Light-assisted 3D bioprinting of tough hydrogels in biomedical applications

Abstract

Light-assisted 3D-bioprinted hydrogels are at the forefront of tissue engineering and biomedical manufacturing due to their precise spatiotemporal controllability and tunable physicochemical properties. However, the mechanical requirements of target organ tissue pose challenges to 3D-bioprinted bio-functional tissue, which must balances biocompatibility, printability, and mechanical strength to replicate the native regenerative microenvironments for functional restoration. This review explores photosensitive polymers with bond chemistry for strengthening hydrogel-based tissue constructs, offering an examination of their potential for facilitating tissue reconstruction through light-assisted 3D bioprinting techniques. We compare the cross-linking chemistry, bond energetics, and resulting mechanics of natural and synthetic hydrogels. Additionally, various light-assisted 3D bioprinting methods for tough hydrogels and engineered living systems are summarized and compared in terms of their practical applications. Importantly, we highlight the critical challenge of enhancing mechanical toughness while balancing the printability, toughness, and biocompatibility of hydrogels via the use of covalent bonds, dynamic covalent bonds, reversible non-covalent interactions, and hybrid bond networks. Furthermore, we discuss emerging applications of light-printed tough hydrogel scaffolds in regenerative medicine and in cartilage, bone, tendon, skull, musculoskeletal, and dental applications. Future prospects and challenges associated with methods of toughing 3D-bioprinted hydrogels are also discussed for guiding future biomedical engineering efforts.

Graphical abstract

Light-assisted 3D bioprinting strategies for hard tissue regeneration: from material design to clinical application.

1. Introduction

Effective tissue regeneration and reconstruction require scaffolds that provide mechanical performance capable of supporting and guiding cell proliferation, lineage commitment, and specialized functions [1,2]. Cells interpret matrix mechanics, such as elasticity, viscoelastic stress relaxation, and anisotropy, by transducing forces into transcriptional programs via integrin, cytoskeleton, and nucleus connections [[1], [2], [3]]. Foundational research has revealed that substrate elasticity solely governs mesenchymal stem cell lineage specification [4], whereas Yes-associated protein (YAP) and transcriptional co-activator with PDZ-binding motif (TAZ) (YAP/TAZ) act as nuclear mechanotransducers that integrate matrix stiffness, cellular morphology, and actomyosin tension to modulate gene expression [3,5,6]. In contrast, mechanical incompatibility between an implant and its target tissue can cause inadequate proliferation, abnormal differentiation, or fibrotic remodeling that hinders tissue repair [1,2]. These principles inform first-order design guidelines for regenerative matrices: (ⅰ) ensure mechanical compatibility of the scaffold with the target tissue across pertinent strain rates; (ⅱ) incorporate programmable viscoelasticity and anisotropy; and (ⅲ) develop time-evolving mechanical properties that maintain appropriate load bearing during degradation and integrate with extra cellular matrix (ECM) remodeling [7]. Emerging dynamic materials, such as stress-relaxing hydrogels and sliding hydrogels that undergo rapid local network reconfiguration, enhance maturation and functional integration, and emphasize that mechanical design should be regarded with equally importance as biochemical signals in any regenerative approach [8,9].

The emergence of three-dimensional (3D) bioprinting has revolutionized the landscape of regenerative medicine, tissue engineering, and personalized therapy. With this technology, complex biological structures can be fabricated with precise spatial deposition of biomaterials and living cells, providing a high degree of control over architecture, cell distribution, and microenvironmental cues within the structure that more closely mimics native tissue [10]. Light-assisted 3D bioprinting encompasses advanced manufacturing techniques that use light to selectively solidify photo-sensitive bioinks to create complex three dimensional structures from a with wide range of biomaterials with high resolution and rapid curing rates [11,12]. Among the diverse range of printable biomaterials, hydrogels have attracted significant attention due to their high water content, tunable physicochemical properties, excellent biocompatibility, and close similarity to the native extracellular matrix (ECM).

However, simultaneously achieving biocompatibility, printability, and mechanical robustness-a triad that is often mutually exclusive-remains a persistent challenge in 3D bioprinting [13]. Ensuring cell viability is a fundamental prerequisite for successful bioprinting [14,15]. With the ability to maintain a cell friendly environment during and after printing, hydrogels have become the most prominent class of bioinks. However, conventional hydrogels, although widely used, frequently suffer from inadequate mechanical strength and poor structural stability, particularly under load bearing conditions or prolonged implantation, such as in tendon or cartilage and bone regeneration. Naturally derived polysaccharide and protein hydrogels, such as gelatin, alginate, chitosan, and hyaluronic acid, are extensively used due to their excellent bioactivity, ECM-like structure, and inherent cell interactive functions. Yet, the mechanical performance of hydrogels remains severely limited. For example, the compressive modulus of gelatin, alginate, and hyaluronic acid is typically less than 10 kPa [16], while that of chitosan rarely exceeds 20 kPa [17]. Many synthetic polymer-based hydrogels, including poly (ethylene glycol) (PEG), polyacrylamide (PAAm), and poly (vinyl alcohol) (PVA), offer good chemical tunability and photopolymerization compatibility. However, the Young's modulus of these materials are only on the order of a few kPa to several hundred kPa, and their fracture energy is generally below 100 J/m², which limits their ability to match the mechanical requirements of native load bearing tissue, such as bone, cartilage, and tendon. As a result, these weak hydrogels often deform or rupture under stress, compromising the hydrogel construct fidelity and impeding tissue regeneration. In recent years, advanced designs have been introduced to bridge this gap, such as double network architectures, reversible dynamic crosslinking, energy dissipating interpenetrating networks, and nanocomposite reinforcement These engineered hydrogels demonstrate superior tensile strength, crack resistance, and elastic recovery while maintaining favorable cytocompatibility and printability, rendering them highly attractive for tissue engineering applications involving bone, cartilage, and tendon regeneration.

Recent reviews on tough hydrogel bioprinting have explored various aspects of this technology, including the following: the overall strategic development of 3D bioprinting in China with a focus on the core functions and evolution of 3D bioprinting and the research to clinic pathway with regulatory and quality safeguards [13]; the rational design strategies of hydrogel-based bioinks and their emerging clinical applications [18]; fundamental photopolymerization mechanisms and light-based 3D printing platforms for biomaterials [19]; the design of bioprintable tough hydrogels emphasizing bioink formulation, printability, and degradation properties and 4D printing trends [20]; and the molecular-and structure-level design principles of tough hydrogels using double networks, slide ring systems, and energy-dissipative architectures in biomedical applications such as soft robotics and bioadhesives [21]. Additionally, a comprehensive classification of 3D printing technologies highlighted how rheological tuning, interpenetrating networks, and post processing methods enabled structurally sophisticated and mechanically resilient hydrogel constructs to be fabricated [22]. Furthermore, several specialized reviews have focused on functional microrobots [23], light-assisted bioprinting for vascular tissue engineering [24], 3D-printed hydrogels for antibacterial applications [25], and the critical evaluation of photoinitiators and phototoxicity in biocompatible photosensitive polymers [26]. However, these reviews predominantly emphasize material development and process optimization, with insufficient exploration of the fundamental chemical origins that govern the interaction among mechanical strength, printability, and biocompatibility, the three properties that frequently impose conflicting constraints on hydrogel-based bioprinting. This review presents a chemistry-based paradigm to elucidate how particular bonding motifs and network combinations influence the relationship between mechanical integrity, printability, and cellular compatibility. We propose a “bond-type property” paradigm based on a comprehensive analysis of covalent, dynamic covalent, and dynamic reversible interactions, and discuss how the chemical characteristics of bonds govern the printability and mechanical stability of hydrogels, thereby providing a cohesive framework for future material designs.

The review provides a comprehensive discussion of recent methods used to toughen 3D-bioprinted photocurable hydrogels, considering the competing demands of biocompatibility, printability, and high mechanical strength (Fig. 1). Types of bond crosslinking (covalent, dynamic covalent, dynamic non-covalent/physical interactions) different hybrid approaches (network combinations, nanocomposites, external treatments), and emerging applications in bone, cartilage, tendon, skull, musculoskeletal, and dental engineering are examined, along with future directions and translational challenges of clinically viable multifunctional bioinks.

Fig. 1.

Overview of light-assisted 3D bioprinting strategies for hard tissue regeneration: from material design to clinical application.

2. Overview of light-assisted 3D bioprinting technologies

Light-based 3D printing technologies, often referred to as vat photopolymerization (VPP) methods, leverage light to selectively polymerize liquid resins, enabling intricate 3D structures to be fabricated with high precision. These techniques are particularly well suited for bioprinting due to their fine spatial control and rapid crosslinking, which are crucial for maintaining cell viability and structural integrity during fabrication [[27], [28], [29]].

Stereolithography (SLA), one of the earliest VPP techniques, uses a UV laser to cure a photopolymerizable resin layer by layer. While offering high resolution, the point-by-point scanning process of SLA can be time consuming for large constructs. Digital light processing (DLP) has emerged as a faster alternative, which uses a digital micromirror device to project an entire 2D image of a layer simultaneously, thereby significantly increasing printing speed with submillimeter and even multi-material patterning when combined with suitable optics and vat control [[28], [29], [30], [31]]. This parallel processing capability renders DLP highly attractive for high-throughput bioprinting, allowing complex geometries with submillimeter fidelity to be rapidly fabricated [32]. Continuous liquid interface production (CLIP) further enhances printing speed by maintaining a continuous liquid interface, preventing oxygen inhibition, and allowing the printed object to be continuously pulled from the resin bath [33]. Process innovations such as vibration-assisted VPP reduce pixelation artifacts and enhance surface smoothness in mask projection systems [34], whereas two-photon polymerization enables VPP to achieve micro- and nanoscale features and has been used to generate 4D micro-architectures pertinent to bio-interfaces [35]. This approach eliminates the need for layer-by-layer separation, dramatically accelerating the printing process. However, the reliance of CLIP on an oxygen permeable membrane and the continuous extraction of the structure can challenge very soft hydrogels that may exhibit structural instability when removed from a liquid suspension.

Beyond the established SLA, DLP, CLIP, and LAE systems, several emerging paradigms are expanding the technological landscape of hydrogel biofabrication, particularly in terms of speed, cell viability, multimaterial integration, and patient-specific tissue repair. Computed axial lithography (CAL), also known as tomographic volumetric printing, enables layer-free fabrication by projecting precomputed 2D light fields into a rotating photosensitive resin or bioink. The target geometry is formed when the accumulated light dose exceeds the polymerization threshold throughout the desired 3D volume. Compared with conventional layer-by-layer VPP, CAL can substantially shorten printing time, reduce interlayer defects, and fabricate complex internal cavities and soft hydrogel architectures, and therefore is an attractive bioprinting technology for volumetric biological constructs and soft devices [36,37]. Multi-wavelength orthogonal photoprinting provides another means of increasing material and functional complexity. By matching different wavelengths with specific photoinitiators, photoresponsive groups, or photochemical pathways, a single printing system can achieve selective polymerization, deprotection, color switching, and local property modulation. This technology is particularly useful for single-vat multimaterial printing, spatially patterned hydrogels, constructs with stiffness gradients, and multi-responsive bioinks [38,39]. In situ bioprinting is emerging as a clinically relevant technology, in which crosslinking is triggered directly within the nozzle, at the nozzle outlet, or at the defect site. This reduces the dependence on highly viscous inks and enables low-viscosity precursors to gel rapidly after deposition or in vivo placement. Therefore, in situ bioprinting is a promising technology for use in minimally invasive repair, irregular defect fitting, and patient-specific tissue reconstruction [[40], [41], [42]]. Acoustic and levitation-assisted biofabrication are not strictly light-assisted printing methods, but they provide useful hybrid or auxiliary methods in hydrogel biofabrication. Acoustic fields can manipulate droplets, particles, and cells in a non-contact manner to induce spatial positioning, aggregation, transport, levitation, and complex interface assembly before subsequent gelation. These techniques may complement light-triggered hydrogel fixation by improving fragile cell handling, multi-droplet organization, and free space patterning [[43], [44], [45]]. Together, these emerging technologies complement conventional VPP and LAE platforms by addressing several persistent challenges in light-assisted bioprinting, including fabrication speed, cell-friendly processing, multimaterial patterning, complex internal architectures, and clinically adaptable deposition.

Light-assisted extrusion (LAE) transcends traditional vat technology, where the latter relies on a reservoir of photosensitive resin and layerwise curing (e.g., stereolithography or DLP), by combining microscale material extrusion with illumination at or near the nozzle that causes freshly deposited filaments to gel almost instantaneously at the air-ink or bath-ink interface. This capability facilitates free space tool paths, overhangs, and architected lattices that conventional extrusion systems have difficulty achieving, thereby supporting in situ and robotic bioprinting workflows [14,15,46]. DLP and LAE can be integrated with digitally addressable lighting and multi-nozzle and -valve feeds for rapid, on-the-fly multi-material changeovers, while maintaining filament fidelity and interfacial registration [30,31,47,48]. Practical limitations affecting these modalities include a limited range of visible light, inadequate performance of existing cytocompatible initiator and co-initiator systems (e.g., eosin Y, Ru/SPS) in terms of initiation efficiency and cytotoxicity, the need to control oxygen inhibition kinetics and light scattering in cell-laden inks, and stringent restrictions on the irradiance dose to prevent phototoxicity while ensuring adequate conversion for sufficient mechanical integrity [[10], [11], [12],20]. From a materials perspective, in addition to biocomposites such as methacrylated chitosan-gelatin blends designed for tissue engineering, VPP- and LAE-compatible hydrogel systems now encompass double network and other dissipation-rich configurations to enhance hydrogel strength without compromising printability [16,17]. Additionally, recent national and sector-wide evaluations examined the use of DLP/LAE chains in regenerative contexts [13,48]. At the device level, these light-assisted methods have produced functional soft system, including thermoresponsive “smart-window” hydrogels [49], ultra-fast photothermal shape-memory microrobots [50], hard/soft switchable hydrogels [51], and organohydrogel strain-sensor platforms [52]. This demonstrates how digitally controlled curing and precise chemistry can simultaneously optimize fidelity, responsiveness, and toughness in printed constructs.

