Search

The eddy-current method has been employed to realize the real-time measurement of thickness distribution and polishing endpoint detection of metal films during the metal chemical mechanical polishing (CMP) process. However, accuracy and sensitivity in thickness measurement can be significantly impacted by a variation in lift-off distance (about 3.5 mm), attributable to the progressive wear of the polishing pad. This paper proposes a novel triple-coil sensor system, integrated with an alternating-current bridge and designed for high-precision metal film thickness measurement. Subsequently, the measurement performance of the bridge output voltage was scrutinized using a theoretical model and a trans-dimensional finite element analysis model of electromagnetic fields and circuit coupling. The correlations between the amplitude, phase, and the ratio of the real part to the imaginary part of the output voltage, in context with film thickness, were determined at assorted excitation frequencies and lift-off distances. Both theoretical equations and simulation results affirmed that the phase and ratio were more resistant to lift-off distance variations than the amplitude, and a linear relationship was identified between the ratio and film thickness. Furthermore, the thickness measurement performance of the amplitude and phase, when the triple-coil was imbalanced, was dissected via numerical simulation. A profound understanding of the proposed system was provided and beneficial for the practical applications in real-time thickness measurements of metal films during the metal CMP process. [ABSTRACT FROM AUTHOR]

The synthesis of CeO2 nanoparticles for CeO2 based slurry gains continuous emphasis on improving its performance in the chemical mechanical polishing of dielectric materials. Urea was selected to dominate the growth and morphology during the calcination process. Thermogravimetry experiments were used to analyze the the decomposition behavior. Particle morphology and size were analyzed. Crystalline phase information and surface valence were used to compare the differences in surface physical and chemical properties of ceria by different synthesis process. The CeO2 nanoparticles synthesized with urea were dispersed in water as slurry. The particle sizes of CeO2 were measured by dynamic light scattering. The Zeta potential of CeO2 dispersion were measured to show dispersing performance. The CeO2 nanoparticles synthesized with urea condensation show good monodisperse properties. The material removal rate of silicon oxide and surface quality after chemical mechanical polishing were selected to evaluate the chemical mechanical polishing performance. The CeO2 nanoparticles synthesized with urea condensation not only yielded better surface quality results than the commercial slurry but also showed a 153% (pH = 4) and 100% (pH = 10) increase in the material removal rate of silicon oxide compared to commercial. [ABSTRACT FROM AUTHOR]

Cobalt (Co), with its low resistivity, superior adhesion property, and void-free seamless fill ability, is being considered by the industry as a promising candidate to replace the conventional tungsten contact plugs for middle-of-the-line and the wiring metal copper interconnects for back-end-of-the-line. Chemical mechanical polishing (CMP), whose demand is greatly increasing as device dimensions shrink to 7-nm node, has been recognized as one of the indispensable manufacturing approaches for the fabrication of multilevel metal interconnection structures. In addition to meeting the appropriate removal rates and removal selection of heterogeneous materials, excellent wafer surface quality and minimized residual particles, also have become the criteria in the developing process of CMP slurries for Co interconnects. This work investigates the optimization process of the final CMP step for Co interconnects by introducing surfactants in hydrogen peroxide based slurries. Various types of surfactants having different functional groups were employed and compared thoroughly by material removal rates (MRRs) and selectivity. The chemical and mechanical mechanisms of various surfactants during CMP process have been clearly revealed by various approaches, including polishing and static etching experiments, electrochemical analysis, tests of absolute zeta potentials, and particle sizes, as well as wafer surface energy measurements. Research results show that Co MRRs are directly correlated with functional groups of various surfactants. Moreover, absolute zeta potentials of abrasive particles can be deeply influenced by different types of surfactants, where the correlation between absolute zeta potentials and Co MRRs presents strong negative. On the other hand, direct positive correlation between oxide MRRs and particle sizes affected by corresponding surfactants can be summarized. Meanwhile, no obvious corresponding relationship between contact angles and MRRs was found, indicating MRRs were barely affected by wettability of wafer surfaces. At last, the selected surfactants were applied to 12-inch wafers for verification, including MRRs, removal selectivity, defects and surface roughness. Each of these surfactants could make Co and TEOS surface Ra maintain at around below 0.7 nm and 0.3 nm, with a number of residual particles lower than 600 particle/wafer. [ABSTRACT FROM AUTHOR]

