Bulk metallic glass (amorphous metal) components

Executive Summary

Bulk metallic glasses (BMGs) are alloys that are cooled fast enough to avoid crystallization, freezing into a non-crystalline, amorphous atomic structure. The technology matters because amorphous metals can combine high yield strength, high hardness, a relatively large elastic strain limit, and the ability to be processed like a plastic in certain supercooled-liquid processing windows. The most famous early demonstration alloy is the zirconium-titanium-copper-nickel-beryllium system, particularly the alloy with composition Zr\(_{41.2}\)Ti\(_{13.8}\)Cu\(_{12.5}\)Ni\(_{10.0}\)Be\(_{22.5}\), often associated with the “Vitreloy 1” demonstration alloy Tier 1, Peker & Johnson 1993, DOI 10.1063/1.110520.

However, a technical capability is not the same as a commercial moat. The foundational patent covering beryllium-bearing bulk amorphous alloys, US 5,288,344 A, granted 1994-02-22, is a granted patent but is now expired. Peer-reviewed mechanical-property literature shows that bulk metallic glasses are strong but usually show near-zero macroscopic tensile ductility at room temperature because deformation localizes in shear bands Tier 1, Schuh et al. 2007, DOI 10.1016/j.actamat.2007.01.052. Liquidmetal Technologies’ public SEC filings show a company with limited revenue and going-concern-style risk disclosures, which is a significant contradiction when compared with marketing material that emphasizes moldable, production-ready amorphous parts Tier 2, SEC EDGAR LQMT archive.

This note deliberately distinguishes granted patents from patent applications and peer-reviewed papers from preprints. No arXiv preprint is used as a substitute for a journal article. The technical/performance claims below are cited to granted patents and DOI-verified journal publications. Adoption claims are cited to SEC filings and company documents, with self-reported items marked accordingly.

Source Key

A patent is evidence of recorded legal claims; it is not independent proof that a product performs as claimed or that a process is commercially viable.

What Bulk Metallic Glass Is

A conventional metal has a crystalline lattice with grains, grain boundaries, dislocations, and other microstructural defects. A bulk metallic glass is an alloy that is cooled from the melt fast enough to suppress crystallization, producing a metastable amorphous solid with short-range atomic order but no long-range crystalline lattice. The result is a material with no grain boundaries and no conventional dislocation-moderated plasticity Tier 1, Schuh et al. 2007, DOI 10.1016/j.actamat.2007.01.052.

The key enabling alloy family in this note is the beryllium-bearing Zr-Ti-Cu-Ni-Be family disclosed in the expired granted patent US 5,288,344 A, Peker and Johnson, granted 1994-02-22. The same lineage is peer-reviewed in Peker and Johnson, Applied Physics Letters 63, 2342 (1993), DOI 10.1063/1.110520. The patent and paper are from the same research lineage, so any load-bearing claim that rests only on both of these should be tagged [SINGLE-SOURCE] for institutional independence. Where possible, this note cross-checks with independent reviews from different institutions, such as Schuh et al. 2007 and Schroers, Advanced Materials 22, 1566 (2010), DOI 10.1002/adma.200902776.

Processing Physics: Why BMG Is Hard to Manufacture

The central processing problem for BMGs is avoiding crystallization. Crystallization is a kinetic competition: if the cooling rate is too slow, atomic diffusion allows the alloy to order into crystals. If the cooling rate is above a critical cooling rate, \(R_c\), the liquid can instead freeze into a glass at the glass transition temperature, \(T_g\).

Above \(T_g\), but below the crystallization temperature, \(T_x\), a BMG-forming alloy can exist in a supercooled liquid region. In this window, the material can flow viscously under moderate pressure, enabling net-shape processing through injection molding, die casting, and thermoplastic forming Tier 1, Schroers 2010, DOI 10.1002/adma.200902776. The foundational patent US 5,288,344 A claims beryllium-bearing alloys that can be cooled at low enough rates to form bulk sections rather than only thin ribbons.

The classic demonstration alloy from the Peker and Johnson lineage has a reduced glass temperature, \(T_g/T_l\), reported near 0.67, which is close to the empirical 2/3 design rule for glass-forming ability Tier 1, Peker & Johnson 1993, DOI 10.1063/1.110520. The exact supercooled liquid region and critical cooling rate should be read from the original paper; in this note, the quantitative calculation treats \(R_c\) as an assumed input because the precise value is not reproduced here from memory alone. Exact numeric values from the paper should be treated as needs-primary-source-validation if quoted without the PDF in hand.

Patent Landscape

5.1 Foundational granted patent

• Patent number: US 5,288,344 A
• Title: Beryllium bearing amorphous metallic alloys formed by low cooling rates
• Inventors: William L. Johnson and Atakan Peker
• Grant date: 1994-02-22
• Source: patents.google.com
• Access date: 2026-02-09

This is a granted patent, not an application. It supports the claim that beryllium-bearing Zr-Ti-Cu-Ni-Be alloys were disclosed as bulk-glass formers. However, because it was filed in the early 1990s and granted in 1994, its 20-year term has expired. Therefore, it supports historical feasibility and alloy-design lineage, but it does not provide a present-day exclusive moat for the original Vitreloy family.

