Digital technologies in the circular economy: a bibliometric analysis

In recent years, the circular economy (CE) has emerged as a transformative solution to tackle resource scarcity, environmental degradation and rising costs by focusing on reuse, recycling and regenerative practices to promote sustainability and innovation (MacArthur, 2013; Korhonen et al., 2018; Corvellec et al., 2022; Velenturf and Purnell, 2021). Supported by technological advancements, shifting consumer preferences and stricter regulatory frameworks, circular models reduce waste, extend product lifecycles and unlock new economic opportunities across various industries (MacArthur, 2013; Sánchez-García et al., 2024).

Digital technologies (DT) play a critical role in enabling and scaling CE practices, as they provide the infrastructure, data and connectivity required to optimize resource use, reduce waste and support sustainable operations (Rizvi et al., 2021; Banihashemi et al., 2024; Liu et al., 2022). The concurrent rise of digital innovation and the increasing urgency of CE transitions underscore the importance of understanding the intersection between these two domains. As sustainability challenges intensify, the integration of DT into CE strategies is gaining prominence as a critical pathway toward systemic change.

Despite growing scholarly interest, research in this area remains fragmented. Existing studies often approach the topic from diverse disciplinary perspectives and employ predominantly qualitative methods, particularly Systematic Literature Reviews (SLRs). For instance, Okorie et al. (2018) developed an early framework linking digital tools to circular strategies while Cagno et al. (2021) analyzed digital enablers through the ReSOLVE framework, identifying gaps in the integrated CE approaches. In addition, Chauhan et al. (2022) synthesized key enabling technologies, barriers to adoption and emerging digital business models such as Product-Service Systems. These reviews collectively highlight the fragmented nature of existing research and emphasize the importance of a comprehensive, data-driven perspective and this bibliometric analysis aims to contribute to this canon.

While these reviews offer valuable insights through content synthesis, they tend to lack a comprehensive, quantitative perspective on the intellectual structure of the field. Notably, according to our observation, only one bibliometric review to date has systematically examined the DT–CE interface, leaving a significant gap in the literature.

Given the rapid evolution of digital solutions and their expanding role in CE, there is a pressing need for a bibliometric analysis that can quantitatively map the landscape. Such an approach complements existing qualitative reviews by identifying research trends, influential contributors, thematic clusters and structural patterns within the literature (Donthu et al., 2021). Moreover, it enables a clearer understanding of how knowledge in this domain is organized and how it has evolved over time.

In response to this gap, the present study conducts a comprehensive bibliometric analysis of 722 Scopus-indexed publications from 2016 to early 2025. Using established bibliometric techniques – citation analysis, co-authorship networks, keyword co-occurrence and thematic evolution – it provides an integrated and data-driven overview of the field. This methodology allows for the identification of emerging research directions and offers a foundation for both academic exploration and practical application of digital innovations in support of circular economy transitions.

The study makes a timely and significant contribution by extending the depth and scope of previous reviews. It captures recent, policy-relevant developments, such as digital product passports, and applies a broader set of bibliometric tools to offer a dynamic and nuanced understanding of research progression. Furthermore, it introduces a novel conceptual model linking digital technologies with CE principles, providing a structured framework for interpreting the field's thematic evolution. Together, these contributions reinforce the study's originality and relevance to ongoing academic, industrial and policy discussions on sustainable development.

The subsequent sections of this paper are organized as follows: First, the paper presents the theoretical background of the study, discusses the research gap and then outlines the research design, which serves as a guiding framework for the research process. Next, it delves into the methodology, offering a structured approach used to conduct the study. Following this, the paper examines the knowledge roots (past), maps the current landscape (present) and proposes how DT and CE might develop further (future). Finally, the paper concludes by emphasizing its originality and contributions, addressing its limitations and providing recommendations for future research.

Core concepts are the foundational constructs that form the basis of a study on the CE or theoretical framework based on the study of Le et al. (2024):

1. Waste Elimination: The circular economy promotes waste elimination by designing products and processes that minimize material use, prevent pollution and integrate renewable energy, reducing environmental impact across the product lifecycle.

2. Resource Utilization: The circular economy maximizes material value through product longevity, repair and recycling, reducing raw material use and enhancing sustainability.

3. Value Retention: The circular economy promotes value retention by encouraging reuse, remanufacturing and recycling to extend product lifecycles and minimize waste and resource depletion.

The CE principles guide the design, production and consumption of products and services to eliminate waste, maximize resource efficiency and retain value across the entire lifecycle (Le et al., 2024).

