Search

Owing to the advanced fabrication technology of silicon, silicon waveguides are particularly attractive for implementing all-optical signal processing devices and switches. Therefore, in this paper, a silicon-on-silica waveguide that consists of four slots arranged in the shape of letter E is proposed to be employed as the building block for simulating fundamental all-optical logic gates (AOLGs), including XOR, AND, OR, NOT, NOR, NAND, and XNOR, at 1.55 μm telecommunications wavelength. The operation concept of these logic gates relies on the constructive and destructive interference that results from the phase difference induced by optical beams that are incident on the E-shaped waveguide. The performance of the target logic gates is assessed against the contrast ratio (CR) metric. Moreover, the dependence of the spectral transmission on the device's key operating parameters is investigated and assessed. Compared to other reported designs, the results obtained by conducting simulations using the finite-difference-time-domain in lumerical commercial software show that the proposed waveguide can operate at a higher speed of 80 Gb/s and attain higher CRs of 36, 39, 35.5, 28.8, 30, 38, and 36.7 dB for logic XOR, AND, OR, NOT, NOR, NAND, and XNOR, respectively. This suggests that by using the proposed scheme, AOLGs could be realized more feasibly with greater performance and faster operation toward satisfying the present and future needs of light wave circuits and systems. [ABSTRACT FROM AUTHOR]

Structure-engineered silicon exhibits a wealth of unique optical properties below its bandgap, which holds promise for mid-infrared and terahertz applications such as photodetection, thermophotovoltaics, radiative cooling, and spectroscopy. In this paper, we investigate enhancement mechanisms of sub-bandgap absorption of black silicon fabricated into periodic pyramids. Our measurements indicate that the pyramid structure leads to an enhanced broadband absorption in the wavelength region from 1.5 to 13.07 μm with an efficiency of over 80%. The broadband absorption enhancement is shown to originate from the Rayleigh–Wood anomaly, localized magnetic plasmonic resonance, and graded-index effect, which together facilitate the interaction between light and free-carriers in silicon. These results are helpful for understanding the interaction between light and black silicon. [ABSTRACT FROM AUTHOR]

Precise control of dopant placement is crucial for the reproducible, and reliable, nanoscale semiconductor device fabrication. In this paper, we demonstrate an atomic force microscopy (AFM) probe assisted localized doping of aluminum into an n-type silicon (100) wafer to generate nanoscale counter-doped junctions within two nanometers of the silicon-air interface. The local doping results in changes in electrostatic potential, which are reported as contact potential difference, with nanoscale spatial resolution. In contrast to the literature where nano-mechanical defects in, or contaminants on, silicon substrates can result in measurable changes in the chemical potential of the near-surface, additional thermal treatment was needed to electrically activate the aluminum dopants in our current work. Unfortunately, the thermal activation step also caused the dopants to diffuse and geometric distortions in the doped area, i.e., broadening and blurring of the electrically distinct areas. The results from optimization efforts show that the "active" dopant concentration depended primarily on the thermal anneal temperature; additional AFM-tip dwell time during the aluminum implantation step had no meaningful impact on the electrical activity of the doped sites. [ABSTRACT FROM AUTHOR]

AlN-on-Si can be considered as a model system for heteroepitaxial growth of highly mismatched materials. Indeed, AlN and Si drastically differ in terms of chemistry, crystalline structure, and lattice parameters. In this paper, we present a transmission electron microscopy and grazing incidence X-ray diffraction study of the microstructure of AlN layers epitaxially grown on Si (111) by molecular beam epitaxy. The large interfacial energy due to the dissimilarities between AlN and Si results in a 3D Volmer-Weber growth mode with the nucleation of independent and relaxed AlN islands. Despite a well-defined epitaxial relationship, these islands exhibit in-plane misorientations up to 6°–7°. We propose a model which quantitatively explains these misorientations by taking into account the relaxation of the islands through the introduction of 60° a-type misfit dislocations. Threading dislocations (TDs) are formed to compensate these misorientations when islands coalesce. TD density depends on two parameters: the islands' misorientation and density. We show that the former is related to the mismatch between AlN and Si, while the latter depends on the growth parameters. A large decrease in TD density occurs during the 3D growth stage by overlap and overgrowth of highly misoriented islands. On the other hand, the TD density does not change significantly when the growth becomes 2D. The proposed model, explaining the misorientations of 3D-grown islands, may be extended to other (0001)-oriented III-nitrides and more generally to any heteroepitaxial system exhibiting a 3D Volmer-Weber growth mode with islands relaxed thanks to the introduction of mixed-type misfit dislocations. [ABSTRACT FROM AUTHOR]

In the analysis of highly doped silicon, energy band gap narrowing (BGN) and degeneracy effects may be accounted for separately, as a net BGN in conjunction with Fermi-Dirac statistics, or lumped together in an apparent BGN used with Boltzmann statistics. This paper presents an experimental study of silicon highly doped with boron, with the aim of evaluating the applicability of previously reported BGN models. Different boron diffusions covering a broad range of dopant densities were prepared, and their characteristic recombination current parameters J0 were measured using a contactless photoconductance technique. The BGN was subsequently extracted by matching theoretical simulations of carrier transport and recombination in each of the boron diffused regions and the measured J0 values. An evaluation of two different minority carrier mobility models indicates that their impact on the extraction of the BGN is relatively small. After considering possible uncertainties, it can be concluded that the BGN is slightly larger in p+ silicon than in n+ silicon, in qualitative agreement with theoretical predictions by Schenk. Nevertheless, in quantitative terms that theoretical model is found to slightly underestimate the BGN in p+ silicon. With the two different parameterizations derived in this paper for the BGN in p+ silicon, both statistical approaches, Boltzmann and Fermi-Dirac, provide a good agreement with the experimental data. [ABSTRACT FROM AUTHOR]

