Bismuth-based nanomaterials with enhanced radiosensitivity for cancer diagnosis and treatment

Abstract

Radiotherapy is one of the most common and effective clinical treatments for tumors, but how to reduce its side effects to achieve better therapeutic outcomes remains a significant challenge. As a heavy metal, bismuth (Bi) is low-cost, safe, and possesses a high X-ray attenuation coefficient, offering new opportunities to overcome these limitations. With recent advances in nanotechnology and nanomedicine, Bi-based nanoradiosensitizers have been extensively explored for enhancing tumor radiosensitization to achieve advanced diagnosis and treatment by taking advantage of their ease of preparation and modification, high stability, low cost, and excellent biocompatibility. However, the use of Bi-based nanoradiosensitizers remains in the early stages of clinical translation. In this review, we summarize the mechanisms of interaction between X-ray and Bi-based nanoradiosensitizers, discuss smart preparation and modification strategies for achieving enhanced radiotherapy sensitization effects, address material safety and biodistribution, and outline recent research advances in radiotherapy-based synergistic diagnosis and treatment. Finally, we will discuss the challenges and research priorities facing Bi-based nanoradiosensitizers to advance their clinical application development.

---

Tianhao Xing

Tianhao Xing received her Bachelor's degree in Packaging Engineering from Zhengzhou University in 2022. Currently, she is pursuing her master's degree under the supervision of Prof. Wanwan Li at the State Key Lab of Metal Matrix Composites, School of Materials Science and Engineering, Shanghai Jiao Tong University, Shanghai, China. Her current research interests focus on the synthesis of Bi-based nanomaterials and their applications in tumor diagnosis and treatment.

Xujiang Yu

Xujiang Yu received his PhD degree in Materials Science and Engineering from Shanghai Jiao Tong University in 2018. Then he worked as a postdoc fellow in Shanghai Jiaotong University from 2018 to 2021. Now he is an associate professor at the School of Materials Science and Engineering at Shanghai Jiao Tong University. His research focuses on the development of novel functional materials and their applications in medical diagnostics and tumor therapy.

Wanwan Li

Wanwan Li obtained his PhD in materials science from Shanghai University of China in 2004; then he joined the School of Materials Science and Engineering of Shanghai Jiao Tong University in 2005, where he was promoted to professor in 2013. From 2012 to 2013, he joined the Laboratory of Molecular Imaging and Nanomedicine at the National Institute of Biomedical Imaging and Bioengineering as a visiting scholar. His research concerns the design and synthesis of functional micro-/nanomaterials and their applications in biomedical diagnosis & therapy.

---

1. Introduction

Cancer poses a severe threat to human health and lifespan, remaining a major challenge in the medical field.¹ At present, radiotherapy (RT) is one of the most widely used clinical cancer treatments,² with about half of cancer patients undergoing it during treatment.³ The efficacy of RT largely depends on the radiosensitivity of cancer cells. However, inadequate X-ray energy deposition at tumor locations limits the therapeutic efficiency of conventional RT, inevitably causing damage to surrounding tissues.^4,5 Given these challenges, radiosensitizers have emerged to enhance tumor cell sensitivity to radiation while minimizing damage to normal cells.⁶ Other therapeutic modalities might be used in conjunction with this strategy to enhance rates of tumor control and treatment substantially.⁷

Bismuth (Bi) is a heavy metal element characterized by good physical stability, low toxicity, and low cost.⁸ In the late 18th century, Bi compounds were discovered to possess antibacterial properties and were authorized for clinical use in treating various gastrointestinal diseases.^9,10 These compounds have been validated for therapeutic efficacy and safety,¹¹ laying the foundation for subsequent research on Bi-based materials in the medical field. With the rapid development of nanotechnology, nanomaterials with high atomic number elements have been developed as radiosensitizers owing to their exceptional radiation deposition capabilities. Bi exhibits abundant reserves in the Earth's crust (approximately 2 times that of gold) and significantly lowers extraction and processing costs compared to precious metals like gold and tantalum, making it suitable for large-scale production and industrial applications.¹² Furthermore, Bi, the non-radioactive element with the highest atomic number (Bi: 83, Au: 79, Ta: 73, Hf: 72, Gd: 64), possesses an exceptionally high X-ray attenuation coefficient.¹³ These inherent advantages endow Bi-based nanomaterials with outstanding capabilities for integrating cancer diagnosis with radiation therapy.^14–16

The leading advantages of Bi-based nanomaterials for cancer diagnostic and radiotherapeutic platforms include the following aspects: (1) Bi, the least toxic heavy metal with minimal bioreactivity and low cost, offers economic and safety advantages for Bi-based nanomaterials in biomedical applications. (2) The inherent properties of Bi (high X-ray attenuation coefficient) enable Bi-based nanomaterials to be applied in computed tomography (CT) and RT. (3) A high specific surface area and convenient synthesis methods provide Bi-based nanomaterials with the capability for customized functionalization and tunable composite structure construction, making them suitable for biomedical applications. (4) Bi-based nanomaterials are potential candidates for photothermal therapy (PTT) and photoacoustic imaging (PAI) due to their unique physicochemical and optoelectronic characteristics, such as configurable band gaps and high near-infrared (NIR) absorption.¹⁷ Although Bi-based nanomaterials hold significant potential in RT, numerous challenges exist in clinical application, such as difficulty in controlling radiation dose, insufficient tumor-targeted delivery efficiency, and the absence of safety assessment standards. The superior therapeutic outcomes achieved through enhanced radiosensitization could potentially reduce treatment doses and minimize side effects on normal organs. With the increasing demand for highly effective cancer therapies, there is also a need to design and precisely engineer nanotherapeutic platforms. Therefore, more clinical data and research findings on the radiosensitizing effects of Bi-based nanomaterials are needed. Although numerous studies on tumor treatment using Bi-based nanomaterials have been summarized,^18–20 new and significant works continue to emerge, further demonstrating their future potential. For instance, Li et al.²¹ have recently revealed the biological application potential of sodium bismuthate containing Bi with variable valence states, making it necessary to revisit and consolidate the relevant research. Moreover, there has been no review focusing on how to obtain Bi-based nanomaterials with enhanced radiosensitivity from an engineering strategy perspective.

In this review, we aim to summarize design strategies for Bi-based nanoradiosensitizers and elaborate on their emerging theranostic implementations in tumor RT (Scheme 1). This review focuses on the need for radiosensitization of Bi-based nanomaterials, progressively refining the construction of cancer diagnostic and radiotherapeutic platforms through fundamental structural design, modification improvements, and feasibility studies. Firstly, we systematically discuss the manufacturing strategy for high-efficiency Bi-based nanoradiosensitizers with superior performance from the perspectives of synthesis and modification. This includes metal doping, heterojunction construction, defect engineering, valence state regulation, size control, and surface modification. The protocols for verifying the radiosensitizing effects and biological safety are also discussed. We also highlight the advantages of combining RT with other established therapies (chemotherapy (CHT), PTT, immunotherapy) or novel approaches (gas therapy, etc.), as well as the importance and prospects of CT-guided tumor RT. Finally, the prospects and challenges of Bi-based nanoradiosensitizers in tumor RT applications are introduced. We hope this review offers new perspectives for the research and development of Bi-based nanoradiosensitizers in the field of tumor RT, thereby accelerating their clinical translation.

Scheme 1 Schematic diagram of the preparation and applications in synergistic and CT-guided radiotherapy of Bi-based nanoradiosensitizers.

2. Mechanism of Bi-based nanomaterials for radiotherapy

RT utilizes high-energy ionizing radiations (such as X-rays, γ-rays, or other electron beams) to inhibit the proliferation of tumor cells, thereby causing cell death (Fig. 1).^22,23 High-energy radiation will break the chemical bond in the molecule and cause direct harm to the cellular DNA or protein after it has been applied to the tumor tissue. Through the photoelectric effect, it can also ionize oxygen and water molecules in the tissue, producing a lot of cytotoxic reactive oxygen species (ROS) that harm cellular DNA and proteins in an indirect manner, leading to the death of cancer cells.²⁴ Localized ionizing radiation also induces immunogenic cell death (ICD).^25,26 This process prompts damaged tumor cells to release damage-associated molecular patterns (DAMPs) and stimulate dendritic cells (DCs) maturation, thereby activating T cells to combat tumor progression and recurrence effectively.

Fig. 1 Mechanism of radiation-induced cell death. Left: DNA damage by ROS. Right: ICD derived from RT.

High-energy radiation with high tissue penetration enables RT to control the progression of various tumors, particularly deep tissue. However, some tumor cells have comparatively higher radiation resistance because of a variety of variables, including the expression of DNA repair enzymes and anti-apoptotic proteins, as well as the particular tumor microenvironment (TME), including pancreatic cancer, glioblastoma, and advanced (stage IV) tumors.^27–29 The TME features characteristics including weak acidity, hypoxia (O₂), overexpression of H₂O₂ and glutathione (GSH), etc., which diminish the therapeutic efficacy of RT.^30,31 Furthermore, while tumor issues are destroyed and killed under X-ray irradiation, surrounding healthy tissue cells are also damaged. The therapeutic efficacy of conventional RT is limited by its negative effects. The strategies for achieving radiation enhancement using nanomaterials under high-energy radiation exposure were proposed decades ago.^32,33 The physical enhancement effect of nanomaterials on X-rays fundamentally stems from three interaction pathways with X-rays (photoelectric effect, Compton scattering, and Rayleigh scattering).³⁴ Among these, materials with high atomic number exhibit strong X-ray attenuation capabilities (Bi: 5.74, Au: 5.16, Pt: 4.99, and Ta: 4.3 cm² kg⁻¹ at 100 keV),¹³ significantly enhancing local energy deposition efficiency. The photoelectric effect is proportional to the third power of the atomic number, which is significant in the interaction of high-energy radiation with materials.³⁵ Compared to other types of nanoparticles (NPs), it has been found that Bi-based NPs are able to exhibit sufficient deposition of radiation energy, larger absorption and scattering effects at tumor sites, thereby achieving a more considerable dose augmentation.³⁶ When exposed to X-ray radiation, Bi, a heavy metal with a high atomic number (Z = 83), exhibits a considerable capacity for photo-electric absorption, producing a lot of secondary electrons with a short range, which will deposit more energy without inducing any chemical or biological effects. In the meantime, these particles interact with surrounding water molecules, generating ROS to disrupt the cellular DNA, thereby enhancing RT.

While lowering the radiation toxicity to healthy tissues, the application of Bi-based nanomaterials can increase the effects of X-ray irradiation in malignant tissues. Furthermore, preclinical studies confirm that Bi and its compounds exhibit lower renal toxicity than other metals and demonstrate higher cytotoxic effects on cancer cells than on normal cells.^37,38 Leveraging the advantages of Bi in medical applications, researchers have developed numerous tumor radiotherapeutic platforms using various Bi-based nanomaterials. We summarized representative research work from the past five years (Table 1). For example, ovalbumin (OVA)-biomineralized Bi₂S₃ NPs loaded on DCs,³⁹ human-serum-albumin-based KBiF₄ nanoclusters,⁴⁰ biodegradable lipid-camouflaged Bi-based nanoflowers,⁴¹etc. Some of them may also enhance tumor cell sensitivity to X-rays by regulating cellular pathways and cell cycles, improving cellular hypoxia, inhibiting tumor angiogenesis, and modulating immune responses. In addition, to achieve Bi-based nanoplatforms with enhanced RT effects, researchers have conducted studies from multiple aspects, including synthesis engineering, surface modification, and biosafety. We will discuss this further in the next section.

Table 1 Synthesis methods for the functionalized Bi-based nanoradiosensitizers along with their synergistic cancer radiotherapy

Materials	Synthesis methods	X-ray dose/Gy	Tumor	Applications	Administration	Ref.
Abbreviations: PEG = polyethylene glycol; PVP = poly(vinylpyrrolidone); HB = a new NIR photosensitizer; PCM = phase change material; aVEGF = anti-vascular endothelial growth factor; BSA = bovine serum albumin; AD-MSCs = adipose-derived mesenchymal stromal cells; OVA = ovalbumin; DC = dendritic cell; 5-ALA = 5-aminolevulinic acid; CUR = curcumin; PDA = polydopamine; DOX = doxorubicin; PAA = polyacrylic acid; PAH = poly(allylamine hydrochloride); BP = black phosphorus; HSA = human serum albumin; RT = radiotherapy; PTT = photothermal therapy; CDT = chemodynamic therapy; IMT = immunotherapy; CHT = chemotherapy; RDT = radiodynamic therapy; SDT = sonodynamic therapy; CT = computed tomography; PAI = photoacoustic imaging; MRI = magnetic resonance imaging; IR = infrared thermal imaging; FL = fluorescence imaging; i.v. = intravenous injection; i.t. = intratumoral injection; i.p. = intraperitoneal injection.
Elemental Bi-based nanomaterials
PtBi/Pt–PEG	Metal doping	6	4T1	RT/PTT/CT/PAI/IR	i.v.	42
BiNPs@SiO2@BamCS/PCM	—	6	4T1	RT/PTT/PAI	i.t.	43
HB@VHMBi–Gd	Metal doping	4	SUNE-1	RT/PDT/MRI/FL	i.v.	44
pFMBi	Metal doping	6	4T1	RT/PTT/CDT/CT/MRI	i.v.	45
DP-BNF@Lat-MPs	—	10	4T1	RT/PTT/CT/IR	Inhalation	41
Bi@PP/miR339	—	6	KYSE30R	RT/CT	i.v.	46
Bi@Au NDs	Heterostructure	6	CT26	RT/IMT	i.v.	14
aVEGF-BiGd	—	10	McA-RH7777	RT/CT/MRI/FL	i.v.	47
Binary Bi-based nanomaterials
Bi2Se3–MnO2@BSA	Metal doping	8	4T1	RT/CT/MRI	i.v.	48
Bi/Bi2O3−x	Defect engineering	6	4T1	RT/PTT/PAI	i.v.	49
Bi2S3@BSA–Au	Heterostructure	4	4T1	RT/CDT/CT	i.v.	50
Bi2S3@BSA–Fe3O4–FA	Heterostructure	4	4T1	RT	i.v.	51
Bi2Te3 nanoplates	Size control	4	HeLa	RT/CT/PAI	i.v.	52
AD-MSCs/Bi2Se3	—	6	A549	RT	i.v.	53
Bi2S3@OVA@DC	—	6	Treg	RT/IMT	i.t.	39
BiOCl/Cu@PVP	Metal doping	6	4T1.2	RT/CDT	i.v.	54
Bi2O3/CS@5-ALA-CUR	—	2	4T1	RT/CHT/CT	i.v.	55
Alg–Bi2S3@BSA	—	4	4T1	RT	i.p.	56
BiF3–Bi2O3−x: Sx–PEG	Heterostructure	4	4T1	RT/SDT/IMT	i.v.	57
Fe3O4@Bi2S3	Metal doping	6	GL261	RT/MRI	i.v.	15
Pt@Bi2Se3–RGD	—	4	4T1	RT/IMT/PAI	i.v.	58
Bi2O3–TiO2@PDA–DOX	Heterostructure	4	4T1	RT/CHT/SDT/CT/PAI	i.v.	59
Ternary Bi-based nanomaterials
BiVO4/Fe3O4@PDA	Metal doping	6	KB	RT/PTT/CT/PAI/MRI	i.t.	60
PEGylated Bi2S3@mBixMnyOz–DOX	Metal doping	8	4T1	RT/CHT/MRI	i.v.	61
BiOCl@PAA	—	6	4T1	RDT(RCT)	i.t.	62
Bi2O2CO3	Defect engineering	4	Panc02	RT	i.t.	63
Bi2O2CO3/20%Yb/0.1%Er–Ce6–PEG	Metal doping	2	Panc02	RT/PDT	i.t.	64
BiF3:10%Yb@BiOI–PEG	Heterostructure	6	4T1	RT/SDT	i.v.	65
PAH–BiOCl@BP	Heterostructure	1	4T1	RT/RDT/PTT	i.v.	66
Cu3BiS3–BP@PEI	Heterostructure	2	4T1	RT/RDT/CDT/PTT	i.t.	67
KBiF4@HSA	—	6	SHZ-88	RT/DECT	i.t.	40
Bi2WO6–BP	Heterostructure	1	4T1	RT/RDT/PTT	i.t.	68
NaBi^VO3–PEG	High valence	4	4T1	RT/CT/IMT	i.v.	21
Bi-based NPs loaded with R848	—	4	CT26	RT/IMT	i.v.	69

---

3. Construction strategies of Bi-based nanoplatforms

Remarkable progress has been achieved in the development of nanotherapeutic radiosensitizers, yet numerous challenges persist in clinical applications. Particularly, affected by issues including poor blood circulation, rapid clearance, and insufficient tumor targeting of nanoscale radiosensitizers, traditional RT often fails to achieve the desired fertility outcomes. In consequence, scientists are exploring various engineering strategies to develop Bi-based NPs with enhanced radiosensitizing effects and excellent biocompatibility, including metal doping, forming heterostructures, defect engineering, valence state regulation, size control, and surface modification. Then, the radiotherapy efficacy of these materials is validated through performance characterization and in vitro/in vivo biological experiments.

