Next Article in Journal
Implementing the Circular Economy in the European Union and Spain: Links to the Low-Carbon Transition
Previous Article in Journal
Research on the Mechanism and Prediction Model of Pressure Drive Recovery in Low-Permeability Oil Reservoirs
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Energies
Volume 17
Issue 21
10.3390/en17215254
energies-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Review

A Review on Machine Learning-Aided Hydrothermal Liquefaction Based on Bibliometric Analysis

by
Lili Qian
1,*,
Xu Zhang
1,
Xianguang Ma
1,
Peng Xue
1,
Xingying Tang
2,
Xiang Li
3 and
Shuang Wang
1
1
School of Energy and Power Engineering, Jiangsu University, Zhenjiang 212013, China
2
Guangxi Key Laboratory on the Study of Coral Reefs in the South China Sea, School of Marine Sciences, Guangxi University, Nanning 530004, China
3
Taizhou DongBo New Materials Co., Ltd., Taizhou 225312, China
*
Author to whom correspondence should be addressed.
Energies 2024, 17(21), 5254; https://doi.org/10.3390/en17215254
Submission received: 10 August 2024 / Revised: 16 October 2024 / Accepted: 18 October 2024 / Published: 22 October 2024
(This article belongs to the Section B: Energy and Environment)
Browse Figures
Versions Notes

Abstract

:
Hydrothermal liquefaction (HTL) is an effective biomass thermochemical conversion technology that can convert organic waste into energy products. However, the HTL process is influenced by various complex factors such as operating conditions, feedstock properties, and reaction pathways. Machine learning (ML) methods can utilize existing HTL data to develop accurate models for predicting product yields and properties, which can be used to optimize HTL operation conditions. This paper presents a bibliometric review on ML applications in HTL from 2020 to 2024. CiteSpace, VOSviewer, and Bibexcel were used to analyze seven key bibliometric attributes: annual publication output, author co-authorship networks, country co-authorship networks, co-citation of references, co-citation of journals, collaborating institutions, and keyword co-occurrence networks, as well as time zone maps and timelines, to identify the development of ML in HTL research. Through the detailed analysis of co-occurring keywords, this study aims to identify frontiers, research gaps, and development trends in the field of ML-aided HTL.

1. Introduction

Contemporary society is confronted with severe challenges, such as energy shortages, increasing demand, environmental impacts, and climate change [1,2]. As our reliance on energy rises substantially, developing sustainable and environmentally friendly energy solutions has become increasingly urgent. Biomass energy has garnered significant attention as a renewable and sustainable alternative energy source [3]. Hydrothermal liquefaction (HTL) can directly convert biomass such as algae, sludge, and lignocellulosic materials into biocrude with high energy density, thus avoiding the need for energy-intensive drying processes [4]. HTL could effectively address energy shortage issues. However, due to the complexity and non-linearity of biomass conversion, optimizing the HTL process to improve biocrude yield and quality presents a considerable challenge.
The development of machine learning (ML) technologies has provided researchers in the HTL field with advanced tools for predicting and optimizing HTL parameters. ML algorithms such as Artificial Neural Networks (ANN), Random Forest (RF), Support Vector Machine (SVM), Gaussian Process Regression (GPR), Generalized Additive model (GAM), Extreme Gradient Boosting (XGB) and Gradient Boosting Regression (GBR) have been proven particularly effective in predicting product characteristics and optimizing reaction conditions. The development of these models offers the potential to enhance biomass conversion efficiency and advance environmentally sustainable HTL technologies [5,6]. ML has made impressive progress in HTL research, spanning a wide range of areas from predicting biocrude yield and composition to optimizing reaction kinetics and exploring novel catalysts. This interdisciplinary approach has garnered extensive attention, as evidenced by the notable increase in publications. Bibliometric analysis is a method that uses available literature data to map the knowledge landscape of a specific field. This approach examines the quantity and quality of publications to assess the research development and identify unexplored areas. Many studies have utilized bibliometric analysis in various academic fields through visualization patterns. Elgarahy et al. [7] conducted bibliometric mapping to identify the latest studies using “bioenergy/biofuel” as the keywords. Ocampo et al. [8] used bibliometric methods to review the growth of literature, research trends, keywords, and collaboration relationships among different countries on HTL. Bassoli et al. [9] conducted bibliometric analysis on HTL of algae for biocrude production, examining research progress, trends, and future directions. Yang et al. [10] bibliometrically analyzed the keywords from literature on biomass HTL from 2006 to 2018. The keywords of “biomass”, “biofuel”, “microalga”, “catalyst”, and “reaction condition” appeared with the highest frequency.
However, the existing bibliometric studies have analyzed the HTL field as a whole, with no specific bibliometric reviews focusing on the application of ML algorithms in HTL research. This paper provides a review of ML-assisted HTL research from 2020 to 2024. CiteSpace, VOSviewer, and Bibexcel were used to analyze scientific literature datasets, visualizing the bibliometric parameters. Annual publication outputs, contributions from authors, countries, institutions, and journals, as well as keyword analysis were comprehensively reviewed. These visualizations will aid in understanding the progress and research directions of ML applications in HTL.

