Search

Here, we report significant pressure-modulate optoelectronic properties of NiPS3. Upon compression, NiPS3 exhibited a photocurrent increase of five orders of magnitude over the initial value. Interestingly, when NiPS3 was finally decompressed to ambient conditions, the photocurrent could maintain a two-order-of-magnitude enhancement. In addition, the spectral response range was extended to the near-infrared spectral range (up to 1650 nm) under high pressure. Raman and XRD measurements and theoretical calculations revealed significant enhancement in both interlayer and intralayer interactions during compression, leading to a remarkable modulation of the optoelectronic properties. By applying pressure, giant enhancement of photocurrent and tunable spectral response range were achieved in NiPS3, which provides a potential way to modify the optoelectronic properties for materials.

We present NeRFVS, a novel neural radiance fields (NeRF) based method to enable free navigation in a room. NeRF achieves impressive performance in rendering images for novel views similar to the input views while suffering for novel views that are significantly different from the training views. To address this issue, we utilize the holistic priors, including pseudo depth maps and view coverage information, from neural reconstruction to guide the learning of implicit neural representations of 3D indoor scenes. Concretely, an off-the-shelf neural reconstruction method is leveraged to generate a geometry scaffold. Then, two loss functions based on the holistic priors are proposed to improve the learning of NeRF: 1) A robust depth loss that can tolerate the error of the pseudo depth map to guide the geometry learning of NeRF; 2) A variance loss to regularize the variance of implicit neural representations to reduce the geometry and color ambiguity in the learning procedure. These two loss functions are modulated during NeRF optimization according to the view coverage information to reduce the negative influence brought by the view coverage imbalance. Extensive results demonstrate that our NeRFVS outperforms state-of-the-art view synthesis methods quantitatively and qualitatively on indoor scenes, achieving high-fidelity free navigation results., Comment: 10 pages, 7 figures

Simultaneous study of magnetic and electrocatalytic properties of cobaltites under extreme conditions expands understanding of physical and chemical processes proceeding in them with the possibility of their further practical application. Therefore, La0.6Sr0.4CoO3 (LSCO) nanopowders have been synthesized at different annealing temperatures tann = 850, 875, 900 C, and their multifunctional properties have been studied comprehensively. As tann increases, the rhombohedral perovskite structure of the LSCO becomes more single-phase, whereas average particle size and dispersion grow. Co ions are in mixed valence states, including major Co3+ and Co4+ components. It has been found that the LSCO-900 shows two main Curie temperatures, TC1 and TC2, associated with a particle size distribution. As an external hydrostatic pressure P increases, average and increase from 253 and 175 K under ambient pressure to 268 and 180 K under P = 0.8 GPa, respectively. At the same time, the antiferromagnetic temperature TAFM and blocking temperature TB also increase from 145 and 169 K to 158 and 170 K, respectively. The increment of for the smaller and bigger particles is sufficiently high and equals 10 and 13 K/GPa, respectively. The magnetocaloric effect in the LSCO-900 nanopowder is relatively weak but with an extremely wide peak > 50 K that makes this composition interesting to be used as one of the components of the composite expanding its working temperature window. Moreover, all LSCO samples showed excellent electrocatalytic performance for the overall water splitting (OER) process (overpotentials only 265-285 mV at a current density of 10 mA/cm2) with minimal values for LSCO-900. Based on the XPS data, it was found that the formation of a dense amorphous layer on the surface of the particles ensures high stability as a catalyst (at least 24 h) during electrolysis in 1 M KOH electrolyte., Comment: 40 pages, 19 figures, 9 tables, 90 references, article

Supplementary Material 1

Mill scale scrap, which contains vast amounts of valuable metals, is a solid waste produced in the iron and steel industry. Conventional mill scale scrap treatment methods for metal extraction are characterized by high energy consumption and low value addition. In this study, co-treatment of mill scale scrap and manganese ore via the oxidization roasting-magnetic separation process was investigated for the synchronous preparation of higher-value materials and recovery of valuable metals. Thermodynamic and magnetism analyses indicated that a higher temperature (>1100 °C) and a MnO2/Fe2O3 molar ratio of 0.75–1 are essential for the preparation of manganese ferrite. The experimental validation revealed that soft magnetic manganese ferrite powders with a purity of 97.5 wt% were obtained when the test was conducted at 1300 °C for 120 min, followed by a two-stage grinding and magnetic separation process; the corresponding yield and the Mn and Fe recoveries were 78.99 wt%, 86.14 wt%, and 84.60 wt%, respectively. During the oxidization process, [Fe2+]O was initially oxidized to the anti-form spinel-type structure of [Fe3+][Fe2+Fe3+]O4, and thereafter, it reacted with the decomposition product of [Mn3+][Mn2+Mn3+]O4 to form a hybrid spinel-type structure [Me2+xMe3+l-x][Me2+1-xMe3+1+x]O4 (Me refers to Mn and Fe) via the Mn2+/Fe2+/Mn3+/Fe3+ ions exchange at the tetrahedral and octahedral sites. Moreover, the as-purified ferrite can be used as an ingredient for the preparation of high-performance MnZn ferrite.

Materials with unique structures can exhibit different properties and are widely studied in the preparation of new materials. Herein we reported a hydrothermal method to fabricate a layered nickel silicate/reduced graphene oxide (NiSiO/RGO) nanocomposite with an interesting dandelion-like structure. The morphology, composition, and electrochemical performance of RGO, NiSiO, and NiSiO/RGO were comparatively investigated in the current work. The results showed that the NiSiO/RGO nanocomposite has a dandelion-like hollow core–shell structure and shows good electrochemical performance. Compared with NiSiO, the original discharge capacity of NiSiO/RGO increased from 1291.6 mA h g−1 to 1653.9 mA h g−1; meanwhile, the reversible specific capacity of NiSiO/RGO increased from 649.6 mA h g−1 to 691.4 mA h g−1 after testing at a current density of 100 mA g−1 for 100 cycles. Moreover, the prepared NiSiO/RGO maintained a coulombic efficiency of about 97% after the initial charging and discharging cycle. This unique hollow dandelion-like structure enhanced the electrical conductivity and further resulted in lower diffusion resistance and higher reversible capacity. This work demonstrated that the layered NiSiO/RGO with an interesting dandelion-like structure can act as an alternative anode material for lithium-ion batteries.

We consider the problem of domain adaptation in LiDAR-based 3D object detection. Towards this, we propose a simple yet effective training strategy called Gradual Batch Alternation that can adapt from a large labeled source domain to an insufficiently labeled target domain. The idea is to initiate the training with the batch of samples from the source and target domain data in an alternate fashion, but then gradually reduce the amount of the source domain data over time as the training progresses. This way the model slowly shifts towards the target domain and eventually better adapt to it. The domain adaptation experiments for 3D object detection on four benchmark autonomous driving datasets, namely ONCE, PandaSet, Waymo, and nuScenes, demonstrate significant performance gains over prior arts and strong baselines.