3.1. Synthesis of Bi-based nanoradiosensitizers

Numerous methods have been used to synthesize Bi-based nanomaterials for cancer diagnosis and treatment, including hydrothermal/solvothermal synthesis,^49,52 thermal injection,⁴³ ion exchange,⁷⁰ and sol–gel synthesis,⁴⁷ and chemical reduction synthesis.⁷¹ When controlling product morphology and structure (e.g., nanorods, porous materials) is required, hydrothermal/solvothermal synthesis can be selected. By choosing different solvents, the polarity and coordination environment of the material can also be regulated, facilitating surface modification. The hydrothermal method is an ideal approach for synthesizing high-quality single-crystal Bi₂Te₃.⁵² When employing an ethylenediaminetetraacetic acid (EDTA)-assisted hydrothermal synthesis strategy, increasing the EDTA/Bi³⁺ molar ratio transforms the morphology of Bi₂Te₃ from irregular plate-like structures (approximately 150 nm) to smooth disc-like (approximately 30 nm) and uniform plate-like structures. However, the hydrothermal/solvothermal method requires the use of hazardous high-pressure reactors as reaction vessels and demands high reaction temperatures.⁶⁶ In contrast, the chemical reduction method employs more convenient equipment and features simpler operation. For example, Lei et al.⁷² employed sodium borohydride as a reducing agent to rapidly reduce bismuth nitrate under mild heating conditions, preparing ultrafine poly(vinylpyrrolidone) (PVP)-coated Bi nanodots. Bi-based nanomaterials synthesized via different methods and experimental parameters exhibit significant variations in structure and properties. Therefore, appropriate preparation methods can be selected based on specific clinical application requirements to construct Bi-based nanoplatforms with radiosensitizing properties, thereby achieving more efficient tumor RT. By employing engineering strategies such as metal doping, forming heterostructures, defect engineering, valence state regulation, size control, and surface modification, nanostructures can be further engineered to accommodate the application requirements for enhancing radiotherapy sensitization.

3.1.1. Metal doping. The incorporation of metal elements is a common method for modifying nanomaterials. By seeding metal ions or metallic NPs, not only can the physical structure of Bi-based nanomaterials be optimized, but synergistic effects can also be achieved to enhance the performance of the nanoscale system. Current efforts to modify metallic materials with nanomaterials primarily focus on metal doping and the formation of hetero-structures. Metal doping refers to the incorporation of trace metal elements or NPs into a host material, modulating the physical, chemical, and functional properties of the material by altering its atomic arrangement, electronic structure, or defect state. This operation can reduce the band gap and inhibit the mass recombination of photo-generated carriers, expanding applications of electronics, energy, and medical fields.^73–75

Doping metal elements into Bi-based materials not only enhances RT efficacy but also enables the integration of additional therapeutic approaches and imaging modalities in tumor RT. The effects produced by doping different types of metal ions also vary significantly. For instance, catalytically active ions such as Fe²⁺ and Cu²⁺ can convert H₂O₂ into toxic ROS through Fenton or Fenton-like reactions, thereby leading to cancer cell death, a process termed chemodynamic therapy (CDT).⁷⁶ Li et al.⁵⁴ developed a Bi-based Cu²⁺-doped BiOCl as a flower-like nanotherapeutic platform (BCHN) (Fig. 2a and b). PVP and Cu²⁺ are immobilized on the surface/pores of NBOF via electrostatic interactions (Fig. 2c and d). Under alkaline conditions, H₂O₂ can form coordination bonds with Cu²⁺ to achieve loading. This nanoplatform achieved oxidative stress-enhanced CDT catalyzed by Cu²⁺, while simultaneously catalyzing oxygen production from self-supplied H₂O₂ to alleviate tumor hypoxia. Through redox and coordination pathways, Cu²⁺ and BiOCl also interacted with GSH in tumor tissues, respectively, inhibiting the clearance of ROS by GSH, which induced the biodegradation of materials. The nanoplatform enables radiotherapy enhancement via synergistic CDT and TME reprogramming. In addition to single-metal doping, multi-metal co-doping can be employed to construct multifunctional nanoradiosensitizers. Fe²⁺ and Mn²⁺ co-doped Bi NPs (pFMBi NPs) were assembled as multifunctional radiosensitizers through an ultrafast redox process (Fig. 2e and f).⁴⁵ The excellent Fenton reaction-mediated therapeutic capability of pFMBi NPs is attributed to Fe²⁺ and Mn²⁺. It was indicated that pFMBi NPs also exhibit time-dependent production of •OH radicals, capable of sustaining ROS generation over a sufficient duration to kill cancer cells. Additionally, the magnetic resonance imaging (MRI) performance of iron and manganese ions complemented the inherent CT imaging of the high-Z element, yielding higher imaging sensitivity than either imaging method alone.

Fig. 2 (a) Scheme of the preparation of BCHN. (d) Transmission electron microscopy (TEM) image of BCHN NPs. (c) Energy dispersive spectrometer (EDS) maps of Bi, O, Cl, and Cu, and overlay of all elements of BCHN NPs. (d) High-resolution X-ray photoelectron spectroscopy (XPS) spectrum of Cu in BCHN NPs.⁵⁴ Copyright 2021, Elsevier. (e) Scheme of the synthesis of pFMBi NPs. (f) TEM image of pFMBi NPs.⁴⁵ Copyright 2022, American Chemical Society. (g) TEM image of Bi₂O₂CO₃ nanosheets. (h) Energy transfer in the Bi₂O₂CO₃ nanosheets with different Er³⁺ doping ratios (upper for low Er³⁺ doping, below for high Er³⁺ doping). (i) Luminescence spectrum of Bi₂O₂CO₃:20%Yb/0.1%Er excited at 980 nm and UV-visible absorption spectrum of Ce6.⁶⁴ Copyright 2024, Wiley-VCH GmbH.

Lanthanide ions, with unique physicochemical properties, offer significant advantages in the development of nanoreagents and have been extensively studied in biological fields such as tumor therapy,⁷⁷ biosensing,⁷⁸ optical imaging,⁷⁹ and antibacterial applications.⁸⁰ Lanthanide-doped upconversion NPs can transform deep-penetrating NIR light into high-energy light, and excite organic/inorganic photosensitizers to generate ROS through energy transfer.⁸¹ By doping rare-earth ions into the two-dimensional (2D) ultrathin Bi₂O₂CO₃ nanosheet (Fig. 2g) and loading it with the Ce6 photosensitizer, a novel composite upconversion luminescent nanomaterial was successfully developed for combined application in tumor radiosensitization and photodynamic therapy (PDT).⁶⁴ With increasing Er³⁺ concentration, the probability of energy transfer between Yb³⁺ and Er³⁺ also improves, leading to a shift from red-dominated emission to green-dominated emission (Fig. 2h). The upconversion emission spectrum of Bi₂O₂CO₃:20%Yb/0.1%Er, doped with low Er³⁺ concentration, when excited by a 980 nm laser, exhibits a significant spectral overlap with the UV absorption spectrum of Ce6 at 660 nm (Fig. 2i). This indicates its potential as a resonant energy transfer medium for photodynamic synergistic RT. In vitro and in vivo experiments demonstrated that the nanosheets suppress tumor growth with a low-dose X-ray radiation (2 Gy), effectively reducing damage caused by high-dose radiation and NIR light. Other ions possess distinct therapeutic effects. For instance, the Na⁺-induced surge in cellular osmotic can pressure triggers pyroptosis.⁸² It is conceivable that NPs doped with Na⁺ or K⁺ could achieve this effect through the biodegradation-mediated release of ions.

3.1.2. Heterojunction construction. Bi-based nanosemionductors have emerged as a research hotspot in the semiconductor industry for their unique photoelectric properties, which are widely used in opto-electronic devices,⁸³ photocatalysis,⁸⁴ biomedicine,⁸⁵ and other fields. However, single-component nano-semiconductors suffer from limitations such as narrow band gaps and excessively rapid electron–hole recombination rates, which also impair photocatalytic performance.¹⁷ Constructing heterojunctions represents one of the direct and effective methods to suppress electron–hole recombination. A heterojunction is the formation of the interface region created by the close bonding of two or more different materials. Typically formed between semiconductors, it can also result from the combination of a semiconductor with a metal or insulator. The performance limitations of nanomaterials can be overcome by constructing heterojunctions, endowing them with novel optoelectronic and catalytic properties along with enhanced stability.⁸⁶ Since the effectiveness of electron–hole pair separation is most closely correlated with photocatalytic activity, heterojunction fabrication is essential for fostering charge separation. Through an anion exchange method, a BiOI@Bi₂S₃ heterojunction could be formed in situ on the surface of hydrothermally synthesized BiOI (Fig. 3a).⁷⁰ This heterojunction addresses the limitation of excessively rapid electron–hole recombination rates and further enhances the formation of electron–hole pairs, leading to higher photocurrent and photocatalytic performance compared to BiOI. Under X-ray irradiation, BiOI@Bi₂S₃ heterostructure generates high-energy electrons, which react with surrounding H₂O and O₂ to produce ROS (Fig. 3b and c). The Schottky barrier is formed owing to the band bending at a metal/semiconductor interface when metal is in contact with the semiconductor, accelerating charge separation through an internal electric field. Hamed Nosrati et al.⁵⁰ prepared a Bi₂S₃@BSA–Au semiconductor–metal heterojunction by in situ growth of gold NPs on the Bi₂S₃@BSA surface. By incorporating auxiliary drugs (methotrexate (MTX) and CUR), a novel multifunctional theranostic system for tumor RT was developed (Bi₂S₃@BSA–Au–BSA–MTX–CUR) (Fig. 3d). Bi₂S₃@BSA, Bi₂S₃@BSA–Au, and Bi₂S₃@BSA–Au–BSA–MTX–CUR prepared using these methods exhibit small, uniform, and spherical sizes (Fig. 3e). In vivo experiments demonstrate that the Au–Bi₂S₃ semiconductor heterojunction can significantly enhance free radicals through the Schottky barrier, thereby boosting its inherent radiosensitizing effect, and can also improve the contrast of the CT image.

Fig. 3 (a) Scheme of the preparation of BSA-coated BiOI@Bi₂S₃ semiconductor heterojunction NPs (SHNPs). (b) Fluorescence spectra of ROS detection by sodium terephthalate under different treatments. (c) The mechanism of X-ray-induced PDT by SHNPs.⁷⁰ Copyright 2017, WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim. (d) Synthesis process of Bi₂S₃@BSA–Au–BSA–MTX–CUR and its tumor ablation mechanism. (e) TEM images and size distribution histograms of of Bi₂S₃@BSA (i), Bi₂S₃@BSA–Au (ii) and Bi₂S₃@BSA–Au–BSA–MTX–CUR (iii).⁵⁰ Copyright 2022, Elsevier.

Under X-ray radiation, Bi-based nanomaterials inevitably undergo corrosion, disrupting their inherent structural characteristics and reducing their unique physicochemical capabilities for radiosensitization. Zhou and co-workers prepared a Bi nanoparticle/graphene oxide heterostructure via an in situ growth approach, functionalized with PVP to form a nanoplatform (PVP–Bi/GO) (Fig. 4a).⁷¹ TEM and high-resolution TEM images reveal highly crystalline BiNPs with diameters of approximately 10 nm uniformly grown on the surface of GO nanosheets (Fig. 4b). Owing to enhanced hole scavenging at the graphene/metal interface, PVP–Bi/GO exhibited superior corrosion resistance during X-ray exposure in contrast to pure Bi NPs, with no significant morphological or microstructural changes observed (Fig. 4c). Analysis of radio-catalytic processes caused by PVP–Bi/GO using reactive red 240 as an effective indicator suggests that this heterostructure could provide more energetic electrons and holes through graphene-assisted charge transfer under X-ray irradiation (Fig. 4d and e). The radio-current enhancement of PVP–Bi/GO was more than seven times greater than that of PVP–Bi, further demonstrating this effect. By combining radiocatalysis with the TME modulation, a significant improvement of cancer RT is realized both in vitro and in vivo.

Fig. 4 (a) Schematic illustration of the mechanism of radiosensitizing effects related to corrosion resistance by PVP–Bi/GO. (b) TEM image (i) and high-resolution TEM image (ii) of PVP–Bi/GO. (c) TEM and XPS of PVP–Bi and PVP–Bi/GO under X-ray irradiation with a single dose of 0 or 18 Gy. (d) Mechanism of X-ray interaction and e_aq⁻ generation induced by PVP–Bi/GO, (i) no interaction, (ii) Rayleigh scattering, (iii) pair production, (iv) photoelectric effect, and (v) Compton scattering. SE: secondary electrons, SE1: photoelectron, SE2: Auger electron, and SE3: Compton electron. (e) Fading phenomenon of reactive red 240 after treatment with PVP–Bi (left) and PVP–Bi/GO (right).⁷¹ Copyright 2020, American Chemical Society.

The integration of specific components with nano-semiconductors to form a composite nanomaterial heterojunction with outstanding physicochemical properties enables the integration of multiple diagnostic methods and therapeutic approaches. Furthermore, the recombination efficiency of electron–hole pairs is also influenced by factors such as the crystallinity, structure, morphology, specific surface area, and valence state regulation of nanosemiconductor crystals.

3.1.3. Defect and Bi valence state regulation. Bi-based nanomaterials, such as BiO_2−x,^87,88 BiOCl,⁸⁹ and Bi₂WO₆,⁹⁰ have attracted a lot of interest because of their remarkable photocatalytic ability attributed to defects. With rapid advancements in fabrication techniques, the optical, electrical, and catalytic capabilities of Bi-based nanomaterials can be modified by adjusting defect types and concentrations,⁹¹ through employing chemical reduction, metal particle deposition, light irradiation, and thermal treatment. The existence of defects, particularly oxygen vacancies, can promote electron–hole separation, reduce the bandgap, and broaden the valence band edge.^92–94 Oxygen vacancies formed on material surfaces can change their electronegativity or produce localized electrons, which encourages redox processes on semiconductor surfaces and the production of ROS. Shi et al.⁶³ successfully synthesized ultrathin 2D Bi₂O₂CO₃ nanosheets with a thickness of 2.8 nm using sodium oleate as a template and surface protectant (Fig. 5a and b). Under X-ray irradiation, enriched surface oxygen vacancies and the separation of electron–hole pairs from the bandgap of Bi nanosheets promote the generation of abundant ¹O₂, enhancing RT efficiency. With the assistance of Bi nanosheets, the tumor growth suppression rate increased from 49.88% to 90.76% after radiation exposure. Notably, Bi₂O₂CO₃ nanosheets undergo gradual degradation after treatment, preventing the accumulation of metal ions within the body.

