Search

The metal-free synthesis of graphene films on Si substrates, the most common commercial semiconductors, is of paramount significance for graphene application on semiconductors and in the field of electronics. However, since current research mainly uses C-H gas as the carbon source in chemical vapor deposition (CVD), and Si does not have a catalytic effect on the decomposition and adsorption of C-H gas, it is challenging to prepare high-quality graphene on the Si surface directly. In this work, we report the growth of graphene directly on Si without metal catalysis by CVD using CO was selected as the carbon source. By controlling the growth temperature (1000–1150 °C), a process of 2–5 layers of graphene growth on silicon was developed. The electrical performance results showed that the graphene film had a sheet resistance of 79 Ω/sq, a resistivity of 7.06 × 10−7 Ω·m, and a carrier migration rate of up to 1473.1 cm2 V−1·S−1. This work would be a significant step toward the growth of graphene on silicon substrates with CO as the carbon source.

Magnetic domain formation in two-dimensional (2D) materials gives perspectives into the fundamental origins of 2D magnetism and also motivates the development of advanced spintronics devices. However, the characterization of magnetic domains in atomically thin van der Waals (vdW) flakes remains challenging. Here, we employ X-ray photoemission electron microscopy (XPEEM) to perform layer-resolved imaging of the domain structures in the itinerant vdW ferromagnet Fe

WC7Co/Ti6Al4V composite coatings are deposited on the pure Ti substrate by pulse laser cladding (LC). During the laser melting process, the decomposition of WC7Co particles will lead to the evolution of microstructure and phases, which is directly related to the wear resistance and mechanism of composite coating. The microstructural evolution, phase compositions and interface reaction of WC7Co/Ti6Al4V composite coating were examined by scanning electron microscopy, energy-dispersive spectrum and X-ray diffraction (XRD). The hardness of different structures and abrasive resistance of composite coating were measured. The results show that the typical microstructure of LC WC7Co/Ti6Al4V composite coating can be classified into dissolved WC7Co composite structure and un-dissolved WC7Co structure. According to XRD results, there are Ti solid solution, W, TiC, VC, Co3W3C and secondary W2C in composite coating. The eutectic structure formed by the dissolved WC7Co particles consisted of W, W2C, TiC and β-Ti solid solution. The mean hardness of different structures exhibits a significant gradient distribution in composite coating. A reaction layer composed of TiC, W and W2C is also generated onto the interface between un-dissolved WC7Co particles and Ti6Al4V alloy matrix. The abrasive mechanisms of WC7Co/Ti6Al4V composite coating are mainly adhesive wear and oxidation wear during the dry sliding process.

Magnetic domain formation in two-dimensional (2D) materials gives perspectives into the fundamental origins of 2D magnetism and also motivates the development of advanced spintronics devices. However, the characterization of magnetic domains in atomically thin van der Waals (vdW) flakes remains challenging. Here, we employ X-ray photoemission electron microscopy (XPEEM) to perform layer-resolved imaging of the domain structures in the itinerant vdW ferromagnet Fe5GeTe2which shows near room temperature bulk ferromagnetism and a weak perpendicular magnetic anisotropy (PMA). In the bulk limit, we observe the well-known labyrinth-type domains. Thinner flakes, on the other hand, are characterized by increasingly fragmented domains. While PMA is a characteristic property of Fe5GeTe2, we observe a spin-reorientation transition with the spins canting in-plane for flakes thinner than six layers. Notably, a bubble phase emerges in four-layer flakes. This thickness dependence, which clearly deviates from the single-domain behavior observed in other 2D magnetic materials, demonstrates the exciting prospect of stabilizing complex spin textures in 2D vdW magnets at relatively high temperatures.