Data is a fundamental building block for LiDAR perception systems. Unfortunately, real-world data collection and annotation is extremely costly & laborious. Recently, real data based LiDAR simulators have shown tremendous potential to complement real data, due to their scalability and high-fidelity compared to graphics engine based methods. Before simulation can be deployed in the real-world, two shortcomings need to be addressed. First, existing methods usually generate data which are more noisy and complete than the real point clouds, due to 3D reconstruction error and pure geometry-based raycasting method. Second, prior works on simulation for object detection focus solely on rigid objects, like cars, but VRUs, like pedestrians, are important road participants. To tackle the first challenge, we propose FPA raycasting and surrogate model raydrop. FPA enables the simulation of both point cloud coordinates and sensor features, while taking into account reconstruction noise. The ray-wise surrogate raydrop model mimics the physical properties of LiDAR's laser receiver to determine whether a simulated point would be recorded by a real LiDAR. With minimal training data, the surrogate model can generalize to different geographies and scenes, closing the domain gap between raycasted and real point clouds. To tackle the simulation of deformable VRU simulation, we employ SMPL dataset to provide a pedestrian simulation baseline and compare the domain gap between CAD and reconstructed objects. Applying our pipeline to perform novel sensor synthesis, results show that object detection models trained by simulation data can achieve similar result as the real data trained model.

Hemorrhage in the basal ganglia is a common type of intracerebral hemorrhage and has high mortality and poor prognosis. In our study, we aimed to evaluate surgical outcomes and functional recovery after evacuation of hematoma using either craniotomy or endoscopy.We analyzed retrospective data from 58 patients with basal ganglia hemorrhage who were treated with hematoma evacuation using either craniotomy or endoscopy. Magnetic resonance imaging and a navigation system were used for calculating hematoma volume and for navigation during surgery. Clinical information and surgical outcomes were recorded. At 6-month follow-up, the recovery of neurologic function and the results of the Aphasia Battery of Chinese test were assessed.The endoscopy group showed lower intraoperative blood loss (75.36 ± 45.56 vs. 462.67 ± 120.08 mL, P 0.001), shorter operation time (1.59 ± 0.30 vs. 4.17 ± 0.86 hours, P0.001), and a higher hematoma clearance rate (0.93% ± 0.05% vs. 0.88% ± 0.13%, P = 0.04) than the craniotomy group, respectively. No significant differences in mortality were identified, but a trend toward lower mortality in the endoscopy group was apparent (7.14% in the endoscopy group vs. 16.67% in the craniotomy group, P = 0.43). Assessment of neurologic recovery indicated significant differences in the modified Rankin Scale grades between the 2 groups (χEvacuation by endoscopy had better surgical outcomes, recovery of neurologic function, and aphasia recovery than evacuation by craniotomy. It appears that endoscopy is the surgical treatment of choice for middle-aged and elderly patients with a basal ganglia hemorrhage volume of35 mL.

Dyes are intensively used in textile and dyeing industries, and substantial volumes of organic wastewater with residual dye require treatment before discharges to public waterways. Flotation separation is an efficient and widely used method for the treatment of massive organic dye wastewaters. The key scientific problems for dye flotation separation lie in the mineralization transformation of dissolved dye to tangible flocs. In this work, a high-efficiency removal of hazardous azo dye Congo red (CR) from simulated wastewaters via metal ions chelation flocculation followed by flotation separation was proposed. It's demonstrated that CR can be chelated by the trivalent metal ions, including Al(III), Fe(III), and its mixture to form hydrophobic flocs, and then the flocs were efficiently removed via flotation in a microbubble column. The effects of chelation flocculation and flotation separation conditions on the removal efficiencies of CR, COD, and chromaticity from CR simulated wastewaters were optimized. Chelation effect of CR by trivalent metal ions was in this order: Al(III)+Fe(III)>Fe(III)>Al(III). The chelation mechanism suggested that CR molecules gradually changed from hydrazones to electronegative azo with the increase of pH to 6–7, and electrostatic attraction between the Al3(OH)45+ or Fe(OH)2+ with the CR was favorable for the chelation reaction, in which the metal ions chelated with N atoms on naphthalene ring and amino groups of CR. Over 99% CR was removed under the optimal chelation and flotation conditions: chelation by composite Al(III)/Fe(III) with a concentration of 25 mg/L at pH of 7 for 25min; followed by flotation with SDS concentration of 20 mg/L and air flow rate of 50 mL/min for 20min. Under this condition, the COD and chromaticity removal efficiency were over 96% and 98%, respectively, and the turbidity was lower than 0.1 NTU, meeting the water discharge requirement. Eventually, resourceful utilization of flotation sludge via calcination was conducted to prepare Al–Fe spinel refractory material.

he success of supervised LiDAR perception methods relies on the availability of large sets of labeled point cloud data, for which the labeling process is costly and time consuming. Given unpaired LiDAR datasets of similar sizes from two domains, with one (source) containing task-specific labels e.g. 3D bounding boxes for all frames, but only a small percentage of frames being labeled in the other (target) domain, it is challenging to train a model that generalizes well on validation data from the target domain. In this paper we propose a novel LiDAR few-shot domain adaptation architecture and training strategy to address this challenge. Our method is based on adapting a task-specific network (3D object detector) to work within the CycleGAN framework modified to operate with LiDAR features, and on the joint end-to-end training of generators, discriminators, and task-specific layers. To overcome nonconvergence issues we propose a training strategy that introduces a mechanism to delay the joint learning between the generators/discriminators and the task-specific network by allowing them to start learning independently, while slowly introducing joint learning as they converge, hence avoiding instability during the early stages of the training. Our proposed integrated architecture enables a direct way to evaluate the performance of the model instead of feeding pre-computed generated data into a separate pretrained model. We include an experimental section where we evaluate our proposed architecture on the publicly available KITTI and Nuscenes datasets, as well as on our own labeled dataset. We present useful mean average precision plots that illustrate the benefits of our domain adaptation architecture as a function of number of labeled target domain frames.

