Search

The steady-state, space-charge-limited piezoresistance (PZR) of defect-engineered, silicon-on-insulator device layers containing silicon divacancy defects changes sign as a function of applied bias. Above a punch-through voltage ($V_t$) corresponding to the onset of a space-charge-limited hole current, the longitudinal $\langle 110 \rangle$ PZR $\pi$-coefficient is $\pi \approx 65 \times 10^{-11}$~Pa$^{-1}$, similar to the value obtained in charge-neutral, p-type silicon. Below $V_t$, the mechanical stress dependence of the Shockley-Read-Hall (SRH) recombination parameters, specifically the divacancy trap energy $E_T$ which is estimated to vary by $\approx 30$~$\mu$V/MPa, yields $\pi \approx -25 \times 10^{-11}$~Pa$^{-1}$. The combination of space-charge-limited transport and defect engineering which significantly reduces SRH recombination lifetimes makes this work directly relevant to discussions of giant or anomalous PZR at small strains in nano-silicon whose characteristic dimension is larger than a few nanometers. In this limit the reduced electrostatic dimensionality lowers $V_t$ and amplifies space-charge-limited currents and efficient SRH recombination occurs via surface defects. The results reinforce the growing evidence that in steady state, electro-mechanically active defects can result in anomalous, but not giant, PZR., Comment: 9 pages, 8 figures

Spins in the ``semiconductor vacuum'' of silicon-28 ($^{28}\mathrm{Si}$) are suitable qubit candidates due to their long coherence times. An isotopically purified substrate or epilayer of $^{28}\mathrm{Si}$ is required to limit the decoherence pathway caused by magnetic perturbations from surrounding $^{29}\mathrm{Si}$ nuclear spins ($I=1/2$), present in natural Si ($^{\text{nat}}\mathrm{Si}$) at an abundance of 4.67%. We isotopically enrich surface layers of $^{\text{nat}}\mathrm{Si}$ by sputtering using high fluence ${}^{28}{\text{Si}}^{\ensuremath{-}}$ implantation. Phosphorus (P) donors implanted into one such $^{28}\mathrm{Si}$ layer with $\ensuremath{\sim}3000$ ppm $^{29}\mathrm{Si}$, produced by implanting 30 keV ${}^{28}{\text{Si}}^{\ensuremath{-}}$ ions at a fluence of $4\ifmmode\times\else\texttimes\fi{}{10}^{18}\phantom{\rule{4pt}{0ex}}{\mathrm{cm}}^{\ensuremath{-}2}$, were measured with pulsed electron spin resonance, confirming successful donor activation upon annealing. The monoexponential decay of the Hahn echo signal indicates a depletion of $^{29}\mathrm{Si}$. A coherence time of ${T}_{2}=285\ifmmode\pm\else\textpm\fi{}14\phantom{\rule{4pt}{0ex}}\ensuremath{\mu}\mathrm{s}$ is extracted, which is longer than that obtained in ${}^{\text{nat}}\text{Si}$ for similar doping concentrations and can be increased by reducing the P concentration in the future. Guided by simulations, the isotopic enrichment was improved by employing one-for-one ion sputtering using 45 keV ${}^{28}{\text{Si}}^{\ensuremath{-}}$ implanted with a fluence of $2.63\ifmmode\times\else\texttimes\fi{}{10}^{18}\phantom{\rule{4pt}{0ex}}{\mathrm{cm}}^{\ensuremath{-}2}$ into ${}^{\text{nat}}\text{Si}$. This resulted in an isotopically enriched surface layer $\ensuremath{\sim}100$ nm thick, suitable for providing a sufficient volume of $^{28}\mathrm{Si}$ for donor qubits implanted into the near-surface region. We observe a depletion of $^{29}\mathrm{Si}$ to 250 ppm as measured by secondary ion mass spectrometry. The impurity content and the crystallization kinetics via solid phase epitaxy are discussed. The $^{28}\mathrm{Si}$ layer is confirmed to be a single crystal using transmission electron microscopy. This method of Si isotopic enrichment shows promise for incorporation into the fabrication process flow of Si spin-qubit devices.