5.2 Distinction from applications

No patent application is used in this note as Tier 1 proof of novel exclusivity. The strongest IP claim available is the granted patent above, and even that is expired. For any current commercial moat claim, a reader would need to retrieve and verify current Liquidmetal or third-party granted patents, not applications or press releases.

Peer-Reviewed Evidence and Contradictions

6.1 Processability claim

Peker and Johnson reported a highly processable Zr-based alloy in a peer-reviewed Applied Physics Letters paper DOI 10.1063/1.110520. The same alloy concept is described in the granted patent US 5,288,344 A. Because the patent and paper share institution and inventors, this is one lineage. Independent support for general BMG processability comes from Schroers’ 2010 review in Advanced Materials, which covers injection molding, thermoplastic forming, and net-shape processing DOI 10.1002/adma.200902776. The load-bearing claim that BMGs can be net-shape processed in the supercooled liquid region is therefore cross-verified by two independent source groups.

6.2 Mechanical performance claim

Peer-reviewed review literature states that BMGs can exhibit yield strengths approaching theoretical limits and elastic strain limits near 2%, but that plastic deformation is highly localized in shear bands, producing low macroscopic tensile ductility in many monolithic BMGs at room temperature Tier 1, Schuh et al. 2007, DOI 10.1016/j.actamat.2007.01.052. This is a critical contradiction with popular marketing that describes amorphous alloys as simply “strong and tough.” High compressive strength and high hardness are often accompanied by shear-band-dominated brittle-like tensile behavior.

Any claim that a specific BMG component solves the ductility problem would require a peer-reviewed, DOI-verified measurement from a second institution. That is not supplied here. Tag [SINGLE-SOURCE] if a commercial site asserts higher toughness than peer-reviewed literature.

6.3 Commercial-readiness contradiction

Liquidmetal Technologies’ SEC filings describe a company with limited commercial revenue and recurring losses. The exact wording should be pulled from the SEC EDGAR archive; this note does not quote the filing text from memory Tier 2, SEC EDGAR LQMT archive, accessed 2026-02-09. This contrasts with Liquidmetal’s own technology page, which describes injection-molded amorphous metal parts as a production technology Tier 2, self-reported, [liquidmetal.com . The discrepancy is not proof that the technology fails; it is evidence that technical capability and commercial scale are separate claims.

Company Filings Cross-Reference

The principal US-listed pure-play is Liquidmetal Technologies, Inc., whose SEC EDGAR filings are retrievable here:

• SEC EDGAR company archive for LQMT 10-K filings: sec.gov
• SEC EDGAR company archive for LQMT 8-K filings, including disclosures around a historic Apple license: sec.gov

Observed discrepancy: The company’s technical/marketing materials emphasize moldability and material advantages, while its SEC risk factors emphasize revenue concentration, limited operating history, and uncertainty about commercialization. A patent is evidence of recorded claims; a 10-K is evidence of actual audited financial condition. Both should be read together before assigning a moat.

Worked Quantitative Calculations

8.1 Calculation 1: Maximum section thickness from critical cooling rate

The goal is to estimate the maximum slab thickness that can bypass crystallization if the cooling rate must be at or above \(R_c\).

Formula:

\[

L_{\max} \approx \sqrt{\frac{10 \alpha \Delta T}{R_c}}

\]

This is a Fourier-conduction order-of-magnitude estimate. The factor 10 is a common engineering approximation for the first eigenvalue, \(\pi^2 \approx 10\). It assumes ideal surface quenching, no latent heat of crystallization, and no heterogeneous nucleation.

Inputs:

Calculation:

\[

10 \alpha \Delta T = 10 \times 3 \times 10^{-6} \times 700 = 0.021 \ \text{m}^2

\]

\[

\frac{10 \alpha \Delta T}{R_c} = \frac{0.021}{10} = 0.0021 \ \text{m}^2

\]

\[

L_{\max} = \sqrt{0.0021} \approx 0.0458 \ \text{m} = 45.8 \ \text{mm}

\]

Sensitivity:

Interpretation: The estimate is an ideal conduction limit. Real castings are smaller than ideal mainly because of mold-interface resistance, fluid flow, and heterogeneous nucleation. If the actual published critical cooling rate for Vitreloy 1 is lower than 10 K/s, the ideal thickness would be larger; if higher, smaller. The calculation demonstrates why thickness is a scarce resource in BMG component design.

8.2 Calculation 2: Injection-molding cooling-time estimate for a small consumer part

Assumed component: a small frame-like part with volume \(2 \ \text{cm}^3\), surface area \(100 \ \text{cm}^2\), density \(6.8 \ \text{g/cm}^3\).

