Sustainable Natural Product Drug Discovery: Biodiversity Protection, Ethical Sourcing, AI Screening, Green Chemistry, and Therapeutic Innovation

Download PDF

INTRODUCTION

Natural products have contributed substantially to modern pharmacotherapy as direct medicines, molecular templates, pharmacophores, and sources of structurally distinctive chemical matter. Their continued relevance reflects the capacity of biological systems to generate complex compounds that occupy chemical and functional spaces not always reproduced efficiently by conventional synthetic collections. This historical contribution provides a strong rationale for maintaining natural product research as part of contemporary drug discovery, but it does not establish that all natural-product-derived programs are environmentally sustainable, ethically sourced, safe, developable, or therapeutically successful [1].

The biological origins of natural products create responsibilities that extend beyond compound identification and activity testing. Medicinal plants, fungi, microorganisms, and other biological resources occur within taxonomic, evolutionary, ecological, geographic, and cultural contexts that shape their availability and appropriate use. Modern biodiversity science can strengthen discovery by improving species authentication, clarifying biological relationships, connecting specimens with collections and distribution data, and identifying gaps in knowledge about source organisms [2]. These capabilities are particularly important when chemical records are detached from voucher specimens, when closely related species are substituted in trade, or when a promising compound is associated with a biologically vulnerable or poorly characterized source.

Natural-product discovery is also exposed to environmental change. Climatic shifts, habitat modification, altered species distributions, land-use pressure, and changes in ecological interactions may affect the abundance, chemistry, accessibility, and reproducibility of medicinal resources. Sustainable research strategies therefore require attention to conservation status, ecological vulnerability, plant parts used, regeneration capacity, geographical variation, and the possibility that increasing research or commercial demand could intensify pressure on already constrained resources [3]. These concerns are not limited to large-scale harvesting; repeated small collections, poorly documented sampling, destructive acquisition of roots or bark, and duplicated extraction across research groups may collectively create avoidable burdens.

Technological advances offer opportunities to reduce some of these pressures, but their contribution must be interpreted cautiously. Natural product databases, metabolomics, genomics, artificial intelligence, advanced analytical chemistry, green extraction, biocatalysis, and engineered production can improve prioritization and broaden access to complex molecules, yet they do not replace source governance, biological validation, safety assessment, or translational evidence [4]. The objective of this article is therefore to propose an original sustainability framework that integrates biodiversity protection, ethical and traceable sourcing, access and benefit-sharing principles, respectful treatment of traditional knowledge, AI-assisted prioritization, extraction minimization, green chemistry, responsible production, therapeutic evaluation, lifecycle thinking, and explicit claim boundaries within a unified natural product drug discovery pathway.

Sustainability in natural product discovery

Sustainability in natural product discovery is best understood as a framework condition governing how biological resources, knowledge, data, chemical processes, and therapeutic opportunities are selected and managed. It requires the reconciliation of healthcare and discovery objectives with biodiversity conservation, responsible resource use, equitable relationships, and the long-term availability of medicinal species and fungi [5]. This definition differs from descriptions based solely on natural origin, renewable branding, a single lower-hazard solvent, community acknowledgment, or the discovery of a pharmacologically interesting molecule. A natural compound can originate from a threatened species, an extract described as green can depend on ecologically damaging harvesting, and an ethically described collaboration can remain inadequately traceable or governed.

A sustainability assessment must therefore begin before extraction or screening. It should consider the identity and distribution of the source organism, the ecological condition of its habitat, the vulnerability of relevant populations, the plant or organism part required, the geographic concentration of supply, and the potential effect of future demand. Global assessments of medicinal resources demonstrate that exposure to environmental change can be examined systematically and can reveal geographically differentiated vulnerabilities that may be overlooked in compound-centered research [6]. Such assessment does not prove that a particular population is declining, but it can identify uncertainty, direct field verification, and prevent candidate advancement from proceeding without awareness of ecological context.

Commercial availability is also an insufficient proxy for sustainable supply. Species used in established herbal markets may lack comprehensive conservation evaluation, may be affected by exploitation, or may enter supply chains through substitutions and poorly documented intermediaries. Evidence from the herbal-medicine sector indicates that conservation status and evaluation coverage can remain incomplete even where species are already in commercial use [7]. Sustainability frameworks should therefore treat missing conservation information as a decision uncertainty rather than as evidence of low risk. Where vulnerability cannot be resolved, the appropriate response may include source substitution, use of archived material, cultivation studies, lower-demand analytical approaches, biotechnology-enabled production, or discontinuation of the candidate.