1. Design for Circularity: Product design should enable disassembly, reuse and recycling by emphasizing modularity, durability and recyclable materials to support circular economy goals.

2. Product Longevity: Circular economy promotes extending product life through maintenance, repair, refurbishment and upgradability to reduce waste and resource consumption.

3. Material Recovery: Circular economy emphasizes recovering and reusing materials through recycling, waste management systems and recovery technologies to reduce waste and conserve resources.

4. Sustainability Integration: Implementing circular economy effectively requires aligning values, operations and decisions with sustainability principles to foster environmental responsibility and long-term development.

5. Collaboration and Innovation: Collaboration across the value chain drives innovation, enabling new circular solutions, business models and technologies through shared resources, knowledge and best practices.

Based on the literature as mentioned above (Le et al., 2024), the key digital technologies used in the CE are as follows:

1. Internet of Things (IoT): A computational system that collects and exchanges data acquired from electronic devices (Okorie et al., 2018). It consists of sensors and actuators connected by networks to monitor or manage objects (Pagoropoulos et al., 2017).

2. Big Data/Analytics: Technologies that assist in extracting business value through the analysis of data (Pagoropoulos et al., 2017). It involves analyzing large volumes of data for insights and decision-making (Wynn and Jones, 2022).

3. Artificial Intelligence (AI): Intelligence demonstrated by machines, involving tasks such as learning and problem-solving (Wynn and Jones, 2022). AI-associated digital technologies could increase energy efficiency and facilitate the circular economy (Wynn and Jones, 2022).

4. Blockchain: A secure, immutable, traceable and visible public ledger system. It can reduce supply chain inefficiencies through its tracking and authentication capabilities (Wynn and Jones, 2022).

5. Cloud Computing: Using a network of remote servers hosted on the internet to store, manage and process data rather than a local server or personal computer (Chauhan et al., 2022).

6. Smart Tags/RFID: Smart tags are small devices that integrate memory, data processing and communication capabilities (Demestichas and Daskalakis, 2020). Smart tags were often used with Radio-frequency identification (RFID), which is a data collection technology that uses electromagnetic fields to automatically identify and track tags attached to objects (Pagoropoulos et al., 2017). Networked smart tags and RFID systems provide complete product lifecycle information (Pagoropoulos et al., 2017).

7. Additive Manufacturing: Additive manufacturing, especially 3D printing, represents agile and connected prototyping of parts or products on a large scale, enabling customization (Okorie et al., 2018). It can improve material recovery rates in plastic recycling (Liu et al., 2021).

8. Cyber-physical systems (CPS): Computational systems that collect and exchange data acquired from electronic devices and are listed as key technologies of Industry 4.0 (Okorie et al., 2018).

9. Automation/Robotics: The use of technology to automate tasks and processes, often involving machines that can perform complex actions (Wynn and Jones, 2022). Robot technology in waste management can improve disposal efficiency (Liu et al., 2021).

10. Digital Twin: Closely related to CE design, digital twin is a digital representation of a physical entity or system, enabling simulation and analysis (Liu et al., 2021).

Focus on integrating DT like IoT, AI, blockchain and big data analytics into the CE has been growing in recent years, recognizing their potential to enable and accelerate sustainable circular strategies, as explored in numerous reviews (Table 1).

As illustrated in Table 1, the existing literature highlights the integration of DT into the CE through various methodologies and objectives. Pagoropoulos et al. (2017) and Okorie et al. (2018) explored the role and maturity of DTs such as IoT and Industry 4.0 in CE transitions while Liu et al. (2022) addressed how DT functions enhance circularity and Uçar et al. (2020) identified DT roles for CE transition through case studies. Barriers to a smart CE, which is an advanced form of the circular economy that integrates digital technologies and smart systems, were analyzed by Trevisan et al. (2023) and Demestichas and Daskalakis (2020). Additionally, reviews by Cagno et al. (2021) and Chauhan et al. (2022) examined DTs within CE frameworks such as ReSOLVE and CE implementation pathways. Bibliometric insights by Liu et al. (2021) mapped trends, while Neri et al. (2024) and Hariyani et al. (2024) explored DT contributions and mediators in CE adoption.

The research questions of this study, which are related to uncovering the foundations, current trends, collaboration and predicting research directions, are as follows.

Together, these three questions structure the investigation. RQ1 establishes historical context, RQ2 examines the present landscape and RQ3 explores future possibilities. The following diagram (Figure 1) illustrates this sequential relationship, which underpins the study's narrative.