The effects of (NH4)2ZrF6 concentration, voltage and treating time on the corrosion resistance of ZK61M magnesium alloy micro-arc oxidation coatings were studied by orthogonal experiments. The SEM result shows that the surface roughness and porosity of MAO coatings increased with (NH4)2ZrF6 concentration, voltage and treating time as a whole, except the porosity decreased with treating time. EDS, XRD and XPS analysis show that (NH4)2ZrF6 was successfully incorporated into coatings by reactive incorporation, coatings are dominantly composed of ZrO2, MgO, MgF2 and amorphous phase Mg phosphate. Potentiodynamic polarization was used to evaluate the corrosion property of coatings. When the concentration of (NH4)2ZrF6 is 6 g/L, the voltage is 450 V, and the treating time is 15 min, the coating exhibits the best corrosion resistance which corrosion current density is four magnitudes lower than substrate attributed to the incorporation of ZrO2 and the deposition of MgF2 in the micropores.

Micro-arc oxidation (MAO) coatings were obtained from an AZ91D magnesium alloy at different oxidation times (5, 10, 15, and 20 min), using a zirconium salt electrolyte system, with (NH4)2ZrF6 as the main salt. The morphology of the coatings was studied using a scanning electron microscope (SEM) and confocal laser scanning microscopy (CLSM). Energy dispersive spectrometry (EDS), X-ray diffractometry (XRD), and X-ray photoelectron spectroscopy (XPS) were employed to determine the type of element and the composition of its phase. The potentiodynamic polarization curve (PDP) was applied to illustrate the corrosion resistance of the coatings. We found the coatings had minor porosity and the best compactness when the MAO treatment time was 10 min. The coatings mainly comprised MgO, ZrO2, MgF2, and Zr3O2F8 phases and amorphous magnesium phosphate. Among the MAO coatings prepared in this experiment, the 10 min coating had the lowest corrosion current density (Icorr), and the Icorr was 4.864 × 10−8 A/cm2, which was three orders of magnitude lower than the uncoated AZ91D magnesium alloy.

The zirconia (ZrO2) as a precursor layer for the diffusion barrier coating (DBC) was deposited by electron beam physical vapor deposition (EB-PVD) between the Rene N5 superalloy substrate and NiCrAl coating. The high temperature exposure for the N5/ZrO2/NiCrAl system was carried out at 1000 °C for 5 h, 100 h and 200 h in atmosphere to investigate the structure evolution of the ZrO2 precursor layer and the elements interdiffusion of the system compared with the N5/NiCrAl system. The results showed that a sandwich structural DBC (α-Al2O3/rich-Zr/α-Al2O3) was formed through redox reaction between zirconia and aluminum under high temperature, leading to the transformation of N5/ZrO2/NiCrAl system into the N5/DBC/NiCrAl system. It was found that the sandwiched DBC could effectively suppress the interdiffusion of elements compared to the N5/NiCrAl system.

A ceramic particle reinforced glass matrix composite coating with a novel Ni + CrAlYSiN cermet bond-coat and an AlN diffusion barrier was prepared on K417G superalloy, and its oxidation behavior at 1000 °C was investigated. The oxidation rate of the coating was significantly reduced compared with the Ni + CrAlYSiN coating, because the glass-based top-coat promoted selective oxidation of Al within the bond-coat due to the oxygen barrier effect. Furthermore, cyclic oxidation resistance of the coating was improved by the low expansion bond-coat while weakened by the microstructure degradation of the top-coat. Additionally, the AlN diffusion barrier effectively restrained interdiffusion between bond-coat and substrate.

Deep learning (DL) has been increasingly adopted by a variety of software-intensive systems. Developers mainly use GPUs to accelerate the training, testing, and deployment of DL models. However, the GPU memory consumed by a DL model is often unknown to them before the DL job executes. Therefore, an improper choice of neural architecture or hyperparameters can cause such a job to run out of the limited GPU memory and fail. Our recent empirical study has found that many DL job failures are due to the exhaustion of GPU memory. This leads to a horrendous waste of computing resources and a significant reduction in development productivity. In this paper, we propose DNNMem, an accurate estimation tool for GPU memory consumption of DL models. DNNMem employs an analytic estimation approach to systematically calculate the memory consumption of both the computation graph and the DL framework runtime. We have evaluated DNNMem on 5 real-world representative models with different hyperparameters under 3 mainstream frameworks (TensorFlow, PyTorch, and MXNet). Our extensive experiments show that DNNMem is effective in estimating GPU memory consumption.

