ASM International (Netherlands)

Silicon oxynitride films with varying oxygen/nitrogen ratio were grown from SiH4, N2O, and NH3 by means of a plasma-enchanced chemical vapor deposition process. The elemental composition of the deposited films was measured by a variety of high-energy ion beam techniques. To determine the chemical structure we used Fourier transform infrared absorption spectroscopy and electron-spin resonance. Ellipsometric data and values for mechanical stress are also reported. We show that the entire range of compositions from silicon oxide to silicon nitride can be covered by applying two different processes and by adjusting the N2O/NH3 gas flow ratio of the respective processes. It is suggested that the N2O/SiH4 gas flow ratio is the major deposition characterization parameter, which also controls the chemical structure as far as the hydrogen bonding configuration is concerned. We found that the films contain significant amounts of excess silicon and that the mechanical stress in the oxynitrides is lower than in plasma nitride. The electron-spin density is low (∼1017/cm3) in all samples.

The anneal behavior of plasma-enhanced chemical vapor deposited silicon oxynitride films has been studied using Fourier transform infrared absorption spectroscopy, nuclear reaction analysis, and electron-spin resonance. The anneal temperature range was 500–1000 °C. It is observed that the oxynitrides which contain only N–H bonds are thermally stable in the temperature range under study. The layers which also contain Si–H bonds are considerably less thermally stable. Abundant hydrogen effusion from these layers is observed at temperatures as low as 600 °C, accompanied by cracking and shrinkage of the films. It is suggested that the coexistence of both Si–H and N–H bonds offers the possibility for cross linking and that consequently the decomposition temperature of both types of bonds is lowered. Evidence for the occurrence of cross linking is found in the infrared difference spectra. Consistently, the silicon unpaired electron density does not increase upon annealing. The Si–H and N–H bands effectively shift towards higher wave numbers upon annealing at higher temperatures. This is ascribed to the inhomogeneity in bond strength, which in turn is related to a variation in electronegativity of the surrounding groups.

In this paper, we introduce a variable-gain control strategy for mechanical ventilators in the respiratory systems. Respiratory systems assist the patients who have difficulty breathing on their own. For the comfort of the patient, fast pressure buildup (and release) and a stable flow response are desired. However, linear controllers typically need to balance between these conflicting objectives. In order to balance this tradeoff in a more desirable manner, a variable-gain controller is proposed, which switches the controller gain based on the magnitude of the patient flow. The effectiveness of the control strategy is demonstrated in experiments on different test lungs.

Abstract Striving for nanometer‐sized solid‐state single‐photon sources, we investigate atom‐like quantum emitters based on single germanium‐vacancy (GeV) centers isolated in crystalline nanodiamonds (NDs). Cryogenic characterization indicated symmetry‐protected and bright (>10 6 counts/s with off‐resonance excitation) zero‐phonon optical transitions with up to 6‐fold enhancement in energy splitting of their ground states as compared to that found for GeV centers in bulk diamonds (i.e. up to 870 GHz in highly strained NDs vs. 150 GHz in bulk). Utilizing lithographic alignment techniques, we demonstrate an integrated nanophotonic platform for deterministic interfacing plasmonic waveguides with isolated GeV centers in NDs, which enables 10‐fold enhancement of single‐photon decay rates along with the emission direction control by judiciously designing and positioning a Bragg reflector. This approach allows one to realize the unidirectional emission from single‐photon dipolar sources, thereby opening new perspectives for the realization of quantum optical integrated circuits.

We investigate the influence of thermal conditions during solid phase epitaxial regrowth (SPER) on the electrical activation level of boron in preamorphized silicon, both with respect to heating ramp rates and the use of low temperature preanneals. Enhancement of electrically active boron concentration by 36% is observed for activation with the fastest ramp rate (487°C∕s) compared to the slowest one (1°C∕s). An important clustering pathway occurs within the amorphous silicon phase (during low temperature preanneal) prior to completion of the SPER process. In these junctions boron deactivation during isochronal post-annealing is almost independent on the maximum boron activation level.

In this paper the thermal behavior of a diffusion furnace is studied. The ultimate goal is to achieve control of the wafer temperatures. It is shown how some previous results from the literature on the behavior of wafer temperatures in a diffusion furnace are incomplete and partly erroneous. A control algorithm has been derived which achieves much better wafer-to-wafer uniformity than conventional control.