Recently, the technology node shrunk below 7 nm and beyond; cobalt (Co) and its low resistivity, superior adhesion property, and excellent electro-migration performance pledge to transform the integrated circuits' landscape in several areas, especially in interconnects and logic contacts. The present work proposes a weakly alkaline polishing slurry for Co interconnects, mainly consisting of colloidal silica, hydrogen peroxide, glycine, and 1,2,4-triazole at pH 8.1. The feature of tunable high material removal rate (greater than 1000–2000 Å/min) and fitting removal selectivity (~ 160) between Co and oxide dielectric could be realized by a moderately variable concentration of glycine complexing agent. Moreover, several analytical techniques, including electrochemical analysis, X-ray photoelectron spectroscopy survey, and adsorption isotherm characterization, were used to understand the interaction mechanism during Co material removal process. The results reveal that 1,2,4-triazole could adsorb on the Co oxide layers via physisorption, achieving an inhibition effect to avoid Co excessive corrosion. Furthermore, the oxidization of hydrogen peroxide, the complexation of glycine zwitterions, and the abrasive mechanical action are critical to maintaining the appropriate material removal rate. Based on the results with research-level wafers, the optimized Co bulk slurry was ultimately applied to the production-level of Co and oxide dielectric wafers. The residual particle, haze condition, and surface profile after polishing have been researched simultaneously. [ABSTRACT FROM AUTHOR]

Person attribute recognition, i.e., the prediction of a fixed set of semantic attributes given an image of a person, becomes an important topic in the field of computer vision. Recently, methods based on convolutional neural networks have shown outstanding performance in this area. They usually employ a CNN network to mine the shared feature representation followed by several layers for attribute classification. To improve the representation ability of the model, many methods element-add or concatenate coarse and fine feature maps to fuse information at different feature levels. However, these methods didn't fully exploit the interaction of multi-level convolutional feature maps for person attribute analysis and not consider the correlation of attributes for the same person. In this paper, we introduce a kind of correlation feature, which exploits the high order interaction of coarse and fine feature maps to capture the robust feature representation from multi-level convolution layers as the image representation for person attribute recognition. Moreover, we propose an intraperson attribute loss to explicitly model the correlation of attributes for the same person. We experiment our proposed model on CIFAR-10 dataset, Berkeley Human Attributes dataset, PA-100 K dataset, and experimental results show the better performance of the feature representation and the effectiveness of intra-person attribute loss. [ABSTRACT FROM AUTHOR]

Cobalt (Co) has been applied as one of the most promising candidates of barrier metals for copper (Cu) interconnects. The present work describes the static etching and chemical mechanical polishing process of Cu and Co, which were conducted by potassium persulfate (K2S2O8) as an oxidizer at various pH values. It was found that compared with the conventional oxidizer hydrogen peroxide (H2O2)-based slurries, the K2S2O8-based slurries exhibited a relatively high Co removal rate, as well as the diminished particulate contamination and excellent post-etching morphology. A slurry consisting of 3 vol% colloidal silica, 10 mM K2S2O8, and 5 mM benzotriazole (BTA) produced a Co removal rate of ~ 127 Å/min at pH 10, along with a removal rate selectivity of ~ 1 between Cu and Co films. Based on the data of X-ray photoelectron spectroscopy, scanning electron microscopy, and nano-scratch depth tests, the interplaying mechanisms of chemical and mechanical on Cu and Co removal rates in K2S2O8-based slurries were investigated. [ABSTRACT FROM AUTHOR]

In this paper, we make full use of the complementarity between low-rank feature and part-based feature and present a novel method extracting discriminative parts with flexible number from low-rank features for action recognition. The proposed method avoids some intermediate processing steps (e.g., actor segmentation, body tracking) required by many traditional methods and can greatly avoid memorizing background information suffered by traditional part-based methods. In addition, traditional part-based methods usually set a fixed and identical number of discriminative parts for all action categories neglecting the differences of recognizing complexity among different action categories. On the contrary, we automatically extract discriminative parts with flexible number for each action category by introducing group sparse regularizer into our model, which is more reasonable and effective. In our method, we first extract low-rank features of all action sequences and transform them into corresponding low-rank images. Then, we densely sample each low-rank image into a large number of parts in multi-scale and represent each part into a feature vector. Afterward, our model automatically learn a set of discriminative part detectors with flexible number for each action category. We further define new similarity constraints to force the responses of detected parts from the same class more similar and consistent and that from different class more different. Finally, we define a corresponding recognition criterion to perform final action recognition. The efficacy of the proposed method is verified on three public datasets, and experimental results have shown the promising results of our method for human action recognition. [ABSTRACT FROM AUTHOR]