Building on existing literature and our previous work (Le et al., 2024), we propose the Review Scope and Conceptual Foundations framework (Figure 2) to provide a structured lens for interpreting the bibliometric findings. This framework defines the boundaries of the review, clarifies its thematic focus and introduces the core concepts and theoretical foundations that underpin the analysis. Specifically, it synthesizes three interrelated components frequently discussed at the intersection of digital technologies (DT) and the circular economy (CE): core concepts, principles and digital technologies. Rather than serving as a model for empirical testing, the framework functions as a conceptual guide to structure the subsequent analysis and discussion.

This study employs a bibliometric analysis to investigate existing research on the integration DTs within the CE. As mentioned in Figure 3, the methodology is structured into the following steps (Donthu et al., 2021): Data collection, study screening, study selection, final selection and bibliometric analysis.

Data Collection. Scopus was selected as the sole database due to its comprehensive coverage of peer-reviewed journals across relevant fields (e.g. engineering, environmental science, information systems) and its compatibility with bibliometric tools including VOSviewer and Bibliometrix (Burnham, 2006). Using a single, high-quality source ensures consistency in metadata and citation structures (Donthu et al., 2021). The potential limitation regarding database coverage is explicitly discussed in the Limitations section. The search string was designed to include key terms such as “digital technologies” and “circular economy,” and their synonyms in titles, abstracts and keywords. After this step, there were 1,230 papers selected.

Study Screening and Selection. This review employs a bibliometric analysis protocol designed to quantitatively map the research landscape, rather than synthesize content qualitatively (Donthu et al., 2021). Our approach emphasizes identifying publication trends, co-authorship networks, keyword co-occurrence and thematic evolution – core elements of structure-focused reviews. This protocol is particularly suitable for capturing the dynamic and interdisciplinary nature of research on DTs in the CE. Moreover, the study limited the timeframe from 2016 to 2025 to focus on the recent surge in research linking digital technologies with circular economy strategies, coinciding with the rise of Industry 4.0 and sustainability-driven digital transformation initiatives. Following an exploratory search, the initial 1,230 publications were screened by year (2016–2025) and language (English) to yield 951 papers. Further filtering by document type (articles, proceedings, chapters) resulted in 930 potentially relevant publications, retrieved using the refined Scopus query structured as follows:

(TITLE-ABS-KEY(“circular economy” OR “circularity” OR “resource efficiency” OR “closed loop” OR “resource recovery” OR “waste management” OR “remanufacturing”) AND TITLE-ABS-KEY(“digital technology” OR “digitalisation” OR “digitalization” OR “industry 4.0” OR “I4.0” OR “IoT” OR “internet of things” OR “artificial intelligence” OR “AI” OR “machine learning” OR “blockchain” OR “big data” OR “analytics” OR “cyber physical system” OR “CPS” OR “digital twin” OR “additive manufacturing” OR “3D printing”)) AND PUBYEAR >2015 AND PUBYEAR <2026 AND (LIMIT-TO (DOCTYPE, “ar”) OR LIMIT-TO (DOCTYPE, “cp”) OR LIMIT-TO (DOCTYPE, “ch”)) AND (LIMIT-TO (LANGUAGE, “English”))

Final Selection. The 930 papers obtained through objective filtering then underwent a final relevance screening. This step was essential to focus the corpus on studies directly addressing the DT-CE intersection, as the broad search inevitably included papers with only tangential mentions. Two authors independently reviewed these papers against predefined inclusion/exclusion criteria linked to our Review Scope and Review Scope framework and specific research questions. Papers were retained (n = 722) only if they substantively met both core requirements: (1) explicit discussion of specific DT applications (e.g. IoT, AI, Blockchain, specific tools beyond generic digitalization) impacting or enabling CE principles/strategies (e.g. waste elimination, resource utilization, design for circularity) AND (2) demonstration that both DT and CE were central themes, not peripheral context. Papers focusing solely on DT or CE without a clear interactive link, or on related but out-of-scope topics, were excluded. Inter-rater discrepancies were resolved through consensus discussion with the third author. The resulting 722-paper dataset represents publications judged highly relevant to the study's core questions.

Bibliometric Analysis. The bibliometric analysis was conducted using VOSviewer and R Bibliometrix packages (Arruda et al., 2022; Derviş, 2019). These tools enabled the extraction of citation data, co-authorship networks, keyword co-occurrence and thematic clusters. Visualization techniques were utilized to identify patterns and relationships, such as influential authors, journals, institutions and geographic regions (Arruda et al., 2022; Derviş, 2019).