2. Methodology

2.1. Data Collection and Processing

The research data were sourced from the Web of Science and Google Scholar. Initially, an advanced search was conducted on the Web of Science using the keywords “machine learning” and ”hydrothermal liquefaction”, which yielded only a few dozen results with limited relevant articles. In contrast, Google Scholar returned over three hundred relevant papers. The majority of these papers were from Elsevier, ACS, and Springer. Given the limited number of papers before 2020, the time span for this review was set from 2020 to 2024, with the search conducted on 20 June 2024. A total of 366 journal papers were retrieved from the Google Scholar database. To improve the accuracy of data sources and the relevance of the literature, the 366 articles were further screened to exclude those unrelated to the topic, resulting in 152 valid papers as shown in the Supplementary Material. These valid articles were then exported in plain text format and processed with CiteSpace 6.2.R4 and VOSviewer 1.6.20 for visual analysis, while Bibexcel, Version 2016-02-20 was used to analyze the current state of research, progress, and trends in the field of ML-aided HTL.

2.2. Bibliometric Visualization Analysis

The dataset was obtained in plain text format and imported into CiteSpace 6.2.R4 and VOSviewer. CiteSpace was developed by Dr. Chen from Drexel University, PA, USA [11], while VOSviewer was developed by Van Eck and Waltman from the Centre for Science and Technology Studies (CWTS) at Leiden University, The Netherlands [12]. Both software tools were developed using the Java programming language. After the data files were imported, parameters such as time range, titles, abstracts, authors, citations, and countries were selected. These data were then analyzed to generate knowledge maps with bibliometric attributes, including annual publication output, author co-authorship networks, country co-authorship networks, co-citation networks of references, co-citation networks of journals, collaborating institutions, keyword co-occurrence networks, time zone maps, and timeline.

3. Results and Discussion

3.1. Annual Publication Output Analysis

After the screening of papers related to ML-assisted HTL, 152 relevant papers were identified. Figure 1 shows the publications from 2020 to June 2024. The earliest paper was published by the Colosi et al. [13] from USA. ML methods were combined with a life cycle assessment method to predict the product fraction yield from hydrothermal treatment. It was found that RF had better predictive accuracy than multiple linear regression and regression tree algorithms.
As shown in Figure 1, a total of 3 papers were published in 2020 and 11 papers in 2021. Subsequently, the publication in this field experienced rapid growth. In 2022, 35 papers were published, and in 2023, 54 papers were published. In the first half of 2024 alone, the number of publications has already surpassed the total number for 2022, with 44 papers. The noticeable increasing trend in publication indicates that the ML-aided HTL is an important research interest. The most recent article was published by the Li et al [14]. They applied Automated Machine Learning (AutoML) methods to predict four gases (CO2, CH4, CO, and H2) produced during the HTL process. The AutoML method could automatically perform hyperparameter optimization and train multiple single-target models. The study found that the Gradient Boosting Machine model provided the best prediction performance, achieving a testing R2 of over 0.86.

3.2. Journal Contribution Analysis

Table 1 shows the top six journals with the most publications on ML-assisted HTL from 2020 to 2024, along with their respective journals. It was observed that in recent years, most of the journals belonged to the energy sector. Among them, the journals with 10 or more publications were Bioresource Technology and Fuel, with 18 and 15 papers, respectively.
The citation data for journals reveal that among the journals listed in Table 1, the most frequently cited journal is the Chemical Engineering Journal, with an average citation count of 47. Three articles published after 2023 had an average citation of 17. For the two journals with the most publications, Bioresource Technology had an average citation of 32, with 7 papers published after 2023 having an average citation of 20. This indicates Bioresource Technology paid significant attention to ML-assisted HTL. Among the 18 papers, there were 11 experimental papers and 7 review papers. Fuel has an average citation of 22.2 for its paper, with 11 papers published after 2023 having an average citation of 17. This journal included 11 experimental papers and 4 review papers.

3.3. Citation Analysis

Table 2 shows the top five papers with the highest citations on ML-aided HTL. The top 2 most highly cited papers were published by the Leng et al. [15,16] in 2021, with the citation of 100 and 97, respectively. Table 2 also includes 1 paper published after 2023 by the Naqvi et al. [17], which has 90 citations, ranking it fourth. This study was a review paper on the application of ML in thermochemical conversion.
The most-cited paper by Leng et al. [15] established multi-task ML models of Decision Regression Tree (DRT), RF, and GBR to achieve biocrude with high yield and low nitrogen content. More importantly, ML-based reverse optimization was performed to guide the high-quality biocrude production. This indicated that the protein content of 27.67%, lipid content of 43.72%, carbohydrate content of 19.90%, temperature of 303 °C, and reaction time of 45 min were favorable. The second most-cited paper focused on algae biomass and predicted the yield, oxygen content, and nitrogen content of algae bio-oil. GBR performed better in both single-target and multi-target tasks, with a training R2 reaching up to 0.90. In comparison to the characteristics of algae, temperature and residence time had a greater impact on bio-oil yield, oxygen content, and nitrogen content [16].