This paper reconsiders the mathematical formulation for the conventional nonparabolic band model and discusses how the nonparabolicity of the conduction band impacts the effective masses of electrons that are confined within the barriers. Since the conventional simplified model for band nonparabolicity does not include the external potential effect as a perturbation, it is examined whether this perturbation can be implemented into the conventional model for usefulness. When the perturbation energy is smaller than the unperturbed energy, the insertion of the perturbation energy term into the conventional expression for the nonparabolic band model is reasonably valid. It is demonstrated that this approximation is actually accurate given a sub-10-nm-thick Si layer surrounded by an SiO2 layer. A major part of this discussion is focused on the low-dimensionality electron system confined with insulator barriers so this study contributes to the analysis of future electron devices. To support this consideration, this paper proposes an analytical expression for the conduction-band effective mass of electrons including the term of band nonparabolicity. [ABSTRACT FROM AUTHOR]

Correct identification of local configurations of dopants and point defects in random alloys poses a challenge to both computational modeling and experimental characterization methods. In this paper, we propose and implement a computationally efficient approach to address this problem. Combining special quasirandom structures, virtual crystal approximation, and real-space lattice static Green's functions, we are able to calculate, at moderate computational cost, the local phonon density of states (LPDOSs) of impurities in a random alloy crystal for system sizes, surpassing the capabilities of a conventional, cubic-scaling, density functional theory. We validate this method by showing that our LPDOS predictions of substitutional silicon in GaAs and InAs are in excellent agreement with the experimental data. For the case study, we investigate a variety of local configurations of Si and Se substitutional dopants and cation vacancies in quasirandom In 0.5 Ga 0.5 As alloys. In all cases, the impurity LPDOS in a random alloy exhibits qualitatively different signatures from those in the pure binary compounds GaAs and InAs. Specifically, they are characterized by a wide continuous band (rather than narrow discrete peaks) of vibrational modes at frequencies typically higher than the bulk modes, a sign of coupling between localized vibrations of the impurity and those of its random neighboring host atoms. The accuracy and computational cost of this approach open a way to the simulation of impurities in random structures on a large scale and the prediction of vibrational signatures of alloys with defects. [ABSTRACT FROM AUTHOR]

A response to a commentary on the paper "A Model for Internal Photoemission at High-k Oxide/Silicon Energy Barriers" published within the issue is presented. He describes the results presented by the study authors on photoelectric emission from silicon surfaces. He also examines the assertion that the modified Powell equation used would lead to a different fraction of electrons which may potentially surmount the interface barrier.

Silicon solar cells containing boron and oxygen are one of the most rapidly growing forms of electricity generation. However, they suffer from significant degradation during the initial stages of use. This problem has been studied for 40 years resulting in over 250 research publications. Despite this, there is no consensus regarding the microscopic nature of the defect reactions responsible. In this paper, we present compelling evidence of the mechanism of degradation. We observe, using deep level transient spectroscopy and photoluminescence, under the action of light or injected carriers, the conversion of a deep boron-di-oxygen-related donor state into a shallow acceptor which correlates with the change in the lifetime of minority carriers in the silicon. Using ab initio modeling, we propose structures of the BsO2 defect which match the experimental findings. We put forward the hypothesis that the dominant recombination process associated with the degradation is trap-assisted Auger recombination. This assignment is supported by the observation of above bandgap luminescence due to hot carriers resulting from the Auger process. [ABSTRACT FROM AUTHOR]

III-V optoelectronics grown epitaxially on Si substrates have large networks of dislocations due to a lattice constant mismatch between the device layers and the substrate. Recombination-enhanced dislocation glide (REDG) allows these dislocations to move and increase in length during device operation, which degrades performance. In this paper, we study REDG dynamics of threading dislocations in situ in (In)AlGaAs double heterostructures grown on Si substrates using scanning electron microscopy cathodoluminescence. The driving force for REDG arises due to the coefficient of thermal expansion differences between Si and the III-V layers leading to large residual strains in the films. Tracking of threading dislocations as moving dark spot defects reveals glide characteristics that vary based on the nature of the dislocation. Remarkably, the alloying of a few atom percent of indium using metamorphic structures arrests threading dislocation glide by more than two orders of magnitude. Finally, we present REDG-based filtering as a pathway to reducing the threading dislocation density in select areas, removing a large fraction of the mobile dislocations. Together, these techniques will enable the understanding of dislocation–dislocation and carrier–dislocation interactions that have so far remained elusive during device operation, leading to reliable III-V integrated optoelectronics on silicon. [ABSTRACT FROM AUTHOR]

For the ⟨ 1 1 ¯ 0 ⟩ screw dislocation (nondissociated) in silicon, there are three types of core structures that are, respectively, referred to as the shuffle-set dislocation (A-core), glide-set dislocation (C-core), and mix-set dislocation (B-core). Each type of core further displays a planar or nonplanar (fourfold) configuration. In this paper, in the context of the generalized Peierls theory, the energy functional and equation of equilibrium of a screw dislocation with a nonorthogonal fourfold structure is derived and is used to investigate the structure and energy of the ⟨ 1 1 ¯ 0 ⟩ screw dislocation in silicon. We find that the energy of the shuffle-set dislocation with a fourfold core is considerably lower than that of the other types of cores. Thus, the shuffle-set dislocation with the fourfold core is the most stable. The glide-set dislocation with a planar core has the highest energy and is the most unstable. As the most stable structure, the shuffle-set screw dislocation with the fourfold core was responsible for wavy slip traces observed at low temperature. [ABSTRACT FROM AUTHOR]

This paper reports the magnetic properties of 2 μm thick electroplated isotropic L10 CoPt films on silicon at temperatures ranging from 300 K to 790 K, as well as the room-temperature properties of the films after various thermal cycles. Electroplated equiatomic CoPt layers require a post-deposition annealing typically at ∼973 K to induce L10 ordering so that they exhibit hard magnetic properties at room temperature. However, the influence of temperature on these post-deposition annealed films is an important consideration for their utility in end applications. Here, a reversible temperature coefficient of remanence of -0.11% K-1 is measured along with a maximum operating temperature of 400 K (recovery to 95% of the initial remanence). The maximal energy density of the films is reduced by 50% at a temperature of 550 K. However, the original room-temperature magnetic properties are shown to be fully recoverable by remagnetization after various thermal cycles-800 K for 15 min in Ar, 373 K for 168 h in air, and 358 K for 168 h in air at 85% relative humidity. These investigations indicate that the electroplated L10 CoPt layers, like bulk, exhibit good thermal robustness, satisfying a key requirement for their use in end applications, such as magnetic microsystems. [ABSTRACT FROM AUTHOR]

