Physical properties of low-dimensional systems

Quantum Transport

Introduction

The invention of the STM strongly contributed to the possibility to tailor and analyze the electronic properties of single nanostructures of adatoms on semiconductor or metal surfaces. A famous example is provided by the quantum corral experiments, in which atomic manipulations were used to assemble closed atomic structures, in which the surface state electrons are confined to a standing wave pattern [1,2].
In this project we study electrical properties and electrical transport phenomena of Platinum nanowires on a Ge(001) substrate by using scanning tunneling microscopy and spectroscopy.

Experimental details

For the experiments we use a UHV system with a background pressure < 5x10-11 mbar, equipped with an OMICRON STM-1 and a reflection high-energy electron diffraction (RHEED) setup (see figure 1).


Figure 1.

Ge(001) surfaces are cut from nominally flat 3 inch by 0.5 mm, single-side-polished n-type wafers. Samples are mounted on Mo holders and contact of the samples to any other metal during preparation and experiment has been carefully avoided. The Ge(001) samples are cleaned by 800 eV Ar+ ion sputtering and annealing at 1100(± 25) K. After several cleaning cycles the Ge(001) samples are atomically clean and exhibit a well ordered (2x1)/c(4x2) domain pattern (see figure 2) [3,4].


Figure 2.

Subsequently an equivalent of 0.25 monolayer of platinum is deposited onto the surface at room temperature. Platinum is evaporated by resistively heating a W wire wrapped with high purity Pt (99.995%). Afer Pt-deposition the sample is annealed at 1050(±25) K for 10 minutes and then cooled down to room temperature by radiation quenching before placing it into the STM for observation.

Research

We observed the formation of one-dimensional, defect- and kink-free Pt nanowires with a cross-section of only one atom and lengths up to hundreds of nanometers [5]. Figure 3 shows a 150x150 nm2 STM image of these wires. Figure 4 shows 3D 10x10 nm2 inzoom on the Pt nanowires. The thickness of the wires is 0.4 nm and the interwire distance is 1.6 nm, i.e. 4 times the lattice constant of Ge(001). The well-ordered structure of the Pt nanowires make them promising candidates for growing nanostructures. They offer a perfect case system for quantum transport studies in one dimension. Both this research pathways are followed in this project. We study quantum transport by STS measurements at room temperature as well as at lower temperatures. On the other hand we try to decorate the Pt nanowires with other molecules, like for instance CO. This pathway may provide some tools for molecular electronics.

Quantum Confinement between the Pt nanowires

The Pt nanowires act as barriers for the surface electrons. Therefore the surface electrons can be confined between the Pt nanowires [6]. This confinement can be measured by use of lock-in techniques, which enables us to measure simulteanously the topography and the derivative of the tunnel current (dI/dV). Fig. 5a shows the topography of the investigated sample. Defects in the nanowires or the underlying terrace are indicated by ellipses. A spatial map of the derivative is shown in fig. 5b. The intensity fades away in the presence of defects. In the lower image the topography (orange) and the dI/dV (purple) map are combined in a 3-dimensional representation. It's crystal clear that the confinement of the electronic state disappears near the indicated defects.


Figure 5.

Spatial distribution of the confined states

The spatial distribution of the confined states can also be measured with our STM [7]. The distribution follows nicely the contours of a sinusoid, with a maximum in the middle of the trough for the ground state in the 2.4 nm wide well (see figure 6, the dotted line is fit). In the first excited state a minimum occurs in the middle of the trough, as expected from the simple particle-in-a-box model (see figure 7, the black dotted line is corrected for contributions from the ground state, the red dotted line is a fit).

Peierls instability of atomic chains

An interesting property of a metallic atomic chain is that it can exhibit a Peierls instability, which turns the chain from a conductor into an insulator, accompanied by a periodicity doubling [8,9]. A Peierls distortion can be understood by considering a monatomic chain with nearest neighbour distance a (see fig. 8a). With one valence electron per atom the band will be filled exactly to the Fermi level (red part of the solid line) and the chain will, hence, be metallic. If we now allow periodicity doubling by a small displacement of every second atom (fig. 8b), the edges of the new Brillouin zone will coincide with the Fermi wave vector. The formation of a gap will now gain electronic energy (counteracted by an increase of the elastic energy) and the chain will be insulating. This transition takes place at 0 K for ideal one-dimensional conductors. For quasi one-dimensional structures (such as our Pt chains, which are coupled to a substrate), the Peierls transition temperature lies above 0 K. STM measurements of the Pt chains at 4.7 K clearly show the presence of a Peierls transition in the Pt chains on Ge(001) [10]. In figure 9, the periodicity doubling from a 2 x periodicity to a 4 x periodicity is observable in chains a-c. Remarkably, isolated chains and chains at the edge of a patch of chains, have apparently a more pure one-dimensional character, and hence a lower Peierls transition temperature, since they maintain the original periodicity. The observation of the conductor-to-insulator transition is complicated by the metallic nature of the underlying terrace. Figure 10 displays dI/dV curves measured on top of the Pt chains at 4.7 K. Although the chains with a 2 x periodicity certainly have a more distinct metallic character, it is too premature to conclude from this figure that the chains with a 4 x periodicity are insulating. In an attempt to separate the contribution from the chains themselves to the tunnel current from the contribution from the underlying terrace, we measured the inverse decay length, kappa. This quantity can provide information about the spatial origin of the tunnelling electrons [11]. The larger kappa, the larger the region from which the tunnelling electrons are gathered by the STM tip. In figure 11 the kappa measurements on top of the Pt chains at 4.7 K are displayed. It is clear from this graph that the tunnelling electrons originated from a much wider area in the case of chains with a 4 x periodicity compared to the chains with a 2 x periodicity. This is consistent with the picture in which the latter are undistorted, while the chains with a 4 x periodicity have already undergone a Peierls distortion.


Figure 8.


Figure 9.

References

[1] M.F. Crommie, C.P. Lutz, D.M. Eigler, Science 262 (1993), 218.
[2] E.J. Heller, M.F. Crommie, C.P. Lutz, D.M. Eigler, Nature 369 (1994), 464.
[3] H.J.W. Zandvliet, Phys. Rep. 388 (2003), 1.
[4] O. Gurlu, H.J.W. Zandvliet, B. Poelsema, Phys. Rev. Lett. 93 (2004), 066101.
[5] O. Gurlu, A.O.A. Adam, H.J.W. Zandvliet, B. Poelsema, App. Phys. Lett. 83 (2003), 4610.
[6] N. Oncel, A. van Houselt, J. Huijben, A-S. Hallbäck, O. Gurlu, H.J.W. Zandvliet and B. Poelsema, Phys. Rev. Lett. 95 (2005), 116801.
[7] A. van Houselt, N. Oncel, B. Poelsema and H.J.W. Zandvliet, Nano Lett. 6 (2006), 1439.
[8] R.E. Peierls, Quantum theory of solids, Oxford: At the Clarendon Press (1955)
[9] R.E. Peierls, More Surprises in Theoretical Physics, Princeton NJ: Princeton University Press (1991)
[10] A. van Houselt, T. Gnielka, J.M.J. Aan de Brugh, N. Oncel, D. Kockmann, R. Heid, K.-P. Bohnen, B. Poelsema and H.J.W. Zandvliet, Surf. Sci. 602 (2008), 1731.
[11] R.J. de Vries, A. Saedi, D. Kockmann, A. van Houselt, B. Poelsema and H.J.W. Zandvliet, Appl. Phys. Lett. 92 (2008), 174101.