查看更多>>摘要：Paleogeographic reconstructions are of key importance for understanding the history of continental breakups and amalgamations during Earth's history.A special case is the history of the Asian continent,which,compared to other continents,consists of several large cratons and numerous smaller continental blocks.The history of the assembly of South China remains controversial in terms of the timing,Late Neoproterozoic or Early Paleozoic,and the participating continental blocks,e.g.Yangtze,Cathaysia,India and Australia.Detrital rutile U-Pb dating has significant potential with regard to deciphering tectonic settings as rutile frequently crystallizes during orogenic events associated with the processes of collision and subduction.Detrital zircon U-Pb dating is a perfect instrument for identifying the provenance characteristics and age characteristics of sedimentary sources.An integration of these two methods of dating offers better opportunities for reconstructing tectonic settings.This paper presents a first attempt to reconstruct the Neoproterozoic to Early Paleozoic tectonic history and paleogeography of the whole South China based on U-Pb geochronology of rutile and zircon and Hf-in-zircon isotopes from Lower Jurassic Baitianba Formation sedimentary rocks of the western margin of the Yangtze Block,a major part of South China.Our obtained integrated U-Pb rutile and zircon age data show three main age populations of 960-940 Ma,630-610 Ma and 530-520 Ma.The new U-Pb detrital rutile and zircon ages,compared with former data from Gondwana and Australia,suggest that Yangtze amalgamated with Cathaysia to form South China during the Sibao orogeny at 960-940 Ma.The detrital rutile and zircons of the new 630-610 Ma age group could have been delivered from western Australia during the Late Neoproterozoic to Cambrian Paterson-Petermann orogeny.The abundant 530-520 Ma detrital rutile and zircon ages fit well with the coeval zircon age populations recorded in Gondwana-derived terranes,like India and Indochina.
查看更多>>摘要：In 2023,the CO2 growth rate was 3.37±0.11 ppm at Mauna Loa,which was 86％above that of the previous year and hit a record high since observations began in 1958,while global fossil fuel CO2 emissions only increased by 0.6％±0.5％.This implies an unprecedented weakening of land and ocean sinks,and raises the question of where and why this reduction happened.Here,we show a global net land CO2 sink of 0.44±0.21 GtC yr-1,which is the weakest since 2003.We used dynamic global vegetation models,satellite fire emissions,an atmospheric inversion based on OCO-2 measurements and emulators of ocean biogeochemical and data-driven models to deliver a fast-track carbon budget in 2023.Those models ensured consistency with previous carbon budgets.Regional flux anomalies from 2015 to 2022 are consistent between top-down and bottom-up approaches,with the largest abnormal carbon loss in the Amazon during the drought in the second half of2023(0.31±0.19 GtC yr-1),extreme fire emissions of 0.58±0.10 GtC yr-1 in Canada and a loss in Southeast Asia(0.13±0.12 GtC yr-1).Since 2015,land CO2 uptake north of 20°N had declined by half to 1.13±0.24 GtC yr-1 in 2023.Meanwhile,the tropics recovered from the 2015-2016 El Nino carbon loss,gained carbon during the La Nina years(2020-2023),then switched to a carbon loss during the 2023 El Nino(0.56±0.23 GtC yr-1).The ocean sink was stronger than normal in the equatorial eastern Pacific due to reduced upwelling from La Nina's retreat in early 2023 and the development of El Nino later.Land regions exposed to extreme heat in 2023 contributed a gross carbon loss of 1.73 GtC yr-1,indicating that record warming in 2023 had a strong negative impact on the capacity of terrestrial ecosystems to mitigate climate change.
