Application of machine learning to analyze ozone sensitivity to influencing factors: A case study in Nanjing, China.

Ground-level ozone (O3) has been an emerging concern in China. Due to its complicated formation mechanisms, understanding the effects of influencing factors is critical for making effective efforts on the pollution control. This study aims to present and demonstrate the practicality of a data-driven technique that applies a machine learning (ML) model coupled with the SHapley Additive exPlanations (SHAP) approach in O3 simulation and sensitivity analysis. Based on hourly measured concentrations of O3 and its major precursors, as well as meteorological factors in a northern area of Nanjing, China, a Light Gradient Boosting Machine (LightGBM) model was established to simulate O3 concentrations in different seasons, and the SHAP approach was applied to conduct in-depth analysis on the impacts of influencing factors on O3 formation. The results indicated a reliable performance of the ML model in simulating O3 concentrations, with the coefficient of determination (R2) between the measured and simulated larger than 0.80, and the impacts of influencing factors were reasonably evaluated by the SHAP approach on both seasonal and diurnal time scales. It was found that although volatile organic compounds (VOCs) and nitrogen oxides (NOx), as well as temperature and relative humidity, were generally the main influencing factors, their sensitivities to O3 formation varied significantly in different seasons and with time of the day. This study suggests that the data-driven ML model is a practicable technique and may act as an alternative way to perform mechanism analysis to some extent, and has immense potential to be applied in both problem research and decision-making for air pollution control.

Interface Engineering Enables Wide-Temperature Li-Ion Storage in Commercial Silicon-Based Anodes.

Silicon-based materials have been considered potential anode materials for next-generation lithium-ion batteries based on their high theoretical capacity and low working voltage. However, side reactions at the Si/electrolyte interface bring annoying issues like low Coulombic efficiency, sluggish ionic transport, and inferior temperature compatibility. In this work, the surface Al2 O3 coating layer is proposed as an artificial solid electrolyte interphase (SEI), which can serve as a physical barrier against the invasion of byproducts like HF(Hydrogen Fluoride) from the decomposition of electrolyte, and acts as a fast Li-ion transport pathway. Besides, the intrinsically high mechanical strength can effectively inhibit the volume expansion of the silicon particles, thus promoting the cyclability. The as-assembled battery cell with the Al2 O3 -coated Si-C anode exhibits a high initial Coulombic efficiency of 80% at RT and a capacity retention ratio up to ≈81.9% after 100 cycles, which is much higher than that of the pristine Si-C anode (≈74.8%). Besides, the expansion rate can also be decreased from 103% to 50%. Moreover, the Al2 O3 -coated Si-C anode also extends the working temperature from room temperature to 0 °C-60 °C. Overall, this work provides an efficient strategy for regulating the interface reactions of Si-based anode and pushes forward the practical applications at real conditions.

[Rapid prediction of flavonoid content in Epimedium sagittatum by infrared spectroscopy].

Epimedii Folium is a well-known Chinese herbal medicine with the effect of nourishing kidney and strengthening Yang. Its main active ingredients are flavonoids. In this study, 60 samples of Epimedium sagittatum were collected for the determination of total flavonoids(TF) including the total amount of epimedin A, epimedin B, epimedin C, and icariin(abbreviated as ABCI) specified in the Chinese Pharmacopoeia as well as rhamnosylicariside Ⅱ and icariside Ⅱ. The calibration parameters of "first derivativemultiva-riate scattering correction in 1 900-650 cm~(-1) band(4-point smoothing)" and "first derivativestandard normal variable correction in 4 000-650 cm~(-1) full band(4-point smoothing)" were confirmed respectively. The quantitative model was established via Fourier infrared spectroscopy plus attenuated total reflection(FTIR-ATR) accessory combined with partial least squares(PLS) method and then used to predict the flavonoid content of 11 validation sets. The average prediction accuracy for ABCI in calibration set and validation set was 98.985% and 96.087%, respectively. The average prediction accuracy for TF in calibration set and validation set was 98.998% and 94.771%, respectively. These results indicated that FTIR-ATR combined with PLS model could be used for rapid prediction of flavonoid content in E. sagittatum, with the prediction accuracy above 94.7%. The establishment of this method provides a new solution for the detection of a large number of E. sagittatum samples.

[Content correlation of eight phenolic acids and flavonoids in different medicinal parts of PA-type Perilla germplasms].