This paper provides a detailed description explaining how to calculate the relation between the silicon Raman frequency and local stress or strain in the silicon, applied to stress measurements in microelectronics. This relation is well known for measurements from the (100) surface of silicon. However, it is often used in the wrong way, neglecting non-zero stress tensor elements. Especially, in current 3D microelectronics technology, where the stress caused by through Si vias or microbumps is of large importance, the vertical stress component, which highly affects the measured Raman frequency shift, is often erroneously neglected. In addition, the equations for the (100) surface are also often used incorrectly for cross-sectional measurements from a (110) surface. In this paper, different ways to calculate the relation between Raman frequency and triaxial stress, and the related Raman peak intensities, are discussed in detail. [ABSTRACT FROM AUTHOR]

The scientific issue of this paper is the formation of the initial surface roughening during black silicon (b-Si) preparation by maskless SF 6 /O 2 plasma texturing. In detail, the authors investigate a novel approach whether merely substrate temperature dependent surface mechanisms and plasma particle diffusion are sufficient to theoretically obtain anisotropic etching. For that, a quasi-2D model is developed including the relevant mechanisms such as (i) etching, (ii) the deposition of the masking layer SiO x F y , (iii) plasma particle transport, and (iv) heat diffusion. Further on, a linear stability analysis is applied, firstly, to reveal theoretical conditions for anisotropic etching and, secondly, to qualitatively evaluate the impact of the model parameters on the texturing range. The evaluation shows that plasma particle diffusion along the surface is the main factor for nano-roughening. Additionally, the experimentally expected strong dependency of the texturing on the substrate temperature is confirmed and other extracted dependencies can be correlated to experimental observations. With that, a novel model is introduced explaining the initial b-Si roughening without taking into account surface removal by directed ions. [ABSTRACT FROM AUTHOR]

A practical challenge in slicing of low-cost multi-crystalline silicon (mc-Si) wafers by the fixed abrasive diamond wire sawing process is increased wire consumption due to greater wear of the diamond compared to slicing of the more expensive mono-crystalline silicon (mono-Si) wafers. In this paper, we present the results of scribing of mc-Si and mono-Si materials with two conical tip diamond indenters of the same geometry to understand the possible reasons for increased diamond wear in cutting of multi-crystalline silicon. Specifically, the scribing forces and the diamond indenter wear produced in scribing of the two silicon materials are analyzed. The results show that the forces generated in scribing of mc-Si are higher than in scribing of mono-Si. The higher forces in scribing of mc-Si are consistent with the corresponding higher tip radius of curvature (due to wear) of the diamond indenter compared to the tip radius produced in scribing of mono-Si. Scanning electron microscopy and confocal microscopy of the diamond indenters show that wear is primarily due to physical micro-fracture and blunting of the diamond. Raman spectroscopy shows evidence of stress-induced phase transformation of the diamond and the formation of compressive residual stress in the diamond. Plausible physical reasons, including the role of material inhomogeneity in mc-Si, for the wear of diamond during scribing are given. [ABSTRACT FROM AUTHOR]

This paper presents a comparative analysis of the structure and photoluminescence (PL) of Si containing dislocations introduced by thermal shock or ion implantation. To study the structure of dislocation cores and their interaction with point defects, we used a high-resolution transmission mode during in situ electron irradiation in the JEM4000EX operating at 400 kV. An appropriate PL spectrum was obtained on dislocated Si after electron irradiation by an external pulse gun operating at 350 kV. This resulted only in an increase in the D2 line intensity that correlated with the formation of metastable interstitial defects on {001}, {111}, and {113} planes near various dislocation cores during in situ electron irradiation, regardless of their types. However, the D1 line relates to a core structure of a shuffle Lomer dislocation consisting of 5/7 atomic rings, which occurs when two shuffle 60° dislocations (so-called S1 type [Pizzagalli et al., Phys. Rev. Lett. 103, 065505 (2009)]) combine with each other in intersecting {111} planes in plastically deformed Si or when an extrinsic Frank partial dislocation transforms into a perfect one in an ion-implanted layer. [ABSTRACT FROM AUTHOR]

Low-cost shortwave infrared detectors have great potential for emerging civilian night-vision applications. This paper reports the characteristics of Ge0.89Sn0.11 photodiodes monolithically grown on a Si substrate that holds great promise for those applications. At room temperature, the 500 μm diameter active area device demonstrated a longwave cutoff of 2.65 μm and a responsivity of 0.32 A/W at 2 μm, which corresponds to an external quantum efficiency of 20% without any contribution from the Ge buffer layer. The measured peak specific detectivity at 300 K and 77 K is 1.7 × 109 Jones and 4.3 × 109 Jones, respectively. The specific detectivity at 77 K is only one-order-of-magnitude lower than that of the market dominating extended-InGaAs photodiode. The detailed device analysis indicated that the 700-nm thick fully relaxed high-quality GeSn absorbing layer and the modified depletion region lead to the above-mentioned device performance. [ABSTRACT FROM AUTHOR]