3.4. Network of Countries Analysis

Figure 2 shows the publications and total citations on ML-aided HTL classified by countries. The bar chart represents the number of publications, while the line chart represents total citations.
Figure 2 clearly demonstrates the contribution and impact of each country in the field of ML-aided HTL. The 152 papers used in this study were from 32 countries. As shown in Figure 2, China was the leading country in terms of publications, with 42 papers, accounting for 27.6% of the total, and a total of 726 citations. India was the second highest, with 21 papers, representing 13.8% of the total, and 297 citations. The third-ranked country was the United States, with 11 papers, accounting for 7.2% of the total, and 216 citations. China, India, and the United States, due to their large populations and higher energy demands, showed greater interest in ML-aided HTL.
Figure 3 shows the cluster co-authorship relationships among countries. The size of the circular nodes represents the number of publications, with larger nodes indicating a larger number of publications. The lines between nodes represent the strength of collaboration, with thicker lines indicating closer cooperation. The colors of the nodes represent different clusters. Judging from Figure 3, it is evident that many countries have published papers in ML-aided HTL, as ML can optimize HTL reaction conditions at a low cost, improve experimental efficiency, and help countries to address their energy demands and environmental challenges. As shown in Figure 3, China has closer collaborations with Canada and Singapore and also maintains a certain level of cooperation with other Asian, Middle Eastern, and European countries. India has closer ties with South Korea and the United States, while the United States has stronger collaborations with Thailand and India.

3.5. Co-Authorship Networks and Citation Analysis

Table 3 lists the authors and their publications and citations in ML-aided HTL. Leng and Li, Hailong each published 9 papers with a total of 256 citations, as they have been collaborating in this field for many years. Peng and Zhang each published 8 papers with a total of 242 citations, and they have also been collaborating in this field. Li, Jie published 5 papers with 201 citations, while Liu and Tippayawong each published 4 papers, with 180 and 98 citations, respectively. Yang, Wang, Gupta, and Katongtung each published 3 papers, with the citations of 180, 100, 112, and 69, respectively. These authors have made significant contributions to the development of this field.
Figure 4 shows co-authorship networks. Leng and Li were the top two authors with the highest citation counts and publication numbers. There was a strong connection between these two authors in the co-authorship network, with many papers being jointly authored by them [5,15,20,21,22,23,24,25,26,27,28,29,30,31,32,33]. Among these co-authored papers, Leng was the first author of nine papers, while Li was the first author of two papers. In Figure 4, the largest blue cluster representing Xu, the largest green cluster representing Li, and the largest yellow cluster representing Liu showed close collaboration with the largest red cluster representing Leng. A total of four papers were jointly authored by these four clusters [15,17,21,22]. This indicates that Leng and Li have made significant contributions to ML-aided HTL. Additionally, some scholars who are not shown in Figure 4 have also made substantial contributions to this field. For example, as shown in Table 3, the Tippayawong’s group from Thailand published four papers with a total of 98 citations. Katongtung, one member of this group was also published a paper as the first author in 2022 and another in 2023. In 2022, Katongtung et al. used 325 sets of data from HTL of wet biomass and developed Extreme Gradient Boosting (XGB), RF, Kernel Ridge Regression (KRR), and Support Vector Regression (SVR) models to predict the yield and heating value of biocrude. Among these models, the XGB model performed the best, achieving testing R2 values of 0.87 and 0.81 for biocrude yield and heating value, respectively [19]. In 2023, Katongtung et al. combined principal component analysis and ML to predict biocrude yield from biomass [34].

