CO adsorption on Pt-induced Ge nanowires

Authors: Danny E. P. Vanpoucke and Geert Brocks
Journal: Phys. Rev. B 81, 235434 (2010)
doi: 10.1103/PhysRevB.81.235434
IF(2010): 3.774
export: bibtex
pdf: <Phys.Rev.B> <arXiv>

Abstract

Using density-functional theory, we investigate the possible adsorption sites of CO molecules on the recently discovered Pt-induced Ge nanowires (NWs) on Ge(001). Calculated scanning tunneling microscope (STM) images are compared to experimental STM images to identify the experimentally observed adsorption sites. The CO molecules are found to adsorb preferably onto the Pt atoms between the Ge nanowire dimer segments. This adsorption site places the CO molecule in between two nanowire dimers, pushing them outward along the NW direction, blocking the nearest equivalent adsorption sites. This explains the observed long-range repulsive interaction between CO molecules on these Pt-induced nanowires.

Pt-induced nanowires on Ge(001): A density functional theory study

Authors: Danny E. P. Vanpoucke and Geert Brocks
Journal: Phys. Rev. B 81, 085410 (2010)
doi: 10.1103/PhysRevB.81.085410
IF(2010): 3.774
export: bibtex
pdf: <Phys.Rev.B> <arXiv>

Abstract

We study formation of the nanowires formed after deposition of Pt on a Ge(001) surface. The nanowires form spontaneously after high-temperature annealing. They are thermodynamically stable, only one atom wide and up to a few hundred atoms long. Ab initio density functional theory calculations are performed to identify possible structures of the Pt-Ge(001) surface with nanowires on top. A large number of structures are studied. With nanowires that are formed out of Pt or Ge dimers or mixed Pt-Ge dimers. By comparing simulated scanning tunneling microscopy images (STM) with experimental ones we model the formation of the nanowires and identify the geometries of the different phases in the formation process. We find that the formation of nanowires on a Pt-Ge(001) surface is a complex process based on increasing the Pt density in the top layers of the Ge(001) surface. Most remarkably we find the nanowires to consist of germanium dimers placed in troughs lined by mixed Pt-Ge dimer rows.

Density functional theory study of Pt-induced Ge(001) reconstructions

Authors: Danny E. P. Vanpoucke and Geert Brocks
Journal: Phys. Rev. B 81, 035333 (2010)
doi: 10.1103/PhysRevB.81.035333
IF(2010): 3.774
export: bibtex
pdf: <Phys.Rev.B> <arXiv>

Abstract

Pt deposited on a Ge(001) surface spontaneously forms nanowire arrays. These nanowires are thermodynamically stable and can be hundreds of atoms long. The nanowires only occur on a reconstructed Pt-Ge-surface where they fill the troughs between the dimer rows on the surface. This unique connection between the nanowires and the underlying substrate make a thorough understanding of the latter necessary for understanding the growth of the nanowires. In this paper we study possible surface reconstructions containing 0.25 and 0.5 of a monolayer of Pt. Comparison of calculated scanning tunneling microscope (STM) images to experimental STM images of the surface reconstruction reveal that the Pt atoms are located in the top layer, creating a structure with rows of alternating Pt-Ge and Ge-Ge dimers in a c(4×2) arrangement. Our results also show that Pt atoms in the second or third layer cannot be responsible for the experimentally observed STM images.

Ab Initio study of Pt Induced Nanowires on Ge(001)

Authors: Danny E.P. Vanpoucke
Ph.D. Thesis at University of Twente, The Netherlands
date: September 11th, 2009
Promoters Prof. Dr. Paul J. Kelly and Dr. Geert H. L. A. Brocks
doi: 10.3990/1.9789036528733
ISBN: 978-90-365-2873-3
#pages 193
export: bibtex
pdf: <PhD.Thesis> <UTwente>
research page with more information

Abstract

The aim of this thesis: “Ab Initio Study of Pt Induced Nanowires on Ge(001)”, is to model the experimentally observed ‘Pt nanowires’ on Ge(001). These one-atom-thick wires can be hundreds of nanometers long while remaining defect and kink free, providing the ultimate wire any chip designer dreams of. However, experiments show the wires not to be conducting; on the contrary, one-dimensional states are discovered between the wires. To model these nanowires, we combine state of the art density functional calculations with calculated scanning tunneling microscope (STM) images. First, the β-terrace substrate is modeled, showing a checkerboard pattern of Pt-Ge and Ge-Ge surface dimers in a Ge(001)-reconstructed surface.

Starting from this substrate model, different models with increasing Pt density are developed in an iterative fashion showing increasing agreement with the experimentally observed nanowires. We show that, contrary to previous assumptions, the observed wires are not Pt atoms but Ge atoms, explaining the lacking conductivity. The germanium nanowires consist of Ge dimers located in a Pt-lined trough. In addition, the 4×1 periodicity observed in the nanowire-arrays is traced back to the bonds of the Ge nanowire dimers to an extra Pt atom at the bottom of the trough, resulting in the buckling of the nanowires dimers.

In the last part of the thesis we investigate the adsorption of CO on the Ge nanowires under study. The observed adsorption of CO seems to contradict our proposed model due to the high sticking probability of CO on Pt, where it is low on Ge. We show that no contradiction exists. The CO molecules bind to the Pt atoms in the surface, but because they are tilted toward the nanowires, the resulting STM images give the impression that they are located on top of the nanowire giving rise to the apparent contradiction. In this last study, we also discover a very stable CO adsorption configuration in which the CO molecules remain invisible for STM, but could allow for the formation of one-dimensional molecular chains. This would open the door to one-dimensional molecular electronics.

