A/Prof Michelle Simmons - University of New South Wales
Participating University of New South Wales Node Researchers:
Dr Neil Curson
Dr Lars Oberbeck
Mr Martin Brauhart
Mr Albert McMaster
Prof Robert Clark
Mr Johnson Goh (PhD)
Mr Toby Hallam (PhD)
Mr Frank Ruess (PhD)
Mr Steven Schofield (PhD)
Collaborating Centre Researchers:
A/Prof Steven Prawer - University of Melbourne
Dr Chris Pakes - University of Melbourne
Dr Alberto Cimmion - University of Melbourne
Prof David McKenzie - University of Sydney
Dr Nigel Marks - University of Sydney
Mr Hugh Wilson - University of Sydney
Dr Marilyn Hawley - Los Alamos National Laboratory, USA
Dr Geoff Brown - Los Alamos National Laboratory, USA
Dr Holger Grube - Los Alamos National Laboratory, USA
The objective of this Program is to fabricate the basic building blocks of a scaleable quantum computer using 31P nuclei (nuclear spin I=1/2), as the qubits, in isotopically pure 28Si (I=0). For qubit operation we require an atomically precise array of P nuclei embedded in silicon approximately 200 Angstroms apart, such that the donor electron wavefunctions can overlap, and an insulating barrier to isolate them from the surface control A and J-gates. Using a combination of scanning tunnelling microscopy (STM) and molecular beam epitaxy (MBE) we employ a "bottom-up" strategy to build the device atomic layer by atomic layer. The final device structure, with the qubit architecture is shown in figure 1.
Figure 1: Si based quantum computer .
Figure 2 shows the detailed "bottom-up" fabrication strategy to build the Kane quantum computer.
- Step 1 The clean, defect-free Si(001) surface is exposed to atomic hydrogen (H) which passivates the surface to form a monolayer (ML) of hydrogen resist.
- Step 2 The STM tip is then used to desorb single H atoms from the resist, thereby exposing the underlying Si substrate.
- Step 3 During subsequent exposure of the surface to phosphorus-containing molecules, such as phosphine (PH3) gas single PH3 molecules adsorb into the hole in the resist bonding directly to the silicon surface .
- Step 4 A critical annealing step incorporates the P atoms from the PH3 molecule into the Si surface while leaving the hydrogen resist layer unaffected. This step is important since it changes the weakly physisorbed phosphorus atom in the phosphine molecule with only one covalent bond to the surface into an incorporated phosphorus atom in the silicon surface with three covalent bonds. This helps to secure the phosphorus atom in its patterned location providing a stronger resistance to thermal diffusion in subsequent processing stages.
Steps 1-4 have already been achieved within the program. In particular, we have demonstrated the controlled incorporation of single P atoms into the Si(001) surface with atomic-scale precision, see figure 3 .
- Step 5 The hydrogen resist layer is removed from the surface without destroying the ordered array of phosphorus atoms.
- Step 6 The phosphorus atoms are then encapsulated in a few monolayers of silicon grown at room temperature to minimise dopant diffusion [4 ,5 ] out of the patterned arrays.
- Step 7 Subsequent rapid annealing reduces the defect density in the silicon layer and flattens the surface allowing us to image the dopants beneath the surface. A key factor in the silicon encapsulation process is to minimise heating of the sample as this may cause the phosphorus atoms to thermally diffuse or segregate to the surface, thereby destroying the carefully created STM patterned array.
- Steps 8-10 The last three steps include Si growth at higher temperatures, the growth of an insulating barrier layer and registration of the P qubits to aligned surface metal gates. Each of these processes are currently being tackled at present.
Figure 2: "Bottom-up" fabrication process
Figure 3: Controlled single P atom incorporation.
The fabrication of registered P qubit arrays has only become possible with the design of a combined customised STM-SEM/MBE system shown in figures 4 (a) and (b). This system incorporates several unique features including the ability to image sample sizes of 1 cm2, an SEM attached to the STM to allow precise tip positioning on the sample surface and observation of nanometer features on the surface, an optical position read-out to allow reproducible positioning of the STM tip several hundred nanometers from a registration marker on the surface (figure 4 (a)) and a full 4" SiGe capability for device quality silicon, sample uniformity and accurate growth rate calibration (figure 4 (b)).
Figure 4a: Customised combined STM-SEM system
Figure 4b: SiGe MBE system
After fabrication of the P qubit array it is then necessary to transfer the sample to the Semiconductor Nanofabrication Facility for subsequent A and J-gate fabrication before being measured in the National Magnet Laboratory. Important collaborative aspects of this project involve close liason with researchers at Los Alamos National Laboratory (USA) on atom lithography and buried dopant imaging. In addition, parallel collaborative programs on single ion implantation and scanning probe imaging are undertaken at the University of Melbourne node.
At present excellent progress is being made in steps 8-10 with the registration of buried P atom structures to electrical contacts on the surface and the hydrogen removal and Si encapsulation of P atoms with buried dopant imaging. This work is at present being patented and overcomes many of the remaining challenges for the realisation of a scaleable solid-state quantum computer design.
 B. E. Kane, Nature 393, 133 (1998).
 S. R. Schofield, N. J. Curson, M. Y. Simmons, F. J. Ruess, T. Hallam, L. Oberbeck, and R. G. Clark, cond-mat/0307599, Phys. Rev. Lett. (in press).
 L. Oberbeck, N. J. Curson, M. Y. Simmons, R. Brenner, A. R. Hamilton, S. R. Schofield, and R. G. Clark, Appl. Phys. Lett. 81, 3197 (2002).
 L. Oberbeck, N. J. Curson, T. Hallam, M. Y. Simmons, R. G. Clark, and G. Bilger,