Feature engineering for improved machine-learning-aided studying heavy metal adsorption on biochar.

Due to the broad interest in using biochar from biomass pyrolysis for the adsorption of heavy metals (HMs) in wastewater, machine learning (ML) has recently been adopted by many researchers to predict the adsorption capacity (η) of HMs on biochar. However, previous studies focused mainly on developing different ML algorithms to increase predictive performance, and no study shed light on engineering features to enhance predictive performance and improve model interpretability and generalizability. Here, based on a dataset widely used in previous ML studies, features of biochar were engineered-elemental compositions of biochar were calculated on mole basis-to improve predictive performance, achieving test R2 of 0.997 for the gradient boosting regression (GBR) model. The elemental ratio feature (H-O-2N)/C, representing the H site links to C (non-active site to HMs), was proposed for the first time to help interpret the GBR model. The (H-O-2N)/C and pH of biochar played essential roles in replacing cation exchange capacity (CEC) for predicting η. Moreover, expanding the coverages of variables by adding cases from references improved the generalizability of the model, and further validation using cases without CEC and specific surface area (R2 0.78) and adsorption experimental results (R2 0.72) proved the ML model desirable. Future studies in this area may take into account algorithm innovation, better description of variables, and higher coverage of variables to further increase the model's generalizability.

Machine learning for hydrothermal treatment of biomass: A review.

Hydrothermal treatment (HTT) (i.e., hydrothermal carbonization, liquefaction, and gasification) is a promising technology for biomass valorization. However, diverse variables, including biomass compositions and hydrothermal processes parameters, have impeded in-depth mechanistic understanding on the reaction and engineering in HTT. Recently, machine learning (ML) has been widely employed to predict and optimize the production of biofuels, chemicals, and materials from HTT by feeding experimental data. This review comprehensively analyzed the application of ML for HTT of biomass and systematically illustrated basic ML procedure and descriptors for inputs and outputs of ML models (e.g., biomass compositions, operation conditions, yield and physicochemical properties of derived products) that could be applied in HTT. Moreover, this review summarized ML-aided HTT prediction of yield, compositions, and physicochemical properties of HTT hydrochar or biochar, bio-oil, syngas, and aqueous phase. Ultimately, future prospects were proposed to enhance predictive performance, mechanistic interpretation, process optimization, data sharing, and model application during ML-aided HTT.

Machine learning assisted predicting and engineering specific surface area and total pore volume of biochar.

Biochar produced from pyrolysis of biomass is a platform porous carbon material that have been widely used in many areas. Specific surface area (SSA) and total pore volume (TPV) are decisive to biochar application in hydrogen uptake, CO2 adsorption, and organic pollutant removal, etc. Engineering biochar by traditional experimental methods is time-consuming and laborious. Machine learning (ML) was used to effectively aid the prediction and engineering of biochar properties. The prediction of biochar yield, SSA, and TPV was achieved via random forest (RF) and gradient boosting regression (GBR) with test R2 of 0.89-0.94. ML model interpretation indicates pyrolysis temperature, biomass ash, and volatile matter were the most important features to the three targets. Pyrolysis parameters and biomass mixing ratios for biochar production were optimized via three-target GBR model, and the optimum schemes to obtain high SSA and TPV were experimentally verified, indicating the great potential of ML for biochar engineering.

Machine learning prediction of nitrogen heterocycles in bio-oil produced from hydrothermal liquefaction of biomass.

Hydrothermal liquefaction (HTL) of high-moisture biomass or biowaste to produce bio-oil is a promising technology. However, nitrogen-heterocycles (NH) presence in bio-oil is a bottleneck to the upgrading and utilization of bio-oil. The present study applied the machine learning (ML) method (random forest) to predict and help control the bio-oil NH, bio-oil yield, and N content of bio-oil (N_oil). The results indicated that the predictive performance of the yield and N_oil were better than previous studies, achieving test R2 of 0.92 and 0.95, respectively. Acceptable predictive performance (test R2 of 0.82 and RMSE of 7.60) for the prediction of NH was also achieved. The feature importance analysis, partial dependence, and Shapely value were used to interpret the prediction models and study the NH formation mechanisms and behavior. Then, forward optimization of NH was implemented based on optimal predictive models, indicating the high potential of ML-aided bio-oil production and engineering.

Machine learning predicting wastewater properties of the aqueous phase derived from hydrothermal treatment of biomass.

Hydrothermal treatment (HTT) is a potential technology for producing biofuel from wet biomass. However, the aqueous phase (AP) is generated inevitably in the process of HTT, and studies are lacking on the detailed exploration of AP properties. Therefore, machine learning (ML) models were built for predicting the pH, total nitrogen (TN), total organic carbon (TOC), and total phosphorus (TP) of the AP based on biomass feedstock and HTT parameters. Results showed that the gradient boosting decision tree (average testing R2 0.85-0.96) can accurately predict the above wastewater properties for both single- and multi-target models. ML-based feature importance indicated that nitrogen content of biomass, solid content, and temperature were the top three critical features for pH, TN, and TP, while those for TOC were reaction time, lipid, and temperature. This ML approach provides new insights to understand the formation and properties of the HTT AP by ML rather than time-consuming experiments.

