The records retrieved from the Web of Science Core Collection were exported in plain text format using the “Full Record and Cited References” option. Due to the platform’s export limit, the records were divided into three files. The records from Scopus were exported in CSV format, containing the available bibliographic metadata. The citation counts used in the bibliometric analyses were obtained directly from the metadata provided by the respective databases. After consolidating the Web of Science files, the records were merged with those from Scopus, forming a single dataset for the subsequent stages of deduplication and analysis.

The keyword co-occurrence networks were generated using VOSviewer version 1.6.20. The bibliometric analyses were performed using R version 4.5.0, RStudio version 2026.01.1+403, and the Bibliometrix package (version 5.0.1) through its Biblioshiny interface. The exported files were converted using the convert2df() function, with the parameters dbsource = “wos” and format = “plaintext” for Web of Science records, and dbsource = ‘scopus’ and format = “csv” for Scopus records. The files from Web of Science were initially consolidated into a single dataset and subsequently integrated with the Scopus records. The VOSviewer analysis was performed using the "Create a map based on bibliographic data" option, with the "All Keywords" field selected as the unit of analysis and the Full Counting method. A minimum occurrence threshold of 54 was adopted, resulting in 6,292 identified keywords, of which 20 met the inclusion criterion and were selected for network construction. No thesaurus file was used. Variations in spelling, abbreviations, singular/plural forms, and equivalent terms were standardized during the metadata cleaning stage prior to network construction. Unless otherwise specified, the default VOSviewer settings were adopted for normalization, clustering, and network layout (Aria and Cuccurullo 2017; van Eck and Waltman 2010; Moral-Muñoz et al. 2020).

Duplicate records were identified and removed using the `duplicatedMatching()` function, with the document title as the comparison field (Field = “TI”), approximate matching (exact = FALSE), and a minimum similarity threshold of 0.95 (tol = 0.95). This procedure made it possible to identify records with identical or highly similar titles, including minor differences in spelling, punctuation, and indexing between the two databases. No separate deduplication step based exclusively on DOI or the first author’s last name was employed. When records had titles similar enough to meet the established similarity criterion, only one record was retained in the consolidated dataset.

After integration and deduplication, the metadata was verified to minimize any remaining inconsistencies. Author and institutional names were standardized when necessary, reducing ambiguities arising from spelling variations and differences in indexing across databases. When discrepancies were identified between metadata from Scopus and Web of Science, the record considered most complete was retained for subsequent analyses.

After integrating the databases, the metadata was standardized to reduce obvious inconsistencies between records from Scopus and Web of Science. This process included correcting spelling variations, obvious duplicates, and minor differences in the spelling of institution and author names, while preserving, whenever possible, the original nomenclature provided by the indexing databases. Neither an external institutional thesaurus nor a specific dictionary was used for complete entity disambiguation. Thus, organizations indexed as university systems, national scientific information platforms, or other aggregated entities remained as recorded in the databases when it was not possible to unequivocally establish their correspondence with individual institutions. Similarly, the identification of authors was based on the standardized metadata provided by the databases; no complete manual disambiguation was performed for authors with identical last names and initials.

Since this study consists of a bibliometric analysis based exclusively on bibliographic metadata, no manual screening of titles, abstracts, or full-text articles was performed. Figure 1 depicts only the process of data retrieval, application of filters, database integration, removal of duplicates, and preparation of the dataset; it does not correspond to the eligibility flowchart of a systematic review.

After completing all stages of retrieval, integration, deduplication, and standardization of the metadata, the final bibliographic corpus consisted of 1,597 unique publications, forming the database used in all bibliometric and scientific mapping analyses conducted in this study.

The evolution of scientific output over the analyzed period is shown in Figure 2. The bibliometric survey identified a total of 1,597 documents published between 2014 and 2025, distributed across 538 sources, including peer-reviewed journals and conference proceedings.