Fig. 5 (a) Schematic illustration of the crystal structure of Bi₂O₂CO₃ nanosheets (left) and the singlet oxygen generation mechanism upon sensitization with Bi₂O₂CO₃ nanosheets (right). (b) HRTEM image (left) and selected-area electron diffraction pattern (right) of the Bi₂O₂CO₃ nanosheets.⁶³ Copyright 2024, Wiley-VCH GmbH. (c) TEM (left) and HRTEM (right) images of BiO_2−x nanosheets. (d) AFM image of BiO_2−x nanosheets. (e) XPS spectra of O 1s of BiO_2−x and Bi₂O₃.⁹⁸ Copyright 2020, American Chemical Society. (f) Schematic diagram of the template-assisted synthesis of NaBi^VO₃ and modification for NaBi^VO₃–PEG. (g) TEM image of NaBi^VO₃. Insert is a HRTEM image. (h) High-resolution XPS spectra of Bi 4f (left), and O 1s (right) of Bi^IIIO_x and NaBi^VO₃. (i) Schematic illustration showing the mechanism by which NaBi^VO₃–PEG undergoes a redox reaction with H₂O₂ to produce •OOH, along with its influence on the generation of •OH and ¹O₂.²¹ Copyright 2025, Springer Nature.

The ground-state electron configuration [Xe]4f¹⁴5d¹⁰6s²6p³ endows Bi with unique semi-metallic properties and diverse oxidation states (Bi^III/Bi^V).⁹⁵ Bi generally can lose electrons from its 6p³ orbital to form the +3 oxidation state. Due to the stability of the 6s² electron pair, Bi^III is the most common and stable oxidation state,⁹⁶ as in compounds like Bi₂O₃ and BiCl₃. Further loss of two extra electrons from the 6s² orbital yields the +5 oxidation state. Through defect engineering and dimensional design, the electronic structure of Bi-based nanomaterials can be tuned to enhance charge transport. BiO_2−x with mixed valence states (Bi^III and Bi^V) with its surface specifically containing oxygen vacancies that serve as photoinduced electron capture centers, thereby suppressing the recombination of photo-generated carriers.⁹⁷ BiO_2−x bulk material prepared by hydrothermal synthesis was broken up into near-monolayer ultrathin nanosheets by ultrasonication, which were then functionalized using Tween 20 to create multifunctional nanoradiosensitizer (T-BiO_2−x nanosheets) (Fig. 5c and d).⁹⁸ X-ray photoelectron spectroscopy (XPS) was employed to characterize the chemical states and oxygen vacancies within BiO_2−x nanosheets. Compared to the O 1s spectrum of Bi₂O₃, BiO_2−x exhibits higher binding energy peaks, likely attributed to low oxygen coordination accompanied by abundant oxygen vacancies (Fig. 5e). On the one hand, BiO_2−x nanosheets exhibit peroxidase-like activity, efficiently decomposing H₂O₂ into O₂ at tumor sites, alleviating hypoxia-induced RT inefficiency. On the other hand, BiO_2−x nanosheets with semiconductor properties in the presence of oxygen vacancies achieve more efficient photocatalytic activity compared to Bi₂O₃, generating more ROS within tumors after X-ray irradiation. The substantial tumor-suppressive effects of T-BiO_2−x nanosheets following exposure to X-rays are also confirmed by in vivo therapeutic experiments.

Compounds with different oxidation states can precisely achieve therapeutic targets such as antibacterial, anticancer, and anti-inflammatory effects by regulating the redox state, coordination structure, and metabolic pathways of metal ions.^99,100 Photovoltaic properties can be enhanced by controlling the valence state of metallic elements within NPs. Liu et al.⁹⁴ developed Bi₄O₇ nanosheets simultaneously containing high-valent Bi ions and Ovs, forming an intermediate band that narrowed the bandgap, and significantly enhanced the photo-induced carrier response and separation efficiency. It is indicated that inorganic high-valent oxygen salts can generate ROS through various spontaneous reactions, leveraging the inherent redox properties of high-valent elements.^101,102 For example, Na₂S₂O₈ NPs containing high-valent sulfur can undergo gradual degradation to generate Na⁺ and S₂O₈²⁻in situ, which subsequently convert into toxic •SO₄²⁻ and •OH radicals, thereby triggering antitumor immune-therapy.¹⁰² Although Bi can form the +5 oxidation state by losing two more electrons from its 6s² orbital, its stability is significantly lower than that of Bi^III. This is because the ionization energy of the 6s² electron pair is extremely high, and exciting the 6s electron to the empty 6d orbital requires absorbing a large amount of energy.^103,104 This leads to the preparation of most pentavalent Bi-based compounds also being relatively complex. Therefore, our group created a novel method for synthesizing sodium bismuthate. We produced NaBi^VO₃ from Bi^IIIO_x NPs using template-assisted etching, which produced a consistent hollow spherical nanoflower-like shape (Fig. 5f and g).²¹ This flexible oxidizing Bi^V nanoplatform (NaBi^VO₃–PEG), modified with a PEGylated phospholipid outer layer, spontaneously produces a large amount of ROS in the presence of endogenous H₂O₂ and O₂ and without the need for exogenous excitation. As shown in Fig. 5h, high-resolution XPS spectra of Bi 4f reveal the absence of the typical trivalent Bi^III peak, while XPS curve fitting of the O 1s spectrum confirms the presence of corresponding lattice oxygen (O_L) in NaBi^VO₃. The acidic conditions of the TME promote the hydrolysis and structural collapse of NaBi^VO₃–PEG. Concurrently with this process, single-electron transfer mediates the conversion of Bi^V to Bi^III, promoting lattice oxygen migration and the release of sodium counterions, while generating •OH and ¹O₂ (Fig. 5i). In vitro experiments further verified that only tumor cells co- incubated with NaBi^VO₃–PEG exhibited substantial ROS generation leading to DNA damage. Additionally, high-valent Bi forms complexes with GSH, an antioxidant molecule overexpressed in tumors, leading to GSH depletion and achieving a synergistic antitumor effect with RT.¹⁰⁵ The reaction between reductive GSH and NaBi^VO₃–PEG produces the byproduct glutathione disulfide, which only reduces the rate of •OH generation but has no effect on ¹O₂. Regulation of the defect and valence state of Bi can enhance the efficacy of Bi-based nanomaterials in RT, increasing their potential for use in cancer detection and treatment.

By precisely controlling defects (such as oxygen vacancies and lattice distortions) and the valence states of Bi (Bi^III/Bi^V) in Bi-based nanomaterials, their electronic structure and surface chemical activity can be significantly optimized. On one hand, defect sites act as “energy traps” to enhance X-ray absorption and scattering efficiency, promoting photoelectron–hole pair separation and boosting ROS generation capacity, thereby directly intensifying DNA damage in tumor cells. On the other hand, dynamic valence state transitions (e.g., reduction of Bi^V to Bi^III under high GSH conditions in the TME) enable controlled drug release and synergistic PTT/RT effects, while reducing systemic toxicity through valence-state-dependent metabolic pathways. This greatly expands the clinical translation potential of Bi-based nanomaterials.

3.1.4. Size control. The radiosensitizing effect of Bi-based nanomaterials is significantly impacted by their morphology and size, depending on the interactions among the nanomaterials, radiation, and biological systems. Dose enhancement factor (DEF) is used to quantify the radiation enhancement effect of nanomaterials or other sensitizers on tumor regions. Currently, the DEF value can be obtained through theoretical Monte Carlo simulations and calculations based on actual experimental data.

It is indicated that through Monte Carlo simulations, under low-energy radiation (50 keV), sheet-shaped Bi₂O₃ NPs exhibit higher DEF compared to spherical Bi₂O₃ NPs, yielding a superior radiation enhancement effect.¹⁰⁶ This is because, as the longitudinal dimensions of NPs increase, the probability of secondary electrons being reabsorbed within the emitting particles also rises. The majority of electrons in sheet-like NPs can escape from their surface and enter the surrounding environment, which results in more severe radiation damage. Although larger NPs exhibit more effective interactions with X-rays, the diameter of Bi NPs exhibits an inverse relationship with DEF, also mainly ascribed to the self-absorption effect of secondary electrons inside the NPs (Fig. 6a).¹⁰⁷ To date, the majority of Monte Carlo simulations in X-ray radiation enhancement applications have focused on gold NPs, while more systematic research is needed on Bi-based NPs, which possess a higher atomic number. The results from Monte Carlo simulations can be leveraged to design highly efficient nanoradiosensitizers tailored to different therapeutic requirements in the future. It should also be noted that a smaller nanoparticle size may reduce the probability of X-ray interactions, while chemical interactions must also be considered. The aforementioned work on ultrathin 2D Bi₂O₂CO₃ nanosheets compared the ROS-generating capacity of sheets with average lateral dimensions of 200, 400, and 600 nm after irradiation to investigate the size-dependent radiosensitizing effect (Fig. 6b and c).⁶³ Nanosheets with a lateral dimension of 200 nm exhibit greater absorption changes after extended X-ray exposure, attributed to the faster kinetics of small-sized nanosheets in generating high levels of ROS. These results suggest that accurate control of the size and shape of Bi-based NPs can lead to improved radiation efficacy.

Fig. 6 (a) Schematic diagram of the self-absorption effect of NPs of different sizes illuminated by the same incident beam.¹⁰⁷ Copyright 2025, IOP Publishing. (b) TEM image of the Bi₂O₂CO₃ nanosheets. (c) The decrease in efficiency of DPBF upon irradiation with different-sized Bi₂O₂CO₃ nanosheets.⁶³ Copyright 2024, Wiley-VCH GmbH. (d) Schematic diagram of the EDTA-assisted hydrothermal synthesis of BT NPs, along with their size-dependent radiosensitizing effects and in vivo metabolic regulation mechanisms. (e) ESR measurement of ¹O₂ generated by BT NPs under X-ray irradiation using TEMP as a trapping agent. (f) ESR measurement of •OH generated by BT NPs under X-ray irradiation using DMPO as a trapping agent.⁵² Copyright 2022, American Chemical Society.

In addition, nanomaterials with different morphologies and sizes exhibit significantly distinct properties and generate distinct categories of ROS. Song et al.⁵² used a hydrothermal method assisted by EDTA to create polymorphic Bi₂Te₃ nanoplates (BT NPs). BT NPs with different sizes and shapes were produced by merely changing the molar ratio of EDTA/Bi³⁺. Research on radiotherapy sensitization mechanisms indicates that larger-sized BT NPs1 readily undergo hole-preferred catalysis under X-ray irradiation, converting OH⁻ into •OH, whereas BT NPs4 with smaller size exhibit accelerated decay kinetics, generating more ¹O₂ to enhance RT efficacy (Fig. 6d and e). By adjusting experimental parameters including acidity, temperature, reactant feed ratio, and reactant concentration, as well as balancing nucleation and growth rates of crystal grains, the shape and size of nanomaterials can be controlled.^108,109 Given the ability to customize experimental conditions, various parameters can be adjusted during the preparation process to obtain the desired nanocomposites for different applications in radiosensitization.

3.2. Surface modification

The excellent biocompatibility and dispersibility of nano-radiosensitizers are crucial for in vivo tumor therapy and detection. However, due to their high specific surface area, NPs tend to agglomerate in physiological environments, hindering therapeutic efficacy within the body. Various polymers are utilized to augment the biocompatibility and dispersibility of Bi-based NPs, thereby prolonging their circulation time in the bloodstream.

Modifying various polymers with Bi-based NPs can enhance their biocompatibility and dispersibility, thereby further prolonging the circulation time of NPs. For instance, polymers such as chitosan,⁵⁵ PDA,¹¹⁰ PVP,^72,111 PEG,¹⁰⁵ and poly(lactic-co-glycolic acid)¹¹² can be utilized, which are modified onto the surface of Bi-based NPs through chemical bonding or physical adsorption, while the newly introduced functional groups expand the capacity for further modification of the NPs. Functionalizing NPs by attaching specific molecules (such as ligands or biomolecules) to their surfaces can endow them with the abilities of targeted delivery, imaging, or responsiveness. It is indicated that folate acid (FA), anti-vascular endothelial growth factor antibody (aVEGF), tumor-homing peptides, and other biomolecules can bind to specific receptors on the surface of cancer cells, which can enhance the endocytosis of Bi-based NPs in tumor sites, substantially reducing side effects in normal tissues. Bi-based NPs with surface-conjugated FA-modified amphiphilic PEG (PFA) molecules can specifically target FA receptors on breast cancer tumor cell surfaces, promoting accumulation at tumor sites and improving the therapeutic effectiveness of RT (Fig. 7a and b).^113–115 aVEGF appears to be the primary mediator of angiogenesis in hepatocellular carcinoma. Active targeting is accomplished through the aVEGF, which specifically binds to aVEGF receptors expressed on tumor endothelial cells.¹¹⁶ Mishra and co-workers successfully employed aVEGF-modified BiGd NPs for tumor-specific CT imaging and enhanced RT.⁴⁷ Using microscopy to track the interaction between NPs and MCA-RH7777 liver cancer cells, it is revealed that only the targeted aVEGF-modified BiGd NPs bound to MCA-RH7777 cells (Fig. 7c). LyP-1 is a homing peptide that binds to the p32 protein overexpressed on breast cancer 4T1 cells, and the small-molecule LyP-1 can be rapidly cleared by the kidneys. Yu et al.¹¹⁷ used oleylamine as a reducing agent to create ultrafine Bi NPs (3.6 nm) (Fig. 7d). After conjugation with tumor-homing LyP-1, these NPs demonstrated high cellular uptake and accumulation within tumor tissues. Furthermore, nearly complete clearance of the NPs was observed in mice after 30 days (Fig. 7e).

Fig. 7 (a) Schematic illustration of the preparation of BiPt–PFA NPs.¹¹³ Copyright 2020, American Chemical Society. (b) Schematic diagram of the structure of Bi@Mn–DTX–PFA.¹¹⁵ Copyright 2021, Wiley-VCH GmbH. (c) Fluorescence microscopy images show the binding of NPs to cells due to the presence of the red fluorescent dye RhB with NPs. Control, nontargeted pR-BiGd NPs, and blocking demonstrate DAPI-stained nuclei (blue) only.⁴⁷ Copyright 2025, American Chemical Society. (d) Scheme of the preparation and of multifunctional Bi NPs and the therapy mechanism. (e) Long-term biodistribution of prepared Bi NPs in the mice's major organs at every stage following dosing.¹¹⁷ Copyright 2017, American Chemical Society.

In addition to the above modification methods, the following methods can also be used: using cell membranes to create a bionic covering that prevents the immune system from eliminating NPs,¹¹⁸ encapsulating mesoporous silica on the surface to load drugs for controlled release,⁴³ and modifying pH-responsive polymers to achieve high expression efficiency of drugs in the TME,⁴⁶etc. Surface modification can be used to further amplify the advantages of Bi-based nanoradiosensitizers and enhance the effects of tumor RT and diagnosis. Compared to other metals, Bi exhibits lower toxicity, making it more suitable for in vivo applications. Surface modification of Bi-based nanoradiosensitizers has expanded their clinical value in tumor RT and diagnosis. On one hand, the modified NPs, characterized by excellent biocompatibility and dispersibility, can be passively enriched in the bloodstream via intravenous injection, thereby improving biodistribution. On the other hand, their targeting capability enables specific tumor therapy, significantly reducing damage to normal tissues and minimizing long-term toxicity.