Flexible piezoelectric thin films on polymeric substrates provide advantages in sensing, actuating, and energy harvesting applications. However, direct deposition of many inorganic piezoelectric materials such as Pb(Zrx,Ti1-x)O3 (PZT) on polymers is challenging due to the high temperature required for crystallization. This paper describes a transfer process for PZT thin films. The PZT films are first grown on a high-temperature capable substrate such as platinum-coated silicon. After crystallization, a polymeric layer is added, and the polymer-PZT combination is removed from the high-temperature substrate by etching away a release layer, with the polymer layer then becoming the substrate. The released PZT on polyimide exhibits enhanced dielectric response due to reduction in substrate clamping after removal from the rigid substrate. For Pb(Zr0.52,Ti0.48)0.98Nb0.02O3 films, release from Si increased the remanent polarization from 17.5 μC/cm2 to 26 μC/cm2. In addition, poling led to increased ferroelastic/ferroelectric realignment in the released films. At 1 kHz, the average permittivity was measured to be around 1160 after release from Si with a loss tangent below 3%. Rayleigh measurements further confirmed the correlation between diminished substrate constraint and increased domain wall mobility in the released PZT films on polymers. [ABSTRACT FROM AUTHOR]

The thermal rectification has potential application in the thermal management of electronics and energy saving. This paper investigates the thermal rectification in silicon nanofilm with cone cavity by the non-equilibrium molecular dynamics simulation. The results show that the thermal resistance caused by the cone cavity is on the order of 10-9 m2 KW-1. The thermal resistance of cone cavity differs averagely by about 11% from 300 K to 1100 K when the heat flow direction is turned over. The phonon wave packet dynamic simulations indicate that transverse phonons are generated by the cavity scattering of longitudinal phonon. When the phonon transport direction is reversed, the average transmission coefficient is changed by about 14% in all the frequency range and is regarded as the main reason for the thermal rectification in silicon film with cone cavity. The thermal rectification depends on the aspect ratio of the cone cavity. [ABSTRACT FROM AUTHOR]

We study the influence of phonon boundary scattering on the transport of thermal energy in semiconductor nanowires from micro to nano length scales. We use a kinetic theory model based on the Boltzmann transport equation that accurately calculates the reduction of the phonon mean free paths by considering their directional dependence and the fundamental statistical definition of the mean free path. As a result, our model does not use phenomenological formulas to account for the reduction of the phonon mean free paths due to boundary scattering. The transport of thermal energy is also fully divided into that carried by different polarizations by separating phonon group velocities and relaxation times for transverse and longitudinal phonons. We study the correctness of using frequency independent versus frequency dependent models for describing the specularity of the nanowire boundary. We also examine the validity of the assumption that phonons in the semiconductor nanowire maintain their bulk phonon dispersion relations and that modifications to the dispersion relations due to phonon confinement effects can be neglected. The thermal conductivities of silicon nanowires are calculated for different length scales and temperatures and good agreement is obtained with experiments. The theoretical results in this paper can be used to understand and quantitatively predict heat transport in nanowires, which is critical for increasing the efficiency of thermoelectric and electronic devices. [ABSTRACT FROM AUTHOR]

In this paper, a comparison of three different models for the conductivity and mobility is given for the case of silicon nanowire devices in the presence of electron-phonon scattering. The consistency of all three models in the case of homogeneous nanowires is demonstrated. The scattering limited conductivity and mobility is a well defined quantity in this case. For nonhomogeneous systems like triple-gate nanowires FETs, these scattering limited quantities are no longer well defined for very short gate lengths. The quality of the underlying assumptions and the physical interpretation of the differences in the resulting transport characteristics are discussed. [ABSTRACT FROM AUTHOR]

Nanosecond (ns) laser ablation of silicon has important applications in electronics, photovoltaics, and many other areas. However, most of the previous work is mainly focused on ultraviolet (UV) or visible ns lasers, and the investigation is very limited for infrared lasers (e.g., at 1064 nm), which have the advantage of much lower cost and external energy consumption. In this paper, a thermal model has been developed and experimentally verified for ablation of silicon by 200 ns and 1064 nm laser pulses. This kind of work has been rarely reported in literature. The model predictions agree reasonably well with experimental measurements. The investigation indicates that surface evaporation should be the dominant physical mechanism of material removal for laser silicon ablation under the studied conditions. It has also been found that the physical processes of multipulse laser-silicon interaction at 1064 nm may strongly depend on laser pulse number and repetition rate (the pulse-to-pulse temporal distance). [ABSTRACT FROM AUTHOR]

The mechanisms of amorphization for crystalline Si (c-Si) induced by ultraviolet femtosecond laser irradiation are described in this paper. The wavelength of the laser pulse was 267 nm, which is the third harmonics of a Ti:sapphire laser. We performed a laser scanning microscopy and a transmission electron microscopy for surface and structural analysis and imaging pump-probe measurements to investigate the dynamics of the process. From the analyses, we confirmed that the thickness of the amorphized layer was quite uniform and there is no lattice defect under the amorphized section. The thickness of the amorphous Si (a-Si) layer was 7 nm and the threshold fluence of the amorphization was 44 mJ/cm2. From the Imaging Pump-Probe measurement it was revealed that the melting time is less than 1 ns and ultra high speed melting and re-solidification process was occurred. The melting depth estimated by the Imaging Pump-Probe measurement was 7 nm. The melted portion completely corresponded to the amorphized section. [ABSTRACT FROM AUTHOR]

In this paper an interferometric setup called “thermoelastic microscopy” has been used to measure the vibration response of the semiconductor cantilevers under modulated laser thermal source. The small vibration amplitude has been detected well below 10 pm magnitude in accurate adjustment conditions. The results showed that experimental responses have a good agreement with the theoretical ones. Also from the analysis we could conclude that the signal amplitude has a power functional dependence on the modulation frequency and that the phase angle linearly depended on the square root of modulation frequency. [ABSTRACT FROM AUTHOR]