There is a general upward trend in annual scientific output, highlighting the growing interest in the valorization of lignocellulosic waste and the development of sustainable materials for environmental remediation (Fang et al. 2020; Meirelles et al. 2025). Despite some fluctuations over the period, such as a decline between 2020 and 2021, the overall trend indicates a substantial expansion of research activity.

Starting in 2022, a sharp increase in the number of publications is evident, reaching its peak in 2025. This trend reflects intensified global research efforts focused on biochar-based materials and TiO₂ applications, particularly in adsorption and photocatalysis processes for contaminant removal (Ahmad et al. 2014).

The average of 28.39 citations per document was calculated based on the citation counts available in the Scopus and Web of Science metadata, which were consolidated into the integrated dataset. The average age of the publications was 4.43 years, indicating that the corpus consists predominantly of recent studies (Table 2).

In terms of document types, the bibliographic corpus consisted predominantly of original scientific articles (1,487 documents), supplemented by 110 conference papers (Table 2). This distribution highlights the predominance of results published in peer-reviewed journals, while conference papers represent a smaller, yet relevant, portion of the scientific output analyzed (Fang et al. 2020; Meirelles et al. 2025).

In terms of authorship structure, a total of 6,250 authors contributed to the dataset, with an average of 5.62 co-authors per paper, reflecting a high level of collaboration. Furthermore, approximately 28.55% of the publications involve international co-authorship, highlighting the global and interdisciplinary nature of the research field (Table 2).

Overall, the results demonstrate that research on biochar and TiO₂-based composites has expanded significantly, driven by the growing demand for sustainable technologies aimed at environmental remediation, waste valorization, and the development of advanced materials.

The observed geographic distribution (Figure 3) may be related to factors such as existing scientific capacity, national research priorities, available infrastructure, and interest in technologies aimed at utilizing lignocellulosic waste (Silalertruksa et al. 2022; Pereira et al. 2025). However, since this study was exclusively bibliometric in nature, indicators of agricultural production, biomass availability, investments in research and development, or scientific funding were not analyzed. Thus, any association between these factors and scientific productivity should be interpreted only as a plausible hypothesis, not as an established causal relationship.

Figure 4 shows the geographic distribution of scientific output, while Table 3 summarizes the ten countries with the highest number of documents and their respective indicators of international collaboration. The results were obtained based on the country of affiliation of the corresponding author, according to the indicators generated by the Bibliometrix software package (Aria and Cuccurullo 2017). The Single Country Publications (SCP) indicator represents documents whose authors belong exclusively to a single country, while Multiple Country Publications (MCP) corresponds to documents produced in collaboration between authors from different countries (Sweileh et al. 2016). Thus, the values presented represent the classification based on the corresponding author’s country and not a complete count of all national affiliations present in each document.

China led in scientific output, with 359 publications (22.5%), followed by India, with 259 publications (16.2%), and Brazil, with 234 publications (14.7%). Together, these three countries accounted for more than half of the publications in the analyzed corpus, highlighting their prominent role in research related to biochar production from sugarcane waste and its environmental applications (Lyu et al. 2018; Domingues et al. 2020).

Indicators of international collaboration revealed significant differences among countries. China produced 257 domestically authored publications (SCP) and 102 publications involving international collaboration (MCP), demonstrating high scientific productivity combined with significant participation in international research networks. India (216 SCP and 43 MCP) and Brazil (185 SCP and 49 MCP) also recorded high numbers of publications, though with a predominance of studies conducted at the national level. In contrast, countries such as Pakistan (60.0% MCP) and Australia (50.0% MCP) showed higher proportions of international collaboration relative to their total scientific output, indicating greater integration into global research networks.

These results show that a higher number of publications does not necessarily imply greater internationalization of scientific research. While some countries focus their research output on domestic collaborations, others have a higher relative share of research conducted through international cooperation, highlighting different strategies for scientific collaboration.