Semantic segmentation of point clouds is a key component of scene understanding for robotics and autonomous driving. In this paper, we introduce TORNADO-Net - a neural network for 3D LiDAR point cloud semantic segmentation. We incorporate a multi-view (bird-eye and range) projection feature extraction with an encoder-decoder ResNet architecture with a novel diamond context block. Current projection-based methods do not take into account that neighboring points usually belong to the same class. To better utilize this local neighbourhood information and reduce noisy predictions, we introduce a combination of Total Variation, Lovasz-Softmax, and Weighted Cross-Entropy losses. We also take advantage of the fact that the LiDAR data encompasses 360 degrees field of view and uses circular padding. We demonstrate state-of-the-art results on the SemanticKITTI dataset and also provide thorough quantitative evaluations and ablation results., 10 pages, 4 figures

Electrolytic manganese anodic slime (EMAS) with high contents of manganese and lead is a well-known refractory secondary resource because of its complex constituents and structure. Traditional stockpiling treatment of EMAS not only causes resource waste, but also leads to long-term environmental pollution. In this study, SO2 roasting followed by acid leaching process was put forward to extract and separate manganese and lead from the EMAS. Thermodynamic analysis showed that a coexistence area of MnSO4 and PbSO4 was present in both the roasting and leaching processes. During the roasting process, manganese and lead oxides were sulfated to MnSO4 and PbSO4 by SO2. Then, MnSO4 and PbSO4 could be separated in the following acid-leaching process owing to their different solubilities in acid solution. The leaching efficiencies of 92.5 wt.% for Mn and only 3.21 wt.% for Pb were obtained under optimal conditions. This study provides an alternative approach to utilize the EMAS.

Investigations on the MnO2-Fe2O3 system roasted in air has been reported in our previous work. This study further investigated the MnO2-Fe2O3 system roasted under CO-CO2 atmosphere. Extensive investigations were concentrated on the reduction of simplex iron oxides or manganese oxides, and little attention were paid on the reduction of MnO2-Fe2O3 system regarding to interactive reactions between them. In this work, it was found that spinel-type MnxFe3−xO4 with high magnetism formed easily under CO-CO2 atmosphere. The reduction and thermodynamic analyses of pure MnFe2O4 were also researched to better understand the reduction behaviors of MnO2-Fe2O3 system. Phase study showed that a series of Mn-Fe composite oxides, including MnxFe3−xO4 and (MnO)y(FeO)1−y, generated during the reduction of MnO2-Fe2O3 system. MnxFe3−xO4 was readily generated under CO content of 2.5–25 vol% at 1000 °C. With further increase of CO content, MnxFe3−xO4 was reduced to (MnO)y(FeO)1−y and then to MnO and metallic iron. Reduction of manganese oxides, iron oxides and manganese ferrites happened concurrently during the reduction of MnO2-Fe2O3 system. And the reduction of MnO2, Fe2O3, Fe3O4 and MnFe2O4 were compared by TG and thermodynamic analyses. In addition, the morphology evolution and magnetism change of the MnO2-Fe2O3 system reduced under different CO contents were also studied.

Lead (Pb) ions are chronically detected in soil, underground and natural water, there is a need for low-cost in situ remediation techniques. A novel mineral based magnetic adsorbent, consisting of natural vanadium, titanium-bearing magnetite particles coated by humic acid (abbr. VTM-HA), was developed for removing Pb(II) from wastewater. In this study, adsorption-desorption characteristics and mechanisms of Pb (II) on the VTM-HA adsorbent were investigated. The regeneration and reuse feasibility of the VTM-HA adsorbent were also conducted. The adsorption tests further verified that Pb(II) was removed rapidly and efficiently by the VTM-HA adsorbent. Moreover, Pb (II) adsorbed on the VTM-HA adsorbent could be easily eluted using a small amount of acidic eluent, and the desorption of Pb (II) could reach 99.3 wt% in 0.1 M HNO3 (pH = 1). After five adsorption-desorption regeneration cycles, the magnetism and Pb(II) adsorption capacity of the regenerated VTM-HA adsorbent almost kept unchanged. The results indicated that the VTM-HA adsorbent had excellent adsorption capacity and regenerative ability, which could be used as ideal adsorbents for removing Pb (II) from acid wastewater in commercial application. The mechanisms of Pb (II) adsorption and desorption were explored by FTIR, XPS analysis and DFT calculation.

There are abundant ferruginous manganese ores (abbr. Fe-Mn ores) in many parts of the world. For example, almost half of the manganese ore resources in India are ferruginous and more than 73% of the manganese ores in China belong to Fe-Mn ores with a low Mn/Fe mass ratio ( Numerous approaches are reported to realize the selective extraction and separation of Mn and Fe from the Fe-Mn ores. The extraction and separation technologies cover physical beneficiation, chemical beneficiation (hydrometallurgy and pyrometallurgy) and physico-chemical combined processes. This present work reviews the technical principles, parameters and recovery efficiencies of diverse processes on the aspect of selective extraction and separation of Mn and Fe from Fe-Mn ores. This review can provide guidance for selecting appropriate methods to exploit Fe-Mn ores or other secondary resources containing Fe and Mn oxides. The authors also put forward a new route to produce manganese ferrite materials using Fe-Mn ores as the raw materials.

Purification and recovery of strategic W/Mo/Re from wastewaters face a dual concern for resource recovery and environmental pollution. In this work, a novel manganese ferrite nanoparticles (MFO-NPs) was prepared from the low-grade manganese ore via pyrometallurgical transformation followed by nano sanding. The effects of pH, initial concentration, and reaction time on adsorption behaviors of simplex and mixed W/Mo/Re by MFO-NPs were investigated. Single anion adsorption indicated that molybdate and tungstate adsorption were consistent with the Langmuir isotherm model with the highest adsorption capacities of 41.985 mg/g and 85.61 mg/g, respectively; whereas rhenate was not adsorped. Selective separation indicated that W was firstly separated by MFO-NPs adsorption at pH of 5 or acid concentration of 0.5 mol/L. After that, Mo was further adsorped by MFO-NPs at pH of 2-4 while Re was kept in the raffinate. The possible selective adsorption mechanism was based on the difference of electrostatic adsorption, polar adsorption, and complexing reaction between MFO-NPs and W/Mo/Re anions, which had various ionic forms under different pH values. Compared with the MFO-NPs prepared by wet chemistry methods, the novel MFO-NPs proposed in this work not only showed satisfactory adsorption capacity but had huge potential for the selective separation of W, Mo, and Re.

Bentonite is the most common binder for the vanadium, titanium-bearing magnetite (VTM) pellet production. However, inorganic bentonite will reduce the iron grade of the finished products because ab...

Calcium stannate (Ca2SnO4) is a common dielectric ceramic that is generally prepared by a high-temperature solid-state method in an air atmosphere with a roasting temperature of more than 1300 °C. In this study, Ca2SnO4 was much more easily synthesized in a CO-CO2 atmosphere at a relatively low temperature of less than 1000 °C for 30 min. We comparatively investigated the formation behavior and microwave dielectric properties of Ca2SnO4 synthesized from tin dioxide (SnO2) and calcium carbonate (CaCO3) by a solid-state method in air and CO-CO2 atmospheres using X-ray diffraction, scanning electron microscopy, inductively coupled plasma atomic emission spectroscopy, and vector network analysis. The formation behavior of calcium stannate indicated that the reactions between SnO2 and CaCO3 in CO-CO2 and air atmospheres were both controlled by three-dimensional diffusion. In a CO-CO2 atmosphere, the reaction had a higher reaction rate constant (k) and lower apparent activation energy (E). In addition, Ca2SnO4 was synthesized by a low-temperature solid-state method in a CO-CO2 atmosphere. The results indicated that Ca2SnO4 ceramics had a higher dielectric constant and lower dielectric tangent loss than those synthesized by the high-temperature solid-state method in air. However, the synthesis temperature and time were reduced by more than 300 °C and 8 h, respectively.