This study applies four core bibliometric techniques aligned with the three research questions (RQs). To address RQ1 (works, key authors and citations), co-citation analysis was used to uncover intellectual structures that have shaped the digital technologies–circular economy (DT–CE) field. For RQ2 (emerging themes, collaboration, clusters), most cited papers and frequent words, bibliographic coupling and co-authorship analysis identified current research clusters, institutional collaborations and dominant contributors. Co-word analysis complemented this by mapping thematic concentrations and keyword trends. Finally, to explore RQ3 (future research directions), co-occurrence network, thematic map and thematic evolution forecasted potential research frontiers and underexplored niches.

The evolution of papers per year shows a steady increase over time, with notable growth observed in recent years and continuing into early 2025 (Figure 4). Starting with just nine papers in 2016, production rose gradually, reaching 45 by 2020. A sharper increase followed, with 93 papers in 2021 and 121 in 2023. By 2024, a significant leap to 203 papers occurred, reflecting rapid expansion, likely due to improved efficiency or higher demand.

The subject areas related to this study demonstrate the interdisciplinary nature (Figure 5). Environmental Science leads with 301 papers, reflecting the focus on sustainability, followed by Engineering (286) and Computer Science (209), highlighting technological innovations related to digital technologies. Energy (185) and Social Sciences (149) emphasize renewable systems and socio-economic impacts, while Business (148) and Economics (71) explore financial and operational strategies. Fields including Decision Sciences (57), Materials Science (33) and Chemistry (24) focus on optimizing processes and materials, while contributions from Psychology (12) and Arts and Humanities (5) address human and cultural dimensions, showcasing the diverse yet uneven engagement across disciplines.

Citation analysis helps highlight the most influential publications, authors and journals by evaluating metrics such as total citations and H-index, offering insight into research impact and visibility (Donthu et al., 2021).

As shown in Table 2, among the top 20 sources, the Journal of Cleaner Production stands out with 2,901 citations for 36 papers, yielding an impressive 80.58 citations per paper and the journal's H-index of 354 (metrics based on Scopus data as of March 2025). Sustainability (Switzerland), though leading in volume with 53 papers, has a lower impact per paper (29.85 citations, H-index 207). Technological Forecasting and Social Change achieves the highest average impact per paper at 107.22 with 965 citations from just nine papers (H-index 209). Other high-impact journals include Resources, Conservation and Recycling (51.65 citations/paper) and Waste Management (47.90).

Bibliometric indicators are grounded on the premise that the impact of a scientific paper can be assessed by the number of times it is cited by other publications. Based on its analysis, this study identifies 391 unique authors, each contributing between one and five publications to the field (Table 3). Among the authors in the field, several stand out with more than 100 total citations, indicating significant scholarly influence. For example, Surajit Bag leads with 803 citations from four papers, averaging an exceptional 200.75 citations per paper, highlighting both high productivity and impact. Furthermore, Yigit Kazancoglu has 550 citations across three papers (183.33 per paper), followed by Charbel J. Chiappetta Jabbour with 333 citations from three papers (111.00 per paper).

Our bibliometric analysis addresses the research questions sequentially: co-citation analysis identifies works (RQ1), bibliographic coupling reveals current trends and collaborations (RQ2) and co-word analysis highlights emerging themes and future directions (RQ3). This combined approach offers a comprehensive map of the field's past development, current state and potential trajectories.

To uncover the foundations of a research field, bibliometric analysis offers powerful tools for exploring its intellectual structure and historical evolution. Focusing on the past, this approach seeks to identify the key contributions and knowledge bases that have shaped the integration of DT into the EC. Central to this analysis is co-citation analysis, which reveals thematic clusters and intellectual linkages by examining how often two works are cited together (Surwase et al., 2011).

Co-citation analysis reveals relationships through shared citations. Document co-citation links related publications, journal co-citation helps identify core journals and author co-citation is widely used to map a field's intellectual structure and thematic connections (Surwase et al., 2011).

Figure 6 presents the author co-citation network, revealing influential authors and distinct intellectual clusters within the DT-CE research landscape. The prominence of Sarkis, indicated by the largest node size, confirms his foundational role and significant influence across the field. His work is central and highly influential in the intersection of DT and the CE, as his publications are frequently referenced with other scholars' work.

