The quest to build a quantum computer has been inspired by the recognition of the formidable computational power such a device could offer. In particular silicon-based proposals, using the nuclear or electron spin of dopants as qubits, are attractive due to the long spin relaxation times involved, their scalability, and the ease of integration with existing silicon technology. Fabrication of such devices however requires atomic scale manipulation -an immense technological challenge. We demonstrate that it is possible to fabricate an atomically-precise linear array of single phosphorus bearing molecules on a silicon surface with the required dimensions for the fabrication of a siliconbased quantum computer. We also discuss strategies for the encapsulation of these phosphorus atoms by subsequent silicon crystal growth. (To appear in Phys. Rev. B Rapid Comm.) 03.67. Lx, 68.37.Ef, A quantum bit (or qubit) is a two level quantum system that is the building block of a quantum computer. To date the most advanced realisations of a quantum computer are qubit ion trap 1 and nuclear magnetic resonance 2-4 systems. However scaling these systems to large numbers of qubits will be difficult 5 , making solidstate architectures 6 , with their promise of scalability, important. In 1998 Kane proposed a novel solid state quantum computer design 7 using phosphorus 31 P nuclei (nuclear spin I = 1/2) as the qubits in isotopically-pure silicon 28 Si (I = 0). The device architecture is shown in Fig. 1a, with phosphorus qubits embedded in silicon approximately 20 nm apart. This separation allows the donor electron wavefunctions to overlap, whilst an insulating barrier isolates them from the surface control A and J gates. These A and J gates control the hyperfine interaction between the nuclear and electron spins and the coupling between adjacent donor electrons respectively. For a detailed description of the computer operation refer to Kane 7 . An alternative strategy using the electron spins of the phosphorus donors as qubits has also been proposed 8 .One of the major challenges of this design is to reliably fabricate an atomically-precise array of phosphorus nuclei in silicon -a feat that has yet to be achieved in a semiconductor system. Whilst a scanning tunnelling microscope (STM) tip has been used for atomic scale arrangement of metal atoms on metal surfaces 9 , rearrangement of individual atoms in a semiconductor system is not straightforward due to the strong covalent bonds involved. As a result, we have employed a hydrogen resist strategy outlined in Fig. 1b. Here the array is fabricated using a resist technology, much like in conventional lithography, where the resist is a layer of hydrogen atoms that terminate the silicon surface. An STM tip is used to selectively desorb individual hydrogen atoms, exposing the underlying silicon surface in the required array. STM induced hydrogen desorption has been developed and refined over the past ten years 10 and has been proposed 11 for the assembly of atomically-ordered device structure...

We present a method for measuring single spins embedded in a solid by probing two-electron systems with a single-electron transistor ͑SET͒. Restrictions imposed by the Pauli principle on allowed two-electron states mean that the spin state of such systems has a profound impact on the orbital states ͑positions͒ of the electrons, a parameter which SET's are extremely well suited to measure. We focus on a particular system capable of being fabricated with current technology: a Te double donor in Si adjacent to a Si/SiO 2 interface and lying directly beneath the SET island electrode, and we outline a measurement strategy capable of resolving singleelectron and nuclear spins in this system. We discuss the limitations of the measurement imposed by spin scattering arising from fluctuations emanating from the SET and from lattice phonons. We conclude that measurement of single spins, a necessary requirement for several proposed quantum computer architectures, is feasible in Si using this strategy.

No abstract

No abstract

We report direct detection of light emission from single-crystal silicon cleaved in vacuum. The emission occurs in at least two wavelength regions, one above 1 eV energy (A) and the other between 0.25 and 0.36 eV (/?). Intensities for about 0.1-cm 2 cleaved areas are in the region of 10 12 photons in region A and 10 M photons in region B. The B radiation is of significantly longer duration than the A radiation. Some of the latter can be explained by radiative bulk band recombination, and the B radiation by radiative surface band recombination. The duration of the B radiation shows that the surface-state gap is indirect.PACS numbers: 78.60.Mq, 73.20.AtWe report direct detection of light emission from cleavage of single-crystal silicon. The light is of at least two wavelength groups in the infrared. A previous account 1 mentioned visible-light emission from silicon and other materials subjected to sandblasting but this appears to be unrelated to true cleavage luminescence which is extremely weak. This work appears to provide the first report and measurement of such.The motivation for the work was to obtain information about the structure of the cleaved surface of silicon. Neither low-energy electron diffraction 2 nor scanning tunneling microscopy 3 can distinguish between the Pandey ^-bonded chain reconstruction 4 or the three-bondscission 5,6 (TBS) model, both of which feature surfaceatom chains along [TlO] directions that are consistent with optical and ion scattering data. 5 The models differ significantly, however, in subsurface structure and in the manner in which they arise. 5 Upon cleavage, the surface-atom reconstructions that are known to occur from the ideal structure will involve changes in bond energies. Such changes may be radiative. The bulk band structure of silicon features an indirect band gap so that radiative transitions are much weaker and slower than from direct-gap materials. Nevertheless, they have been observed both in forwardbiased pn junctions 7 and in reverse-biased junctions. 8 In the latter, microplasmas and high-energy radiation are involved, but in the former, which appears to be the more relevant case, band-gap radiation, less exciton and phonon energies, has been detected.In the case of the cleaved surface, there may be radiation during the reconstruction process and subsequently when the surface band structure is established. A search for any such was undertaken. The radiation is expected to be in the range up to the bond-breaking energy which is about 2.1 eV, or half the heat of sublimation of crystalline silicon. No photon detector has its maximum sensitivity over the entire range so the first experiments were conducted with a liquid-nitrogen-cooled InSb photoconductive detector of peak detectivity about 10 ll cmHz 1/2 W ~l, with peak sensitivity at about 0.3 eV and low-energy cutoff through the sapphire window at about 0.21 eV. Subsequently, 1-cm 2 Si photovoltaic detectors of peak detectivity about 10 cm Hz ,/2 W _1 and lowenergy cutoff at about 1 eV were also used, and later ...