Molybdenum (Mo) is a strategic metal element, and recovery of Mo from low concentration Mo-containing solution is significant to alleviate the Mo resources shortage. In this study, the treatment of low-concentration Mo-containing solution by Fe3+ addition was investigated by batch experiment, and the effects of main parameters including molar ratio of n(Mo):n(Fe), pH value, initial concentrations, and reaction time on the recovery of Mo were studied. The results showed that Mo is precipitated rapidly, with a Mo recovery of 100% in 5 min when the pH range is 4–6 and molar ratio of n(Mo):n(Fe) is greater than 1:3, indicating the reaction between Mo and Fe readily approaches chemical equilibrium. Moreover, the initial concentration of Mo in the solution has little influence on the recovery of molybdenum. This technology has potential to be applied for the enrichment and recycling of molybdenum, while at the same time meeting the industrial discharge standard for the effluent.

Autonomous robotic systems and self driving cars rely on accurate perception of their surroundings as the safety of the passengers and pedestrians is the top priority. Semantic segmentation is one the essential components of environmental perception that provides semantic information of the scene. Recently, several methods have been introduced for 3D LiDAR semantic segmentation. While, they can lead to improved performance, they are either afflicted by high computational complexity, therefore are inefficient, or lack fine details of smaller instances. To alleviate this problem, we propose AF2-S3Net, an end-to-end encoder-decoder CNN network for 3D LiDAR semantic segmentation. We present a novel multi-branch attentive feature fusion module in the encoder and a unique adaptive feature selection module with feature map re-weighting in the decoder. Our AF2-S3Net fuses the voxel based learning and point-based learning into a single framework to effectively process the large 3D scene. Our experimental results show that the proposed method outperforms the state-of-the-art approaches on the large-scale SemanticKITTI benchmark, ranking 1st on the competitive public leaderboard competition upon publication., Comment: 10 pages, 6 figures, 4 tables

Electrode polarization potential is an important factor of cell voltage during electrowinning, which can affect the total energy consumption of the zinc hydrometallurgy process. In this work, we creatively use a metal-free oxygen evolution catalyst of water electrolysis to the hydrometallurgy system. An amino-rich hierarchical-network carbon(Amino-HNC) electrocatalyst was prepared by a simple two-step method of amino-assisted polymerization and carbonization process. The surface morphology, phase composition, and electrochemical properties were characterized. The results show that Amino-HNC is an electrocatalyst for oxygen evolution reaction with an amino-rich network structure, and the oxygen evolution overpotential in 0.5 mol L−1 H2SO4 electrolyte is 389 mV (@ 10 mA cm−2). Meanwhile, the prefabricated catalyst is directly used as an anode in the zinc electrowinning system (50 g L−1 Zn2+ + 150 g L−1 H2SO4) to measure the cell voltage. Compare with the traditional pure lead anode, the cell voltage decreased in a short time during the zinc electrowinning process.

Polymetallic ferruginous manganese ores contain a variety of valuable metals, it is significant to realize the efficient utilization. Spinel-type composite ferrites with strong magnetism can be prepared from the polymetallic ferruginous manganese ores, while the impurities in the ores have certain impact on the properties of the ferrites. This study was mainly concentrated on the behavior of Al during the formation process of ferrites. It has been found that during the oxidation roasting process, Al element has occupied the octahedral gap (sit B) in spinel structure of ferrites, which reduced the magnetism property of ferrite. However, SiO2 can improve the magnetism properties of Al-bearing manganese ferrites owing to that SiO2 was able to transfer Al out of B sit of spinel via the chemical driving force generated by the formation of corresponding silicates from the SiO2 and Al oxides. This work demonstrated that Al impurity can be regulated by adding SiO2.

Semantic Segmentation is a crucial component in the perception systems of many applications, such as robotics and autonomous driving that rely on accurate environmental perception and understanding. In literature, several approaches are introduced to attempt LiDAR semantic segmentation task, such as projection-based (range-view or birds-eye-view), and voxel-based approaches. However, they either abandon the valuable 3D topology and geometric relations and suffer from information loss introduced in the projection process or are inefficient. Therefore, there is a need for accurate models capable of processing the 3D driving-scene point cloud in 3D space. In this paper, we propose S3Net, a novel convolutional neural network for LiDAR point cloud semantic segmentation. It adopts an encoder-decoder backbone that consists of Sparse Intra-channel Attention Module (SIntraAM), and Sparse Inter-channel Attention Module (SInterAM) to emphasize the fine details of both within each feature map and among nearby feature maps. To extract the global contexts in deeper layers, we introduce Sparse Residual Tower based upon sparse convolution that suits varying sparsity of LiDAR point cloud. In addition, geo-aware anisotrophic loss is leveraged to emphasize the semantic boundaries and penalize the noise within each predicted regions, leading to a robust prediction. Our experimental results show that the proposed method leads to a large improvement (12\%) compared to its baseline counterpart (MinkNet42 \cite{choy20194d}) on SemanticKITTI \cite{DBLP:conf/iccv/BehleyGMQBSG19} test set and achieves state-of-the-art mIoU accuracy of semantic segmentation approaches.

A fast and accurate panoptic segmentation system for LiDAR point clouds is crucial for autonomous driving vehicles to understand the surrounding objects and scenes. Existing approaches usually rely on proposals or clustering to segment foreground instances. As a result, they struggle to achieve real-time performance. In this paper, we propose a novel real-time end-to-end panoptic segmentation network for LiDAR point clouds, called CPSeg. In particular, CPSeg comprises a shared encoder, a dual-decoder, and a cluster-free instance segmentation head, which is able to dynamically pillarize foreground points according to the learned embedding. Then, it acquires instance labels by finding connected pillars with a pairwise embedding comparison. Thus, the conventional proposal-based or clustering-based instance segmentation is transformed into a binary segmentation problem on the pairwise embedding comparison matrix. To help the network regress instance embedding, a fast and deterministic depth completion algorithm is proposed to calculate the surface normal of each point cloud in real-time. The proposed method is benchmarked on two large-scale autonomous driving datasets: SemanticKITTI and nuScenes. Notably, extensive experimental results show that CPSeg achieves state-of-the-art results among real-time approaches on both datasets., Comment: Accepted at ICRA 2023