3. Bioresins for light-assisted 3D bioprinting

The successful fabrication of complex, functional biological constructs via 3D bioprinting hinges on a sophisticated interplay between the printing technology and the bioink formulation. Light-assisted bioprinting techniques provide high resolution and speed, while the hydrogel bioink design determines the printability of the final construct and its mechanical and biological characteristics.

3.1. Bioresin selection: natural vs. synthetic hydrogels

The selection of bioink material is paramount in 3D bioprinting, as it dictates the printability and mechanical properties of the printed construct and the biological response of the encapsulated cells and the host tissue upon implantation. Hydrogels, due to their high water content and resemblance to the ECM, are the most commonly used bioinks [53,54]. They can be broadly categorized into naturally derived and synthetic hydrogels, each differing systematically in mechanical performance for reasons rooted in network architecture and cross link chemistry. Natural gels tend to form heterogeneous, labile networks, whereas synthetics enable more defect-lean, programmable networks with predictable mechanics.

3.1.1. Mechanics of naturally derived hydrogels and synthetic hydrogels

Natural polymers, such as gelatin, gelatin methacryloyl (GelMA), hyaluronic acid (HA/HAMA), alginate, and chitosan, frequently gel via physical or ionic interactions at the network level, producing heterogeneous meshes with a high number of network defects. In contrast, synthetic systems can be created from well-defined macromers and step-growth photochemistry (e.g., thiol-ene, thiol-norbornene) to approach optimal network connectivity with fewer defects, higher moduli, and better reproducibility than natural polymers [[55], [56], [57], [58]]. In terms of crosslink energetics, synthetic gels can incorporate dynamic covalent motifs (such as boronate esters and Schiff bases) to achieve strain-stiffening and self-healing without compromising baseline strength, while natural gels often rely on reversible hydrogen bonding, hydrophobic associations, or Ca²⁺ “egg-box” coordination that relax under load [54,59]. The difference between these types of gels is further increased by light-curing kinetics, in which oxygen inhibition during radical photopolymerization may weaken interlayers or undercure surfaces [57,60]. Finally, swelling mechanics coupling is pronounced in highly hydrated natural gels, which show strong poroelastic dissipation and lower instantaneous load-bearing. In contrast, “ideal” or dual-network synthetic designs can decouple hydration and toughness to deliver superior combinations of strength and toughness [61].

3.1.2. Naturally derived hydrogels

Natural materials, such as gelatin, hyaluronic acid, alginate, and chitosan, provide cell recognition motifs that activate receptor-mediated migration, proliferations, and matrix synthesis [62,63]. In 3D bioprinting, chemically modified, cross-linkable derivatives of these biopolymers (e.g., GelMA, HAMA, methacrylated alginate, chitosan derivatives) are preferred because they pair this bioactivity with controllable crosslinking and rheology, supporting adhesion, migration and differentiation [29,64]. For example, hyaluronic acid hydrogels have been explored for their dynamic properties that promote stem cell differentiation and morphogenesis into complex lymphatic networks [65]. Chitosan, a polysaccharide derived from chitin, has also been successfully modified to be photo-crosslinkable for DLP printing, demonstrating excellent biocompatibility and printability for complex 3D hydrogel structures [66]. These materials often provide a biomimetic environment that closely replicates the native tissue microenvironment, which is crucial for successful tissue regeneration.

To endow naturally derived hydrogels with photo-crosslinkable properties, photoreactive groups are commonly introduced onto their polymer backbones through chemical functionalization. Typical methods include methacrylation, acrylation, allylation, norbornene modification, and thiol-ene clickable modifications of amino, hydroxyl, and carboxyl groups [27,67,68]. Methacrylation or methacrylamide modifications of gelatin-based systems enables gelatin methacryloyl and methacrylamide hydrogels to form, such as GelMA, in which gelation behavior, stiffness, degradation, and cellular responses can be tuned by the degree of functionalization and the network architecture [[69], [70], [71], [72]]. In addition, allylated gelatin enables thiol-ene photocrosslinking and has been developed as a platform bioink for multiple 3D biofabrication technologies [73].

Alginate can be functionalized with methacrylate, glycidyl methacrylate, RGD peptides, or norbornene groups to improve light-triggered covalent network formation, shape fidelity, controlled degradation, and cellular interactions [[74], [75], [76], [77], [78]]. This functionalization is particularly important because purely ionically crosslinked alginate hydrogels may exhibit long-term instability due to the loss of ionic crosslinkers under physiological conditions [79]. Hyaluronic acid can also be converted into photo-crosslinkable HA-based networks through methacrylation or glycidyl methacrylate modifications, conferring this ECM-derived polysaccharide with tunable covalent crosslinking while retaining its biological relevance [80]. Similarly, chitosan can be endowed with photoreactivity through methacrylate or glycidyl methacrylate modifications, enabling DLP printing of biocompatible hydrogel constructs [66]. Silk fibroin, although mechanically attractive, generally exhibits very gradual gelation kinetics and poor processability in its native form. Functionalization with photoreactive groups can generate photocurable silk fibroin derivatives with improved gelation speed, printability, and structural stability [81].

Photo-crosslinkable dECM-based hydrogels are another important class of naturally derived bioinks. dECM hydrogels retains tissue-specific structural proteins, glycosaminoglycans, and cell interaction motifs, but unmodified dECM hydrogels usually exhibit weak mechanical stability, slow gelation, and poor shape fidelity. To address these limitations, dECM can be methacrylated or crosslinked through visible light-mediated processes, thereby retaining tissue-specific ECM components while improving gelation speed, structural stability, and printing fidelity [[82], [83], [84], [85]]. Representative examples include kidney-derived ECM bioinks for renal tissue formation [82], cartilage-derived ECM bioinks for auricular cartilage engineering [83], light activated dECM bioinks with enhanced mechanical integrity [84], and graphene oxide-embedded dECM hydrogels for 3D bioprinting applications [85]. Photo-crosslinkable modifications provide a key bridge between the biological advantages of natural polymers and the manufacturing requirements of light-assisted 3D bioprinting. By introducing photoreactive groups into naturally derived hydrogels, faster gelation, improved shape retention, tunable mechanical properties, and better spatial control of the hydrogels can be achieved while preserving their inherent bioactivity.

However, a significant drawback of many naturally derived hydrogels is their intrinsically poor mechanical properties, as these hydrogels are often soft and easily degradable, which limits their application in load bearing tissue [21,64]. Their mechanical properties can also be highly variable depending on the source and purification methods, leading to inconsistencies in printed constructs. Furthermore, the rapid degradation rates of some natural hydrogels can compromise the long-term structural integrity of engineered tissue, particularly in dynamic physiological environments. To overcome these limitations, reinforcement techniques are often necessary to improve the hydrogel biomechanical properties, which can involve chemical modifications or the incorporation of reinforcing agents [64].

3.1.3. Synthetic hydrogels

Synthetic polymers have excellent mechanical properties, are highly adjustable, and are readily available. Synthetic hydrogels, typically derived from polymers such as polyethylene glycol (PEG), poly (lactic acid) (PLA), and poly (caprolactone) (PCL), allow precise control over their chemical composition, mechanical properties, and degradation rates [27,29]. Unlike natural hydrogels, synthetic hydrogels can be engineered with specific functional groups to allow for tunable photo-crosslinking, enabling predictable and reproducible material properties. For example, PEG-based hydrogels can be designed as ideal network hydrogels with dynamic covalent boronate ester crosslinks that exhibit biomimetic strain stiffening and self-healing properties^59. Hydrogel tunability allows the hydrogel mechanical properties to be matched to the mechanical needs of various tissues in specific tissue engineering contexts.

Despite their high tunability and reproducibility, synthetic hydrogels often lack the inherent bioactivity of natural materials. They may not contain the necessary cell adhesion ligands or growth factors to support robust cellular functions, and therefore cell viability can be reduced and tissue integration may be suboptimal in vivo. Moreover, some synthetic monomers and photoinitiators used in synthetic hydrogel fabrication can exhibit cytotoxicity, and therefore their concentrations and light exposure parameters must be carefully determined to ensure biocompatibility [27].

In photopolymerization-based bioprinting, potential cytotoxicity may arise from the chemical identity of monomers and photoinitiators and from the light-triggered radical generation process [86,87]. Potential toxic species include residual photoinitiators, unreacted monomers or macromers, photoinitiator-derived free radicals, reactive oxygen species (ROS), fragmented by-products, and long-term degradation products. [[86], [87], [88]]. These species may damage cellular proteins, lipids, and DNA, induce oxidative stress, and impair cell viability or cell-specific functions during and after photo-crosslinking [86,[89], [90], [91], [92]]. Therefore, cytocompatibility should be considered across multiple exposure stages, including prepolymerization contact with monomers and photoinitiators, radical generation during illumination, early leaching of residual components, and long-term exposure to degradation products [87,88].

The toxicity of photoinitiators is strongly dependent on initiator type, concentration, wavelength, irradiation dose, exposure time, and cell type [[93], [94], [95], [96], [97], [98], [99], [100], [101], [102], [103]]. Commonly used photoinitiators, such as lrgacure 2959 and LAP, can support high cell viability within optimized concentration and exposure windows, whereas excessive initiator concentrations or light doses may increase ROS generation, reduce cell survival, and alter cell morphology [97,99,100]. TPO-based systems require particular caution because free TPO may show strong dose-dependent cytotoxicity, particularly under irradiation. Therefore, optimizing photoinitiator concentration, light intensity, exposure duration, and post-curing removal of residual species is essential for balancing polymer conversion, mechanical integrity, and cytocompatibility.

The absence of intrinsic b 11 ioactivity often requires bioactive molecules, peptide ligands, degradable linkers, or surface functionalization incorporated into the gel to promote desired cellular responses. In addition, applying efficient visible-light photochemistry, reducing an unnecessary radical burden, improving monomer conversion, removing residual photoinitiators, and incorporating antioxidant or radical-scavenging components when appropriate may further reduce photopolymerization-associated toxicity. However, these approaches should be carefully optimized because excessive radical scavenging may slow gelation and compromise mechanical properties.

3.1.4. Hybrid approaches and composite bioinks

To capture the best of both worlds, hybrid bioinks combining natural and synthetic polymers are increasingly being developed. These formulations are designed to achieve a balance between biocompatibility, bioactivity, and mechanical robustness. For example, hyaluronic acid-alginate hydrogel inks were used in a “printing in liquid” technique coupled with in situ precipitation of nanoparticles, in which the ionic crosslinking of alginate provided initial structural stability, and photo curable hyaluronic acid improved the mechanical and physiological stability of hydrogel [104]. This two-step crosslinking method, combining different types of bonds, was used to fabricate nanocomposite hydrogels with significantly improved mechanical strength, biostability, and biological performance compared to pure hyaluronic acid hydrogels [104].

In addition to polymer blends, composite hydrogels incorporating various additives, such as nanoparticles, microparticles, or fibers, are gaining traction for enhancing mechanical strength and introducing additional functionalities [105,106]. Nanoparticle-reinforced hydrogels, for example, have exhibited remarkable mechanical reinforcement by improving particle hydrogel interactions, particularly when particles were incorporated in situ during printing rather than preloaded, which can reduce printability [104]. Multi-material bioprinting, which can be used to fabricate heterogeneous multicellular constructs through the precise deposition of different bioinks, further expands the possibilities of replicating the complex compositions and diversity of native tissue [107]. These hybrid and composite bioinks represent a crucial step towards developing bioinks that can simultaneously meet the stringent requirements of printability, mechanical toughness, and biological functionality in light-assisted 3D bioprinting.

4. Engineering photopolymerized hydrogel toughness through bond chemistry

Engineering hydrogel toughness for load bearing use is fundamentally a bond level design problem. Tough 3D-printing methods can be organized by crosslink class (Table 1, Table 2): (i) irreversible covalent networks formed via radical photopolymerization (e.g., GelMA, PEGDA) [69,108], thiol-ene click reactions [109,110], and redox photo-crosslinking [19,111,112] that deliver high stiffness and patterning fidelity but have little capacity for network rearrangement (typical bond energies ∼220–570 kJ/mol) [113,114], potentially limiting cellular adaptation; (ii) dynamic covalent motifs such as imines, hydrazones, oximes, boronate esters, disulfides, and Diels–Alder adduct exchange under pH, redox, thermal, or ligand cues, enabling stress relaxation, self-healing, shape memory, and programmable degradation while maintaining moderate stability (∼80–300 kJ/mol) [[115], [116], [117], [118]]; and (iii) dynamic non-covalent interactions, including hydrogen bonding, electrostatics, hydrophobic association, host-guest recognition, π-π stacking, and metal-ligand coordination, that respond to pH, redox, temperature, or glucose, provide reversible shear-thinning, rapid self-recovery, and cell-friendly viscosities suited to extrusion (generally <200 kJ/mol [[119], [120], [121], [122], [123]], tunable via multivalency or coordination). Selecting among these mechanisms requires balancing mechanical integrity, printability (shear thinning versus feature fidelity), and cytocompatibility. Static covalent networks maximize strength and resolution, dynamic covalent chemistries add adaptability without forfeiting robustness, and physical bonds provide rheological control and recovery. In practice, bond chemistry should be matched to the biofabrication modality and target tissue function and not treated as a one size fits all solution. In this review, the mechanical performance of hydrogels was systematically evaluated in terms of Young's modulus (E), compressive modulus (E_c), compressive strength (σ_c), and tensile strength (σ_t). To enable a direct comparison across studies, all values are expressed in kilopascals (kPa).