In this study, 10 PA-type Perilla germplasms were selected to detect the content of two phenolic acids, i.e., rosmarinic acid(RA) and caffeic acid(CA), and six flavonoids, including scutellarin-7-O-diglucuronoside(SDG), luteolin-7-O-diglucuronoside(LDG), apigenin-7-O-diglucuronoside(ADG), scutellarin-7-O-glucuroside(SG), luteolin-7-O-glucuroside(LG), and apigenin-7-O-glucuroside(AG) in leaves, stems, and fruits. The total content of phenolic acids and flavonoids in leaves was 3.991-12.028 mg·g~(-1) and 12.309-25.071 mg·g~(-1), respectively, which was much higher than that in stems(0.586-2.015 mg·g~(-1) and 0.879-1.413 mg·g~(-1), respectively) and fruits(0.004-2.222 mg·g~(-1) and 0.651-1.936 mg·g~(-1), respectively). RA was detected in five fruit samples, and RA content between leaves and fruits showed a significant negative correlation in the other five samples. For flavonoids, only LG and LDG could be detected in stems, and SG and SDG were not detected in fruits, while other flavonoids were not detected in some samples. The content of total flavonoids and LG in leaves and fruits was significantly positively correlated, and the content of LG in stems and fruits was significantly positively correlated. In 10 stem samples, seven met the standard that the content of RA in the stem should be not less than 0.1% specified in the Chinese Pharmacopoeia(2020 edition). Only one fruit sample reached the standard of RA content in the fruit not less than 0.25% specified in the Chinese Pharmacopoeia.

[Study on morphological classification and chemical-type of Perilla frutescens cultivated germplasm].

Fifty cultivated Perilla seeds were collected all over the country and planted in Beijing experiment field for morphology and chemical-type researches. Twenty morphological characteristics were selected and observed, and the essential oil from leaves was extracted by steam distillation and analyzed by GC-MS to confirm chemical-types. There were significant diversities in plant height, leaf color and morphology, and fruit color and weight. Clustering analysis was carried out based on these morphological characteristics. Six types were divided with their chemical-type designated. Type Ⅰ: Six germplasms, attributed to P. frutescens var. crispa, with dwarf plants, thin creased purple leaf, named Crispa, their chemical types were diversified, including EK, PAPK, PA and PK. Type Ⅱ: Six germplasms, attributed to P. frutescens var. crispa, plants were taller than type I and with thin and creased green leaf, named Big Crispa, all PK type. Type Ⅲ: Seventeen germplasms, attributed to P. frutescens var. frutescens with leaf color upside green and underside purple, tall plant and wide distribution all over the China, named Ordinary Frutescens, all PK. Type Ⅳ: Four germplasms, attributed to P. frutescens var. acuta with tall plant and small seed, named Acuta, all PK. Type Ⅴ: Seven germplasms, attributed to P. frutescens var. frutescens with green leaves, tall plants and long clusters, named Long-spike Frutescens, all PK. Type Ⅵ: Ten germplasms, attributed to P. frutescens var. frutescens with big, thick and creased leaf, named Thick-leaf Frutescens, including PK, PP, PL and PA. The morphological classification of this paper would lay the foundation for the taxonomic naming and following evaluation of the Perilla germplasm resources.This study also showed that there was no correspondence but a certain correlation between volatile oil chemical-types and subspecies classification and morphological characteristics of Perilla.

[Morphological characteristics of Chinese Perilla fruits].

The morphological traits of 55 Chinese Perilla fruit samples (size, 100 grains weight, color, hardness, surface ridge height) are described and the statistically analyzed. It can be divided into 6 categories by cluster analysis, namely: Ⅰ, big grain (diameter 1.5 mm above and 100 grains weight above 0.16 g), low ridge, hard; Ⅱ, big grain, low ridge, soft; Ⅲ, big grain, high ridge, soft, fruit; Ⅳ, big grain. high ridge, gray brown or dark brown; Ⅴ, small grain (diameter 1.5 mm below and 100 grain weight 0.16 g below), low ridge, hard, dark brown; Ⅵ, small grain, low ridge, hard, yellow brown. The 38 fruit samples were planted, among which 31 ones were P. frutescens var. frutescens, 4 ones P. frutescens var. crispa and 3 ones P. frutescens var. acuta. By chemotype classification, they were 29 PK type, 3 PA type, 2 PL type, 2 PP type, 1 EK type and 1 PAPK type. According the description of herb Perillae Fructus in China Pharmacopoeia, the plant originates from P. frutescens var. frutescens. In contrast, not all fruits of P. frutescens var. frutescens have accord features. The fruits with white pericarp are mainly from P. frutescens var. frutescens with purple leaves. The materials with small grain, low ridge, hard, yellow brown or dark brown, are likely to be PA type and mainly P. frutescens var. crispa.

[Research on effects of chemotype and components of Perilla frutescens leaf volatile oil I: different phenological periods].