The SiC/SiO2 interface is a central component of many SiC electronic devices. Defects intrinsic to this interface can have a profound effect on their operation and reliability. It is therefore crucial to both understand the nature of these defects and develop characterization methods to enable optimized SiC-based devices. Here we make use of confocal microscopy to address single SiC/SiO2-related defects and show the technique to be a noncontact, nondestructive, spatially resolved and rapid means of assessing thequality of the SiC/SiO2 interface. This is achieved by a systematic investigation of the defect density of the SiC/SiO2 interface by varying the parameters of a nitric oxide passivation anneal after oxidation. Standard capacitance-based characterization techniques are used to benchmark optical emission rates and densities of the optically active SiC/SiO2-related defects. Further insight into the nature of these defects is provided by low-temperature optical measurements on single defects.

The bismuth (Bi) substitutional donor in silicon (Si) is an attractive qubit candidate for quantum computing proposals due to its large Hilbert space, clock transitions, and potential to couple to superconducting flux qubits. Single-qubit control, coupling, and readout by surface nanocircuitry requires a Bi depth of ∼20 nm in Si. This can be achieved using ion implantation of ∼25 keV Bi. This work explores the activation properties of Bi implanted at 26 keV with fluences of 1×1014 and 6×1012cm-2 into both crystalline and preamorphized Si. The Bi electrical activation yield was measured over a broad range of annealing conditions using resistivity and Hall effect measurements, enabling optimal annealing strategies to be proposed for the different implant parameters. For the high and low fluences, the maximum Bi activation yields achieved were 64% and 46%, respectively. Above a critical thermal budget, a substantial fraction of Bi forms electrically inactive complexes in the high fluence sample only. The substitutional fraction and diffusion of high fluence Bi was quantified, with diffusion coefficients D0=4.0±0.5 and 7.5±0.5cm2s-1 found for implantation into crystalline and preamorphized Si, respectively, using Rutherford backscattering spectrometry. To demonstrate the successful activation and quantum control of near-surface implanted Bi, the full hyperfine spectrum of these donors is obtained using continuous-wave electron spin resonance at 25 K, supporting the suitability for Bi donor qubits.

We analyze the electron spin relaxation rate $1/T_1$ of individual ion-implanted $^{31}$P donors, in a large set of metal-oxide-semiconductor (MOS) silicon nanoscale devices, with the aim of identifying spin relaxation mechanisms peculiar to the environment of the spins. The measurements are conducted at low temperatures ($T\approx 100$~mK), as a function of external magnetic field $B_0$ and donor electrochemical potential $��_{\rm D}$. We observe a magnetic field dependence of the form $1/T_1\propto B_0^5$ for $B_0\gtrsim 3\,$ T, corresponding to the phonon-induced relaxation typical of donors in the bulk. However, the relaxation rate varies by up to two orders of magnitude between different devices. We attribute these differences to variations in lattice strain at the location of the donor. For $B_0\lesssim 3\,$T, the relaxation rate changes to $1/T_1\propto B_0$ for two devices. This is consistent with relaxation induced by evanescent-wave Johnson noise created by the metal structures fabricated above the donors. At such low fields, where $T_1>1\,$s, we also observe and quantify the spurious increase of $1/T_1$ when the electrochemical potential of the spin excited state $|\uparrow\rangle$ comes in proximity to empty states in the charge reservoir, leading to spin-dependent tunneling that resets the spin to $|\downarrow\rangle$. These results give precious insights into the microscopic phenomena that affect spin relaxation in MOS nanoscale devices, and provide strategies for engineering spin qubits with improved spin lifetimes., 14 pages, 8 figures. Version 2 has extensive text revisions, typo corrections, and two extra authors