Titanium alloy is a major structural material with excellent high specific strength in aerospace applications. Cubic boron nitride (cBN) is a synthetic wear-resistant material with high hardness, similar to that of diamond, that is used in mechanical cutting and grinding. In addition, the thermal stability of cubic boron nitride particles is much better than that of diamond. In order to further enhance the wear resistance of the Ti6Al4V alloy, the laser cladding (LC) technology characteristics of metallurgical bonding were used to prepare cubic boron nitride/Ti6Al4V and Ni-plated cubic boron nitride/Ti6Al4V composite coatings on Ti6Al4V substrates in this paper. However, in the laser molten pool, it is difficult to retain the raw properties of cubic boron nitride particles under laser radiation. Both composite coatings were analyzed using scanning electron microscopy (SEM), energy dispersive spectroscopy (EDS), X-ray diffraction (XRD), and X-ray photoelectron spectroscopy (XPS). The microstructures and interface bonding between cubic boron nitride particles and the Ti6Al4V matrix were examined using SEM, and the wear resistance and the worn track morphology of the composite coatings were evaluated using the ball-on-disc wear test and step profiler (WTM-2E). The results indicated that the Ni-plated cubic boron nitride/Ti6Al4V composite coating showed fewer thermal defects in comparison with the cubic boron nitride/Ti6Al4V coating. The Ni plating on the surface of cubic boron nitride particles was able to avoid the generation of thermal cracking of the cubic boron nitride particles in the composite coating. The TiN reaction layer was formed between the cubic boron nitride particles and Ti6Al4V matrix, which effectively prevented the further decomposition of the cubic boron nitride particles. The XRD and XPS results confirmed that the TiN reaction layer formed between the cubic boron nitride particles and Ti6Al4V. The Ni plating on the surface of the cubic boron nitride particles was also beneficial for increasing the wear resistance of the composite coating.

The van der Waals ferromagnet Fe

A composite coating composed of intermetallic compounds, Al&ndash, Si alloys, and an oxide ceramic layer was prepared on TA2 substrate by hot-dipping Al&ndash, Si alloy and micro-arc oxidation (MAO) methods. The microstructure and composition distribution of the resulting hot-dipped Al&ndash, Si alloy layer and MAO-caused ceramic layer were studied by scanning electron microscope (SEM) and energy dispersive spectrum (EDS). In addition, the phase composition of the diffusion layer obtained by the Al&ndash, Si alloy hot-dipping procedure was investigated by electron backscattered diffraction (EBSD), and the phase structure of the MAO-treated layer was studied by X-ray diffraction (XRD) analysis and X-ray photoelectron spectroscopy (XPS). The MAO method can make the hot-dipped Al&ndash, Si alloy layer in-situ oxidized to form a ceramic layer. Finally, a three-layer composite coating composed of a diffusion layer formed by the Ti&ndash, Al&ndash, Si interdiffusion, an Al&ndash, Si alloy layer and a ceramic layer was prepared on TA2 substrate. Compared with TA2 substrate, the TA2 sample with a three-layer composite coating has larger friction coefficient and less abrasion loss. The three-layer composite coating can significantly improve the wear resistance of TA2. A technical composite method was developed to the low cost in-situ growth of alumina-based ceramic wear-resistant coatings on TA2 substrate.