Oily ferroalloy scraps generated from machinery honing enterprises are typical hazardous municipal materials that release benzene-series volatile organic compounds (VOC), which endanger human physical and mental health. Therefore, harmless treatment and resource reuse for these hazardous materials is urgent. In this study, the VOC emission, and pyrolysis and de-oiling behaviors of oily honing scrap was first characterized to evaluate the environmental risks. Smelting separation was then proposed to economically and eco-friendly recover valuable metals from the de-oiled ferroalloy scraps. The thermodynamics of Al2O3-SiO2 and CaO-Al2O3-SiO2 systems was calculated to optimize the slag formation. Results showed that after de-oiling and smelting with CaO addition, the hazardous VOC are removed, and the valuable metals are recovered in ferroalloy state. Under optimum conditions, a crude Fe-Mo-Cu alloy with Fe, Mo and Cu recoveries of 98.5 wt%, 97.9 wt%, and 98.4 wt% were obtained. In addition, the slag containing few toxic elements and VOC can be used for silicate cement production. Pyrolysis, de-oiling behaviors and mechanism for slagging and growth of Fe-Mo-Cu alloy during smelting were discussed via various testing techniques, and leaching toxicity of the cleaned slag was also characterized in this work. This process is also applicative to recover metals from various honing ferroalloy scraps.

In object detection for autonomous driving and robotic applications, conventional RGB cameras often fail to sense objects under extreme illumination conditions and on texture-less surfaces, while LIDAR sensors often fail to sense small or thin objects located far from the sensor. For these reasons, an intuitive and obvious choice for perception system designers is to install multiple sensors of different modalities to increase (in theory) the detection robustness. In this paper we focus on the analysis of an object detector that performs early fusion of RGB images and LIDAR 3D points. Our goal is to go beyond the intuition of simply adding more sensor modalities to improve performance, and instead analyze, quantify, and understand the performance differences, strengths and weaknesses of the object detector under three different modalities: 1) RGB-only, 2) LIDAR-only, and 3) Early fusion (RGB and LIDAR), and under two key scene variables: 1) Distance of objects from the sensor (density), and 2) Illumination (Darkness). We also propose methodologies to generate 2D weak semantic training masks, and a methodology to evaluate the object detection performance separately at different distance ranges, which provides a more reliable detection performance measure and correlates well with object LIDAR point density.

Benzohydroxamic acid (BHA) is intensively used as collector in the mineral processing industry, and thus tremendous hazardous beneficiation wastewater containing BHA organic contaminant needs urgent treatment. In this work, an efficient removal of BHA organic contaminants from various industrial beneficiation wastewaters via the chelation, flocculation, and precipitation flotation process was studied. The removal efficiency of BHA, COD and turbidity of the purified water, and the chelation mechanism were investigated by optical microscope, SEM, FTIR, XPS, Zeta potential, and thermodynamic analyses. The results showed that organic BHA contaminant can be chelated firstly by Cu(II) to form hydrophobic BHA-Cu precipitate effectively via the “O-O” chelation and a five-membered chelation ring was formed under weak alkaline conditions. The generated precipitate was then successfully removed using CTAB as surfactant and NP-40 as frother in a microbubble flotation column. Over 94% BHA was removed under the optimal chelation, flocculation and flotation conditions: chelation with Cu(II) concentration of 30 mg/L and pH of 8 for 30 min; flocculation with CTAB concentration of 0.2 mg/L for 20 min; followed by flotation with NP-40 frother of 0.6 mg /L and air flow rate of 100 mL/min for over 40 min. Under this condition, 83.91% COD was removed, and the residuel COD and turbidity of the purified water were lower than 25 mg/L and 0.1 NTU, respectively, which met the wastewater discharge requirement. It’s demonstrated that the precipitate flotation can be employed as an effective method to purify industrial wastewater containing BHA. In addition, the interaction mechanism of Cu(II) and BHA under different pH values was also discussed in this work.

Hazardous metal pollution became a severe environmental issue in China. An efficient precipitation-flotation process was developed to achieve fast removal for metal-ions from wastewater. Structure and strength of precipitate particles/flocs significantly influence the flotation removal of metal-ions. Formation and growth-evolution of precipitate flocs in precipitate flotation were studied by stage analysis of precipitate particles-formation, flocs-regulation and flotation separation. The results demonstrate that early formed precipitates MHA(humics-metal complexing particles) have small particle size, high fractal dimension, low strength and recovery factor. The addition of Fe3+ and CTAB(cetyl trimethyl ammonium bromide) reagents make the precipitate particles aggregated to flocs(MHA-Fe, MHA-Fe-CTAB) much more large, loose, coarse, and small-density. The final generated MHA-Fe-CTAB flocs are hard to be broken up, easy to be recovered and efficient to be separated by flotation process. The flotation removal of MHA-Fe-CTAB flocs is clearly higher than that of MHA or MHA-Fe. The flotation results of MHA-Fe-CTAB are as follows: flotation removal of 98.7 ± 0.40%–99.9 ± 0.10%, residual TOC of 0.96 ± 0.38–1.35 ± 0.41 mg/L and turbidity of 0.44 ± 0.09–0.63 ± 0.16 NTU. Introducing Fe3+ and CTAB reagents into flotation solution contributes to the growth-evolution of precipitate flocs, which could intensify the metal-ions removal via precipitate flotation process and result in more ideal purification indexes for metal-containing wastewater.

Consolidation of the ferruginous manganese ore fines is of vital importance for manganese extraction in the smelting process. The main impurities in the Fe–Mn ores are Ca and Si constituents and they have a significant effect on the consolidation of iron and manganese oxides. In this study, characterization on the behaviors of Ca and Si constituents during the roasting process of Fe–Mn ores were investigated by thermodynamic, fusion temperatures, and phase identification analyses. The results showed that the Ca and Si constituents can decrease the fusion temperature of the MnO2–Fe2O3 binary system dramatically due to the formation of low-melting complicate silicates from the CaO, SiO2, and partial manganese and iron oxides. The high-melting solid manganese ferrites phase generated from the remaining manganese and iron oxides were bonded and solidified by the liquid phase. This provided favorable mineralization conditions for the consolidation of Fe–Mn ore.

