Substantial contribution of transported emissions to organic aerosol in Beijing.

Haze in Beijing is linked to atmospherically formed secondary organic aerosol, which has been shown to be particularly harmful to human health. However, the sources and formation pathways of these secondary aerosols remain largely unknown, hindering effective pollution mitigation. Here we have quantified the sources of organic aerosol via direct near-molecular observations in central Beijing. In winter, organic aerosol pollution arises mainly from fresh solid-fuel emissions and secondary organic aerosols originating from both solid-fuel combustion and aqueous processes, probably involving multiphase chemistry with aromatic compounds. The most severe haze is linked to secondary organic aerosols originating from solid-fuel combustion, transported from the Beijing-Tianjing-Hebei Plain and rural mountainous areas west of Beijing. In summer, the increased fraction of secondary organic aerosol is dominated by aromatic emissions from the Xi'an-Shanghai-Beijing region, while the contribution of biogenic emissions remains relatively small. Overall, we identify the main sources of secondary organic aerosol affecting Beijing, which clearly extend beyond the local emissions in Beijing. Our results suggest that targeting key organic precursor emission sectors regionally may be needed to effectively mitigate organic aerosol pollution.

Quantifying HONO Production from Nitrate Photolysis in a Polluted Atmosphere.

The photolysis of particulate nitrate (pNO3-) has been suggested to be an important source of nitrous acid (HONO) in the troposphere. However, determining the photolysis rate constant of pNO3- (jpNO3-) suffers from high uncertainty. Prior laboratory measurements of jpNO3- using aerosol filters have been complicated by the "shadow effect"─a phenomenon of light extinction within aerosol layers that potentially skews these measurements. We developed a method to correct the shadow effect on the photolysis rate constant of pNO3- for HONO production (jpNO3- â†’ HONO) using aerosol filters with identical chemical compositions but different aerosol loadings. We applied the method to quantify jpNO3- â†’ HONO over the North China Plain (NCP) during the winter haze period. After correcting for the shadow effect, the normalized average jpNO3- â†’ HONO at 5 °C increased from 5.89 × 10-6 s-1 to 1.72 × 10-5 s-1. The jpNO3- â†’ HONO decreased with increasing pH and nitrate proportions in PM2.5 and had no correlation with nitrate concentrations. A parametrization for jpNO3- â†’ HONO was developed for model simulation of HONO production in NCP and similar environments.

Recent Progress in Free Radical Transformations of Allenamides.

Allenamides are special allenes, and the unique reactivity, selectivity (both stereoselective and regionally selective) and stability of allenamides have been widely studied. In this review, the development of the free radical transformation of allenamides over the last few years will be summarized. This review discusses in detail in three parts: intermolecular radical addition to C- X (X = N, S, O, Se) bonds, metal salt mediated cyclization of allenamides, and photocatalytic cyclization of allenamides. In addition, reasonable details of the mechanisms are provided for the vast majority of these transformations.

Non-fibril amyloid aggregation at the air/water interface: self-adaptive pathway resulting in a 2D Janus nanofilm.

The amyloid states of proteins are implicated in several neurodegenerative diseases and bioadhesion processes. However, the classical amyloid fibrillization mechanism fails to adequately explain the formation of polymorphic aggregates and their adhesion to various surfaces. Herein, we report a non-fibril amyloid aggregation pathway, with disulfide-bond-reduced lysozyme (R-Lyz) as a model protein under quasi-physiological conditions. Very different from classical fibrillization, this pathway begins with the air-water interface (AWI) accelerated oligomerization of unfolded full-length protein, resulting in unique plate-like oligomers with self-adaptive ability, which can adjust their conformations to match various interfaces such as the asymmetric AWI and amyloid-protein film surface. The pathway enables a stepwise packing of the plate-like oligomers into a 2D Janus nanofilm, exhibiting a divergent distribution of hydrophilic/hydrophobic residues on opposite sides of the nanofilm. The resulting Janus nanofilm possesses a top-level Young's modulus (8.3 ± 0.6 GPa) among amyloid-based materials and exhibits adhesive strength two times higher (145 ± 81 kPa) than that of barnacle cement. Furthermore, we found that such an interface-directed pathway exists in several amyloidogenic proteins with a similar self-adaptive 2D-aggregation process, including bovine serum albumin, insulin, fibrinogen, hemoglobin, lactoferrin, and ovalbumin. Thus, our findings on the non-fibril self-adaptive mechanism for amyloid aggregation may shed light on polymorphic amyloid assembly and their adhesions through an alternative pathway.