3.6. Co-Occurring Keyword Analysis

Figure 5 is a keyword time zone map from CiteSpace. The purple, cyan, green, and red diamonds represent the years of 2021, 2022, 2023, and 2024, respectively. The keywords in 2020 are not depicted since only two papers were available. The keywords that appeared in 2021 will be overlaid with diamonds of different colors in the subsequent years where they appeared, stacked on the year of their first appearance.
From Figure 5, it can be observed that from 2021 to 2024, “machine learning”, “hydrothermal liquefaction”, “model”, “biomass”, “microalgae”, and “co-liquefaction” have been consistently key topics of discussion. Regarding “microalgae”, since microalgae was the most important feedstock of HTL process, the data from algal biomass were frequently utilized to develop ML models. Concerning “co-liquefaction”, in 2021, Bhaskar et al. [35] used ANN model to predict yields of biocrude, aqueous phase product, biochar, gas, and energy, as well as carbon recovery rate from co-HTL of algae and lignocellulosic biomass. In 2022, “random forest”, “artificial intelligence”, “extreme gradient boosting”, and “gaussian process regression” were the keywords, suggesting the emphasis on the model selection. Among these models, Random Forest is the most widely used model for its good predictive performance and with no need for feature scaling or normalization of the inputs. In 2023, “cellulose”, “lignin”, and “hemicellulose” gained attention, indicating the feedstock of HTL expanding from microalgae to lignocellulosic biomass. In 2024, “heterogeneous catalysts” and “model component” were introduced for the first time. Using ML models to predict catalytic HTL processes and HTL of model components will be a future research trend.
Figure 6 shows the keyword timeline map from CiteSpace. Similar to time zone maps, purple, cyan, green, and red diamonds represent the years of 2021, 2022, 2023, and 2024, respectively. The keywords that appearing in subsequent years are overlaid with diamonds of different colors. The horizontal axis represents the years, and the vertical axis represents 10 major clusters. These clusters are: biocrude oil (#0), synergy (#1), biocrude-oil (#2), bio-oil (#3), lignocellulosic (#4), multidimensional analysis (#5), hydrothermal reaction (#6), techno-economic analysis (#7), algorithm optimization (#8), and biorefinery (#9). Keyword analysis is an effective method for understanding the evolution of research fields and future trends [36].
Figure 7 presents the co-occurrence keyword map derived from 2022 to 2024, featuring a total of 158 keywords. The keywords with higher frequencies indicate that they are currently gaining significant attention in ML-aided HTL and may become new trends or hot topics. For example, “machine learning” (33 occurrences) and “hydrothermal liquefaction” (31 occurrences) are prominent. Keywords with the third and fourth highest frequency, “biomass” (23 occurrences) and “microalgae” (21 occurrences), represent the primary feedstock during HTL. The gradient colors from deep blue to green and to yellow in Figure 7 represent the years when different keywords began to be studied. These keywords are key focuses in ML-aided HTL research.
In early 2022, clusters represented by deep blue, including “biocrude oil”, “algorithm optimization”, “prediction and optimization”, “biocrude yield”, and “extreme gradient boosting”, were prominent. Among these keywords, “prediction and optimization” represented the critical application of ML-aided HTL. ML could be directly applied to predict the yields and properties of HTL products. Using feature analysis, SHAP analysis and partial dependence analysis, the effects of the biomass compositions and operation conditions could be revealed. Further refinement of feedstock compositions and reaction conditions could be achieved through ML-based reverse optimization. Clusters represented by light blue included “model”, “microalgae”, “yield”, “random forest”, “co-liquefaction”, “biomass”, “components”, and “residue”. The keywords of “microalgae” and “model” were hotspots in mid-2022. In 2023, the clusters represented by green included “biofuel”, “biocrude”, “optimization”, “conversion”, “sensitivity analysis”, “protein”, and “kinetic model”. In 2024, “heterogeneous catalyst” and homogeneous “Na2CO3” identified by the yellow cluster were introduced for the first time. Yang et al. [37] analyzed the impact of Na2CO3 on HTL of biomass model compounds and used an Adaboost model to predict the biocrude yield in the presence of Na2CO3. The training and testing R2s reached 0.96 and 0.8, respectively. Feature importance analysis revealed that the lipid content and Na2CO3 concentration had the most significant influence on biocrude yield, surpassing the effects of temperature, reaction time, and other reaction conditions.

4. Conclusions

This paper discusses the application of machine learning (ML) in hydrothermal liquefaction (HTL) and its research trends.
  • Since 2020, the application of ML in HTL has grown rapidly, particularly in predicting biocrude yield and optimizing reaction conditions. The number of papers significantly increased from 3 in 2020 to 49 in the first half of 2024, indicating sustained growth and research interest in this area.
  • Bioresource Technology and Fuel are the journals with the most publications on ML-aided HTL, with Bioresource Technology particularly remarkable in terms of citations. Highly cited articles are predominantly published by Leng, Naqvi and Aghbashlo, focusing mainly on the prediction and optimization of biocrude yield and quality.
  • China, India, and the United States are leading in ML-assisted HTL research, ranking high in both publication numbers and citations. These three countries also maintain close collaborative relationships with other nations.
  • Keyword analysis indicates that “machine learning”, “hydrothermal liquefaction”, “model”, “biomass”, “microalgae”, and “co-liquefaction” are significant research focuses. The selection of ML models has become a research hotspot. Using different ML models to predict catalytic HTL processes will be a future research trend.
To better harness ML methods in the field of HTL, it is essential to improve the predictive performance and generalizability of ML models. More experimental data are needed to increase the size of the training dataset. Additionally, individual HTL or co-HTL data from non-catalytic, homogeneous, and heterogeneous catalytic experiments can be integrated to increase the model generalizability. Through generalized ML models, research on biomass HTL processes is expected to be accelerated.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/en17215254/s1, Table S1: Dataset used on machine learning-aided hydrothermal liquefaction studies.