Ruthenium is the most promising material for the barrier layer used for the sub 14 nm technology node in integrated circuits manufacturing. Potassium periodate (KIO)-based slurry is used in the chemical mechanical planarization (CMP) process of the barrier layer. However, the electrochemical and corrosion properties of ruthenium have not been investigated in such slurry. In this paper, the electrochemical and corrosion behaviors of ruthenium in KIO solutions were investigated under static conditions but at different pH values by potentiodynamic polarization and electrochemical impedance spectroscopy measurements, combined with surface chemical analysis using auger electron spectroscopy. Moreover, to study wear enhanced corrosion during CMP, tribocorrosion experiments were carried out to monitor the current density changes during and after mechanical scratching. The results show that at pH 6, ruthenium forms a relatively thick and heterogeneous surface film composed of RuO·2HO/RuO, showing a high corrosion resistance and it exhibits a quick repassivation after mechanical scratching. At pH 4, ruthenium shows a passivation behavior with formation of a uniform and conductive oxide like RuO·2HO. It should be noted that there is a possible formation of RuO toxic gas under this condition, which should be avoided in the actual production. However, at pH 11, ruthenium exhibits no considerable passivity and the corrosion proceeds uniformly. [ABSTRACT FROM AUTHOR]

Although traditional bag-of-words model, together with local spatiotemporal features, has shown promising results for human action recognition, it ignores all structural information of features, which carries important information of motion structures in videos. Recent methods usually characterize the relationship of quantized spatiotemporal features to overcome this drawback. However, the propagation of quantization error leads to an unreliable representation. To alleviate the propagation of quantization error, we present a coding method, which considers not only the spatial similarity but also the reconstruction ability of visual words after giving a probabilistic interpretation of coding coefficients. Based on our coding method, a new type of feature called cumulative probability histogram is proposed to robustly characterize contextual structural information around interest points, which are extracted from multi-layered contexts and assumed to be complementary to local spatiotemporal features. The proposed method is verified on four benchmark datasets. Experiment results show that our method can achieve better performance than previous methods in action recognition. [ABSTRACT FROM AUTHOR]

Machine vision provides an efficient way for the automatic crack detection of civil structures. However, it is still very challenging to extract the small cracks embedded in noisy background. Even some very recent methods require manual intervention or omission of crack width. In this paper, we aim at extracting such inconspicuous cracks automatically with width information preserved. The basic idea of the proposed method is to assign the pixel points to some arbitrarily shaped clusters, and then sift out the crack clusters according to their elongated shapes. Treating each gray-level image as a parametric surface, we devise an anisotropic clustering algorithm that exploits the geometric properties of the surface. By virtue of the geometric representation and the anisotropy, this algorithm solves the problem of separating adjacent objects while simultaneously grouping the fragments of a crack into the same cluster. Moreover, the globally convex segmentation model is incorporated into our method, serving as a tool that provides appropriate candidate points and important parameters for the clustering procedure. Experimental results on real images demonstrate that the cracks extracted by our method are very similar to manually traced ground truth cracks and thus can be used for measuring the widths of real cracks. [ABSTRACT FROM AUTHOR]

Motion representation is a challenging task in human action recognition. To represent motion, most traditional methods usually require certain intermediate processing steps such as actor segmentation, body tracking, and interest point detection, which make these methods sensitive to errors caused by these processing steps. In this paper, motivated by the successful recovery of low-rank matrix using robust principal component analysis (RPCA), we present a novel motion representation method for action recognition by extracting refined low-rank features of RPCA. Compared with the traditional methods, our method does not require the intermediate processing steps mentioned above. Unfortunately, with traditional λ, RPCA is incapable of extracting the discriminative information of motion in action videos, thus we first conduct extensive experiments to determine a feasible parameter λ suitable for action recognition. Then, we perform RPCA with this λ to obtain the low-rank images including the discriminative information of motion. To represent characteristic of the obtained low-rank images, we define two descriptors [i.e., edge distribution histogram (EDH) and accumulated edge distribution histogram (AEDH)] to refine the low-rank images. Finally, a support vector machine is trained to classify human actions represented by EDH or AEDH features. The efficacy of the proposed method is verified on three public datasets, and experimental results have shown the promising results of our method for human action recognition. [ABSTRACT FROM AUTHOR]