These findings indicate that a high volume of publications does not necessarily correspond to greater international collaboration. While China, India, and Brazil dominate total output, countries such as Pakistan and Australia show greater integration into global research networks.

The predominance of SCP among leading countries suggests that a large portion of research is still conducted within national frameworks. However, the substantial presence of MCP highlights the growing internationalization of the field and the importance of collaboration among countries for advancing research on biochar and TiO₂-based materials (Shi et al. 2022; El-Sabban et al. 2025).

Overall, the results show that scientific output in this field is geographically concentrated and increasingly interconnected, reflecting the combined influence of resource availability, scientific capacity, and global collaboration in addressing environmental challenges and developing sustainable technologies.

The institutional analysis (Figure 5) reveals that scientific output is concentrated in a limited number of universities and research systems, highlighting the central role of key organizations in advancing this field of research. Among the institutional entities recorded in the databases consulted, the Egyptian Knowledge Bank had the highest number of associated documents. This result should be interpreted with caution, as it reflects how affiliations are indexed in bibliographic databases and may include national information platforms or aggregated institutional systems, followed by Guangxi University and the Indian Institute of Technology (IIT System). Other relevant collaborators include the University of Agriculture Faisalabad, the State University System of Florida, the University of Florida, and the Chinese Academy of Sciences (Chen et al. 2023).

These results indicate a strong geographic alignment between leading institutions and countries with high scientific output, particularly in Asia and the Middle East. Institutions in India, China, Pakistan, and Egypt play a prominent role, reflecting growing investment in research related to environmental technologies, biomass utilization, and advanced materials (Chen et al. 2023).

The presence of institutions such as the University of São Paulo (USP) and Khon Kaen University further highlights the global distribution of research efforts, although their contribution is comparatively smaller than that of leading institutions (Schellekens et al. 2018; Kumalayanti et al. 2025). This suggests that, although scientific output is distributed globally, it remains concentrated in specific, high-output research systems.

From the perspective of collaboration, the structure of institutional networks reflects the patterns observed at the country level, with research predominantly conducted within well-established national frameworks (Shi et al. 2022; El-Sabban et al. 2025). This is particularly evident in countries with strong academic infrastructure, where institutions maintain high internal productivity while gradually expanding international collaboration.

In bibliometric studies, identifying institutional collaboration networks is essential for understanding how scientific knowledge is structured and disseminated among research groups, revealing patterns of cooperation and the organization of academic output. Overall, the findings indicate that research on biochar and TiO₂-based composites is driven by a core of highly productive institutions, located primarily in emerging and developing economies, where environmental challenges and resource availability stimulate scientific innovation (El-Sabban et al. 2025). Institutional networks have helped accelerate the development of sustainable technologies for environmental remediation and biomass utilization (Aria and Cuccurullo 2017).

Figure 6 presents a treemap of the most frequent keywords identified in the bibliographic dataset. Only terms that met the minimum occurrence threshold established for the analysis were included in the visualization. The percentages shown correspond to the relative frequency of each keyword compared to the total number of occurrences of the keywords displayed in the treemap, allowing for a comparison of their representativeness within the selected dataset (Aria and Cuccurullo 2017; Cobo et al. 2011).

An analysis of keyword occurrence indicates that terms such as biochar, adsorption, pyrolysis, and sugarcane bagasse are among the most frequently reported in the literature, reflecting the strong focus on the production and application of biochar derived from lignocellulosic waste for environmental purposes (Arumugam and Dhandapani 2025; Shikuku and Jemutai-Kimosop 2020; Shi et al. 2022; Souza et al. 2025).

The prominence of these terms confirms that the thermochemical conversion of biomass via pyrolysis and its application in adsorption processes constitute the central focus of research in this field. This trend is consistent with recent studies highlighting the efficiency of biochar-based materials in removing pollutants, particularly in water and wastewater treatment systems (da Silva et al. 2025; Cai et al. 2025).

