Vapor-liquid-solid growth of unconventional nanowires

arXiv:2604.09959 · cond-mat.mtrl-sci · Submitted 2026-04-10 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Condensed Matter. The day in condensed matter: materials, phases and the measurements under them.

Kai: We're looking at the paper: "Vapor-liquid-solid growth of unconventional nanowires".

Mira: Here's the gist of the paper. Vapor-liquid-solid (VLS) growth has achieved a level of synthetic control in group IV and III-V semiconductor nanowires that surpasses any other class of one-dimensional materials.

Kai: However, for oxides, carbides, and chalcogenides—collectively termed non-conventional nanowires—comparable deterministic control has yet to be realized despite their predicted functional properties and broad application potential.

Mira: This review systematically examines the divergence between these material classes by surveying literature on VLS and VLS-related synthesis across the stages of growth: precursor creation and delivery, seed particle formation and evolution, and nucleation/growth.

Kai: A consistent theme emerges: the three stages of VLS growth are intrinsically coupled.

Paper summary: Kai: So we've covered that this paper is looking at where vapor-liquid-solid growth succeeds for group four and three materials but struggles for oxides, carbides, and chalcogenides. The authors claim to examine the literature on these non-conventional nanowires by following the sequence of growth: precursor creation, seed particle formation, and nucleation and growth itself.

Mira: Exactly. They argue that even though we see one-dimensional structures in these other materials reported a lot, they lack deterministic control over morphology, crystal phase, and composition compared to the conventional systems.

Lev: The thesis is that the divergence between these material classes needs to be understood by looking at which step of the growth process is causing the issue.

Kai: They focus heavily on precursor chemistry and delivery as a major hurdle, noting there's a lack of precursors for oxides, carbides, and chalcogenides that match both VLS dissociation temperatures and high-vacuum compatibility.

Mira: That means we can't easily get the right starting point materials into the reactor for these specific types of nanowires.

Lev: And they also highlight inadequate reactor control, specifically that many horizontal tube furnaces don't allow independent temperature setting for different precursors or between the precursor and the growth substrate.

Kai: Those delivery issues mean that accessible precursor chemistries and reactor temperature profiles define the seed compositions we can form, which severely limits our accessible seed design space.

Mira: The resulting gap in deterministic synthesis persists because experiments designed to close this problem haven't been performed for these material classes yet.

Lev: And they move into the second stage: seed particle formation and evolution. They point out that the choice and evolution of the seed particle is central to this whole problem, encompassing its material choice, chemical composition during growth, and its dynamics.

Kai: This means even if we have a good precursor delivery system, if the resulting seed behaves unpredictably under growth conditions, we're stuck.

Mira: They contrast this with conventional systems where binary phase diagrams helped predict seed selection based on droplet state manipulation; no such equivalent rule exists here for these non-conventional materials.

Conclusion: Kai: So, looking at the full picture of this paper, 'Vapor-liquid-solid growth of unconventional nanowires,' the authors are really emphasizing that we need a new way to think about seeding these materials.

Mira: They're not just saying VLS works for some things and not others; they're systematically examining *why* they diverge by tracing the problem through the entire growth sequence, from precursors to the final structure.

Lev: The implication is that we need a comprehensive understanding of how precursor chemistry, seed behavior, and interfacial kinetics all interact when growing these complex materials.

Kai: For someone who just listens to this, it means that instead of just trying different materials blindly for oxide or carbide nanowires, we need a framework based on the growth process itself.

Mira: It's about moving beyond simple binary rules for seeds and developing a way to predict how alloyed and multicomponent droplets will behave during growth.

Lev: The challenge boils down to understanding "the particle at the tip." It’s figuring out its state, composition, and movement while it grows under those specific conditions.

Kai: So the paper is basically saying that solving this requires resolving open questions about seed particles for materials we already demonstrated, before we can even hope to design better seeds for new quantum applications.

Virginia Tech

cond-mat.mtrl-sci

Submitted: 2026-04-10

Updated: 2026-09-25

Comments: 45 pages, 10 figures, 2 tables

DOI: 10.1021/acs.chemmater.6c01372

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 81/100

The gist: Vapor-liquid-solid (VLS) growth has achieved a level of synthetic control in group IV and III-V semiconductor nanowires that surpasses any other class of one-dimensional materials.

Key concepts

Vapor-liquid-solid (VLS) growth
A synthesis method where vaporized material dissolves into a liquid droplet at the tip of a substrate, which then precipitates to form a solid nanowire. This technique has been highly successful for group IV and III-V semiconductors but is challenging for other materials.
Non-conventional nanowires
This category includes materials like oxides, carbides, and chalcogenides that are not the standard group IV or III-V semiconductors. Despite predicted useful properties, achieving deterministic control over their growth using VLS remains difficult.
Seed particle bottleneck
The seed particle—the initial droplet at the tip where growth begins—is identified as the primary obstacle. Understanding its material choice, composition during and after growth, and how it behaves is crucial because it dictates which final structures can be formed.
Binary eutectics
These are simple phase diagrams that predict seed selection based on mixing two components into a single liquid state. The paper argues that these rules are insufficient for non-conventional materials, suggesting a need for more complex frameworks to guide growth.