Concerning the clustering of authors, Figure 6 displays several distinct clusters of authors, indicated by different colors (red, blue, green):

1. The blue cluster forms a dense core, uniting highly co-cited authors such as Sarki, Govindan, Kazançoğlu and Bag. The strong interconnections suggest a well-established theoretical base within this group.

2. The red cluster is another prominent group featuring authors including Liu, Y. Wang, Li J., Chen, Li X. and Z. Wang. While these authors may not be among the most globally cited, the predominance of Chinese names might suggest a regional focus or a specific thematic concentration. Further investigation into the specific works of these co-cited authors is warranted to fully delineate this cluster's thematic identity.

3. The green cluster is led by some of the most influential authors, such as Terzi, Rosa, Ramakrishna and Sassanelli. This cluster likely represents another distinct set of researchers contributing to this field, possibly with a different focus compared to the red and blue clusters.

Overall, the distinct author clusters suggest that the field, while interconnected, is shaped by several distinct research streams or schools of thought. Recognizing these different intellectual anchor points helps researchers effectively position their contributions within the existing scholarly landscape.

The journal co-citation network (Figure 7) maps the influential publication venues and delineates the core thematic pillars of DT-CE research. Three major clusters emerge, representing distinct but interconnected knowledge domains.

1. Green Cluster – Environmental and Sustainability Science: This cluster is characterized by journals such as Journal of Cleaner Production, International Journal of Production Research and International Journal of Production Economics and is strongly aligned with the core concepts of the research model. Its central position and the prominence of the Journal of Cleaner Production underscore the foundational importance of environmental management and process optimization perspectives in this field.

2. Blue Cluster – Recycling Technologies: This cluster includes journals such as Resources, Conservation and Recycling, Journal of Industrial Ecology and Waste Management and is closely aligned with the principles of the research model, particularly Material Recovery and Product Longevity. The cluster's focus on resource efficiency, waste valorization and industrial process optimization reflects key circular economy strategies aimed at extending product lifecycles, enabling recycling systems and fostering technical innovations that reduce material loss and environmental impact.

3. Red Cluster – Technology Foresight: This cluster is anchored by journals such as Sustainability, Procedia CIRP and Engineering and is strongly associated with the Digital Technologies identified in the research model. This cluster reflects research on digital transformation, Industry 4.0 and business model innovation in CE, driven by technologies including IoT, AI, blockchain, CPS and digital twins, which form the backbone of digitally-enabled circular strategies.

Having established the foundations (RQ1), we now examine present-day trends (RQ2). This section analyzes the current research landscape, mapping thematic clusters, collaboration networks and the state of DT-CE research using cited document analysis, frequent words, bibliographic coupling and co-authorship networks.

Most cited papers serve as intellectual pillars representing works in the evolution of the research field and trend indicators reflecting dominant paradigms and emerging areas (Paunović et al., 2024). Table 4 presents the ranking of the Top 10 Most Cited that serves as the intellectual backbone of the research domain, with Kristoffersen et al. (2020) and Upadhyay et al. (2021) leading in total citations.

These papers are core intellectual drivers in DT and CE, combining strategic frameworks, digital enablers and empirical methods. High TC/year indicates strong recent relevance – especially Chauhan et al. (2022) leads with 94.5 citations/year, indicating it is a rapidly influential paper despite being recent. Moreover, Upadhyay (2021) and Bag et al. (2021) follow closely, showing sustained high relevance in integrating digital transformation and circular economy.

Most frequent words highlight a research domain's key themes, concepts and methodological approaches (Figure 8). In this study, high-frequency author keywords such as circular economy, sustainable development, recycling, waste management and digital technologies indicate the field's core focus and evolving research directions. These keywords reflect both foundational environmental concerns and growing digital integration, helping researchers identify dominant topics and thematic clusters.

Analyzing keyword frequency over time (Figure 9) reveals evolving research focus. While terms such as “circular economy,” “sustainable development” and “recycling” grew steadily from 2016–2025, “digital technologies” and “information management” surged notably after 2020, highlighting increasing digital integration and suggesting a paradigm shift towards digitally-enabled circular practices.

Clustering by bibliographic coupling identifies papers that share similar reference lists, thereby revealing contemporary research fronts and thematic cohesion. Unlike co-citation (which reflects intellectual roots), coupling highlights ongoing scholarly conversations (Gheno, n.d.). Figure 10 presents a strategic diagram based on bibliographic coupling, mapping the current research fronts (themes actively citing similar recent literature) by their centrality (X-axis) and density/impact (Y-axis).