3.3. Verification of radiosensitizing effects of Bi-based nanomaterials

Verification of radiosensitizing effects is a core component in the development and application of nanoradiosensitizers, achievable through physicochemical mechanisms and in vivo/in vitro experiments. The DEF described in the previous section mainly quantifies the physical dose gain of materials with high atomic number following X-ray irradiation, influenced by nanoparticle size, composition, concentration and radiation energy.¹¹⁹ The sensitization enhancement ratio (SER), on the other hand, examines biological interactions while calculating the ratio of radiation doses needed to produce the same biological impact with and without NPs. This metric is widely used to evaluate the actual therapeutic efficacy of Bi-based nanoradiosensitizers in complex biological environments. SER is defined as the comparison between the radiation dose required to achieve the same level of cell killing between single radiotherapy (D_x) and radiotherapy combined with a radiosensitizer .

This requires plotting cell surviving fraction (SF) curves under both conditions using clonogenic assays and selecting the same survival endpoint for calculation. The fitted formula of SF is as shown in eqn (2):

SER can be calculated using the above two equations and the average lethal dose (D₀) and the quasi-threshold dose (D_q) parameters derived from actual experimental data. A higher SER indicates a stronger ability of the radiosensitizer to enhance the efficacy of RT. The radiosensitizing effect of nanomaterials is tested using the in vivo cell colony formation assay, which is recognized as a “gold standard” for evaluating the efficacy of RT (Table 2). Ma et al.¹²¹ developed a nanoplatform (NP@PVP), comprising bismuth nitrate and a cisplatin prodrug. Through this approach, the calculated SER of free cisplatin and the coordination nanoplatform were 1.78 and 2.29, respectively, providing reliable data supporting the superior tumor-killing efficacy of NP@PVP during X-ray irradiation. Additionally, Xiang et al.¹²² integrated hypoxia-activated prodrugs and CeO₂ nanoenzymes into mesoporous silica-coated Bi₂O₃ (BDS NPs), encapsulated within cell membranes, to create a biomimetic hybrid radiosensitizer (BDSCA@M NPs). The colony formation assay showed that at the dose required for 10% survival rate, the SER of BDSCA@M NPs was approximately 23% higher than that of BDS NPs, demonstrating the most significant anti-cancer effect. This approach not only allows direct comparison of radiosensitizing effects among materials with different components or even different materials, but also enables comparison of the sensitization effects of the same material on different tumor cells. Liu et al.¹²⁰ prepared Bi/Se co-doped nanoradiosensitizers loaded with the anti-angiogenic drug lenvatinib (Len) (Bi/Se–Len NPs). Due to their excellent radiosensitizing properties, Bi/Se–Len NPs demonstrated potent antitumor activity in RT for hepatocellular carcinoma. Particularly, their radio-sensitizing effects varied among different cell lines. It is indicated that the SER for Bi/Se–Len NPs acting on Hep3B cells and SK-Hep1 cells are 2.36 and 3.12, respectively. This significant discrepancy may relate to the ability of Len to target epidermal growth factor receptors, which are highly expressed in SK-Hep1 cells.

Table 2 Bi-based nanoradiosensitizers and their characteristics in radiotherapy

Materials	Size	Concentration	Therapy	Tumor	SER	Ref.
Fe3O4@Bi2S3 NPs	72.3 ± 3.9 nm	100 µg mL^−1	RT	GL261	2.11	15
Bi2O2CO3 nanosheets	200 nm	200 µg mL^−1	RT	Panc02	2.18	63
Bi/Se–Len NPs	120 nm	—	RT	Hep3B	2.36	120
				SK-Hep1	3.12
PtBi/Pt–PEG nanoplates	Lengths of ∼41 nm and widths of ∼5.2 nm	20 ppm	RT	4T1	1.1	42
NP@PVP NPs	190 nm	With a Pt concentration of 2 µg mL^−1	RT	EMT-6	2.29	121
BDSCA@M NPs	60 nm	—	RT	4T1	1.383	122

---

Under X-ray irradiation, Bi-based nanoradiosensitizers deposit greater radiant energy, accelerating the process of water molecule dissociation and generating abundant ROS including •OH, H₂O₂, and O₂⁻. The detection of ROS levels can be achieved by measuring the molecular structure of ROS or their reaction products, thereby indirectly reflecting the radiosensitizing effect.¹⁵

4. Cancer diagnosis and treatment based on multifunctional Bi-based nanoradiosensitizers

Despite significant advances in RT sensitization research, the therapeutic efficacy and prognosis of tumor RT remain significantly constrained by issues such as radiation tolerance associated with monotherapy. This challenge can be addressed through combination therapy, integrating RT with other treatment modalities. The integration of two or more therapies is not merely a simple combination but a synergistic approach that maximizes efficacy while minimizing side effects, achieving a “1 + 1 > 2” effect. Bi-based nanomaterials with high Z elements (Bi) offer advantages such as high stability, straightforward synthesis, and ease of functionalization, which can be engineered into versatile multimodal therapeutic platforms.

4.1. Chemo-radiotherapy

The combination of CHT and RT is one of the most essential strategies in clinical cancer treatment. Most of the chemotherapeutic drugs used clinically are small-molecule compounds, some of which not only serve as chemotherapeutic agents but also possess radiosensitizing effects. At the tissue level, radiation can enhance vascular permeability, allowing more drug transport to tumors.¹²³ At the cellular level, CHT drugs can enhance radiation sensitivity through various pathways, including the inhibition of the DNA repair mechanism, the promotion of oxygen-free radical formation, the suppression of cell cycle progression, causing cells to arrest at a radiation-sensitive stage, and the induction of cell apoptosis.

Currently, frequently used CHT drugs involve DOX, cisplatin, and paclitaxel.¹²⁴ While these CHT drugs can rapidly destroy tumors, most do not specifically target tumor sites. Instead, they distribute throughout the body and are easily cleared, inevitably causing adverse effects on normal tissues. Mesoporous Bi-based NPs or mesoporous silica can improve drug biodistribution when loaded with CHT agents as carriers.^43,110 Through passive or active targeting, these drug-loaded NPs can accumulate at tumor sites, reducing required doses, mitigating tumor cell resistance, and enhancing therapeutic efficacy. Bi₂S₃ nanorods serve as partially sacrificial templates to rapidly form “yolk–shell” structured nanosystems at room temperature (Fig. 8a).⁶¹ The yolk–shell nanostructure featuresadditional pore space between the core and shell, enabling the integration of the CHT drug DOX into a single nanosystem for highly efficient drug loading. Based on research data from concurrent chemo-radiotherapy, specific drug-radiotherapy combinations can effectively enhance tumor radiosensitivity, thereby reducing radiation dose and exposure time.¹²⁵ However, therapeutic safety must be considered when selecting combination drugs. Natural compounds with cancer-fighting properties can amplify the chemical and biological mechanisms that induce cancer cell death. Furthermore, due to their antioxidant and protective effects, they may counteract or mitigate the harmful impacts of metallic NPs.³⁶ The natural polyphenolic compound CUR not only possesses anti-inflammatory, antioxidant, and anticancer properties but also offers radiation protection for normal tissues and radiosensitizing effects at tumor sites.¹²⁶ CUR binds to the HER2 receptor overexpressed in breast cancer to achieve active targeting, especially in the SKBr-3 cell line. Dastgir and co-workers employed chitosan-modified Bi₂O₃ NPs as radiosensitizers while simultaneously serving as a carrier for the chemotherapeutic agent CUR and the radiosensitizer 5-ALA, enabling precise delivery to tumor sites (Fig. 8b).⁵⁵ SKBr-3 cells exhibited significant cytotoxic levels when co-cultured with Bi₂O₃/CS@5-ALA–CUR at various concentrations, while normal cells showed no marked decrease in viability (Fig. 8c–e). In vivo studies in tumor-bearing mice demonstrated that Bi₂O₃ NPs loaded with CUR significantly enhanced the efficacy of chemo-radiotherapy when supplemented with X-rays, compared to single components (Fig. 8f). Building upon this work, this research team further developed CUR-loaded Bi₂O₃ NPs functionalized with polycyclodextrin (PCD) and glucose (Glu) for chemoradiotherapy applications (Fig. 9a).¹²⁷ Real-time PCR was performed to assess the impact on cell death-related gene expression levels. TP53 is a tumor suppressor gene whose upregulation indicates that cells may be under stress or damaged, while BAX and CASP3 play key roles in promoting apoptosis. As shown in Fig. 9b–d, compared to the negative control, the expression levels of TP53, BAX and CASP3 were significantly elevated in the Bi₂O₃@PCD–CUR–Glu NPs + X-ray radiation group. According to the intracellular ROS level measurements using the fluorescent probe DCFH-DA (Fig. 9e–g), the interaction between radiation and NPs with CUR reagent promotes ROS generation, further enhancing radiosensitivity.

Fig. 8 (a) Schematic diagrams of the synthetic procedure for PEGylated Bi₂S₃@mBi_xMn_yO_z–DOX nanosystem (PBmB–DOX).⁶¹ Copyright 2022, Elsevier. (b) The synthesis of Bi₂O₃/CS@5-ALA–CUR NPs and the therapy mechanism in vivo. Cell viability of SKBr-3 cancer cells for 1 day (c) and 2 days (d) under different treatment processes. (e) Cell viability of NIH3t3 healthy cells for 1 day and 2 days under different treatment processes. (f) Tumor volumes change under different treatments.⁵⁵ Copyright 2024, Elsevier.

Fig. 9 (a) The synthetic procedure of Bi₂O₃@PCD–CUR–Glu NPs and improving the X-ray attenuation and radiosensitivity of tumor cells. (b) Detection of TP53, BAX, and CASP3 expression levels in the Bi₂O₃@PCD–CUR–Glu NPs, and Bi₂O₃@PCD–CUR–Glu NPs + X-ray radiation groups, in contrast to the NC group. (c) Green fluorescence intensity of X-ray irradiation in conjunction with intracellular ROS generation under different treatments. (d) Fluorescence images of cancer cells for control, CUR, Bi₂O₃@PCD–Glu, and Bi₂O₃@PCD–CUR–Glu NPs. (e) The flow cytometric histogram profiles of ROS.¹²⁷ Copyright 2024, Elsevier.

4.2. Thermo-radiotherapy

Oxygen molecules are critical for enhancing tumor cell radiosensitivity and inducing DNA damage.^128,129 However, the majority of solid tumor microenvironments exist under hypoxic conditions, which significantly impedes the efficacy of RT. PTT is a therapeutic method that utilizes NIR lasers with a specific wavelength range to irradiate the tumor sites, converting light energy into intense heat to induce tumor cell apoptosis.¹³⁰ The thermal effect induced by PTT enhances nanoparticle uptake efficiency and blood oxygen concentration, promoting increased ROS production and heightening cellular radiosensitivity, thereby achieving more efficient damage to tumor DNA.^131,132 Notably, the limited tissue penetration depth of NIR laser light restricts its therapeutic effectiveness for deep tumor tissues. Therefore, it is of great significance to combine hyperthermia with RT in cancer treatment. Bi-based NPs are highly suitable for combined RT and PTT treatment because of their intrinsic X-ray deposition capacity and photothermal conversion properties.

Bi-based nanomaterials such as elemental Bi,¹¹⁷ Bi₂S₃,¹³³ Bi₂Se₃,¹³⁴ and Cu₃BiS₃⁶⁷ possess narrow bandgaps and high photothermal efficiencies, and have been studied as photothermally active agents in combined RT and PTT treatment. Cheng et al.¹³⁵ firstly demonstrated that Bi₂S₃ nanorods-mediated thermotherapy could suppress HIF-1 expression, leading to reduced hypoxia and enhanced oxygenation, which further amplified the antitumor efficacy of RT. Under X-ray irradiation, the SER of Bi₂S₃ is 1.22, increasing to 1.34 under NIR irradiation, indicating that the thermal effect effectively achieves radiosensitization (Fig. 10a). Poly(ADP-ribose) polymerase (PARP) and Rad51, as DNA repair-associated proteins, participate in the DNA repair process. As shown in Fig. 10b, Bi₂S₃ nanorods-mediated hyperthermia significantly decreased PARP and Rad51 activity, which can substantially disrupt DNA repair and amplify ionizing radiation-induced DNA damage.

Fig. 10 (a) Colony formation survival rate assessment of cells treated with X-ray irradiation alone versus X-ray irradiation combined with Bi₂S₃ nanorods. (b) Representative images of PARP and RAD51 immunofluorescence staining in serial sections of 4T1 cells following physiological saline/Bi₂S₃ + NIR treatment (left). Histograms: rad 51 and PARP quantification from identified pictures (right). P values were calculated by the student's test: *p < 0.05, ***p < 0.001.¹³⁵ Copyright 2017, Ivyspring International Publisher. (c) The mechanism of Bi₂S₃–MoS₂ heterostructure (BMNPs) for imaging-guided synergistic therapy. (d) The photothermal profiles of BMNPs at varying concentrations in aqueous dispersions and DI water following laser irradiation. (e) The linear time data during the cooling period and the negative natural logarithm of the drive temperature. The temperature curve of the BMNPs aqueous solution was recorded and allowed to cool naturally (inserted dataset). (f) Representative photographs of mice in different treatment groups on day 0 and day 28, and comparative analysis of relative tumor volume. (g) The weight of the tumor excised from mice using different treatments. (h) The survival rate of mice using different treatments.¹³⁶ Copyright 2020, Elsevier.

The photothermal conversion performance of Bi-based NPs can be further enhanced through the synthetic engineering mentioned above. Fei Gao and colleagues synthesized heterogeneous NPs composed of Bi₂S₃ and MoS₂ (BMNPs) to achieve improved PTT therapeutic effects by boosting photothermal conversion efficiency,¹³⁶ thereby further enhancing radiosensitizing effects (Fig. 10c). Compared to Bi₂S₃ NPs, BMNPs exhibit superior photothermal conversion capability when exposed to 808 nm laser irradiation, with conversion efficiencies of 35.8% and 28.1%, respectively (Fig. 10d and e). By inducing synergistic effects, bone marrow NPs reduced the quasi-threshold X-ray dose while enhancing sensitivity by 17.9%. This effect was probably attributed to elevated tumor oxygen levels and ROS-induced DNA damage, which further inhibits tumor proliferation. Compared with the control group, the tumor suppression rates achieved by PTT and RT using BMNPs were 41.5% and 32.9% (Fig. 10f and g). However, PTT/RT based on BMNPs completely eliminated tumors within 7 days, with no recurrence observed over the subsequent 20 days. Twenty-eight days after treatment, mice injected with BMNPs and simultaneously exposed to X-rays and laser irradiation achieved a 100% survival rate (Fig. 10h). The NIR-II window spans wavelengths from 1000 to 1700 nm, exhibiting higher spatial resolution, deeper biomatrix penetration, and lower optical absorption and scattering characteristics compared to the NIR-I window (700–900 nm). Bi/Bi₂O_3−x NPs with oxygen vacancy defects were synthesized, which showed strong NIR light absorption and photothermal properties under 1064 nm laser irradiation, enabling enhanced synergistic NIR-II PTT/RT.⁴⁹

Heat shock proteins (HSPs) are rapidly produced by tumor cells in response to high temperatures in order to protect the tumor area from heating-induced damage.¹³⁷ Therefore, inhibiting the function of HSPs can enhance the efficacy of photothermal-enhanced RT. Modifying glycyrrhizic acid, a drug that inhibits HSPs, with Bi-based NPs enhances the thermosensitivity of cancer cells and improves the efficacy of photothermal-enhanced RT.^138,139 Furthermore, thermal injury and inflammatory reactions in adjacent healthy tissues can be prevented by targeting the modification to the tumor site.

