals meet, forming an atomic scale contact controlled by the gate potential (blue curves), as
which is bridging the gap. commonly observed in transistors. As calculations
While silver islands are deposited in the junction we have shown [
], for atomic-scale silver contacts a
monitor the conductance between the two electrodes. quantized conducting “on-state” of 1 G0 corresponds
to a single-atom contact.
When we set the gate potential to an intermediate
“hold” level between the “on” and the “off ” potentials, the
currently existing state of the atomic switch remains
stable, and no further switching takes place. is is
demonstrated in Fig. 5 both for the “on-state” of the
switch (le arrow) and for the “off-state” of the switch
(right arrow). us, the switch can be reproducibly
operated by the use of three values of the gate potential for
“switching on”, “switching off ” and “hold”. ese results
give clear evidence of a hysteresis when switching
between the two quantized states of the switch. It can be
explained by an energy barrier which has to be
overcome when performing the structural changes within
the contact when switching from the conducting to the
non-conducting state of the switch and vice versa.
e results indicate that switching occurs by a reversibly
rearrangement of the contacting group of atoms
between two different stable configurations with a
potential barrier between them. For silver the observed
quantum conductance levels appear to coincide with
integer multiples of the conductance quantum [
e observed integer conductance levels of the switch
are determined by the available bistable junction
conformations, similar to the observation of preferential
atomic configurations in metallic clusters
corresponding to “magic numbers” . By snapping into ‘magic’
bistable conformations, such energetically preferred
junctions configurations are mechanically and
thermally stable at room temperature, and they are
reproducibly retained even during long sequences of
Reproducible switching in the above cases was always
performed by opening and closing a quantum point
contact, i.e. by switching between a quantized
conducting state and a non-conducting state. However, it was
not clear if this kind of gate-electrode controlled
switching is also possible between two different conducting
states of one and the same contact. Such kind of
switching would involve two different stable contact
configurations on the atomic scale, between which
reversible switching would occur even without ever
breaking the contact.
Such multi-level logics and storage devices on the
atomic scale would be of great interest as they allow a more
efficient data storage and processing with a smaller
number of logical gates. By developing a modified
procedure of fabrication, a multi-level atomic quantum
transistor was obtained, allowing the gate-controlled
switching between different conducting states.
Instead of setting the lower threshold where the
dissolution process is stopped by the computer, to a value
near 0 Go, the lower threshold was set at a value above
the desired quantized conductance of the lower of the
two “on-state” levels [
Figure 6 demonstrates the operation of such a
twolevel transistor: A controlled change of the gate
potential UG leads to a controlled switching of the
conductance of the quantum point contact between two
different quantized conducting states, exhibiting
conductance levels of 1 G0 and 3 G0, respectively. Sharp
transitions are observed between the two levels. No
intermediate steps or staircase-like structures in
conductance are observed in the diagram. e
transitions are instantaneous within the time resolution of the
diagram of Fig. 6 (50 ms).
Conclusions and perspectives
e development of the single-atom transistor represents
a first demonstration of the functionality of a transistor
on the atomic scale. is is of great interest, as there were
many previous demonstrations of passive devices such as
atomic-scale and molecular resistors. However, there
was a lack of an actively switching device such as a
transistor on the atomic scale. e atomic transistor as an • ey are a first demonstration of an all-metal transistor
actively controllable device, which reproducibly operates without the use of any semiconductor, the lack of a
at room temperature, is filling this gap. band gap allowing operation at very low voltages.
Atomic transistors represent a new class of devices Such devices provide a number of advantages: ey
poswhich show remarkable properties: sess extremely nonlinear current-voltage characteristics,
• ey allow the switching of an electrical current by the desirable in many applications, and they can be
manugeometrical relocation of individual atoms rather than factured using conventional, abundant, inexpensive and
by locally changing electronic properties as done in non-toxic materials. At the same time, the devices open
conventional transistors. perspectives for electronic switching at ultrafast
frequen• ey represent quantum switches, the levels between cies: although the switching time in our current
which the switching occurs being given by funda- investigations is limited by the response time of the
elecmental laws of quantum mechanics. tronic measuring setup (3-5 microseconds), the intrinsic III
fe Taur eS
“off-state” and a
on-state” at 4 G0.
level can be kept
stable, if UG is
kept at a “hold”
level (see arrows).
