Search

With the rapid growth in the number of passenger cars (PCs) in China over the past decades, more than ten million tons of used tires have already become solid wastes and subsequently caused serious environmental issues. Due to the presence of synthetic rubber in PC tires, waste PC tires cannot be disposed through rubber reclaiming technology. Thus, waste PC tires have become one of fastest growing solid wastes in China. First, the current disposal capacity of the pyrolysis method, regarded as a promising technology for the disposal of waste PC tires, is surveyed and compared with other disposal methods mentioned in previous papers. Second, this work establishes a model to predict the total number of waste PC tires in the next five years depending on the rate of PC growth and current waste tire disposal capacity. Moreover, pyrolysis is evaluated on 15 collected waste PC tires selected from the most representative tire brands in the Chinese market. The corresponding results imply that ∼68.5% of S was into oil and ∼44.3% N and large amount of heavy metals resided in solid carbon which severely limit further applications. Finally, a new pyrolysis technology is introduced that may represent a solution to the limits in the application of tire disposal methods and relief for the coming waste tire crisis.

The production of aromatic hydrocarbons from the waste tire pyrolysis attracts more and more attention because of its tremendous potential. Based on styrene-butadiene rubber (SBR), which is the main rubber in the waste passenger car tires, this work studies the temperature influence on primary pyrolysis product distribution by experimental techniques (Py-GC/MS, TG-MS), and then, the formation mechanism of monocyclic aromatic hydrocarbons (MAHs) observed in the experiment was analyzed by first-principles calculations. The experimental results show that the MAHs during the pyrolysis mainly include styrene, toluene, and xylene, and subsequent calculations showed that these compounds were formed through a series of primary and secondary reactions. The formation pathways of these typical MAHs were studied via the reaction energy barrier analysis, respectively. It shows that the MAHs were not only derived from the benzene ring in the SBR chain but also generated from short-chain alkenes through the Diels-Alder reaction. The obtained pyrolysis reaction mechanism provides theoretical guidance for the regulation of the pyrolysis product distribution of MAHs.

This paper investigates robot manipulation based on human instruction with ambiguous requests. The intent is to compensate for imperfect natural language via visual observations. Early symbolic methods, based on manually defined symbols, built modular framework consist of semantic parsing and task planning for producing sequences of actions from natural language requests. Modern connectionist methods employ deep neural networks to automatically learn visual and linguistic features and map to a sequence of low-level actions, in an endto-end fashion. These two approaches are blended to create a hybrid, modular framework: it formulates instruction following as symbolic goal learning via deep neural networks followed by task planning via symbolic planners. Connectionist and symbolic modules are bridged with Planning Domain Definition Language. The vision-and-language learning network predicts its goal representation, which is sent to a planner for producing a task-completing action sequence. For improving the flexibility of natural language, we further incorporate implicit human intents with explicit human instructions. To learn generic features for vision and language, we propose to separately pretrain vision and language encoders on scene graph parsing and semantic textual similarity tasks. Benchmarking evaluates the impacts of different components of, or options for, the vision-and-language learning model and shows the effectiveness of pretraining strategies. Manipulation experiments conducted in the simulator AI2THOR show the robustness of the framework to novel scenarios., Comment: 8 pages, 3 figures, 3 tables

N2O is a greenhouse gas and a candidate oxidant. Volatile organic pollutants (VOCs) have caused great harm to the atmospheric ecological environment. Developing the technique utilizing N2O as the oxidant to oxidize VOCs to realize the collaborative purification has significant importance and practical value for N2O emission control and VOC abatement. Therefore, the study of N2O catalytic oxidation of tert-butanol based on zeolite catalysts was carried out. A series of molecular sieves, including FER, MOR, ZSM-5, Y, and BEA, were selected as the catalyst objects, and the 1.5% wt Fe and Co were, respectively, loaded on the zeolite catalysts via the impregnation method. It was found that the catalytic performance of BEA was the best among the molecular sieves. Comparing the catalytic performance of Fe-BEA under different load gradients (0.25~2%), it was found that 1.5% Fe-BEA possessed the best catalytic activity. A series of characterization methods showed that Fe3+ content in 1.5% Fe-BEA was the highest, and more active sites formed to promote the catalytic reaction. The α-O in the reaction eventually oxidized tert-butanol to CO2 over the active site. The Co mainly existed in the form of Co2+ cations over Co-BEA samples; the 2% Co-BEA possessing higher amounts of Co2+ exhibited the highest activity among the prepared Co-BEA samples.