Author Contributions

Conceptualization, L.Q.; methodology, L.Q. and X.Z.; software, X.Z.; validation, L.Q. and X.Z.; formal analysis, L.Q.; investigation, X.Z., X.M. and P.X.; resources, L.Q. and S.W.; data curation, L.Q. and X.Z.; writing—original draft preparation, L.Q. and X.Z.; writing—review and editing, L.Q., X.T., X.L. and S.W.; visualization, L.Q. and X.Z.; supervision, L.Q.; project administration, L.Q.; funding acquisition, L.Q. and X.L. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the National Natural Science Foundation of China, grant number 52206252, the Fundamental Research Funds for the Central Universities, grant number 2242024k30025, the Key Laboratory of Energy Thermal Conversion and Control of Ministry of Education, Southeast University, grant number etcc202404, and the Taizhou Science and Technology Support Programme, grant number TS202202. The APC was funded by MDPI.

Data Availability Statement

No additional data are available.

Conflicts of Interest

Author Xiang Li was employed by the company Taizhou DongBo New Materials Co., Ltd. The remaining author declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

References

  1. Pitchai, C.; Vedanarayanan, M.; Chen, C.-M.; Gopalakrishnan, S.M. Enhanced electrochemical efficiency of the open porous sandrose structured electrocatalyst for sustainable hydrogen and oxygen evolution reactions. Int. J. Hydrogen Energy 2024, 72, 755–763. [Google Scholar] [CrossRef]
  2. Pitchai, C.; Gopalakrishnan, S.M.; Chen, C.-M. Ultra-efficient nitrogen-doped carbon dots-supported nickel sulfide as a platinum-free electrocatalyst for overall water splitting in basic medium. Energy Fuels 2024, 38, 2235–2247. [Google Scholar] [CrossRef]
  3. Usman, M.; Cheng, S.; Boonyubol, S.; Cross, J.S. From biomass to biocrude: Innovations in hydrothermal liquefaction and upgrading. Energy Convers. Manag. 2024, 302, 118093. [Google Scholar] [CrossRef]
  4. Wang, Y.; Qian, L.; Yang, D.; Gong, Y.; Yuan, C.; Hu, Y.; Gu, H.; Sun, P.; Wang, S. Integration of hydrothermal liquefaction of Cyanophyta and supercritical water oxidation of its aqueous phase products: Biocrude production and nutrient removal. Sci. Total Environ. 2024, 914, 169835. [Google Scholar] [CrossRef] [PubMed]
  5. Zhang, W.; Chen, Q.; Chen, J.; Xu, D.; Zhan, H.; Peng, H.; Pan, J.; Vlaskin, M.; Leng, L.; Li, H. Machine learning for hydrothermal treatment of biomass: A review. Bioresour. Technol. 2023, 370, 128547. [Google Scholar] [CrossRef]
  6. Guirguis, P.M.; Seshasayee, M.S.; Motavaf, B.; Savage, P.E. Review and assessment of models for predicting biocrude yields from hydrothermal liquefaction of biomass. RSC Sustain. 2024, 2, 736–756. [Google Scholar] [CrossRef]
  7. Elgarahy, A.M.; Hammad, A.; El-Sherif, D.M.; Abouzid, M.; Gaballah, M.S.; Elwakeel, K.Z. Thermochemical conversion strategies of biomass to biofuels, techno-economic and bibliometric analysis: A conceptual review. J. Environ. Chem. Eng. 2021, 9, 106503. [Google Scholar] [CrossRef]
  8. Ocampo, E.; Beltrán, V.V.; Gómez, E.A.; Ríos, L.A.; Ocampo, D. Hydrothermal liquefaction process: Review and trends. Curr. Res. Green Sustain. Chem. 2023, 7, 100382. [Google Scholar] [CrossRef]
  9. Bassoli, S.C.; da Fonseca, Y.A.; Wandurraga, H.J.L.; Baeta, B.E.L.; de Souza Amaral, M. Research progress, trends, and future prospects on hydrothermal liquefaction of algae for biocrude production: A bibliometric analysis. Biomass Convers. Bior. 2023, 1–16. [Google Scholar] [CrossRef]
  10. Yang, J.; Hong, C.; Xing, Y.; Zheng, Z.; Li, Z.; Zhao, X.; Qi, C. Research progress and hot spots of hydrothermal liquefaction for bio-oil production based on bibliometric analysis. Environ. Sci. Pollut. Res. 2021, 28, 7621–7635. [Google Scholar] [CrossRef]
  11. Chen, C. CiteSpace: A Practical Guide for Mapping Scientific Literature; Nova Science Pub Inc.: Hauppauge, NY, USA, 2016. [Google Scholar]