Identifying Driving Factors of Atmospheric N2O5 with Machine Learning.

Dinitrogen pentoxide (N2O5) plays an essential role in tropospheric chemistry, serving as a nocturnal reservoir of reactive nitrogen and significantly promoting nitrate formations. However, identifying key environmental drivers of N2O5 formation remains challenging using traditional statistical methods, impeding effective emission control measures to mitigate NOx-induced air pollution. Here, we adopted machine learning assisted by steady-state analysis to elucidate the driving factors of N2O5 before and during the 2022 Winter Olympics (WO) in Beijing. Higher N2O5 concentrations were observed during the WO period compared to the Pre-Winter-Olympics (Pre-WO) period. The machine learning model accurately reproduced ambient N2O5 concentrations and showed that ozone (O3), nitrogen dioxide (NO2), and relative humidity (RH) were the most important driving factors of N2O5. Compared to the Pre-WO period, the variation in trace gases (i.e., NO2 and O3) along with the reduced N2O5 uptake coefficient was the main reason for higher N2O5 levels during the WO period. By predicting N2O5 under various control scenarios of NOx and calculating the nitrate formation potential from N2O5 uptake, we found that the progressive reduction of nitrogen oxides initially increases the nitrate formation potential before further decreasing it. The threshold of NOx was approximately 13 ppbv, below which NOx reduction effectively reduced the level of night-time nitrate formations. These results demonstrate the capacity of machine learning to provide insights into understanding atmospheric nitrogen chemistry and highlight the necessity of more stringent emission control of NOx to mitigate haze pollution.

Time Crystal in a Single-Mode Nonlinear Cavity.

Time crystal is a class of nonequilibrium phases with broken time-translational symmetry. Here, we demonstrate the time crystal in a single-mode nonlinear cavity. The time crystal originates from the self-oscillation induced by a linear gain and is stabilized by a nonlinear damping. We show in the time crystal phase there are sharp dissipative gap closing and pure imaginary eigenvalues of the Liouvillian spectrum in the thermodynamic limit. Dynamically, we observe a metastable regime with the emergence of quantum oscillation, followed by a dissipative evolution with a timescale much longer than the oscillating period. Moreover, we show there is a dissipative phase transition at the Hopf bifurcation, which can be characterized by the photon number fluctuation in the steady state. These results pave a new promising way for further experiments and deepen our understanding of time crystals.

Key drivers of the oxidative potential of PM2.5 in Beijing in the context of air quality improvement from 2018 to 2022.

The mass concentration of atmospheric particulate matter (PM) has been continuously decreasing in the Beijing-Tianjin-Hebei region. However, health endpoints do not exhibit a linear correlation with PM mass concentrations. Thus, it is urgent to clarify the prior toxicological components of PM to further improve air quality. In this study, we analyzed the long-term oxidative potential (OP) of water-soluble PM2.5, which is generally considered more effective in assessing hazardous exposure to PM in Beijing from 2018 to 2022 based on the dithiothreitol assay and identified the crucial drivers of the OP of PM2.5 based on online monitoring of air pollutants, receptor model, and random forest (RF) model. Our results indicate that dust, traffic, and biomass combustion are the main sources of the OP of PM2.5 in Beijing. The complex interactions of dust particles, black carbon, and gaseous pollutants (nitrogen dioxide and sulfur dioxide) are the main factors driving the OP evolution, in particular, leading to the abnormal rise of OP in Beijing in 2022. Our data shows that a higher OP is observed in winter and spring compared to summer and autumn. The diurnal variation of the OP is characterized by a declining trend from 0:00 to 14:00 and an increasing trend from 14:00 to 23:00. The spatial variation in OP of PM2.5 was observed as the OP in Beijing is lower than that in Shijiazhuang, while it is higher than that in Zhenjiang and Haikou, which is primarily influenced by the distribution of black carbon. Our results are of significance in identifying the key drivers influencing the OP of PM2.5 and provide new insights for advancing air quality improvement efforts with a focus on safeguarding human health in Beijing.

Fast Generation of GHZ-like States Using Collective-Spin XYZ Model.