The chlorine (Cl2) drying technology using ionic liquids (ILs) as absorbents was proposed for the first time and systematically investigated from the molecular level scaled up to the industrial level. The hygroscopic IL [EMIM][CH3SO3] was screened as a suitable absorbent from 238 potential IL candidates consisting of 14 cations and 17 anions, by calculating the Cl2 and H2O solubility and separation selectivity of Cl2 to H2O in different ILs based on the COSMO-RS model. The microscopic atomic and molecular insights into the separation mechanisms were deeply revealed by using COSMO-RS model analyses (i.e., σ-profiles, σ-potentials, excess enthalpies, entropies, and Gibbs free energies) and quantum chemistry calculation (binding energies and weak interaction analyses). The Cl2 solubility in pure IL and H2O + IL systems were predicted by the COSMO-RS model, and the results agree with the microscopic mechanism identification. Moreover, the strict equilibrium stage model employed with the COSMO-RS model parameters was built to perform the process simulation, and continuous Cl2 drying with ILs was conceptually designed and optimized at industrial scale. It was confirmed that [EMIM][CH3SO3] is a very promising absorbent leading to a less IL amount, a much lower energy consumption than the other IL [EMIM][BF4], which has a very bright industrialization potential used for Cl2 drying technology.

Direct oxidation of methane to methanol (DMTM), constituting a major challenge for C1 chemistry, has aroused significant interest. The present work reports the synergistic effect of neighboring [Fe]--[Cu] cations, which can significantly boost the CH3OH productivity (100.9 and 41.9 → 259.1 μmol∙g−1cat∙h−1) and selectivity (0.28 and 17.6% → 71.7%) of the best performing Fe0.6%Cu0.68%-BEA (relative to monomeric Fe1.28%- and Cu1.28%-BEA) during the continuous H2O-mediated N2O DMTM. The combined experimental (in situ FTIR, D2O isotopic tracer technique) and theoretical (DFT, ab initio molecular dynamics (AIMD)) studies reveal deeper mechanistic insights that the synergistic effect of [Fe]--[Cu] can not only significantly favor active O production (ΔG = 0.18 eV), but also efficiently motivate the reaction following a H2O proton-transfer route (ΔG = 0.07 eV), eventually strikingly promoting CH3OH productivity/selectivity. Generally, the proposed strategy by employing the synergistic effect of bimetallic cations to modify DMTM activity would substantially favor other highly efficient catalyst designs.

Contemporary grasp detection approaches employ deep learning to achieve robustness to sensor and object model uncertainty. The two dominant approaches design either grasp-quality scoring or anchor-based grasp recognition networks. This paper presents a different approach to grasp detection by treating it as keypoint detection in image-space. The deep network detects each grasp candidate as a pair of keypoints, convertible to the grasp representationg = {x, y, w, {\theta}} T , rather than a triplet or quartet of corner points. Decreasing the detection difficulty by grouping keypoints into pairs boosts performance. To promote capturing dependencies between keypoints, a non-local module is incorporated into the network design. A final filtering strategy based on discrete and continuous orientation prediction removes false correspondences and further improves grasp detection performance. GKNet, the approach presented here, achieves a good balance between accuracy and speed on the Cornell and the abridged Jacquard datasets (96.9% and 98.39% at 41.67 and 23.26 fps). Follow-up experiments on a manipulator evaluate GKNet using 4 types of grasping experiments reflecting different nuisance sources: static grasping, dynamic grasping, grasping at varied camera angles, and bin picking. GKNet outperforms reference baselines in static and dynamic grasping experiments while showing robustness to varied camera viewpoints and moderate clutter. The results confirm the hypothesis that grasp keypoints are an effective output representation for deep grasp networks that provide robustness to expected nuisance factors., Comment: 25 pages, 12 figures, 14 tables