Long-range surface plasmons traveling on thin metal films have demonstrated promising potential in subwavelength waveguide applications. In work toward device applications that can leverage existing silicon microelectronics technology, it is of interest to explore the propagation of surface plasmons in a metal-oxide-semiconductor geometry. In such a structure, there is a high refractive index contrast between the semiconductor (n≈3.5 for silicon) and the insulating oxide (typically n≈1.5-2.5). However, the introduction of dielectrics with disparate refractive indices is known to strongly affect the guiding properties of surface plasmons. In this paper, we analyze the implications of high index contrast in 1D layered surface plasmon structures. We show that it is possible to introduce a thin dielectric layer with a low refractive index positioned next to the metal without adversely affecting the guiding quality. In fact, such a configuration can dramatically increase the propagation length of the conventional long-range mode. While this study is directed at silicon-compatible waveguides working at telecommunications wavelengths, this configuration has general implications for surface plasmon structure design using other materials and operating at alternative wavelengths. [ABSTRACT FROM AUTHOR]

Compact low-loss delay lines using a low-resistivity silicon substrate are designed and investigated in this paper. The studied lines have metal-insulator-semiconductor (MIS) substrate profiles. Substrate parameters such as layer thickness and resistivity are chosen in a way to allow for low-loss signal propagation in the slow-wave region. Subsequently, meander MIS microstrip lines are designed and their layout parameters are optimized to maximize the delay while maintaining the component area small. [ABSTRACT FROM AUTHOR]

When measuring I-V characteristics and carrier lifetimes in quasi-steady-state (QSS) conditions, it is important to consider the time dependence of the charge due to excess carriers within the device. This paper shows that the space-charge region present in pn-junction devices and in many lifetime test structures can produce a significant capacitive effect when measuring the low voltage and low carrier density range of QSS I-V curves. Both computer modeling and experiments show that the junction capacitance is particularly significant in the case of low-resistivity silicon wafers, but it can also be noticeable in intermediate and high-resistivity samples. The paper demonstrates that the static I-V characteristics can be accurately reconstructed using a simple analytical model for the space-charge region. It thus fills a gap in the understanding of the low injection range of QSS voltage and lifetime measurements. [ABSTRACT FROM AUTHOR]

Nanoelectromechanical systems (NEMS) can be designed and characterized by understanding the interaction and coupling between the mechanical, electrical, and the van der Waals energy domains. In this paper, we present physical models and their numerical simulation for coupled electrical and mechanical analysis of silicon NEMS. A nonlinear continuum elastic model is employed for mechanical analysis. The material properties required in the continuum model are extracted from molecular-dynamics simulations. We present three electrostatic models—namely, the classical conductor model, the semiclassical model, and the quantum-mechanical model, for electrostatic analysis of NEMS at various length scales. The electrostatic models also account for the corrections to the energy gap and the effective mass due to the strain in the silicon nanostructure. A continuum layer approach is introduced to compute the van der Waals forces. The coupling between the mechanical, electrical, and the van der Waals energy domains as well as their numerical implementation is described. Numerical results are presented for several silicon NEM switches to understand the static electromechanical pull-in behavior. [ABSTRACT FROM AUTHOR]

Erratum: "Numerical simulation and validation of subsurface modification and crack formation induced by nanosecond-pulsed laser processing in monocrystalline silicon" [J. Appl. Phys. In our previously published paper, there are two typographical errors in Table I. (1) There is an error in the units for electron mobility. 35 and 36GPaFracture toughness37KIC0.9MPa-m1/2 ht 1 a I m SB e sb i , the electron rest mass. [Extracted from the article]

In this paper, we study the recombination mechanisms in ion-implanted black silicon (bSi) emitters and discuss their advantages over diffused emitters. In the case of diffusion, the large bSi surface area increases emitter doping and consequently Auger recombination compared to a planar surface. The total doping dose is on the contrary independent of the surface area in implanted emitters, and as a result, we show that ion implantation allows control of emitter doping without compromise in the surface aspect ratio. The possibility to control surface doping via implantation anneal becomes highly advantageous in bSi emitters, where surface passivation becomes critical due to the increased surface area. We extract fundamental surface recombination velocities Sn through numerical simulations and obtain the lowest values at the highest anneal temperatures. With these conditions, an excellent emitter saturation current (J0e) is obtained in implanted bSi emitters, reaching 20 fA/cm2±5 fA/cm2 at a sheet resistance of 170 X/sq. Finally, we identify the different regimes of recombination in planar and bSi emitters as a function of implantation anneal temperature. Based on experimental data and numerical simulations, we show that surface recombination can be reduced to a negligible contribution in implanted bSi emitters, which explains the low J0e obtained. [ABSTRACT FROM AUTHOR]

The electrical performance a solar cell is determined from direct measurements of the current voltage relationship, while the so-called implied current-voltage measurements are often performed to estimate the performance of partially processed samples. Implied current voltage measurements are commonly obtained from quasi steady state photoconductance and quasi steady state photoluminescence measurements, where the implied voltage is inferred from the average excess carrier density. As will be shown here, this approach can be problematic due to the presence of excess carriers that do not contribute to the terminal voltage. These carriers are referred to as voltage independent carriers, a concept that is not widely known or generally accepted. This paper provides the theoretical background for the distinction of voltage dependent and voltage independent carriers. It is shown that the relative impact of voltage independent carriers on implied voltage data depends strongly on device parameters and on the illumination wavelength. Practical limits are provided for these parameters for which the voltage independent carriers can be neglected and for which an implied voltage thus accurately reflects the junction voltage. [ABSTRACT FROM AUTHOR]

The presence of copper contamination is known to cause strong light-induced degradation (Cu-LID) in silicon. In this paper, we parametrize the recombination activity of light-activated copper defects in terms of Shockley-Read-Hall recombination statistics through injection- and temperature dependent lifetime spectroscopy (TDLS) performed on deliberately contaminated float zone silicon wafers. We obtain an accurate fit of the experimental data via two non-interacting energy levels, i.e., a deep recombination center featuring an energy level at Ec - Et = 0:48 - 0:62 eV with a moderate donor-like capture asymmetry (k = 1:7 - 2:6Þ and an additional shallow energy state located at Ec - Et = 0:1 - 0:2 eV, which mostly affects the carrier lifetime only at high-injection conditions. Besides confirming these defect parameters, TDLS measurements also indicate a power-law temperature dependence of the capture cross sections associated with the deep energy state. Eventually, we compare these results with the available literature data, and we find that the formation of copper precipitates is the probable root cause behind Cu-LID. [ABSTRACT FROM AUTHOR]

