Bulk glasses are solid materials whose atoms are frozen in a disordered, liquid-like arrangement rather than the neat crystalline lattice you find in ordinary metals or ceramics. The term “bulk” distinguishes them from thin films and ribbons by a practical threshold: they are thick enough to hold in your hand and use as structural parts, generally at least a millimeter across and often much larger. Most research and commercial interest centers on bulk metallic glasses (BMGs), alloys that combine the strength of metal with the formlessness of window glass, but the family also includes non-metallic compositions like chalcogenide glasses used in optics. What makes them fascinating, and frustrating, is that their unusual atomic disorder gives them remarkable mechanical and chemical properties while simultaneously making them difficult to manufacture at useful sizes.
What Makes a Glass “Bulk”
Scientists have been making metallic glasses since the 1960s, but the early versions were vanishingly thin: ribbons and foils produced by splashing molten metal onto a spinning copper wheel at cooling rates of a million degrees per second or more. That extreme cooling was necessary to freeze the atoms before they could organize into crystals. The material was genuinely glassy, but you couldn’t make a gear or a phone case out of a foil thinner than a human hair.
The breakthrough came in the 1990s, when researchers found alloy compositions that resisted crystallization so strongly they could be cooled at rates below 100 degrees per second and still remain amorphous, with dimensions of a centimeter or more.1Intermetallics. Bulk metallic glasses That jump from micrometer-thin ribbons to centimeter-scale rods and plates is the dividing line between metallic glass and bulk metallic glass. Some compositions have since been cast into fully amorphous pieces up to 75 mm thick.2Proceedings of the Japan Academy, Series B. Stabilization of Supercooled Liquid and Opening-up of Bulk Glassy Alloys At those dimensions, the material stops being a laboratory curiosity and starts being a potential engineering material.
The Recipe for Disorder
Not just any combination of metals will form a glass when cooled. Decades of experimentation have converged on three empirical rules that alloy designers follow to maximize the chance of getting an amorphous solid instead of a crystalline one. The alloy needs at least three different elements. Those elements should differ in atomic size by more than about 12 percent. And they should have negative heats of mixing, meaning their atoms prefer each other’s company over clustering with atoms of the same kind.3Proceedings of the Japan Academy, Series B. Stabilization of Supercooled Liquid and Opening-up of Bulk Glassy Alloys
The logic behind these rules is intuitive once you picture what happens as liquid metal cools. If all the atoms are roughly the same size and only one or two elements are present, they easily stack into a crystal lattice, the way identical marbles settle into neat rows in a box. Throw in atoms of very different sizes and chemical affinities, and the liquid becomes confused: no single crystal structure satisfies everyone, so the atoms jam into a disordered arrangement and stay there. Researchers have classified the BMGs discovered so far into seven groups based on atomic size difference, mixing energy, and where the elements sit on the periodic table.4Materials Transactions. Classification of Bulk Metallic Glasses by Atomic Size Difference, Heat of Mixing and Period of Constituent Elements and Its Application to Characterization of the Main Alloying Element
The most commercially explored families include zirconium-based, copper-based, iron-based, and palladium-based systems. Zirconium alloys (often with copper, nickel, aluminum, and sometimes titanium or beryllium) have been the workhorses of the field because they form glasses relatively easily at modest cooling rates and offer a good balance of properties. Iron-based BMGs attract attention for their magnetic behavior, and precious-metal systems like palladium-based glasses are prized as model systems for fundamental research even though nobody is going to make budget consumer products out of palladium.
What the Atoms Actually Look Like Inside
Saying the structure is “disordered” doesn’t mean the atoms are scattered randomly like gas molecules. There is short-range order: atoms cluster into small motifs, often icosahedral clusters where one atom sits at the center surrounded by a shell of neighbors. In a copper-zirconium-aluminum BMG, for instance, researchers resolved the atomic-level structure and identified copper- and aluminum-centered icosahedral clusters as the basic building blocks. Adding even a small percentage of aluminum to a simpler copper-zirconium binary glass dramatically increased the number of these full icosahedra and how extensively they connected to each other.5PubMed. Atomic level structure in multicomponent bulk metallic glass
Those interconnected clusters are part of what makes the glass resistant to crystallization. The icosahedral arrangement has fivefold symmetry, which is incompatible with the repeating, space-filling patterns that crystals require. The more of these clusters the liquid contains, the harder it is for the atoms to reorganize into a crystal during cooling. This is essentially why the composition rules work: the right mix of atom sizes and chemical interactions promotes icosahedral packing in the liquid, which then gets frozen in place when the alloy solidifies.