4.3. Immuno-radiotherapy

In recent years, immunotherapy has garnered significant attention, offering novel approaches to tumor treatment. Unlike CHT and RT, immunotherapy activates the immune function through the body's inherent immune system to clear tumor cells.¹⁴⁰ However, tumor heterogeneity can lead to suboptimal efficacy of immunotherapy and may even trigger drug resistance and recurrence. As described above, RT can induce ICD in tumor cells. However, the immunosuppressive TME significantly limits the efficacy of immunotherapy.^141,142 As a consequence, the combination of immunotherapy and radiation therapy can overcome some limitations of monotherapy.

In immunotherapy, DCs can precisely present antigens and efficiently activate naive T cells. Nevertheless, to activate immune responses, DCs undergo in vitro engineering modifications, and their mature phenotype may be difficult to maintain in physiological environments, causing inadequate antigen expression and presentation. NPs can serve as co-delivery systems encapsulating antigens and adjuvants to address this issue.¹⁴³ Yu and co-workers utilized antigen OVA as a protein template to prepare Bi₂S₃@OVA via a biomineralization method, loaded on DCs to construct Bi₂S₃@OVA@DC vaccines (Fig. 11a and b).³⁹ Compared to DCs pulsed with free OVA, this vaccine can induce robust immune responses. Bi₂S₃@OVA@DC synergistically promotes tumor-associated DC maturation, which raises the proportion of CD8⁺ T cells and effectively stimulates the production of OVA-specific IgG1/IgG2α antibody. Based on the investigation, the optimal combination regimen was selected: B16F10–OVA melanoma mice received subcutaneous injection of the Bi₂S₃@OVA@DC vaccines, followed by 6 Gy X-ray irradiation 24 hours later. Analysis of the results (Fig. 11c–e) indicates that the Bi₂S₃@OVA@DC vaccines significantly promote the maturation of DCs in lymph nodes, further activate more CD8⁺ T cells, reduce the content of regulatory Treg cells in vivo, and ultimately inhibit melanoma progression when combined with RT. Extensive research indicates that the anti-angiogenic agent Len promotes tumor vascular remodeling to alleviate hypoxia while simultaneously enhancing the recruitment of T cells and CD8⁺ T cells.^144,145 It also potentiates the efficacy of immunotherapies such as anti-PD1 antibodies. As mentioned in our previous discussion, the SER testing of Bi/Se–Len NPs confirmed their excellent radiosensitizing effects.¹²⁰ Owing to the loaded Len, this nanomedicine demonstrated outstanding properties in alleviating hypoxia and immunosuppression. Therefore, tumor growth was significantly suppressed following the injection of agents, indicating excellent radiosensitizing effects. These studies have demonstrated that Bi-based NPs can serve as carriers for immunotherapeutic agents, enabling effective drug delivery, inducing immune responses, and enhancing radiosensitization. However, the mechanisms of interaction between Bi-based NPs and biological agents, as well as potential side effects, require further investigation.

Fig. 11 (a) Synthesis of Bi₂S₃@OVA@DC vaccines. (b) Confocal images showing Bi₂S₃@OVA NPs endocytosis. Cell membranes, nuclei, and NPs were stained with DiO (green), DAPI (blue), and Cypate (red), respectively. (c) The proportion of CD80⁺/CD86⁺ mature DCs. (d) The proportion of CD3⁺/CD8⁺ T cells. (e) The proportion of Treg cells (CD4⁺/CD25⁺/FOXP3⁺).³⁹ Copyright 2022, John Wiley and Sons. (f) Enhanced radiotherapy sensitization and immunogenicity in tumor treatment and metastasis suppression of Bi@Au NDs. (g) Normalized HMGB1 expression in CT26 cells (upper) and detection of ATP secretion by the luciferin-based ATP assay kit (below) after different treatments. (h) Flow cytometry analysis of DC maturation in the lymph nodes of mice. NS: no significance, *p < 0.05, **p < 0.01, ***p < 0.001.¹⁴ Copyright 2025, American Chemical Society.

As an immunostimulatory strategy, radiotherapy can induce ICD in tumor, activating the immune system.¹⁴⁶ Without combining with immunotherapeutic agents, Bi-based nanomaterials can enhance antitumor immune responses and radiotherapy efficacy simultaneously. Shen et al.¹⁴ constructed Bi@Au nanodots (Bi@Au NDs) with a Schottky heterojunction to suppress tumor metastasis by sensitizing RT and activating systemic immunity (Fig. 11f). Under X-ray irradiation, these nanodots effectively promoted the substantial generation of ROS and the rapid depletion of GSH, thereby enhancing radiosensitizing effects. Compared to the control group, the Bi@Au NDs + X-ray group promoted the release of the “find me” signal HMGB1 from CT26 cells (Fig. 11g), facilitating the phagocytosis of apoptotic cells and enhancing immunogenicity. Adenosine triphosphate secretion from CT26 cells, which affects DNA repair, was also significantly increased (Fig. 11h). These data collectively demonstrate the ability of Bi@Au NDs to enhance the efficacy of RT, triggering potent ICD and promoting systemic immunity.

The ability of bacteria to specifically target hypoxic tumors, coupled with their immunogenicity and engineerability, makes them ideal therapeutic agents and drug delivery vehicles for immunotherapy.¹⁴⁷ For instance, bacterial components such as peptidoglycan, flagella, and DNA can be recognized by pattern recognition receptors on immune cells, thereby triggering corresponding immune responses. Thus, bacterial-based immunotherapies are viewed as an innovative treatment approach that activates and modulates the host immune system to recognize and attack cancer cells. Bifidobacterium infantis, a probiotic widely used in digestive disorders, exhibits excellent biocompatibility and can target hypoxic regions within tumors.¹⁴⁸ Xiao et al.⁶⁹ leveraged the superior properties of Bifidobacterium infantis to deliver Bi-based NPs (BPBR) loaded with the immunotherapeutic resiquimod (R848) to tumor sites (Fig. 12a). They utilized the bacterium to attach the nanomaterials to its surface. BPBR biodistribution in tumor-bearing mice demonstrated bacterial accumulation at tumor sites, with higher concentrations than in liver and kidneys (Fig. 12b). Notably, it is indicated that BPBR co-incubated directly with mouse bone marrow-derived DCs showed no observed differences in DC maturation and immune-active factors compared to R848, while both were enhanced following X-ray irradiation (Fig. 12c). Immunofluorescence staining of mouse tumors and cytokine release assays (IFN-β, IL-6, TNF-α) demonstrate that the combination of BPBR and X-rays achieves a potent antitumor immune response (Fig. 12d–g). This work demonstrates the superior radio-immunotherapeutic efficacy of Bi-based NPs conjugated with bacteria, providing valuable insights for designing novel Bi-based nanomaterial-biological hybrid drug delivery systems. Furthermore, the engineered conjugation of the anaerobic bacterium Clostridium butyricum with Bi-based nanomaterials amplified X-ray-induced tumor killing and reprogrammed the immunosuppressive TME, alleviating pancreatic tumor metastasis and achieving satisfactory therapeutic outcomes.¹⁴⁹

Fig. 12 (a) Mechanism of BPBR-enhanced immuno-radiotherapy. (b) Analysis of the distribution of bacteria within the mouse by examining bacterial colonization in the liver, kidneys, other organs, and tumor sites. (c) Representative flow cytometry plots of CD11c⁺/CD80⁺/CD86⁺ cells in BMDCs after treatments with CT26 cell supernatant. (f) Enzyme-linked immunosorbent assay was used to detect cytokine expression, including interferon-β (IFN-β) (d), interleukin-6 (IL-6) (e), and tumor necrosis factor-α (TNF-α) (f). (g) Immunofluorescent expression of CD4⁺ T cells in tumors under different treatment conditions.⁶⁹ Copyright 2025, Royal Society of Chemistry.

With the ability of bacteria to target tumors and induce immunogenicity, precise delivery of Bi-based nanomaterials can be achieved to enhance radiotherapy sensitization, induce immunogenic cell death, and modulate the TME. However, during the preparation of nanomedicines, it is essential to balance bacterial replication activity to prevent overgrowth and avoid systemic toxicity.

4.4. Trimodal radiotherapy

As described above, the combination of both therapies produces stronger antitumor efficacy than either treatment alone, thereby more thoroughly eliminating tumors and suppressing recurrence. To further enhance tumor treatment outcomes, the rapid advancement of nanomaterials in cancer therapy has led to the development of a trimodal synergistic treatment model, including RT/PTT/PDT,⁷⁰ RT/PTT/RDT,^66,133 RT/PTT/QT,¹⁵⁰ RT/CDT/CHT,⁶¹etc.

PDT is a non-invasive photo-activated treatment that utilizes photosensitizers, light of specific wavelengths, and oxygen to selectively destroy diseased tissues.^151,152 The RT/PTT/PDT trimodal therapy can be accomplished by integrating photosensitizers or photothermal agents with Bi-based NPs into a single system. The mechanism of PDT involves activating photosensitizers with excitation light of specific wavelengths, which react with oxygen to generate cytotoxic ROS that induce tumor cell death, which also benefits the RT enhancement. Consequently, the tri-modality therapy demonstrates significantly greater therapeutic efficacy compared to monotherapy or any combination of dual-modality treatments. The preceding section described a nanoradiosensitizer, BSA-coated BiOI@Bi₂S₃ heterojunction NPs (SHNPs),⁷⁰ applied in trimodal therapy (RT/PTT/PDT) (Fig. 13a). The high photothermal conversion efficiency of SHNPs enhances oxygen levels within the TME, ultimately inducing cancer cell death through synergistic killing effects from laser and X-ray radiation. Based on therapeutic outcomes in the in vivo mouse tumor model, the most striking finding was that the group treated with SHNPs combined with X-rays and NIR light achieved near-complete tumor suppression (Fig. 13b and c).

Fig. 13 (a) Schematic diagram of SHNPs enhancing the efficiency of RT/PDT/PTT. (b) Tumor volume growth curves of mice after different treatment processes. (c) The weights of tumors from the mice after different treatment processes.⁷⁰ Copyright 2017, WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim (d) Schematic diagrams of the preparation (A) and the tumor ablation mechanism (B) of the PAH–BiOCl@BP heterojunction. (e) Fluorescence images with Calcein-AM/PI after different treatments, scale bar = 100 nm.⁶⁶ Copyright 2025, American Chemical Society.

The use of ionizing radiation to excite photosensitizers and generate ROS for therapeutic purposes is regarded as a novel treatment strategy termed radiodynamic therapy (RDT). By constructing heterojunction nanostructures, an additional internal electric field is established to promote the separation of electron–hole pairs, thereby enhancing the therapeutic efficacy of RDT.¹³³ Black phosphorus (BP) nanosheets exhibit broad-spectrum light absorption and outstanding photothermal conversion efficiency, having been applied in various tumor therapies. Consequently, the development of multimodal therapeutic platforms combining Bi-based NPs with BP for concurrent RT and thermotherapy represents a current research trend.^66,67,153 Fu et al.⁶⁶ developed a novel multifunctional PAH–BiOCl@BP heterojunction nanoreagent for RT/RDT/PTT (Fig. 13d). After modifying BP, the unique subband structure and strong exciton effect of BP endow this nano-reagent with enhanced radiative catalytic efficiency (Fig. 13e). As indicated in Fig. 13f, the most pronounced red fluorescence was observed beneath tumor cells undergoing synergistic treatment, confirming the superior antitumor properties of the combined RDT, PTT, and RT mediated by the PAH–BiOCl@BP heterojunction. Gas therapy represents a novel green strategy in cancer treatment, leveraging a series of critical gas messenger molecules.¹⁵⁴ As a vasodilatory gas signaling molecule, NO improves hypoxia in tumor cells, thereby enhancing their sensitivity to RT. Researchers designed and synthesized a multifunctional Bi-based nanomedicine by functionalizing Bi NPs with S-nitrosothiol groups.¹⁵⁰ It is indicated that X-ray irradiation induces S–N bond cleavage, simultaneously triggering the release of massive NO to enhance tumor killing efficacy.

The application of Bi-based nanomaterials containing high-Z elements in RT has been extensively studied. Leveraging the excellent photothermal conversion efficiency of Bi-based nanomaterials, numerous studies have combined near-infrared light with X-rays to achieve multimodal therapy that enhances radiation sensitization.^136,155,156 Furthermore, synergistic effects between X-rays and other physical energy fields (such as ultrasound or magnetic fields) can be harnessed to surpass the efficiency of single-modality radiation sensitization. For instance, the radiosensitizing agent BiF₃–Bi₂O_3−x:S_x, composed of BiF₃ and S-doped Bi₂O₃, exhibits a narrower bandgap and improved charge carrier mobility under ultrasonic excitation,⁶⁵ further enhancing RT efficacy. Although combination therapies with RT hold significant potential in tumor treatment, their interaction mechanisms remain complex and warrant further investigation.

4.5. CT-guided multimodal radiotherapy

Despite tremendous efforts in the fight against cancer over the past decades, accurately diagnosing and treating early-stage cancers remains a challenge.^157,158 In recent years, research on nanoplatforms with both diagnostic and therapeutic functions has seen extensive development. A single material system can achieve precise tumor diagnosis and treatment, which effectively reduces drug dosage and minimizes side effects while enhancing tumor treatment efficacy.

CT technology, as one of the commonly used diagnostic methods in clinical practice, is a non-invasive imaging technique that offers advantages such as high resolution, no depth limitations, and diverse image processing capabilities. Bi-based nanomaterials are ideal radiosensitizers for CT imaging and RT.^159,160 As shown in Fig. 14a, Zhou et al.⁴⁶ developed a versatile nano-platform based on Bi₂S₃ nanoblossoms (Bi@PP) for efficient delivery of miR339 and the improvement of radiation resistance. miR339 directly targets USP8 to suppress its expression, thereby inhibiting tumor cell stemness and radioresistance. Additionally, the CT imaging capability provided by Bi enables the observation of the biodistribution of Bi@PP/miR339, allowing for the determination of the optimal timing for X-ray administration and guiding the therapeutic process (Fig. 14b and c). Following combined treatment with Bi@PP/miR339 and X-ray irradiation, tumor growth was nearly completely suppressed (Fig. 14d and e), demonstrating significantly superior efficacy compared to other treatment groups. This nanoplatform exhibits outstanding bioimaging performance, high drug loading efficiency, excellent biocompatibility, and safety.

Fig. 14 (a) Schematic diagram of the preparation of Bi@PP/miR339 and its properties of overcoming stemness and radioresistance. (b) Quantitative analysis of fluorescence intensity in tumor, heart, lung, liver, spleen and kidney tissues. (c) In vivo CT images of the biodistribution after injection with Bi@PP/miR339. Tumor tissue images (d) and tumor weight analysis (e) under different treatments to validate the efficacy of Bi@PP/miR339 combined with radiotherapy. Statistical analysis was performed by a two-tailed, unpaired Student's t test, *p < 0.05, **p < 0.01, ***p < 0.001.⁴⁶ Copyright 2024, American Chemical Society.