Terminology

Summary

Vapor-liquid-solid (VLS) growth has achieved a level of synthetic control in group IV and III-V semiconductor nanowires that surpasses any other class of one-dimensional materials. p. 0

Where the paper says it

1 Vapor-liquid-solid growth of unconventional nanowires Thang Pham* and Arindom Nag Department of Materials Science and Engineering, Virginia Tech, Blacksburg, VA 24060 * Corresponding author: thangpham@vt. edu Abstract Vapor-liquid-solid (VLS) growth has delivered a level of synthetic control in group IV and III-V semiconductor nanowires that remains unmatched in any other class of one-dimensional materials.

However, for oxides, carbides, and chalcogenides—collectively termed non-conventional nanowires—comparable deterministic control has yet to be realized despite their predicted functional properties and broad application potential. p. 0
Where the paper says it

For Si, Ge, and GaAs, decades of mechanistic study have made diameter, growth direction, crystal phase, and axial composition into design parameters. For oxides, carbides, and chalcogenides, including emerging van der Waals materials, which we refer to collectively as non-conventional nanowires, one-dimensional structures have been reported in abundance, but comparable deterministic control has not been achieved despite their predicted functional properties and broad application potential.

This review systematically examines the divergence between these material classes by surveying literature on VLS and VLS-related synthesis across the stages of growth: precursor creation and delivery, seed particle formation and evolution, and nucleation/growth. p. 0
Where the paper says it

This review examines where and why the two groups diverge. We survey and categorize the literature on VLS and VLS-related synthesis of non-conventional nanowires, following the sequence of the growth process itself, namely precursor creation and delivery, seed particle formation and evolution, and nucleation and growth of one-dimensional structures.

A consistent theme emerges: the three stages of VLS growth are intrinsically coupled. p. 2

Where the paper says it

Across these sections, a consistent picture emerges. The three stages of VLS growth are coupled rather than independent.

Precursor chemistry and reactor design dictate reachable seed compositions and temperatures, while the resulting seed particle governs which structures, phases, and compositions can form. p. 2
Where the paper says it

The three stages of VLS growth are coupled rather than independent. Precursor chemistry and reactor design set which seed compositions and growth temperatures can be reached, and the seed particle in turn governs which structures, phases, and compositions can form.

The principal bottleneck restricting deterministic synthesis across non-conventional nanowires is identified as the seed particle itself—encompassing its material choice, chemical composition during and after growth, and its dynamics. p. 2
Where the paper says it

Precursor chemistry and reactor design set which seed compositions and growth temperatures can be reached, and the seed particle in turn governs which structures, phases, and compositions can form. Within this coupling, we identify the seed particle, from the choice of seed material to its chemical composition during and after growth and its dynamics, as the principal bottleneck restricting deterministic synthesis of non-conventional nanowires.

The authors pinpoint two major obstacles in precursor creation and delivery:

  1. Limited Precursor Availability: There is a lack of suitable molecular precursors for oxides, carbides, and chalcogenides that meet both the appropriate dissociation temperature for VLS growth and compatibility with high-vacuum environments. This contrasts sharply with group IV (Si, Ge) and III-V (GaAs, InAs) systems where precursor selection benefits from existing thin-film deposition research. p. 8

    Where the paper says it

    9 (i) an appropriate dissociation temperature for VLS growth, and (ii) compatibility with high-vacuum environments. For nanowires in group IV (Si, Ge) and III-V (GaAs, InAs) systems, molecular precursor selection benefits directly from decades of research in thin-film deposition of the same materials.

  2. Inadequate Reactor Control: Commonly used horizontal tube furnaces lack independent temperature control for different precursors or between precursors and the growth substrate.

These constraints exert their influence downstream: accessible precursor chemistries and reactor temperature profiles define the seed compositions that can be formed and stabilized, thereby limiting the accessible seed design space. p. 9

Where the paper says it

These constraints matter most through their downstream effect. The accessible precursor chemistries and reactor temperature profiles define which seed compositions can be formed and stabilized during growth, and therefore how much of the seed design space can be explored at all.

The gap in deterministic synthesis persists because experiments designed to close it have not yet been performed for these material classes. p. 2
Where the paper says it

Within this coupling, we identify the seed particle, from the choice of seed material to its chemical composition during and after growth and its dynamics, as the principal bottleneck restricting deterministic synthesis of non-conventional nanowires. We support this assessment throughout the review and discuss potential solutions, including new precursor and source chemistries, reactor architectures with independently controlled precursor delivery, and in situ growth experiments adapted to these material classes.