Copper chemical mechanical polishing (CMP) in barrier layer slurries with periodate as oxidant has not been intensively studied. This work presents an investigation into copper tribocorrosion in potassium periodate-based slurries during CMP. The research focused on copper tribocorrosion behavior, the surface chemical and electrochemical reaction products, and the electrochemical mechanism during CMP. The copper surface film was characterized by Raman spectra experiments. Tribocorrosion tests combined with CMP chemical experiments were conducted to study the tribochemical behavior and wear-accelerated corrosion effect. The results show that copper corrosion is more severe in acid solutions than in alkaline conditions. The copper surface is mainly passivated with copper oxides, copper hydroxide, and copper iodide. Small amounts of copper iodate, copper periodate, and iodine could be detected under specific pH conditions. Abrasion could help to get a uniform passivation film on copper surface consisting of copper oxides, copper hydroxide, and copper iodide only. The material loss due to wear-accelerated corrosion during CMP was also investigated. The results show that under weakly alkaline conditions (pH 9 and pH 10), the wear-corrosion effect plays an important role in the total material loss due to corrosion. The wear-accelerated corrosion is mainly caused by the exposure of more cathodic reaction sites to the slurry for participation in the redox reaction and the local galvanic corrosion during the CMP process. The high wear corrosion proportion of the total corrosion (Δ I/ I) could help to obtain a better surface quality and desirable material removal rate during CMP. [ABSTRACT FROM AUTHOR]

Cytotoxic T-lymphocyte antigen 4 ( CTLA-4) is a potent immunoregulatory molecule that suppresses antitumor response by downregulating T cell activation. The most studied CTLA-4 +49A/G polymorphism has been associated with various cancers risks. However, little is known about the association between CTLA-4 +49A/G polymorphism and bladder cancer risk. A hospital-based case-control study was conducted in 300 patients with bladder cancer and 300 healthy controls matched with age and sex. The CTLA-4 +49A/G polymorphism was genotyped using polymerase chain reaction-restriction fragment length polymorphism. Patients with bladder cancer had a significantly lower frequency of CTLA-4 +49GG genotype [odds ratio (OR) = 0.44, 95 % confidence interval (CI) = 0.23, 0.85; P = 0.01] and G allele (OR = 0.73, 95 % CI = 0.56, 0.96; P = 0.02) than healthy controls. When stratifying by the stage, grade, and histological type of bladder cancer, we found no statistical association. This is the first study to highlight the significant association between CTLA-4 +49A/G polymorphism and bladder cancer risk. Additional studies are needed to confirm this finding. [ABSTRACT FROM AUTHOR]

In this paper, the effect of slurry injection position on removal rate and nonuniformity was experimentally investigated on a 300-mm polisher developed by the authors, and the effect of slurry injection position on slurry fluid pressure distribution during polishing was also measured by a novel in situ fluid pressure measurement system integrated in the polisher. In situ pad conditioning was employed in all experiments. The results showed that removal rate, nonuniformity, and fluid pressure distribution were influenced dramatically by the slurry injection position. The slurry injection position determines the slurry distribution, and therefore, the removal rate profile and fluid pressure distribution were affected. [ABSTRACT FROM AUTHOR]

A finite element analysis (FEA) model is developed for the chemical-mechanical polishing (CMP) process on the basis of a 12-in five-zone polishing head. The proposed FEA model shows that the contact stress non-uniformity is less dependent on the material property of the membrane and the geometry of the retaining ring. The larger the elastic modulus of the pad, the larger contact stress non-uniformity of the wafer. The applied loads on retaining ring and zone of the polishing head significantly affect the contact stress distribution. The stress adjustment ability of a zone depends on its position. In particular, the inner-side zone has a high stress adjustment ability, whereas the outer-side zone has a low stress adjustment ability. The predicted results by the model are shown to be consistent with the experimental data. Analysis results have revealed some insights regarding the performance of the multi-zone CMP. [ABSTRACT FROM AUTHOR]

Presents an approach to calculate the sound field generated by the vibrating surface of a body with arbitrary shape moving at a small Mach number. Derivations from united aerodynamics and aeroacoustic approach; Sound pressure distribution on surface and in far field for the sphere with vibrating surface; Sound pressure distribution.