Table 1.

Bond chemistries of tough hydrogel bioink classes, representative chemistries, network structures, key advantages and limitations, and compatible 3D printing types.

Type of bonds	Interaction of types			Advantages	Limitations	3D printing types
Covalent bonds	Radical photo-polymerization	Thiol-ene click	Redox photo-crosslinking	Rapid curing; High strength; Excellent shape fidelity.	Brittleneless; Limited energy dissipation; No self-healing; prone to fatigue.	High resolution; Versatile materials.
Dynamic covalent bonds	Imine	Hydrazone	Oxime	Shelf-healing; Stress relaxation; Re-processability; Bioinspired adaptability.	Trade-off in strength/stability; Tunable kinetics; Biocompatible stimuli required.	Fine resolution; Low viscosity bioinks.
	Boronate ester	Disulfide	Reversible Diels-Alder
Dynamic noncovalent bonds/Physical bonds	Hydrogen bond	Electrostatic interaction	Hydrophobic association	Rapid shelf-healing; High toughness; Excellent biocompatibility.	Environmentally sensitive; Prone to creep; Lower long term stability.	Extrusion based printing.
	Host-guest	π-π stacking	Coordination complex

Table 2.

Bond chemistries of tough hydrogel bioink classes and approximate single bond energies.

Type of bonds	Bond energies	Bond energy breakdown
Covalent bonds	220-570 kJ/mol
Dynamic covalent bonds	27.2-477 kJ/mol
Dynamic noncovalent bonds/Physical bonds	0.1-300 kJ/mol

4.1. Covalent bonds for mechanical strength

Traditional hydrogels are typically formed through irreversible covalent crosslinking, which creates a stable, permanent polymer network. These covalent bonds provide high stiffness and strength, making them suitable for applications requiring robust structural integrity. Here, we review the mechanical properties of hydrogels fabricated via light-assisted extrusion and vat photopolymerization bioprinting.

Using light-assisted extrusion bioprinting, multifunctional hydrogels have been engineered by integrating bioactive molecules into multinetwork architectures. For example, GelMA hydrogels loaded with curcumin designed for use in diabetic wound healing significantly reduced ROS-induced apoptosis in ADSCs while achieving a Young's modulus of 100–150 kPa [124]. Liu et al. [125] developed a skin inspired hydrogel using acrylic acid and polypyrrole with a compressive modulus of 800 kPa and tensile strength exceeding 125 kPa, and with self-healing, stretchability, and conductivity characteristics appropriate for electronic skin applications. A microgel based biphasic (MB) hydrogel comprising gelatin microgels and GelMA precursors was used to generate vascularized organ constructs via the sequential printing in a reversible ink template (SPIRIT) method, with an elastic modulus of 8.55 ± 2.21 kPa and failure stress of 236.96 ± 2.67 kPa [126]. For osteochondral interface regeneration, tyramine-modified alginate crosslinked with gelatin under visible light achieved a compressive modulus of 8.41 kPa (pre-swelling) and 1.46 kPa (post-swelling) [127]. A maleic acid based LP-capped gel showed excellent extrusion fidelity and mechanical robustness with compressive strength reaching ∼1500 kPa, suitable for bioprinting load bearing constructs [128]. Quince seed mucilage-based hydrogels fabricated using the FRESH method demonstrated tunable compressive strength (0.6–1.2 kPa), offering a cost-effective solution for soft tissue engineering using plant-derived biomaterials [129].

Bioprinting via vat photopolymerization has enabled precise control over material architecture and mechanics. Methacrylated silk fibroin, with an 87.3%degree of substitution, supported osteoblast growth and achieved compressive moduli ranging from 12 kPa to 96 kPa, thus demonstrating tunable stiffness suitable for bone applications [130]. Chitosan bioink (CHI-MA) hydrogels exhibited a compressive modulus of 910 kPa, a strength of ∼150 kPa, very fine resolution (∼150 μm) [66], and high optical sensitivity under 405-nm light, with potential for soft tissue scaffolding. In cultured meat applications, GelMA crosslinked with IMA allowed bovine fibroblasts to differentiate into myogenic and adipogenic lineages in steak-like constructs with compressive modulus of 50.4 kPa^131. Strength tunable GelMA systems achieved tensile moduli from 1 kPa to 1 MPa, tensile strength up to 1358.6 kPa, and compressive moduli from 0.23 to 1081.9 kPa in the precise mechanical tuning of human umbilical vein endothelial cell (HUVEC)-laden projection-based bioprinting [131]. A silk fibroin based hydrogel sealant (Sil-MAS) with a tensile strength of ∼450 kPa exhibited fast curing and compatibility with NIH/3T3 cells, demonstrating its potential for clinical sealant applications^81.

These covalent bonds provide high stiffness and strength, and therefore are suitable for applications requiring robust structural integrity. The bond energy of covalent bonds, ranging from 220 to 570 kJ/mol (Table 1), is significantly higher than that of other types of bonds, indicating that more energy is required to break covalent bonds [132]. The density of covalent crosslinks directly influences the mechanical properties of a hydrogel, a higher crosslinking density generally leads to increased stiffness and tensile strength [21]. However, this comes at a cost: purely covalent networks often have limited elasticity and are prone to brittle fracture, as they lack mechanisms for dissipating energy under stress [133]. Once a covalent bond breaks, it does not readily reform, causing irreversible damage that can result in a rapid loss of mechanical integrity. This characteristic makes conventional covalently crosslinked hydrogels less suitable for tissue that experiences repetitive mechanical deformations or requires remodeling capabilities.

4.2. Dynamic covalent bonds for mechanical strength

In polymer networks, dynamic covalent bonds act as sacrificial bonds that enhance energy dissipation and stress relaxation, providing a basis for advanced bionic material design [134,135]. Dynamic covalent chemistry is a revolutionary approach to hydrogel design combining the stability of covalent bonds and the adaptability of dynamic interactions [54,133,136,137]. Unlike static covalent bonds, dynamic covalent bonds can reversibly break and reform in response to external stimuli (e.g., light, temperature, pH, mechanical stress). This allows the polymer network to rearrange, conferring the material with unprecedented properties, such as self-healing, stress relaxation, reprocessability, and adaptability [54,133,[136], [137], [138], [139]]. This dynamic nature is highly desirable in bioprinting applications, as it allows hydrogels to mimic the viscoelastic behavior of native tissue, accommodate cellular remodeling, and withstand mechanical fatigue.

Several specific dynamic covalent chemistries have been successfully integrated into photopolymerized hydrogels for 3D bioprinting (Table 1 for the network structure of the dynamic covalent hydrogels).

• (1)Boronate esters. Dynamic boronate esters provide tunable bond exchange in adaptive, catalyst free 3D printing resins, with room temperature bond exchange, interfacial welding, and post-printing functionalization. A combination of dynamic and static crosslinkers can balance mechanical stability with vitrimer-like reprocessability by increasing dynamic crosslinker content. This approach accelerates stress relaxation and promotes reversible strain stiffening while supporting self-healing, interfacial welding, and glucose-responsive behaviors in biomedical applications [59,[140], [141], [142], [143]].
• (2)Imine bonds. UV crosslinkable imine containing vitrimers based on vanillin and dimer fatty diamine cure rapidly, are rigid and thermally stable, and can be reprocessed within minutes. Increasing the imine content accelerates stress relaxation and enhances rigidity, supporting recyclable photoresins [144].
• (3)Thia-Michael addition. Reversible thia-Michael crosslinks (distinct from irreversible thiol-maleimide) enable pH and temperature tunable kinetics that directly control strain stiffening, providing a means of incorporating adaptive mechanical responses into hydrogels [145].
• (4)Disulfide bonds. UV induced disulfide formation and reduction provides both spatially and temporally controlled reversible switching between thiol and disulfide states. This capability supports dynamic photopatterning and on demand interfacial property tuning [146].
• (5)Hindered urea bonds. Photo printable networks with hindered urea bonds permit on-demand shape reconfiguration via homolytic exchange without altering the topology of the hydrogel, and can be converted to urethanes to modulate the hydrogel mechanics, enabling “print-once, program-many” workflows [147].
• (6)Double Michael addition to alkynones. Dynamic covalent dextran hydrogels formed by thiol-alkynone double Michael addition are injectable, self-healing, and biocompatible [148]. The reversible covalent crosslinks formed during the process directly influence the mechanical strength and strain stiffening behaviors, enabling tunable stiffness under a variety of conditions. These hydrogels function as predictable, sustained release depots for peptide vaccines.

In extrusion bioprinting, dynamic covalent, ionic, and enzymatic methods can be used to tailor the hydrogel rheology and post print mechanics. For example, an injectable, granular hyaluronic acid system combining norbornene and hydrazide modified chains with aldehyde-hyaluronic acid formed reversible hydrazone crosslinks that enabled shear thinning, self-healing, and cohesive jamming, with printed constructs that reached a compressive modulus of about 9 kPa and a failure stress of about 7 kPa. The granular hydrogel facilitates high levels of post-injection cell invasion, for example into myocardial tissues [149].

Bioprinting via vat photopolymerization (including DLP and volumetric approaches) can provide precise control over crosslink density and network topology, and dynamic chemistries can support post print reprogramming. A hindered urea bond network based on poly (propylene glycol) acrylate was formulated for use in DLP, with a Young's modulus of approximately 1.16 × 10⁶ kPa and a tensile strength of about 5.55 × 10⁴ kPa with slight reductions after bond isomerization. The reversible network permitted on demand mechanical reconfiguration suited to adaptive devices such as soft robotics and shape memory systems [147]. A synthetic photoresin based on norbornene-functionalized polyvinyl alcohol blended with polyethylene glycol diacrylate used in volumetric DLP, delivered rapid curing, reconfigurable mechanics, and a compressive modulus of 25.1 ± 1.7 kPa to provide high resolution printing of wearable soft constructs and responsive sensors [150].

Dynamic covalent bonds introduced into light-based 3D printing resins have shown promise in overcoming the limitations of conventional thermosets, which are crosslinked by permanent covalent bonds and therefore have limited adaptability and restricted reprocessability [137,138]. These reversible crosslinks produce crosslinked printed polymers with chemical and physical recyclability, self-healing capabilities, and degradability [138]. Current research on the mechanical properties of dynamic covalent bonds has centered largely on rheological assessments, whereas a systematic quantification of their compressive and tensile behaviors remains less explored.

4.3. Dynamic (non-covalent) bonds and supramolecular interactions

In addition to dynamic covalent bonds, non-covalent interactions are crucial for conferring hydrogels with viscoelasticity, shear-thinning, and self-healing properties. These interactions, including hydrogen bonding, host-guest interactions, ionic bonds, and hydrophobic associations, are inherently reversible and typically have lower bond energies compared to covalent bonds (Table 1 for the network structure of the dynamic hydrogels). With this reversibility, bonds can be rapidly broken and reformed, producing time dependent mechanical responses that are characteristic of many native biological tissues [151].

The ability of non-covalent interactions to provide reversible crosslinks is particularly advantageous in biofabrication. Hydrogels formed through physical crosslinks often exhibit shear-thinning behavior, in which their viscosity decreases under shear stress, allowing them to flow easily through a printing nozzle or be shaped during printing. Once the shear stress is removed, the bonds rapidly reform, leading to gelation and shape retention. This characteristic is highly desirable for extrusion based bioprinting and for handling soft materials in vat photopolymerization [152]. Furthermore, the dynamic nature of these bonds supports self-healing, in which damaged hydrogel networks spontaneously repair themselves, restoring the mechanical integrity of the construct. This is critical for improving the robustness and longevity of engineered tissue. Because these reversible, self-healing interactions underlie the rheological and biological foundation of associated bioinks, the following section classifies representative systems according to printing modality (extrusion vs. vat photopolymerization) and photochemical mechanism, thus explicitly linking the type of bond to the mechanics and functions that can be achieved by the hydrogel.

In a study of extrusion printed silk-gelatin scaffolds formed by horseradish peroxidase mediated crosslinking of tyramine-functionalized silk fibroin and gelatin, a compressive modulus of 384 kPa and tensile strength of approximately 8000 kPa, were achieved, and the constructs were able to maintain human adipose-derived stem cell (ADSC) viability and chondrogenic differentiation [153]. Nutrient perfusable microchannel hydrogels have been produced by coaxial extrusion of sodium alginate with calcium chloride and ionotropic gelatin, which supported L929 fibroblast proliferation and had a tensile strength of 116 kPa [154]. Chitosan-gelatin bioinks ionically crosslinked with tripolyphosphate and printed under mild conditions were used to produce uniaxial scaffolds with a compressive modulus of 0.79 kPa and good cytocompatibility [155]. A chitosan-based composite reinforced with cellulose nanocrystals and hydrophobic physical associations exhibited a yield stress of 0.585 kPa, Young's modulus of 0.128 kPa, and storage modulus of 13 kPa, and supported mouse embryo osteoblast precursor cell (MC3T3-E1) osteogenic differentiation [156]. Viscoll collagen (chemically unmodified) showed excellent extrusion fidelity and structural stability with a storage modulus of 4.72 kPa, while maintaining NIH 3T3 viability [157]. A decellularized ECM bioink from porcine liver microparticles blended with gelatin formed a physically crosslinked hydrogel with a compressive modulus around 6 kPa that retained printability and HUVEC compatibility [158].

Vat photopolymerization (including DLP and volumetric approaches) has been used to polymerize a rotaxane hydrogel from acrylated β-cyclodextrin (host) and ethylene glycol diacrylate (guest), which formed a supramolecular network with a storage modulus exceeding 35 kPa and compressive strength below 40 kPa, providing rapid deformation and biosensing for wearables [159].