The van der Waals ferromagnet Fe5GeTe2 has a Curie temperature TC of about 270 K, which can be raised above room temperature by tuning the Fe deficiency content. To achieve insights into its ferromagnetic exchange, we have studied the critical behavior by measuring the magnetization in bulk Fe5GeTe2 crystal around the ferromagnetic to paramagnetic phase transition. The analysis of the magnetization by employing various techniques including the modified Arrott plot, Kouvel-Fisher plot and critical isotherm analysis achieved a set of reliable critical exponents with TC = 273.7 K, beta = 0.3457, gamma = 1.40617, and delta = 5.021, suggesting a three-dimensional magnetic exchange with the distance decaying as J(r) ~ (r)$^-4.916, which is close to that of a three-dimensional Heisenberg model with long-range magnetic coupling., 22 pages, 2 tables, 6 figures

High strength-to-weight ratio, commendable biocompatibility and excellent corrosion resistance make Ti alloys widely applicable in aerospace, medical and marine industries. However, these alloys suffer from serious biofouling, and may become vulnerable to corrosion attack under some extreme marine conditions. The passivating and biofouling performance of Ti alloys can be attributed to their compact, stable and protective films. This paper comprehensively reviews the passivating and biofouling behavior, as well as their mechanisms, for typical Ti alloys in various marine environments. This review aims to help extend applications of Ti alloys in extremely harsh marine conditions.

A glass-based composite coating incorporating YSZ, corundum, and Ni3Al was prepared on IC10 superalloy with a sputtering Hf-modified Ni3Al bond-coat. Sealing effect of the glass-based top-coat and reactive elemental effect of Hf facilitated the formation of columnar α-Al2O3 at 1050°C. Furthermore, a discontinuous gahnite interlayer formed at the α-Al2O3/top-coat interface due to interfacial reactions. However, because of the discontinuous microstructure of gahnite, the α-Al2O3 suffered durative attack by the active glass matrix, which caused severe internal oxidation of Hf and Al within the bond-coat. In addition, the oxygen barrier ability of the top-coat changed with its microstructure evolution.

Aluminium layers were coated onto the surface of pure titanium using hot-dip aluminising technology, and then the aluminium layers were in situ oxidised to form oxide ceramic coatings, using the micro-arc oxidation (MAO) technique. The microstructure and composition distribution of the hot-dip aluminium coatings and ceramic layers were studied by using scanning electron microscopy and energy-dispersive X-ray spectroscopy. The phase structure of the MAO layers was studied using X-ray diffraction. The surface composition of the MAO layer was studied by X-ray photoelectron spectroscopy. The wear resistance of the pure titanium substrate and the ceramic layers coated on its surface were evaluated by using the ball-on-disc wear method. Therefore, aluminising coatings, which consist of a diffusion layer and a pure aluminium layer, could be formed on pure titanium substrates using the hot-dip aluminising method. The MAO method enabled the in-situ oxidation of hot-dip pure aluminium layers, which subsequently led to the formation of ceramic layers. Moreover, the wear resistance values of the ceramic layers were significantly higher than that of the pure titanium substrate.

The growth of a plasma electrolytic oxide (PEO) coating on Al was widely considered as an ejection of molten alumina on the surface, induced by a localized breakdown of the solid coating. However, evidences presented here indicated no breakdown of the insulating coating in the PEO process, based on the X-ray diffraction, scanning electron microscopy, transmission electron microscopy and energy dispersive spectrometry investigations. There are abundant channels in the coating layer from surface to substrate. However, the SEM observes indicating that the channels in the outer layer with a low-density distribution and a micron-sized diameter, the channels extending into the internal layer with a high-density distribution and a nano-sized diameter. In addition, a continuous thin amorphous alumina layer (AAL) was found between coating and substrate. An intermediate state of the AAL before transition to crystalline, with unusually thick character (∼2.7 μm), is reported for the first time, which presented a typical cross-sectional characteristic of the ionic migration processes. Based on these microstructure observations, a growth mechanism is proposed here to explain the structural features, in which the discharge ignited in the gas-filled channels passing through both the outer layer and the internal layer. The main voltage drop of applying the hole sample loaded on the AAL at a magnitude of ∼1V·nm-1 due to excellent conductivity of discharge plasmas in the channels. Under a high electrical field, the AAL was thickening, ultimately squeezing the coating expanding outward. Abundant moving micro-discharges and localized expansions could result in the growth of the whole coating.