In this paper, we present experimental evidence on the change of the phonon spectrum and vibrational properties of a bulk material through phonon hybridization mechanisms. The phonon spectrum in a finite material is strongly affected by the presence of free surfaces, which is the addition of a contribution from an essentially two-dimensional crystal. The phonon spectrum of a bulk material can hence be altered by a hybridization mechanism between confined phonon modes in nanostructures introduced on the surface of a bulk material and the underlying bulk phonon modes. We measured the heat capacities of bare and surface-structured silicon substrates originating from the same silicon wafer. Then, we deduced important features of the phonon spectra of the samples investigated through a rigorous analysis of the measured heat capacity curves. The results show that the shape and size of the nanostructures made on the surface of the bulk substrate have a strong effect on the phonon spectrum of the bulk material. [ABSTRACT FROM AUTHOR]

An evaporated Al layer is known as an excellent rear metallization for highly efficient solar cells, but suffers from incompatibility with a common solder process. To enable solar cellinterconnection and module integration, in this work the Al layer is complemented with a solder stack of TiN/Ti/Ag or TiN/NiV/Ag, in which the TiN layer acts as an Al diffusion barrier. X-ray photoelectron spectroscopy measurements prove that diffusion of Al through the stack and the formation of an Al2O3 layer on the stack's surface are responsible for a loss of solderability after a strong post-metallization anneal, which is often mandatory to improve contact resistance and passivation quality. An optimization of the reactive TiN sputter process results in a densification of the TiN layer, which improves its barrier quality against Al diffusion. However, measurements with Xray diffraction and scanning electron microscopy show that small grains with vertical grain boundaries persist, which still offer fast diffusion paths. Therefore, the concept of stuffing is introduced. By incorporating oxygen into the grain boundaries of the sputtered TiN layer, Al diffusion is strongly reduced as confirmed by secondary ion mass spectroscopy profiles. A quantitative analysis reveals a one order of magnitude lower Al diffusion coefficient for stuffed TiN layers. This metallization system maintains its solderability even after strong post-metallization annealing at 425 °C for 15 min. This paper thus presents an industrially feasible, conventionally solderable, and longterm stable metallization scheme for highly efficient silicon solar cells. [ABSTRACT FROM AUTHOR]

Presents information on a study which constructed a set of linear and of parabolic rate constants within a linear-parabolic model of oxide growth on silicon in dry oxygen. Methodology of the study; Results and discussion; Conclusions.

The article presents the author's response to comments on the study "Intrinsic point defect incorporation in silicon single crystals grown from a melt, revisited" which appeared in a 2012 issue. He states that the equation used in the paper to estimate the stress is easy to use, yields the right order of magnitude and provides an idea of the independence on the crystal radius. He also mentions the factors which need to be taken into account when using the Voronkov criterion.

This paper discusses the effects of laser fluence on silicon modification by four-beam laser interference. In this work, four-beam laser interference was used to pattern single crystal silicon wafers for the fabrication of surface structures, and the number of laser pulses was applied to the process in air. By controlling the parameters of laser irradiation, different shapes of silicon structures were fabricated. The results were obtained with the single laser fluence of 354 mJ/cm2, 495 mJ/cm2, and 637 mJ/cm2, the pulse repetition rate of 10 Hz, the laser exposure pulses of 30, 100, and 300, the laser wavelength of 1064 nm, and the pulse duration of 7-9 ns. The effects of the heat transfer and the radiation of laser interference plasma on silicon wafer surfaces were investigated. The equations of heat flow and radiation effects of laser plasma of interfering patterns in a four-beam laser interference distribution were proposed to describe their impacts on silicon wafer surfaces. The experimental results have shown that the laser fluence has to be properly selected for the fabrication of well-defined surface structures in a four-beam laser interference process. Laser interference patterns can directly fabricate different shape structures for their corresponding applications. [ABSTRACT FROM AUTHOR]

This paper reports the results of atomistic and ab initio simulation of several different tilt grain boundaries in silicon. The boundary structures obtained with genetic algorithm turned out to have no coordination defects, i.e., all silicon atoms restored their tetrahedral coordination during the structure optimisation. That concerns previously known symmetric Σ5 (130), Σ3 (211), and Σ29 (520) boundaries and previously unknown asymmetric Σ9 (̅255)/(̅211), Σ3 (̅255)/(211), and Σ13 (790)/(3 11 0) structures. An extensive study has been performed on defect segregation on the boundaries, including neutral vacancy and carbon, phosphorus, and boron impurities. A clear correlation has been revealed between the segregation energy of the defect and local geometry of the boundary site where the defect is segregated. The authors propose a simple purely geometric model for evaluating approximate segregation energies of the listed defects. [ABSTRACT FROM AUTHOR]

Historically, two microscopic electron scattering calculation methods have been used to fit macroscopic electron mobility data in n-type silicon. The first method was performed using a static system that included long-range electron-plasmon scattering; however, the well-known Born approximation fails in this case when dealing with electron-impurity scattering. In the second method, sophisticated numerical simulations were developed around plasmon-excited potential fluctuations and successfully reproduced the mobility data at room temperature. In this paper, we propose a third method as an alternative to the first method. First, using a fluctuating system, which was characterized on the basis of our recently experimentally extracted plasmon-excited potential fluctuations, the microscopic calculations reveal enhanced short-range scattering of electrons by phonons and ionized impurities due to increased electron temperature and increased screening length, respectively. The increased hot electron population makes the Born approximation hold, which eases the overall calculation task substantially. Then, we return to the static system while incorporating plasmon-enhanced impurity scattering. The resulting macroscopic electron mobility shows fairly good agreement with data over wide ranges of temperatures (200-400 K) and doping concentrations (1015-1020cm-3). Application of the proposed method to strained silicon is also demonstrated. [ABSTRACT FROM AUTHOR]