Ultra-shallow p+ junctions formed by solid phase epitaxial regrowth (SPER) have promise for sub-65 nm CMOS technologies. Due to above-equilibrium solid solubilities and minimal diffusion, such junctions can far outperform spike-annealed junctions in terms of resistance, abruptness, and depth. However, the low-temperature annealing does not dissolve the end of range defects creating concerns for junction leakage in the device. In this work, we show how SPER junctions can be optimized to meet the ITRS junction profile and low-power leakage requirements of the 45 nm CMOS node [International Technology Roadmap for Semiconductors (Semiconductor Industry Association, San Jose, CA, 2001)]. Diode leakage is shown to decrease with Ge amorphization depth and B dose and energy. Leakage is shown to increase dramatically with the background doping level. Increasing the regrowth, or post-annealing, thermal budget improves leakage and can be optimized to avoid deactivation. The inclusion of a preanneal does not affect the junction leakage, however co-implanting F increases leakage. The influence of each is explained using various physical and electrical characterization techniques. Optimizing these parameters gives a junction of 9 nm with an Rs of 698 Ω/sq and area leakage of 1E-6 A/cm2 with HALO doping.

A model, based on surface energy minimization under nonequilibrium conditions, is presented to describe the evolution of the amorphous silicon (a-Si) topography near the hemispherical grained silicon. The evolution of the depletion area can be explained by a combination between capture of silicon (Si) atoms at the grain boundary and energy minimization of the surrounding a-Si surface. Grain depletion dependence on annealing time was measured by means of transmission electron microscopy. The simulated results agree well with the real observations. This approach is presented as a first step in physically based modeling of HSG formation.

We investigated the effect of hydroxyethyl cellulose (HEC) concentration in the alkaline slurry on the surface qualities of polished silicon wafers after touch polishing, such as the number of remaining particles and the haze level. They significantly decreased and saturated over the addition of a 0.2 wt % HEC. We attributed these surface quality enhancements to a formation of nanometer-sized passivation layers onto the silicon wafer and colloidal silica surfaces during touch polishing. The formation of passivation layers were shown by the results of investigations with an average particle size, the zeta potential of abrasives, the contact angle, and X-ray photoelectron spectroscopy measurements.

AbstractUsing a sample of 4,148 convertibles issued over 1990–2009 by companies listed in 35 countries, the authors exploited worldwide differences in short-sale constraints to examine whether short selling by convertible arbitrageurs creates downward pressure on convertible issuers' stock prices. They found that short-sale constraints have a positive effect on issue-date abnormal stock returns, which suggests that a substantial part of the stock price effect of convertible issues is attributable to convertible arbitrageurs. Convertible bonds are securities that can be converted, at the option of the holder, into a fixed number of the issuer's ordinary shares. Convertible arbitrage hedge funds have played an important role in the convertible bond market, especially since the beginning of the 21st century. These hedge funds combine long positions in convertibles with short positions in the underlying stock.We exploited worldwide differences in short-sale constraints to examine whether convertible arbitrage short selling creates downward pressure on convertible issuers' stock prices. Because arbitrage hedge funds are unable to execute their hedging strategy in markets that are short-sale constrained, we used the existence of short-sale constraints as a proxy for the presence of convertible arbitrage hedge funds in a market. We hypothesized that convertibles issued by companies listed in countries where short selling is legally restricted are associated with more favorable issue-date stock price effects than are convertibles issued in countries where short selling is allowed and practiced. We tested this hypothesis with a sample of 4,148 convertible bonds issued over 1990–2009 by companies listed in 35 countries. In line with our hypothesis, we found that short-sale constraints have a positive effect on issue-date abnormal stock returns. We further found that this effect is stronger in years with higher hedge fund involvement and for offerings expected to induce more arbitrage short selling.In addition, our study maps the global convertible bond market as completely as permitted by publicly available data sources and offers new insights into the determinants of the negative stock price reaction associated with convertible bond offerings. Previous papers have attributed this negative reaction to the signaling content of convertible bond issues. Our approach allowed us to estimate the magnitude of downward price pressure around convertible bond offerings that is attributable to the actions of convertible arbitrageurs rather than to the negative signal inferred from the convertible bond announcement. Our findings suggest that both academics and practitioners who analyze post-2000 convertible bond announcement effects are likely to overstate the negative announcement effects when they fail to control for the short-sale pressure of convertible bond arbitrageurs. On a more general level, our study suggests that stock price behavior around corporate financing events can be substantially affected by short-selling regulations.