A precursory stack, composed of magnetron-sputtered Al top-layer, microarc-oxidized TiO 2 interlayer and pure Ti substrate as bottom-layer, was used to fabricate oxidation-resistant coatings via reactive synthesis method that involves a vacuum-diffusional annealing at 600 °C and a following pre-oxidation treatment at 900 °C. The compositional and structural evolution from precursory stack to oxidation-resistant coating was investigated by EDS, XRD and SEM, respectively. The anti-oxidation performance of the coating was evaluated by cyclic oxidation test at 900 °C. The results indicate that a Ti–Al gradient structure preliminarily generates at coating/substrate interface for decreased thermal mismatch during vacuum-diffusional annealing, because of the controllable diffusion of Al atoms from top-layer to Ti substrate via porous TiO 2 interlayer. On the other hand, in the process of subsequent pre-oxidation treatment, an Al 2 O 3 rich outer coating is formed accompanied with the consumption of TiO 2 interlayer, while the Ti–Al gradient structure is eventually achieved in the reactively synthesized coating. The oxidation resistance of Ti samples with the coating is about 12 times higher than that of bare Ti, demonstrated by a 50 cycle test.

A hot-dip aluminised layer was prepared on a Mo1 substrate and the obtained layer was oxidised in situ by a micro-arc oxidation (MAO) technique to form an oxide ceramic coating. The obtained hot-dip aluminised layer had a two-layer structure overall: the outer layer was a pure aluminium layer, and the inner layer was a diffusion layer. The hot-dip aluminising temperature directly influenced the structure and phase composition of the obtained diffusion layer. The diffusion layer of the hot-dip aluminised layer at 710 °C was divided into three layers. The inner, intermediate, and outer layers were mainly composed of Al8Mo3, Al4Mo, and Al12Mo phases, respectively. The diffusion layer of the hot-dip aluminised layer at 750 °C was divided into two layers: the inner layer was mainly composed of an Al8Mo3 phase, and the outer layer of Al4Mo. After the hot-dip aluminised layer was treated by MAO, a three-layer coating was obtained. The outer layer was a ceramic one obtained by MAO, which was mainly composed of Al2O3, mullite, and unoxidized aluminium. The middle one was pure aluminium obtained by hot-dip aluminising and without MAO. The inner layer was an Al–Mo diffusion layer.

Ceramic coatings with high emissivity were fabricated on TC4 alloys by micro-arc oxidation technique (MAO) in mixed silicate and phosphate electrolytes with varying KMnO₄ addition. The microstructure, phase and chemical composition were characterized by scanning electron microscope (SEM), X-ray diffraction (XRD), and X-ray photoelectron spectroscopy (XPS), and the infrared emissivity of the MAO coatings was measured in a waveband of 5-20 μm. The results show that the thickness of the coatings increased with the addition of KMnO₄, but the roughness of the coatings first decreased and then increased slightly due to the inhibitory effect of KMnO₄ on Na₂SiO₃ deposition. The main phase composition of the coatings was anatase and rutile TiO₂, amorphous form of SiO₂ and MnO₂. The infrared emissivity value of the coatings strongly depended on KMnO₄ concentration, the coating formed at the concentration of 0.8 g/L KMnO₄ reached the highest and an average of up to 0.87 was observed.

Most recommendation algorithms assume that an account represents a single user, and capture a user’s interest by what he/she has preferred. However, in some applications, e.g., video recommendation on smart TVs, an account is often shared by multiple users who tend to have disparate interests. It poses great challenges for delivering personalized recommendations. In this paper, we propose the concept of profile coherence to measure the coherence of an account’s interests, which is computed as the average similarity between items in the account profile in our implementation. Furthermore, we evaluate the impact of profile coherence on the quality of recommendation lists for coherent and incoherent accounts generated by different variants of item-based collaborative filtering. Experiments conducted on a large-scale watch log on smart TVs conform that the profile coherence indeed impact the quality of recommendation lists in various aspects—accuracy, diversity and popularity.