Strength Without Ductility
The mechanical profile of bulk metallic glasses reads like a wish list with one glaring asterisk. They are remarkably strong, often exceeding 1.5 to 2 gigapascals in compressive strength. They can elastically deform to about two percent strain before yielding, roughly double what conventional crystalline metals allow. And they are extremely hard, making them resistant to scratching and wear. But they have essentially zero tensile ductility: pull on a BMG sample and it fractures almost immediately after yielding, with no warning.6PubMed Central. Bulk metallic glass composite with good tensile ductility, high strength and large elastic strain limit
In a crystalline metal like steel or aluminum, plastic deformation happens through the movement of dislocations, line defects in the crystal lattice that can slide past each other gradually. Bulk metallic glasses have no crystal lattice, so they have no dislocations. Instead, when stress exceeds a critical level, deformation concentrates into extremely thin shear bands, typically nanometers wide, where atoms rearrange catastrophically. One or two of these bands can run across the entire sample and cause fracture almost instantly. The process is driven primarily by stress rather than by heat: experiments coating BMG samples with low-melting-point metals like tin and indium showed that the coatings did not melt near individual shear bands, only near the final fracture surface where bands concentrated.7PubMed Central. Localized shear deformation and softening of bulk metallic glass: stress or temperature driven? This confirmed that individual shear bands form under mechanical stress, with significant heating happening only as a secondary consequence when many bands pile up.
The brittleness problem is the single biggest barrier to structural applications. A bridge beam or an aircraft bracket that fails without warning is unacceptable, no matter how strong the material is. The research community has attacked this from several angles. One strategy is to create BMG composites, embedding a crystalline phase (like a ductile metal dendrite) within the glassy matrix. The crystalline inclusions interrupt shear bands and force the material to form many small bands instead of one catastrophic one, which dramatically improves tensile ductility while retaining much of the strength and elastic strain limit.8PubMed Central. Bulk metallic glass composite with good tensile ductility, high strength and large elastic strain limit
Corrosion and Wear Resistance
One of the most practically useful features of BMGs is their resistance to chemical attack. Crystalline metals corrode preferentially along grain boundaries, dislocations, and other defects because these sites are more chemically reactive. Bulk metallic glasses have none of these features: no grain boundaries, no dislocations, no second-phase particles segregating at interfaces. Their chemical homogeneity leaves corrosion with nowhere easy to start.9Elsevier / Journal of Alloys and Compounds. Corrosion behavior of Zr–Cu–Ni–Al bulk metallic glasses in chloride medium
Wear resistance follows a similar logic. The high hardness means the surface is difficult to scratch, and the absence of grain boundaries eliminates weak paths for material removal. Testing on zirconium-based BMGs has shown that heat treatment to introduce some structural heterogeneity can actually improve wear performance further. In one study, treated samples had about a third lower wear rate in dry sliding compared to as-cast samples, and the gap persisted in saltwater conditions as well.10MDPI. The Tribological Behaviors in Zr-Based Bulk Metallic Glass with High Heterogeneous Microstructure The combination of corrosion resistance and wear resistance makes BMGs attractive for components exposed to harsh environments, such as surgical instruments, chemical processing equipment, and marine hardware.
Soft Magnets and Other Functional Properties
Iron-based bulk metallic glasses have attracted special attention for their magnetic behavior. A good soft magnetic material magnetizes easily in a weak field and demagnetizes just as easily when the field is removed, which is exactly what you want in transformer cores, inductors, and other power-conversion components. Crystalline silicon steel has dominated this space for decades, but BMGs can match or beat it in certain respects. Ternary iron-boron-based BMGs have demonstrated saturation magnetization of 1.56 tesla with coercivity below 40 A/m, meaning they magnetize strongly and release that magnetization with very little energy loss.11Applied Physics Letters. Soft magnetic ternary iron-boron-based bulk metallic glasses Their electrical resistivity, measured above 200 microhm-centimeters, is also an advantage: high resistivity suppresses eddy currents, reducing energy waste as heat in alternating-current applications.