The Greenberger-Horne-Zeilinger (GHZ) state is a key resource for quantum information processing and quantum metrology. The atomic GHZ state can be generated by one-axis twisting (OAT) interaction H_{OAT}=χJ_{z}^{2} with χ the interaction strength, but it requires a long evolution time χt=π/2 and is thus seriously influenced by decoherence and losses. Here we propose a three-body collective-spin XYZ model which creates a GHZ-like state in a very short timescale χt∼lnN/N for N particles. We show that this model can be effectively produced by applying Floquet driving to an original OAT Hamiltonian. Compared with the ideal GHZ state, the GHZ-like state generated using our model can maintain similar metrological properties reaching the Heisenberg-limited scaling, and it shows better robustness to decoherence and particle losses. This Letter opens the avenue for generating GHZ-like states with a large particle number, which holds great potential for the study of macroscopic quantum effects and for applications in quantum metrology and quantum information.

The Significant Role of New Particle Composition and Morphology on the HNO3-Driven Growth of Particles down to Sub-10 nm.

New particle formation and growth greatly influence air quality and the global climate. Recent CERN Cosmics Leaving OUtdoor Droplets (CLOUD) chamber experiments proposed that in cold urban atmospheres with highly supersaturated HNO3 and NH3, newly formed sub-10 nm nanoparticles can grow rapidly (up to 1000 nm h-1). Here, we present direct observational evidence that in winter Beijing with persistent highly supersaturated HNO3 and NH3, nitrate contributed less than ∼14% of the 8-40 nm nanoparticle composition, and overall growth rates were only ∼0.8-5 nm h-1. To explain the observed growth rates and particulate nitrate fraction, the effective mass accommodation coefficient of HNO3 (αHNO3) on the nanoparticles in urban Beijing needs to be 2-4 orders of magnitude lower than those in the CLOUD chamber. We propose that the inefficient uptake of HNO3 on nanoparticles is mainly due to the much higher particulate organic fraction and lower relative humidity in urban Beijing. To quantitatively reproduce the observed growth, we show that an inhomogeneous "inorganic core-organic shell" nanoparticle morphology might exist for nanoparticles in Beijing. This study emphasized that growth for nanoparticles down to sub-10 nm was largely influenced by their composition, which was previously ignored and should be considered in future studies on nanoparticle growth.

Health risks and sources of trace elements and black carbon in PM2.5 from 2019 to 2021 in Beijing.

A comprehensive health risk assessment of PM2.5 is meaningful to understand the current status and directions regarding further improving air quality from the perspective of human health. In this study, we evaluated the health risks of PM2.5 as well as highly toxic inorganic components, including heavy metals (HMs) and black carbon (BC) based on long-term observations in Beijing from 2019 to 2021. Our results showed that the relative risks of chronic obstructive pulmonary disease, lung cancer, acute lower respiratory tract infection, ischemic heart disease, and stroke decreased by 4.07%-9.30% in 2020 and 2.12%-6.70% in 2021 compared with 2019. However, they were still at high levels ranging from 1.26 to 1.77, in particular, stroke showed the highest value in 2021. Mn had the highest hazard quotient (HQ, from 2.18 to 2.56) for adults from 2019 to 2021, while Ni, Cr, Pb, As, and BC showed high carcinogenic risks (CR > 1.0×10-6) for adults. The HQ values of Mn and As and the CR values of Pb and As showed constant or slight upwards trends during our observations, which is in contrast to the downward trends of other HMs and PM2.5. Mn, Cr, and BC are crucial toxicants in PM2.5. A significant shrink of southern region sourcesof HMs and BCshrank suggests the increased importance of local sources. Industry, dust, and biomass burning are the major contributors to the non-carcinogenic risks, while traffic emissions and industry are the dominant contributors to the carcinogenic risks in Beijing.

The relationship between mobile phone addiction and time management disposition among Chinese college students：A cross-lagged panel model.

Previous research has identified a negative association between mobile phone addiction and time management disposition among college students; however, the direction of this relationship remains divergent. This study utilized a cross-lagged panel model to elucidate the directionality of the relationship between mobile phone addiction and time management disposition. A total of 466 college students completed two measures at seven-month intervals. The findings revealed a prevalence of mobile phone addiction at 10.94 % and 13.73 % in the two surveys. Notably, both mobile phone addiction and time management disposition demonstrated stability over time. Furthermore, a discernible negative bidirectional relationship was observed between the two. The present findings underscore the importance of timely intervention for college students facing challenges in mobile phone usage and time management.