  12. Van Eck, N.J.; Waltman, L. VOSviewer Manual Version 1.6.8; Leiden University: Leiden, The Netherlands, 2018. [Google Scholar]
  13. Cheng, F.W.; Porter, M.D.; Colosi, L.M. Is hydrothermal treatment coupled with carbon capture and storage an energy-producing negative emissions technology? Energy Convers. Manag. 2020, 203, 112252. [Google Scholar] [CrossRef]
  14. Zhang, W.; Ai, Z.; Chen, Q.; Chen, J.; Xu, D.; Cao, J.; Kapusta, K.; Peng, H.; Leng, L.; Li, H. Automated machine learning-aided prediction and interpretation of gaseous by-products from the hydrothermal liquefaction of biomass. Sci. Total Environ. 2024, 945, 173939. [Google Scholar] [CrossRef] [PubMed]
  15. Li, J.; Zhang, W.; Liu, T.; Yang, L.; Li, H.; Peng, H.; Jiang, S.; Wang, X.; Leng, L. Machine learning aided bio-oil production with high energy recovery and low nitrogen content from hydrothermal liquefaction of biomass with experiment verification. Chem. Eng. J. 2021, 425, 130649. [Google Scholar] [CrossRef]
  16. Zhang, W.; Li, J.; Liu, T.; Leng, S.; Yang, L.; Peng, H.; Jiang, S.; Zhou, W.; Leng, L.; Li, H. Machine learning prediction and optimization of bio-oil production from hydrothermal liquefaction of algae. Bioresour. Technol. 2021, 342, 126011. [Google Scholar] [CrossRef]
  17. Khan, M.; Raza Naqvi, S.; Ullah, Z.; Ali Ammar Taqvi, S.; Nouman Aslam Khan, M.; Farooq, W.; Taqi Mehran, M.; Juchelková, D. Applications of machine learning in thermochemical conversion of biomass—A review. Fuel 2023, 332, 126055. [Google Scholar] [CrossRef]
  18. Shafizadeh, A.; Shahbeig, H.; Nadian, M.H.; Mobli, H.; Dowlati, M.; Gupta, V.K.; Peng, W.; Lam, S.S.; Tabatabaei, M.; Aghbashlo, M. Machine learning predicts and optimizes hydrothermal liquefaction of biomass. Chem. Eng. J. 2022, 445, 136579. [Google Scholar] [CrossRef]
  19. Katongtung, T.; Onsree, T.; Tippayawong, N. Machine learning prediction of biocrude yields and higher heating values from hydrothermal liquefaction of wet biomass and wastes. Bioresour. Technol. 2022, 344, 126278. [Google Scholar] [CrossRef]
  20. Leng, L.; Zhang, W.; Chen, Q.; Zhou, J.; Peng, H.; Zhan, H.; Li, H. Machine learning prediction of nitrogen heterocycles in bio-oil produced from hydrothermal liquefaction of biomass. Bioresour. Technol. 2022, 362, 127791. [Google Scholar] [CrossRef]
  21. Leng, L.; Zhang, W.; Liu, T.; Zhan, H.; Li, J.; Yang, L.; Li, J.; Peng, H.; Li, H. Machine learning predicting wastewater properties of the aqueous phase derived from hydrothermal treatment of biomass. Bioresour. Technol. 2022, 358, 127348. [Google Scholar] [CrossRef]
  22. Liu, T.; Zhang, W.; Xu, D.; Leng, L.; Li, H.; Wang, S.; He, Y. Predicting co-liquefaction bio-oil of sewage sludge and algal biomass via machine learning with experimental optimization: Focus on yield, nitrogen content, and energy recovery rate. Sci. Total Environ. 2024, 920, 170779. [Google Scholar] [CrossRef]
  23. Ai, Z.; Zhang, W.; Yang, L.; Chen, H.; Xu, Z.; Leng, L.; Li, H. Investigation and prediction of co-pyrolysis between oily sludge and high-density polyethylene via in-situ DRIFTS, TGA, and artificial neural network. J. Anal. Appl. Pyrol. 2022, 166, 105610. [Google Scholar] [CrossRef]
  24. Leng, L.; Yang, L.; Lei, X.; Zhang, W.; Ai, Z.; Yang, Z.; Zhan, H.; Yang, J.; Yuan, X.; Peng, H.; et al. Machine learning predicting and engineering the yield, N content, and specific surface area of biochar derived from pyrolysis of biomass. Biochar 2022, 4, 23. [Google Scholar] [CrossRef]
  25. Leng, L.; Li, T.; Zhan, H.; Rizwan, M.; Zhang, W.; Peng, H.; Yang, Z.; Li, H. Machine learning-aided prediction of nitrogen heterocycles in bio-oil from the pyrolysis of biomass. Energy 2023, 278, 127967. [Google Scholar] [CrossRef]
  26. Leng, L.; Zhou, J.; Li, T.; Vlaskin, M.; Zhan, H.; Peng, H.; Huang, H.; Li, H. Nitrogen heterocycles in bio-oil produced from hydrothermal liquefaction of biomass: A review. Fuel 2023, 335, 126995. [Google Scholar] [CrossRef]