Furthermore, the presence of terms such as “activated carbon” (172 occurrences, 3%); “sorption” (170 occurrences, 3%); and “carbon” (167 occurrences, 3%) reinforces the relevance of carbonaceous materials in environmental remediation technologies. Terms such as “water” (136 occurrences, 2%) further highlight the environmental scope of the research, indicating applications in aquatic and terrestrial systems (Alharbi et al. 2025).

To explore the intellectual structure of the field, a three-field plot was constructed using the Bibliometrix package, as shown in Figure 7. The visualization relates authors (AU), keywords (DE), and scientific sources (SO), allowing us to identify the connections between these elements in the analyzed corpus. To facilitate interpretation, the plot was constructed using the eight authors with the highest number of publications, the eight most frequent keywords, and the eight most represented journals. The width of the connections is proportional to the number of documents shared between the connected elements, highlighting the strength of the relationships among authors, research topics, and publication outlets (Boyack and Klavans 2010; van Eck and Waltman 2010; Zupic and Čater 2015; Waltman et al. 2010).

The central column highlights the most frequent keywords, with biochar, adsorption, pyrolysis, and sugarcane bagasse occupying prominent positions, confirming their role as the main research topics. On the left side, authors with the highest number of publications in the corpus stand out, such as Gao B, Wang Jj, and Agarwal S, who are associated with these topics. On the right side, the journals most represented in the dataset are shown, such as the “Journal of Environmental Chemical Engineering”, “Bioresource Technology”, and “Biomass Conversion and Biorefinery”, highlighting the main publication outlets in the field (Yao et al. 2018; Luo et al. 2023).

To complement the analysis of keyword frequency, co-occurrence networks were generated using the VOSviewer software, as shown in Figure 8 (a–b). The cluster map in Figure 8(a) was used to identify conceptual relationships among the most frequently occurring keywords, while the overlay map in Figure 8(b) represents the average year of publication associated with each term, allowing for the identification of temporal trends in the evolution of the literature (Fang et al. 2020; Khare et al. 2021). It should be noted that warmer shades indicate keywords related to more recent publications and should not be interpreted, in isolation, as evidence of greater scientific relevance.

Figure 8a (cluster map) reveals the organization of the field into interconnected thematic clusters. The term “biochar” occupies a central and highly connected position in the network, indicating its role as one of the main structuring elements of the research domain and its association with multiple topics under investigation (Loc 2023; Bhakta et al. 2023).

One prominent cluster is associated with bagasse, which shows strong connections to adsorption and titanium dioxide. This configuration points to a line of research related to the development of hybrid materials based on biochar and TiO₂ for adsorption and photocatalysis processes (Leichtweis et al. 2021; Tohamy et al. 2023). The observed connections highlight the thematic association between these terms in the analyzed literature, indicating an expanding research topic; however, this alone does not allow us to infer its degree of scientific consolidation.

Another relevant cluster includes terms such as pyrolysis, biomass, and soil amendment, which are associated with the production, characterization, and environmental application of biochar derived from agricultural waste. This cluster reflects the fundamental processes underlying biochar generation and its use in both environmental remediation and agricultural improvement (Borges et al. 2020; Barragán-Mantilla et al. 2023).

Additionally, smaller clusters related to contaminant removal (e.g., dyes and heavy metals) and material characterization can be observed, indicating the diversification of research topics and the emergence of more specialized applications (Cai et al. 2025).

Figure 8b (temporal overlay map) illustrates the evolution of research topics over time. Terms represented in warmer shades are associated with more recent average publication years, indicating themes that have begun to appear more frequently in recent studies. In contrast, early research, represented by cooler colors, is more strongly associated with fundamental topics such as biomass, carbon, and pyrolysis processes, reflecting the early stages of knowledge development in the field (Ahmad et al. 2014).