The size limitation remains relevant here. A transformer core needs to be large, and iron-based BMGs are among the hardest to cast in bulk because iron is an excellent crystal-former. The iron-boron glasses mentioned above could be made amorphous in rods at least 1 mm in diameter and plates at least 0.5 mm thick.12Applied Physics Letters. Soft magnetic ternary iron-boron-based bulk metallic glasses That’s fine for small power electronics but far from the massive laminated cores in utility transformers. Still, in consumer electronics, sensors, and micro-electromechanical systems, those dimensions are more than adequate.
Shaping Bulk Glasses After Casting
One of the more counterintuitive properties of BMGs is that they can be reshaped after they’re made, if you heat them back to the right temperature. Every glass has a transition temperature where it softens into a viscous supercooled liquid. In that state, the material flows like warm taffy and can be pressed, blown, or stamped into complex shapes using modest forces. The window between the glass transition temperature and the crystallization temperature is the supercooled liquid region, and in some alloy systems it spans more than 100 degrees, giving a comfortable processing window.
Iron-based BMGs, for example, show superplastic behavior in this region. Hot compression tests on one iron-based glass revealed flow stress dropping systematically with increasing temperature and decreasing strain rate, with the material behaving in a way consistent with superplasticity in the supercooled liquid region.13Elsevier. Thermoplastic deformation behavior of a Fe-based bulk metallic glass within the supercooled liquid region This thermoplastic forming approach has been explored for making micro-parts, surface textures, and complex geometries that would be impossible to achieve by casting alone. The catch is time: spend too long at elevated temperature and the atoms start organizing into crystals, destroying the glassy structure and the properties that come with it.
3D Printing Amorphous Metal
Additive manufacturing, particularly laser powder bed fusion, has emerged as a promising route around the traditional size limitations of BMGs. In this process, a laser selectively melts thin layers of metallic glass powder, and each tiny melt pool solidifies so quickly that the atoms don’t have time to crystallize. Cooling rates in laser powder bed fusion can reach on the order of a hundred million degrees per second, far exceeding what’s needed even for alloys with moderate glass-forming ability.14Europe PMC / MDPI Materials. Mechanical Properties of Bulk Metallic Glasses Additively Manufactured by Laser Powder Bed Fusion: A Review
The appeal is obvious: you could build complex-shaped BMG parts layer by layer, bypassing the need to pour molten alloy into a mold and cool it uniformly. Research on 3D-printed zirconium-based BMGs has shown that the process can produce highly dense structures, above 99.8 percent of theoretical density, while retaining an amorphous structure as measured by X-ray diffraction. The technique also offers a surprising degree of control: by adjusting laser parameters, researchers can tune the degree of structural relaxation and even dial in small amounts of nanocrystallinity throughout a part, potentially tailoring mechanical properties for specific applications.15Additive Manufacturing. Semi-analytical and experimental heat input study of additively manufactured Zr-based bulk metallic glass: Insights into nano- and global-scale relaxation and crystallization
The field is still working through challenges. Each layer of the print is reheated when the next layer is deposited, and that repeated thermal cycling can relax or partially crystallize the material in ways that are hard to predict. But the trajectory is encouraging, and additive manufacturing may ultimately be the technology that moves BMGs from specialty components into wider engineering use.
Using Machine Learning to Find New Compositions
Discovering a new BMG composition has traditionally been slow, empirical work: melt a candidate alloy, quench it, check whether it’s amorphous, repeat with a different composition. The design space is enormous. Even limiting yourself to practical elements, the number of possible ternary, quaternary, and quinary combinations runs into the millions. Machine learning has begun to accelerate this search by training models on the database of known glass-forming alloys and then predicting which unexplored compositions are most likely to form bulk glasses.
One study trained multiple machine learning models on known metallic glasses and then swept the entire composition space spanned by ternary combinations of 24 practical elements, identifying novel compositions with a predicted probability above 95 percent of being a BMG.16Acta Materialia. Machine learning versus human learning in predicting glass-forming ability of metallic glasses The interesting wrinkle is that these predictions sometimes pointed to compositions very different from anything in the training data, suggesting that human intuition and established empirical rules might be missing pockets of composition space where good glass formers lurk. The approach doesn’t replace experimental validation, but it narrows the search from millions of candidates to a manageable handful.