Blast furnace slag (BFS) is a typical solid waste generated in the steel production. Most of previous investigations have reported that the BFS has been used as traditional construction materials (cement, roadbed filling, concrete). In this study, a novel use route for BFS was proposed to prepare high-temperature composite phase change materials (C-PCMs) for thermal energy storage. Three typical inorganic PCMs (NaNO3, Al and Na2SO4 with different operating temperature) were blended with the pre-ground BFS to fabricate BFS-based C-PCMs by means of a mixing and sintering process. The results showed that NaNO3 had excellent chemical compatibility with BFS and the prepared C-PCMs had perfect phase change performance. The enthalpies of NaNO3/BFS C-PCMs was 65.53 J/g with melting point 300.5 °C and the super-cooling was only 0.1 °C. Furthermore, the NaNO3/BFS C-PCMs could retain good thermal reliability after 100 thermal cycles, which presented a potential application in the thermal energy storage system. In addition, the morphological structure, thermal reliability and heat transfer property of the NaNO3/BFS C-PCMs were characterised by using SEM, TGA and TG-DSC.

Magnetic reduction roasting followed by magnetic separation process is reported as a simple route to realize separation of Mn and Fe from ferruginous manganese ores (Fe-Mn ores). However, the separation and recovery of Mn and Fe oxides are not very effective. This work clarified the underlying reason for the poor separation and also proposed some suggestions for the magnetic reduction process. In this work, the effect of temperature on the magnetic reduction roasting – magnetic separation of Fe-Mn ore was investigated firstly. Then the reduction behaviors of MnO2-Fe2O3 system and MnO2-Fe2O3-10 wt.%SiO2 system under 10 vol.% CO-90 vol.% CO2 at 600–1000 °C were investigated by XRD, XPS, SEM-EDS, VSM, DSC and thermodynamics analyses. Reduction and separation tests showed that higher reduction temperature was beneficial to the recovery of iron while it’s not in favor of the recovery of manganese when the temperature was over 800 °C. The formation of composite oxide MnxFe3−xO4 with strong magnetism between the interface of the MnO2 and Fe2O3 particles leaded to the poor separation of iron and manganese. In addition, the formation mechanism of MnxFe3−xO4 from MnO2 and Fe2O3 as well as the interface reaction reduced under 10 vol.% CO was discussed in this study. Finally, some suggestions were recommended for the magnetic reduction roasting for utilizing the Fe-Mn ores effectively.

Carbonization of pure MnO2 by CH4-H2 gas mixture with various CH4/(CH4 + H2) molar ratios in the temperature range of 900–1250 °C had been performed in our previous work. This study further investigated the carbonization and alloying process of MnO2 by CH4-H2 gas mixture in the presence of Fe2O3 at 900–1300 °C using XRD, optical microstructure, SEM-EDS and phase diagram analyses. Carbonization test of pure Fe2O3 was also conducted to understand the effect of Fe2O3 on the carbonization of MnO2 more explicitly. It was found that Fe2O3 could be carbonized to Fe3C at 900–1150 °C. The Fe2O3 briquettes began to melt at 1100 °C and a large amount of Fe-C melts appeared as the temperature went up to 1150 °C. When the MnO2 + 25 wt%Fe2O3 briquettes were reduced at 900–1000 °C in CH4-H2 gas mixture, the Fe2O3 was preferentially carbonized to Fe-C phase while the MnO2 was only reduced to MnO. The formation of Fe-C phase was beneficial to the carbonization of MnO since the carbonization temperature of MnO can decrease from 1100 °C to 1000 °C. The carbonization degree of MnO was dramatically accelerated over 1100 °C due to the transformation of Fe-C phase from solid state to melts. Morphology analysis indicated that the MnO core was wrapped by the shell of Fe-C melts. The alloying process of Mn-Fe-C was conducted through the elements migration of Fe and Mn between the core and shell. At 1300 °C, the Mn-Fe-C alloy turned to melt and aggregated since the highest melting temperature of the Mn-C-25 wt% Fe system was only 1274 °C. Finally, a ferromanganese crude alloy with 68–72 wt% Mn, 20–25 wt% Fe and 6.0–8.0 wt% C was obtained when the MnO2 + 25 wt% Fe2O3 briquettes were carbonized by CH4-H2 gas mixture at 1300 °C for 60 min.

Solid-state reaction method is a common and effective technique to synthesize ferrites. This work investigated the phase transformation of MnO 2 and Fe 2 O 3 system roasted at 500–1400 °C in air atmosphere to understand the formation process of manganese ferrite. The results showed that the formation of manganese ferrite (Mn x Fe 3− x O 4 ) was derived from the reaction between Fe 2 O 3 and Mn 3 O 4 (the decomposition product of MnO 2 ). Below 900 °C, MnO 2 firstly decomposed to Mn 2 O 3 and then to Mn 3 O 4 , and Fe 2 O 3 was seldom reacted with Mn 2 O 3 and Mn 3 O 4 . When the temperature went up to 1000 °C, Fe 2 O 3 easily reacted with Mn 3 O 4 to generate manganese ferrite. The reaction degree was enhanced dramatically with the rising of temperature. Moreover, the x value in the Mn x Fe 3− x O 4 increased from 0 to 1 from 900 °C to 1400 °C. In other words, the higher the temperature was, the closer the Mn x Fe 3− x O 4 was to MnFe 2 O 4 . Thermodynamic analysis of MnO 2 -Fe 2 O 3 system under different O 2 partial pressures was carried out to further explain the formation mechanism.

The proven global tin reserves were reported to be approximately 4.7 million tons (Mts) in 2016, and among these resources, only approximately 2.2 Mts can be recovered economically. The original tin deposits will be exhausted in several years, therefore, tin-bearing secondary resources, such as tin alloy, tin anode slime, e-wastes, tin slag and tin-bearing tailings, will become the primary source from which tin can be extracted. Many investigations have been conducted on the recovery of tin from these tin-bearing materials. However, the separation and recovery approaches of tin vary significantly, since the content and phase compositions of tin are totally different in these secondary resources. This paper reviews these methods of extracting and separating tin from different kinds of tin-bearing secondary resources.

A novel method for Na 2 SnO 3 preparation from SnO 2 and Na 2 CO 3 roasted under CO-CO 2 atmosphere in a solid state has been successfully developed. In this study, the formation kinetics of Na 2 SnO 3 was investigated. The formation of Na 2 SnO 3 was controlled by the first-order interfacial chemical reaction at 800 °C under 10 vol% CO content; as the temperature increased to 850 °C–900 °C, it was controlled by the three-dimension (Ginstling and Brounshtein) diffusion through the product layer at the early reaction stage. The rate constant increased with the increasing of temperature. Furthermore, the rate controlling characterizations of three-dimension diffusion were also discussed.