Ohmic contacts to AlGaN grown on sapphire substrates have been previously demonstrated for various compositions of AlGaN, but contacts to AlGaN grown on native AlN substrates are more difficult to obtain. In this paper, a model is developed that describes current flow through contacts to Si-doped AlGaN. This model treats the current through reverse-biased Schottky barriers as a consequence of two different tunneling-dependent conduction mechanisms in parallel, i.e., Fowler-Nordheim emission and defect-assisted Frenkel-Poole emission. At low bias, the defect-assisted tunneling dominates, but as the potential across the depletion region increases, tunneling begins to occur without the assistance of defects, and the Fowler-Nordheim emission becomes the dominant conduction mechanism. Transfer length method measurements and temperature-dependent current-voltage (I-V) measurements of Ti/Al-based contacts to Si-doped AlGaN grown on sapphire and AlN substrates support this model. Defect-assisted tunneling plays a much larger role in the contacts to AlGaN on sapphire, resulting in nearly linear I-V characteristics. In contrast, contacts to AlGaN on AlN show limited defect-assisted tunneling appear to be only semi-Ohmic. [ABSTRACT FROM AUTHOR]

This paper describes the properties of Ge1-ySny light emitting diodes with a broad range of Sn concentrations (y=0.0-0.11). The devices are grown upon Si(100) platforms using ultra-low temperature deposition of highly reactive Ge and Sn hydrides. The device fabrication adopts two new photodiode designs which lead to optimized performance and enables a systematic study of the effects of strain relaxation on emission efficiency. In contrast with n-Ge/i-Ge1-ySny/p-Ge analogs, which in most cases contain two defected interfaces, our designs include a p-layer with composition Ge1-zSnz chosen to be z1-xSnx alloy is also used for the n layer, with compositions in the 0≤x≤y range, so that defected and non-defected n/i interfaces can be studied. The electroluminescence spectra vs the Sn content y in the intrinsic layer of the diodes exhibit a monotonic shift in the emission wavelength from 1550 nm to 2500 nm. On the other hand, the emission intensities show a complex dependence that cannot be explained solely on the basis of Sn concentrations. Detailed theoretical modeling of these intensities makes it possible to extract recombination lifetimes that are found to be more than three times longer in samples in which strain relaxation has not occurred at the n-i interface, demonstrating the existence of a large nonradiative contribution from the relaxation defects. This finding is particularly significant for direct gap diodes with y>0.09, for which it is practically impossible to avoid strain relaxation in n-Ge/i-Ge1-ySny/p-Ge analogs. The new designs introduced here open the door to the fabrication of highly efficient electrically pumped systems for applications in future generations of integrated photonics. [ABSTRACT FROM AUTHOR]

This paper describes the development of an efficient and accurate dynamic hysteresis model that combines the Cauer circuit representations with the play model. The physical meaning of the standard Cauer circuit is discussed and is used to derive a mathematical representation of hysteretic inductors. The iron-loss and hysteresis loops of silicon steel that were obtained using the proposed model agree with experimental data measured under sinusoidal and pulse width modulation excitations. [ABSTRACT FROM AUTHOR]

In this paper, we study the effect of random discrete dopants on the performance of a 6.6 nm channel length silicon FinFET. The discrete dopants have been distributed randomly in the source/drain region of the device. Due to the small dimensions of the FinFET, a quantum transport formalism based on the non-equilibrium Green's functions has been deployed. The transfer characteristics for several devices that differ in location and number of dopants have been calculated. Our results demonstrate that discrete dopants modify the effective channel length and the height of the source/ drain barrier, consequently changing the channel control of the charge. This effect becomes more significant at high drain bias. As a consequence, there is a strong effect on the variability of the oncurrent, off-current, sub-threshold slope, and threshold voltage. Finally, we have also calculated the mean and standard deviation of these parameters to quantify their variability. The obtained results show that the variability at high drain bias is 1.75 larger than at low drain bias. However, the variability of the on-current, off-current, and sub-threshold slope remains independent of the drain bias. In addition, we have found that a large source to drain current by tunnelling current occurs at low gate bias. [ABSTRACT FROM AUTHOR]

Ultrashort pulse lasers are promising tools for numerous measurement purposes. Among other benefits their high peak powers allow for efficient generation of wavelengths in broad spectral ranges and at spectral powers that are orders of magnitude higher than in conventional light sources. Very recently this has been exploited for the establishment of sophisticated measurement facilities for electrical characterization of photovoltaic (PV) devices. As the high peak powers of ultrashort pulses promote nonlinear optical effects they might also give rise to nonlinear interactions with the devices under test that possibly manipulate the measurement outcome. In this paper, we present a comprehensive theoretical and experimental study of the nonlinearities affecting short circuit current (ISC) measurements of silicon (Si) solar cells. We derive a set of coupled differential equations describing the radiation-device interaction and discuss the nonlinearities incorporated in those. By a semi-analytical approach introducing a quasi-steady-state approximation and integrating a Green's function we solve the system of equations and obtain simulated ISC values. We validate the theoretical model by ISC ratios obtained from a double ring resonator setup capable for reproducible generation of various ultrashort pulse trains. Finally, we apply the model to conduct the most prominent comparison of ISC generated by ultrashort pulses versus continuous illumination. We conclude by the important finding that the nonlinearities induced by ultrashort pulses are negligible for the most common ISC measurements. However, we also find that more specialized measurements (e.g., of concentrating PV or Si-multijunction devices as well as highly localized electrical characterizations) will be biased by two-photon-absorption distorting the ISC measurement. [ABSTRACT FROM AUTHOR]