The purpose of a cluster in a clustering map is to represent a group of closely related nodes that share strong conceptual or citation links. Table 5 identifies five thematic clusters in DT and CE research, ranging from core topics like circular economy and sustainability to emerging areas such as digital product passports. These clusters reflect thematic maturity, niche focus and growing innovation trends.

The publication data reveal significant global contributions to DT-CE research (Table 6). Italy emerges as the most prolific country in this dataset with 89 papers, followed by the UK (79), India (78), China (68) and Germany (61). Several other nations also demonstrate substantial output, including Spain (47), the USA (40), the Netherlands (34) and Sweden (32). Further contributions from countries such as Brazil (27), France (27) and Finland (24) underscore the broad international engagement. Overall, these figures highlight a diverse and competitive global research landscape in this field.

The geographical distribution of citations (Figure 11) reveals variations in impact. While Italy leads in paper volume, its average citations per paper (TC/Paper) are lower (22.01), whereas the USA shows high impact (50.38 TC/Paper) from fewer papers. The UK and the Netherlands demonstrate a balance of volume and impact (over 45 TC/Paper), while high-output countries such as India and China show moderate citation averages. (Note: Averages for countries with very few papers, including South Africa, should be interpreted cautiously). Additionally, Figure 11 illustrates international collaboration links, identifying China, India, USA, Germany and the UK as key hubs, with notable partnerships like China-UK/USA and India-USA/Europe.

To predict future research directions (RQ3), this section analyzes the field's conceptual structure using co-word analysis. Based on keyword co-occurrence patterns, we utilize the co-occurrence network, thematic map and thematic evolution outputs (Bibliometrix approach) to identify emerging themes and potential gaps (Derviş, 2019).

Figure 12 presents a keyword co-occurrence network to visualize how frequently and closely keywords appear together in publications, thus revealing thematic clusters, conceptual linkages and emerging areas in a research field (Derviş, 2019).

As presented in Figure 12, two main clusters emerge.

1. Cluster 1 (Red): The dominant terms in this cluster are recycling and waste management. Other significant keywords include decision-making, sustainability, economics, environmental impact and environmental technology. This cluster appears to be centered on the more traditional aspects of the circular economy, focusing on waste treatment, recycling processes and their economic and environmental implications. The presence of artificial intelligence and information and communication technology suggests the application of digital tools to optimize recycling and waste management processes.

2. Cluster 2 (Blue): This cluster is heavily focused on the intersection of circular economy, sustainable development and digital technologies. Prominent related terms include information management, supply chains, information and communication technology, supply chain management, blockchain, IoT and Industry 4.0. This suggests a theme centered around how digital technologies are being applied to improve supply chains' and industrial processes' circularity and sustainability, potentially leveraging concepts like blockchain and IoT for better information management. The presence of terms like product design, building information modeling and digital product passport indicates a focus on incorporating circular economy principles from the design stage and using digital tools for product lifecycle tracking.

The thematic map in Figure 13 provides a dynamic view of the research landscape, categorizing themes based on their centrality (relevance) and density (development).

1. Cluster 1 - Green: This cluster includes keywords “economic aspect” and “human”. Its position signifies a niche theme – one that is relatively well-developed internally (higher density) but less connected to the rest of the field (lower centrality). This suggests a specialized research stream focusing perhaps on the socio-economic factors, human behavior aspects or cost-benefit analyses related to CE and potentially DT, operating somewhat apart from the core technical or strategic discussions.

2. Cluster 2 - Blue: This cluster contains the keywords “waste-management” and “sustainability”. This position characterizes it as a basic theme – highly central and relevant to the entire field, but perhaps less internally developed or cohesive as a distinct research program (lower density). These keywords represent foundational pillars or cross-cutting concepts that are essential starting points or integration points for much of the research in the DT-CE domain.

3. Cluster 3 - Red: This cluster features “sustainable development” and “life cycle”. Its low centrality and low density suggest this theme is either emerging and not yet fully integrated into the field's core structure, or potentially declining or fragmenting. The presence of those broad terms here, rather than in the basic or motor quadrants, is somewhat surprising and might indicate their use in specific, less connected contexts within the analyzed literature, or perhaps represent older foundational concepts losing thematic momentum compared to newer DT-focused terms (which are notably absent as major themes in this specific map visualization).