Mass:

\[

V = 2 \times 10^{-6} \ \text{m}^3

\]

\[

m = \rho V = 6800 \times 2 \times 10^{-6} = 0.0136 \ \text{kg}

\]

Sensible heat removed:

\[

Q = m c_p \Delta T

\]

\[

Q = 0.0136 \times 400 \times 700 = 3808 \ \text{J}

\]

Specific heat \(c_p = 400 \ \text{J kg}^{-1}\text{K}^{-1}\) is an assumed engineering value and should be verified against measured thermal data for the specific alloy; it is labeled needs primary source validation.

Heat transfer rate:

\[

\dot{Q} = h A \Delta T_{\text{avg}}

\]

\[

\dot{Q} = 500 \times 0.01 \times 500 = 2500 \ \text{W}

\]

Here \(h = 500 \ \text{W m}^{-2}\text{K}^{-1}\) is an assumed metal-mold interface value, not a measured BMG-specific value.

Cooling-only time:

\[

t_{\text{thermal}} = \frac{Q}{\dot{Q}} = \frac{3808}{2500} \approx 1.5 \ \text{s}

\]

Sensitivity:

• If \(h = 200 \ \text{W m}^{-2}\text{K}^{-1}\), \(\dot{Q} = 1000 \ \text{W}\), so \(t_{\text{thermal}} \approx 3.8 \ \text{s}\).
• If part volume doubles to 4 cm³, \(t_{\text{thermal}} \approx 3.0 \ \text{s}\) at \(h = 500\).
• Real cycle time is dominated by mold motion, part handling, and tool life. If overhead is 15 s and cooling is 1.5 s, effective cycle is ~16.5 s, or about 218 parts/hour per cavity.

Interpretation: Thermal removal is not the economic bottleneck if BMG can be injected into a chilled mold. The bottleneck is tooling, part complexity, alloy cost, and the thickness/crystallization constraint from Calculation 1.

Parameter Table

Unit Economics / Business Model

10.1 Who monetizes

The observed monetization models in the public record are:

1. Licensing/IP monetization. Liquidmetal historically entered a license arrangement with Apple for the consumer-electronics field. The event is disclosed in Liquidmetal’s 8-K history, available via SEC EDGAR Tier 2, SEC EDGAR LQMT 8-K archive. A license can produce lump-sum or royalty revenue without the licensor scaling manufacturing.
2. Prototype and development services. Liquidmetal’s public disclosures describe building samples/prototypes for customers, which is a service-and-NRE model rather than high-volume component sales Tier 2, SEC EDGAR LQMT 10-K archive.
3. Component sales. The technology promises net-shape parts that could reduce machining and assembly, but the SEC record shows this has not yet matured into a large, stable revenue stream.

10.2 Margins

No audited segment-level gross-margin number is cited in this note. Any specific margin estimate would be fabricated unless pulled directly from a filing; therefore margin figures are marked needs primary source validation. Structurally, a specialty process with high tooling cost, expensive alloy feedstock, and limited competition can support high technical margins, but low volume can destroy economic margins through overhead absorption.

10.3 Who captures value

If BMG parts scale in consumer electronics, the likely value capture is:

• OEM/integrator, such as Apple, captures the product-price premium from thinner/stronger components and fewer assembly steps.
• Licensor/process developer, such as Liquidmetal, captures fees or royalties, but only if the IP and know-how survive scrutiny.
• Feedstock supplier, such as a beryllium or high-purity alloy producer, captures input margin indirectly.
• Contract manufacturer may capture fabrication margin if the process can be qualified at a foundry.

The pure-play component maker does not necessarily capture the majority of the value; the OEM often does.

Disconfirming Evidence

The bulk-metallic-glass component thesis would be weakened or falsified by any of the following:

1. No audit-grade revenue growth. If Liquidmetal’s 10-K filings continue to show minimal product revenue and customer concentration, the commercial-moat claim fails even if the material works technically SEC EDGAR LQMT archive.
2. Embrittlement confirmed in independent tests. If peer-reviewed testing from an institution unaffiliated with the patent holder shows that monolithic BMG components consistently fail in tension via shear bands, then product applications requiring toughness become unviable Schuh et al. 2007, DOI 10.1016/j.actamat.2007.01.052.
3. Thickness limit incompatible with components. If the critical cooling rate of commercially viable alloys restricts section thickness to a few millimeters or less, then the target structural and frame applications are not feasible under the calculation above.
4. High cycle time or low yield. If molding trials show that tool life, fill defects, or handling overhead produce no cost advantage over CNC-machined aluminum or injection-molded magnesium, then unit economics fail.
5. Absence of current IP protection. If no current granted patent with meaningful claims is identified beyond the expired US 5,288,344 A, then there is no legal moat, only know-how and trade secrets US 5,288,344 A.
6. OEM abandonment. If the Apple relationship produces no product-scale purchasing volume over time, then the largest observable demand pull disappears SEC EDGAR LQMT 8-K archive.