Social sustainability is inseparable from ecological stewardship because medicinal resources are frequently embedded within local livelihoods, customary practices, and community knowledge systems. Community-level research has shown that conservation incentives, economic benefits, resource use, and local participation can interact in ways that influence the persistence and governance of medicinal plants [8]. Responsible discovery should consequently include traceability, community engagement, source-country collaboration, supply-chain transparency, equitable participation, and procedures for responding to grievances or changing conditions. These elements must be connected with AI-assisted prioritization, green chemistry, safety evaluation, developability, and lifecycle review so that environmental or social burdens are not shifted from one stage of discovery to another.

Biodiversity protection and ethical sourcing

Biodiversity protection requires discovery teams to assess more than whether sufficient material can be obtained for an initial experiment. Relevant questions include whether collection could affect a threatened or unevaluated species, whether destructive plant parts are required, whether habitat is already under pressure, whether projected development demand could exceed regenerative capacity, and whether the source can be replaced through cultivation, renewable harvesting, microbial production, or synthesis. Ethical sourcing adds a parallel requirement for verifiable origin, documented custody, responsible supplier relationships, and transparent terms governing access and downstream use. Case-based evidence from biodiversity trade demonstrates that traceability, sustainable resource use, value-chain organization, and benefit-sharing arrangements must be considered together rather than treated as interchangeable claims [9]. The use of artificial intelligence or digital biological information does not remove provenance and benefit-sharing concerns, because computational systems may still depend on biological materials, associated knowledge, sequence information, chemical records, or data derived from governed resources [10].

Access and benefit-sharing principles require formal, context-sensitive governance rather than informal assurances that a resource was publicly available or scientifically interesting. Operational experience with genetic resources shows that access conditions, documentation, institutional responsibilities, transfer arrangements, downstream use, and benefit expectations can become complex across custodians, researchers, repositories, and commercial actors [11]. A sustainability framework should therefore include a governance checkpoint capable of identifying when institutional, source-country, provider, or community review is required. Material transfer governance may be incorporated where supported, with records linking the physical material, permitted uses, derivatives, associated data, onward transfers, retention periods, and restrictions. These measures support accountability but should not be represented as universal evidence of legal compliance or ethical certification.

Traditional knowledge can help identify culturally important species, preparations, therapeutic hypotheses, and underexplored biological relationships, but its scientific usefulness does not extinguish the rights or interests of knowledge holders. Responsible use requires attention to knowledge provenance, consent or authorization where applicable, culturally appropriate disclosure, community participation, attribution, protection against decontextualization, and independent evaluation of identity, quality, pharmacology, and safety. Scholarship on indigenous knowledge and medicinal plants emphasizes both its potential contribution to therapeutic research and the challenges of preservation, validation, ownership, and equitable use [12]. Traditional use should therefore be treated as a source of hypotheses and contextual evidence, not as proof of efficacy, safety, consent, ownership transfer, or unrestricted permission to commercialize.

Source-country collaboration can strengthen sustainability when it includes substantive scientific participation, shared capability, locally relevant research priorities, transparent data arrangements, and continuing engagement rather than symbolic affiliation. The development of an Ecuadorian herbal pharmacopoeia illustrates how traditional knowledge, biodiversity information, scientific monographs, linguistic accessibility, institutional participation, and biodiscovery objectives may be organized within a nationally grounded evidence system [13]. Such models are context dependent and do not automatically establish clinical validity, regulatory acceptance, or universal transferability, but they demonstrate the value of connecting biological resources and associated knowledge with source-country institutions and validation processes. Table 1 summarises biodiversity protection and ethical sourcing requirements for sustainable natural product drug discovery.