The cam and tappet are key components of the internal combustion engine valve train system. Due to the alternating high-frequency cyclic of high contact stress, the cam and tappet are prone to abnormal wear and failure. In this work, Cr12W steel was selected as the material of the roller tappet in valve train. Then, it was treated by ion nitriding treatment. The fretting wear tests were carried out under laboratory conditions. The wear mechanisms of nitrided and untreated specimens which rubbed with GCr15 steel were investigated under dry friction and oil lubrication, respectively. It is found out that the predominant wear mechanisms of nitrided specimens under dry friction are slight abrasive wear and adhesive wear, whereas under oil lubrication, it is mainly spalling. The wear mechanisms of untreated specimens under dry friction are identified as adhesive wear and oxidative wear, while it is mainly abrasive wear under oil lubrication.

The mechanical properties of plasma-enhanced magnetron sputtering Si–C–N hard coatings with various compositions are characterized. The effect of chemical composition on the microstructure and properties of coating is investigated. The results show that the microstructure and mechanical properties of Si–C–N coatings are very sensitive to chemical composition. The nanocrystalline/amorphous composite structure is beneficial to the coating's mechanical properties. It also reveals that Si–C–N coating with low Si and high C concentrations has the highest hardness (≥40 GPa) and the best wear property with dry friction coefficient about 0.2.

SiC W /SiC composites without an intergranular glassy phase are prepared by the CVI process. The microstructure and high temperature strength of SiC W /SiC composites are investigated. The high temperature strength of SiC W /SiC composites depends on the interfacial bonding strength between matrix and whiskers. The matrix is effectively strengthened and toughened by the whiskers, and SiC W /SiC composites show a constant strength of 475 ± 32 MPa below 1000 °C. The flexural strength gradually declines from 475 ± 32 MPa to 208 ± 15 MPa at the temperature from 1000 °C to 1500 °C due to the decreased whisker-matrix interfacial bonding strength. The interfacial bonding strength is too low to transfer stress from the matrix to the whiskers, and SiC W /SiC composites show a constant strength of 208 ± 15 MPa at temperatures from 1500 °C to 1800 °C. The fracture mode of SiC W /SiC composites changes from the toughened fracture below 1000 °C to the brittle fracture above 1000 °C.

37.2 vol.% SiCP/SiC and 25.0 vol.% SiCW/SiC composites were prepared by chemical vapor infiltration (CVI) process through depositing SiC matrix in the porous particulate and whisker preforms, respectively. The particulate (or whisker) preforms has two types of pores; one is small pores of several micrometers at inter-particulates (or whiskers) and the other one is large pores of hundreds micrometers at inter-agglomerates. The microstructure and mechanical properties of 37.2 vol.% SiCP/SiC and 25.0 vol.% SiCW/SiC composites were studied. 37.2 vol.% SiCP/SiC (or 25.0 vol.% SiCW/SiC) consisted of the particulate (or whisker) reinforced SiC agglomerates, SiC matrix phase located inter-agglomerates and two types of pores located inter-particulates (or whiskers) and inter-agglomerates. The density, fracture toughness evaluated by SENB method, and flexural strength of 37.2 vol.% SiCP/SiC and 25.0 vol.% SiCW/SiC composites were 2.94 and 2.88 g/cm3, 6.18 and 8.34 MPa m1/2, and 373 and 425 MPa, respectively. The main toughening mechanism was crack deflection and bridging.

A two-process model was developed according to the infiltration-induced structural evolvement in the porous fiber preforms and the physicochemical phenomena involved in the Isothermal Chemical Vapor Infiltration (ICVI) process. The structural evolvement was modeled by representing the void space by a small pore among filaments and a large pore among bundles. The mathematical model was implemented by finite-element method to investigate the effect of the preform structure on the deposition uniformity, processing time and final average density. Calculation results imply that the fiber preform structure plays a dramatic role in determining the deposition uniformity, processing time and final average density. The fiber preforms consisted of filaments and bundles of moderate diameters are preferable by combined consideration of the highest final average density, the relatively low density gradient and the relatively short infiltration time.