Carrier recombination at the metal-semiconductor contacts has become a significant obstacle to the further advancement of high-efficiency diffused-junction silicon solar cells. This paper provides the proof-of-concept of a procedure to reduce contact recombination by means of enhanced metalinsulator- semiconductor (MIS) structures. Lightly diffused n+ and p+ surfaces are passivated with SiO2/a-Si:H and Al2O3/a-Si:H stacks, respectively, before the MIS contacts are formed by a thermally activated alloying process between the a-Si:H layer and an overlying aluminum film. Transmission/scanning transmission electron microscopy (TEM/STEM) and energy dispersive x-ray spectroscopy are used to ascertain the nature of the alloy. Idealized solar cell simulations reveal that MIS(n+) contacts, with SiO2 thicknesses of ~1.55 nm, achieve the best carrierselectivity producing a contact resistivity qc of ~3 mΩ cm2 and a recombination current density J0c of ~40 fA/cm2. These characteristics are shown to be stable at temperatures up to 350 °C. The MIS(p+) contacts fail to achieve equivalent results both in terms of thermal stability and contact characteristics but may still offer advantages over directly metallized contacts in terms of manufacturing simplicity. [ABSTRACT FROM AUTHOR]

Co-implantation, with overlapping implantation projected ranges, of Si and doping species (P, As, and B) followed by a thermal annealing step is a viable route to form doped Si nanocrystals (NCs) embedded in silica (SiO2). In this paper, we investigate optical characterizations of both doped and un-doped Si-NCs prepared by this method. The effective NC presence in the oxide layer and their crystallinity is verified by Raman spectrometry. Photoluminescence (PL) and PL excitation measurements reveal quantum confinement effects and a gradual PL quenching with increasing dopant concentrations. In un-doped NC, the measured Stokes shift remains constant and its value ~0.2 eV is almost twice the Si-O vibration energy. This suggests that a possible radiative recombination path is a fundamental transition assisted by a local phonon. PL lifetime investigations show that PL time-decays follow a stretched exponential. Using a statistical model for luminescence quenching, a typical NC diameter close to 2 nm is obtained for As- and P-doped samples, consistent with our previous atomic probe tomography (APT) analyses. APT also demonstrated that n-type dopant (P and As) are efficiently introduced in the NC core, whereas p-type dopant (B) are located at the NC/SiO2 interface. This last observation could explain the failure of the luminescence-quenching model to determine NC size in B-doped samples. All together, these experimental observations question on possible different carrier recombination paths in P or As doped NC compared to B one's. [ABSTRACT FROM AUTHOR]

Phosphorus (P) or boron (B) atoms can be doped at temperatures as low as 80 to 350 °C, when crystalline silicon (c-Si) is exposed only for a few minutes to species generated by catalytic cracking reaction of phosphine (PH3) or diborane (B2H6) with heated tungsten (W) catalyzer. This paper is to investigate systematically this novel doping method, "Cat-doping", in detail. The electrical properties of P or B doped layers are studied by the Van der Pauw method based on the Hall effects measurement. The profiles of P or B atoms in c-Si are observed by secondary ion mass spectrometry mainly from back side of samples to eliminate knock-on effects. It is confirmed that the surface of p-type c-Si is converted to n-type by P Cat-doping at 80 °C, and similarly, that of n-type c-Si is to p-type by B Cat-doping. The doping depth is as shallow as 5 nm or less and the electrically activated doping concentration is 1018 to 1019 cm-3 for both P and B doping. It is also found that the surface potential of c-Si is controlled by the shallow Cat-doping and that the surface recombination velocity of minority carriers in c-Si can be enormously lowered by this potential control. [ABSTRACT FROM AUTHOR]

In this paper, a fluence-dependent antireflection performance is presented from ion-beam fabricated nanofaceted-Si surfaces. It is also demonstrated that these nanofacets are capable of producing room temperature ultra-violet and blue photoluminescence which can be attributed to inter-band transitions of the localized excitonic states of different Si-O bonds at the Si/SiOx interface. Timeresolved photoluminescence measurements further confirm defect-induced radiative emission from the surface of silicon nanofacets. It is observed that the spectral characteristics remain unchanged, except an enhancement in the photoluminescence intensity with increasing ion-fluence. The increase in photoluminescence intensity by orders of magnitude stronger than that of a planar Si substrate is due to higher absorption of incident photons by nanofaceted structures. [ABSTRACT FROM AUTHOR]

Low-temperature (≤200 °C) epitaxial growth yields precise thickness, doping, and thermal-budget control, which enables advanced-design semiconductor devices. In this paper, we use plasmaenhanced chemical vapor deposition to grow homo-epitaxial layers and study the different growth modes on crystalline silicon substrates. In particular, we determine the conditions leading to epitaxial growth in light of a model that depends only on the silane concentration in the plasma and the mean free path length of surface adatoms. For such growth, we show that the presence of a persistent defective interface layer between the crystalline silicon substrate and the epitaxial layer stems not only from the growth conditions but also from unintentional contamination of the reactor. Based on our findings, we determine the plasma conditions to grow high-quality bulk epitaxial films and propose a two-step growth process to obtain device-grade material. [ABSTRACT FROM AUTHOR]

In a recently published paper, suggestions have been contained concerning close relations existing between the position of the inflection point on the capacitance-voltage, C(VG), characteristic of the metal-oxide-semiconductor (MOS) structure and the position of the flat-band voltage of this structure. In this article, a systematic, quantitative analysis of this problem is presented. The analysis of the relations existing between the voltage of the inflection point VINFL on the capacitance-voltage, C(VG), characteristic and the flat-band voltage, VFB, of a MOS structure is conducted for ideal and non-ideal (but with negligible density of interface states, Dit) MOS structures on silicon and silicon carbide substrates. It is found that, in general, there is a significant difference between the VINFL and the VFB values. The VINFL - VFB difference is found to depend primarily on the substrate doping density ND and on the dielectric layer thickness tOX. This difference is calculated for a range of ND and tOX values for structures with negligible Dit values on silicon and silicon carbide substrates. Experiments are described which were conducted to allow comparison of calculated and experimentally obtained VINFL values of different MOS structures on Si and SiC substrates. Results of these experiments support the quantitative results of our analysis for structures with negligible densities of interface states. As expected, for structures with higher Dit values, significant differences are found between calculated and experimentally obtained VINFL values. [ABSTRACT FROM AUTHOR]