III operation time is expected to be limited by the
atomicscale rearrangement within the junction (picoseconds),
opening perspectives for ultra-high frequency operation.
Because the switching process is achieved with very small
gate potentials in the millivolt range, the power
consumption of such devices is by orders of magnitude lower than
that of conventional semiconductor-based electronics.
Although the development of the single-atom transistor
just marks the beginning of actively controlled
electronics on the atomic scale, it opens fascinating perspectives
for quantum electronics and logics based on individual
atoms. e development of a first, simple integrated
] and a multilevel quantum transistor  are
first encouraging steps in this direction.
We thank Robert Maul, Wolfgang Wenzel and Gerd
Schön for stimulating discussions, for the excellent
Example of a
levels of 1 G0 and
3 G0, respectively.
of the atomic
curves) is directly
controlled by the
gate potential (UG)
cooperation and for corresponding theoretical work.
is work was supported by the DFG-Center for
Functional Nanostructures and by the Baden-Württemberg
Foundation within the Network of Excellence on
Functional Nanostructures, Baden-Württemberg.
About the authors
Prof. omas Schimmel is Professor of Physics and
Joint Institute Director at the Institute of Applied
Physics, University of Karlsruhe. He is at the same time
heading a research group at the Institute of
Nanotechnology at the Research Center Karlsruhe, Karlsruhe
Institute of Technology (KIT). Prof. Schimmel is
Scientific Director of the Network of Excellence “Functional
Nanostructures” and Editor in Chief of the Beilstein
Journal of Nanotechnology.
Dr. Christian Obermair and Dr. Fang-Qing Xie are
Senior Scientists in the same group, their research
focusing on quantum transport and atomic-scale
electronics. Dr. Obermair is Administrative Director of the
above research network. For further information see
Correspondence should be addressed to . Schimmel
 F.-Q. Xie , R. Maul , A. Augenstein , Ch. Obermair, E.B. Starikov , G. Schön, Th. Schimmel, W. Wenzel, Nano Lett. 8 ( 12 ), 4493 ( 2008 ).
 N. Agraїt , A. L. Yeyati , J. M. van Ruitenbeek , Phys. Rep . 377 , 81 ( 2003 ).
 E. Scheer , N. Agraıt , JC. Cuevas, A. Levy Yeyati , B. Ludoph , A. Martin-Rodero , G. Rubio Bollinger , J. M. van Ruitenbeek , and C. Urbina , Nature 394 , 154 ( 1998 ).
 A. I. Mares , J. M. van Ruitenbeek , Phys. Rev. B 72 , 205402 ( 2005 ).
 C. Z. Li , A. Bogozi , W. Huang , N. J. Tao , Nanotechnology 10 , 221 ( 1999 ).
 F.-Q. Xie , L. Nittler, Ch. Obermair, Th. Schimmel, Phys. Rev. Lett . 93 , 128303 ( 2004 ).
 F.-Q. Xie , Ch. Obermair and Th. Schimmel, Solid State Communications 132 , 437 ( 2004 ).
 D.P.E. Smith , Science 269 , 371 ( 1995 ).
 K. Terabe , T. Hasegawa , T. Nakayama , and M. Aono , Nature 433 , 47 ( 2005 ).
 F.-Q. Xie , R. Maul , S. Brendelberger , Ch. Obermair, E. Starikow , W. Wenzel, G. Schön, Th. Schimmel, Appl. Phys. Lett . 93 , 043103 ( 2008 ).
 F.-Q. Xie , Ch. Obermair and Th. Schimmel, in: R. Gross et al. (eds.), Nanoscale Devices - Fundamentals and Applications . Springer, 153 - 162 , 2006 .
 M. N. Huda and A. K. Ray , Phys. Rev. A 67 , 013201 ( 2003 ).
 F.-Q. Xie , R. Maul , Ch. Obermair, G. Schön, W. Wenzel, Th. Schimmel, Advanced Materials 22 , 2033 ( 2010 ).
 Th. Schimmel , F.-Q. Xie , Ch. Obermair, Patent pending, US 2009195300 .