With the integration and advancement of disciplines, single-modality image-guided collaborative cancer treatment is gradually giving way to multi-modality image-guided multifunctional cancer treatment approaches. Multi-imaging-guided multifunctional tumor treatment enables more precise tumor localization. CT is primarily used for examining bone structures, while MRI offers superior detail in soft tissues.¹⁶¹ By constructing Bi-based nanomaterials with MRI capability through synthetic engineering, such as ion doping or loading MRI contrast agents, the inherent weakness of CT in detecting soft tissues can be overcome, achieving complementary imaging effects.^45,162,163 Two-dimensional Bi₂Se₃ nanosheets were used to deposit ultrafine Cu_2−xSe NPs for preparing heterojunction nanocomposites (BCP NCs).¹³⁴ The Bi₂Se₃ layer enhances radiation energy deposition at tumor sites, thereby exhibiting superior CT imaging capability (Fig. 15a and b). The MRI capability of BCP NCs is attributed to the effective T₁ MR contrast agent Cu²⁺ (Fig. 15a). As shown in Fig. 15c, BCP NCs exhibit markedly enhanced T₁-weighted MR images with high signal intensity. The potent NIR-II absorption of NCs further supports their utility as agents for infrared imaging. Following irradiation with 1064 nm laser or X-rays, BCP NCs significantly reduced tumor cell viability. Apoptotic cells were detected via membrane protein V/PI double staining. Tumor cells co-incubated with NPs exhibited the highest apoptosis rate (48.1%) following X-ray and laser irradiation, confirming enhanced therapeutic efficacy with combined BCP NCs treatment (Fig. 15d). Concurrently, the results of combined therapy in mouse tumors demonstrated excellent tumor growth inhibition.

Fig. 15 (a) In vitro CT images (left) and T₁-weighted MRI (right) of BCP NCs dispersion in different concentrations. (b) In vivo CT images of mice after injection. (c) In vivo T₁-weighted MRI of mice after injection at different times. (d) Analysis of 4T1 tumor cells apoptosis by flow cytometry under different treatments.¹³⁴ Copyright 2022, Elsevier. (e) In vivo CT, MR, and PA images of mouse tumors pre- and post-injections of BiVO₄/Fe₃O₄@PDA NPs. (f) Relative tumor volume curves of mice. P-values were calculated by one-way ANOVA: *p < 0.05, **p < 0.01.⁶⁰ Copyright 2021, Springer Nature.

PAI is a novel diagnostic technology that utilizes the photoacoustic effect for imaging, currently at a critical stage of transition from research to clinical application.^164,165 In another study, Wang and co-workers constructed BiVO₄/Fe₃O₄@PDA superparticles (SPs) for CT/PA/MR multimodal imaging and RT/PTT synergistic therapy.⁶⁰ The Hounsfield unit values of this nanoplatform are comparable to those of the clinically used CT contrast agent iopamidol, at 28.2136 HU mL mg⁻¹ and 25.6570 HU mL mg⁻¹, respectively. Based on the superparamagnetic properties of Fe₃O₄ NPs, the estimated r₂ value of SPs is 186 mM⁻¹ s⁻¹, outperforming current commercial contrast agents such as Resovist (143 mM⁻¹ s⁻¹) and FeliMag (93 mM⁻¹ s⁻¹). Following intratumoral injection of the SPs, the tumor's photoacoustic signal significantly intensified. Results indicate the immense potential of SPs in multimodal imaging, capable of integrating the strengths of various techniques to provide complementary information for precise diagnosis (Fig. 15e). Therapeutic outcomes in tumor-bearing mice confirmed that synergistic treatment (RT/PTT) enhances efficacy compared to any single modality (Fig. 15f). It is worth noting that while further raising the percentage of BiVO₄ within SPs can significantly enhance their CT imaging performance, this may come at the expense of PAI and MRI capabilities.

Numerous multifunctional therapeutic nanoplatforms for tumor RT guided by multiple imaging modalities have been constructed through methods such as structural composites. However, it is essential to consider the interactions between the physical properties of constituent components, minimize negative effects, maximize the efficacy of each component, and achieve the optimal design of Bi-based nanoradiosensitizers.

5. Pharmacokinetics and toxicity of Bi-based nanomaterials

Given the immense potential and growing demand for Bi-based nanomaterials in biomedical applications, it is imperative to thoroughly evaluate the cytotoxicity of the designed Bi-based nanoradiosensitizers before clinical translation. Studies have confirmed that even under high-dose conditions, Bi-based NPs produce extremely low or negligible toxicity to cells.¹³ However, as the synthesis technology of nanomaterials continues to be updated and iterated, the pharmacokinetics and toxicity of Bi-based NPs prepared through different synthesis pathways to blood, cells, and tissues remain unpredictable. Under the current laboratory conditions, we can conduct pharmacokinetic studies of nanomaterials and evaluate their in vitro and in vivo toxicity through characterization and biological experiments:

(i) Study the blood circulation half-life of nanomaterials through methods such as material tracking, and investigate the degradation behavior of materials in an in vitro physiological environment simulation.

(ii) Measure the cell viability of different types of tumor cells at different concentrations to study the cytotoxicity of nanomaterials.

(iii) Assess the biological half-life and in vivo toxicity of nanomaterials by measuring Bi levels in mouse liver/kidney/spleen tissues and comparing histological staining results.

(iv) Analyze biochemical indicators in the blood to assess in vivo toxicity.

The biodistribution and clearance pathways of NPs within the body are closely correlated with their size, which affects the radiosensitization efficacy of nanomaterials. Due to the unique pathological structure of tumor tissues, such as abnormal tumor neovascularization and impaired lymphatic function, NPs can more readily penetrate into tumor tissues and become restricted from being excluded, thus remaining in the lesion for a long time. This phenomenon is referred to as the enhanced permeability and retention (EPR) effect.¹⁶⁶ Utilizing the EPR effect, nanoradiosensitizers can be passively targeted and enriched at tumor sites, thus enhancing the efficacy of RT as well as minimizing systemic toxicity. Larger NPs (< 200 nm) preferentially accumulate in tumor tissues through EPR effects, but small-sized NPs (< 8 nm) can be readily removed by renal filtration without causing a substantial accumulation in tumors.¹¹⁷ Bi-based nanomaterials exhibit varying half-lives, which are closely related to the hydrodynamic size and surface modification of each nanomaterial. Pharmacokinetic analysis of BT NPs of different sizes revealed that larger BT NPs1 displayed a longer elimination half-life (t_1/2β), suggesting that nanoparticle size significantly influences their kinetic behavior in blood.⁵² It is also important to consider that excessively large sizes may induce biotoxicity.¹⁶⁷ We summarized the pharmacokinetic studies on Bi-based nanoradiosensitizers over the past five years, covering data on particle size, dynamic size, and the in vivo accumulation and distribution of nanomaterials (Table 3). The circulation and biodistribution of BCHN were analyzed by measuring Bi content in different mouse organs.⁵⁴ The relatively long half-lives (t_1/2α = 2.22 h, t_1/2β = 44.46 h) facilitated passive targeting of BCHN to tumor sites. Bi levels in major organs (including heart, liver, spleen, lungs, and kidneys) exhibited time-dependent clearance trends. However, due to the EPR effect and prolonged half-life, BCHN exhibited sustained accumulation in tumors, peaking at 6 hours. In vivo fluorescence imaging revealed that PBmB–DOX accumulated in tumor sites 6 hours post-injection with a maximum efficiency of 4.9%, significantly decreasing after 12 hours but maintaining relatively stable levels over two days.⁶¹ This was hypothesized to result from rapid degradation of the PBmB–DOX shell fraction within the TME. Bi-based nanomaterials can be gradually excreted by mice within 30 days, with no significant damage observed in major organs, indicating low biotoxicity.

Table 3 Pharmacokinetic studies of Bi-based nanoradiosensitizers

Materials	Size	Mean hydrodynamic size	Blood circulation half-life	Tumor	Time of maximal tumor accumulation/h	Ref.
Abbreviations: t1/2α = terminal half-life; t1/2β = elimination half-life.
PtBi/Pt–PEG nanoplates	Lengths of ∼41 nm and widths of ∼5.2 nm	60 nm	—	4T1	48	42
BiOCl/Cu@PVP NPs (BCHN)	—	606.4 nm	t 1/2α = 2.22 h	4T1.2	6	54
			t 1/2β = 44.46 h
Bi2Se3@Cu2−xSe NPs	75 nm	91–142 nm	t 1/2α = 1.49 h	4T1	6	134
			t 1/2β = 8.62 h
PEGylated Bi2S3@mBixMnyOz–DOX NPs (PBmB–DOX)	—	111.1 nm	t 1/2α = 0.22 h	4T1	6	61
			t 1/2β = 10.38 h
Bi2Te3 nanoplates (BT NPs)	BT NPs1 with lengths of ∼150 nm	—	t 1/2β = 42.19 h	HeLa	12	52
	BT NPs4 with lengths of ∼30 nm and widths of ∼10 nm	—	t 1/2β = 19.18 h		8
Cell membrane-coated Fe3O4@Bi2S3 NPs	—	211.6 nm	t 1/2α = 2.73 h for Fe	GL261	24	15
			t 1/2α = 1.99 h for Bi
NaBi^VO3–PEG NPs	120 nm	—	t 1/2α = 5.71 ± 0.53 h	4T1	48	21

---

In addition, the surface modification methods mentioned above can be employed to achieve longer blood retention times for nanomaterials, enable targeted delivery to specific treatment sites, and regulate metabolic pathways, thereby further improving the in vivo biodistribution and biosafety of Bi-based nanomaterials. An ideal radiosensitizer requires both prolonged retention and easy clearance from the body after treatment, and degradable Bi-based nanomaterials may overcome this challenge.^{54,63,113,115} This characteristic provides a feasible reference for the clinical translation of Bi-based nanomaterials.

6. Summary and outlook

In summary, this paper reviews some widely used construction strategies for Bi-based nanomaterials for use in tumor diagnosis and RT. Employing various synthetic engineering and functionalization techniques, such as metal doping, heterojunction construction, size control, defect engineering, valence state regulation, and surface modification, can enhance the performance of Bi-based nanomaterials in radiosensitizing efficiency, tumor targeting, biocompatibility, and imaging capabilities. This allows Bi-based nanomaterials to play a more significant role in tumor RT. We also summarize the assessment strategies of radiosensitizing effects and biosafety currently adopted by most studies for reference. Extensive experiments have demonstrated that modified Bi-based nanomaterials, when combined with multiple therapeutic modalities including CHT, PTT, and immunotherapy, exhibit impressive therapeutic effects under X-ray irradiation.

Despite significant progress in the field of tumor RT using Bi-based nanomaterials, several challenges remain to be addressed before achieving clinical translation:

(i) The properties of Bi-based nanomaterials require precise regulation: physicochemical parameters such as size, shape, and concentration of Bi-based nanomaterials directly influence RT dose and blood circulation, and are also key factors determining their toxicity levels. There is an urgent need to design simpler, more controllable Bi-based nanomaterials for radiosensitization and to systematically investigate the relationship between the properties of Bi-based nanomaterials, experimental parameters, and potential toxicity. Furthermore, the performance of Bi-based nanomaterials can be enhanced by adding various functional groups to their surfaces. Further investigation is needed into the effects of modified compounds on the immunogenicity, pharmacokinetics, and biodistribution of Bi-based nanomaterials.

(ii) Tumor-targeted delivery efficiency also requires improvement: existing Bi-based nanomaterials typically achieve tumor accumulation rates below 10% ID per g via the EPR effect. A significant portion of NPs is cleared by the reticuloendothelial system, limiting radiosensitization efficacy and increasing systemic toxicity risks. More effective strategies are needed to enhance targeted delivery efficiency, such as employing specific ligand modifications or external stimuli to disrupt the tumor vascular barrier.

(iii) Lack of scalable preparation and quality control standards: currently designed and studied Bi-based nanomaterials are synthesized in laboratories, exhibiting issues such as inconsistent size distribution between batches and variations in surface modification stability. These challenges hinder meeting clinical requirements for drug purity, reproducibility, and long-term storage stability. Scalable preparation methods must be developed to reduce organic solvent usage and lower production costs, while establishing comprehensive quality control systems.

(iv) Unclear biosafety assessment criteria: Bi-based nanocomposites involve multi-step synthesis and purification processes, and significant variations in parameters such as synthesis, modification, dosage, and animal models can complicate safety assessments. The optimal radiation dose and strategy for inducing antitumor immunity remain incompletely understood. Currently, most in vivo experiments rely on constructing small animal models for short-term studies, lacking long-term safety assessments. A series of systematic biosafety assessments needs to be conducted in larger animal models (such as dogs, pigs, and monkeys). Therefore, to promote the clinical translation of Bi-based nanomaterials, it is necessary to conduct more in-depth research on their biosafety and establish standardized evaluation procedures.

Therefore, ideal Bi-based nanomaterials must incorporate biodegradability, an appropriate half-life, efficient accumulation within tumors (with high tumor cell uptake rates), scalable production, and optimal radiosensitizing effects to achieve the goal of minimal side effects. Future development of Bi-based nanomaterials in the field of radiation effect enhancement can also focus on the following innovative directions to further overcome existing technical bottlenecks:

(i) Accelerate the development efficiency of Bi-based nanomaterials through artificial intelligence and high-throughput computing. By integrating machine learning models such as random forests and neural networks to construct predictive models for “the structure and radiosensitization efficiency of Bi-based nanomaterials,” the screening cycle for new materials can be significantly shortened.^168,169 Combined with human pharmacokinetic models, this approach predicts the in vivo distribution and radiation response of Bi-based NPs with varying sizes and surface modifications, providing a theoretical foundation for designing personalized RT regimens.¹⁷⁰

(ii) Create Bi-based nanomaterials suitable for different types of radiothration therapy. Clinical RT methods include intensity-modulated radiation therapy, hyperfractionated RT, and stereotactic body radiation therapy etc. These novel approaches may potentially expand the structural design and application scope of Bi-based nanomaterials.

(iii) Integrate the design of Bi-based nanomaterials with clinical techniques. The development of nanotechnology and nanomaterials in RT will continue to evolve toward biological precision, accelerating the clinical translation process. In the future, by integrating patient-specific information such as genetic and metabolic data, treatment plans can be stratified and biologically guided to achieve personalized optimization of parameters including radiation dose, delivery techniques, and fractionation schedules.

Research into the tumor treatment mechanisms of nanomaterials remains at a relatively early stage. Further efforts are needed to investigate the mechanisms of Bi-based nanomaterials in RT and synergistic RT systems. Although deeper advancement is required across design, preparation, and safety assessment to advance Bi-based nanomaterials into clinical trials, we believe scientists and engineers will strive to overcome limitations and ultimately realize the clinical application of these materials for the benefit of humanity.

Author contributions

Tianhao Xing: conceptualization and writing – original draft preparation; Xujiang Yu: writing – review and editing; Wanwan Li: supervision and funding acquisition.

Conflicts of interest

The authors declare no conflict of interest.

Data availability

Data sharing is not applicable to this review article as no new data were created or analyzed in this study.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China, Project No. 82372089, 82272823, 82402435, and Inner Mongolia Autonomous Region-Shanghai Jiao Tong University Science and Technology Cooperation Special Project, Project No. KJXM2023-02-01, Natural Science Foundation of Shanghai (25CL2900900, 24141900402, 23ZR1434600). The authors also thank Zhejiang Orient Gene Biotech Co., Ltd for their support.