Investigating the levels, spatial distribution, and trophic transfer patterns of short-chain chlorinated paraffins in the Southern Bohai Sea, China.

The marine environment of the southern Bohai Sea is severely polluted by short-chain chlorinated paraffins (SCCPs). To improve understanding of how SCCPs occur and of how they migrate, are transformed, and transferred in this area, we collected seawater, sediment, and organism samples, and determined the SCCP contents using a new approach based on high-resolution mass spectrometry. The ΣSCCP concentrations in the seawater, sediment, and organism samples ranged from 57.5 to 1150.4 ng/L, 167.7-1105.9 ng/g (dry weight), and 11.4-583.0 ng/g (wet weight), respectively. Simulation of the spatial distribution of SCCPs using Kriging interpolation showed that SCCPs were markedly influenced by land-based pollution. Substantial quantities of SCCPs were transported to the marine environment via surface runoff from rivers that passed through areas of major SCCP production. Once discharged from such rivers into the Bohai Sea, these SCCPs were further dispersed under the influence of ocean currents. Furthermore, the logarithmic bioaccumulation factor that varied from 2.12 to 3.20 and the trophic magnification factor that reached 5.60 (r2 = 0.750, p < 0.01) suggest that organisms have the ability to accumulate and biomagnify SCCPs through the food chain, which could potentially present risks to both marine ecosystems and human health.

Comparative analysis for the impacts of VOC subgroups and atmospheric oxidation capacity on O3 based on different observation-based methods at a suburban site in the North China Plain.

The persistent O3 pollution in the Beijing-Tianjin-Hebei (BTH) region remains unresolved, largely due to limited comprehension of O3-precursor relationship and photochemistry drivers. In this work, intraday O3 sensitivity evolution from VOC-limited (volatile organic compound) regime in the forenoon to transition regime in the late afternoon was inferred by relative incremental reactivity (RIR) in summer 2019 at Xianghe, a suburban site in BTH region, suggesting that VOC-focused control policy could combine with stringent afternoon NOx control. Then detailed impacts of VOC subgroups on O3 formation were further comprehensively quantified by parametric OH reactivity (KOH), O3 formation potential (OFP), as well as RIR weighted value and O3 formation path tracing (OFPT) approach based on photochemical box model. O3 episode days corresponded to stronger O3 formation, depicted by higher KOH (10.4 s-1), OFP (331.7 µg m-3), RIR weighted value (1.2), and F(O3)-OFPT (15.5 ppbv h-1). High proportions of isoprene and OVOCs (oxygenated VOCs) to the total KOH and the OFPT method were demonstrated whereas results of OFP and RIR-weighted presented extra great impacts of aromatics on O3 formation. The OFPT approach captured the process that has already happened and included final O3 response to the original VOC, thus reliable for replicating VOC impacts. The comparison results of the four methods showed similarities when utilizing KOH and OFPT methods, which reveals that the potential applicability of simple KOH for contingency VOC control and more complex OFPT method for detailed VOC- and source-oriented control during policy-making. To investigate propulsion of VOC-involved O3 photochemistry, atmospheric oxidation capacity (AOC) was quantified by two atmospheric oxidation indexes (AOI). Both AOIp_G (7.0 × 107 molec cm-3 s-1, potential AOC calculated by oxidation reaction rates) and AOIe_G (8.5 µmol m-3, estimated AOC given redox electron transfer for oxidation products) were stronger on O3 episode days, indicating that AOC promoted the radical cycling initiated from VOC oxidation and subsequent O3 production. Result-oriented AOIe_G reasonably characterized actual AOC inferred by good linear correlation between AOIe_G and O3 concentrations compared to process-oriented AOIp_G. Therefore, with continuous NOx abatement, AOIe_G should be considered to represent actual AOC, also O3-inducing ability.

Synthesis and functionalization of scalable and versatile 2D protein films via amyloid-like aggregation.

