In order to address the evolving emission characteristics of oxygenated volatile organic compounds(OVOCs),it is essential to develop adsorbent materials specifically designed for the efficient adsorption of OVOCs with...In order to address the evolving emission characteristics of oxygenated volatile organic compounds(OVOCs),it is essential to develop adsorbent materials specifically designed for the efficient adsorption of OVOCs with large kinetic diameters.In this study,we used co-pyrolysis to prepare a series of graded porous carbon materials with well-developed micropores by adjusting the doping ratios of root nodules and pretreated cellulose.The material with root nodule to cellulose mass ratio of 1:1(TCC-RN-1)exhibited the highest saturated adsorption capacity for butyl acetate(834 mg/g).This can be attributed to enhanced pore size distribution from nodule doping,which facilitates the development of a micropore-graded structure.Additionally,the nodules acted as auxiliary activating agents that enhanced the KOH micropore regulation effect during the activation stage,resulting in the highest micropore volume(0.863 cm^(3)/g).The doping of root nodules facilitated the formation of additional defects on the surface of the porous carbon material,leading to a more disordered arrangement that improved pollutant adsorption.Furthermore,TCC-RN-1 demonstrated good thermal stability in an air atmosphere,main-taining a butyl acetate adsorption capacity exceeding 95%after five adsorption-desorption cycles.This indicates its favorable potential for industrial applications.展开更多
碳纳米管(CNT)因其优异的物理性能被广泛认为是一种有效的挥发性有机化合物(VOCs)吸附剂。然而,由于团聚效应产生传质阻力,会影响CNT的吸附性能。微纤复合材料作为一种新型结构材料,有效克服了这一问题。文中采用化学气相沉积(CVD)法在...碳纳米管(CNT)因其优异的物理性能被广泛认为是一种有效的挥发性有机化合物(VOCs)吸附剂。然而,由于团聚效应产生传质阻力,会影响CNT的吸附性能。微纤复合材料作为一种新型结构材料,有效克服了这一问题。文中采用化学气相沉积(CVD)法在纸状烧结不锈钢纤维载体(PSSF)表面分别直接生长CNT和硫掺杂碳纳米管(SCNT),成功制备了具有梯度多孔结构的CNT/PSSF和SCNT/PSSF复合材料。通过扫描电子显微镜(SEM)、透射电子显微镜(TEM)、热重分析技术(TG)、拉曼光谱(Raman)、傅里叶变换红外光谱仪(FT-IR)和X射线光电子能谱(XPS)对其进行系统表征。固定床反应器中的吸附实验结果表明,SCNT/PSSF因其表面丰富的缺陷和酸碱官能团,展现出优异的动态吸附性能。此外,通过构建由颗粒CNT和SCNT/PSSF组成的结构化固定床,研究了微纤复合材料在二甲苯吸附过程中的传质增强机理。与纯5 cm CNT结构床相比,4 cm CNT+1 cm SCNT/PSSF床层的利用率(ΔZ)提高了8.1%。结构化固定床的吸附过程可以被Yoon-Nelson模型良好拟合。展开更多
随着深度学习技术的发展,光学字符识别(optical character recognition,OCR)技术逐步从传统方法转向基于深度神经网络的端到端学习模型,涌现出大量具备高准确率和强泛化能力的OCR大模型。多模态大模型通过融合视觉、语言等多种感知通道...随着深度学习技术的发展,光学字符识别(optical character recognition,OCR)技术逐步从传统方法转向基于深度神经网络的端到端学习模型,涌现出大量具备高准确率和强泛化能力的OCR大模型。多模态大模型通过融合视觉、语言等多种感知通道,提高了模型在复杂场景下的理解与生成能力,而多任务统一大模型则通过构建通用架构,简化了模型设计,提升了多个OCR任务的处理效率。本文回顾了OCR和多模态学习领域的最新技术进展,重点介绍了OCR大模型在多模态学习和多任务统一模型中的应用与前沿进展。此外,本文还分析了OCR增强的多模态大模型、文档理解多模态大模型和针对特定OCR任务的多模态大模型的现状与挑战,探讨了OCR大模型面临的技术瓶颈和未来发展方向,包括提升分辨率处理能力、改进视觉标记压缩、增强结构化图形符号和复杂版面结构的感知与理解等,展望了其在文档数字化、程序自动化测试和智能教育等方面的广泛应用潜力。展开更多
基金supported by the National Natural Science Foundation of China(No.52370112).