Clean and deep separation of molybdenum (Mo) and rhenium (Re) from the industrial solutions with ultra-low concentrations are extremely difficult. In current study, various low concentrations of Mo and Re are efficiently separated from the industrial solutions, including lixiviums, eluents and the waste waters, via vapidly stepwise selective coagulation and flocculation precipitation. In-situ generated metal hydroxyl coagulants in the solution have selective coagulation of series isopolymolybdates, while the cetyl-trimethyl-ammonium-bromide (CTAB) has a strong combination with rhenate. The effects of metal ion types, dosages, pH value, and reaction time on the separation behaviors of Mo and Re from the solution are systematically investigated via thermodynamic calculations, SEM-EDS, FT-IR and XPS analyses. It’s found that all the in-situ generated metal hydroxyl coagulants of Cu(II), Al(III), Fe(II) and Fe(III) have selective coagulation of Mo while the Fe(III) hydroxyl coagulant (FeHC) shows the best adsorption performance. After selective coagulation by FeHC, 99.9%Mo is separated while 95.8%Re is kept in the raffinate after filtration. Then the 98.9%Re is further recovered by CTAB flocculation precipitation. The selective coagulation mechanism of Mo and Re by the FeHC is discussed. In addition, this novel technology is applicative for the deep separation of Mo and Re from various low-concentration Mo,Re-bearing industrial solutions.

According to the laws of thermodynamics, materials normally exhibit contraction or expansion along the directions of the applied pressure or tension. Here, we show that a man-made cocrystal of a metallofullerene and highly energetic cubane, with strained sp3 bonding, may exhibit an anomalous negative volume compressibility. In this cocrystal, the freely rotating fullerene Sc3N@C80 acts as a structural building block while static cubane molecules fill the lattice interstitial sites. Under high pressure, Sc3N@C80 keeps stable and preserves the crystalline framework of the materials, while the cubane undergoes a progressive configurational transformation above 6.5 GPa, probably promoted by charge transfer from fullerene to cubane. A further configurational change of the cubane into a low-density configuration at higher pressure results in an anomalous pressure-driven lattice expansion of the cocrystal (∼1.8% volume expansion). Such unusual negative compressibility has previously only been predicted by theory and suggested to appear in mechanical metamaterials.

Traditional utilization of ferruginous manganese ore has mainly concentrated on the separation of Mn and Fe. However, an alternative process of novel uphill reaction diffusion followed by magnetic separation was developed for the synchronous recovery of manganese and iron as the form of ferrite ceramics from the ferruginous manganese ore in this work. The phase reconstruction, interfacial uphill reaction diffusion, ferrite size quantification and separation behavior were characterized via various means of XRD, XPS, TEM, VSM, SEM-EDS, optical microscopy, microhardness and thermodynamic analyses. In the phase reconstruction course, manganese and iron oxides reacted to form ferrites, whereas the impurities were enriched in liquid silicate phases due to the uphill reaction diffusion. Meanwhile, the ferrite particles grew and the liquid silicates aggregated owing to the intensified reaction diffusion by CaO addition. During the water quenching process, microcracks at the cleavage plane between the silicate and ferrites are preferentially generated due to fast cooling shrinkage of liquid silicates. Huge differences in hardness, magnetism, fusibility and cooling shrinkage of ferrites and silicates provided excellent mineralogical conditions for the successful grinding and magnetic separation of ferrites. Eventually, a soft magnetic manganese ferrite precursor with recoveries of 89.2 wt% iron and 85.6 wt% manganese was obtained. The interfacial reaction mechanisms for the formation of ferrites and silicates were also discussed in this work.

High-P oolitic hematite is recognised as one of the most refractory mineral resources because of the special oolitic structure. Conventional selective reduction roasting is characterised by high energy consumption and large greenhouse gas emissions. In this study, a clean utilisation of refractory high-P oolitic hematite via phase reconstruction was proposed to make the P for crop fertiliser and Fe for ferrite ceramic. Phase reconstruction indicated that the oolitic and finely disseminated structure was dramatically changed by the oxidisation roasting with Mn ore at 1100°C–1250 °C. After reconstruction, 85 wt%–90 wt% Fe and Mn were concentrated in the magnetic ferrite phase whilst 90 wt%–100 wt% P, K, Ca, Si, Mg, and Al impurities were enriched in the nonmagnetic P-bearing tailings. After separation, the separated Mn ferrite and P-bearing tailings after activation could be used as ingredients in ceramic and fertiliser production. The as-purified ferrite showed good electromagnetism performance whilst almost all nutrients in the P-bearing tailings were activated via high-temperature melting and water quenching and can thus be readily absorbed by the crop roots. Moreover, the phase reconstruction mechanism and feasible granular fertiliser preparation from P-bearing tailings were discussed. All elements in the high-P oolitic hematite were utilised without emission of greenhouse gases.

Humics are considered as an ideal surfactant for removing cupric ions from mining wastewater by ion flotation process. In this study, ultra-efficient enhancement of surface interactions of humics with cupric ions were systematically investigated via combining DFT computations and MD simulations. The results confirm that carboxyl groups play a crucial role in surface interactions, and surface complexation capacities of cupric ions are intensively relevant with carboxyl group contents. Therefore, large number of carboxyl groups are introduced into humic surfaces (MHA) through the esterification reaction between phenolic hydroxyl groups and citric acid, which is quite beneficial to enhance surface interactions of humics with cupric ions. The interactions of humics/MHA with cupric ions strengthen gradually as the number of deprotonated carboxyl groups increases. Owing to existing abundant carboxyl groups on MHA surfaces, the surface interactions are much stronger than that of humics. Unlike humics, cupric ions can form strong chelate with carboxyl groups of MHA. Additionally, new insights into the surface interaction mechanisms were explored, which are dominated by surface complexation, combining the effects of electrostatic interactions. The above-mentioned results demonstrate that it is feasible to enhance surface interactions by increasing the number of carboxyl groups of humic surfaces. This study could provide a new idea for designing efficient flotation reagents and improving removal efficiency of metal ions at molecular scales.

SiMn alloy is the largest consumption product in the Mn series alloys. Manganese silicates generated during the pre-treatment procedures of manganese ores are the main ingredient for SiMn alloy production. However, conventional cognition is that manganese and silicon in the silicates are respectively reduced to Mn7C3 and SiC by carbon, and then the carbides interact to form alloy. In current study, pure MnSiO3 was synthesized firstly and then a new perspective on the interface reaction and morphology evolution for the reduction of pure MnSiO3 were investigated via XRD, SEM, XPS and thermodynamics analyses. Interface reaction showed that the MnSiO3 was firstly reduced to manganese carbides and SiO2, and then the SiO2 was further reduced by the carbides to generate SiMn alloy. The intermediate product Mn7C3 played a crucial role since the reduction of SiO2 by newly born Mn7C3 preferentially conducted than the carbon reductant. Morphology evolution of the interface between C and MnSiO3 as well as thermodynamic calculations of Mn-Si-O and Mn-Si-C-O systems were minutely clarified in this work to clarify the reduction mechanism. In addition, decomposition of SiMn alloy to SiC ceramic at 1700 °C was an important inducement to illustrate the frangibility of industrial SiMn alloy product.