High-entropy bulk metallic glasses represent another frontier being explored partly through computational methods. These alloys contain five or more principal elements in roughly equal proportions, borrowing the design philosophy from high-entropy crystalline alloys. The massive atomic size mismatch in such systems can push lattice distortion past the point where a crystal structure is stable, collapsing the arrangement into an amorphous one.17PubMed Central. Designing High Entropy Bulk Metallic Glass (HE-BMG) by Similar Element Substitution/Addition Whether these materials deliver property combinations that simpler BMGs cannot is still an open question.
The Scaling Problem
For all their impressive laboratory properties, bulk metallic glasses have had a rocky road to commercial viability. The fundamental issue is cost. Making BMGs requires high-purity feedstock, precise alloy control, and specialized casting or processing equipment. Early commercial ventures, notably Liquidmetal Technologies, found that manufacturing processes were not yet refined enough to produce price-sensitive commodity products cost-effectively.18Materials Today. Features The case for bulk metallic glass The company’s own leadership acknowledged that unpredictable customer adoption cycles and intense pricing pressure made it difficult to compete in commodity markets.
The result has been a pivot toward niche, high-value applications where the unique properties justify the premium. Consumer electronics casings, precision watch components, medical device parts, and sporting goods (notably golf club faces) have been the most visible commercial uses. In these markets, the combination of high strength, elasticity, surface finish quality, and corrosion resistance provides enough of an advantage over conventional metals to absorb the higher manufacturing cost. But nobody is building car body panels or bridge girders out of BMG, and that situation is unlikely to change soon without a fundamental shift in production economics.
Beyond Metals
The phrase “bulk glass” doesn’t always mean metallic. Chalcogenide glasses, based on sulfur, selenium, and tellurium with additions of elements like germanium, arsenic, and antimony, are an entirely separate family of bulk glasses with their own set of applications. These glasses are transparent from the visible spectrum well into the infrared, making them valuable for thermal imaging optics, fiber-optic cables for mid-infrared wavelengths, and chemical sensors.19Elsevier. Optical properties and applications of chalcogenide glasses: a review Their low phonon energy means they don’t absorb infrared light as strongly as oxide glasses like silica, which is why your standard window glass is opaque in the thermal infrared while a chalcogenide lens can see through it clearly.
Chalcogenide glasses are also used in phase-change memory devices, the technology behind rewritable optical discs and some types of solid-state storage. In these applications, tiny spots of the glass are switched between amorphous and crystalline states using laser or electrical pulses, with the two states representing digital ones and zeros. The speed and reversibility of this switching depend on the same glass-forming physics that governs BMGs: how easily the material transitions between ordered and disordered atomic arrangements. The overlap in underlying science, even though the materials and applications are wildly different, is one reason researchers in both communities pay attention to each other’s work.
Biomedical Glasses
Magnesium-based bulk metallic glasses have drawn interest as biodegradable implant materials. Magnesium corrodes naturally in the body, which is a liability in most engineering contexts but an advantage for temporary implants like bone screws or stents that should dissolve after the tissue heals. The amorphous structure of a magnesium-zinc-calcium BMG gives it higher strength and more uniform degradation compared to crystalline magnesium alloys, where corrosion tends to attack grain boundaries unevenly and can cause the implant to weaken unpredictably. Researchers have explored adding titanium particles to these glasses to form composites that fine-tune the degradation rate and mechanical strength.20Journal of Alloys and Compounds. Degradation behavior and mechanical strength of Mg-Zn-Ca bulk metallic glass composites with Ti particles as biodegradable materials
The concept of a structural material that is intentionally designed to disappear runs against every instinct in traditional materials engineering, but for biomedical applications it’s exactly what surgeons want. A patient who receives a biodegradable BMG bone screw avoids a second surgery to remove the hardware. The challenge is controlling the degradation timeline: too fast and the implant fails before the bone heals, too slow and you might as well have used a permanent material. Alloy design in this space is essentially glass-forming-ability optimization with the additional constraint that every element must be biocompatible, which rules out many of the heavy metals that make the best non-biological BMGs.