Two-dimensional (2D) protein films can be used to modify the properties of surfaces, and find applications predominantly in the fields of biomaterials, lithography, optics and electronics. However, it is difficult to produce scalable homogeneous and robust protein films with an easy, low-cost, green and efficient method. Further challenges include encapsulating and releasing functional building blocks in the film without inactivating them, and maintaining or improving the bioactivities of proteins used for the formation of the films. Here we detail the process to prepare large 2D protein films with user-defined features and structures via the amyloid-like aggregation of commonly synthesized proteins. These films can be synthesized at meter scales, have high interface adhesion, high functional expansibility and tunable functional properties, obtained by controlling the position of the disulfide bond breakage. For example, we can retain or even enhance the natural antibacterial, biomineralization and antifouling activity of proteins involved in film formation, and the properties can also be expanded through the physical blending or chemical grafting of additional functional blocks on the surface of the film. A 2D protein film can be prepared in ~3 h using four alternative coating techniques: immersion, transfer, hydrogel stamping and spraying. The characterization process of the film requires ~5 d. The procedure can be carried out by users with basic expertise in materials science.

A New Exopolysaccharide of Marine Coral-Associated Aspergillus pseudoglaucus SCAU265: Structural Characterization and Immunomodulatory Activity.

Recent studies have found that many marine microbial polysaccharides exhibit distinct immune activity. However, there is a relative scarcity of research on the immunomodulatory activity of marine fungal exopolysaccharides. A novel water-soluble fungal exopolysaccharide ASP-1 was isolated from the fermentation broths of marine coral-associated fungus Aspergillus pseudoglaucus SCAU265, and purified by Diethylaminoethyl-Sepharose-52 (DEAE-52) Fast Flow and Sephadex G-75. Structural analysis revealed that ASP-1 had an average molecular weight of 36.07 kDa and was mainly composed of (1→4)-linked α-D-glucopyranosyl residues, along with highly branched heteropolysaccharide regions containing 1,4,6-glucopyranosyl, 1,3,4-glucopyranosyl, 1,4,6-galactopyranosyl, T(terminal)-glucopyranosyl, T-mannopyranosyl, and T-galactopyranosyl residues. ASP-1 demonstrated significant effects on the proliferation, nitric oxide levels, and the secretion of cytokines TNF-α and IL-6 in macrophage RAW264.7 cells. Metabolomic analysis provided insights into the potential mechanisms of the immune regulation of ASP-1, suggesting its involvement in regulating immune function by modulating amino acid anabolism, particularly arginine synthesis and metabolism. These findings provide fundamental scientific data for further research on its accurate molecular mechanism of immunomodulatory activity.

[Influence of Ammonium Sulfate and Ammonium Nitrate on the Properties of Ambient PM2.5 in Zhenjiang].

Recently, the contribution of inorganic salts (nitrates in particular) to the mass concentration of particulate matter with an aerodynamic diameter of less than 2.5 µm (PM2.5) has been increasing across China. However, it is urgent to understand how the increased inorganic salts affect the crucial properties of PM2.5. Here, we conducted continuous field observations at Zhenjiang Ecology and Environment Protection Bureau from January 1 to December 31, 2021. The mass concentrations of ammonium sulfate[(NH4)2SO4] and ammonium nitrate (NH4NO3) were calculated using different methods. The contributions of (NH4)2SO4 and NH4NO3 to the extinction coefficient, hygroscopic growth, and acidity of PM2.5 were discussed in detail. Our results demonstrated that the mean mass concentrations of (NH4)2SO4 and NH4NO3 during the study period were (6.5±4.5) and (15.0±13.3) µg·m-3, which contributed (20.5±18.2)% and (34.5±18.4)% to the mass concentration of PM2.5, respectively. The total extinction coefficient of PM2.5 was (224.5±194.2) Mm-1, in which NH4NO3 was the largest contributor[(40.1±20.9)%] followed by (NH4)2SO4[(19.1±10.8)%]. (NH4)2SO4 and NH4NO3 were also the dominant contributors to the hygroscopic growth of PM2.5. In particular, NH4NO3contributed from (53.8±13.4)% to (61.6±14.6)% to the aerosol water content of PM2.5 under pollution conditions. Thus, NH4NO3 was a key air pollutant to be targeted for further improving the visibility and air quality in Zhenjiang in the future. However, the reduction in the precursors of NH4NO3 would lead to an increase in aerosol acidity, particularly in the spring and winter seasons. Our results help us understand the evolution of air quality and the related impacts and also provide important information on air quality improvement in Zhenjiang in the future.

Electrocatalytic oxidation of gaseous toluene in an all-solid cell using a foam Ti/Sb-SnO2/ß-PbO2 anode.