Although dynamic non-covalent bonds have excellent responsiveness and adaptability, their mechanical strength is generally lower than that of covalent or dynamic covalent bonds. Therefore, they generally must be integrated with stronger crosslinking mechanisms to achieve the desired toughness in load bearing applications. The interplay between the kinetics of bond breaking and reformation and the overall density of these dynamic interactions dictate the viscoelastic properties and the extent of toughness that can be achieved in these hydrogels.

4.4. Hybrid approaches for mechanical strength

To address the mechanical limitations of hydrogels used in biofabrication, particularly for brittleness and low moduli, hybrid designs involving covalent, dynamic, and physical bond crosslinking, as well as composite reinforcement, have become increasingly more common.

4.4.1. Light-assisted extrusion bioprinting

In one study, GelMA/PLA hybrid scaffolds fabricated by extrusion reached a compressive strength of 54.26 × 10³ kPa and modulus about 98.73 × 10³ kPa in personalized ear constructs, and were suitable for auricle reconstruction [160]. Bioconcrete inks, consisting of A-C composite bioinks with eletrosprayed GelMA microgels (A component) embedded in a GelMA prepolymer matrix (C component), achieved a modulus of 2.61 × 10³ kPa with the potential for large scale bone regeneration [41]. Dual covalent composite hydrogels using N-acryloyl glycinamide (NAGA), GelMA, and lignin nanoparticles reached a compressive modulus of 650 kPa and strength of around 7.0 × 10³ kPa, and provided cartilage repair with antioxidative capacity [161]. A marine-inspired microfluidic chitosan hydrogel had a modulus of 1.51 kPa with high printability, appropriate for soft tissue [162]. Loosely pre-crosslinked cellulose hydrogels designed for osteogenic applications exhibited a compressive modulus of 2.9 × 10³ kPa and tensile modulus of 6.6 × 10³ kPa [163]. Gelatin-fibrinogen muscle mimetic constructs had a tensile modulus of 147.7 kPa and an aligned myofiber architecture [164]. A MeTro/GelMA hybrid hydrogel for vascular tissue printing had a modulus of 49.2 kPa with elastic recovery [165]. Hyaluronic acid-alginate-calcium phosphate nanocomposites for bone exhibited a modulus of 6.4 kPa and toughness of 359 J/m^{2 104}. Cellulose nanocrystals (CNC) and alginate enhanced the printability of inks and reached a modulus of 52.6 kPa to support chondrocytes [166]. PEI-silica-alginate composites used for in situ bone scaffold printing attained a Young's modulus of 1.84 × 10⁴ kPa [167]. Decellularized myocardium GelMA-MeHA bioinks crosslinked by microbial transglutaminase were used in cardiac tissue formation, with a modulus up to 18.4 kPa [168]. Recyclable poly (N-isopropylacrylamide)-based direct ink writing (PNIPAM-based DIW) composites with carbon nanotubes or MXene achieved a modulus up to 1.0 × 10⁵ kPa and high fidelity temperature responsiveness [169]. Carboxymethyl chitosan (CMCS)/functionalized carbon nanotube (CNT) hydrogels designed for strain sensing had a tensile strength of 123.97 kPa and a compressive strength of 8.66 kPa [170]. Dual crosslinked gelatin-ZIF-8-CaO₂ hydrogels for oxygen-releasing bone regeneration scaffolds reached a modulus of 5.92 × 10³ kPa and strength of 2.34 × 10^3 172. GelMA-oxidized dextran (OD) hybrids conferred injectable constructs with a compressive strength around 100 kPa [171]. Conductive microfiber hydrogels for wearable electronics comprising PAA, alginate, and graphene-oxide achieved a tensile strength of 550 kPa with stable phase control [172]. GelMA-methacrylated chondroitin hydrogels printed at physiological temperature exhibited a modulus of 700 kPa with high-fidelity curing [173].

4.4.2. Vat photopolymerization

Peptide-reinforced GelMA with rigid peptide nanorods for osteochondral regeneration were generated using DLP, achieving a compressive strength up to 712.7 kPa [174]. DLP-printed PVA/AAm hydrogels exhibited a tensile strength from 20 to 10.46 × 10⁴ kPa, bridging soft hydrogel to plastic-like regimes [175]. A CMAV-based resin combining reprintability and self-healing with a tensile strength of 32.6 × 10³ kPa and modulus of 2.09 × 10⁶ kPa, was developed as a sustainable, high performance DLP platform [176]. Temperature-controlled projection stereolithography was used to produce PAAm-κ-carrageenan double networks with a tensile stress of 200 kPa and elongation exceeding 2400% [177]. DLP-printed GelMA-PLA auricle scaffolds reached a compressive strength of 63.43 × 10³ kPa for cartilage reconstruction applications [178]. Cardiac decellularized left ventricle (H-dECM)-GelMA bioinks printed by DLP exhibited a compressive modulus around 6.5 kPa with robust cell support [179], while methacrylated hyaluronic acid-GelMA systems provided high-fidelity photocuring with a modulus of 176.7 kPa [180].

These examples suggest that hybrid-network and composite design methods offer the broadest mechanical tunability among light-assisted printed hydrogels. Their advantage arises from combining a photo-crosslinked covalent framework with dynamic or sacrificial bonds, interpenetrating networks, fillers, microgels, or hierarchical architectures, thereby improving shape fidelity, energy dissipation, and damage tolerance. However, these improvements are accompanied by important trade-offs. The higher polymer content, filler loading, or network complexity may increase viscosity, reduce light penetration, enhance scattering, require higher photoinitiator doses, and compromise cytocompatibility. Therefore, hybrid and composite hydrogels should not be evaluated solely by their peak modulus or strength, but instead by how effectively they balance mechanical reinforcement, printability, and biological compatibility in tissue-specific applications.

To further compare the mechanical performance based on different types of bonds, we benchmarked Young's modulus (E), compressive modulus (E_c), compressive strength (σ_c), and tensile strength (σ_t) across bond classes (covalent bond, dynamic covalent bond, physical bond, and hybrid bond) and printing modalities (light-assisted extrusion and vat photopolymerization). All values were converted to kPa to be easily compared. Fig. 2 summarizes these ranges and distributions. Static covalent networks generally achieve the highest stiffness and strength; dynamic covalent systems provide intermediate mechanics with enhanced stress relaxation; and physical networks exhibit lower moduli and strength but superior shear thinning and recovery. Hybrid designs spanned the broadest window, often reaching load bearing targets while retaining printability and cytocompatibility.

Fig. 2.

Mechanical performance of hydrogels fabricated via light-assisted 3D bioprinting as a function of crosslinking mechanisms. (a, c) Young's modulus (E/E_c) and compressive/tensile strength (σ_c/σ_t) of hydrogels fabricated by light-assisted extrusion 3D printing. (b, d) Young's modulus (E/E_c) and compressive/tensile strength (σ_c/σ_t) of hydrogels printed by vat photopolymerization. Hydrogels are categorized by their dominant crosslinking types: covalent bonds (orange), physical bonds (green), hybrid bonds (blue), and dynamic covalent bonds (purple). Each black square represents data from a reported hydrogel formulation. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

The mechanical robustness of hydrogels remains a critical determinant for their appropriateness for biomedical applications, particularly in load bearing tissue. Recent advances in light-assisted 3D bioprinting, including vat photopolymerization and light-assisted extrusion, have enabled precise spatial control over hydrogel polymerization, yet the mechanical characteristics are largely dictated by the underlying crosslinking mechanism (a summary is provided in Table 3, Table 4).

Table 3.

Light assistant 3D printing via diverse chemical bonding strategies.

Types of bonds	Materials	Young's modulus (kPa)	Compression/tensile strength (kPa)	3D printing	Chemical structure	biocompatibility	Reference
Covalent	QSGM	Ec: 1.2 kPa	σc: 0.6 ± 0.1 to 1.2 ± 0.3 kPa	FRESH	Covalent	BMDCs	129
	Alg-Tyr-gel, CMC-Tyr-gel	Ec: 8.41 kPa (Sb) 1.46 kPa (Sa)		Extrusion	Covalent	hBMSCs	127
	Cur-GelMA	Ec: 100∼150 kPa		Extrusion	Covalent	ADSCs	124
	AA with DCh-PPy20	Ec: 800 kPa	σc: 60∼65 kPa σt: 10∼12.5 kPa	Extrusion	Covalent		125
	GelMA	E: 8.55 ± 2.21 kPa	σt: 236.96 ± 2.67 kPa	SPIRIT	Covalent		126
	Lipoic acid and PEG		σc: ∼1500 kpa	Extrusion	Covalent	MC3T3-E1	128
Physical	Chitosan and gelatin	Ec: 0.79 kPa		Extrusion	Physical	L929	155
	Chitosan and CNCs	E: 0.128 kPa		Extrusion	Physical	MC3T3-E1	156
	Viscoll collagen	Ec: 21.5 ± 1.4 kPa		Extrusion	Physical	NIH 3T3	157
	Silk-gelatin	Ec: 384 kPa	σc: ∼8000 kPa	Extrusion	Physical	hADSCs	153
	PVA/PEI/LiCl	E: 259 kPa	σt: 1.800 kPa	Extrusion	Physical		181
	dECM -gelatin	Ec: ∼6 KPa		extrusion	Physical	hHUVEC	182
	NaAlg		σt: 116 kPa	Extrusion	Physical	L929	154
Dynamic covalent	NOrHA/Hyd-NOR or HA/AID-NOrHA	Ec: ≈9 kPa	σc: ≈7 kPa	Extrusion	Dynamic covalent	HUVEC and MSCs	183
	GM-HAc/HAc/Cap	Ec: 6.4 kPa	σc: 359 kPa				104
	GelMA & MCMA	Ec: 700 kPa		Extrusion	dual covalent		173
Hybrid	Alginate, gellan gum, and cationic silica	Ec: 100∼120 kPa		Extrusion	Dual physical	Chondrocyte	184
	GelMA with (NH4)2SO4		σt: 6600 kPa	Extrusion	External treatment		185
	AAm, CC, CaCl2, ZIF-8-CaO2	Ec: 5920 kPa	σc: 2340 kPa	Extrusion	Dual covalent & composite	MC3T3-E1	186
	PNIPAM/Mxene or PNIPAM/CNT	E: ∼100000 kPa	σt: ∼1000 kPa	Extrusion	Dual physical bonds		169
	GelMA/PLA	Ec: 98730 kPa	σc: 54.26 ± 12.25 × 103 kPa	Extrusion	Hybrid covalent	Cartilage fragments	160
	Bioconcrete/GelMA	Ec: 2608 Kpa		Extrusion	Composite covalent	BMSCs	41
	NAGA/GelMA/LPN	Ec: 650 kPa	σc: ∼7000 kPa	Extrusion	Dual covalent Composite	BMSCs	161
	Chitosan- PHEC	Ec: 1.51 kPa		Extrusion	Physical Covalent	3T3-L1	187
	Cellulose	Ec: 2900 kPa E: 6600 kPa		Extrusion	Hybrid	MC-3T3	163
	Gelatin & fibrinogen, Alginate & Thrombin	E:147.7 Kpa		Extrusion	Dual physical	C2C12	164
	MeTro/GelMA	Ec: 49.2 ± 8.7 kPa		Extrusion	Dual covalent	HUVECs	165
	Cellulose Nanocrystals - Alginate	Ec: 52.6 Kpa		Extrusion	Dual covalent	rMSCs	166
	NaAlg/PEI/Silica	E: 18370 kPa			Dual physical	human nasoseptal chondrocytes	167
	dhECM/GelMA, GelMA-MeHA	Ec: 9.9 ± 2.6 kPa (w/o mTGase), 18.4 ± 2.8 kPa (w/mTGase)		Extrusion	Dual Physical Composite		168
	CMCS & fCNT	Ec: 8.8×102 kPa	σc: 8.66 kPa σt: 123.97 kPa	Extrusion	Composite	Human fibroblasts	170
	GelMA OD	Ec: 13 kPa ∼15 kPa	σc: ∼100 kPa	Extrusion	Covalent hybrid		171
	PAA, GO, PAA-Alginate, CaSO4		σt:∼550 kPa	Extrusion	Physical and Covalent		172
	HEMA	Ec: 363000 kPa		Extrusion	Nanocomposite	L929	188
	Me-Gel Me-HA	Ec: 1.16 ± 0.96 kPa			Dual covalent	Breast cancer cells and lung fibroblasts	189

Table 4.

Vat photopolymerization based on 3D printing via diverse chemical bonding strategies.