Front cover of the PhD thesis.

The formation of Self-Assembled Nanowire Arrays on Ge(001): a DFT study of Pt Induced Nanowire Arrays

Authors: Danny E. P. Vanpoucke and Geert Brocks
Book title: Symposium Z–Computational Nanoscience–How to Exploit Synergy between Predictive Simulations and Experiment
proceeding: Mater. Res. Soc. Symp. Proc. 1177, 1177-Z03-09 (2009)
doi: 10.1557/PROC-1177-Z03-09
export: bibtex
pdf: <MRS Proceeding> <arXiv>

Abstract

Nanowire (NW) arrays form spontaneously after high temperature annealing of a sub monolayer deposition of Pt on a Ge(001) surface. These NWs are a single atom wide, with a length limited only by the underlying beta-terrace to which they are uniquely connected. Using ab-initio density functional theory (DFT) calculations we study possible geometries of the NWs and substrate. Direct comparison to experiment is made via calculated scanning tunneling microscope (STM) images. Based on these images, geometries for the beta-terrace and the NWs are identified, and a formation path for the nanowires as function of increasing local Pt density is presented. We show the beta-terrace to be a dimer row surface reconstruction with a checkerboard pattern of Ge-Ge and Pt-Ge dimers. Most remarkably, comparison of calculated to experimental STM images shows the NWs to consist of germanium atoms embedded in the Pt-lined troughs of the underlying surface, contrary to what was assumed previously in experiments.

Formation of Pt-induced Ge atomic nanowires on Pt/Ge(001): A density functional theory study

Authors: Danny E. P. Vanpoucke and Geert Brocks
Journal: Phys. Rev. B 77, 241308 (2008)
doi: 10.1103/PhysRevB.77.241308
IF(2008): 3.322
export: bibtex
pdf: <Phys.Rev.B> <arXiv> <UTwentePublications>

Abstract

Pt deposited onto a Ge(001) surface gives rise to the spontaneous formation of atomic nanowires on a mixed Pt-Ge surface after high-temperature annealing. We study possible structures of the mixed surface and the nanowires by total energy density functional theory calculations. Experimental scanning-tunneling microscopy images are compared to the calculated local densities of states. On the basis of this comparison and the stability of the structures, we conclude that the formation of nanowires is driven by an increased concentration of Pt atoms in the Ge surface layers. Surprisingly, the atomic nanowires consist of Ge instead of Pt atoms.

f-orbitals

3D f-Orbitals

3D gnuplot-gif-animations of the f-orbitals S03(θ,φ), S23(θ,φ) and
S33(θ,φ). In the images presented, the blue part represents the positive phase, and the red part the negative phase. Note that in gnuplot, the spherical coordinate θ is defined as π/2 – θ. Other than that the definitions of φ and θ coincide with those used in Griffiths’ Introduction to Quantum Mechanics.

For those interested: animations in gnuplot are only available for gnuplot versions > 4.0 (which at the moment of making these animations, was still in beta version).

S03(θ,φ)


Animated f-orbitals (gnuplot) of the S03 function. Different phases indicated in red and blue.

S23(θ,φ)


Animated f-orbitals (gnuplot) of the S23 function. Different phases indicated in red and blue.

S33(θ,φ)


Animated f-orbitals (gnuplot) of the S33 function. Different phases indicated in red and blue.

d-orbitals

3D d-Orbitals

3D Maple-images of the d-orbitals S02(θ,φ), S12(θ,φ) en S22(θ,φ). Note that the spherical coordinates (θ and φ) used by Maple are reversed compared to the definitions used in Griffiths’ Introduction to Quantum Mechanics (the latter being the more standard definition in physics and mathematics courses).

S02(θ,φ)

>  plot3d(abs(3*cos(phi)*cos(phi)-1),theta=0..Pi,phi=0..2*Pi,
grid=[60,60],coords=spherical,axes=frame,labels=[x,y,z]);

Maple 3D representation of atomic d-orbitals defined by the S02 function.

S12(θ,φ)

>  plot3d(abs(sin(phi)*cos(phi)*cos(theta)),theta=0..2*Pi,phi=0..Pi,
grid=[60,60],coords=spherical,axes=frame,labels=[x,y,z]);

Maple 3D representation of atomic d-orbitals defined by the S12 function.

S22(θ,φ)

>  plot3d(abs(sin(phi)*sin(phi)*cos(2*theta)),theta=0..2*Pi,phi=0..Pi,
grid=[60,60],coords=spherical,axes=frame,labels=[x,y,z]);

Maple 3D representation of atomic d-orbitals defined by the S22 function.

Effect of exchanging θ and φ

Maple assumes the first angle given is the angle in the xy-plane; the second angle is with regard to the z-axis. This makes that you have to be very careful when giving Maple the θ and φ angles, and make sure that their definitions are the same. If the definitions
are reversed: I.e. if we use the variable θ as the variable φ and vice versa, the resulting plots become something quite different. This goes for all available plotting programs (Maple, gnuplot…); make sure you certain that what you think you enter is also what the program thinks you have entered. If not you could end up with surprising results. The same images as above, but now with θ and φ exchanged:


Wrong Maple 3D representation of atomic d-orbitals defined by the S02 function. The theta and phi angles are exchanged.
Wrong Maple 3D representation of atomic d-orbitals defined by the S12 function. The theta and phi angles are exchanged.

Wrong Maple 3D representation of atomic d-orbitals defined by the S22 function. The theta and phi angles are exchanged.