We believe that the most important task in the development of high-/spl kappa/ gate dielectrics is to engineer the interface to assure high enough mobility and reliability. Considering the 100-nm node, Al/sub 2/O/sub 3/ would appear to be the most promising candidate in terms of chemical and thermal stability, barrier offset, and compatibility with the conventional CMOS process. The integration of Al/sub 2/O/sub 3/ gate dielectrics in sub-100 nm-FETs has already been demonstrated; however, the resulting electron mobility was only a quarter the value for a FET with SiO/sub 2/ gate dielectric (D. Buchanan et al., Tech. Digest IEDM, p. 223, 2000; J.H. Lee et al., ibid., p. 645, 2000). We have clarified the mechanism by which mobility is thus degraded, both experimentally and theoretically.

The limits of using B or BF <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> alone in forming ultrashallow junctions have been reached for the 90 nm CMOS generation. In this paper we evaluate the use of Ge and F co-implants to extend conventional implantation and spike anneal to the 65 nm CMOS technology node. In this work we show that the F co-implant can improve the abruptness of the B junction, while the single Ge usually degrades it. The use of Ge co-implanted with F gives the best junction abruptness - less than 5nm/decade. The best trade-off between junction depth (Xj) and sheet resistance (R <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">sheet</sub> ) is achieved by deep Ge pre-amorphization and deep co-implantation of F. A comparison between slow and fast ramp-up is made. Significant improvement for the junction activation, its depth and abruptness is obtained by spike anneal with fast ramp-up for B junctions with Ge and F co-implantation.

The recently introduced Direct Bonded Heterogeneous Integration (DBHi) Si bridge technology has chips connected by a bridge housed inside a cavity or trench that is precisely machined in the chip-carrier laminate. The size, thickness, and placement centrality of the bridge are critical parameters that are necessary to prevent interference with the laminate substrate trench. In this study, we compared laser and saw dicing processes for the DBHi bridge technology. Different laser dicing recipes are developed and optimized. The laser and saw dicing process are compared for dimensional tolerances, edge quality, and die strength of the bridge chips. The die strength is affected by the edge defects on the top and bottom surfaces adjacent to the edges and is measured using a three-point bend fixture in a Universal Testing Machine (UTM). The laser dicing process is improved to equalize the die strength at both the surfaces as well as to achieve equal or higher die strength than blade dicing specifically for the thinner wafer thicknesses.

Nanothermometers can detect changes in the local temperature in living cells and in vivo, revealing fundamental biological properties. Despite the exploration of different temperature-responsive materials, the design and development of temperature-sensing probes with high brightness and high sensitivity remain a daunting challenge. Here, we employed the UiO-66 type metal–organic frameworks (MOFs) to anchor UNCPs on the surface of the MOFs for constructing MOF@UCNPs nanohybrids. The in situ composite method with MOFs leads to the coordination interaction between the ligands and the surface of UCNPs, enabling controlled composite formation between different MOFs and UCNPs. Remarkably, the surface interaction favors the anomalous thermo-enhanced luminescence, achieving a 35-fold enhancement of UiO-66@NaYF4:Yb/Tm at 413 K. Furthermore, these MOF@UCNPs nanohybrids with thermo-enhanced luminescence are developed as multifunctional biological probes for bioimaging and intracellular temperature sensing, demonstrating a high thermal sensitivity of 1.92% K–1 in the physiological temperature range. Based on these findings, temperature monitoring of the local position was successfully carried out by the designed MOF@UCNPs nanoprobes in vivo. These findings underscore the potential of MOF@UCNPs nanohybrids, opening up new avenues for the development of a multifunctional platform for biological analysis.