Types of bonds	Materials	Young's modulus (kPa)	Compression/tensile strength (kPa)	3D printing	Chemical structure	biocompatibility	Reference
Covalent	GelMA	Ec: 50.4 kPa		DLP	Covalent	BEFS-teton-MyoD, BEFS-PPARγ2	190
	SF-MA	Ec: 12 kPa ∼96 kPa		DLP	Covalent	Osteoblasts	130
	GelMA	E: 1 kPa∼ 1000 kPa Ec: 0.23 ∼1081.9 kPa	σt: 3.24 to 1358.6 kPa	DLP	Covalent	HUVECs	131
	chitosan-MA	Ec: 910 Kpa	σc: ∼150 Kpa	DLP	Covalent	HUVECs	66
	GMA to Sil-MAS		σt: ∼450 Kpa	DLP	Covalent	NIH/3T3	81
	IBOA		σt: ∼8000 kPa	DLP	Covalent		191
Physical	CD/PEG CD/PAAm	Ec: 35 ∼40 kPa	σt: ∼78.1 kPa	DLP	Physical		159
Dynamic covalent	HBA & HUBM	Isomerization Ec: 1.16 ± 0.06 × 106 kPa (before) Ec: 1.12 ± 0.06 × 106 kPa (after)	Isomerization: σc: 55.50 ± 6.3 × 103 kPa (before) σc: 53.1 ± 5.2 × 103 kPa (after)	DLP	Dynamic		147
	BDG CROSS	E: 1.4∼1.6×106 kPa	σt: 35000∼40000 kPa	DLP	Dynamic		144
	PVA & PEG2SH	Ec: 25.1 ± 1.7 kPa		DLP	Dynamic covalent bond	C2C12	150
Hybrid	GelMA & PRNs Nanorods	Ec: 712.7 kpa		DLP	Composite		174
	PVA/PAAm	E: 5.92×105 kPa	σt: 0.020 ± 0.002 kPa to 104.58 ± 4.4 × 103 kPa	DLP	Covalent hybrid		175
	CMAV	E: 2.09 × 106 kPa	σt: 32600 kPa	DLP	Dual dynamic covalent		176
	κ-Carrageenan/PAAm		σt: 200 kPa, σc: 15000 kPa	TOPS	Hybrid		177
	HAMA/GelMA	Ec:176.7 ± 8.5 kPa (before), 42.9 ± 3.8 kPa (after)		DLP	Dual covalent	fC2C12	180
	GelMA/PLA		σc: 63430 kpa	DLP	Covalent hybrid	Chondrocyte	178
	dECMs/GelMA	Ec: 6.5 kPa		DLP	Physical & covalent	hiPSC	179
	AAC & ACMO	E: 2.61 ± 0.08 × 106 kPa		DLP	Dual dynamic covalent		192

∗Abbreviation.

-Materials.

QSGM: Quince seed mucilage/glycidyl methacrylate, Alg-Tyr-gel: Alginate modified with tyramine and gelatin, CMC-Tyr-gel: Cellulose modified with tyramine and gelatin, Cur-GelMA: Curcumin-incorporated gelatin methacryloyl, AA: Acrylic acid, GelMA: Gelatin methacryloyl, PEG: Polyethylene Glycol, CNCs: cellulose nanocrystals, PVA/PEI: poly(vinyl alcohol)/Poly(ethyleneimine), NaAlg: Sodium Alginate, dECM: Decellularized extracellular matrix, Hyd-NorHA: NorHA was then separately modified with either hydrazide groups, Ald-NorHA: aldehyde groups, NorHA: HA was modified with norbornene groups, GM-HAc/HAc/Cap: glycidyl methacrylate-hyaluronic acid/hyaluronic acid/CaP, MCMA: methylcellulose methacrylate, AAm: Acrylamide, CC: carboxymethylcellulose, PNIPAM: poly(N-isopropylacrylamide), CNT: carbon nanotube, PLA: polylactic acid, NAGA: N-acryloyl glycinamide, LPN: Laponite XLG, PHEC: chitosan-based hydrogels via a phenolated, polyelectrolyte complex, MeTro: methacryloyl-substituted recombinant human tropoelastin, MeHA: methacrylated hyaluronic acid, CMCS: carboxymethyl chitosan and f-CNTs: unzipped carbon nanotubes, OD: oxidized dextran, PAA: Polyacrylamide, GO: graphene oxide, HEMA: hydroxyethyl methacrylate, MeHA and MeGel: methacrylic anhydride (Sigma) with HA (0.5%) and gelatin (10%) solution, SF-MA: methacrylated silk fibroin, Sil-MAS: silk fibroin sealant, IBOA: Isobornyl acrylate, BDG: Bifunctional Vanillin-Based Building Block Bridge, CROSS: The Trifunctional Vanillin-Based Cross-Linker (CROSS), HUBM: hindered urea containing bismethacrylate, HBA: 4-hydroxybutyl acrylate, PEG2SH: α,ω-bis(mercapto)-poly(ethylene glycol), PRNs: peptide-based rigid nanorods, CMAV: Cystamine methacrylate vanillin, AAC: Hydrolysable acetal acrylate cross-linker, ACMO: hydrophilic 4-acryloylmorpholine monomer.

-Cells.

BMDCs: Bone marrow-derived DCs, hBMSCs: Human bone marrow mesenchymal stem cells, ADSCs: Adiposederived stem cells, MC3T3: Mouse osteoblast-like bone cells, L929: Mouse fibroblast cells, NIH3T3: The embryonic mouse fibroblast cell line, HUVEC: Human Umbilical Vein Endothelial Cells, 3T3-L1: fibroblast, MC-3T3 cell: Mouse bone cells, C2C12: The mouse myoblast cell line, hiPSC: Human Induced Pluripotent Stem Cells.

As illustrated in Fig. 2, hydrogels crosslinked by physical interactions (e.g., ionic bonding, hydrogen bonding, hydrophobic associations) have limited mechanical performance, with Young's moduli and compressive and tensile strengths generally below 10²–10³ kPa. In contrast, covalently crosslinked hydrogels exhibit improved stiffness and strength, with a Young's modulus up to ∼10⁴ kPa and strength in the 10³–10⁵ kPa range. However, these networks often suffer from brittleness and poor energy dissipation.

The term “tough hydrogel” should not be simply equated with a hydrogel possessing a high elastic modulus. Stiffness, strength, and toughness describe different mechanical characteristics. Stiffness reflects resistance to elastic deformation and is commonly represented by Young's modulus or the compressive modulus; strength reflects the maximum stress that a material can withstand before failure and is commonly represented by tensile or compressive strength; and toughness describes the ability of a material to resist crack initiation, crack propagation, fatigue damage, and irreversible structural failure through energy-dissipating mechanisms [61,132,193]. Therefore, in a strict fracture mechanics sense, hydrogel toughness should ideally be evaluated by fracture energy, work of fracture, tearing energy, hysteresis, fatigue threshold, cyclic durability, and damage tolerance, rather than by modulus or strength alone [193,194].

Mathematically, the toughness density $U_T$ can be defined as the energy absorbed per unit volume before rupture, which can be expressed as the area under the stress–strain curve:

$$U_T={\int }_0^{{\epsilon }_f}\sigma d\epsilon$$

where $U_T$ is the toughness density, $\sigma$ is the stress, and ${\epsilon }_f$ is the strain at failure. We also introduced the fracture energy $\Gamma$ as a fracture mechanics parameter that describes resistance to crack propagation. Therefore, from a strictly mechanical perspective, hydrogel toughness should ideally be evaluated by fracture energy, work of fracture, tearing energy, hysteresis, fatigue threshold, cyclic durability, and damage tolerance, and not by modulus or strength alone.

From a network design perspective, tough hydrogels are generally produced either by generating more homogeneous network architectures that reduce the stress concentration or by introducing energy-dissipating mechanisms into the polymer network. These design principles can be implemented through double networks or interpenetrating networks architectures, slide-ring topologies, highly entangled networks, nanocomposite reinforcement, and hierarchical or anisotropic structures [21,194]. Of these, double network hydrogels provide a classic example: a rigid and brittle first network dissipates mechanical energy through sacrificial fracture or local yielding, while a soft and ductile second network preserves the overall structural integrity and delays catastrophic crack propagation [61].

Importantly, the meaning of “tough” should also be interpreted according to the specific biomedical application. Soft tissue generally requires compliant matrices in the kPa range, whereas hard or mineralized tissue, such as cartilage, bone, and dental tissue, requires substantially higher mechanical performance because of its mineralized ECM, hierarchical architecture, and load-bearing functions [195]. Recent reports on hard tissue organoids further highlighted that recreating hard tissue-like microenvironments requires integrated biochemical cues, matrix stiffness, mineralized structures, and mechanical signals [195]. Accordingly, in this review, tough hydrogels are broadly defined as mechanically reinforced, highly hydrated polymer networks that exhibit high resistance to deformation, mechanical failure, and structural damage through their rational network design and energy-dissipating mechanisms. In hydrogels generated by light-assisted 3D bioprinting, mechanical reinforcement must be balanced with photo-crosslinking efficiency, light penetration, printing fidelity, nutrient diffusion, degradation behavior, and cytocompatibility.

In 3D bioprinting, biocompatibility evaluations must extend far beyond determining the short-term survival of encapsulated cells (e.g., live/dead staining). In tough hydrogels, a critical yet often overlooked factor is the regulation by the printed matrix on cellular behaviors, fate, and functions, a process governed by cellular mechanotransduction. The native ECM is a highly dynamic system that continuously delivers both static (e.g., invariable stiffness) and dynamic (e.g., viscoelasticity and stress relaxation) biophysical cues to resident cells [196]. Therefore, to engineer strong living scaffolds for load-bearing musculoskeletal tissue regeneration, the structural and mechanical properties of tough hydrogels must be deliberately designed. According to our previously described chemical classification, the evolution from static to dynamic mechanotransduction can be precisely tailored through the choice of covalent, physical, dynamic covalent, and hybrid bonds.

4.4.2.1. Densely crosslinked covalent networks and cellular confinement

Traditional tough hydrogels often rely on dense covalent networks to achieve high mechanical strength. These networks present static mechanical cues due to their high bulk stiffness and invariable elasticity [196]. While a stiff static substrate can strongly promote cellular mechanosensing and direct MSCs toward osteogenic differentiation in 2D environments, it presents a paradox in 3D bioprinting. Within the 3D space, these static and overly dense covalent networks act as physical barriers. Without the ability to yield or dynamically remodel, static tough hydrogels severely restrict cell spreading, migration, and new ECM deposition, ultimately hindering long-term tissue maturation.

4.4.2.2. Reversible non-covalent interactions for stress relaxation and cell remodeling

To reconcile the trade-off between macroscopic toughness and microscopic cellular confinement, reversible non-covalent interactions (e.g., hydrogen bonds, host-guest interactions, or electrostatic interactions) can be introduced to provide highly dynamic mechanical cues. Physical bonds endow the hydrogel with rapid stress relaxation and shear-thinning properties. Dynamic stress relaxation allows cells to physically deform the matrix by exerting traction forces, which actively promote integrin clustering, focal adhesion maturation, and downstream mechanotransduction pathway activation (e.g., YAP/TAZ nuclear translocation). For example, 3D-bioprinted nano-laponite hydrogel constructs dissipated energy dynamically while presenting specific mechanosensing cues that activated key signaling cascades, including the PI3K/AKT pathway, promoting osteogensis [197].

4.4.2.3. Dynamic covalent networks for adaptable viscoelasticity

Dynamic covalent bonds (such as hydrazone, Schiff base, and boronate ester bonds) offer an ideal balance between the robust stability of static covalent bonds and the rapid relaxation of physical bonds. They provide adaptable viscoelastic matrices with tunable stress-relaxation times, thereby determining how readily the matrix yields to cell generated traction forces. Cells encapsulated within dynamic covalent networks can trigger force-induced network reorganization and local softening or remodeling of the hydrogel network through bond exchange mechanisms. This adaptability is particularly crucial in the fabrication of dynamic hydrogels for 3D-bioprinted cartilage organoids [198], as it permits continuous cellular remodeling and ECM synthesis while providing the necessary structural support over time. Furthermore, the integration of stimuli-responsive motifs (e.g., photothermos-responsive bioinks) can improve printability during extrusion while preserving the dynamic remodeling capabilities required for cell growth post-printing [199].

4.4.2.4. Hybrid bond networks for biomimetic heterogeneous tissue constructs

Different types of tissue inherently have specific mechanical demands and static–dynamic balances. For example, osteochondral regeneration requires scaffolds that accommodate the gradient transition from compliant cartilage to stiff subchondral bone. By integrating hybrid bond networks (e.g., interpenetrating networks of covalent and physical bonds), 3D-bioprinted scaffolds can achieve a biomimetic static–dynamic synergy [200]. Through spatially controlled, light-assisted printing, the covalent network component can provide robust static load-bearing support in the osseous region, while the physical or dynamic covalent components simultaneously maintain a highly dynamic, stress-relaxing microenvironment in the chondrogenic region. Consequently, future tough bioinks should be designed to form scaffolds with more than static structural support after printing and crosslinking. These scaffolds should integrate hierarchical architectures with dynamic, cell-remodelable microenvironments to support heterogeneous and biomimetic tissue-specific regeneration [201].

Notably, hybrid bond networks generally show broadest mechanical coverage among the summarized systems. By combining a photo-fixed covalent backbone with dynamic covalent or non-covalent sacrificial interactions, hybrid systems simultaneously combine crosslinking methods, which combine physical and covalent interactions, consistently outperform single mode networks. Hybrid crosslinked systems exhibit exceptional mechanical profiles, with Young's moduli and strength values exceeding 10⁵ kPa under both light-assisted extrusion and DLP. The synergistic nature of hybrid bonding simultaneously provides reinforcement and flexibility, leading to tougher and more durable hydrogels. An emerging class of dynamic covalent bonds can further enhance DLP-printed hydrogel performance by introducing reversible covalent linkages (e.g., Schiff base, disulfide exchange). These provide energy dissipation, self-healing, and stress relaxation, which have synergistically increase strength levels above 10⁶ kPa. Table 5 summarizes the relationships among the network strategy, dominant bond type, approximate modulus/strength range, native tissue or ECM applicability, and printing advantages. Specifically, Table 5 compares physical and non-covalent networks, permanent covalent networks, dynamic covalent networks, and hybrid bond networks, and relates their typical mechanical regimes to representative types of tissue, including soft ECM-like tissue, reinforced soft tissue, and dynamic and remodeling tissue, along with appropriate scaffolds for use in cartilage, osteochondral, tendon, ligament, and bone restoration.

Table 5.

Correlation between bond types, mechanical properties, and tissue-specific applications in vat-based bioprinting.

Network strategy	Dominant bond type	Modulus/strength range	Native tissue/ECM relevance	Printing advantages
Physical/non- covalent	Ionic bonds H-bonds Host-guest interactions	E or σ < 102–103 kPa	Soft ECM-like tissues [202], e.g., brain [203], liver [204], lung [205], muscle [[206], [207], [208]].	Self-healing, shear-thinning, cell-friendly processing
Permanent covalent	Irreversible photo-crosslinked bonds, e.g., methacrylate/acrylate	E up to 103–104 kPa σ can reach 103–105 kPa	Reinforced soft tissues [202], e.g.,skin [209], cornea [210], cartilage-related matrices [211]	High fidelity, rapid shape fixation, structural integrity
Dynamic covalent	Reversible bonds, e.g., Schiff base, hydrazone, boronate ester	Generally10 [2]–103 kPa, depending on exchange kinetics	Dynamic ECM-like tissues [202], e.g., muscle, cartilage-related remodeling tissues [212,213]	Stress relaxation, rapid shape fixation, self-healing, adaptive remodeling, tunable viscoelasticity
Hybrid bond	Double networks, IPNs, fillers, covalent + dynamic or sacrificial bonds	Broadest range; Often > 103 kPa	Hard tissues [202], e.g., Cartilage [214], tooth [215,216], tendon [217]/ligament [218], bone-related scaffolds [219]	Energy dissipation, high toughness, damage tolerance

Overall, hybrid and dynamic covalent networks represent a promising means of balancing printability with mechanical integrity in photo-crosslinked hydrogels. These systems can produce bioinks suitable for engineering soft to hard tissue interfaces and structurally resilient implants.

4.5. Design principles: translating bond-type toughening into light-assisted 3D bioprinting

Bond-centric toughening achieved through permanent covalent networks, dynamic covalent bonds, dynamic non-covalent interactions, and hybrid designs, has significantly expanded the mechanical design possibilities of hydrogels. However, high mechanical performance must be balanced against the strict constraints of light-assisted 3D bioprinting cnotaining living cells. In addition to conventional metrics, such as strength and modulus, biofabrication has three critical constraints: (i) printability, which is affected by viscosity range, shear-thinning behavior, and photo-crosslinking kinetics; (ii) cytocompatibility of all components, including monomers, crosslinkers, photoinitiators, and fillers; and (iii) mechanobiological function, which is the capacity of the material to facilitate cell sensing, remodeling, and migration within a viscoelastic matrix over relevant time scales [[220], [221], [222]].

Purely covalent, densely crosslinked networks provide superior stiffness and shape integrity but typically function as virtually elastic solids with constrained stress relaxation. This can lead to a “frozen” microenvironment in which cell viability is preserved, but essential processes such as spreading, migration, and matrix remodeling are significantly impeded, despite a bulk modulus congruent with that of normal tissue [223]. Dynamic covalent chemistries can somewhat mitigate this constraint by incorporating exchangeable bonds that regulate stress relaxation and plasticity in physiological settings. However, their working range is frequently limited by very specific pH levels, catalysts, or temperature conditions. Dynamic non-covalent interactions are particularly appealing in bioinks because of their intrinsic reversibility, functionality under mild aqueous conditions, and shear-thinning and self-healing properties. However, when used in isolation, they often do not have sufficient long-term stiffness and fatigue resistance necessary for load bearing applications [224].

Hybrid bond networks therefore represent the most viable approach for balancing these conflicting requirements. These networks comprise a percolated covalent network that forms a durable elastic backbone, maintaining the printed structure, and overlaid dynamic covalent or non-covalent subnetworks that provide adjustable viscoelastic stress relaxation and energy dissipation [225]. In light-assisted 3D bioprinting, this equilibrium can be meticulously tuned to accomplish three primary objectives: (1) the photo-responsive covalent network allows on-demand shape retention; (2) the dynamic network relaxation times coincide with the timescales of cellular traction and matrix remodeling, thereby enabling mechano-sensing and remodeling rather than confining cells within a solely elastic environment [223]; and (3) the printed macro-architecture distributes mechanical stresses and provides structural direction for tissue-specific functions.

Therefore, assessments of bond-type toughening methods for bioinks must go beyond simple enhancements in modulus, and instead should be evaluated according to their arrangement within a 3D bioprinting window that fulfills the criteria for printability, toughness, and biocompatibility. Although techniques such as increasing polymer concentrations or substantially loading inorganic filler may produce remarkable mechanical properties, they frequently hinder light penetration, flow characteristics, and cell migration. Covalent-dynamic hybrid systems with a moderate solid content, characterized by reversible sacrificial contacts that relax, preserve a viscoelastic, cell-permissive milieu. These overarching principles connect bond-level chemistry with functional tissue structures to inform the selection and integration of network types for sophisticated bioprinting applications.

5. Fitting photo-crosslinkable tough hydrogels into the 3D bioprinting window

Implementing these design concepts requires the photo-crosslinkable robust hydrogels be precisely tailored to fit within a multidimensional constraint framework. Expanding upon the “biofabrication window” concept, these requirements are known to be optimally characterized as two partially decoupled, yet interrelated windows: a printing stage window dictated by bioink rheology and process-induced stresses [226], and a cultivation stage window regulated by the mechanics and dynamics of the crosslinked construct.

5.1. Rheological window and print fidelity

The principal limitation of robust hydrogels in light-assisted bioprinting is rheology. Extrusion-based printing, structural printability, and viability have conflicting requirements. A high viscosity and yield stress enhance strand stability; however, to preserve the viability of various cell types, shear and extensional stresses must be minimized during extrusion, with less than 100 Pa typically recommended [221,227,228].

In light-assisted extrusion, print fidelity is governed by three rheological parameters: shear thinning viscosity, yield stress, and modulus recovery kinetics, rather than by a singular ideal viscosity. Recent techniques have extended the time window for the photocurable network to develop a shape retaining structure by dissociating flow from the form support [226,229]. Low-viscosity inks can produce high-fidelity filaments using methods such as in-nozzle ionic crosslinking, embedded support bath printing, or rapid post-extrusion photo-crosslinking [227,229,230]. Toughness is closely correlated with printability when solidification is achieved using a photo-network “lock” that overlays a dynamic, dissipative subnetwork. A representative design integrated a dynamic subnetwork (ionic/physical or dynamic covalent) with a light-set network, which allowed intricate structures to be extruded that retained significant deformability and damage tolerance post-curing. The reversible bonds facilitated hysteretic dissipation, and the photo-network maintained structural characteristics and inter-filament connections post-curing (Fig. 2c and d).

In vat photopolymerization (VPP), accuracy specifically refers the geometric fidelity and shape retaining ability of the printed part and is determined by the spatiotemporal development of the photocurable network. Based on formation theory, effective double-bond conversion (EDBC) and related metrics, such as average crosslinking steps (ACS) and density of molecular network (DMN), more accurately forecast the development of a shape-retaining percolated network compared to total conversion [111]. Paradoxically, a slow crosslinking effect occurs, in which diminished light intensity and prolonged exposure enhance the EDBC and DMN by promoting propagation rather than premature termination, thus increasing network uniformity and alleviating internal stress. Projection-based methodologies such as formation-stabilization printing, that use rapid, high intensity exposure to delineate geometry, followed by a gradual, low intensity formation phase, can improve resolution, mechanical integrity, and biocompatibility [111].

5.2. Elasticity–stability window: Transitioning form soft matrices to load-bearing scaffolds

In addition to printing, hydrogels must meet an elasticity–stability threshold that aligns with the target tissue. Native tissue has a broad stiffness spectrum, ranging from approximately 1 kPa in the brain to several GPa in bone [4]. Thus, there is a trade-off in engineered matrices: extremely pliable hydrogels (<10 kPa) are optimal for soft tissue morphogenesis but lack the stability for macroscopic load bearing, whereas highly rigid hydrogels (>1 MPa) maintain form but frequently inhibit cell proliferation, spreading, and migration.

To avoid a frozen microenvironment associated with an excessively high DMN [111], hybrid designs intentionally decouple short-term elasticity (regulated by the covalent backbone) from long term relaxation (dictated by dynamic bonds). Within this design framework, load bearing, cytocompatible hydrogels are typically engineered with the following characteristics: (1) elastic moduli on the order of 10²–10⁴ kPa, a range commonly reported for tough or hybrid hydrogels that balance mechanical support with cellular permissiveness [231]; (2) dissipation mechanisms that increase fracture energy and fatigue resistance [61,193]; and (3) adequate viscoelastic relaxation so that cellular behavior is not restricted.

5.3. Relaxation biological window: matching viscoelasticity and cytocompatible photocuring

Ultimately, hydrogels must perform within a relaxation window in a biological context in which network viscoelasticity and photochemistry effectively facilitate cellular activity. Innovative research has demonstrated that, at a constant modulus, hydrogels with accelerated stress relaxation promote more MSC spreading, proliferation, and osteogenic differentiation relative to simply elastic gels [223].

Dynamic bonds (such as imine, boronic esters, and metal-ligand interactions) in light-printed hydrogels allow the relaxation spectrum to be precisely regulated. Their characteristic periods can be adjusted to align with the minute-to-hour ranges pertinent in cell traction and adhesion turnover. This highlights that relaxation behavior, in addition to the modulus, dictates biological permissibility [224].

Cytocompatible photochemistry is equally essential. Limiting photoinitiator concentrations or using efficient crosslinking techniques in VPP are necessary to minimize radical-induced damage and apoptosis [232]. By integrating cell-compatible photochemistry with dynamic mechanics, perfusable structures and resilient endothelial networks (CD31) can be created within constructs, improving mechanical strength while maintaining cell viability and vascularization potential (Fig. 3a and b) [234].

Fig. 3.

(a) The proposed double network dynamic gelatin hydrogel combines a reversible hydrazone crosslinked dynamic network and a free radical-induced methacrylate crosslinked non-dynamic network. (b) Representative enlarged view of a 3D-printed vascularized multi-layered lattice construct on day 7. Scale bars, 200 μm²³³. Copyright 2025, Elsevier. (c) Schematic of tough and biocompatible alginate-PEGDA-nanoclay composite hydrogels. (d) Viability of embedded hMSCs in alginate-PEGDA-nanoclay composite hydrogels over 7 days. (Inset) Live (green)/dead (red) cell images after 7 days of encapsulation [233]. Copyright 2015, Wiley. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

5.4. Summary: an integrated bioprinting window for photo-crosslinkable tough hydrogels

Sections 5.1–5.3 delineate a cohesive integrated bioprinting window for photo-crosslinkable tough hydrogels. During the printing phase, bioinks must meet specific rheological criteria to ensure high cell viability and printing fidelity. During the culture phase, constructs exist within an elasticity–stability continuum, adequately rigid for load-bearing while avoiding the frozen state. The relaxation biological window bridges these factors, in which a stress relaxation spectrum underpins cellular processes. In this multifaceted design domain, hybrid networks are engineered at the mesoscale that balance printability, toughness, and biocompatibility within a single construct.

6. Light-assisted bioprinting technologies for biomedical applications

6.1. 3D bioprinted tough hydrogels for cartilage regeneration

Light-assisted bioprinting of tough hydrogels has emerged as a promising means of creating scaffolds that provide mechanical support while facilitating cellular activities essential for cartilage regeneration. Native articular cartilage has an elastic modulus of roughly 2000–10,000 kPa and a toughness on the order of 0.01–0.8 MJ/m³ [235].To realize cartilage-relevant toughness in 3D bioprinting, current research is focused on double-network designs, nanocomposite matrices, and bioinspired structural motifs.

Dynamic bioinks are most frequently modified through double-network hydrogels. This often entails the use of primary reversible crosslinking to enhance printability, with the dynamic network subsequently reinforced through secondary crosslinking. In a study by Wang et al. [236], a double-network was used to enhance the mechanical performance of hyaluronic acid-based hydrogels in 3D extrusion bioprinting (Fig. 4a and b). The primary network was formed through dynamic covalent hydrazone bonds between hydrazide and aldehyde-modified hyaluronic acid, imparting the hydrogel with shear-thinning and self-healing properties essential for printability and shape fidelity. A secondary network was introduced by incorporating norbornene functionalized hyaluronic acid, which was subsequently crosslinked via thiol-ene photopolymerization. This system was therefore photo-stiffened and photo-patterned. Using this double network method, the elastic modulus increased by ∼300%, with a ∼70% reduction in hydrogel erosion and high cell viability (>80%) post-printing. The study demonstrated that combining dynamic covalent chemistry with orthogonal photo-crosslinking is a promising means of fabricating mechanically robust, cytocompatible, and spatially tunable scaffolds suitable for cartilage tissue engineering applications. Roh et al. [237] optimized a previously introduced oxidized hyaluronate-glycol chitosan-adipic acid dihydrazide (OHA-GC-ADH) dynamic bioink system with a dynamically crosslinked imine and hydrazone bond primary network. The addition of a secondary network to alginate crosslinked with calcium ions enhanced the mechanical properties and structural stability of the printed bioink, resulting in cartilage constructs with well-defined structures supporting prolonged cell culture.

Fig. 4.

(a, b) Concept and microenvironment of a double network (DN) dynamic hydrogel comprising a permanent covalent framework and a reversible, dynamic network. The dynamic network provides cellular interaction cues that support adhesion and spreading, while the permanent network improves print fidelity and dimensional stability [236]. Copyright 2018, Wiley. (c) Workflow for preparing an NSG hydrogel in which PRNs (nanorods) function as nanoscale struts within the network. The schematic indicates nanorod reorientation and axial deformation under a compressive strain. (d) Compressive mechanical response of NSG hydrogels: representative stress–strain curves and the corresponding toughness values. Statistics: ∗∗∗p = 0.0006, ∗∗∗∗p < 0.0001; n = 3 independent experiments [174]. Copyright 2025, Nature Publishing Group. (e) Synthetic scheme of the 3D SF–GT hydrogel scaffold. (f) An in vivo assessment of joints harvested at 12 and 16 weeks showing regenerated cartilage tissue [153]. Copyright 2021, Elsevier.

By covalently grafting peptide-based rigid nanorods into a GelMA network, DLP nanorod-supported GelMA (NSG) hydrogels converted reversible nanorod deformation into efficient energy dissipation, which increased the compressive strength to 712.7 ± 78.6 kPa and compressive toughness to 60.8 ± 2.2 kJ/m³, approximately 10 fold and 5 fold higher than that of pristine GelMA, respectively, while maintaining photo-induced curability, fatigue resistance, and biocompatibility (Fig. 4c and d) [174]. The compressive strength of the hydrogel was higher than that of other composite hydrogels (∼500 kPa), including graphene oxide [238], carbon nanotubes (CNTs) [239], hydroxyapatite nanowire [240] biphasic calcium phosphate nanoparticles [241], and laponite [242]. Additionally, in vivo, the peptide-based rigid nanorod (PRN)-reinforced constructs promoted osteochondral defect repair and were readily integrated into an additive manufacturing workflow.

Additionally, Trachsel et al. engineered a double network bioink by combining sortase, an enzymatically crosslinked poly (2-ethyl-2-oxazoline) peptide, with ionic alginate. A small amount of cellulose nanofibrils tuned the rheology for extrusion, enabling high fidelity printing of chondrocyte-laden constructs with >90% viability and mechanically reinforced scaffolds for cartilage engineering [225].

In a study by Li et al. [153], a 3D printable silk-gelatin (SF-GT) hydrogel scaffold with a tunable macroporous architecture was developed for cartilage regeneration. The ink was fabricated by blending silk fibroin with tyramine conjugated gelatin and then cross-linked under mild conditions using an HRP/H₂O₂ enzymatic system (Fig. 4e and f). The ink gelated at low temperature and maintained high cell viability. By adjusting printing parameters, scaffolds with different pore sizes and geometries were produced for selective nutrient diffusion and cell infiltration. Two cell seeding methods, a conventional cell suspension and pre-formed mesenchymal stem cell (MSC) aggregates, were compared. Aggregate seeding significantly increased cell retention, cell spatial distribution, and chondrogenic differentiation, and formed hyaline-like cartilage rich in collagen II and glycosaminoglycans. In a rabbit osteochondral defect model, SF-GT scaffolds loaded with MSC aggregates promoted superior cartilage repair at both 12 and 16 weeks, producing neocartilage with native zonal organization and appropriate mechanical properties. These results highlighted that enzymatically cross-linked SF-GT hydrogels seeded with aggregates offer a promising translational platform for engineering high quality cartilage.

These findings highlighted the importance of tailoring the bioink composition to meet both mechanical and biological requirements in cartilage regeneration applications. Collectively, these innovative studies demonstrated the potential of using light-assisted bioprinting to fabricate complex, functional scaffolds that replicate both the structure and biological performance of cartilage, providing new opportunities for clinical cartilage repair and regeneration.

6.2. 3D bioprinted tough hydrogel scaffolds for bone regeneration

Light-assisted bioprinting of tough hydrogels has emerged as a promising means of creating scaffolds that provide mechanical support and facilitate cellular activities essential for bone regeneration. Hydrogels with tunable mechanical properties can be produced by incorporating photo-crosslinkable materials into the bioprinting process, which is crucial for load bearing bone tissue engineering applications. For example, Wang et al. [243]. demonstrated the potential of a novel bioink composed of gelatin methacrylate, alginate methacrylate, and hydroxyapatite to support the long-term cultivation of bioprinted bone organoids. promote multicellular differentiation, and increase the bone repair capability in an in vivo setting. Similarly, Kim et al. [244]. developed a methacrylated silk fibroin-based bioink suitable for DLP bioprinting applications that exhibited bone tissue-like viscoelastic behavior and supported cell proliferation (Fig. 5d–f). This underscores the importance of integrating bioactive components into hydrogel formulations to achieve optimal outcomes in bone regeneration. Additionally, Yang et al. [246]. introduced a Janus periosteum designed to mimic the structure and function of the natural periosteum, which modulated cells to increase osteogenesis and angiogenesis, highlighting the importance of innovative scaffold design in bone repair. Chatterjee's group developed silk fibroin-derived bioinks for DLP-based 3D bioprinting of scaffolds with compressive moduli ranging from ∼20 kPa to ∼96 kPa [130]. These findings highlight the importance of tailoring the bioink composition to meet both mechanical and biological requirements in osteogenic applications. Collectively, these innovations demonstrate the potential of light-assisted bioprinting to fabricate complex, functional scaffolds that replicate both the structure and biological performance of bone, providing new opportunities for clinical bone repair and regeneration.

Fig. 5.

(a) SEM micrographs with corresponding energy dispersion spectroscopy (EDS) elemental maps showing the surface morphology and elemental distribution of GelMA/AlgMA/HAP-printed scaffolds. (b) Compressive stress–strain response of the constructs. (c) Reconstructed micro-CT (μCT) volumes of defects in standard deviation (SD) rats [243]. Copyright 2024, Wiley. (d) Scheme of SF modifications with glycidyl methacrylate (GMA). GMA was covalently grafted onto SF, introducing pendant vinyl groups that acted as UV-crosslinkable sites. (e) Representative compressive stress-strain curve acquired under the stated test conditions. (f) Live/dead assay of NIH/3T3 cells encapsulated within hydrogels for 14 days (live cells, green; dead cells, red); scale bar = 500 μm²⁴⁴. Copyright 2018, Nature Publishing Group. (g) Schematic of a DLP-printable bioink and its use to promote in situ formation of stem cell spheroids. (h) Illustration of the cell distribution within the gels and the post-printing dextran removal process. (i) Additional representative compressive stress-strain profiles of the hydrogels [245]. Copyright 2025, Elsevier. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Dong et al. [247] prepared a nanocomposite hydrogel composed of 15% (w/v) GelMA and 8% (w/v) laponite for facile extrusion-based 3D printing. A highly concentrated and uniformly dispersed laponite suspension was obtained by stirring at low temperature (0°C), which markedly enhanced the hydrogel's rheological properties, degradation stability, and mechanical strength. The resulting ink, with excellent printability, shape fidelity, and cytocompatibility, significantly promoted the proliferation and osteogenic differentiation of bone marrow derived mesenchymal stem cells (BMSCs). These findings demonstrated that GelMA-laponite nanocomposite hydrogels have great promise as a bioink for bone tissue regeneration applications. For example, Cebe et al. [248] introduced nanosilicate particles into methacrylated chitosan (MAC) to augment the compressive strength of 3D printed bone tissue scaffolds, with a maximum compressive strength of approximately 15 MPa, which approached the compressive strength of cancellous bone. Using this reinforced hydrogel, the scaffolds were printed as mesh-like structures, and osteoblast precursor cells were subsequently seeded onto the scaffold. Greater formation of biominerals in the bone tissue was observed after culturing for 21 d in methacrylated chitosan-laponite (MAC-Lp) scaffolds than in methacrylated gelatin-laponite (MAG-Lp) scaffolds.

In light-assisted printing, pore architecture is as important as bulk chemistry. Total porosity, pore size, interconnectivity, surface curvature and spatial gradients govern both perfusion (nutrient and oxygen transport) and the mechanical competence of bone mimetic scaffolds. A case in point is an in situ spheroid formed by an emulsion-templated DLP process, in which a GelMA/dextran bioink concentrated cells during curing, and the subsequent leaching of dextran left smooth, concave microcavities that induced spheroid formation (Fig. 5g–i). This niche enhanced proliferation and the osteogenic and angiogenic potential of dental pulp stem cells (DPSCs) and supported dentin repair while preserving print fidelity [245].

6.3. 3D bioprinted tough hydrogel scaffolds for tendon regeneration

Light-assisted bioprinting of tough hydrogels is rapidly gaining traction as a versatile method for fabricating tendon mimetic scaffolds that recapitulate both the hierarchical mechanics and the dynamic cellular microenvironment of native tendon tissue. By leveraging photo-crosslinkable bioinks with high tensile strength, fatigue resistance, and an anisotropic architecture, recent studies have demonstrated unprecedented control over scaffold geometry and biochemical cues critical parameters for guiding tenogenic differentiation and aligning collagen deposition.

For example, inspired by Chinese ramen, Yao et al. [185], extended their “print-train-crosslink (PTC)” approach to a methacrylated gelatin-silk fibroin (GelMA–SilMA) system (Fig. 6a). After extrusion printing, the scaffold underwent cyclic mechanical stretching in a kosmotropic (NH₄)₂SO₄ solution to induce β-sheet crystallization and molecular alignment, followed by 405-nm photo-crosslinking to lock the oriented structure (Fig. 6b). The resulting scaffold achieved a tensile modulus of ∼6.5 MPa and an ultimate strength of 12.4 MPa (Fig. 6c).

Fig. 6.

(a) Workflow illustrating the fabrication of the PTC hydrogel. (b) Resulting architecture, with high strength and hierarchical, multiscale anisotropy. (c) Representative uniaxial stress–strain response of the tough hydrogel, highlighting its large strain deformability and energy dissipation. (d) Typical images showing the transplantation sites at 2- and 4-week time points. (e) Quantification of centrally nucleated fibers per visual field 4 weeks post-transplantation [185]. Copyright 2025, Springer Nature.

These values aligned with the mid-range of these properties in the human patellar tendon (5–100 MPa) and exhibited viscoelastic J-curve behavior suitable for early-stage load-bearing tendon repair. Notably, macroscopic inspection at harvest showed clearer defect depressions in the untreated hydrogel scaffold, whereas both the control hydrogel and PTC-treated scaffolds displayed more evident volume restoration with only limited residual hydrogel remaining (Fig. 6d). Consistent with this gross recovery, histological outcomes indicated a larger regenerated fiber caliber and a higher fraction of centrally nucleated nascent fibers in the PTC-treated scaffolds, supporting accelerated early stage myogenesis within the defect region (Fig. 6e). Collectively, these advances underscore the transformative potential of light-assisted bioprinting to produce tendon specific scaffolds that integrate high mechanical integrity via mechanically trained DN architectures and spatiotemporal delivery of tenogenic cues. These multimodal hydrogels promise to accelerate functional tendon regeneration while minimizing fibrotic encapsulation, paving the way for patient-specific, load-ready constructs in orthopedic and sports medicine applications.

6.4. 3D-bioprinted tough hydrogel scaffolds for skull regeneration

Light-assisted 3D printing of tough double network hydrogels is also being leveraged to create elastic cranial scaffolds that reconcile high mechanical protection with the compliance required to interface with brain tissue after traumatic injury. Chen et al. extrusion printed a sodium alginate/acrylamide precursor under 365-nm UV to form a covalent poly (acrylamide) network, which was then soaked in CaCl₂ to generate a secondary ionic alginate-Ca²⁺ network to engineer a porous lattice scaffold that coupled MPa level stiffness with large, reversible deformation and efficient nutrient transport [249]. Following Ca²⁺ crosslinking (Fig. 7a), the printed construct exhibited a Young's modulus of 1.07 MPa, an elongation at break of 117.39% (Fig. 7b), and a compressive strength that increased from 0.44 to 4.64 MPa (Fig. 7c), which was adequate to buffer intracranial pressure fluctuations while remaining far more compliant than conventional rigid cranial implants. In a rat cranial defect model, the printed scaffolds matched the defect geometry well during implantation (Fig. 7d), with no infection or mortality observed. Micro-CT analyses indicated minimal self-repair in untreated control rats and incomplete regeneration in rats treated with hydrogels without adequate Ca²⁺, whereas the Ca²⁺-conditioned DN scaffolds promoted mineralized tissue deposition that aligned with the printed struts and promoted bone formation, even in the defect center (Fig. 7e and f), consistent with channel-guided regeneration and osteopromotive Ca²⁺ release. Sears et al. developed an extrusion-printable GelMA/κ-carrageenan/nanosilicate ionic–covalent entanglement ink that produced fully infilled, UV-crosslinked constructs with enhanced mechanical properties (78 ± 17 kPa modulus; 40 kJ m⁻³ toughness) while maintaining adequate compliance for press-fit craniofacial implantation. After conditioning with iP-hMSC-derived ECM, the constructs exhibited over 90% survivability, with high cellular infiltration inside the scaffold, osteogenic differentiation, and mineralization [250].

Fig. 7.

(a) Schematic diagram of the fabrication route of a double network hydrogel scaffold, in which the printed lattice is first photo-crosslinked and then subjected to Ca²⁺ treatment to generate the second network. (b) Representative tensile stress–strain curves of the hydrogels before and after Ca²⁺-induced crosslinking, highlighting the increased strength and elongation after the ionic network was formed. (c) Compressive stress–strain responses of the hydrogels with and without Ca²⁺ crosslinking, showing the markedly improved compressive load-bearing capacity of the double network system. (d) Visualization of calvarial bone regeneration. (e) Coronary micro-CT sections of the surgical site after healing for 4 and 8 weeks (scale bar: 1 mm). (f) Three-dimensional micro-CT renderings of the defect region at 4 and 8 weeks (scale bar: 300 μm) [249]. Copyright 2019, Wiley.

6.5. 3D-bioprinted tough hydrogel scaffolds for musculoskeletal regeneration

Light-assisted bioprinting is increasingly being applied to ligament regeneration, in which the main requirements are printable architectures with anisotropy and crimp-like guidance, with sufficient mechanical stability to endure early loading while maintaining high cytocompatibility. For example, Weng et al. developed a mechanically reinforced and self-healing natural bioink by blending methacrylated collagen peptide (COPMA) with xanthan gum (XG) to form an interpenetrating network in which UV-triggered covalent crosslinking was coupled with dynamic hydrogen bonding (Fig. 8a), enabling rapid photo-fixation while retaining recoverable interactions. This dual-crosslinking structure demonstrated improved compressive mechanics compared with COPMA alone, with increases in the maximum stress to 38.7 ± 3.2 kPa and Young's modulus to 27.4 ± 2.0 kPa (Fig. 8c). The printed hMSC-laden constructs also maintained high cytocompatibility, as evidenced by the predominantly live cells (Fig. 8b) and sustained cellular activity over 28 days, and an interconnected 3D cellular network and high COL-1 deposition under differentiation conditions (Fig. 8d and e) [251]. Focusing on architectural mechanics rather than bulk reinforcement, Lin et al. used a freeform reversible embedding of suspended hydrogel-enabled extrusion printing method to fabricate microfibrous collagen scaffolds that were photo-stabilized by crosslinking in 0.02% riboflavin under 365-nm UV for 30 min. Under a physiologically relevant laminar shear load in a flow bioreactor, compared with straight fibers, the crimp-like waveform fibers showed reduced deformation and a greater load tolerance, with periodontal ligament cells remaining viable and exhibiting greater spreading, adhesion, and viability under shear [252].

Fig. 8.

(a) COPMA/XG hydrogel preparation and crosslinking scheme. (b) Bioprinting and 28-day culture of hMSC-laden COPMA/XG constructs (proliferation vs. differentiation media). (c) Compressive stress–strain curves of COPMA₁₅ and COPMA₁₅/XG_3.5 hydrogels. d) Volumetric rendering of the cellular network in different hydrogels on days 1 and 28 post-culture. Scale bar: 100 μm. e) Percentage of the COL-I fluorescence area within the constructs on days 1 and 28, expressed relative to the total area (n ≥ 3; ns: not significant; ∗p > 0.05, ∗∗p < 0.01, ∗∗∗∗p < 0.0001) [251]. Copyright 2025, Wiley.

6.6. 3D-bioprinted tough hydrogel scaffolds for dental regeneration

Light-assisted extrusion bioprinting is also actively used for periodontal regeneration, with a goal of combining a controllable fiber architecture with rapid photocuring for shape fixation, while maintaining sufficient mechanical support and high cytocompatibility during early remodeling. For example, Wang et al. extrusion printed a dextran/GelMA aqueous two-phase emulsion (ATPE) bioink reinforced with trace phosphorylated cellulose nanofibrils, and subsequently subjected it to photocuring to produce durable microporous constructs (Fig. 9a). The Ca²⁺(X) Mp_ATPE hydrogelfollowing Ca²⁺-assisted crosslinking exhibited a high compressive Young's modulus and maximum stress (higher than other samples tested; 30.5 ± 4.6 kPa, 276.5 ± 52 kPa, respectively) (Fig. 9b). The printed MC3T3-E1 constructs exhibited high viability (>70%) across all samples for a duration of 7 days, suggesting good biocompatibility (Fig. 9c), with cytoskeletal staining indicating spreading during culture [254]. Additionally, Miao et al. extrusion printed a GelMA/sodium alginate bioink containing bioactive glass microspheres, followed by quick post-printing dual curing to stabilize the scaffold [255]. The cell-laden constructs demonstrated excellent cytocompatibility, with live/dead straining indicating high cell viability with a cell population that increased over time. Growth factors (BMP2 or PDGF) loaded in the construct further enhanced proliferation at days 3 and 5 compared with scaffolds without growth factors. In addition to filler-reinforced formulations, ECM-informed photocurable bioinks provide biochemical guidance while maintaining compatibility with living cells during printing and photocuring. Yang et al. developed an ECM using a light-cured periodontal module method, in which cell viability exceeded 85% on day 3 and attained 92% by day 14 post-printing, thus demonstrating substantial tolerance to the printing and light stabilization process (Fig. 9d). ECM incorporated into the alveolar bone module enhanced post-printing survival and sustained greater viability on day 7 (90.11 ± 1.83% compared to 83.71 ± 1.84% for GelMA) [253] (see Fig. 10).

Fig. 9.

(a) Schematic illustration of a bioprinting process using p-ATPE bioink. (b) Compression stress–strain curves of hydrogels after Ca²⁺ crosslinking. (c) Viability of MC3T3-E1 cells in the bioprinted hydrogels on days 1, 3, and 7²⁵³. Copyright 2024, Wiley. (d) Schematic overview of 3D-bioprinted periodontal modules for periodontal regeneration made with GelMA/dECM bioink encapsulating human dental follicle cell (DFCs) [253]. Copyright 2023, Wiley.

Fig. 10.

Key future directions for light-assisted 3D bioprinting of tough hydrogels.

7. Future prospects and challenges

Despite rapid progress in the bond chemistry used in the light-assisted 3D bioprinting of tough hydrogels, mechanically robust, printable, and cytocompatible constructs are challenging to generate because material formulation, photocuring physics, and cell-matrix interactions are coupled across multiple length scales. In this review, we discussed how covalent networks, dynamic covalent bonds, reversible non-covalent interactions, and hybrid energy-dissipating systems can improve hydrogel toughness while maintaining printability and cytocompatibility [11,19,20,193]. Nevertheless, achieving robust mechanical performance in light-assisted bioprinted hydrogels remains challenging because the mechanical performance of the hydrogels are governed by coupled material, optical, and biological processes across multiple length scales. Therefore, future studies should move beyond simply increasing the bulk mechanical strength of tough hydrogels and focus on four specific directions: photocuring heterogeneity into cell-laden bioinks; generating gradient structures through programmable light modulation; achieving a post-printing balance among toughness, porosity, and degradation; and prioritizing translational validation, including scaffold-minimized biofabrication(Fig. 10).

7.1. Cell-laden photocuring heterogeneity and reciprocal cell-matrix interactions

A key challenge of photosensitive cell-laden bioinks is that cells are not passive components. A high cell density can scatter light, alter the oxygen distribution, and disturb radical-mediated polymerization, leading to local undercuring, a heterogeneous crosslinking density, and reduced printing fidelity. For example, high cell density DLP bioprinting often must balance the cell density and printing resolution. Refractive index matching has been used to reduce cell-induced light scattering and improve vascularized tissue fabrication [256]. Deep learning-based optical correction has also been proposed to compensate for cell-induced light scattering during light-based bioprinting [257].

Future work should therefore establish quantitative relationships among cell density, light transport, curing kinetics, local mechanics, and cell fate. In addition to reporting the bulk modulus or live/dead staining results, studies should map the local stiffness and matrix remodeling around cells. Particle-tracking microrheology, Brillouin microscopy, fluorescence force sensors, and 3D traction force microscopy may elucidate the mechanism by which encapsulated cells remodel reversible bonds and dynamic networks after printing. This research focus directly addresses the balance among toughness, printability, and compatibility.

7.2. Programmable light modulation for gradient and interfacial tough hydrogels

Many tough hydrogel bioinks are mechanically reinforced but structurally homogeneous, whereas native tissue often has gradients, anisotropy, and soft-hard interfaces. This mismatch represents another major challenge for future research. A concrete example is the osteochondral interface, in which the cartilage region comprises a hydrated, low-friction, and relatively compliant matrix, while the subchondral bone region has a higher stiffness with mineral-associated support. Simply stacking two hydrogels with different stiffnesses to recapitulate this interface may cause a swelling mismatch, weak stress transfer, and interfacial delamination.

Future studies should use light as a spatial programming tool in addition to a curing trigger. Programmable light exposure is currently used to regulate material composition and bond chemistry and can be extended in the future to local network formation, material heterogeneity, and mechanical gradients. Recent advances in wavelength-selective multi-material 3D printing [258] and lithographic crystallinity regulation [259] further demonstrated that the light wavelength, dose, and exposure sequence can be used to program material composition, network structure, crystallinity, and local mechanical properties. Although these approaches have not yet been fully translated to cell-laden tough hydrogel bioprinting, they provide useful design principles for constructing interfacially integrated and stiffness gradient hydrogel systems. Grayscale DLP can deliver varying local light doses and thereby tune the crosslinking density or material properties within a single vat [260,261]. Volumetric bioprinting and computed axial lithography may further reduce layer-by-layer defects and create complex internal channels or curved interfaces [262]. However, these methods still require better control over oxygen inhibition, radical diffusion, optical attenuation, and cell-induced scattering. Thus, future research should develop predictive models of the light dose and resulting network mechanics by combining absorber tuning, multi-wavelength photochemistry, grayscale compensation, and interfacial tie-layer chemistry to fabricate mechanically integrated gradient constructs.

7.3. Balancing post-printing toughness, porosity, degradation, and fatigue

The third challenge appears after printing. Bioinks and printing methods that increase hydrogel toughness, such as a high polymer concentration, a high crosslinking density, double networks, and sacrificial bonding, often generate dense networks. These dense structures may improve the initial modulus and shape fidelity but restrict nutrient diffusion, oxygen transport, cell migration, vascular ingrowth, and tissue replacement. Conversely, highly porous or channelized structures improve biological permeability but at the cost of mechanical integrity.

Future tough hydrogel design should therefore integrate mechanical reinforcement with hierarchical porosity and controlled degradation. For example, bone or osteochondral repair constructs may require a reinforced region to provide early load support along with interconnected pores or perfusable channels to support vascular invasion and new tissue formation. High cell density tissue manufactured with embedded vascular channels has already shown the importance of coupling a dense cellular organization with a perfusable architecture [263]. Possible solutions for light-assisted tough hydrogels include phase-separated networks, sacrificial templates, microgel-based bioinks, embedded channels, and topology-optimized pore architectures.

Degradation kinetics should also be considered in the long-term compatibility of a scaffold. A scaffold that degrades too slowly may block tissue replacement, whereas a rapidly degrading scaffold may lose mechanical support before achieving functional regeneration. Therefore, in addition to the initial compressive modulus, future studies should also evaluate fatigue-crack growth, cyclic loading stability, swelling-induced softening, degradation-associated mechanical loss, and interfacial fatigue at 37°C under hydrated conditions in the presence of -enzymes [222,264].

7.4. Next-generation biofabrication and translational validation

Future studies should also integrate material design with clinically relevant manufacturing and evaluation. AI and machine learning can help optimize bioink formulations, photoinitiator concentrations, absorber content, and exposure parameters. However, if models are trained only on characteristics such as strand width, surface roughness, or printing fidelity, they will not predict other critical properties such as fracture energy, fatigue threshold, stress relaxation, degradation behavior, or cell function. AI-assisted bioprinting therefore requires physics-informed constraints, curated datasets, real-time optical and mechanical feedback, and destructive wet-state mechanical validation [257].

In addition to conventional bioprinting using hydrogel-based bioinks, scaffold-free and scaffold-minimized direct cell printing are emerging as alternative methods. Early cell-only bioink systems showed that living cells can be printed in a photocrosslinkable supporting bath without being carried by a conventional biomaterial ink [265]. More recently, biomaterial-minimalistic photoactivated bioprinting using acrylate-modified cell membranes allowed living cells at near-physiological densities, up to approximately 10⁻⁹ cells mL⁻¹, to serve directly as DLP-printable bioinks [42]. Spheroid-based methods have also supported high cell density tissue fabrication through cell aggregate fusion and scalable positioning [266,267].

However, scaffold-free designs usually have limited initial mechanical stability and are unlikely to replace tough hydrogel bioinks for cartilage, bone, tendon, or osteochondral repair in the near term. A more realistic direction is hybrid biofabrication, in which cell-dense units provide biological function and induce endogenous ECM formation, while temporary tough hydrogel frameworks provide early mechanical support, spatial guidance, and controlled degradation. To facilitate clinical translation, future studies should establish standardized protocols to evaluate printing fidelity, pore stability, local modulus mapping, fatigue resistance, degradation products, immune responses, vascularization, and long-term integration in relevant animal models [222,264].

In summary, the future of light-assisted 3D-bioprinted tough hydrogels extends beyond a goal of maximizing toughness. The key research directions will focus on bond chemistry, photocuring physics, structural architecture, cellular remodeling, and standardized validation to achieve constructs that are printable, mechanically durable, biologically permissive, and clinically translatable.

8. Conclusion

Light-assisted bioprinting of tough hydrogels combines rapid photopolymerization with engineered network topologies to resolve the classical trade-off among mechanical robustness, printability, and cytocompatibility. Photoinitiated cross-linking supplies the spatiotemporal precision required for stereolithography, digital light processing, and continuous liquid interface systems, while hierarchical toughening motifs, such as double networks and ideal networks, dynamic covalent and reversible physical bonds, nanocomposite reinforcement, and field-guided alignment create effective energy dissipation pathways that enhance fracture resistance without compromising printing resolution. Through these synergistic molecular chemistries and processing methods, photocurable hydrogels can be tailored to replicate demanding tissue environments, from calcium phosphate-reinforced composites with bone-like stiffness and porosity to aligned fibrous constructs that reproduce tendon-like anisotropy, all under cytocompatible light exposure. Looking forward, the integration of artificial intelligence-assisted design of materials and processes with four-dimensional architectures capable of dynamic adaptation, along with embedded sensing and self-healing functionalities, will be crucial for advancing tough photocurable hydrogels form laboratory demonstrations to clinically approved patient-specific implants and next generation soft robotic systems.

CRediT authorship contribution statement

Haiyan Yin: Writing – original draft. Meimei Xiong: Data curation. Sen Yang: Validation. Liang Chen: Validation. Yanjun Wang: Validation. Wei Cui: Validation. Feng Luo: Validation. Changjin Huang: Supervision. Jie Tao: Supervision. Rui Liu: Supervision.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

No data was used for the research described in the article.