Overall, a combined analysis of keyword frequency, structural relationships, and temporal trends demonstrates that the scientific literature is organized around three main research areas: (i) biochar production via pyrolysis, (ii) application in adsorption processes for contaminant removal, and (iii) development of functional and hybrid materials for environmental remediation. Furthermore, the growing emphasis on biochar/TiO₂ hybrid systems highlights increasing technological sophistication (Ahmad et al. 2014; Afshar and Mofatteh 2024).

Although the literature reviewed demonstrates significant growth and increasing thematic diversification, the bibliometric results reveal the persistence of major research gaps related to the valorization of sugarcane waste for biochar production and its integration with functional materials. Recent studies have highlighted the potential of these materials for environmental applications, particularly in the treatment of water and contaminated effluents (Inyang et al. 2015; Park et al. 2019).

One of the main gaps identified relates to the still-limited integration between research focused on biochar production via pyrolysis and that dedicated to the development of advanced composites with photocatalytic properties, particularly those involving TiO₂. Although terms related to adsorption and effluent treatment are consistently present in the co-occurrence network, the term “titanium dioxide” was also identified among the analyzed topics, indicating its application in studies focused on the development of hybrid materials for environmental applications (Cañon-Tafur et al. 2025; Chauhan et al. 2025; Chen and Mao 2007). However, its frequency of occurrence is lower than that observed for the main themes in the field, suggesting that this line of research is still in the process of expansion and development.

Another important aspect concerns the understanding of the synergistic mechanisms between adsorption and photocatalysis. In biochar/TiO₂ composites, biochar can act simultaneously as a structural support, an adsorbent, and an electron transfer mediator, reducing the recombination of electron-hole pairs generated during the photocatalytic process. This combined effect has been identified as one of the factors contributing to increased efficiency in the degradation of organic pollutants in aqueous environments (Du et al., 2024; Azizah et al., 2025).

However, although adsorption and photocatalysis are often used in a complementary manner in biochar/TiO₂ composites, these processes have distinct mechanisms and should not be considered equivalent. While adsorption facilitates the transfer of the contaminant from the aqueous phase to the material’s surface, photocatalysis promotes its chemical transformation through the action of reactive species. Furthermore, the degradation of the original contaminant does not necessarily imply its complete mineralization or a reduction in toxicity, since intermediate products may form during the process. In this context, future studies should consider, in addition to removal efficiency, aspects such as mineralization, byproduct formation, ecotoxicity, biochar regeneration, catalyst recovery, stability over successive cycles of use, potential leaching of metals or nanoparticles, and the assessment of the environmental safety of these materials in real-world applications (Chandra et al. 2021; Chong et al. 2010).

Furthermore, the growing interest in using lignocellulosic waste as a feedstock for biochar production underscores the potential of these materials as an alternative for the valorization of agroindustrial waste within the context of the circular economy and environmental sustainability. The valorization of agricultural and agroindustrial waste enables the production of low-cost adsorbents and contributes to circular economy strategies and sustainable waste management (Verma et al. 2024). In this context, biochar has been extensively investigated as a promising material for the removal of emerging pollutants in aquatic environments, due to its high surface area, the presence of functional groups, and significant adsorption capacity (Dong et al. 2023; Arfasa and Tilahun 2025).

Although biochar is often associated with sustainability and circular economy strategies, its environmental performance depends on various factors throughout its life cycle, and not exclusively on the use of agricultural waste as a raw material. Factors such as carbon yield, material stability, energy demand for pyrolysis, emissions associated with composite synthesis, TiO₂ production, material regeneration, and end-of-life disposal can significantly influence its environmental footprint. Therefore, future studies should integrate life cycle assessments (LCA), techno-economic analyses, and carbon balances to quantify the potential environmental benefits and possible trade-offs between remediation efficiency and process-related impacts (Alighanbari et al. 2025; Arfasa and Tilahun 2025; Fawzy et al. 2022; Li et al. 2024). From an institutional and geographic perspective, although the field is experiencing strong growth in countries such as China, Brazil, and India, the results also point to the need to strengthen international and inter-institutional collaborations, particularly around integrated approaches involving materials science, environmental engineering, and the circular economy. The predominance of publications classified as Single-Country Publications (SCP) suggests that there is still room to expand the internationalization of research and strengthen cooperation networks focused on methodological standardization, comparisons between different raw materials, and the development of more robust protocols for performance evaluation. This trend has also been observed in recent bibliometric studies on heavy metal adsorption, environmental applications of biochar, modification of biochars for aquatic remediation, carbon sequestration, and biochar life-cycle assessment (Admas et al. 2026).