Investigations for next generation contacts of silicon and SiGe devices show that a 2-step nickel salicidation process is favorable over a single step NiSi and over CoSi <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> in every respect and can be introduced easily in existing and advanced not fully depleted CMOS flows once the post silicidation thermal treatments can be kept below 700degC. Partial conversion for the deposited Ni layer to Ni <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2 </sub> Si in a first RTP1 step at temperatures as low as 250degC avoids the reverse linewidth effect and enables superior uniformities over complete conversion. A second RTP2 step at typically 450degC is used to form low resistivity NiSi with less silicon consumption and lower contact resistivities than today's CoSi <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> contacts. Challenging integration issue are peripheral leakage currents, that are likely to be related to undesired low temperature pyramidal NiSi <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2 </sub> formation and spiking

Semi-insulating polycrystalline silicon films have been grown by a low-pressure chemical vapor deposition process between 565 and 623 °C. Rutherford backscattering spectrometry and nuclear reaction analysis were used to determine the layer composition and thickness. The O/Si concentration ratio in the films increases with increasing N2O/SiH4 input for N2O/SiH4&amp;lt;0.6. For N2O/SiH4&amp;gt;0.6, a saturation effect in the oxygen uptake was observed. Furthermore the O/Si ratio increases with increasing growth temperature. The hydrogen concentration increases more than linearly with the N2O/SiH4 input ratio for N2O/SiH4&amp;lt;0.5 up to an amount of 8.5 at. % at 600 °C. For N2O/SiH4&amp;gt;0.5 a saturation effect was observed in the hydrogen uptake. We suggest that the hydrogen is bound to silicon atoms, which are further coordinated with three oxygen atoms. Based on this model, the growth mode of the films at the temperatures considered is discussed.

The use of an in-line electrical metrology tool for evaluation and characterization of high-k dielectrics has been investigated. The equivalent oxide thickness obtained from the KLA-Tencor Quantox agrees well with high-frequency capacitance-voltage (C-V) results. The ACTIV Jg index parameter shows good correlation with the gate current obtained from C-V measurements. Therefore, this tool can be used as a fast and much less expensive evaluation tool for optimization of high-k process conditions and process stability monitoring. © 2003 The Electrochemical Society. All rights reserved.

Over the last years, singulation of thin semiconductor wafers with (ultra) low-K top layer has become a challenge in the production process of integrated circuits. The traditional blade dicing process is encountering serious yield issues. These issues can be addressed by applying a laser grooving process prior to the blade dicing, which is the process of reference nowadays. However, as wafers are becoming thinner, this process flow is not providing the yield and cost required. This paper will discuss the results of a study done on several multi beam laser ablation technologies on thin Si wafers and describe the pro's and con's for each of them.

In order to improve cycle time in the furnace area, two alternatives are evaluated. Mini batch manufacturing and integrated metrology can both save valuable time. Dynamic simulations are used to investigate the influence of both options. The results are a cycle time improvement when going front large batch to mini batch manufacturing. The gain is up to 40% for normal lots and up to 30% for hot lots wafers. The consequence is an increase in the number of required tubes. Cycle time improvement versus costs is analysed. The cost of one hour cycle time gain is determined. Integrated metrology can save approximately 5-10% on the average cycle time.

We have measured electron-spin resonance spectra of silicon-oxynitride layers grown by plasma-enhanced chemical vapor deposition from gas mixtures of SiH4, N2O, and NH3. The analyses were performed on untreated, thermally annealed, and gamma irradiated samples. A relatively small spin density (about 1017 cm−3) is found for the as-deposited layers. Since the resonances remain weak upon annealing, gamma irradiation was performed in order to acquire more structural information on the silicon-oxynitride network. The evolution of the spin density upon annealing is discussed in relation to the hydrogen release during annealing. In the as-deposited and annealed samples, g values were found close to 2.0055. This strongly suggests the presence of silicon clusters in the material. Spin densities remain low upon annealing unless we anneal at temperatures as high as 1000 °C. For these temperatures, smaller g values are observed. Irradiation with gamma rays induces similar signals except for oxygen-rich samples. These show a narrow line at a g value of about 2.001, which can be ascribed to E′ centers. The intensity of this signal falls rapidly with decreasing oxygen content, which implies that these materials are not made up of a mixture of a silicon-nitride and a silicon-oxide phase.