Table 1. Biodiversity Protection and Ethical Sourcing Requirements for Sustainable Natural Product Drug Discovery: Species Conservation, Habitat Pressure, Overharvesting Risk, Traceable Sourcing, Access and Benefit-Sharing Principles, Traditional Knowledge Respect, Community Engagement, Supply-Chain Transparency, Renewable Sourcing, and Governance Boundaries

AI screening and green chemistry

AI-assisted natural product discovery depends on the quality, coverage, standardization, accessibility, and provenance of its underlying evidence. Natural product databases can provide structures, names, source organisms, spectra, bioactivity records, and links among related compounds, but these resources differ substantially in curation, duplication, annotation quality, chemical representation, and interoperability [14]. Database search should consequently be paired with identity checks, source verification, uncertainty labeling, and recognition that absence from a database is not evidence of novelty. Dereplication provides a particularly direct opportunity to reduce unnecessary resource use because mass-spectral searching and related analytical approaches can identify known microbial metabolites before repeated isolation and characterization are undertaken [15]. A tentative spectral match, however, remains an identification hypothesis until supported by appropriate analytical confirmation.

Virtual screening, similarity analysis, machine learning, and other predictive methods can help rank compounds for defined targets, biological endpoints, pharmacokinetic properties, and toxicity concerns. Their sustainability contribution arises primarily from reducing low-value testing, narrowing experimental portfolios, identifying uncertainty, and directing scarce biological material toward higher-priority questions. Reviews of artificial intelligence in natural product drug discovery show that these approaches can support compound prioritization, bioactivity prediction, ADMET assessment, toxicity prediction, and data integration, while also remaining limited by heterogeneous datasets, sparse labels, model bias, inadequate benchmarking, and uncertain applicability domains [16]. AI rankings should therefore be interpreted as decision-support outputs requiring expert review and prospective experimental validation, not as proof of molecular activity, mechanism, safety, therapeutic value, ethical acceptability, or sustainability.

Computational prioritization should be linked to extraction and process design rather than treated as an isolated screening layer. Once a candidate or extract has sufficient evidentiary justification, green chemistry planning can compare extraction methods, solvent hazards, energy requirements, material efficiency, selectivity, purification burden, waste generation, and possibilities for recovery or reuse. Green extraction scholarship emphasizes reduced solvent use, safer inputs, lower energy demand, process intensification, and improved selectivity, but also shows that performance depends on the biological matrix, target compounds, equipment, scale, and complete process configuration [17]. A solvent described as natural, renewable, or alternative may still create toxicity, recovery, land-use, or energy burdens, while a low-energy extraction may be undermined by unsustainable harvesting or inefficient downstream purification.

Where direct extraction is ecologically constrained or chemically inefficient, sustainable synthesis, biocatalysis, microbial production, cell-based production, and synthetic biology may provide alternative routes to selected compounds or precursors. Synthetic-biology approaches can assist pathway discovery, access cryptic chemistry, reconstruct biosynthetic systems, and produce natural products in engineered hosts, but feasibility, productivity, containment, downstream processing, scale, and lifecycle performance remain compound specific [18]. These approaches should therefore enter the framework as conditional production options rather than preferred solutions by default. Waste reduction, energy reduction, process intensification, safety evaluation, and feedback from experimental results should be integrated into route selection. Figure 1 illustrates how AI screening and green chemistry can reduce redundant extraction, prioritize candidates, support safer design, and guide sustainable natural product discovery workflows.

Alt-text description

A conceptual workflow showing natural product evidence flowing through AI screening, dereplication, bioactivity and safety prediction, extraction minimization, green chemistry, sustainable synthesis, validation planning, and sustainability boundaries.

Therapeutic innovation pathways

Therapeutic innovation begins only after chemical identity and bioactivity have been separated from broader claims of medical usefulness. A natural product may produce an assay response yet remain unsuitable because of uncertain target relevance, nonspecific activity, instability, poor exposure, toxicity, limited supply, or an impractical manufacturing route. Reviews of natural bioactive products emphasize that discovery must progress through mechanistic investigation, safety assessment, pharmacokinetic characterization, formulation, and development planning before a compound can be treated as a credible therapeutic candidate [19]. Within a sustainability framework, these requirements also prevent ecological and social burdens from being justified by preliminary activity that may never translate into a viable intervention.

Modern development pathways may involve the original natural compound, a standardized mixture, a semisynthetic derivative, a simplified analogue, a biosynthetically produced molecule, or a pharmacophore-inspired scaffold. Contemporary approaches to plant-derived natural products combine chemical identification, computational screening, analogue generation, formulation, and experimental evaluation to address potency, selectivity, stability, supply, and safety limitations [20]. Responsible candidate prioritization should therefore compare therapeutic relevance with sourcing feasibility, ecological pressure, production alternatives, anticipated process burden, and equity considerations. A more potent analogue is not necessarily a more sustainable option if its synthesis requires hazardous reagents or if its development depends on continued extraction from a vulnerable source.