This experiment researched on three kinds of Perilla frutescens including the widespread PK, PA and rare PL chemotype. The Perilla samples were the mature leaves collected in nutrition, flowering and frutescence three different phenological periods, and at 7 am, 12 pm and 6 pm three day time. The volatile oil was extracted by steam distillationand analyzed by GC-MS, as a result, the three chemotype samples'volatile oil yield was between 0.08% and 0.96%; volatile oil yield of different growth period was as follow: nutrition>flowering>fructescence, and the volatile oil yield of nutrition period: PA type>PK type>PL type. Each chemotype was not affected by the growth and development, indicating that the chemotype is determined by genetic factors. Characteristic and main components of PA and PK type are relatively stable, and the characteristic components of PL type are significantly decreased with the growth. There are still a large number of upstream metabolism components, and the chemical type may have their primitiveness and changeability. The relative content of perillaldehyde, characteristic components of PA type, is basically decreased from morning to night, in all the period. The relative content of perillaketone, characteristic components of PK type, in nutrition and flowering period, when samples were collected at 12 noon is relatively higher than that at 7 am and 6 pm, and contrary to samples collected in frutescence period. The relative content of perillene, characteristic components of PL type, in nutrition and frutescence period are highest at 12 noon, while in flowering period is highest at 6 pm. According to the volatile oil yield and relative content of maincomponents, the best harvest time of PA type is in the morning of the nutrition period; the best harvest time of PK type is in the morning of all the period; and the best harvest time of PL type is at dusk of the nutrition period.

Combined use of ionic liquid and sulfobutylether-ß-cyclodextrin as electrolyte additives for separation and determination of camptothecin alkaloids by CZE.

This work reported that ionic liquid (IL) ([Bmim] [PF6 ]) and sulfobutylether-ß-CD (SBE-ß-CD) were used as electrolyte additives for the separation and determination of camptothecin (CPT) alkaloids by CZE. Separation parameters such as the buffer type, pH, and concentration of the running buffer, the concentration of SBE-ß-CD and IL, temperature, and separation voltage were all investigated in order to achieve the maximum possible resolution. The four analytes were baseline separated within 10 min in capillary at the separation voltage of 15 kV with a running buffer consisting of 20 mM borate buffer, 20 mM IL, and 100 mM SBE-ß-CD at pH 9.0. Under such conditions, good linearity about two orders of magnitudes of peak areas was achieved for the investigated CPT alkaloids with the correlation coefficients ranging from 0.9946 to 0.9985. For all analytes, detection limits (S/N = 3) and quantitation limits (S/N = 10) range from 0.05 to 0.92 µg/mL and 0.17 to 3.06 µg/mL, respectively. The proposed method has not only been successfully applied to the separation and determination of CPT alkaloids but also showed that IL seemed to be a promising additive in CZE separation.

[Perilla resources of China and essential oil chemotypes of Perilla leaves].

This study, based on the findings for Perilla resources, aimed to describe the species, distribution, importance, features, utilization and status of quantitative Perilla resources in China. This not only helps people to know well about the existing resources and researching development, but also indicates the overall distribution, selection and rational use of Perilla resource in the future. According to the output types, Perilla resources are divided into two categories: wild resources and cultivated resources; and based on its common uses, the cultivated resources are further divided into medicine resources, seed-used resources and export resources. The distribution areas of wild resources include Henan, Sichuan, Anhui, Jiangxi, Guangxi, Hunan, Jiangsu and Zhejiang. The distribution areas of medicine resources are concentrated in Hebei, Anhui, Chongqing, Guangxi and Guangdong. Seed-used resources are mainly distributed in Gansu, Heilongjiang, Jilin, Chongqing and Yunnan. Export resource areas are mainly concentrated in coastal cities, such as Zhejiang, Jiangsu, Shandong and Zhejiang. For the further study, the essential oil of leaf samples from different areas were extracted by the steam distillation method and analyzed by GC-MS. The differences in essential oil chemotypes among different Perilla leaves were compared by analyzing their chemical constituents. The main 31 constituents of all samples included: perillaketone (0.93%-96.55%), perillaldehyde (0.10%-61.24%), perillene (52.15%), caryophyllene (3.22%-26.67%), and α-farnesene (2.10%-21.54%). These samples can be classified into following five chemotypes based on the synthesis pathways: PK-type, PA-type, PL-type, PP-type and EK-type. The chemotypes of wild resources included PK-type and PA-type, with PK-type as the majority. All of the five chemotypes are included in cultivated resources, with PA-type as the majority. Seed-used resources are all PK-type, and export resources are all PA-type. The P. frutescens var. frutescens include five chemotypes, with PK-type as the majority. The PK-type leaves of P. frutescens var. acuta are green, while the PA-type leaves are reddish purple. The P. fruteseens var. crispa was mainly PA type with reddish purple leaves. The differences of the main chemotypes provide a scientific basis for distinguishing between Zisu and Baisu in previous literatures. Based on the lung toxicity of PK and the traditional use of Perilla, the testing standard of essential oil and Perilla herb shall be built, and PA type is recommended to be used in traditional Chinese medicine.