Methane and hydrogen reduction has emerged as promising technology for clean and efficient metallurgy. To understand the reduction process of MnO2 by CH4-H2 gas mixture, the effects of various CH4/(CH4 + H2) molar ratios (0–1) and temperatures (1100°C–1250°C) on the reduction of pure MnO2 were researched. Characterization of the solid carbon derived from the decomposition of CH4 gas was also determined. Thermodynamic analysis indicated that the MnO2 was first reduced to MnO by CH4-H2 gas mixture, and then MnO was reduced to Mn7C3 according to this reaction, 7MnO(s) + 10CH4(g) = Mn7C3(s) + 7CO(g) + 20H2(g). Experimental results showed that under the CH4/(CH4 + H2) molar ratio of 0.2, the reduction extent of MnO2 increased with the rising of temperatures. Nevertheless, the deep reduction of MnO to Mn7C3 was retarded even by increasing the temperature and prolonging the time, which was a result of the formation of deposited solid carbon. The free carbon contents in the reduced samples increased with the increasing of the CH4/(CH4 + H2) molar ratio and reduction time. The reduction of MnO2 and formation of carbon can be regulated by adjusting the CH4/(CH4 + H2) molar ratio in a gas mixture. For example, when MnO2 was reduced by 50 vol.%CH4-50%N2, large amounts of MnO and Mn7C3 particles were wrapped by the platelike and rodlike carbon with nanoscale. Yet, the wrapping phenomenon cannot be observed under 10 vol.%CH4-40 vol.%H2-50 vol.%N2.

Phase transformation of MnO 2 and Fe 2 O 3 roasted in air atmosphere at temperatures of 500–1400 °C for synthesizing manganese ferrite (Mn x Fe 3 − x O 4 ) has been reported. In current work, the morphologies and properties of the Mn x Fe 3 − x O 4 products with various x values during the synthesis process were characterized by XRD, XPS, SEM-EDS, AFM, VSM and Vickers-type microhardness analyses to further understand the formation mechanism of Mn x Fe 3 − x O 4 from MnO 2 and Fe 2 O 3 by solid-state reaction. Results showed that the x value in the Mn x Fe 3 − x O 4 gradually increased to 1 with the extension of roasting time from 30 min to 120 min at 1300 °C. Morphology measurements indicated the Mn x Fe 3 − x O 4 particles presented polyhedron structure and the shape of polyhedron particles gradually became regular with the increasing of x value. In addition, it's interesting to find that the polyhedron particles had multi-layer stacking structure and the thickness of each layer was about 5 nm. The saturation magnetization of the Mn x Fe 3 − x O 4 samples increased with the rising of x value. A polyhedral MnFe 2 O 4 (x = 1) product with saturation magnetization of 78.5 emu/g was obtained when the MnO 2 and Fe 2 O 3 briquettes with molar ratio of 1:1 were roasted at 1300 °C for 120 min. Moreover, the Hv hardness of the Mn x Fe 3 − x O 4 increased from 154.8 kg/mm 2 to 339.4 kg/mm 2 with elevating the temperature from 1000 °C to 1300 °C.

Chemical vapor transport (CVT) process is a simple and efficient method for synthesizing nano-SnO products, and previous references reveal that the CVT process is characterized as high temperature and long synthesis time. Our previous studies indicated that SnO2 was easily reduced to solid-state SnO and then volatilized as gaseous SnO under certain CO-CO2 atmosphere, so that a novel CVT process under CO-CO2 atmosphere was proposed in the present work. The nano-SnOs synthesized by the CVT under 15 vol% CO/(CO + CO2) atmosphere were characterized using XRD, SEM, TEM, XPS, etc. The results showed that tetragonal structure nano-SnO diskette and sphere with uniform granularity of 50–100 nm were obtained with high purity. During the CVT process, CO-CO2 mixed gases could reduce SnO2 to SnO and act as a CVT medium to take away the newly formed gaseous SnO, resulting in the decrease of the partial pressure of SnO. In addition, the nano-SnOs could enhance the SCR deNOx activity of activated carbon (AC), and both the reaction temperature window of AC and the NO conversion were increased if the AC was coated by nano-SnOs. Keywords: Nano-SnO, Chemical vapor transport, Reduction volatilization, SCR activity

An investigation on the behavior of SnO 2 in the tin-bearing iron concentrates during the reduction sintering process (RSP) has been conducted by the authors' group. It was found that only about 30 wt.% of tin was removed under the optimal conditions. However, the underlying reason why it was so difficult to remove tin by the RSP was not understood. In this study, the phase evolution of tin oxide, iron oxide and calcium oxide under simulative sintering conditions was investigated by varying the gas compositions, sintering temperature and basicity (CaO/SiO 2 ). The results indicated that the Ca-Sn oxides were prior formed at 900 °C–1000 °C in the reduction atmosphere with CO / (CO + CO 2 ) = 10–40 vol.%, while the phases of Fe-Sn spinel (Fe 3 − x Sn x O 4 ) and Ca-Fe-Sn oxides were easily generated as the temperature increased to 1100 °C–1200 °C. The phases of Fe 2 − 4x/3 Sn x O 3 and Ca-Fe-Sn-Si oxides began to generate at 1100 °C in an oxidative atmosphere with O 2 /(O 2 + CO 2 ) = 25–100 vol.%. In addition, increasing alkalinity accelerated the formation of Ca-Sn oxides under reduction atmosphere, which brought about an adverse effect on the tin removal in the sintering process.

Sodium salts were used in the reduction roasting and magnetic separation process to separate and recover iron and manganese from the high-iron manganese oxide ores to utilize the complex ores. Results showed that Na2S2O3 was the most effective salt. A magnetic concentrate with 86.39 wt% TFe and 96.21 wt% Fe recovery as well as a nonmagnetic product with 54.84 wt% TMn and 85.96 wt% Mn recovery was obtained when the ore sample was reduced at 1100°C for 100 min in the presence of 7 wt% Na2S2O3. In addition, the effects of roasting and separation parameters on the recovery of manganese and iron and the function mechanism of Na2S2O3 were investigated.

A selective reduction volatilization (SRV) process developed by the authors was effective to remove tin from the complex tin-bearing iron concentrates. It is reported that many gangue minerals of quartz, calcite, and fluorite are existent in the concentrates. Our previous studies investigated the effect of quartz and calcite on the removal of tin. In this study, the effect of fluorite (calcium fluoride) was conducted, and pure CaF2 was added into the tin-bearing iron concentrates to investigate how CaF2 affected the tin volatilization behavior under different roasting conditions. The effect of CaF2 on the phase transformation of SnO2 was also determined using SEM-EDS and XRD. The results indicated that SnF2 was much easier to volatilize than SnO under CO-CO2 atmosphere in the presence of CaF2, so the calcium fluoride in the concentrates was beneficial to promote the tin removal during the SRV process.