References

1. R. L. Siegel, K. D. Miller, H. E. Fuchs and A. Jemal, Ca-Cancer J. Clin., 2021, 71, 7–33.
2. C. Allen, S. Her and D. A. Jaffray, Adv. Drug Delivery Rev., 2017, 109, 1–2.
3. W. Jiang, Z. Zhang, M. Ye, S. Pan, G. Huang, T. Chen and X. Zhu, Nano Today, 2022, 46, 101598.
4. Y. Chen, P. Gao, T. Wu, W. Pan, N. Li and B. Tang, Chem. Commun., 2020, 56, 10621–10630.
5. W. Pan, B. Cui, P. Gao, Y. Ge, N. Li and B. Tang, Chem. Commun., 2020, 56, 547–550.
6. R. Cheng, H. Liu, X. Dong, S. Zhu, R. Zhou, C. Wang, Y. Wang, X. Wang, C. Su and Z. Gu, Chem. Eng. J., 2020, 394, 124872.
7. B. Liu, H. Zhou, L. Tan, K. T. H. Siu and X.-Y. Guan, Signal Transduction Targeted Ther., 2024, 9, 175.
8. A. T. Odularu, J. Chem., 2020, 2020, 1–15.
9. S. L. Gorbach, Gastroenterology, 1990, 99, 863–875.
10. R. J. Loffeld, H. V. Potters, E. Stobberingh, J. A. Flendrig, J. P. van Spreeuwel and J. W. Arends, Gut, 1989, 30, 1206.
11. Y. Yang, R. Ouyang, L. Xu, N. Guo, W. Li, K. Feng, L. Ouyang, Z. Yang, S. Zhou and Y. Miao, J. Coord. Chem., 2015, 68, 379–397.
12. Price of Elements, https://en.wikipedia.org/wiki/Prices_of_chemical_elements, (accessed November, 2025).
13. Y. Cheng and H. Zhang, Chem. – Eur. J., 2018, 24, 17405–17418.
14. C. Shen, X. Niu, J. Zhang, S. Wang, J. Chen, F. Xu, Y. Duan, Y. Zhang, L. Weng, Z. Luo and L. Wang, ACS Nano, 2025, 19, 12382–12398.
15. X. Song, L. Xiao, L. Xu, Y. Jiang, W. Fu, B. Chen, W. Zheng, H. Qian and Y. Yu, ACS Appl. Mater. Interfaces, 2025, 17, 21478–21490.
16. M. Varzandeh, M. Karbasi, S. Labbaf, J. Varshosaz, N. Esmaeil and R. Birang, J. Mater. Chem. B, 2025, 13, 7769–7784.
17. Y. Cheng, X. Kong, Y. Chang, Y. Feng, R. Zheng, X. Wu, K. Xu, X. Gao and H. Zhang, Adv. Mater., 2020, 32, 1908109.
18. X. Yu, X. Liu, K. Yang, X. Chen and W. Li, ACS Nano, 2021, 15, 2038–2067.
19. W. Huang, J. Zhu, M. Wang, L. Hu, Y. Tang, Y. Shu, Z. Xie and H. Zhang, Adv. Funct. Mater., 2021, 31, 2007584.
20. M.-A. Shahbazi, L. Faghfouri, M. P. A. Ferreira, P. Figueiredo, H. Maleki, F. Sefat, J. Hirvonen and H. A. Santos, Chem. Soc. Rev., 2020, 49, 1253–1321.
21. Y. Tang, X. Yu, L. He, M. Tang, W. Yue, R. Chen, J. Zhao, Q. Pan and W. Li, Nat. Commun., 2025, 16, 860.
22. L. Lu, M. Sun, Q. Lu, T. Wu and B. Huang, Nano Energy, 2021, 79, 105437.
23. A. A. Scott, A. Polo, E. Zubizarreta, C. Akoto-Aidoo, C. Edusa, E. Osei-Bonsu, J. Yarney, B. Dwobeng, M. Milosevic and D. Rodin, JAMA Network Open, 2022, 5, e2226319.
24. S. Kc and R. Abolfath, Sci. Rep., 2022, 12, 19853.
25. M. B. Bernstein, S. Krishnan, J. W. Hodge and J. Y. Chang, Nat. Rev. Clin. Oncol., 2016, 13, 516–524.
26. B. Chen, Y. Wang, M. Mu, H. Li, C. Feng, S. Xiao, R. Fan, B. Zou and G. Guo, ACS Nano, 2025, 19, 10147–10161.
27. B. M. Scott, C. Gutiérrez-Vázquez, L. M. Sanmarco, J. A. da Silva Pereira, Z. Li, A. Plasencia, P. Hewson, L. M. Cox, M. O’Brien, S. K. Chen, P. M. Moraes-Vieira, B. S. W. Chang, S. G. Peisajovich and F. J. Quintana, Nat. Med., 2021, 27, 1212–1222.
28. Y. Wu, Y. Song, R. Wang and T. Wang, Mol. Cancer, 2023, 22, 96.
29. D. Wang, J. Cao, Y. Chen, L. Zhang, C. Zhou, L. Huang and Y. Chen, BMC Cancer, 2024, 24, 1497.
30. H. Liu, W. Lin, L. He and T. Chen, Biomaterials, 2020, 226, 119545.
31. A. Menegakis, R. Klompmaker, C. Vennin, A. Arbusà, M. Damen, B. van den Broek, D. Zips, J. van Rheenen, L. Krenning and R. H. Medema, Cells, 2021, 10, 610.
32. F. H. James, N. S. Daniel and M. S. Henry, Phys. Med. Biol., 2004, 49, N309.
33. J. F. Hainfeld, D. N. Slatkin and H. M. Smilowitz, Phys. Med. Biol., 2004, 49, N309–N315.
34. E. Seeram, in X-Ray Imaging Systems for Biomedical Engineering Technology: An Essential Guide, ed. E. Seeram, Springer International Publishing, Cham, 2023, pp. 17–30 DOI:10.1007/978-3-031-46266-5_2.
35. Q. Lyu, R. Neph and K. Sheng, Nat. Biomed. Eng., 2023, 7, 323–334.
36. N. N. T. Sisin and W. N. Rahman, BioNanoScience, 2023, 13, 300–321.
37. Y. Liu, C. Shen, X. Zhang, H. Yu, F. Wang, Y. Wang and L. W. Zhang, Toxicol. Ind. Health, 2018, 34, 188–199.
38. G. Yin, S. Zhao, M. Zhao, J. Xu, X. Ge, J. Wu, Y. Zhou, X. Liu, L. Wei and Q. Xu, Ecotoxicol. Environ. Saf., 2024, 278, 116424.
39. H. Yu, H. Guo, H. Zu, H. Ding, S. Lin, Y. Wang, L. W. Zhang and Y. Wang, Aggregate, 2022, 3, e194.
40. Y. Huang, Z. Lu, H. Liu, N. Yang, Y. Chen, C. Wang, W. Huang, J. Hu, D. Wang and D. Yao, J. Nanobiotechnol., 2025, 23, 451.
41. Q. Wang, J. Liu, D. Chen, S. Miao, J. Wen, C. Liu, S. Xue, Y. Liu, Q. Zhang and Y. Shen, Adv. Healthcare Mater., 2023, 12, 2202622.
42. Y. Liu, X. Li, Y. Shi, Y. Wang, X. Zhao, X. Gong, R. Cai, G. Song, M. Chen and X. Zhang, Nanoscale, 2021, 13, 14245–14253.
43. H. Xiang, Y. Wu, X. Zhu, M. She, Q. An, R. Zhou, P. Xu, F. Zhao, L. Yan and Y. Zhao, J. Am. Chem. Soc., 2021, 143, 11449–11461.
44. Z. Liu, P. Wang, F. Xie, J. Chen, M. Cai, Y. Li, J. Yan, Q. Lin and F. Luo, Adv. Healthcare Mater., 2022, 11, 2102060.
45. W. Xie, J. Ye, Z. Guo, J. Lu, X. Gao, Y. Wei and L. Zhao, ACS Appl. Mater. Interfaces, 2022, 14, 21931–21944.
46. X. Zhou, F. Gao, W. Gao, Q. Wang, X. Li, X. Li, W. Li, J. Liu, H. Zhou, A. Luo, C. Chen and Z. Liu, ACS Nano, 2024, 18, 19232–19246.
47. A. Mishra, E. M. D. San Valentin, A. J. R. Barcena, D. K. M. Bolinas, M. R. Bernardino, G. Canlas, K. A. Ricks, J. A. Damasco and M. P. Melancon, ACS Appl. Mater. Interfaces, 2025, 17, 15097–15108.
48. Y. Yao, P. Li, J. He, D. Wang, J. Hu and X. Yang, ACS Appl. Mater. Interfaces, 2021, 13, 28650–28661.
49. X. Huang, F. Zha, J. Zou, Y. Li, F. Wang and X. Chen, Adv. Funct. Mater., 2022, 32, 2113353.
50. H. Nosrati, E. Attari, F. Abhari, M. Barsbay, M. Ghaffarlou, N. Mousazadeh, R. Vaezi, T. Kavetskyy, H. Rezaeejam, T. J. Webster, B. Johari and H. Danafar, Bioact. Mater., 2022, 7, 74–84.
51. H. Nosrati, M. Ghaffarlou, M. Salehiabar, N. Mousazadeh, F. Abhari, M. Barsbay, Y. N. Ertas, H. Rashidzadeh, A. Mohammadi, L. Nasehi, H. Rezaeejam, S. Davaran, A. Ramazani, J. Conde and H. Danafar, Biomater. Adv., 2022, 140, 213090.
52. Z. Song, T. Liu, H. Lai, X. Meng, L. Yang, J. Su and T. Chen, ACS Nano, 2022, 16, 4379–4396.
53. J. Xiao, L. Zeng, S. Ding, Y. Chen, X. Zhang, X.-W. Bian and G. Tian, Adv. Healthcare Mater., 2022, 11, 2200143.
54. Y. Liu, J. Zhang, J. Du, K. Song, J. Liu, X. Wang, B. Li, R. Ouyang, Y. Miao, Y. Sun and Y. Li, Acta Biomater., 2021, 129, 280–292.
55. M. Dastgir, Y. Ayyami, A. Pourfarshid, M. Ghorbani and T. Mortezazadeh, J. Drug Delivery Sci. Technol., 2024, 92, 105279.
56. B. Colak and Y. N. Ertas, ACS Appl. Mater. Interfaces, 2024, 16, 15718–15729.
57. H. Qian, J. Pan, X. Li, J. Du, L. Gu, Y. Miao, H. Gao, J. Zhang and Y. Li, J. Colloid Interface Sci., 2025, 687, 801–816.
58. H. Tan, S. Fu, L. Shen, Q. Lin, W. Li, Y. Ran, Y. Zhao, L. Tan and Y. Hao, Theranostics, 2025, 15, 2720–2736.
59. Z. Zheng, G. R. Williams, H. Guo, Y. Zheng, M. Xiu, Y. Zhang, H. Zhang, K. Wang, J. Xia, Y. Wang and L.-M. Zhu, Int. J. Nanomed., 2025, 20, 5593–5610.
60. Z. Wang, G. Wang, T. Kang, S. Liu, L. Wang, H. Zou, Y. Chong and Y. Liu, J. Nanobiotechnol., 2021, 19, 90.
61. M. Hou, K. Zhu, H. Hu, S. Zheng, Z. Wu, Y. Ren, B. Wu, L. Qi, D. Wu, Y. Xu, C. Yan and B. Zhao, J. Controlled Release, 2022, 347, 55–67.
62. Y. Liao, D. Wang, S. Zhu, R. Zhou, F. Rahbarizadeh and Z. Gu, Nano Today, 2022, 44, 101510.
63. S. Shi, X. Li, Y. Zhang, H. Huang, J. Liu, J. Zhang, Z. Wang, H. Niu, Y. Zhang and Q. Mei, Small, 2024, 20, 2307974.
64. S. Shi, C. Liao, Y. Liu, J. Liu, J. Liu, Y. Zhang, Y. Zhang and Q. Mei, Adv. Healthcare Mater., 2024, 13, 2401586.
65. L. Zhu, G. Chen, Q. Wang, J. Du, S. Wu, J. Lu, B. Liu, Y. Miao and Y. Li, J. Colloid Interface Sci., 2024, 662, 914–927.
66. H. Fu, S. Ren, J. Cheng, Q. Zhang, Y. Xie, Q. Liang and X. Xiao, ACS Appl. Nano Mater., 2025, 8, 4218–4228.
67. H. Fu, Y. Xie, S. Ren, Q. Zhang, J. Cheng, Q. Liang and X. Xiao, ACS Biomater. Sci. Eng., 2025, 11, 930–941.
68. S. Ren, Q. Zhang, H. Fu, J. Cheng, Y. Xie, Q. Liang and X. Xiao, J. Mater. Chem. B, 2025, 13, 3128–3137.
69. S. Xiao, Y. Wang, S. Pan, M. Mu, B. Chen, H. Li, C. Feng, R. Fan, W. Yu, B. Han, N. Chen and G. Guo, J. Mater. Chem. B, 2025, 13, 10871–10885.
70. Z. Guo, S. Zhu, Y. Yong, X. Zhang, X. Dong, J. Du, J. Xie, Q. Wang, Z. Gu and Y. Zhao, Adv. Mater., 2017, 29, 1704136.
71. R. Zhou, X. Liu, Y. Wu, H. Xiang, J. Cao, Y. Li, W. Yin, Y. Zu, J. Li, R. Liu, F. Zhao, Z. Liu, C. Chen, Z. Gu, L. Yan and Y. Zhao, ACS Nano, 2020, 14, 13016–13029.
72. P. Lei, R. An, P. Zhang, S. Yao, S. Song, L. Dong, X. Xu, K. Du, J. Feng and H. Zhang, Adv. Funct. Mater., 2017, 27, 1702018.
73. S. Ma, Int. J. Mod. Phys. B, 2015, 29, 1550160.
74. Z. Sun, W. Chen, J. Liu, B. Yu, C. Jiang and L. Lu, ACS Appl. Mater. Interfaces, 2021, 13, 35568–35578.
75. S. Wojtyła, K. Śpiewak and T. Baran, J. Inorg. Organomet. Polym. Mater., 2020, 30, 3418–3428.
76. M. Liu, H. Wu, S. Wang, J. Hu and B. Sun, J. Mater. Chem. B, 2021, 9, 9413–9422.
77. J.-L. Wang, X.-Y. Hu, C.-G. Han, S.-Y. Hou, H.-S. Wang and F. Zheng, ACS Nano, 2022, 16, 14827–14837.
78. J. V. G. de Faria, L. M. Pontes, B. B. P. Maia, P. P. A. Magrin, N. N. Silveira, R. A. dos Santos, L. A. Rocha and E. J. Nassar, J. Lumin., 2025, 281, 121210.
79. C. Li, X. Li and X. Liu, ACS Appl. Mater. Interfaces, 2022, 14, 10947–10954.
80. A. Karami, L. Toronjo-Urquiza, S. Kidd, C. J. Sumby and J. Bi, ACS Appl. Mater. Interfaces, 2025, 17, 46850–46860.
81. S. Lamon, H. Yu, Q. Zhang and M. Gu, Light: Sci. Appl., 2024, 13, 252.
82. W. Jiang, L. Yin, H. Chen, A. V. Paschall, L. Zhang, W. Fu, W. Zhang, T. Todd, K. S. Yu, S. Zhou, Z. Zhen, M. Butler, L. Yao, F. Zhang, Y. Shen, Z. Li, A. Yin, H. Yin, X. Wang, F. Y. Avci, X. Yu and J. Xie, Adv. Mater., 2019, 31, 1904058.
83. W. Xiang, E. Cronk, J. Wall, L. Li, K. Zhu, J. J. Berry, R. J. Lad, L. Yu and F. Yan, ACS Appl. Mater. Interfaces, 2024, 16, 44988–44996.
84. Y. Jiang, H. Y. Chen, J. Y. Li, J. F. Liao, H. H. Zhang, X. D. Wang and D. B. Kuang, Adv. Funct. Mater., 2020, 30, 2004293.
85. M. Yuan, S. Liang, L. Yang, F. Li, B. Liu, C. Yang, Z. Yang, Y. Bian, P. A. Ma, Z. Cheng and J. Lin, Adv. Mater., 2023, 35, 2209589.