This letter presents a framework to detect and rank affordances of novel objects to assist with robotic manipulation tasks. The framework segments the affordance map of unseen objects using region-based affordance segmentation. Detected affordances define an initial state from which to generate action primitives for manipulation via the planning domain definition language (PDDL). The proposed category-agnostic affordance segmentation approach generalizes learned affordances to unseen objects by utilizing binary classification on proposed instance masks. The predicted pixel-wise level affordances are ranked by KL-divergence, augmenting the available affordance choices for manipulation tasks with non-primary affordances of an object. Experimental results show that the proposed method achieves state-of-the-art performance on affordance segmentation of novel objects, and outperforms baselines on affordance ranking. Actual robotic manipulation scenarios demonstrate the use of affordance detection with PDDL-generated action primitives for task execution. Prediction of ranked affordances on unseen objects provides flexibility to accomplish goal-oriented tasks.

In this study, comparative studies over nine NH3-SCR Cu-zeolites with different topologies were conducted. A clear trend was observed that the NH3-SCR performance was gradually improved with the zeolite-framework structures changing from straight-channel (ZSM-5, Beta, Y) to cage-type (SSZ-13, SSZ-16, SSZ-17), bridged by hybrid-structures (OFF-ERI series: Offretite, ZSM-34, UZM-12). Characterizations of the catalysts with XRD, XRF, XPS, UV–vis, NH3-TPD, and H2-TPR confirmed, under identical catalyst synthesis conditions, that the cage-type samples (Cu-SSZ-13, Cu-SSZ-16, and Cu-SSZ-17) possessed higher concentrations of strong acid sites and isolated Cu cations with respect to the straight-channel samples (Cu-ZSM-5, Cu-Beta, and Cu-Y). The cage-type catalysts showed superior SCR performance as compared to the straight-channel and hybrid types. This study corroborated the consensus from recent literature that a small cage aperture (∼3.8 A) is critical to prevent zeolite dealumination. Moreover, a clear correlation was found between zeolite acid strength and the nature of exchanged Cu species. Cage type zeolites with stronger acidity led to stabilization of isolated Cu ion monomers while dimeric Cu ions form in straight-channel zeolites with lower acidity. Catalysts with dominant monomeric active sites display more desirable SCR activity and selectivity to N2.

This letter presents a deep learning framework to predict the affordances of object parts for robotic manipulation. The framework segments affordance maps by jointly detecting and localizing candidate regions within an image. Rather than requiring annotated real-world images, the framework learns from synthetic data and adapts to real-world data without supervision. The method learns domain-invariant region proposal networks and task-level domain adaptation components with regularization on the predicted domains. A synthetic version of the UMD data set is collected for autogenerating annotated, synthetic input data. Experimental results show that the proposed method outperforms an unsupervised baseline, and achieves performance close to state-of-the-art supervised approaches. An ablation study establishes the performance gap between the proposed method and the supervised equivalent (30%). Real-world manipulation experiments demonstrate use of the affordance segmentations for task execution, which achieves the same performance with supervised approaches.

We consider the task of grasping a target object based on a natural language command query. Previous work primarily focused on localizing the object given the query, which requires a separate grasp detection module to grasp it. The cascaded application of two pipelines incurs errors in overlapping multi-object cases due to ambiguity in the individual outputs. This work proposes a model named Command Grasping Network(CGNet) to directly output command satisficing grasps from RGB image and textual command inputs. A dataset with ground truth (image, command, grasps) tuple is generated based on the VMRD dataset to train the proposed network. Experimental results on the generated test set show that CGNet outperforms a cascaded object-retrieval and grasp detection baseline by a large margin. Three physical experiments demonstrate the functionality and performance of CGNet., 7 pages, 2 figures, Accepted to the ICRA2021