The difference between the co-occurrence network (two clusters) and the thematic map (three clusters) is not unusual and does not affect the reliability of the results and these methods serve different analytical purposes. The thematic map, based on Callon et al.‘s centrality and density measures, identifies clusters by examining how strongly themes are developed and how much they interact with others (Callon et al., 1983). In contrast, co-occurrence analysis groups keywords that frequently appear together, showing dominant topic relationships (Van Eck and Waltman, 2009). Rather than contradicting each other, these methods offer complementary insights and together provide a clearer view of the research field.

The thematic evolution map (Figure 14) traces the trajectory of core research themes across three distinct periods (Derviş, 2019), revealing a clear conceptual progression in the DT-CE field. These periods (2016–2020, 2021–2023, 2024–2025) were demarcated to analyze potential shifts corresponding to the field's early growth phase, its subsequent rapid acceleration (coinciding with increased volume seen in Figure 4 and major policy initiatives such as the 2020 EU CE Action Plan (Johansson, 2021)), and the most recent trends (potentially reflecting newer regulatory pushes like the ESPR proposal). While the resulting map (Figure 13) is visually dense due to the richness of keyword connections, it reveals a clear overarching narrative of conceptual progression:

2016–2020 – Laying the Foundations of Digital CE. During this initial phase, research explored enabling digital technologies including Big Data, Blockchain, IoT and Machine Learning for Circular Economy. Key themes involved applying these tools alongside concepts like Life Cycle Assessment (LCA) and remanufacturing to address waste, material recovery (e.g. via BIM in construction) and measure environmental impacts. Early considerations of business models, sustainability and Industry 4.0 framed the DT-CE link within a broader systemic transformation, laying the conceptual and technological groundwork for future research.

2021–2023 – Strategic Integration and Systemic Alignment. This period shifted towards strategic integration and application. Key digital tools such as AI and Big Data were viewed less in isolation and more as integral components for decision-making, optimization and traceability, often linked to information management and resource efficiency across value chains. Research expanded into spatial dimensions (using BIM, GIS) for urban/territorial CE and focused on operationalization through themes including industrial ecology, resource recovery and supply chain management. The field also diversified, incorporating implementation barriers, marketing and sector-specific issues (e.g. food waste), while maintaining a strong link to sustainable development goals, indicating a phase of broader integration and application. Key terms including information management, circular manufacturing and resource recovery show principles being embedded through collaboration and innovation across value networks, spurred by pandemic-era digital acceleration and ISO 59000 CE standards drafts (Arana-Landin et al., 2024).

2024–2025 – Operationalization, Policy Integration and Sectoral Impact. This 2024–2025 period signifies a shift towards practical implementation, policy alignment and measuring impact. The emergence of digital product passport (DPP) (Faraca et al., 2024) is prominent, reflecting responses to regulations like Ecodesign for Sustainable Products Regulation (ESPR) and driving the use of blockchain, digitalization and digital transformation for traceability and compliance. Research also increasingly focuses on applied business models and assessing circularity to measure value creation and scale up practices. Core digital technologies (AI, ML, IoT, Industry 4.0, BIM) remain central but are now deeply interwoven with executional goals and system integration, grounded in industrial ecology and environmental sustainability. This phase marks an evolution towards tangible, policy-aligned and outcome-driven DT-CE research aimed at real-world transformation.

This study employed bibliometric analysis to map the evolving scholarly landscape at the intersection of DT and the CE, analyzing a corpus of 722 Scopus-indexed publications from 2016 to early 2025. The analysis highlights rapid publication growth post-2020 and identifies influential contributors, journals and collaboration networks within this dataset. Key thematic findings include a demonstrated evolution from foundational CE concepts towards operationalization and, more recently, towards digitally-enabled institutionalization, with technologies such as IoT, AI, Blockchain and Digital Product Passports becoming increasingly central.

The primary contribution of this work is providing an updated (2016–2025) structural overview of the DT-CE research field based on the analyzed corpus. By mapping key themes, their evolution and prominent players, this study offers a quantitative perspective complementing existing qualitative reviews, potentially aiding researchers in navigating this dynamic area. Identifying high-impact journals and keywords may also offer guidance. To enhance practical and research relevance, the study highlights key findings, offers actionable recommendations, such as using AI and blockchain for traceability, and suggests future research on digital product passports and technology adoption, ensuring alignment with the thematic analysis and supporting circular economy transitions.