86. W. Wang, S. Mei, H. Jiang, L. Wang, H. Tang and Q. Liu, Chin. J. Catal., 2023, 55, 137–158.
87. Z. Liu and J. Xu, Appl. Surf. Sci., 2020, 511, 145460.
88. L. Yang, B. Tian, Y. Xie, S. Dong, M. Yang, S. Gai and J. Lin, Adv. Mater., 2023, 35, 2300648.
89. C. Liu, Y. Xiang, X. Dong, Y. Wang, J. Niu and N. Zheng, Sep. Purif. Technol., 2025, 361, 131424.
90. S. Wang, X. Zheng, H. Shi and Y. Chen, Colloids Surf., A, 2025, 719, 136981.
91. J. Huang, T. Lin, L. Lin, G. Ma, Z. Zhang, S. Handschuh-Wang, A. Meng, P. Han and B. He, ACS Appl. Energy Mater., 2024, 7, 10710–10720.
92. L. Wang, D. Lv, F. Dong, X. Wu, N. Cheng, J. Scott, X. Xu, W. Hao and Y. Du, ACS Sustainable Chem. Eng., 2019, 7, 3010–3017.
93. R.-P. Li, S.-M. Yu and Z.-Y. Zhao, Appl. Surf. Sci., 2023, 640, 158392.
94. J. Liu, S. Guo, H. Wu, X. Zhang, J. Li and K. Zhou, J. Mater. Sci. Technol., 2021, 85, 1–10.
95. C. Gu, S. Zhu and Z. Gu, Coord. Chem. Rev., 2024, 517, 215988.
96. M. Mato and J. Cornella, Angew. Chem., Int. Ed., 2024, 63, e202315046.
97. J. Li, X. Wu, W. Pan, G. Zhang and H. Chen, Angew. Chem., Int. Ed., 2017, 57, 491–495.
98. H. Liu, R. Cheng, X. Dong, S. Zhu, R. Zhou, L. Yan, C. Zhang, Q. Wang, Z. Gu and Y. Zhao, Inorg. Chem., 2020, 59, 3482–3493.
99. W. Wang, Y. Xu, Y. Tang and Q. Li, Adv. Mater., 2024, 37, 2416122.
100. A. Islam, B. L. Rodrigues, I. M. Marzano, E. C. Perreira-Maia, D. Dittz, M. T. Paz Lopes, M. Ishfaq, F. Frézard and C. Demicheli, Eur. J. Med. Chem., 2016, 109, 254–267.
101. L. Hou, F. Gong, Z. Han, Y. Wang, Y. Yang, S. Cheng, N. Yang, Z. Liu and L. Cheng, Angew. Chem., Int. Ed., 2022, 61, 202208849.
102. Y. Liu, W. Zhen, Y. Wang, S. Song and H. Zhang, J. Am. Chem. Soc., 2020, 142, 21751–21757.
103. J. Autschbach, J. Chem. Phys., 2012, 136, 150902.
104. P. Pyykko, Chem. Rev., 1988, 88, 563–594.
105. R. Peng, G. Ainiwa, Y. Luo, K. Zhang, R. Zhang, B. Duan, X. Xiang, W. Lv, L. Zhu, J. Wang, X. Liu, Z. Fu, K. Wang and L. Dong, Small, 2025, 21, 2501860.
106. Z. Alyani Nezhad, G. Geraily, F. Hataminia, W. Parwaie, H. Ghanbari and S. Gholami, Med. Phys., 2021, 48, 1417–1426.
107. S. Fernandes Alba, L. Pessoa da Silva, G. Corati Touguinha, M. Salomon Alva Sanchez, T. Alves Pianoschi and H. Trombini, Phys. Med. Biol., 2025, 70, 175007.
108. E. Bárdos, V. Márta, L. Baia, M. Todea, G. Kovács, K. Baán, S. Garg, Z. Pap and K. Hernadi, Appl. Surf. Sci., 2020, 518, 146184.
109. W. Huang, J. Zhu, M. Wang, L. Hu, Y. Tang, Y. Shu, Z. Xie and H. Zhang, Adv. Funct. Mater., 2020, 31, 2007584.
110. J. Guo, B. Chen, M. Zhao, W. Chen, S. Qiu, M.-D. Li and L. Dang, ACS Appl. Bio Mater., 2025, 8, 3525–3532.
111. P. Jia, J. Tu, H. Shen, Y. Jiang, Q. Zhang, W. Xue, M. Liu, J. Liu, Y. Miao, R. Ouyang and S. Zhou, Mater. Today Bio, 2025, 32, 101680.
112. Y. Huang, M. Ma, S. Chen, J. Dai, F. Chen and Z. Wang, RSC Adv., 2014, 4, 26861–26865.
113. J. Zhang, Y. Liu, X. Wang, J. Du, K. Song, B. Li, H. Chang, R. Ouyang, Y. Miao, Y. Sun and Y. Li, ACS Appl. Mater. Interfaces, 2020, 12, 57768–57781.
114. A. DeCarlo, C. Malardier-Jugroot and M. R. Szewczuk, Bioconjugate Chem., 2021, 32, 512–522.
115. J. Liu, J. Zhang, K. Song, J. Du, X. Wang, J. Liu, B. Li, R. Ouyang, Y. Miao, Y. Sun and Y. Li, Small, 2021, 17, 2101015.
116. Y. Yang, S. Q. Li and L. S. Ye, World J. Gastroenterol., 2022, 28, 6034–6044.
117. X. Yu, A. Li, C. Zhao, K. Yang, X. Chen and W. Li, ACS Nano, 2017, 11, 3990–4001.
118. J. Deng, S. Xu, W. Hu, X. Xun, L. Zheng and M. Su, Biomaterials, 2018, 154, 24–33.
119. J. C. L. Chow and S. Jubran, Micromachines, 2023, 14, 1230.
120. J. Liu, J. Chen, H. Liu, K. Zhang, Q. Zeng, S. Yang, Z. Jiang, X. Zhang, T. Chen, D. Li and H. Shan, ACS Appl. Mater. Interfaces, 2021, 13, 42473–42485.
121. Y.-C. Ma, X.-F. Tang, Y.-C. Xu, W. Jiang, Y.-J. Xin, W. Zhao, X. He, L.-G. Lu and M.-X. Zhan, Biomater. Sci., 2021, 9, 3401–3409.
122. H. Xiang, X. Li, H. Wei, X. Zu, Y. Wan, Z. Cheng, T. Mao, Y. Xiao, G. Ren, J. Lu, P. Fan, H. Yuan, M. Xiong, H. Meng, L. Yan and Y. Zhao, Adv. Mater., 2025, 37, 2501077.
123. D. Zhang, J. He and M. Zhou, Adv. Drug Delivery Rev., 2023, 193, 114642.
124. J. Yang, X. Wang, B. Wang, K. Park, K. Wooley and S. Zhang, Adv. Drug Delivery Rev., 2022, 190, 114525.
125. M. L. Mierzwa, M. K. Nyati, M. A. Morgan and T. S. Lawrence, Oncologist, 2010, 15, 372–381.
126. F. Mohammed, F. Rashid-Doubell, S. Taha, S. Cassidy and S. Fredericks, Int. J. Oncol., 2020, 57, 445–455.
127. M. Yektamanesh, Y. Ayyami, M. Ghorbani, M. Dastgir, R. Malekzadeh and T. Mortezazadeh, Int. J. Pharm., 2024, 659, 124264.
128. E. Khakshour, M. A. Shahram, R. Chaman, H. Shoorei, E. Samami, J. S. Welsh, G. Dhawan, S. A. Javadinia and H. Azimian, J. Radiat. Res. Appl. Sci., 2025, 18, 101450.
129. Z. Ma, Y. Zhang, X. Dai, W. Zhang, M. F. Foda, J. Zhang, Y. Zhao and H. Han, Adv. Mater., 2021, 33, 2104504.
130. D. Zhi, T. Yang, J. O'Hagan, S. Zhang and R. F. Donnelly, J. Controlled Release, 2020, 325, 52–71.
131. Y. Cai, T. Chai, W. Nguyen, J. Liu, E. Xiao, X. Ran, Y. Ran, D. Du, W. Chen and X. Chen, Signal Transduction Targeted Ther., 2025, 10, 115.
132. M. Pontico, M. Conte, F. Petronella, V. Frantellizzi, M. S. De Feo, D. Di Luzio, R. Pani, G. De Vincentis and L. De Sio, Pharmaceutics, 2023, 15, 319.
133. Z. Wang, S. Liu, L. Wang, H. Zou, Z. Wang, X. Tang, W. Feng, Y. Chong, Y. Liu, B. Yang and H. Zhang, ACS Appl. Bio Mater., 2020, 3, 5080–5092.
134. X. Chen, M. Qiu, L. Liu, Q. Ji, Z. Xu, Z. Xiong and S. Yang, J. Colloid Interface Sci., 2022, 625, 614–627.
135. X. Cheng, Y. Yong, Y. Dai, X. Song, G. Yang, Y. Pan and C. Ge, Theranostics, 2017, 7, 4087–4098.
136. F. Gao, D. Wang, T. Zhang, A. Ghosal, Z. Guo, Y. Miao, G. Li, X. Liu, J. Lu, J. Yu, H. Fan and L. Zhao, Chem. Eng. J., 2020, 395, 125032.
137. G. Gao, Y. W. Jiang, Y. Guo, H. R. Jia, X. Cheng, Y. Deng, X. W. Yu, Y. X. Zhu, H. Y. Guo, W. Sun, X. Liu, J. Zhao, S. Yang, Z. W. Yu, F. M. S. Raya, G. Liang and F. G. Wu, Adv. Funct. Mater., 2020, 30, 1909391.
138. Y. Zhang, L. Yang, M. Su, J. Wang, R. Wang, H. Li and Z. Zhang, Mater. Des., 2025, 256, 114320.
139. J. Lan, R. Zeng, Z. Li, X. Yang, L. Liu, L. Chen, L. Sun, Y. Shen, T. Zhang and Y. Ding, Adv. Mater., 2024, 36, 2408511.
140. L. Xue, A. S. Thatte, D. Mai, R. M. Haley, N. Gong, X. Han, K. Wang, N. C. Sheppard, C. H. June and M. J. Mitchell, Nat. Rev. Mater., 2023, 9, 100–118.
141. Y. Lu, S.-S. Wang, M.-Y. Li, R. Liu, M.-F. Zhu, L.-M. Yang, F.-Y. Wang, K.-B. Huang and H. Liang, Acta Pharm. Sin. B, 2025, 15, 424–437.
142. D. J. Irvine and E. L. Dane, Nat. Rev. Immunol., 2020, 20, 321–334.
143. H. Liu, K. D. Moynihan, Y. Zheng, G. L. Szeto, A. V. Li, B. Huang, D. S. Van Egeren, C. Park and D. J. Irvine, Nature, 2014, 507, 519–522.
144. D. Zhang, C. Jiang, X. Zheng, Z. Lin, Q. Zhuang, H. Xie, Y. Liang, Y. Xu, L. Cui, X. Liu and Y. Zeng, Small, 2023, 19, 2207786.
145. Y. Jiang, Y. Huang, X. Hu, G. Chen, R. Chen and X. Xu, Clin. Res. Hepatol. Gastroenterol., 2025, 49, 102642.
146. K. Ni, Z. Xu, A. Culbert, T. Luo, N. Guo, K. Yang, E. Pearson, B. Preusser, T. Wu, P. La Riviere, R. R. Weichselbaum, M. T. Spiotto and W. Lin, Nat. Biomed. Eng., 2022, 6, 144–156.
147. X. Huang, J. Pan, F. Xu, B. Shao, Y. Wang, X. Guo and S. Zhou, Adv. Sci., 2021, 8, 2003572.
148. V. N. Dargenio, F. Cristofori, V. F. Brindicci, F. Schettini, C. Dargenio, S. P. Castellaneta, A. Iannone and R. Francavilla, Nutrients, 2024, 16, 3510.
149. W.-Q. Yao, S.-M. Zhang, X.-Q. Hu, L. He, R. Meng, Y.-T. Lin, W.-H. Chen and X.-Z. Zhang, Mater. Today, 2025, 86, 126–145.
150. F. Zhang, S. Liu, N. Zhang, Y. Kuang, W. Li, S. Gai, F. He, A. Gulzar and P. Yang, Nanoscale, 2020, 12, 19293–19307.
151. P. He, M. Jia, L. Yang, H. Zhang, R. Chen, W. Yao, Y. Pan, Q. Fan, W. Hu and W. Huang, Adv. Mater., 2025, 37, 2418978.
152. J. Yu, J. Wu, J. Huang, C. Xu, M. Xu, C. Z. H. Koh, K. Pu and Y. Zhang, Nat. Commun., 2025, 16, 153.
153. H. Fu, L. Feng, Q. Liang and X. Xiao, Adv. Healthcare Mater., 2025, 14, 2404815.
154. Y. He, M. R. Younis, C. Jiang, G. He, J. He, Y. Zhang, H. Liu, P. Huang and J. Lin, Appl. Mater. Today, 2022, 26, 101313.
155. Y. Chen, G. Zhao, S. Wang, Y. He, S. Han, C. Du, S. Li, Z. Fan, C. Wang and J. Wang, Biomater. Sci., 2019, 7, 3450–3459.
156. Y. Song, Y. Wang, Y. Zhu, Y. Cheng, Y. Wang, S. Wang, F. Tan, F. Lian and N. Li, Adv. Healthcare Mater., 2019, 8, e1900250.
157. L. Pu, R. Dhupar and X. Meng, Cancers, 2025, 17, 33.
158. H. T. Tran, S. Heeke, S. Sujit, N. Vokes, J. Zhang, M. Aminu, V. K. Lam, A. Vaporciyan, S. G. Swisher, M. C. B. Godoy, T. Cascone, B. Sepesi, D. L. Gibbons, J. Wu and J. V. Heymach, Ann. Oncol., 2024, 35, 183–189.
159. Z. Fu, K. Li, H. Wang, Y. Li, J. Zhang, J. Zhou, J. Hu, D. Xie and D. Ni, Acta Biomater., 2023, 166, 615–626.
160. K. Ai, Y. Liu, J. Liu, Q. Yuan, Y. He and L. Lu, Adv. Mater., 2011, 23, 4886–4891.
161. Z. Yang, L. Wang, J. Zhang, J. Liu and X. Yu, Nano Today, 2023, 49, 101799.
162. S. Eom, G. Choi, H. Nakamura and J.-H. Choy, Bull. Chem. Soc. Jpn., 2020, 93, 1–12.
163. K. Luo, J. Zhao, C. Jia, Y. Chen, Z. Zhang, J. Zhang, M. Huang and S. Wang, ACS Appl. Mater. Interfaces, 2020, 12, 22650–22660.
164. H.-Q. Gao, J.-T. Zhang, X.-W. Qi, D. Jiao, Y.-N. Hong, K. Shan, X.-L. Kong and D. Ding, Chin. J. Polym. Sci., 2022, 40, 1090–1100.
165. Z. Li, Y. Hu, Z. Miao, H. Xu, C. Li, Y. Zhao, Z. Li, M. Chang, Z. Ma, Y. Sun, F. Besenbacher, P. Huang and M. Yu, Nano Lett., 2018, 18, 6778–6788.
166. J. Fang, W. Islam and H. Maeda, Adv. Drug Delivery Rev., 2020, 157, 142–160.
167. A. U. Khan, M. Qutob, A. Gacem, M. Rafatullah, K. K. Yadav, P. Kumar, J. K. Bhutto, M. Rehman, S. Bansoid, L. B. Eltayeb, N. Malik, M. A. Ali, M. A. Alreshidi and M. W. Alam, Toxicology, 2024, 509, 153985.
168. D. Yan, A. D. Smith and C.-C. Chen, Nat. Comput. Sci., 2023, 3, 572–574.
169. G. Guo, T. L. Saidi, M. W. Terban, M. Valsecchi, S. J. L. Billinge and H. Lipson, Nat. Mater., 2025, 24, 1726–1734.
170. X. Li, M. Zhan, J. Fang, G. Liu, Y. Tang and W. Li, J. Med. Chem., 2025, 68, 16678–16690.

---