Looking ahead, the results suggest potential for expanding research related to the development of biochar/TiO₂ composites, the surface modification of biochars, the use of semiconductor dopants, and the application of these materials in the removal of emerging contaminants from complex environmental matrices. Promising directions also include the integrated valorization of agro-industrial waste within the context of the circular bioeconomy, the development of multifunctional materials with greater selectivity and stability, and the incorporation of environmental, economic, and technological assessment tools capable of supporting the transfer of these technologies to full-scale applications (Al-Gailani et al. 2025; Badawi et al. 2026).

These findings are consistent with recent review studies that address biochar- and metal oxide-based nanocomposites, surface-modified TiO₂ photocatalysts, the mineralization of emerging contaminants via advanced oxidative processes (AOPs), strategies for standardizing and scaling up these processes, and aspects related to the energy consumption of photocatalytic materials used in environmental remediation (Aziz et al. 2025; Dhila et al. 2025).

This study has some limitations related to the quality and standardization of the metadata provided by the bibliographic databases. Although a basic harmonization of institutional affiliations and authors’ names was performed to reduce obvious inconsistencies, some aggregated entities, such as university systems, national scientific information platforms, and other organizations indexed by the databases, were retained as they appeared in the original records. Furthermore, authors identified only by last name and initials may represent different researchers, a limitation inherent in bibliometric analyses based on metadata. Thus, institutional and author productivity indicators should be interpreted with these limitations in mind (Mongeon and Paul-Hus, 2014; Nowakowska, 2025).

Another limitation relates to citation lag, especially for the most recent publications included in the analysis period (2024–2025), which have not yet had sufficient time to accumulate citations. Furthermore, bibliographic databases are continuously updated, so that new indexings, metadata corrections, and changes to records may result in minor differences if the search strategy is replicated at a later time (Belter 2015).

Although a basic harmonization of author names, institutions, and keywords was performed to reduce obvious inconsistencies, ambiguities arising from different author signature formats, institutional variations, and the lack of complete standardization of descriptors may influence some bibliometric indicators, particularly those related to institutional productivity, scientific collaboration, and co-occurrence networks. Since a formal thesaurus was not used, minor terminological variations may remain in the dataset (Tang and Walsh 2010).

The results also depend on the search strategy used. Although the search was designed to retrieve literature related to biochar derived from sugarcane waste, titanium dioxide, and environmental applications, minor changes in the combination of terms, Boolean operators, or inclusion criteria could yield a different set of documents a characteristic inherent to bibliometric studies (Dagli et al. 2025).

In addition, all analyses were conducted using the full-counting method, which is the default setting in Bibliometrix and VOSviewer. Thus, each document contributes in full to the authors, institutions, and countries involved, which may favor entities with a greater number of co-authors compared to approaches based on fractional counting (Perianes-Rodriguez et al. 2016).

Finally, it is important to note that bibliometric indicators reflect patterns of scientific output, collaboration, and academic impact, but do not, on their own, allow for an assessment of the technological performance of the materials studied or provide evidence of their environmental efficiency (Dagli et al. 2025). Aspects such as adsorption capacity, photocatalytic activity, stability, regeneration, economic viability, and environmental performance must be investigated through experimental studies, life-cycle assessments, and complementary techno-economic analyses (Fawzy et al. 2022).