Integrated metabolomic and genomic approaches can improve novelty assessment, associate compounds with biosynthetic pathways, and focus isolation on chemical entities that are better supported by multiple evidence streams [21]. These methods may reduce untargeted extraction, but predicted biosynthetic relationships and molecular annotations still require structural and functional confirmation. Renewable production may also be considered where selected natural products can be produced through microorganisms, pathway engineering, or related biotechnological systems. Reviews of microbial production show that these platforms can support the manufacture of bioactive natural products and biologics, although productivity, genetic stability, scale-up, fermentation inputs, downstream purification, and process reproducibility remain compound dependent [22].

Therapeutic innovation should consequently be evaluated through a stage-gated process that combines target relevance, mechanistic plausibility, safety, developability, supply resilience, ecological safeguards, ethical governance, and expert review. Retrospective analyses suggest that natural-product-associated development programs can show favorable aggregate patterns across clinical development, but such findings cannot predict the success, safety, or approval of any individual candidate [23]. The proposed framework therefore treats clinical-development trends as contextual evidence rather than candidate-level validation. No discovery score, traditional-use history, mechanistic hypothesis, production route, or sustainability assessment permits a therapeutic claim without the appropriate experimental and clinical evidence.

Proposed sustainability framework

The proposed framework begins with definition of a bounded discovery question, followed by biodiversity and sourcing assessment and an ethical governance review. The discovery question should identify the therapeutic problem, intended biological hypothesis, relevant alternatives, and the minimum biological material and evidence needed to investigate it. Biodiversity assessment then examines species identity, conservation uncertainty, habitat pressure, harvesting method, projected demand, and the feasibility of cultivation, renewable sourcing, or alternative production. Ethical governance examines provenance, access conditions, benefit-sharing principles, traditional knowledge, community relationships, source-country collaboration, transfer conditions, and supply-chain accountability. Engineered production may be evaluated as an alternative to continued plant extraction when a biosynthetic route is technically plausible, as illustrated by synthetic-biology strategies proposed for ginsenoside production [24]. Such substitution remains conditional on comparative ecological, process, safety, and lifecycle evidence.

The next stages establish a natural product evidence foundation and apply AI-assisted prioritization, dereplication, and extraction minimization. Existing chemical records, spectra, biological assays, voucher information, source metadata, and knowledge provenance should be consolidated before new collection is considered. Computational approaches may then rank candidates, identify uncertainty, predict selected bioactivity or safety endpoints, and direct experimental resources toward defined hypotheses. Dereplication determines whether known chemistry can be recognized through archived materials or analytical comparison before additional biomass is processed. Green chemistry planning follows only after a candidate passes these evidentiary and governance gates. Biocatalysis may provide selective and potentially lower-burden routes to complex natural-product-related structures, although the complete route, enzyme production, solvents, cofactors, purification, and scale must be assessed rather than assuming superiority from the use of an enzyme [25].

Mechanistic and therapeutic evaluation, safety evaluation, and developability assessment form the scientific validation core of the framework. Confirmed chemical identity should precede target-engagement studies, and mechanistic claims should be supported by orthogonal, reproducible, and concentration-aware evidence. Safety assessment should address purity, metabolites, exposure, structural alerts, relevant toxicity endpoints, and formulation effects. Developability assessment should consider solubility, stability, pharmacokinetics, manufacturability, supply continuity, and quality control. At the process level, biocatalysis and other catalytic options should be compared through selectivity, reaction conditions, solvent requirements, material inputs, energy demand, waste generation, and downstream processing [26]. These comparisons enable route selection but do not independently establish environmental sustainability or therapeutic suitability.

A therapeutic innovation pathway is selected only after the evidence supports a defensible route for further investigation. The pathway may retain the original compound, modify it, reproduce it through biotechnology, redesign its synthesis, or discontinue it where ecological, ethical, safety, or developability risks outweigh the scientific opportunity. Green extraction and synthesis claims should be evaluated using multidimensional criteria because high yield, fewer steps, solvent substitution, or renewable feedstock can each conceal burdens elsewhere in the process [27]. The framework therefore requires transparent assumptions, relevant material and hazard indicators, uncertainty reporting, and comparison of realistic alternatives. These requirements are proposed decision controls rather than validated sustainability metrics or certification criteria.