  27. Li, H.; Ai, Z.; Yang, L.; Zhang, W.; Yang, Z.; Peng, H.; Leng, L. Machine learning assisted predicting and engineering specific surface area and total pore volume of biochar. Bioresour. Technol. 2023, 369, 128417. [Google Scholar] [CrossRef]
  28. Li, H.; Chen, J.; Zhang, W.; Zhan, H.; He, C.; Yang, Z.; Peng, H.; Leng, L. Machine-learning-aided thermochemical treatment of biomass: A review. Biofuel Res. J. 2023, 10, 1786–1809. [Google Scholar] [CrossRef]
  29. Leng, L.; Kang, B.; Xu, D.; Kapusta, K.; Xiong, T.; Xu, Z.; Fan, L.; Liu, T.; Peng, H.; Li, H. Machine-learning-aided prediction and optimization of struvite recovery from synthetic wastewater. J. Water Process Eng. 2024, 58, 104896. [Google Scholar] [CrossRef]
  30. Leng, L.; Lei, X.; Abdullah Al-Dhabi, N.; Wu, Z.; Yang, Z.; Li, T.; Zhang, W.; Liu, W.; Zhan, H.; Peng, H.; et al. Machine-learning-aided prediction and engineering of nitrogen-containing functional groups of biochar derived from biomass pyrolysis. Chem. Eng. J. 2024, 485, 149862. [Google Scholar] [CrossRef]
  31. Leng, L.; Zhou, J.; Zhang, W.; Chen, J.; Wu, Z.; Xu, D.; Zhan, H.; Yuan, X.; Xu, Z.; Peng, H.; et al. Machine-learning-aided hydrochar production through hydrothermal carbonization of biomass by engineering operating parameters and/or biomass mixture recipes. Energy 2024, 288, 129854. [Google Scholar] [CrossRef]
  32. Yang, Z.; Chen, B.; Zu, H.; Zhang, W.; Ai, Z.; Leng, L.; Chen, H.; Feng, Y.; Li, H. Rapid and accurate identification of effective metal organic frameworks for tetrafluoromethane/nitrogen separation by machine learning. J. Taiwan Inst. Chem. Eng. 2024, 162, 105556. [Google Scholar] [CrossRef]
  33. Zhang, W.; Zhou, J.; Liu, Q.; Xu, Z.; Peng, H.; Leng, L.; Li, H. A novel intelligent system based on machine learning for hydrochar multi-target prediction from the hydrothermal carbonization of biomass. Biochar 2024, 6, 19. [Google Scholar] [CrossRef]
  34. Katongtung, T.; Onsree, T.; Tippayawong, K.Y.; Tippayawong, N. Prediction of biocrude oil yields from hydrothermal liquefaction using a gradient tree boosting machine approach with principal component analysis. Energy Rep. 2023, 9, 215–222. [Google Scholar] [CrossRef]
  35. Sahoo, A.; Saini, K.; Jindal, M.; Bhaskar, T.; Pant, K.K. Co-Hydrothermal Liquefaction of algal and lignocellulosic biomass: Status and perspectives. Bioresour. Technol. 2021, 342, 125948. [Google Scholar] [CrossRef] [PubMed]
  36. Krishnamoorthy, N.; Pathy, A.; Kapoor, A.; Paramasivan, B. Exploring the evolution, trends and scope of microalgal biochar through scientometrics. Algal Res. 2023, 69, 102944. [Google Scholar] [CrossRef]
  37. Ye, W.; Lin, X.; Zhong, X.; Lan, Q.; Yang, G.; Huang, J.; Wu, Y.; Zhao, Z.; He, Q.; Wu, Y.; et al. Unraveling the effects of sodium carbonate on hydrothermal liquefaction through individual biomass model component and machine learning-enabled prediction. Fuel 2024, 358, 130330. [Google Scholar] [CrossRef]
Energies 17 05254 g001
Figure 1. Publication volume and total number of publications on ML-aided HTL from 2020 to June 2024.
Figure 1. Publication volume and total number of publications on ML-aided HTL from 2020 to June 2024.
Energies 17 05254 g001
Energies 17 05254 g002
Figure 2. Publications and total citations of ML-aided HTL from 2020 to June 2024.
Figure 2. Publications and total citations of ML-aided HTL from 2020 to June 2024.
Energies 17 05254 g002
Energies 17 05254 g003
Figure 3. International collaboration on ML-aided HTL from 2020 to June 2024.
Figure 3. International collaboration on ML-aided HTL from 2020 to June 2024.
Energies 17 05254 g003
Energies 17 05254 g004
Figure 4. Visualization map of co-authorship on ML-aided HTL from 2020 to 2024.
Figure 4. Visualization map of co-authorship on ML-aided HTL from 2020 to 2024.