查看更多>>摘要：Atmospheric CO2 growth rate(CGR),reflecting the carbon balance between anthropogenic emissions and net uptake from land and ocean,largely determines the magnitude and speed of global warming.The CGR at Mauna Loa Baseline Observatory reached a record high in 2023.We quantified major components of the global carbon balance for 2023,by developing a framework that integrated fossil fuel CO2 emissions data and an atmospheric inversion from the Global ObservatioN-based system for monitoring Greenhouse GAses(GONGGA)with two artificial intelligence(AI)models derived from dynamic global vegetation models.We attributed the record high CGR increase in 2023 compared to 2022 primarily to the large decline in land carbon sink(1803±197 TgC year-1),with minor contributions from a small reduction in ocean carbon sink(184 TgC year-1)and a slight increase in fossil fuel emissions(24 TgC year-1).At least 78％of the global decline in land carbon sink was contributed by the decline in tropical sink,with GONGGA inversion(1354 TgC year-1)and AI simulations(1578±666 TgC year-1)showing similar declines in the tropics.We further linked this tropical decline to the detrimental impact of El Nino-induced anomalous warming and drying on vegetation productivity in water-limited Sahel and southern Africa.Our successful attribution of CGR increase within a framework combining atmospheric inversion and AI simulations enabled near-real-time tracking of the global carbon budget,which had a one-year reporting lag.
查看更多>>摘要：Global ecosystems face mercury contamination,yet long-term data are scarce,hindering understanding of ecosystem responses to atmospheric Hg input changes.To bridge the data gap and assess ecosystem responses,we compiled and compared a mercury accumulation database from peat,lake,ice and marine deposits worldwide with atmospheric mercury deposition modelled by GEOS-Chem,focusing on trends,magnitudes,spatial-temporal distributions and impact factors.The mercury fluxes in all four deposits showed a 5-to 9-fold increase over 1700-2012,with lake and peat mercury fluxes that generally mirrored atmospheric deposition trends.Significant decreases in lake and peat mercury fluxes post-1950 in Europe evidenced effective environmental policies,whereas rises in East Asia,Africa and Oceania highlighted coal-use impacts,inter alia.Conversely,mercury fluxes in marine and high-altitude ecosystems did not align well with atmospheric deposition,emphasizing natural influences over anthropogenic impacts.Our study underscores the importance of these key regions and ecosystems for future mercury management.
查看更多>>摘要：Due to almost identical boiling points of benzene and cyclohexane,the extraction of trace benzene from cyclohexane is currently performed via the energy-intensive extractive distillation method.Their adsorptive separation by porous materials is hampered by their similar dimensions.Metal-organic frameworks(MOFs)with versatile pore environments are capable of molecular discrimination,but the separation of trace substrates in liquid-phase remains extremely challenging.Herein,we report a robust MOF(NKU-300)with triangular channels decorated with crown ether that can discriminate trace benzene from cyclohexane,exhibiting an unprecedented selectivity of 8615(10)for the mixture of benzene/cyclohexane(v/v=1/1000).Remarkably,NKU-300 demonstrates exceptional selectivities for the extraction of benzene from cyclohexane over a wide range of concentrations of 0.1％-50％with ultrafast sorption kinetics and excellent stability.Single-crystal X-ray diffraction and computational modelling reveal that multiple supramolecular interactions cooperatively immobilise benzene molecules in the triangular channel,enabling superior separation performance.This study will promote the application of advanced sorbents with tailored binding sites for challenging industrial separations.
查看更多>>摘要：Cu-based materials can produce hydrocarbons in CO2 electroreduction(CO2RR),but their stability still needs to be enhanced particularly in acidic media.Metallic Pt is highly stable in both acidic and alkaline media,yet rarely utilized in CO2RR,due to the competitive activity in hydrogen evolution reaction(HER).In this research,abundant thionine(Th)molecules are stably confined within Pt nanocrystals via a molecular doping strategy.The Pt surface is successfully modulated by these Th molecules,and thereby the dominant HER activity is converted to CO2RR activity.CO2 could be electroreduced to CH4 using organic molecule-modified Pt-based catalysts for the first time.Specifically,this composite catalyst maintains more than 100-hour stability in strong acid conditions(pH 1),even comparable to those state-of-the-art CO2RR catalysts.In-situ spectroscopic analysis and theoretical calculations reveal that the molecular modification can decrease the energy barrier for *COOH formation,and guarantee the sufficient local*H near Pt surface.Additionally,the*H derived from H2O dissociation is favorable for the *CO hydrogenation pathway towards *CHO,eventually leading to the formation of CH4.This strategy might be easily applied to microenvironment and interface regulation in other electrocatalytic reactions.