The final stages comprise lifecycle sustainability review, expert and stakeholder review, feedback updating, and explicit claim boundaries. Lifecycle review considers sourcing, cultivation or fermentation, extraction, synthesis, utilities, purification, packaging, transport, use, and end-of-life processes where they are relevant to the proposed product system. Pharmaceutical life-cycle scholarship shows that conclusions depend strongly on the functional unit, system boundaries, inventory quality, toxicity treatment, manufacturing data, and the availability of comparable alternatives [28]. Expert and stakeholder review should integrate ecological, community, chemical, pharmacological, manufacturing, and equity perspectives, while feedback from validation, sourcing, process development, and engagement should be permitted to revise or reverse earlier decisions. The framework supports responsible planning but does not establish legal compliance, ethical certification, environmental validation, clinical effectiveness, regulatory acceptance, or commercialization readiness. Table 2 presents the proposed sustainability framework for natural product drug discovery.

Table 2. Proposed Sustainability Framework for Natural Product Drug Discovery: Biodiversity Protection, Ethical Sourcing, AI-Assisted Prioritization, Dereplication, Extraction Minimization, Green Chemistry, Safety Evaluation, Developability, Therapeutic Innovation, Lifecycle Review, Stakeholder Oversight, Feedback, and Claim Boundaries

Figure 2 presents the proposed sustainability framework for natural product drug discovery, integrating biodiversity protection, ethical sourcing, AI screening, green chemistry, therapeutic innovation, stakeholder oversight, and lifecycle feedback.