Surface segregation constitutes an efficient approach to enhance the alkaline hydrogen evolution reaction (HER) activity of the bimetallic PtxNiy nanoalloys. Herein, we propose a new strategy by utilizing the small gas molecule of H2 as the structure directing agent (SDA) to in-situ induce Pt surface segregations over a series of PtNi5-n samples with extremely low Pt dopings (Pt/Ni = 0.2). Impressively, the sample of PtNi5-0.3 synthesized under 0.3 MPa H2 delivers an unprecedentedly low overpotential of 12.8 mV (-10 mA cm-2) and Tafel slop of 14.1 mV dec-1, which is superior than most of previously reported PtxNiy electrocatalyst. This is closely related to the strong in-situ inducing effect of H2 (0.3 MPa) leading to an obvious Pt-rich@Ni-rich core-shell nanostructure which can efficiently reduce the reaction Gibbs free energies. Finally, the specific mechanistic effects of H2 during PtNi5-n synthesis process are well illustrated based on the combined experimental and theoretical studies.

A deep learning architecture is proposed to predict graspable locations for robotic manipulation. It considers situations where no, one, or multiple object(s) are seen. By defining the learning problem to be classified with null hypothesis competition instead of regression, the deep neural network with red, green, blue and depth (RGB-D) image input predicts multiple grasp candidates for a single object or multiple objects, in a single shot. The method outperforms state-of-the-art approaches on the Cornell dataset with 96.0% and 96.1% accuracy on imagewise and object-wise splits, respectively. Evaluation on a multiobject dataset illustrates the generalization capability of the architecture. Grasping experiments achieve 96.0% grasp localization and 89.0% grasping success rates on a test set of household objects. The real-time process takes less than 0.25 s from image to plan.

Structural effects on various imidazolium–based ionic liquids (ILs) (i.e., [C][A], [C] = [EMIM]+, [BMIM]+, and [OMIM]+; [A] = [BF4]–, [PF6]–, and [Tf2N]–) suitability as drying agents for gas dehydration processes are explored from thermodynamic and dynamic insights by quantum chemistry (QC) calculations and molecular dynamics (MD) simulations. It is found that [EMIM][BF4] is regarded as the most promising drying agent because it exhibits the lowest Henry’s law constant and highest diffusion coefficient of H2O among all ILs. The microscopic mechanism at molecular level is revealed based on QC calculations and MD simulations, and the results demonstrate that the IL (i.e., [EMIM][BF4]) simultaneously with the smallest cation and anion size corresponds to both the strongest hydrogen bond (HB) interaction of H2O–anion and the strongest HB together with van der Waals interactions of H2O–cation. This work provides a valuable guidance from viewpoint of thermodynamics and dynamics for developing and screening novel ILs for gas dehydration.

A human-in-the-loop system is proposed to enable collaborative manipulation tasks for person with physical disabilities. Studies show that the cognitive burden of subject reduces with increased autonomy of assistive system. Our framework obtains high-level intent from the user to specify manipulation tasks. The system processes sensor input to interpret the user's environment. Augmented reality glasses provide ego-centric visual feedback of the interpretation and summarize robot affordances on a menu. A tongue drive system serves as the input modality for triggering a robotic arm to execute the tasks. Assistance experiments compare the system to Cartesian control and to state-of-the-art approaches. Our system achieves competitive results with faster completion time by simplifying manipulation tasks.

SSZ-13 zeolite, a CHA-type aluminosilicate, has attracted wide attention recently because of its extraordinary physiochemical properties and excellent potential for applications in catalysis. In this work, SSZ-13 zeolite was synthesized using choline chloride as a CHA-topology construction template. This new template has a low cost and is environmentally friendly. Specifically it has the appropriate geometry and functional groups for easy self-assembly as dual molecules to form the CHA topology, which was verified by self-prepared organic templates with similar structure and functional groups as well as IR spectroscopy and DFT and potential energy distribution calculations. After ion exchange by Cu cations, the as-synthesized SSZ-13 exhibited an outstanding performance for the selective catalytic reduction of NO and showed ≈100 % NO conversions across a wide temperature window of 150–450 °C.

The front cover artwork for issue 08/2014 is a collaboration between Beijing University of Chemical Technology (P.R. China) and Poitiers University (France). The image shows the principle on catalyst combination for an ideal autoexhaust control covering a broad temperature range. See the Full Paper itself at dx.doi.org/10.1002/cctc.201402514.