Ruijie MaHongxiang LiTop Archie Dela PeñaHeng Wang...
238-247页
查看更多>>摘要：Solid additive engineering has been intensively explored on morphology tuning for highly efficient all-polymer solar cells(all-PSCs),a promising photovoltaic technology towards multi-scenario application.Although the nano-fibrillar network of the active layer induced by additive treatment is confirmed as the key factor for power conversion efficiency(PCE)of all-PSCs,its formation mechanism is not clearly revealed,for lack of precise and convincing real-time observation of crystallization and phase separation during the liquid-to-solid transition process of spin-coating.Herein we report an in-situ grazing incidence wide-angle/small-angle X-ray scattering(GIWAXS/GISAXS)screening that reveals the fact that naphthalene derived solid additives can suppress the aggregation of the polymer acceptor(PY-IT)at the beginning stage of spin coating,which provides sufficient time and space for the polymer donor(PM6)to form the fibril structure.Moreover,guided by this knowledge,a ternary all-polymer system is proposed,which achieves cutting-edge level PCEs for both small-area(0.04 cm2)(also decent operational stability)and large-area(1 cm2)devices.
查看更多>>摘要：Diffusion models,a powerful and universal generative artificial intelligence technology,have achieved tremendous success and opened up new possibilities in diverse applications.In these applications,diffusion models provide flexible high-dimensional data modeling,and act as a sampler for generating new samples under active control towards task-desired properties.Despite the significant empirical success,theoretical underpinnings of diffusion models are very limited,potentially slowing down principled methodological innovations for further harnessing and improving diffusion models.In this paper,we review emerging applications of diffusion models to highlight their sample generation capabilities under various control goals.At the same time,we dive into the unique working flow of diffusion models through the lens of stochastic processes.We identify theoretical challenges in analyzing diffusion models,owing to their complicated training procedure and interaction with the underlying data distribution.To address these challenges,we overview several promising advances,demonstrating diffusion models as an efficient distribution learner and a sampler.Furthermore,we introduce a new avenue in high-dimensional structured optimization through diffusion models,where searching for solutions is reformulated as a conditional sampling problem and solved by diffusion models.Lastly,we discuss future directions about diffusion models.The purpose of this paper is to provide a well-rounded exposure for stimulating forward-looking theories and methods of diffusion models.
查看更多>>摘要：Recently,the multimodal large language model(MLLM)represented by GPT-4V has been a new rising research hotspot,which uses powerful large language models(LLMs)as a brain to perform multimodal tasks.The surprising emergent capabilities of the MLLM,such as writing stories based on images and optical character recognition-free math reasoning,are rare in traditional multimodal methods,suggesting a potential path to artificial general intelligence.To this end,both academia and industry have endeavored to develop MLLMs that can compete with or even outperform GPT-4V,pushing the limit of research at a surprising speed.In this paper,we aim to trace and summarize the recent progress of MLLMs.First,we present the basic formulation of the MLLM and delineate its related concepts,including architecture,training strategy and data,as well as evaluation.Then,we introduce research topics about how MLLMs can be extended to support more granularity,modalities,languages and scenarios.We continue with multimodal hallucination and extended techniques,including multimodal in-context learning,multimodal chain of thought and LLM-aided visual reasoning.To conclude the paper,we discuss existing challenges and point out promising research directions.
NETL
NSTL
万方数据