The two-dimensional (2D) hole gas at the surface of transfer-doped diamond shows quantum-mechanical interference effects in magnetoresistance in the form of weak localization and weak antilocalization (WAL) at temperatures below about 5 K. Here we use the quenching of the WAL by an additional magnetic field applied parallel to the 2D plane to extract the magnitude of the in-plane $g$-factor of the holes and fluctuations in the well width as a function of carrier density. Carrier densities are varied between 1.71 and $4.35\ifmmode\times\else\texttimes\fi{}{10}^{13}\phantom{\rule{0.16em}{0ex}}\mathrm{c}{\mathrm{m}}^{\ensuremath{-}2}$ by gating a Hall bar device with an ionic liquid. Over this range, calculated values of $|g|$ vary between 1.6 and 2.3 and the extracted well-width variation drops from 3 to 1.3 nm rms over the phase coherence length of 33 nm for a fixed geometrical surface roughness of about 1 nm as measured by atomic force microscopy. Possible mechanisms for the extracted variations in the presence of the ionic liquid are discussed.

Diamond, a wide band-gap semiconductor, can be engineered to exhibit superconductivity when doped heavily with boron. The phenomena has been demonstrated in samples grown by chemical vapor deposition where the boron concentration exceeds the critical concentration for the metal-to-insulator transition of nMIT4×1020/cm3. While the threshold carrier concentration for superconductivity is generally well established in the literature, it is unclear how well correlated higher critical temperatures are with increased boron concentration. Previous studies have generally compared several samples grown under different plasma conditions, or on substrates having different crystallographic orientations, in order to vary the incorporation of boron into the lattice. Here, we present a study of a single sample with unchanging boron concentration, and instead modify the charge-carrier concentration by introducing compensating defects via high-energy ion irradiation. Superconductivity is completely suppressed when the number of defects is sufficient to compensate the hole concentration to below threshold. Furthermore, we show it is possible to recover the superconductivity by annealing the sample in vacuum to remove the compensating defects.

The understanding of weak measurements and interaction-free measurements has greatly expanded the conceptual and experimental toolbox to explore the quantum world. Here we demonstrate single-shot variable-strength weak measurements of the electron and nuclear spin states of a P31 single-atom donor in silicon. We first show how the partial collapse of the nuclear spin due to measurement can be used to coherently rotate the spin to a desired pure state. We explicitly demonstrate that phase coherence is preserved with high fidelity throughout multiple sequential single-shot weak measurements and that the partial state collapse can be reversed. Second, we use the relation between measurement strength and perturbation of the nuclear state as a physical meter to extract the tunnel rates between the P31 donor and a nearby electron reservoir from data conditioned on observing no tunneling events. Our experiments open avenues to measurement-based state preparation, steering and feedback protocols for spin systems in the solid state, and highlight the fundamental connection between information gain and state modification in quantum mechanics.

We investigate the dynamics of a strongly-driven, microwave-dressed, donor-bound electron spin qubit in silicon. A resonant oscillating magnetic field $B_1$ is used to dress the electron spin and create a new quantum system with a level splitting proportional to $B_1$. The dressed two-level system can then be driven by modulating the detuning $\Delta\nu$ between the microwave source frequency $\nu_{\rm MW}$ and the electron spin transition frequency $\nu_e$ at the frequency of the level splitting. The resulting dressed qubit Rabi frequency $\Omega_{R\rho}$ is defined by the modulation amplitude, which can be made comparable to the level splitting using frequency modulation on the microwave source. This allows us to investigate the regime where the rotating wave approximation breaks down, without requiring microwave power levels that would be incompatible with a cryogenic environment. We observe clear deviations from normal Rabi oscillations and can numerically simulate the time evolution of the states in excellent agreement with the experimental data.

We describe experiments which demonstrate that carbon atoms introduced into a fused-silica substrate by means of MeV ion implantation can, after suitable annealing, form nanocrystalline diamond. Unlike other methods of creating diamond, the coalescence of the carbon into diamond nanocrystals occurs when the samples are heated in a conventional furnace and does not require the application of high external pressures, or any pre-existing diamond template. Following a dose of 5 × 1016 C/cm2 into fused quartz and after annealing in forming gas (4% hydrogen in argon), perfect cubic diamond crystallites of 5-7 nm diameter are formed. For higher doses, the same annealing treatments produce larger crystallites which are comprised of other varieties of solid carbon phases. We conclude that diamond is the stable form of carbon provided that the crystallite size is sufficiently small (less than 7 nm) and that the nanocrystallites are appropriately surface passivated.