文摘In order to address the evolving emission characteristics of oxygenated volatile organic compounds(OVOCs),it is essential to develop adsorbent materials specifically designed for the efficient adsorption of OVOCs with large kinetic diameters.In this study,we used co-pyrolysis to prepare a series of graded porous carbon materials with well-developed micropores by adjusting the doping ratios of root nodules and pretreated cellulose.The material with root nodule to cellulose mass ratio of 1:1(TCC-RN-1)exhibited the highest saturated adsorption capacity for butyl acetate(834 mg/g).This can be attributed to enhanced pore size distribution from nodule doping,which facilitates the development of a micropore-graded structure.Additionally,the nodules acted as auxiliary activating agents that enhanced the KOH micropore regulation effect during the activation stage,resulting in the highest micropore volume(0.863 cm^(3)/g).The doping of root nodules facilitated the formation of additional defects on the surface of the porous carbon material,leading to a more disordered arrangement that improved pollutant adsorption.Furthermore,TCC-RN-1 demonstrated good thermal stability in an air atmosphere,main-taining a butyl acetate adsorption capacity exceeding 95%after five adsorption-desorption cycles.This indicates its favorable potential for industrial applications.
文摘碳纳米管(CNT)因其优异的物理性能被广泛认为是一种有效的挥发性有机化合物(VOCs)吸附剂。然而,由于团聚效应产生传质阻力,会影响CNT的吸附性能。微纤复合材料作为一种新型结构材料,有效克服了这一问题。文中采用化学气相沉积(CVD)法在纸状烧结不锈钢纤维载体(PSSF)表面分别直接生长CNT和硫掺杂碳纳米管(SCNT),成功制备了具有梯度多孔结构的CNT/PSSF和SCNT/PSSF复合材料。通过扫描电子显微镜(SEM)、透射电子显微镜(TEM)、热重分析技术(TG)、拉曼光谱(Raman)、傅里叶变换红外光谱仪(FT-IR)和X射线光电子能谱(XPS)对其进行系统表征。固定床反应器中的吸附实验结果表明,SCNT/PSSF因其表面丰富的缺陷和酸碱官能团,展现出优异的动态吸附性能。此外,通过构建由颗粒CNT和SCNT/PSSF组成的结构化固定床,研究了微纤复合材料在二甲苯吸附过程中的传质增强机理。与纯5 cm CNT结构床相比,4 cm CNT+1 cm SCNT/PSSF床层的利用率(ΔZ)提高了8.1%。结构化固定床的吸附过程可以被Yoon-Nelson模型良好拟合。
文摘随着深度学习技术的发展,光学字符识别(optical character recognition,OCR)技术逐步从传统方法转向基于深度神经网络的端到端学习模型,涌现出大量具备高准确率和强泛化能力的OCR大模型。多模态大模型通过融合视觉、语言等多种感知通道,提高了模型在复杂场景下的理解与生成能力,而多任务统一大模型则通过构建通用架构,简化了模型设计,提升了多个OCR任务的处理效率。本文回顾了OCR和多模态学习领域的最新技术进展,重点介绍了OCR大模型在多模态学习和多任务统一模型中的应用与前沿进展。此外,本文还分析了OCR增强的多模态大模型、文档理解多模态大模型和针对特定OCR任务的多模态大模型的现状与挑战,探讨了OCR大模型面临的技术瓶颈和未来发展方向,包括提升分辨率处理能力、改进视觉标记压缩、增强结构化图形符号和复杂版面结构的感知与理解等,展望了其在文档数字化、程序自动化测试和智能教育等方面的广泛应用潜力。