1. Newman DJ, Cragg GM. Natural products as sources of new drugs over the nearly four decades from 01/1981 to 09/2019. J Nat Prod. 2020;83(3):770-803. doi:10.1021/acs.jnatprod.9b01285
2. Davis CC, Choisy P. Medicinal plants meet modern biodiversity science. Curr Biol. 2024;34(4):R158-73. doi:10.1016/j.cub.2023.12.038
3. Mykhailenko O, Jalil B, McGaw LJ, Echeverría J, Takubessi M, Heinrich M. Climate change and the sustainable use of medicinal plants: A call for "new" research strategies. Front Pharmacol. 2025;15:1496792. doi:10.3389/fphar.2024.1496792
4. Newman DJ. Natural products and drug discovery. Natl Sci Rev. 2022;9(11):nwac206. doi:10.1093/nsr/nwac206
5. Howes MJR, Quave CL, Collemare J, Tatsis EC, Twilley D, Lulekal E, et al. Molecules from nature: reconciling biodiversity conservation and global healthcare imperatives for sustainable use of medicinal plants and fungi. Plants People Planet. 2020;2(5):463-81. doi:10.1002/ppp3.10138
6. Theodoridis S, Drakou EG, Hickler T, Thines M, Nogués-Bravo D. Evaluating natural medicinal resources and their exposure to global change. Lancet Planet Health. 2023;7(2):e155-63. doi:10.1016/S2542-5196(22)00317-5
7. Flowers I, Lovett JC, Hassall C. Conservation status of species used in the UK herbal medicine industry. Plants People Planet. 2026;1-10. doi:10.1002/ppp3.70154
8. Asigbaase M, Adusu D, Anaba L, Abugre S, Kang-Milung S, Acheamfour SA, et al. Conservation and economic benefits of medicinal plants: insights from forest-fringe communities of Southwestern Ghana. Trees For People. 2023;14:100462. doi:10.1016/j.tfp.2023.100462
9. Giuliani A, Undurraga JT, Dunkel T, Aung SM. Access and benefit sharing and the sustainable trade of biodiversity in Myanmar: the case of Thanakha. Sustainability. 2021;13(22):12372. doi:10.3390/su132212372
10. Lawson C, Englezos E, Rourke M. Access and benefit sharing biological materials for machines: Artificial intelligence, machine learning and deep learning. Plants People Planet. 2025;7(5):1485-97. doi:10.1002/ppp3.70007
11. Ebert AW, Engels JMM, Schafleitner R, van Hintum TJL, Mwila G. Critical review of the increasing complexity of access and benefit-sharing policies of genetic resources for genebank curators and plant breeders—a public and private sector perspective. Plants. 2023;12(16):2992. doi:10.3390/plants12162992
12. Alum EU. The role of indigenous knowledge in advancing the therapeutic use of medicinal plants: challenges and opportunities. Plant Signal Behav. 2024;19(1):2439255. doi:10.1080/15592324.2024.2439255
13. Llivisaca-Contreras S, Naranjo-Morán J, Bastidas-Gálvez M, Jaime-Carvajal J, Muenala-Tituaña M, Manzano-Santana P, et al. The herbal pharmacopoeia of Ecuador: A national model for integrating traditional knowledge and biodiscovery. Front Pharmacol. 2025;16:1662980. doi:10.3389/fphar.2025.1662980
14. Sorokina M, Steinbeck C. Review on natural products databases: where to find data in 2020. J Cheminform. 2020;12:20. doi:10.1186/s13321-020-00424-9
15. Mohimani H, Gurevich A, Shlemov A, Mikheenko A, Korobeynikov A, Cao L, et al. Dereplication of microbial metabolites through database search of mass spectra. Nat Commun. 2018;9:4035. doi:10.1038/s41467-018-06082-8
16. Saldívar-González FI, Aldas-Bulos VD, Medina-Franco JL, Plisson F. Natural product drug discovery in the artificial intelligence era. Chem Sci. 2022;13(6):1526-46. doi:10.1039/D1SC04471K
17. Chemat F, Abert-Vian M, Fabiano-Tixier AS, Strube J, Uhlenbrock L, Gunjevic V, et al. Green extraction of natural products: origins, current status, and future challenges. TrAC Trends Anal Chem. 2019;118:248-63. doi:10.1016/j.trac.2019.05.037
18. Wang J, Nielsen J, Liu Z. Synthetic biology advanced natural product discovery. Metabolites. 2021;11(11):785. doi:10.3390/metabo11110785
19. Aware CB, Patil DN, Suryawanshi SS, Mali PR, Rane MR, Gurav RG, et al. Natural bioactive products as promising therapeutics: A review of natural product-based drug development. S Afr J Bot. 2022;151:512-28. doi:10.1016/j.sajb.2022.05.028
20. Najmi A, Javed SA, Al Bratty M, Alhazmi HA. Modern approaches in the discovery and development of plant-based natural products and their analogues as potential therapeutic agents. Molecules. 2022;27(2):349. doi:10.3390/molecules27020349
21. Caesar LK, Montaser R, Keller NP, Kelleher NL. Metabolomics and genomics in natural products research: complementary tools for targeting new chemical entities. Nat Prod Rep. 2021;38(11):2041-65. doi:10.1039/D1NP00036E
22. Pham JV, Yilma MA, Feliz A, Majid MT, Maffetone N, Walker JR, et al. A review of the microbial production of bioactive natural products and biologics. Front Microbiol. 2019;10:1404. doi:10.3389/fmicb.2019.01404
23. Domingo-Fernández D, Gadiya Y, Preto AJ, Krettler CA, Mubeen S, Allen A, et al. Natural products have increased rates of clinical trial success throughout the drug development process. J Nat Prod. 2024;87(7):1844-51. doi:10.1021/acs.jnatprod.4c00581
24. Son SH, Kang J, Shin YJ, Lee CY, Sung BH, Lee JY, et al. Sustainable production of natural products using synthetic biology: ginsenosides. J Ginseng Res. 2024;48(2):140-8. doi:10.1016/j.jgr.2023.12.006
25. Heath RS, Ruscoe RE, Turner NJ. The beauty of biocatalysis: sustainable synthesis of ingredients in cosmetics. Nat Prod Rep. 2022;39(2):335-88. doi:10.1039/D1NP00027F
26. Sheldon RA, Woodley JM. Role of biocatalysis in sustainable chemistry. Chem Rev. 2018;118(2):801-38. doi:10.1021/acs.chemrev.7b00203
27. Sheldon RA. Metrics of green chemistry and sustainability: past, present, and future. ACS Sustain Chem Eng. 2018;6(1):32-48. doi:10.1021/acssuschemeng.7b03505
28. Chen Z, Lian JZ, Zhu H, Zhang J, Zhang Y, Xiang X, et al. Application of life cycle assessment in the pharmaceutical industry: A critical review. J Clean Prod. 2024;459:142550. doi:10.1016/j.jclepro.2024.142550