The formation of R8 germanium is reported. The β-Sn phase is first induced by the indentation of amorphous germanium (a-Ge) and the resultant phases on pressure release are characterized by Raman scattering. The expected Raman line frequencies for the various phases of Ge are determined from first-principles calculations using density functional perturbation theory of the zone-center phonons in the diamond, ST12, BC8, and R8 Ge phases. In addition to the R8 phase, traces of BC8 may also be present following pressure release.

A micro-Raman scattering study on the linewidth and frequency shift of Si-III and Si-XII produced by indentation is presented over the temperature range 80--300 K. Measurements are compared to the Raman lines originating from the Si-I substrate. The main Si-XII Raman line shows a strong dependence on temperature and can be adequately described by anharmonic terms from the phonon proper self-energy. In contrast, the main Si-III Raman linewidth decreases with increasing temperature. A model related to electron-phonon interactions describes the data well.

The near-surface (0--200 nm) and end-of-range (EOR) (600--1000 nm) structural properties of ion-damaged Pb{sub 2}P{sub 2}O{sub 7} have been determined using ion-channeling and high-performance liquid chromatography. Single crystals of Pb{sub 2}P{sub 2}O{sub 7} were implanted with 1-MeV O{sup +} ions at doses ranging from 10{sup 11} to 10{sup 16} ions/cm{sup 2}. The initiation of amorphization in the near-surface region, rather than at the EOR, where most of the elastic nuclear collisions occur, reveals the importance of relatively weak electronic interactions (1500 eV/nm ion) in producing or enhancing structural alterations leading to the amorphization of ion-beam-irradiated solids. The electronic interactions produced by the 1-MeV O{sup +} ions in the present experiments are about one order of magnitude smaller in energy than that generally thought necessary to produce displacive structural damage in insulating inorganic solids.

The kinetics of intrinsic and dopant-enhanced solid phase epitaxy (SPE) is stud- ied in amorphous germanium (a-Ge) layers formed by ion implantation on Ge substrates. The SPE rates were measured with a time-resolved reflectivity (TRR) system between 300 and 540 degC and found to have an activation energy of (2.15 +/- 0.04) eV. To interpret the TRR measurements the refractive indices of the a-Ge layers were measured at the two wavelengths used, 1.152 and 1.532 ��m. For the first time, SPE rate measurements on thick a-Ge layers (>3 ��m) have also been performed to distinguish between bulk and near-surface SPE growth rate behavior. Possible effects of explosive crystallization on thick a-Ge layers are considered. When H is present in a-Ge it is found to have a considerably greater retarding affect on the SPE rate than for similar concentrations in a-Si layers. Hydrogen is found to reduce the pre-exponential SPE velocity factor but not the activation energy of SPE. However, the extent of H indiffusion into a-Ge surface layers during SPE is about one order of magnitude less that that observed for a-Si layers. This is thought to be due to the lack of a stable surface oxide on a-Ge. Dopant enhanced kinetics were measured in a-Ge layers containing uniform concentration profiles of implanted As or Al spanning the concentration regime 1-10 x1019 /cm-3. Dopant compensation effects are also observed in a-Ge layers containing equal concentrations of As and Al, where the SPE rate is similar to the intrinsic rate. Various SPE models are considered in light of these data., 42 pages, 10 figures, 1 table, accepted for publication in Physical Review B

The detailed structural changes that occur during ion-beam--induced amorphization of a solid have been experimentally determined for the first time using single-crystal surfaces of lead pyrophosphate that were implanted with varying doses of Pb/sup 3+/ ions. For all Pb-implant doses, the structure of the amorphous layer produced by ion bombardment was found to be significantly different than the structure of lead pyrophosphate glass prepared via thermal quenching.