Mineralization of benzene, toluene, and xylene (BTX) with high efficiency at room temperature is still a challenge for the purification of indoor air. In this work, a foam Ti/Sb-SnO2/ß-PbO2 anode catalyst was prepared for electrocatalytically oxidizing gaseous toluene in an all-solid cell at ambient temperature. The complex Ti/Sb-SnO2/ß-PbO2 anode, which was prepared by sequentially deposing Sb-SnO2 and ß-PbO2 on a foam Ti substrate, shows high electrocatalytic oxidation efficiency of toluene (80%) at 7 hr of reaction and high CO2 selectivity (94.9%) under an optimized condition, i.e., a cell voltage of 2.0 V, relative humidity of 60% and a flow rate of 100 mL/min. The better catalytic performance can be ascribed to the high production rate of ⋅OH radicals from discharging adsorbed water and the inhibition of oxygen evolution on the surface of foam Ti/Sb-SnO2/ß-PbO2 anode when compared with the foam Ti/Sb-SnO2 anode. Our results demonstrate that prepared complex electrodes can be potentially used for electrocatalytic removal of gaseous toluene at room temperature with a good performance.

Reactive Chlorine Species Advancing the Atmospheric Oxidation Capacities of Inland Urban Environments.

Chlorine (Cl) radicals from photolabile chlorine species are highly reactive and can affect the fate of air pollutants in the atmosphere. Although several campaigns have been conducted, typically in coastal environments, long-term observations of reactive chlorine species and their impacts on atmospheric oxidation capacities (AOCs) are lacking. Here, we report nearly full-year observations of Cl2 and ClNO2 levels in Beijing and evaluate their impacts on the AOC with a box model coupled with Cl chemistry. Cl radicals promote the circulation of OH-HO2-RO2 by accelerating the OH chain lengths by up to 12.6% on average, hence boosting the AOC, especially in the winter or spring. This promotion effect is nonlinearly dependent on the VOC and NOx concentrations, thus leading to a slight shift in ozone formation from a VOC-sensitive regime to a transition regime with seasonal differences. Given the ubiquitous reactive chlorines in polluted inland urban regions, the AOCs and the formation of secondary pollutants will be underestimated if the reactive chlorine species are neglected.

Synthesis of robust underwater glues from common proteins via unfolding-aggregating strategy.

Underwater adhesive proteins secreted by organisms greatly inspires the development of underwater glue. However, except for specific proteins such as mussel adhesive protein, barnacle cement proteins, curli protein and their related recombinant proteins, it is believed that abundant common proteins cannot be converted into underwater glue. Here, we demonstrate that unfolded common proteins exhibit high affinity to surfaces and strong internal cohesion via amyloid-like aggregation in water. Using bovine serum albumin (BSA) as a model protein, we obtain a stable unfolded protein by cleaving the disulfide bonds and maintaining the unfolded state by means of stabilizing agents such as trifluoroethanol (TFE) and urea. The diffusion of stabilizing agents into water exposes the hydrophobic residues of an unfolded protein and initiates aggregation of the unfolded protein into a solid block. A robust and stable underwater glue can thus be prepared from tens of common proteins. This strategy deciphers a general code in common proteins to construct robust underwater glue from abundant biomass.

The role of NOx in Co-occurrence of O3 and PM2.5 pollution driven by wintertime east Asian monsoon in Hainan.

Clarifying the driving forces of O3 and fine particulate matter (PM2.5) co-pollution is important to perform their synergistic control. This work investigated the co-pollution of O3 and PM2.5 in Hainan Province using an observation-based model and explainable machine learning. The O3 and PM2.5 pollution that occurs in winter is affected by the wintertime East Asian Monsoon. The O3 formation shifts from a NOx-limited regime with a low O3 production rate (PO3) in the non-pollution season to a transition regime with a high PO3 in the pollution season due to an increase in NOx concentrations. Increased O3 and atmospheric oxidation capacity promote the conversion from gas-phase precursors to aerosols. Meanwhile, the high concentration of particulate nitrate favors HONO formation via photolysis, in turn facilitating O3 production. Machine learning reveals that NOx promotes O3 and PM2.5 co-pollution during the pollution period. The PO3 shows an upward trend at the observation site from 2018 to 2022 due to the inappropriate reduction of volatile organic compounds (VOCs) and NOx in the upwind areas. Our results suggest that a deep reduction of NOx should benefit both O3 and PM2.5 pollution control in Hainan and bring new insights into improving air quality in other regions of China in the future.