While Liu et al. (2021) provided a bibliometric review of the digital economy and CE integration, our study offers distinct contributions based on several key differences. Firstly, our analysis covers a more extended and recent timeframe (2016-early 2025 vs Liu et al.'s likely earlier cutoff), capturing the significant acceleration in publications and thematic shifts post-2020, including the emergence of critical topics such as Digital Product Passports, which were less prominent previously. Secondly, our methodology employs a broader range of bibliometric techniques reported in detail, including thematic evolution analysis and detailed mapping of recent collaboration networks, allowing for a more dynamic view of the field's progression beyond the structural overview provided by Liu et al. Thirdly, our study uses co-occurrence and thematic mapping to explicitly link specific digital technologies with core CE principles. This offers a granular view of how distinct DTs enable specific CE strategies, complementing Liu et al.'s broader conceptual work. Our thematic evolution analysis further details the shift towards operationalization driven by specific digital innovations and policies, an insight missing previously. Thus, we provide a necessary, nuanced mapping of the maturing DT-CE landscape, justifying our contribution beyond an incremental update.

At first, key findings reveal rapid publication growth post-2020 and identify influential authors (e.g. Bag, Kazancoglu and Govindan from Table 3) and core works (e.g. Kristoffersen et al. from Table 4) within this literature set. High-impact journals including Sustainability, Journal of Cleaner Production and Resources, Conservation & Recycling are prominent outlets (Table 2). Analysis of frequent keywords (e.g. “circular economy,” “recycling,” “digital technologies” in Figure 8) confirms the field's focus on blending foundational environmental concerns with digital enablers like AI, IoT and blockchain. Network analyses revealed thematic clusters, including emerging niches focused on digital product passports (Table 5), highlighting the coexistence of technology-driven and policy-oriented approaches. Geographically, while European nations such as Italy and the UK, alongside India and China, lead in publication volume, while others including the USA show high citation impact despite lower output (Table 6, Figure 11).

Thus, synthesizing these findings suggests several overarching trends within the analyzed literature. First, the thematic evolution observed suggests moving from foundational concepts towards more data-driven approaches and implementation discussions. Second, while Environmental Science and Engineering remain dominant subject areas (Figure 5), the increasing presence of Business, Social Sciences and Economics points towards a broadening disciplinary perspective incorporating technical, managerial and policy considerations. Third, the growing publication focuses on digital solutions, particularly the post-2020 surge seen in keyword frequency (Figure 9), indicates their increasing importance in scholarly discussions surrounding how circular initiatives are planned and potentially governed.

Moreover, these findings can also be interpreted through the lens of the proposed Review Scope and Conceptual Foundations framework (Figure 2), which outlines Core Concepts, Principles and Digital Technologies based on existing literature. For instance, journal co-citation clusters (Figure 7) broadly correspond to these layers (e.g. environmental journals with Core Concepts, recycling journals with Principles, technology journals with Digital Technologies). Temporal analyses (Figure 9, Figure 14) suggest progression consistent with the model's logic, moving from Core Concepts to Principles and increasingly integrating Digital Technologies over time. While not a formal test, this alignment suggests the model serves as a useful heuristic for understanding the interplay between CE and DT in the literature. Future work using other methods is needed to validate the model's proposed mechanisms.

Finally, the study further emphasizes the prominent role of digital technologies within environmental science, engineering and computer science – disciplines that are pivotal to advancing CE strategies. Environmental science provides sustainability metrics and waste reduction frameworks, engineering drives innovations such as smart manufacturing and resource-efficient design, while computer science underpins data-driven solutions through technologies including AI, IoT, and Blockchain. The high volume of publications in these fields reflects their foundational contributions to digitally-enabled circular transitions. This interdisciplinary convergence highlights the field's strong technical and scientific underpinnings and reinforces the critical role of each domain in supporting sustainable digital transformation.

The findings must be interpreted with significant caution due to several limitations. The reliance on the Scopus database may omit relevant works, and the defined timeframe might miss the newest developments. More significantly, the construction of the final 722-paper corpus involved a manual relevance screening step; while guided by explicit criteria, this process carries an inherent risk of selection bias that could impact the representativeness and generalizability of the results. Furthermore, the quantitative nature of bibliometrics provides limited insight into the practical implementation or real-world nuances of DT adoption in the CE. Finally, data collection occurred in early 2025, potentially excluding later publications.

Acknowledging these constraints, future research should integrate bibliometric insights with qualitative methods including organizational case studies to assess DT adoption, readiness and policy effectiveness. Cross-industry and regional analyses could uncover specific challenges and best practices. Investigating the impact of emerging technologies such as edge computing and refining the understanding of digital twins within CE frameworks also remain important avenues for subsequent research.

This study used only publicly available data from Scopus, thus not requiring formal ethical approval or informed consent.