Energies 17 05254 g004
Energies 17 05254 g005
Figure 5. Keyword time zone map from CiteSpace on ML-aided HTL from 2021 to 2024.
Figure 5. Keyword time zone map from CiteSpace on ML-aided HTL from 2021 to 2024.
Energies 17 05254 g005
Energies 17 05254 g006
Figure 6. Keyword timeline view from CiteSpace on ML-aided HTL from 2021 to 2024.
Figure 6. Keyword timeline view from CiteSpace on ML-aided HTL from 2021 to 2024.
Energies 17 05254 g006
Energies 17 05254 g007
Figure 7. The co-occurrence keyword map of ML-aided HTL from 2022 to 2024.
Figure 7. The co-occurrence keyword map of ML-aided HTL from 2022 to 2024.
Energies 17 05254 g007
Table 1. The top six journals with the most publications on ML-aided HTL from 2020 to June 2024, along with their average citations.
Table 1. The top six journals with the most publications on ML-aided HTL from 2020 to June 2024, along with their average citations.
RankJournalNumber of Papers PublishedAverage CitationsPapers Published After 2023Average Citations for Papers After 2023
1Bioresource Technology1832.44719.71
2Fuel1522.21117.18
3Chemical Engineering Journal747.14317.33
4Energy Conversion and Management726.1444.5
5Science of The Total Environment516.2516.2
6Energy & Fuels32115
Table 2. The top six most-cited papers on ML-aided HTL from 2020 to June 2024.
Table 2. The top six most-cited papers on ML-aided HTL from 2020 to June 2024.
RankTitleCorresponding AuthorsYearPublished JournalCitation CountReference
1Machine learning aided bio-oil production with high energy recovery and low nitrogen content from hydrothermal liquefaction of biomass with experiment verificationLeng2021Chemical Engineering Journal100[15]
2Machine learning prediction and optimization of bio-oil production from hydrothermal liquefaction of algaeLeng2021Bioresource Technology97[16]
3Applications of machine learning in thermochemical conversion of biomass-A reviewNaqvi2023Fuel90[17]
4Machine learning predicts and optimizes hydrothermal liquefaction of biomassLam2022Chemical Engineering Journal89[18]
5Machine learning prediction of biocrude yields and higher heating values from hydrothermal liquefaction of wet biomass and wastesTippayawong2022Bioresource Technology77[19]
Table 3. Authors with a large publication number on ML-aided HTL from 2020 to June 2024.
Table 3. Authors with a large publication number on ML-aided HTL from 2020 to June 2024.
AuthorPublicationsCitations
Leng, Lijian9256
Li, Hailong9256
Peng, Haoyi8242
Zhang, Weijin8242
Li, Jie5201
Liu, Tonggui4180
Nakorn, Tippayawong498
Yang, Lihong3180
Wang, Xiaonan3100
Vijai-Kumar, Gupta3112
Tossapon, Katongtung369
Thossaporn, Onsree251
Jiang, Shaojian2151
Chen, Qingyue261
Mikhail, Vlaskin252
Muhammad-Nouman-Aslam, Khan280
Salman-Raza, Nagvi280
Cheng, Fangwei299
Lisa, Colosi299
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Qian, L.; Zhang, X.; Ma, X.; Xue, P.; Tang, X.; Li, X.; Wang, S. A Review on Machine Learning-Aided Hydrothermal Liquefaction Based on Bibliometric Analysis. Energies 2024, 17, 5254. https://doi.org/10.3390/en17215254

AMA Style

Qian L, Zhang X, Ma X, Xue P, Tang X, Li X, Wang S. A Review on Machine Learning-Aided Hydrothermal Liquefaction Based on Bibliometric Analysis. Energies. 2024; 17(21):5254. https://doi.org/10.3390/en17215254

Chicago/Turabian Style

Qian, Lili, Xu Zhang, Xianguang Ma, Peng Xue, Xingying Tang, Xiang Li, and Shuang Wang. 2024. "A Review on Machine Learning-Aided Hydrothermal Liquefaction Based on Bibliometric Analysis" Energies 17, no. 21: 5254. https://doi.org/10.3390/en17215254

APA Style

Qian, L., Zhang, X., Ma, X., Xue, P., Tang, X., Li, X., & Wang, S. (2024). A Review on Machine Learning-Aided Hydrothermal Liquefaction Based on Bibliometric Analysis. Energies, 17(21), 5254. https://doi.org/10.3390/en17215254

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop