Progresses and Challenges in Additive Manufacturing of Bulk Metallic Glasses

Abstract

Bulk metallic glasses (BMGs) are a type of amorphous metal with a high degree of mechanical strength, elasticity and corrosion resistance, properties that are highly influenced by composition and the processing of the material. BMGs can be applied in advanced engineering fields, such as aerospace, biomedical, MEMS, and industrial applications. Additive manufacturing (AM) is revolutionary in producing intricate BMG parts whilst maintaining the amorphous structure. The current review critically evaluates the recent development in AM of BMGs, such as the development of selective laser melting, electron beam melting, and directed energy deposition, and new classes of hybrid strategies. Enhancements in dimensional accuracy, amorphous retention, microstructural tailoring and functional performance are emphasized along with computational and real-time process optimization strategies to improve overall manufacturing efficiency and material quality. Subsequently, the challenges that still exist are addressed in the review, including crystallization during printing, the buildup of stress, printable thickness, complicated geometries, oxidation, contamination, and heterogeneous amorphous fractions. Lastly, multi-material printing, scalable AM approaches, and AI-assisted design solutions are key features of future perspectives to solve existing restrictions. The review provides an excellent guidance for the researcher and engineer interested in advancing additive manufacturing of BMGs with the best structure–property relations.

1. Introduction

Bulk metallic glasses (BMGs) are a unique category of metallic alloys that form into an amorphous state as they cool much faster out of a molten condition. BMGs do not have any long-range atomic order, nor do they have any grain boundaries and dislocations, compared to conventional crystalline metals, which leads to a combination of mechanical and functional properties [1,2]. They consist of such clusters and motifs as short-range order (SRO) and medium-range order (MRO) that have a powerful impact on deformation mechanisms, crystallization behavior, liquid fragility, and structural stability due to their atomic structure. Even minor changes in the topology of atoms may produce a major influence on ductile behavior, fracture behavior and thermal stability [3,4,5].

Figure 1 presents the thermodynamic and kinetic forces of glass formation and compares pure metals, simple alloys, and bulk metallic glasses (BMGs). The enthalpy-temperature diagram (Figure 1a) shows the supercooled liquid between the temperatures of liquidus (T_l) and of glass transition (T_g). To maintain an amorphous state, crystallization should be prevented.

Figure 1. Transition from supercooled to glassy. (a) A diagram of enthalpy and volume versus temperature embodies the nature of the process of the formation of glasses. When cooled rapidly in a comparatively free medium of impurities, a pure liquid or mixture may be supercooled, that is, cooled substantially below its normal melting or liquidus temperature (T_L), without freezing. (The liquidus temperature of a mixture is like the melting temperature of a pure liquid: it is the highest temperature at which a crystalline phase can exist in a stable state). At T_G, the density of packing is so large that the supercooled liquid will freeze to an amorphous solid, or a glass. (b) A time–temperature–transformation (TTT) diagram further demonstrates the duration which is required to cool off the liquid of T_L to T_G; before it appears as crystals is dependent on the temperature and composition of the liquid. In this case, the orange, red, and green curves indicate the nominal crystallization times of pure metals and simple alloys and the more complicated ones that constitute BMGs. To create a glass, a liquid should be allowed to cool sufficiently fast so that its cooling curve never crosses its crystallization curve; in the case of a pure metal, the R₁ cooling curve would lead to crystallization, but the R₂ cooling curve would produce a glass. BMG critical cooling rate may be almost 12 orders lower than that of pure metal [6]. (Reproduced from Schroers, J. Bulk Metallic Glasses. Phys. Today 2013, 66, 32–37, with the permission of the American Institute of Physics).

According to the time–temperature–transformation diagram shown in Figure 1b, the BMG-forming systems have much slower crystallization kinetics than pure metals and simple alloys and thus can be vitrified at much lower critical cooling rates. It is this kinetic stability which contributes to the high glass-forming capability and bulk processability of multicomponent metallic glasses [6].

The thermodynamic properties of characteristic transition temperatures such as glass transition temperature (T_g), crystallization onset temperature (T_x), and liquidus temperature (T_l) determine the glass-forming ability (GFA) of BMGs. Different thermodynamic indicators, including the lower glass transition temperature (T_rg), the supercooled liquid volume (ΔT_x), and combined criteria suggested in the literature, have been established to determine GFA and amorphous stability. However, these single-parameter descriptors often fail to reliably estimate the maximum attainable amorphous thickness (D_max) or the processability of newly developed alloy systems [3,7,8,9]. As a result, data-driven modeling approaches integrated with thermodynamic analysis have recently emerged to improve alloy design and GFA prediction [10,11,12].

BMGs have drawn significant interest based on their excellent properties such as high hardness, high yield strength, high wear and corrosion resistance, and in some of their compositions, soft magnetic behavior and high elastic energy storage capacity. These benefits have their own limitations though. BMGs are susceptible to crystallization during heating or slow cooling and the vast majority of monolithic assemblies have limited tensile ductility causing catastrophic tensile brittle failure when tensile loads are applied [2,13]. In addition, conventional manufacturing schemes such as casting and intensive post-machining may create flaws on the face, residual stress, and localized thermal treatment that promote devitrification and worsen performance [14].

In this respect, additive manufacturing (AM) has emerged as a paradigm shift in processing strategy for BMGs. AM processes with laser- and electron-beam technology yield very high cooling rates and steep thermal gradients, which are beneficial to retain the amorphous structure. Its advantage over traditional fabrication techniques is that AM enables the fabrication of complex, application-specific geometries because of near net-shape fabrication with less material wastage and minimal post-processing. Architected and functionally graded structures, including energy-absorbing lattices, compliant mechanisms, and lightweight high-stiffness components previously difficult to make using traditional methods are possible due to the inherent strength and hardness of BMGs and AM geometric freedom [9,15].

Despite benefiting due to the supercooled liquid space above T_g to deform BMGs via viscous flow like a polymer, thermoplastic forming (TPF) cannot scale geometries or provide localized design freedom as AM does. In contrast, laser powder bed fusion (LPBF) and laser-directed energy deposition (LDED) have proven to have significant potential in making BMG components with controlled microstructures and tailored properties [16,17,18]. Further improvement of amorphous retention and dimensional precision has also been achieved through state-of-the-art beam shaping, scan strategy optimization, transient thermal control and atmosphere regulation [19,20].

However, residual stresses, devitrification, formation of pores, compositional segregation, oxidation, and dissimilar amorphous fractions remain the major problems in additive manufacturing of BMGs. Devitrification is not usually preferred in additive manufacturing of bulk metallic glasses; even partial crystallization may affect the structural uniformity and the functional performance of the amorphous phase. Such flaws may undermine mechanical integrity and magnetic performance and reliability. Thus, AM processing aims not only at the creation of glassy components, but at the actualization of design-based applications, whereby the amorphous phase is retained at the same time in which the structural and functional performance is enhanced [20,21].

This review provides a comprehensive and critically integrated assessment of recent advances and unresolved challenges in the additive manufacturing of BMGs. It systematically bridges fundamental glass science with AM processing by examining the interdependence between alloy composition, glass-forming ability, crystallization kinetics, and laser- or electron-beam-induced thermal histories. Emphasis is placed on process–structure–property relationships, including the effects of energy density, scan strategy, layer thickness, and thermal accumulation on amorphous retention, defect evolution, and mechanical performance. The review also assesses the relative strengths of LPBF, EBM, and DED methods, points at new hybrid and multi-material methods, and addresses post-processing methods of structural relaxation and stability with respect to enhancement. Lastly, it presents future directions for predictive process window design, manufacturing with the help of digital twins, sustainability, and the large-scale AM fabrication of BMG parts.

2. Fundamentals of Bulk Metallic Glasses

2.1. Glass-Forming Ability (GFA)

Glass-forming ability, commonly abbreviated as GFA, is the property of a metallic alloy that allows it to fight against crystallization during solidification and instead form an amorphous or glassy film. An alloy with high GFA can be made to form glass at quite low cooling rates, whilst a poor glass former requires very rapid quenching, otherwise the alloy forms crystals. The alloy composition has a significant effect on the GFA, since not only the atomic interactions, thermal stability, and local structural arrangements within the melt but also what elements are where and in what amounts all depend on the composition [12].

Thermodynamically, alloys of three or more elements composed of a large atomic size difference and whose negative heats of mixing are strong are likely to have a higher GFA. The greater mismatch in atomic size complicates the ability of atoms to crystallize into an ordered crystal structure, and negative enthalpies of mixing imply that atoms do not tend to separate, but to be mixed closely, stabilizing the disordered phase. These concepts are emphasized by empirical guidelines put forward by Takeuchi and Inoue [22] on the formation of bulk metallic glasses which propose that good glass-forming alloys are multicomponent systems, that the difference in atomic sizes is at least 12%, and that the alloy has negative heats of mixing between its constituent elements. Once these conditions are satisfied, the liquid phase becomes thermodynamically stable to crystallization and this aids the alloy to remain amorphous during cooling [22].

One more significant manner of depicting GFA is through referral to the thermal parameters representing a way the glass responds to heating or cooling. Alloys with a high reduced glass transition temperature ratio to melting point tend to easily form glass since liquid can exist at a large temperature interval before crystallization begins. The fact that there is a large, supercooled liquid area, which is the difference between the temperature of the glass transition and the temperature of crystallization, is also an indicator of good GFA as it indicates that the liquid is able to endure longer without crystallizing. On the same note, decreasing liquidus temperature or forming a deep eutectic mixture is likely to increase GFA, as either decreases the force of crystallization [23,24].

Although thermodynamics is important in presenting the general energy picture, the internal atomic composition of the alloy is also important in determining the ease with which a metallic glass may form. A metallic glass does not have a totally random structure. Rather, it has short-range order; that is, every atom has a clear collection of closest neighbors, and medium-range order, which is expanded by a few spacings between atoms. This local structural formation highly relies on the alloy composition. Two kinds of local atomic clusters are usually present in Zr-Cu-Al bulk metallic glasses. One of them is an icosahedral-like cluster with fivefold symmetry which is not easy to fit into a crystalline lattice. The other is a crystal-like cluster which has fourfold or sixfold symmetries like those in crystal structures [25,26,27].

Another role is played by composition which determines the density and type of atomic packing. It was discovered that addition of alloying components like aluminum enhances the packing density as well as stabilizes the icosahedral order of Zr-Cu-based glasses. This increases the less competitive nature of the respective crystalline phases and increases the thermal stability of the amorphous phase. A lower liquidus temperature composition that forms a deep eutectic mixture is an additional enhancement to GFA because the metastable liquid can be kept longer, and the crystallization driving force decreases [14,28].

Figure 2 illustrates the basic structural and thermodynamic properties that define bulk metallic glass formation. Figure 2a presents structural schematic diagrams of single-element amorphous metal, amorphous alloy, and high-entropy amorphous alloy. It schematically shows that in an amorphous alloy, the atomic packing is disordered and that there is no long-range order despite dense atomic packing, a property of metallic glasses that cannot be found in a crystalline structure [25]. Figure 2b provides a schematic time–temperature–transformation (TTT) diagram which determines the processing windows used in LPBF-processed Fe-based bulk metallic glasses [19]. TTT shows how crystallization may be prevented through maintaining a sufficiently high cooling rate to suppress nucleation, allowing the amorphous form to be retained [9,15,19]. A combination of these structural and kinetic factors identifies the glass-forming capacity and stability of bulk metallic glasses.

Figure 2. Fundamental structural and thermodynamic characteristics of bulk metallic glasses. (a) Schematic representation of the disordered atomic packing in an amorphous alloy compared to other amorphous systems with varying elemental compositional distribution [25] (Reproduced from Han et al., Interdisciplinary Materials (2024), 3, 480–491, licensed under CC BY 4.0). (b) Schematic time–temperature–transformation (TTT) diagram showing processing windows for LPBF-processed Fe-based BMGs [19] (Reproduced from Ren et al., Materials & Design (2025), 252, 113750, licensed under CC BY 4.0).

2.2. Key Thermodynamic and Kinetic Parameters

Characteristic transition temperatures (i.e., onset of crystallization temperature (T_x), glass transition temperature (T_g), and liquidus temperature (T_l)) control the GFA of BMGs thermodynamically. These temperatures are intrinsically dependent on alloy composition and define the thermal stability of both the supercooled liquid and the amorphous phase. Establishing quantitative correlations between these characteristic temperatures and GFA has therefore been a central focus in metallic glass research [29,30].

Indicatively, Lu and Liu [31] introduced the following parameter as an indicator of glass-forming ability:

$$\gamma = {\displaystyle \frac{T_g}{(T_g+T_l)}},$$

where higher values of γ generally correspond to improved resistance to crystallization.

Subsequently, Tripathi et al. [32] proposed a more generalized criterion based on thermodynamic considerations and genetic programming:

$$G_p={\displaystyle \frac{T_g(T_x-T_g)}{{(T_l-T_x)}^2}}.$$

This parameter incorporates both the supercooled liquid region and the resistance to crystallization relative to the liquidus temperature, thereby accounting for liquid-phase stability and glassy-phase persistence simultaneously.

Simpler metrics are also widely employed, including:

$${\mathsf{\Delta }T}_x=\left(T_x{-T}_g\right),$$

and

$$T_{rg}={\displaystyle \frac{T_g}{T_l}},$$

where ΔT_x represents the supercooled liquid region and T_rg denotes the reduced glass transition temperature. Larger values of these parameters are often associated with improved GFA.

Other empirical measurements are the critical cooling rate (R_c) and critical casting diameter (d_c), that measures the lowest possible cooling rate or largest possible amorphous thickness that can be generated to inhibit crystallization, respectively [33].

Nevertheless, none of these parameters are a universally predictive measure of GFA in all alloy systems. Indicatively, ${\mathsf{\Delta }T}_x$ is a measure of thermal stability of the supercooled liquid, but is not explicit as to the liquid-state kinetics, whereas T_rg is a measure of the ease of vitrification without making any explicit reference to the crystallization resistance upon heating or cyclic thermal exposure. The metrics must therefore be adopted as comparative measures and not conclusive predictors and must be supplemented by structural, kinetic and processing-related considerations, especially in additive manufacturing [33,34].

Although parameters such as ΔT_x, T_rg, γ, and Gp provide valuable thermodynamic indicators of glass-forming ability, their practical significance becomes clearer when evaluated in the context of alloy-specific crystallization pathways. In additive manufacturing conditions, where rapid solidification is to be succeeded by repeated thermal cycling, the interaction between inherent thermal stability and kinetic reaction defines the possibility of saving the amorphous structure.

2.3. Crystallization Physics and Thermal–Kinetic Mechanisms

The formation of an amorphous structure in BMGs during additive manufacturing is governed by the competition between atomic diffusion, nucleation, and crystal growth under rapid and highly transient thermal conditions. Vitrification requires suppression of crystallization by exceeding the critical cooling rate, R_c, which depends strongly on alloy composition and GFA [35,36].

2.3.1. Crystallization Kinetics Modeling

Devitrification is modeled using the Johnson–Mehl–Avrami–Kolmogorov (JMAK) equation for non-isothermal conditions, adapted via Nakamura extension for AM thermal cycling [37,38,39]:

$$X_c(T,t)=1-e^{{\left[-{\int }_0^{t}k\left(T\left(\tau \right)\right)d\tau \right]}^n};k(T)=k_0{\mathrm{e}}^{(-{\displaystyle \frac{E_a}{RT}})},$$

where $X_c$ is crystalline fraction, $k_0$ is pre-exponential factor, $E_a$ is activation energy (~300–500 kJ/mol for Zr-BMGs), $n$ is Avrami exponent (2–4 for heat-affected regions (HAZ) nucleation), and the integral captures cumulative reheating exposure [24,39].

2.3.2. Nucleation Rate and Growth Kinetics

From a physical metallurgy perspective, crystallization proceeds through coupled nucleation and growth kinetics [40]. The nucleation rate (I) depends on the thermodynamic driving force and atomic mobility and can be expressed as [38,41]:

$$I=I_0{e}^{(-{\displaystyle \frac{{\mathsf{\Delta }\mathrm{G}}^{*}}{kT}})}.$$

Crystal growth is diffusion-controlled and follows [42,43]:

$$U=U_0{e}^{(-{\displaystyle \frac{Q}{RT}})},$$

where ΔG^* is the nucleation barrier, Q is the activation energy for atomic diffusion, k is Boltzmann’s constant, R is the gas constant, and T is temperature. Under rapid solidification, atomic mobility is significantly reduced, suppressing both nucleation and growth. However, under AM conditions, repeated thermal cycling can locally enhance these processes, particularly within HAZs [41,43,44].

2.3.3. Energy Density and Thermal History Metrics

Process parameters are typically expressed through energy density variants [45,46], such as:

Linear energy density

$$E_L={\displaystyle \frac{P}{v}},$$

Areal energy density

$$E_A={\displaystyle \frac{P}{vh}},$$

And volumetric energy density

$${\mathrm{E}}_{\mathrm{v}}={\displaystyle \frac{\mathrm{P}}{\mathrm{v}\mathrm{h}\mathrm{t}}},$$

where $P$ is laser/electron beam power, $\mathrm{v}$ is scan speed, $h$ is hatch spacing, and $t$ is layer thickness.

2.4. Crystallization and Devitrification Mechanisms in Different BMG Systems

While thermal indicators such as ΔT_x and T_rg quantify intrinsic stability, the actual crystallization behavior of BMGs depends on system-specific nucleation kinetics and sensitivity to reheating. Devitrification in additive manufacturing commonly starts in the HAZ during repeated thermal exposure, as opposed to when it undergoes initial solidification [47]. Table 1 summarizes the comparative thermal stability trends and dominant crystallization mechanisms across representative BMG families.

Table 1. Thermal stability indicators, crystallization behavior, and AM relevance of major BMG systems [9,16,48,49,50,51,52,53,54,55,56,57].

Alloy System	Thermal Stability Trend (ΔTx, Trg)	Dominant Devitrification Mechanism	Sensitivity to Reheating	AM-Relevant Implication
Zr-based	Large ΔTx; moderate–high Trg	Intermetallic nucleation in HAZ; cooling-rate-dependent growth	Moderate	Relatively stable; reheating control required
Fe-based	Smaller ΔTx; lower Trg	Rapid competing phase nucleation	High	Narrow AM processing window
Cu-based	Moderate ΔTx; structure-sensitive Trg	Interface-controlled B2 CuZr precipitation	Moderate	Kinetically sensitive to cooling rate
Ti-based	Moderate ΔTx; tunable via Zr enrichment	Composition-dependent nucleation	Moderate	Lightweight option; thermal management critical
Pd-based	Large ΔTx; high Trg	Delayed nucleation; separated crystallization events	Low–Moderate	High intrinsic stability; cost-constrained
High-entropy BMGs	Emerging; entropy-enhanced stability	Diffusion-suppressed nucleation	Under investigation	Promising tolerance to thermal cycling

It is observed that devitrification that represents the principal microstructural instability limiting additive manufacturing of BMGs can, in both cases, be instigated by either cooling (when the actual cooling rate is below the critical cooling rate) or by subsequent thermal exposure (re-heating, annealing or layer deposition) when crystallization kinetics can be enhanced by structural relaxation, presence of quenched-in nuclei, or chemical heterogeneity [21,58].

It is also observed from the comparison that alloy systems with larger supercooled liquid regions (ΔT_x) and higher reduced glass transition temperatures (T_rg) generally exhibit greater resistance to crystallization under AM-relevant thermal cycles. Nonetheless, intrinsic thermodynamic stability is not enough, since reheating and local thermal accumulation are equally important. Therefore, additive manufacturing of alloys requires a compromise between high GFA indicators and kinetic resistance to cyclic devitrification [59,60,61].

2.5. Comparison of BMGs with Crystalline Metals Regarding AM Processability

Crystalline metals and BMGs have radically different processing sensitivities in AM. The rate of cooling is generally more tolerated by crystalline alloys, since the stability of its phases and its microstructural development, including its grain growth and solidification morphology, is well known and can be controlled. On the contrary, BMGs need cooling rates high enough to inhibit crystallization and maintain the amorphous structure. This means that the processing range of BMGs is much less since volumetric energy input should be optimally balanced to be able to achieve complete densification without creating devitrification [16,61].

The two categories of materials are further distinguished through thermal cycling in the deposition at the layer-by-layer level. Although reheating in crystalline metals is mainly responsible for the change in grain structure and the evolution of residual stress, repeated exposures to heat in BMGs have the potential to cause devitrification, especially in the heat-affected areas. Thus, to achieve successful AM of BMGs, not only the conditions of initial solidification need to be tightly controlled, but also the cumulative thermal condition during the build [62].

Defect sensitivity is also significantly different. Even though porosity, absence of fusion, and cracks are harmful in any metallic system, BMGs are especially susceptible because they do not have dislocation-mediated plasticity. The smallest crystallized fractions or porosity can be stress concentrators, and can therefore localize shear-bands and lead to premature failure. However, AM also presents its own major benefit to BMGs, namely the fact that it allows almost net-shape fabrication and very large scale components, compared to traditional casting processes that are constrained by severe cooling rate limits [47,63,64].

The processability of BMGs in additive manufacturing differs fundamentally from that of conventional crystalline metals due to the requirement to suppress crystallization and maintain amorphous stability under rapid solidification and repeated thermal cycling. However, the relative suitability of crystalline metals and BMGs in additive manufacturing is application dependent. Table 2 highlights the governing differences and clarifies how each material class performs under specific AM constraints.

Table 2. Comparative framework for evaluating crystalline metals and BMGs in additive manufacturing.

Aspect	Crystalline Metals	Bulk Metallic Glasses	AM Processing Considerations	Practical Interpretation
Cooling Rate Requirement [61,65]	Moderate cooling sufficient; microstructure evolves via grain growth	High cooling rate required to suppress crystallization	Cooling must exceed critical cooling rate (Rc)	BMGs require tighter thermal control; crystalline alloys are more process-forgiving
Processing Window [16,66]	Governed mainly by melt pool stability and porosity control	Constrained by both densification and crystallization avoidance	Narrower energy density window for BMGs	Parameter optimization is more sensitive for BMGs
Thermal Cycling Effects [62,67]	Influences grain coarsening and texture	Reheating may trigger devitrification	Full thermal history must be managed for BMGs	BMG builds are more sensitive to layer-by-layer reheating
Defect Sensitivity [15,68]	Dislocation of plasticity provides tolerance to defects	Shear-band dominated failure; defects act as stress concentrators	Stricter defect mitigation required	BMG reliability strongly depends on defect minimization
Geometry and Scale Limitation [47,69]	No intrinsic thickness limit from cooling rate	Casting limited by critical diameter; AM enables localized vitrification	AM can overcome size constraints for BMGs	BMGs benefit significantly from AM’s localized rapid solidification
Failure Mechanism [63,64]	Ductile fracture via plastic deformation	Brittle shear-band fracture	Structural homogeneity critical	BMGs offer higher strength but lower damage tolerance

Thus, although AM poses an extraordinarily promising path to the production of bulk metallic glasses with complicated structures and larger scale, the special demands of the formation and stability of BMGs introduce even more severe limitations than those faced by crystalline metals. Alloy GFA, process cooling rate, effects of thermal cycling, and control of defects should be well controlled in order to achieve success.

3. Additive Manufacturing Techniques for BMGs

The additive manufacturing of BMGs is founded on the various process modalities each possessing varying thermal histories, geometrical capabilities, and quality parameters. Additive manufacturing of BMGs primarily involves laser powder bed fusion (LPBF), electron beam melting (EBM), and directed energy deposition (DED), which permit the local melting and solidification needed to retain the amorphous phase. Figure 3 presents AM techniques of bulk metallic glasses.

Figure 3. Diagrammatic representation of several additive manufacturing methods for bulk metallic glasses: (a) LPBF [50], (b) EBM [70], and (c) DED [71].

Nevertheless, maintaining an entirely glassy structure requires close control of the cooling rate, energy density and cyclic reheating interactions because inadequate quenching or thermal storage may easily initiate crystallization and structural heterogeneity [49,70].

Regarding bulk metallic glass fabrication, LPBF is typically able to achieve the highest cooling rates and hence the highest possible amorphous phase retention and high dimensional accuracy, albeit with high residual stress. EBM offers lower thermal gradients at vacuum and preheating conditions, yet it may be poorly able to cool to fully vitrify. DED has the largest melt pools, slowest cooling, and repeated reheating, making it the most challenging to maintain amorphous stability; therefore, it is best served by strict control of energy input and interlayer temperature [72]. All of them are described, specifying the salient features and advantages and disadvantages of using them in BMGs, in this section.

3.1. Laser Powder Bed Fusion (LPBF)

Laser powder bed fusion (LPBF), also commonly referred to as selective laser melting (SLM), is an additive method of making metal objects in which metal layers are deposited repeatedly on top of each other, thereby creating metallic parts of near-net density. This method is especially important for BMGs as it can attain very high cooling rates of the order of 10⁶ to 10⁸ K/s, which is necessary to inhibit crystallization and maintain an amorphous structure through the solidifying process [52].

The glass-forming capability of alloys applied in LPBF is a decisive variable as to whether amorphous structures can be maintained following rapid solidification. The GFA in metallic glasses like Zr-, Cu-, Ni-, and Nb-based alloys is extremely sensitive to the chemical composition, as it can alter the melting behavior, thermal diffusivity, and resistance of the alloy to nucleation [73]. High-GFA alloys such as Zr_59.3Cu_28.8Nb_1.5Al_10.4 (available commercially as AMZ4) are highly thermodynamically stable and slow to crystallize, and so are good candidates to be processed by LPBF [74].

In the LPBF procedure, the target particles are melted selectively by a focused laser beam in an agreed pattern using a powder mass. New layers are deposited on top of the old one and the intersections of the melted pools produce local heating and cooling of the material. These heat variations affect the local atomic structure, which consequently impacts the amorphous structure as well as the mechanical properties across the build height. Indicatively, Best et al. [54] found that a Zr-based BMG that had undergone LPBF exhibited a progressive change in structure and hardness between the bottom and the top of the build. The lower section directly in contact with the build plate cooled faster as it dissipated more heat and thus became less hard and revitalized in the amorphous state. The middle region, on the other hand, cooled slower and underwent more cyclic reheating to develop a more relaxed and hard glass structure. The most elevated area, which received weaker heating, was of intermediate nature.

This difference in cooling rates and thermal history is a direct indication of the interactions between GFA and composition during the LPBF process. Strongly negative heat of mixing and extensive atomic size incompatibility (characteristic of Zr-Cu-Al-Nb) alloys naturally resist the nucleation of crystal formation, and amorphic survival also relies on heat dissipation capability during laser scanning. Poor cooling or local over-reheating may result in partial crystallization in the HAZs, which introduces porosity defects, embrittlement, or phase separation [75].

Laser power, scan-speed, hatch spacing, and layer-thickness parameters of the process must be finely tuned so as to provide high amorphous retention. As an example, too high energy density can cause local overheating and crystallization, whereas too low energy can cause fusion defects. It has been demonstrated that these parameters are essential to maintain the amorphous fraction, the homogenous microstructure, and the mechanical stability of the BMGs produced by the LPBF at a high level [76]. Also, melt pool geometry and thermal gradients can be controlled by shaping laser beam patterns and scan path optimization (rotating stripe patterns and modified profiles of laser spots), which further improve amorphous structure retention and eliminates residual stresses.

3.2. Electron Beam Melting (EBM)

Electron beam melting (EBM) is an upgraded and developed powder bed fusion technology which involves a sharp, focused, high-energy stream of electrons to selectively fuse and crust interacting layers of metal powder under elevated vacuum conditions. An EBM environment in the vacuum chamber has several significant advantages in BMG processing. Oxidation and contamination in melting and solidification is reduced by the vacuum, which is especially beneficial in relation to reactive or oxygen-sensitive alloys such as zirconium-, titanium- or magnesium-based. They are also subject to oxidation in air at high temperatures, and therefore the vacuum condition of EBM goes a long way in enhancing chemical stability in the fabrication process [28].

The other benefit of EBM is that it allows the uniform preheating of the powder bed before melting. The high preheat temperatures, which are usually in the form of hundreds of degrees Celsius, diminish the temperature gradients between the melt pool and the material around the melt pool. This reduces the residual stress and alleviates warping or cracking effects that are usually evident in high strength and brittle materials, such as BMGs. A low thermal gradient also aids in the release of internal stresses that could otherwise impact the air during cooling. All these increase the dimensional stability and mechanical integrity of the part printed [52].

Nonetheless, as much as EBM has these advantages, the additive manufacturing of amorphous alloys has some challenges associated with it. Its rather low cooling rate relative to LPBF is one of the most significant constraints. Ordinary EBM cooling rates are in the order of 10³–10⁵ K/s, which is not usually slow enough to avoid crystallization in most metallic glass systems that need higher quenching rates than 10⁶ K/s [28]. Larger and deeper melt pools are also prone to be produced by the electron beam due to the increased beam energy and spot size. Such broader melt pools increase the volume and duration of the molten state, which facilitates the diffusion of atoms and may cause subsidiary crystallization in the HAZs. Consequently, BMGs which are processed via EBM are more prone to the emergence of nanocrystals or crystalline phases within the amorphous matrix in cases where the process parameters are not strictly regulated [77].

EBM is also less likely to have a high effective cooling rate than LPBF, and can have large melt pools that may cause greater risk of crystal formation in the heat-affected areas in BMGs. Therefore, to apply EBM to BMGs, the process window definition must be conducted with care in order to maintain the amorphous structure [47].

3.3. Directed Energy Deposition (DED)

Directed energy deposition (DED) is an additive manufacturing technique where metal powder or wire is passed through a focused energy beam (laser or electron) which allows layer-by-layer deposition. In the case of BMGs, DED especially allows a high level of control over the composition: as a result, it is therefore possible to produce graded structures or even composite structures by varying the mix during processing. It can be combined with crystalline metals to produce materials with high strength and enhanced ductility, which is made possible with this flexibility [15].

However, preservation of the amorphous structure during DED is difficult due to the low cooling rates and subsequent heating, which may cause crystallization or segregation. Close monitoring of the process parameters such as laser power, scan speed, powder feed rate and cooling rate is necessary to maintain the glassy phase. Enhanced use of pulsed lasers, beam oscillation, or real-time thermal monitoring can be used to obtain rapid solidification and homogeneous composition. In general, DED provides an opportunity to produce and repair BMG parts with specific properties; however, its efficiency relies on precise thermal management to prevent crystal formation and to guarantee uniformity of the structure [30].

3.4. Hybrid and Multi-Material AM for BMGs

Recent advances in AM have created new opportunities to mix BMGs with other functional or metallic materials to create hybrid or multi-material components. The above processes, facilitated mainly by laser-based technologies including DED and LPBF, are aimed at addressing one of the primary limitations of BMGs, namely, their low ductility, as well as preserving their exceptional strength, wear resilience and corrosiveness.

In such hybrid fabrication systems, it is possible to control the spatial variation in the material composition and microstructure in a single build by using the AM process. For example, a piece of amorphous metal can be printed next to, or progressively changing to, a crystalline alloy piece. This graded structure can better spread loads and avoid disastrous fractures since the crystalline structure can serve as a ductile reinforcement. The experimental examples of the use of Zr-based BMGs have revealed that, when they are combined with crystalline metals like titanium or steel with controlled laser energy input, they produce high levels of metallurgical bonding and damage tolerance [78].

The multi-material printing ability of DED is also particularly beneficial in this situation since it enables the composition of powder or wire feeds to be altered in real time. This allows the creation of functionally graded structures in which the metallic glass is gradually replaced with crystalline alloy in the build direction. Such compositionally graded materials (CGMs) can take advantage of the fact that the amorphous zone is hard and wear-resistant in areas subjected to friction or corrosion, but ductile and tough in load-bearing areas.

In addition to multi-material integration, the use of additive manufacturing has also opened the possibility of producing BMG composites, where the amorphous matrix is actively reinforced with crystalline phases. They are important in preventing the propagation of shear-bands, the primary deformation mechanism, which inhibits monolithic metallic glass plasticity by the presence of these crystalline inclusions, whether formed in situ during laser melting or introduced as secondary particles. What is obtained is a composite material that not only has the high level of strength of the amorphous matrix, but also utilizes the high ductility of the crystalline reinforcements [79].

Laser-based fabrication has the added benefit of control of the volume fraction, morphology, and distribution of the crystalline phase by modulating energy density, scan speed, and cooling rate. As an example, reduced laser intensity and increased scan rates encourage fast solidification and maintain amorphous spaces, and slightly slower cooling may permit the development of nanoscale crystalline precipitates that enhance the matrix, but do not substantially degrade its amorphous character. This partial crystallization where the crystallization rate of a material is controlled, as is shown in laser-processed Zr- and Fe-based BMGs, results in a fine hybrid microstructure with equal strength and ductility.

Nevertheless, such hybrid and multi-material techniques are also very challenging to process. The difference in the thermal and mechanical characteristics of amorphous and crystalline phases produces residual stress, delamination threat, and the probability of devitrification on the interfaces. To obtain a stable amorphous structure in these composites, it is essential to control the thermal gradient of deposition and the heat energy. The amorphous integrity is commonly reinforced with the help of such post-processing treatments as annealing or surface laser remelting to remove internal stresses [30].

3.5. Process-Performance Comparison of Additive Manufacturing Techniques for BMGs

Comparative analysis of the LPBF, EBM, and DED methods in bulk metallic glass fabrication shows clear trade-offs between amorphous retention, dimensional accuracy, and thermogenic effects of the process. LPBF tends to provide better vitrification and surface resolution because of its higher cooling rates, and EBM and DED tend to be more vulnerable to crystallization because of lower quenching efficiency and cumulative reheating which should result in more careful optimization of the parameters [35,80].

Table 3 provides a comparison summary, focusing on the combined control of structural stability and process reliability in bulk metallic glasses obtained through additive manufacturing by cooling rate, the nature of the melt pool, and the amount of energy input. When comparing SLM/LPBF, EBM, and DED for BMG fabrication, several key performance metrics arise.

Table 3. Comparative assessment of LPBF, EBM, and DED for bulk metallic glass fabrication ([35,52,56,60,70,71,72,80,81]).

Metric	LPBF	EBM	DED
Cooling rate	106–108 K/s	103–105 K/s	Lower; large melt pools
Amorphous retention	Nearly full under optimized parameters	Partial; composition-dependent	Limited; HAZ crystallization common
Dimensional resolution	High	Moderate	Lower; post-processing required
Thermal gradient	Steep	Reduced (preheated bed)	High thermal accumulation
Residual stress	High risk	Lower	Moderate–high
Reheating sensitivity	Build-height heterogeneity	Nanocrystallization possible	High devitrification risk
Process window	Narrow but tunable	Narrow; cooling-limited	Very narrow; strict control required
Overall suitability	Most promising for high amorphous fraction	Viable for high-GFA alloys	Suitable for graded/large builds; vitrification challenging

Although all AM methods have their own advantages concerning BMG fabrication, at the present moment, LPBF is the most promising regarding high amorphous fraction and fine geometries. However, both of these methods necessitate careful process design, alloy choice, and post-processing approaches to address the challenges that are inherent in the vitrification of metallic glasses in an additive manufacturing system.

3.6. Overview of Successes and Limitations in AM of BMGs

3.6.1. Successful AM-Fabricated BMG Systems

The successful additive manufacturing of BMGs is largely controlled by the capability to sustain reasonably high cooling rates and reduce the accumulation of heat during layer-by-layer deposition. Examples of alloys (e.g., Zr-, Ti-, Fe-, and Cu-based) that have been reported to form near-amorphous structures under optimal processing conditions are formed by the specific AM techniques utilized (e.g., LPBF, EBM and DED). These successful examples demonstrate the importance of the amalgamation of alloy composition, process variables (e.g., energy density, scan speed), and thermal control in determining high amorphous retention, structural uniformity, and favorable mechanical behavior. Table 4 provides some representative examples of fabricated BMG systems using various AM techniques, including processing conditions, amorphous fraction, and key mechanical properties.

Table 4. Summary of successful additively manufactured BMG systems.

AM Process and Key Notes	System	Amorphous Fraction/Devitrification	Key Properties (Yield Strength, Fracture Toughness, Ductility)	References
LPBF (Optimal energy density 20–33 J/mm3)	Zr-based (AMZ4)	Fully amorphous (>99% relative density)	Flexural strength: 2080 MPa; KIC: ~38 MPa·m1/2; plastic strain: ~6%	[14,38]
LPBF (selective laser melting)	Zr-based (Zr60Fe10Cu20Al10)	96.95% amorphous	Engineering strain: 18.42% (with 4.05% nanocrystals)	[82]
3D printing (synergistic defects engineering)	Zr-based (Zr60.14Cu22.31Fe4.85Al9.7Ag3)	Partial crystallization (<90%)	Yield strength: 1.8 GPa; KIC: ~45 MPa·m1/2; plasticity: >1%	[83]
Laser foil printing (LFP)	Zr-based (Zr52.5Ti5Al10Ni14.6Cu17.9)	Fully amorphous	Properties comparable to as-cast BMG; high yield strength and toughness	[84]
LPBF (interpenetrating microstructure)	Fe-based (FeCoBSiNb-Cu)	~80% in Fe-rich region	High fracture strength; improved compressive plasticity	[85]
LPBF (in situ β-reinforced)	Ti-based (Ti-Zr-Cu)	BMGC (dendrite + amorphous)	Compressive strength: 1.9 GPa; total strain: >13%	[86]
LPBF (XRD/DSC quantified)	Cu-based (CuZr-based)	Amorphous content varies with parameters	Hardness vs. amorphous fraction discussed; specific yield/toughness not detailed	[87]
DED/LMD (laser metal deposition)	Zr-based (AMZ4)	Amorphous/composite structure	Microstructure and mechanical correlation discussed; explicit numeric yield not available	[88]

3.6.2. Failed and Challenging BMG Systems in AM

Although there has been significant improvement, a number of BMG compositions which display superior glass-forming behavior in conventional casting temperature do not maintain their amorphous structure in additive manufacturing. These failures are mainly linked to process-related factors, including cyclic thermal reheating, steep thermal gradients, and defect formation which favor devitrification, cracking, and structural heterogeneity. Specifically, alloys with low ductility or small supercooled liquid volumes are extremely vulnerable to microcracking, HAZ crystallization, and leftover stress concentration during AM processing. The table below details some of the representative cases of failure, revealing the underlying mechanisms which restrict printability, and the important difference between castability and AM processability. Table 5 provides a summary of reported failure cases in additive manufacturing of bulk metallic glasses, including compositions, processing challenges, and underlying failure mechanisms.

Table 5. Reported failure cases in additive manufacturing of BMGs.

Alloy System/Composition	Process	Primary AM Defect/Failure Mode	Underlying Mechanism During LPBF	Consequence on Build/Properties	References
FeCrMoBC	LPBF	Severe cracking	Extremely limited ductility combined with high thermal gradients inherent to LPBF	Parts were too brittle and cracked to undergo meaningful mechanical testing; very narrow processing window	[89]
Fe-based	LPBF	Unavoidable microcracking	Micropores act as stress concentrators; FEM showed local stresses reached up to 4.1 GPa, exceeding the 3.5 GPa fracture strength of the alloy	Unavoidable microcracking leading to structural failure	[90]
Fe68.3C6.9Si2.5B6.7P8.7Cr2.3Mo2.5Al2.1	LPBF	Microcracks	Large temperature gradients between laser-processed and solidified layers generating excessive thermal stress	Limited fracture toughness leading to extensive crack propagation	[91]
Fe43.7Co7.3Cr14.7Mo12.6C15.5B4.3Y1.9	LPBF	Extensive HAZ crystallization	Reheating (laser flashes) from adjacent scan tracks raised local temperatures above the crystallization temperature ($T_x$)	Loss of monolithic amorphous structure triggered structural heterogeneity	[52]
Al86Ni6Y4.5Co2La1.5	LPBF	Severe cracking, porosity, and HAZ crystallization	Highly sensitive thermal processing window; high power caused excessive stress/cracking and increased HAZ width; low power caused incomplete melting/porosity	Exceedingly difficult to prepare a crack-free and fully amorphous build without secondary remelting strategies	[92]
Zr55Cu30Ni5Al10	LPBF	Brittle crystallization phases in HAZ	Temperature in the HAZ exceeded $T_x$ during laser irradiation of adjacent powder bed regions	Premature destructive brittle fracture: yield strength significantly reduced to 1504 MPa compared to cast counterparts	[93]
Zr-based (AMZ4)	LPBF	Catastrophic tensile failure	Lack of fusion (LoF) defects acting as severe stress concentrators	Significant drop in ultimate tensile strength and impact toughness compared to cast samples	[94,95]
Nb-reinforced BMG composite	DED	Nanocrystalline embrittlement	Formation of nanocrystals at the boundaries of the melt pool during laser directed energy deposition	Embrittlement of the monolithic BMG matrix; reduced ductility and plasticity	[96]
Fe-based BMG	DED	Oxygen-induced crystallization	Oxygen uptake during DED processing triggered phase evolution and devitrification	Loss of amorphous structure; structural heterogeneity and mechanical degradation	[97]
Fe-Ti Mixtures	DED	Interlayer cracking and pores	Precipitation of brittle intermetallics and high residual stresses during layer build	Structural embrittlement and high failure risk during processing	[98]
Zr41.2Ti13.8Cu12.5Ni10Be22.5 (Vit 1)	EBM (remelting)	Severe HAZ crystallization	Cyclic reheating in the HAZ raised temperatures above Tx.	Loss of amorphous state in the HAZ while fusion zone remained amorphous	[72]

4. Process–Structure–Property Relationships

4.1. Composition–Stability Relationships

The composition-stability relationships refer to the relationships between the composition and stability parameters of the product. During AM, the time taken to heat and cool a BMG depends on the composition of the material. Deep eutectic alloys, large atomic size discrepancies, and strongly negative heat of mixing tendencies all lead to greater thermal stability and crystallization obstacles, implying an extended process window under the conditions of laser-based AM [74]. High thermal stability is identified by a large supercooled-liquid phase and large reduced glass transition temperature, both of which aid in resisting crystallization during rapid solidification or remelting events.

Pauly et al. [28] used selective laser melting (SLM) on Fe-based glass-forming powders and revealed that slight compositional variations as low as contamination of the material with Cu, Ni, or Ti reduced the crystallization onset temperature. However, optimized laser conditions maintained the amorphous form, proving that the process window is a factor of both intrinsic alloy kinetics and extrinsic factors like energy density, scan speed, and layer thickness. Slower diffusing, higher viscosity alloys in their supercooled-liquid form have a greater ability to withstand such variations.

In contrast to the single cooling used in traditional casting, in AM, alloys are subjected to repeated heat flows. Thus, the compositions applicable to AM should focus on the presence of the glass-forming ability, as well as thermal cycling strength. By altering the Cu/Ni ratio or incorporation of Al, Be or Nb, the liquidus temperature can be adjusted and ΔTₓ expanded to enhance better stability. Alloys containing Zr have been shown to have broader supercooled space and enhance amorphous retention compared to Fe-based systems [18].

The impact of partial crystallization on the crack propagation characteristics of bulk metallic glass specimens produced by LPBF is shown in Figure 4. The etched optical micrographs and subsequent image analysis maps identify the amorphous and crystalline regions and show that the spatial distribution of crystalline tracks has a strong impact on the fractured paths. Local stress distribution is changed by the existence of crystalline regions, and heterogeneities are created which change the path of the crack. Furthermore, cracks are observed to propagate preferentially perpendicular to the crystalline/amorphous interfaces, indicating a strong correlation between crack direction and interface morphology [99].

Figure 4. Schematic of the workflow followed to analyse the paths of preferential crack propagation by image analysis. (a) Etched optical micrograph illustrating, in brown contrast, crystalline regions and in light contrast, amorphous regions; (b,c) binarized images highlighting amorphous (green, (b)) and crystalline (grey, (c)) phases; (d) enlarged view of a region of (c) in which facets along the crystalline/amorphous interfaces have been coloured according to the orientation of their normal direction following the colour coding included in the inset. (e) Histogram comparing the preferential orientations of cracks (red bars, θ) and of the directions perpendicular to the different facets along the crystalline/amorphous interphases (blue bars, θ’) in the LPBF manufactured Kuamet6B2 KS1 sample. [99]. (Reproduced from Rodríguez-Sánchez et al., Materialia (2024) 35:102111 licensed under CC by 4.0).

4.2. Defects and Residual Stresses

The structure and defects formed during build are indicative of composition and process conditions in AM of BMGs, and the defects formed in turn have a strong impact on functional performance. Rapid thermal cycling, steep thermal gradients, and shrinkage in solidification give rise to residual stresses, the magnitude of which depends on the thermal diffusivity, melt viscosity, and phase changes, which depend on alloy chemistry. As an example, alloys of greater modulus or lesser thermal conductivity could be subjected to greater residual stress under the same processing conditions. Surface roughness and porosity also depend on the properties of the powder (which affect these parameters in turn depending on composition, e.g., alloy melting point, oxidation tendency, flowability) and melt-pool stability (due to viscosity, surface tension, and thermal conductivity). High porosity or surface roughness becomes a critical defect in BMGs since there are no dislocations or grain boundaries, so that their plasticity depends upon the formation and propagation of shear transformation zones; a pore or a roughness pit is a stress concentrator leading to disastrous shear banding. In such a way, the composition of the chemical (determining the sensitivity of intrinsic defects and the melt dynamics) and the processing (determining the defect formation) determines the as-built defect topography and, by extension, the final property [35].

Figure 5 illustrates the strong dependence of surface roughness on LPBF processing conditions and post-exposure strategy. The SEM image of the as-built upper surface, shown in Figure 5a, reveals partially melted and unmelted powder particles adhering to the surface, particularly for particle sizes exceeding 50 μm. These particles produce large surface topography differences, established by a confocal surface profile with a height variation of about ±150 μm and roughness value of Ra ≈ 40 μm (Figure 5b). They are caused by the lack of stabilization of melt pools, remitting all the powder particles, which depend on alloy thermal conductivity, viscosity, and laser energy density [99].

Figure 5. Surface roughness of LPBF-built CuCr1Zr samples. (a) SEM image of the as-built upper surface showing partially melted and unmelted powder particles. (b) Confocal surface profile revealing topography variations of approximately ±150 μm (Ra ≈ 40 μm). (c) SEM image after application of a manifold laser exposure strategy on the final layer. (d) Corresponding surface profile showing reduced topography variation of approximately ±50 μm (Ra ≈ 9 μm). The results highlight the strong influence of processing strategy on surface integrity and defect formation [17]. (Reproduced from Jahns et al., Int. J. Adv. Manuf. Technol. (2020) 107:2151–2161, licensed under CC BY 4.0).

The unmelted particles are melted after a manifold laser exposure strategy is applied on the last layer in Figure 5c,d to the point that the height variation is about ±50 μm and the Ra is about 9 μm. This proves that the quality of the surfaces in LPBF depends not only on composition but also on the dynamics within melt pools to process final layers. Such surface anomalies are strain concentrators in brittle or defect-prone materials, e.g., bulk metallic glasses, and may increase the rate of shear localization and crack propagation. Thus, the interplay between alloy chemistry and process parameters directly determines the as-built defect topography and, consequently, the functional performance of the component [17].

4.3. Amorphous Stability and Devitrification

One of the central questions of the AM of BMGs is whether the part is completely amorphous, or whether crystallization has partially occurred; this depends closely on a complicated interaction between processing and alloy composition. Alloys that have very high critical cooling rates or moderately high stability of supercooled liquids tend to form devitrification in the HAZ or elevation zones of the melt pool where the cooling is slower or the melt pool is re-heated. As an example, some Zr-based AM experiments have found a remelted core zone which is amorphous, but an annular HAZ with crystalline precipitates due to slower cooling and nuclei of the substrate [28]. Therefore, the crystal fraction is determined by the distribution of cooling rates, reheating, kinetics of alloy nucleation, and available heterogeneities. Textile crystallization can weaken the desirable high hardness and strength of the glassy phase, though in others it can be contrived (e.g., glass-crystal composites) to provide ductility or toughness where required. The critical outcome is that the structure (fully glassy vs. partially crystallized) is not simply a binary output of AM, but rather a continuous function of processing parameters, alloy chemistry and thermal history.

Figure 6 represents a quantitative representation of the effect on the amorphous fraction of the sample that has undergone the LPBF process as a continuous variable of laser power (P), scan speed (v) and layer thickness (h). The color maps show that only in a limited processing window are completely amorphous structures realized and slight changes in combinations of parameters result in quantifiable partial crystallization. As an illustration, incomplete melting and deficiency of thermal input at lower energy densities (low P or high v) decrease densification and may favor structural heterogeneities. On the other hand, energy densities that are too high lead to higher thermal concentration and reheating, and improve the kinetics of crystallization in the melt pool and HAZ. It was found that thinner layers (e.g., h = 50 µm) are more conducive to higher amorphous fractions over the expanded parameter range since the cooling rates are higher, whereas thicker layers (h = 80–90 µm) have a narrower fully amorphous processing range and prefer to partially crystallize. Significantly, the maps point to the fact that the structural condition is not a two-state (amorphous vs. crystalline) phenomenon, but can continuously change over time as the processing parameters change. The observed gradients in amorphous fraction are a result of local cooling rate distribution, repeated thermal cycling, and crystallization kinetics. These results establish that the spatial and temporal differences in thermal history are the dominant factors in controlling phase formation in additively manufactured bulk metallic glasses, and that alloy chemistry and process control cannot be separated in determining the ultimate microstructure [17].

Figure 6. Amorphous fraction of LPBF-fabricated samples as a function of laser power and scan speed for different layer thicknesses: (a) Print 1; (b) print 2, h = 50 µm; (c) print 2, h = 70 µm; (d) print 2, h = 90 µm [99]. (Reproduced from Rodríguez-Sánchez et al., Materialia (2024) 35:102111. Licensed under CC BY 4.0).

These observations confirm that the mechanical response of additively manufactured BMGs is closely linked to phase formation tendencies. Rather than behaving as purely brittle materials, partially crystallized microstructures can change fracture morphology and crack propagation dynamics. This further supports the notion that structure evolution during processing is continuous and directly governs the resulting mechanical behavior [99].

4.4. Energy–Thermal–Crystallization Framework

To determine a quantitative connection between additive manufacturing parameters of processing and devitrification behavior in bulk metallic glasses, it proves beneficial to differentiate between nominal energy descriptors and the consequent thermal-history factors that are the controlling variables of the crystallization process. The overall response of laser- and electron beam-based AM processes is usually expressed in volumetric energy density (E_v), which is the amount of energy required to process one unit of volume of powder [100].

Nevertheless, even identical values of energy density can give significantly different melt-pool geometries, peak temperatures, and cooling rates, based on beam diameter, absorptivity, scan strategy, and the interlayer dwell time. Hence, energy density can only be considered as a first-order proxy of underlying physics [101]. The governing variables for amorphous retention are instead determined by derived thermal-history metrics, such as peak temperature $T_{peak}$, liquid-state dwell time, effective cooling rate $\mid \mathrm{d}\mathrm{T}/\mathrm{d}\mathrm{t}\mid$, and the cumulative exposure time within the crystallization-sensitive temperature interval between the liquidus and glass-transition temperatures. The thermal descriptors can be quantitatively correlated to crystallization kinetics through non-isothermal models like the Nakamura formulation, which correlates the local fraction of crystallized material to the crystallization rate K(T) along the transient thermal path in AM processing [37,102]. Within this context, two opposing needs can be taken to define the process window of BMG additive manufacturing: (i) adequate energy input to obtain full melting and densification without forming lack-of-fusion defects, and (ii) adequate cooling and limited reheating to prevent devitrification in the heat-affected area.

Table 6 provides a compilation of the correlation between widely used AM process descriptors and the thermal-history variables to which devitrification in bulk metallic glasses is subject. Such nominal parameters as laser power, scan speed, hatch spacing, and layer thickness are frequently published in the literature in the form of energy density measures (linear, areal or volumetric energy density). However, these descriptors by themselves do not specify the thermal pathway that the material undergoes. Rather, they affect physically significant thermal parameters, which include peak temperature, dwell time in liquid state, cooling rate, and cumulative exposure to heat in the heat-affected zone. These thermal descriptors eventually regulate the degree of crystallization by thermo-sensitive nucleation and growth kinetics, which could be modeled in non-isothermal equations like the Nakamura or JMAK equations [37,102,103].

Table 6. Relationship between AM process parameters, thermal history, and crystallization behavior in bulk metallic glasses.

Descriptor	Parameters	Mathematical Definition	Physical Meaning	Influence on BMG Structure	References
Process input	Laser/beam power, P	Input power of energy source (W)	Total heat input to melt pool	Higher power increases melt pool temperature and dwell time, potentially promoting crystallization	[100]
	Scan speed, v	Beam travel velocity (mm/s)	Interaction time between beam and powder	Higher speed reduces heat input and increases cooling rate	[45]
	Hatch spacing, h	Distance between adjacent scan tracks (µm)	Overlap between melt pools	Larger spacing may lead to lack-of-fusion defects	[100]
	Layer thickness, t	Powder layer thickness (µm)	Material volume melted per pass	Thicker layers require higher energy input for full melting	[45]
Energy	Linear energy density, El	$E_l=P/v$	Energy per unit scan length	Useful for single-track analysis and melt-pool stability	[104]
	Volumetric energy density, Ev	Ev = P/(vht)	Energy input per unit volume processed	Widely used metric for LPBF process optimization	[105]
Temperature history	Peak temperature $T_{peak}$	Maximum temperature in melt pool	Determines melting and dissolution of nuclei	High $T_{peak}$ promotes full melting but may increase HAZ size	[104]
	Colling rate, dT/dt	-	-	Temperature gradient over time	-
	Liquid-state dwell time, ${\tau }_{liq}$	Time above liquidus temperature	Indicates duration of melt-pool stability	Longer dwell promotes grain nucleation or crystal growth	[104]
	Reheating exposure time, ${\tau }_{HAZ}$	Time above critical crystallization temperature during reheating	Captures HAZ thermal cycling	Multiple reheating events promote devitrification	[104]
Kinetic Descriptors	Crystallization rate constant, $K\left(T\right)$	Temperature-dependent kinetic constant	Describes nucleation and growth kinetics	Determines crystallization probability during thermal cycle	[106]
	Crystallized fraction, $X_c$	Fraction $X_c$ $X_c=1-e^{[-(\int K\left(T\right) dt{)}^n]}$	Fraction of crystalline phase formed	Determines final amorphous fraction in BMG	[104]
Process Outcome	Amorphous fraction, $X_a$	$X_a=1-X_c$	Degree of glassy structure retained	High $X_a$ corresponds to desired BMG structure	[104]
	Defect regime	Empirical process window	Identifies lack-of-fusion or keyhole regimes	Too low energy → lack of fusion; too high energy → keyhole or devitrification	[107]

4.5. Structure–Performance Relationships

The functional performance of the as-built microstructure of AM-fabricated BMGs includes amorphous fraction, crystalline phase content, residual stress state, porosity distribution, and surface roughness, which are directly connected. Mechanically, increased amorphous fraction translates to increased hardness and increased yield strength, due to the lack of any dislocation, although it can be at the cost of decreased global plasticity and increased brittleness. On the contrary, the presence of ordered crystalline precipitates or shear-band arresting properties can increase plasticity/toughness at the expense of strength [108]. Other thermal performance properties (e.g., stability under load, creep resistance) are also controlled by the microstructure: a crystallized nature may reduce the barrier to shear-band movement or modify thermal expansion inequity, resulting in the relaxation of residual stress or failover. Other functional performances, such as the soft magnetic behavior of Fe-based BMGs or the elastic storage of energy in Zr-based systems, are very sensitive to microstructure purity: any light crystallization or element segregation impairs the magnetic domain structures, or elastic resilience. Therefore, the triad of process–structure–property is particularly closely interconnected in AM of BMGs: it requires a reduction in amorphous fraction and decrease in negative defects to optimize performance; residual stresses and microstructural heterogeneity also have to be reduced [109].

Figure 7 shows the key processing inputs and specimen configurations that underpin the process–structure–property relationships in LPBF-fabricated Zr-based bulk metallic glass (AMZ4). The SEM images of fine and coarse powders (a, b) emphasize their spherical form and particle size distribution, which is significant in the powder flow, packing behavior and melt pool stability during laser processing. The printed 5 × 5 × 5 mm³ cubes (c) reflect that the densification and retention of amorphous phases are sensitive to the choice of processing parameters. The compression specimen shape (d) permits uniform mechanical testing of the manufactured material. In general, it demonstrates the way that the primary powder properties and the processing conditions of the LPBF influence the final microstructure and define the mechanical properties of the material [14].

Figure 7. SEM images of fine (a) and coarse (b) AMZ4 powders used for LPBF, (c) fabricated 5 × 5 × 5 mm³ BMG cubes, and (d) compression specimen geometry (dimensions in mm). Powder characteristics and LPBF process parameters control densification, amorphous structure retention, and mechanical performance [14]. (Reproduced from Li et al., Acta Mater., (2024) 266:119685 under CC BY 4.0).

However, it is worth mentioning that process descriptors like energy density which are commonly in use do not independently dictate the end result properties. Rather, thermal history is affected by the energy density, which also affects the cooling rate, which in turn governs microstructural evolution, such as crystalline dendrite-formation and distribution. Crystalline dendrites are tree-like solidification structures formed during quick cooling and reinforcing phases in the amorphous matrix. The mechanical behavior (strength and ductility) of these dendrites is highly dependent on the volume fraction and morphology of these dendrites. Previous studies by Lu et al. [110] have demonstrated that cooling rate serves as the critical bridging parameter linking processing conditions to microstructure and properties, highlighting the limitations of relying solely on energy density to interpret process–structure–property relationships.

5. Post-Processing and Thermal Stability

Post-processing is important in determining the structural integrity, thermal stability, and long-term performance of additively manufactured BMGs. The metastable amorphous structure of BMGs in contrast to conventional crystalline alloys is extremely temperature sensitive, hence, with controlled heat treatment, the balance between defect mitigation, stress relaxation, and crystallization resistance is easily achieved. Such processes as annealing and hot isostatic pressing (HIP) can be used to increase densification and structural homogeneity, but excessive thermal management can cause relaxation-induced embrittlement or partial devitrification. Thus, the interplay between thermal treatment, structural relaxation, and phase evolution is the keystone for the maximization of the functional reliability of the AM-fabricated BMG components [59,111]. Figure 8 presents the post-processing and thermal stability strategies for BMGs.

Figure 8. Post-processing and thermal stability analysis of BMGs.

5.1. Annealing, Hot Isostatic Pressing and Rejuvenation

Post-processing treatments are important in determining the thermal stability and long-term behavior of BMGs. Traditional annealing at temperatures less than the glass transition temperature implies that the structure relaxes to reduce the amount of free volume and internal lattice stresses, and, consequently, can easily enhance the yield strength and stability of the glassy state. However, when annealed over a long period of time, it can cause embrittlement or the beginning of nanocrystallization unless the super-cooled liquid area is properly controlled. As an example, controlled annealing of a Cu-Zr-Al glass also led to softening through rejuvenation with a higher enthalpy state induced by the treatment path, which demonstrates the duality of the role of heat treatment in the aging or rejuvenation of the amorphous structure [78].

5.2. Structural Relaxation and Phase Evolution

The influence of the post-processing of BMGs can be explained through structural relaxation, redistribution of free volumes, and possible crystallization development. The decreasing volume of free matter and the rearrangement of atomic clusters during annealing causes a decrease in the stored enthalpy energy and the move of the material to a lower energy state, which usually enhances stability but can also decrease toughness or ductility since the shear transformation zones are harder to initiate. Conversely, rejuvenation therapies raise the free volume, the disorder, and the level of energy of the system, and may allow more distributed shear-banding and better ductility, but at the cost of some statical strength and possibly lower thermal stability. Recent exploration of metallic glass heat-treatment paths showed that the aging/rejuvenation balance is highly sensitive to the previous cooling rate, heating rate, and duration of annealing; slower cooling rates or rapid heating/cooling sequences and longer temperature pulses favor aging or rejuvenation, respectively [112]. In addition, microstructural changes in post-processing should be well controlled to prevent devitrification. Although a BMG can be amorphous following AM, further annealing, HIP, or thermal exposure can permit the formation and development of crystalline phases, especially in alloys with only moderate glass-forming capability or low-lying supercooled liquid. The addition of nano-crystallites in the glass matrix may serve as a local stress concentrator, interfere with shear-band propagation, or initiate embrittlement. Crystallization of only a few percent in volume in a BMG has been demonstrated to significantly reduce fracture toughness [113]. The balance between post-processing and thermal stability in BMGs is fragile: both annealing and HIP ease defects in the structure and stabilize it, increasing plasticity, and both suppress undesired microstructural formation (such as crystallization) by exploiting the distinct amorphous nature of the material. In the case of AM-fabricated BMGs, the combination of the as-built microstructure (cooling history, residual stresses and porosity) and post-processing path will prove to be especially important to desired performance.

6. Progress in Additive Manufacturing of BMGs

Over the past few years, additive manufacturing of BMGs has shifted its focus towards demonstrations of the systematic optimization of processes and microstructure. Recent developments in laser- and electron-beam technology, process modeling, and in situ control have greatly increased the feasible processing window to obtain high amorphous retention and enhanced structural reliability. Advances in alloy design, hybrid manufacture strategies, and microstructure control have added to the achievement of improved mechanical performance and functional tenability [114]. Combined with other innovations, these hail a transition to more predictive, scalable and application-oriented manufacturing of BMG components. Figure 9 provides the recent advances in additive manufacturing of BMGs.

Figure 9. Progress in additive manufacturing of BMGs.

6.1. Process Innovations

Within the last five years, the field has shifted away from demonstrations of early feasibility to systematic optimization of the process-material windows, as well as exhibiting new chemistries and architectures. Laser-based powder bed fusion (LPBF/SLM) has historically experienced the highest development, as its tiny melt pools and enormous cooling rates per unit optimally suit kinetic conditions in vitrification; recent reviews and parameter considerations determine how scan plan, energy concentration, and beam shape broaden the practical process window of most Zr, Ni and Nb-bearing BMG compositions [115]. Developed laser beam shaping (flat-top and structured intensity profiles) demonstrated the creation of broader and shallower melt pools, leading to lower peak thermal gradients, decreased HAZ crystallization, higher amorphous retention, and more uniform part density. Complementary numerical studies have shed light on the effect of hatch patterns, layer time and recoating on the distributions of cooling-rates and interlayer reheating, to allow the preparation of predictive process maps within a choice of chemical structures. Oxygen-sensitive alloys are also being used in EBM, as well as in vacuum-based methods, as the absence of oxidation and the reduced amount of trapped gases promote faster cooling (accompanied by a larger melt pool necessitating close control or combined cooling methods), although these techniques have the disadvantages of being slow to cool and requiring significant power. Normally used in functionally graded or repair applications, where compositional control and gradient structures are beneficial, DED currently faces challenges in achieving amorphous builds due to larger melt pools and slower solidification), although progress is being made in this direction. Lastly, first AM demonstrations of previously non-castable chemistries have been reported by several groups (e.g., binary Ni-Nb and some Fe-based commercial MGs), with AM demonstrating the capability to reach compositions with poor castability, and good as-built characteristics when the process conditions are optimized [116].

6.2. Performance Enhancement

The combined effect of hardware advancements, control of processes, and feedstock has enhanced dimensional fidelity and amorphous fraction in printed BMGs. Optimized landing posts (resulting in lower volumetric energy density and higher scan speeds with multiple pass strategies) for LPBF TP3D parts are now commonly able to yield near fully amorphous parts with Zr-based material displaying lower surface roughness and dimensional tolerances closer to those required to pass engineering tests, and post-processing polishing/HIP further reduces porosity and increases fatigue resistance [111,117]. Beam shaping and controlled thermal treatment have been directly related to an increased amorphous fraction and compressive/yield strength in as-built samples; compressive yield strengths have been reported to be 2–3 GPa in dense LPBF components of customized alloys [118]. Simultaneously, closed-loop process controls (e.g., in situ thermography and melt-pool monitoring) have minimized partially devitrified runs by permitting real-time parameter changes when thermal indicators show that there is a risk of HAZ crystallization [20,118].

6.3. Hybrid and Intelligent Manufacturing

Strategies for hybrid manufacturing, such as AM plus subtractive finishing, localized heat sinks, or stepwise pre-heating, are showing to be effective in the trade-off between cooling-rate requirements and dimensional precision and control of residual stress. Multi-material and hybrid prints (e.g., incorporation of ductile crystalline phases or compliant metallic skins with amorphous cores) are beginning to overcome the inherent brittleness of monolithic BMGs and achieve functioning performance gradation [119,120]. Real-time monitoring approaches (thermography, pyrometry, acoustic sensing, melt-pool imaging) combined with digital twins and process models are no longer considered as mere proof of concept but as devices that can be used to eliminate devitrification instances and select adaptive scanning strategies [121]. Physics-informed optimization and machine learning have been employed in observations of high-throughput parameter studies, and accelerate the process of discovering robust parameter windows for new chemistries at decreased cost with less trial and error [122,123].

6.4. Microstructural Engineering

Modern technologies deliberately sacrifice a small, regulated degree of crystallinity in pre-planned second stages to enhance ductility and toughness at the cost of a little of the strength in the glassy phase. This include nanoscale precipitations, which is a controlled, deliberately formed glass-crystal composite and functionally graded structure in which crystalline phases serve as shear-band arrestors or crack-deflecting ligaments [88,124]. The compressive plasticity and energy absorption in architected lattices and bulk specimens have shown considerable improvements in these engineering routes that allow the local compositional and thermal control of the AM. In the case of functional properties (magnetism, elastic energy storage), control of the process that maintains chemical homogeneity and alleviates any elemental segregation during the melting/solidification process is now understood to be just as important as a large amorphous fraction [125].

7. Challenge-Mitigation Framework for BMG-AM

BMG additive manufacturing is similarly subject to the essentially constrained metastable state of the amorphous state despite the impressive progress that has been achieved in terms of processing strategy and microstructural control. Maintaining cooling rates above the critical temperature at the same time as controlling complex thermal histories creates a persistent danger of devitrification, accretion of residual strains, and structural non-homogeneity. In addition, drawbacks in the form of printable thickness, quality of feedstock, and environmental friendliness complicates reproducibility and scalability. These challenges, as illustrated in Figure 10, highlight that successful AM of BMGs needs coordination between the optimization of alloys, thermal regulation, management of geometry, and process stability to retain amorphous integrity across the build volume.

Figure 10. Challenges in additive manufacturing of BMGs.

7.1. Thermal Accumulation and Devitrification

One of the significant issues in additive manufacturing of BMGs is that, due to the accumulation of heat and repeated reheating in the process of making the layers, there is spontaneous crystallization. The localized remelting and heat accumulation caused by each deposited layer cause the effective cooling rate to drop below the critical cooling rate of the alloy, leading to nuclei forming and developing within the HAZs, even in small crystallized regions. This effectively lowers the high strength, elasticity, and corrosion resistance of the amorphous matrix [36].

The highest accumulation occurs in low layers because the conduction of heat to the substrate is reduced, thus devitrification partially occurs. Johnson–Mehl–Avrami and Nakamura models of crystallization show that the hatch pattern, layer time, and laser energy density are critical factors that affect crystallization behavior [78].

Optimizing the laser scanning speed, power, and interlayer dwell time in Fe- and Zr-based BMGs is encouraged, particularly with low-glass-forming alloys [18], as it allows for the control of non-localized crystallization. Mitigation measures include the optimization of scanning strategies, alternating hatch directions, minimization of dwell times, and utilization of beam shaping/active cooling to ensure high effective cooling rates. However, in alloys with a small supercooled liquid region, these may not be adequate. Increased use of sophisticated real-time measurements to monitor temperature and devitrification during printing is encouraged through infrared thermography and in situ X-ray diffraction [28,60].

7.2. Residual Stress and Cracking

The major issue with additive manufacturing of BMGs is the formation of residual stress. The high-speed, non-uniform cooling and sharp thermal gradient of the laser- or electron-beam treatment process create high residual stresses which are not relieved by dislocation movement because BMGs do not creep by shear-band propagation, but creep through localized shear-band propagation [79]. These strains add up in several layers, resulting in crack formation, delamination, and warping, particularly in heavier builds [28].

Amorphous matrix crystallization also increases local stress because of volume changes and phase mismatch, and enhances crack propagation. Experimental observations of Zr- and Fe-based BMGs subjected to high thermal loads indicate that both surface and internal cracking occurs in LPBF-processed substrates [52,126].

Substrate preheating and substrate/scan speed optimization are mitigation methods applied to reduce thermal gradients and prevent devitrification by slowing down cooling rates; melting/stress optimization is used to control melting and stressed islands; remelting runs are used to idealize temperature; the homogenization of temperature is employed to relax stresses; and post-build HIP aids in internal pore closure. Nevertheless, all methods come at a cost: both alloy and thermal management must consider the trade-offs needed to ensure minimal residual stress and amorphous integrity in BMG additively manufactured parts [15,127].

7.3. Geometric and Thickness Limitations

The additive manufacturing of BMGs can be fabricated with maximum printable amorphous thickness (Dₘₐₓ) as a result of the interplay between the critical cooling rate of the alloy and established thermal management in the print. Thin features and walls cool quickly and therefore vitrify with ease; thicker cross-sections have slower heat loss, greater thermal retention, and a higher probability of crystallization. Consequently, the ability to maintain entirely amorphous structures in large or complicated structures remains a bottleneck in central processing [128].

Surface to volume cooling is controlled by small melt pools solidifying in microseconds and larger ones solidifying more slowly, and frequently below the critical cooling rate of the alloy. In the case of Cu-, Fe-, and Zr-based systems there is experimental evidence that amorphous structures can only be maintained in the form of layers up to a few hundred micrometers before devitrification starts. Scanning tracks overlap thermally in thicker builds, resulting in cumulative heating, which in turn results in nanocrystallization along the HAZ [16].

To extend Dₘₐₓ, researchers have adopted beam shaping, high speed scans, and multi-laser remelting plans in order to improve local cooling or heat redistribution. Although these are short-term ways of adding amorphous volume, they can compromise dimensional accuracy and create residual stress or pores. State-of-the-art designs incorporate internal cooling channels or hybrid BMG-crystalline designs, in which amorphous materials provide hardness and corrosion resistance, and crystalline domains manage mechanical load-bearing requirements [54].

Simply, it is still not possible to accomplish large-scale amorphous structures and complex geometries in AM. It requires alloy composition, process parameter, and thermal control plan optimalization to balance all three aspects of cooling kinetics, geometric faithfulness, and amorphous steadiness [129].

7.4. Feedstock and Environmental Control

The positive results of the AM of BMGs are critically dependent on the properties of powder feedstock (particle size distribution, morphology, surface oxides, and internal porosity) and environmental control (purity of inert gases and vacuum). Surface oxides and trapped gases enhance the tendency to porosity and modify melting pool fluid dynamics, which may alter local chemistry; either are likely to catalyze crystalline phase nucleation and worsen functioning properties. One common issue of oxygen-sensitive chemistries (e.g., systems based on Zr-, Ti-, Cu-based metals) is that oxygen is recaptured during atomization, powder handling, or reuse. Maintaining low oxygen content, tightly controlling powder reuse, and adequate storage/handling are required to prevent failures caused by contamination. In addition, feedstock variability between powder batches hinders process transfer and ensures reproducibility [130,131].

7.5. Vertical Structural Inhomogeneity

AM structures often have depth-related changes in amorphous fraction: upper layers (cooling more quickly) can be completely glassy, and mid-height areas or the HAZs partially crystallized because of increased reheating. The result of this vertical inhomogeneity is the creation of spatially varied mechanical and functional qualities (e.g., strength, magnetic response, toughness), which makes it difficult to qualify as an engineering material [132]. This is due to the inhomogeneity of thermal history aspects driven by local geometry (heat sink, overhang) and process settings, and solving this problem requires process (adaptive scan strategy, interlayer cooling control) and material (larger supercooled liquid tolerances, more resistance against annealing-induced crystallization) design. Closed-loop control with real-time observation is potentially effective, but is in its infancy with regard to habitual mitigation of vertical inhomogeneities [130].

7.6. Process Constraints, Mitigation Strategies, and Performance Trade-Offs in AM of BMGs

Additive manufacturing of bulk metallic glasses is constrained by thermal accumulation, residual stresses, geometric limitations, and feedstock quality. These factors collectively influence devitrification, defect formation, and structural integrity. Targeted process interventions such as energy optimization, scan strategy control, and environmental regulation enable improved amorphous retention, though often introduce trade-offs in processing complexity, cost, and dimensional accuracy [15]. The interrelationship between thermal accumulation, defect formation, feedstock quality, and mitigation strategies can be systematically understood through key quantitative metrics, as summarized in Table 7.

Table 7. Key variables, mitigation strategies, and their effects on amorphous retention and defects in AM of BMGs.

Key Quantitative Metric	Intervention	Improved Metric (s)	Potential Trade-Offs	References
Thermal Accumulation and Devitrification
Oxygen content thresholds	Oxygen contamination from AM environment or material oxidation during processing	Increased crystallization susceptibility, reduced GFA	Lower GFA leads to higher crystallization and lower amorphous content	[60]
Allowable crystallized volume fraction	Adjusting laser energy density (Ed) and scanning speed (laser power, scan speed)	Control over amorphous vs. crystalline phase formation	High energy densities induce crystallization; low densities promote amorphousness	[78]
Porosity limits	Reduced porosity via adjusting laser settings, scan strategies, and powder quality (e.g., lower gas absorption)	Improved mechanical properties (e.g., higher tensile strength, fracture toughness)	High porosity weakens strength; too low energy can cause poor densification	[78]
Residual stress	Multiple scan strategies and low-energy passes for stress relief	Reduced residual stresses and crack formation	Increased scan strategies lead to longer processing times and added complexity	[60]
Thermal stability	Control cooling rates by adjusting laser power and scan speed	Improved resistance to devitrification and better phase stability	Increased energy density raises cooling rates, affecting crystallization	[36]
Residual Stress and Cracking
Residual stress	Substrate preheating, scan speed optimization	Reduced residual stress	Increased process time	[126]
	Post-build HIP, remelting	Stress relaxation, reduced crack formation	High cost, potential for dimensional distortion	[79,126]
Crack initiation	Homogenization of temperature, remelting runs	Reduced crack formation	May lead to lower cooling rates and geometrical inaccuracies	[79]
	Stress optimization, optimized scan strategies	Reduced crack propagation	Increased time for optimization	[126]
Geometric and Thickness Limitations
Maximum printable amorphous thickness (Dmax)	Beam shaping, high-speed scans, and multi-laser remelting	Improved cooling rates, enhanced amorphous volume	Reduced geometric accuracy, increased residual stress, possible pore formation	[128]
Cooling rate	Optimization of scan strategies (e.g., hatch pattern, laser energy density)	Better control over cooling rates to maintain amorphous structure	Potential increase in residual stress	[129]
Crystallization temperature (Tx)	Alternate direction of hatch, minimizing dwell times, beam shaping	Higher crystallization temperatures, reduced formation of crystals	Potential decrease in mechanical properties	[54]
Thickness of amorphous layer	Use of beam shaping, higher speed scans, internal cooling channels, hybrid designs	Increased allowable thickness for amorphous regions	Possible structural weaknesses, dimensional inaccuracies	[54]
Feedstock and Environmental Control
Oxygen content in powder	Control oxygen content during powder handling and atomization	Increased GFA, improved mechanical properties	Increased production complexity, added cost for controlling oxygen levels	[130]
Powder flowability	Use of flow aids (e.g., SiO2 nanoparticles) and optimizing powder particle size distribution	Improved powder flowability, better powder layer uniformity during laser processing	Possible introduction of impurities (SiO2), leading to contamination and reduced amorphization	[130]
Porosity limits	Adjust laser power and scan velocity to control porosity in the melt pool	Reduced porosity, better densification, and structural integrity	Excessive energy input can lead to crystallization, reduced geometric fidelity, and higher porosity	[130]
Vertical Structural Inhomogeneity
Viscoelastic heterogeneity	Control of laser scanning patterns and energy density (LED) to optimize melt pool dynamics	Improved mechanical uniformity and homogeneity in phase transition across the structure	Potential increase in process time and complexity; may lead to minor dimensional changes in fine features	[130,132]
Porosity and density distribution	Optimized laser energy density (LED), Marangoni convection control, and gas bubble management in the melt pool	Reduction in porosity, improved relative density and surface finish	Higher energy input can cause uneven melt flow and promote defects like bubble agglomeration	[130,132]
Shear transformation zones (STZs)	Increase cooling rate and reduce the impact of thermal gradients by improving scanning strategies	Enhanced mechanical properties, more stable amorphous structure	Risk of uneven crystallization, reduction in material toughness, and brittleness if not carefully controlled	[132,133]

The combination of energy density, cooling rate, and thermal accumulation is one of the determinants of the practical processability window of the BMGs in the various additive manufacturing methods. Among the other processing methods, LPBF normally features the largest processing window because of its inherently high cooling rates, which allows crystallization to be effectively suppressed as well as maintaining high densification. Conversely, EBM works at greater effective energy contributions because of the preheating of powder beds, which lowers residual stress but raises the vulnerability to partial crystallization. DED processes, which have larger melt pools and slower cooling rates, also need further process improvements including pulsed energy input or beam oscillation in order to exploit viable amorphous processing regimes. The differences indicate the existing trade-offs of thermal stability, process efficiency, and structural integrity in AM of BMGs [19,71,111].

8. Applications of Additively Manufactured BMGs

Additive manufacturing has rendered BMGs no longer laboratory curiosities but performative and architecturally complex materials with increasing industrial importance. Combining high specific strength, high elastic strain limits, corrosion resistance, and design-like fabrication, AM-fabricated BMGs make it possible to create design-driven applications that take advantage of inherent material superiority, as well as geometric freedom. These materials are finding their way into high-value, performance-critical systems as they become more stable when processing and possess greater amorphous retention [49,69,134]. Figure 11 represents the applications of additively manufactured BMGs.

Figure 11. Typical applications of additively manufactured BMGs.

8.1. Aerospace, Defense, and Biomedical Systems

Additive manufacturing of BMGs has opened doors into aerospace, defense, and biomedical applications by exploiting the combination of high strength-to-weight ratio, large elastic strain limit, corrosion resistance, and near net-shape production of the materials. These characteristics render AM-produced BMGs superior when it comes to components requiring an extreme level of mechanical performance and durability in harsh environments [19,72].

BMGs are used in lightweight mechanisms, vibration isolators, energy-absorbing mounts and precision actuators in the aerospace and defense domains since they can store and release large quantities of elastic energy without being permanently damaged. They have high specific strength which tends to exceed 2 GPa at densities below 7 g/cm³, which is useful in reducing the mass of structures without affecting the load-bearing capacity—significant in aerospace linking systems, deployable structures, and impact absorbers. AM laser processes, such as SLM and LPBF, allow fabrication of complex and integrated geometries, which was not feasible previously both by casting process implementation or by machining. To validate the somewhat amorphous alloys, manufacturers have established that AM BMGs can be utilized in space to produce and fix spacecraft parts [19,76].

In the case of BMG’s aerospace application qualification, issues include fatigue resistance, thermal cycling endurance, and radiation stability. The amorphous structure should be preserved when subjected to working thermal loads and cosmic radiation. It is important to optimize the laser energy, scan speed, and interlayer cooling of AM process windows so that crystallization that would diminish mechanical and magnetic properties is not achieved. Zr-, Ti-, and Cu-based BMGs have been shown to possess exceptional biocompatibility, corrosion resistance, and mechanical compatibility with human tissue in biomedical applications. AM enables patient-specific implants, such as porous bone scaffolds and load-bearing fixation devices, where porosity and topology can be tailored to match the modulus of natural bone and minimize stress shielding [69,135].

8.2. MEMS, Sensors, and High-Performance Tooling

Thermoplastic formability and high elastic limits can provide BMGs with a proprietary niche in microsystems: MEMS, micro-actuators, and precise sensors can take advantage of the repeatable elastic deformation, high fatigue limits, and small feature replication of BMGs. MEMS elements with extremely high resolution and functional activity (e.g., micro springs, resonators) have been produced by thermoplastic forming and by micro-scale AM (or hybrid AM + micro-machining) using the capacity of BMGs to perfectly replicate mold structures and store elastic energy without plastic deformation [136]. With tooling and high-strength fixtures, BMGs printed via LPBF can provide services for small tooling inserts with an excellent wear resistance and dimensional accuracy, punches and tooling with high-stiffness jigs and surface finishing, and HIP actions aimed at eliminating porosity, as well as reducing stress concentrators. AM enables the rapidity of the design iteration process by conformal cooling and lattice-reinforced tooling designs which take advantage of the high hardness and fatigue strength of BMGs to increase the service life of the tool in harsh conditions [137].

8.3. Energy and Extreme Environment Components

Additive-manufactured bulk metallic glasses are increasingly being proposed to be used as components in chemically hostile, high-pressure, or wear-intensive environments [138]. The amorphous structure removes grain boundaries which normally provide localities of corrosion in crystalline-type alloys, allowing homogeneous formation of passive film and greater resistance to localized corrosion. Additive manufacturing can be used to make complex fluid handling parts, marine connectors, hydrogen-exposed fittings and chemical processing parts with internal channels and optimized geometries that are challenging to fabricate through casting [72,134].

The high hardness of BMGs and their erosive and sliding wear resistance are used in pump housings and abrasive flow components, and their high hardness and elastic strain limits are utilized in slurry transport systems. It is worth noting that AM provides the local thermo-hydraulic control and compositional adjustment to preserve corrosion resistance and structural integrity. However, long-term thermal treatments at amorphous stability are also of significance, particularly in energy applications and offshore uses where cyclic loads and temperature variation cannot be avoided. Increased improvements in process control and purity of the feedstock will become the force that will allow AM-produced BMGs to graduate from laboratory-level demonstrations and begin years of service under harsh conditions [134,139].

9. Future Perspectives and Research Directions

9.1. Future Perspectives

Alloy design, increased control over processing, and application-driven engineering approaches will be convergent factors that determine the future development of additively manufactured bulk metallic glasses (BMGs), as shown in Figure 12. Integrated design–process frameworks, scalable manufacturing routes, and predictive modeling tools that can be used to stabilize the amorphous structure with complex thermal histories are needed [140,141]. Simultaneously, the research on new opportunities for multi-material builds, ecological manufacturing, and biomedical translation indicate the necessity of a cross-functional cooperation and a common set of qualifications. Further development will be based on the linkage of material innovation to digital manufacturing, real-time tracking, and lifecycle-conscious design to make the deployment of AM-fabricated BMG components reliable and large scale.

Figure 12. Future perspectives for additive manufacturing of BMGs.

9.1.1. Design–Process Integration for Industrial Scale-Up

Design, materials and process engineering should be closely integrated to scale AM-manufactured BMG parts to production needed in an industrial environment based on laboratory demonstrations. Design choices have to encode thermal inertia of vitrification (critical cooling rates, location of heat sinks, division of large cross-sections) such that geometry, scan strategy, and fixture design are not independently designed [140]. Deliberate digital twins that integrate melt-pool thermal models with crystallization kinetics would make it possible to run the process operations at scale, as well as to qualify parts, by forecasting where either reheating or thermal storage would pose a risk to amorphous storage [142].

9.1.2. Multi-Material and Graded Architectures

The exceptional strength/thermoplastic formability of BMGs allows them to be considered in the contexts of multi-material and functionally graded constructions in the sense that distinct parts of a structure require distinct amounts of stiffness, toughness, or functional performance. Glass, crystal composites, ductile-layered structures or localized forms of crystalline reinforcements leading to the arrestment of shear-bands and enhanced global toughness can be produced using AM techniques that allow local compositional control (often tunings) or graded deposition (e.g., hybrid LPBF/DED approaches, powder-to-parcel control). Realizing these concepts at scale requires solving material science challenges (i.e., avoiding detrimental segregation and controlling local cooling rates) and developing robust process methods and monitoring in order to guarantee repeatability [143].

9.1.3. Sustainability and Standardization

The future of additively manufactured BMGs will be based on sustainability and supply-chain resilience. The life-cycle analysis of BMGs highlights the potential benefits, reduction in material waste, and lack of tooling, but distinguishes such obstacles as high energy intensity, the impracticality of reusing the powder, and contamination, which could deteriorate the outcomes of BMGs. Circular economy practices (managed reuse of powder, closed-loop recycling, and energy-saving planning of processes) should be demonstrated as specifically implementable in BMG chemistries because even small substitutes in the degree of oxygen or the form of the particles can result in devitrification or degradation of the functions [144,145]. At the same time, standardized testing techniques, material data reports, and qualification roadmaps (standardized in the ISO/ASTM standards and ANSI/America Makes roadmaps) are needed to facilitate industry adoption: these will provide standard definitions, testing protocols of feedstock and parts, and directions toward certification in regulated industries. Liaisons between standards organizations, research centers, and manufacturing will expedite the dissemination of technologies between demonstration on the benchtop and certified manufacturing [146,147].

9.1.4. Clinical Translation and Biocompatible BMG Design

Biomaterial bulk metallic glass presents considerable potential due to their strength, resistance to corrosion, and their elasticity in compatibility with bone. This is further supported by tribocorrosion research in simulated body fluid that demonstrates the promising wear-corrosion behavior of Zr-based BMGs in physiological conditions [148]. However, these biomedical implementations are not yet widespread due to a number of significant challenges, such as limited printable size, compositional toxicity related to some elements (Ni or Be) in some high-GFA alloys, and the threat of embrittlement due to structural relaxation or thermal treatment.

In addition, the lack of long-term fatigue data, missing in vivo validation, and the further lack of regulatory standardization still limit clinical translation. The creation of fully biocompatible alloy systems; further development of GFA to facilitate the manufacture of components of larger size; improvement in intrinsic toughness; and adoption of more sophisticated surface engineering techniques to enhance osseointegration and biological stability, can therefore be identified as future research priorities. The overall optimization of additively manufactured BMG components, including alloy chemistry, AM processing stability, standardized processing pathways, and overall long-term performance validation, will eventually be necessary to achieve successful biomedical implementation of the additively manufactured component [149].

9.2. Research Directions

The additive manufacturing of BMGs have numerous key routes of development, as indicated by the research directions presented in Table 8. Altogether, the literature indicates a concerted necessity of materials innovation, process optimization, and further mechanistic knowledge. Some of the critical priorities include the large-scale manufacture of fully amorphous, dense, and defect-free components by alloy design refinement, increasing glass-forming capacity, and enhancing thermal stability. At the process level, future work should focus on in situ monitoring, closed-loop control, and optimization of the parameters with the help of AI/ML in order to decrease the number of defects and enhance reliability. Other valuable guidelines are beam-shaping strategies, scaling of laboratory coupons with larger parts, mitigation of oxygen contamination, and enhancement of fatigue performance [110,150]. There is also an increasing interest in predictive multiscale modeling and spatially resolved characterization of structural heterogeneity, as well as predicting crack suppression in marginal glass-forming systems. Lastly, new opportunities exist in functionally graded and multi-material BMG designs and composite AM architectures, in addition to combined AM methods, to learn more about novel AM-compatible glass-forming alloys.

Table 8. Future research directions in additive manufacturing of bulk metallic glasses [15,20,47,54,60,110,150,151].

Future Direction	Research Focus
Alloy and process co-design for robust vitrification	Co-design alloys and process windows with similar glass-forming capability and thermal stability to allow amorphous builds of high density with reduced defects
In situ monitoring and closed-loop process control	Real-time sensing and data-driven control can be used to identify defects during the building process and decide parameters or remelt at the locality before defects build up
Laser beam shaping and energy distribution control	Localized energy and tailored beam shape distribution to control melt-pool behavior and thermal gradients, improving density, amorphization, and mechanical performance
Scale-up and qualification of larger components	Confirm that parameters that have been tested on small coupons still work on large parts, and measure size-dependent changes in thermal history, microstructure and properties
Feedstock quality, oxygen control, and fatigue reliability	Improve powder quality and oxygen uptake, create alloys that are more oxygen tolerant, and create qualification paths based on fatigue to achieve long-term reliability
Multiscale thermal and crystallization modeling	Develop predictive models that couple melt pool flow, heat transfer, reheating cycles, and crystallization kinetics to guide parameter selection and scan strategy
Spatially resolved structure–property characterization	Create maps regarding build height and location-dependent variations in glassy structures and local properties to clarify how repeated thermal cycling affects performance
Crack suppression in lower-GFA systems	Develop preheating, substrate engineering, and multi-step scanning strategies to suppress cracking while maintaining high density and vitrification in lower GFA alloys
Graded, multi-material, and composite BMG architectures	Increase AM pathways of graded, layered, multi-material, and composite BMG systems that are amorphous matrices with ductile or functional phases
High-throughput discovery of AM-compatible glass formers	Screen compositions and find glass-forming alloys that can be used in AM thermal conditions on a large scale with AM using combinatorial and direct-write AM.

10. Conclusions

Additive manufacturing has become a radical route for the creation of BMGs with multifunctional shapes and performance. Recent developments of LPBF, EBM, DED, and hybrid AM methods have dramatically increased amorphous retention and dimensional and mechanical reliability by means of improved thermal control, beam shaping, and process control. The combination of real-time surveillance, computational modeling, and alloy design solutions tailored to AM have enlarged the usefulness of BMGs through bypassing the constraints of traditional thermoplastic forming or casting paths. Although this has been achieved, there are still critical issues such as thermal accumulation-based devitrification, development of residual stress, sensitivity of feedstock, limitations of geometric thickness, and vertical structural inhomogeneity. Scalable and reproducible AM of BMGs would thus necessitate a high level of process–structure–property modeling, control of powder quality, and efficient post-processing treatments of HIP as a remedy to defects and to maintain amorphous integrity. In the future, the integration of information-based alloy designs, self-regulated in situ processing, and sustainable manufacturing plans will be necessary to ensure that AM-fabricated BMGs are not limited to laboratory-level experiments but also provide solid industrial applications. Further convergence of material science, digital manufacturing, and predictive engineering platforms will finally enable strong, scalable and application-ready amorphous metal components in the aerospace, biomedical, energy, and tools industries.