Co-liquefaction of Chlorella and soybean straw for production of bio-crude: Effects of reusing aqueous phase as the reaction medium.

Co-liquefaction was combined with hydrothermal liquefaction (HTL) aqueous phase (AP) recirculation to improve the practicality of HTL process. The Chlorella powder (CL), soybean straw (SS), and their mixture (CS) with ratio 1:1 were processed at 300 °C for 20 min, and the AP was recirculated four times. The yield of CS bio-crude was increased (from 24.28% to 31.83%) by co-liquefaction, but remained stable during AP recirculation. By contrast, the yields were increased for CL bio-crude (from 32.40% to 41.19%), SS hydrochar (from 19.55% to 30.88%), and CS hydrochar (from 9.42% to 14.76%) by recirculation. The elemental analysis, chemical composition analysis, functionality analysis, thermogravimetric analysis, and verification experiments (HTL with model AP components) show the N-containing compounds (e.g., amines) in AP were converted into amides (acylation) for CL bio-crude, into N-heterocycles (Maillard reactions) for CS hydrochar, and into Mannich bases for SS hydrochar, which contributed to the increased yield and N content (from 7.27% to 8.82% for CL bio-crude). Furthermore, the O content of CS bio-crude was decreased (from 15.31% to 12.52%) by recirculation, resulted from the conversion of N-heterocyclic ketones into pyrazine derivates. The decreased O content and comprehensive combustibility index (from 0.306 to 0.177) of CS bio-crude indicate the great potential of this craft combination.

Machine learning prediction and optimization of bio-oil production from hydrothermal liquefaction of algae.

Hydrothermal liquefaction (HTL) of algae is a promising biofuel production technology. However, it is always difficult and time-consuming to identify the best optimal conditions of HTL for different algae by the conventional experimental study. Therefore, machine learning (ML) algorithms were applied to predict and optimize bio-oil production with algae compositions and HTL conditions as inputs, and bio-oil yield (Yield_oil), and the contents of oxygen (O_oil) and nitrogen (N_oil) in bio-oil as outputs. Results indicated that gradient boosting regression (GBR, average test R2 âˆ¼ 0.90) exhibited better performance than random forest (RF) for both single and multi-target tasks prediction. Furthermore, the model-based interpretation suggested that the relative importance of operating conditions (temperature and residence time) was higher than algae characteristics for the three targets. Moreover, ML-based reverse and forward optimizations were implemented with experimental verifications. The verifications were acceptable, showing great potential of ML-aided HTL for producing desirable bio-oil.

Co-culture of fungi-microalgae consortium for wastewater treatment: A review.

The treatment of wastewater by microalgae has been studied and proved to be effective through previous studies. Due to the small size of microalgae, how to efficiently harvest microalgae from wastewater is a crucial factor restricting the development of algal technologies. Fungi-assisted microalgae bio-flocculation for microalgae harvesting and wastewater treatment simultaneously, which was overlooked previously, has attracted increasing attention in the recent decade due to its low cost and high efficiency. This review found that fungal hyphae and microalgae can stick together due to electrostatic neutralization, surface protein interaction, and exopolysaccharide adhesion in the co-culture process, realizing co-pelletization of microalgae and fungi, which is conducive to microalgae harvesting. Besides, the combination of fungi and microalgae has a complementary effect on pollutant removal from wastewaters. The co-culture of fungi-microalgae has excellent development prospects with both environmental and economic benefits, and it is expected to be applied on an industrial scale.

A review on nitrogen transformation in hydrochar during hydrothermal carbonization of biomass containing nitrogen.

Biomass is a type of renewable and sustainable resource that can be used to produce various fuels, chemicals, and materials. Nitrogen (N) in biomass such as microalgae should be reduced if it is used to produce fuels, while the retention of N is favorable if the biomass is processed to yield chemicals or materials with N-containing functional groups. The engineering of the removal and retention of N in hydrochar during hydrothermal carbonization (HTC) of biomass rich in protein is a research hot spot in the past decade. However, the N transformation during HTC has not yet been fully understood. In order to mediate the migration and transformation of N in hydrochar, the present review overviewed i) the characteristics of hydrochar and the original feedstock, ii) the possible N transformation behavior and mechanisms, and iii) the effect of factors such as feedstock and pyrolysis parameters such as temperature on hydrochar N. The high temperature and high protein content promote the dehydration, decarboxylation, and deamination of biomass to produce hydrochar solid fuel with reduced N content, while the Millard and Mannich reactions for lignocellulosic biomass rich in carbohydrate (cellulose, hemicellulose, and lignin) at medium temperatures (e.g., 180-240 °C) significantly promote the enrichment of N in hydrochar. The prediction models can be built based on properties of biomass and the processing parameters for the estimation of the yield and the content of N in hydrochar.