The authors argue that the seed problem is an opportunity rather than a dead end. p. 2

Where the paper says it

The three stages of VLS growth are coupled rather than independent. Precursor chemistry and reactor design set which seed compositions and growth temperatures can be reached, and the seed particle in turn governs which structures, phases, and compositions can form.

For conventional nanowires, binary phase diagrams provided predictive rules for seed selection based on droplet state manipulation; however, no equivalent rule exists for non-conventional materials. p. 1
Where the paper says it

1a): delivery of a vaporphase precursor to a substrate bearing metal particles, formation of a liquid alloy droplet governed by the eutectic composition and temperature, and supersaturation-driven nucleation of the solid at the droplet interface. The binary phase diagram of the seed and nanowire materials rationalizes each step (Fig.

The physical state of the seed during oxide growth remains contested (e. p. 4
Where the paper says it

5 hierarchical structures [108,109]; surfaceguided planar wires [90]; alloyed wires [103] 108, 109 Dichalcogenide monolayer ribbons MoS2, MoSe2, WS2 Salt-generated volatile intermediates (sodium molybdate) [78,82] Molten alkali salt droplets (NaCl, KCl, hydroxides) [78,82,83] Monolayer and few-layer ribbons [82,83]; core-shell architectures [81] 78, 81- 83 Transition metal trichalcogenides NbS3, NbSe3, TiS3 (TaS3, ZrS3 as bulk) Salt-generated volatile intermediates [58]; CVT for bulk crystals [130] Multicomponent Au-alkali halide alloy droplets [58] Wires and ribbons with distinct growth orientations, lateral dimension set by droplet size [58] 58, 130 Emerging quasi-1D vdW and quantum materials, nonseeded to date Ta2Pd3Se8, Ta2Ni3Se8, Nb2Se9, Bi4I4, Sb2S3, GaSeI, InSeI, SnIP, SnTe, Cd3As2, TaAs, WP, BaTiS3, BaZrS3 CVT [140]; liquidprecursor CVD [141]; solid-state growth [136]; uncatalyzed vaporphase crystallization [138,139] None established; in Bi4I4, Au particles act as VS nucleation sites or, at specific compositions, as VLS seeds [138] Wires, ribbons, and quasi-2D sheets selected by growth parameters or seed composition [138,139, 146, 149, 150]; helical and twisted morphologies [137] 136- 142, 146, 149, 150 1. Precursors 1.

g., conflicting VLS/VSS assignments in Au-seeded ZnO), and the compositions enabling trichalcogenide growth correspond to no measured phase diagram, with properties inferred post-growth. p. 9
Where the paper says it

3c) and crystal phase selection (Figs. During the nanowire growth, the seed particle can exist in either a solid or liquid state, corresponding to the vapor-solid-solid (VSS) or vaporliquid-solid (VLS) growth mechanisms, respectively.

The principal bottleneck restricting deterministic synthesis is thus identified as the absence of a predictive and mechanistic understanding of seed particles beyond binary eutectics. p. 0

Where the paper says it

From this comparison we identify seed particles, spanning the choice of seed material, the chemical composition of the seed during and after growth, and its dynamics, as the principal bottleneck restricting deterministic synthesis, with precursor creation and delivery acting as an upstream constraint that narrows the accessible seed design space. We discuss potential solutions, including new precursor chemistries, improved reactor designs, and in situ characterization of growth, that would deepen mechanistic understanding of nanowires already demonstrated by VLS and extend seeded growth to emerging quantum and functional materials, ultimately paving the way for deterministic synthesis and integration of complex one-dimensional nanomaterials.

Progress requires resolving open questions regarding the state, composition, and dynamics of the seed during growth for already demonstrated materials. p. 0
Where the paper says it

From this comparison we identify seed particles, spanning the choice of seed material, the chemical composition of the seed during and after growth, and its dynamics, as the principal bottleneck restricting deterministic synthesis, with precursor creation and delivery acting as an upstream constraint that narrows the accessible seed design space. We discuss potential solutions, including new precursor chemistries, improved reactor designs, and in situ characterization of growth, that would deepen mechanistic understanding of nanowires already demonstrated by VLS and extend seeded growth to emerging quantum and functional materials, ultimately paving the way for deterministic synthesis and integration of complex one-dimensional nanomaterials.

For emerging quantum materials, what is missing is precisely what a seed provides: defined growth location, dimensional control via droplet footprint, and the ability to switch composition mid-growth for heterostructures. p. 2
Where the paper says it

The second is to extend seeded growth to emerging quantum and functional materials, where the controls a seed particle provides, over growth location, dimension, and composition, would add capability to materials so far grown only by other routes. Material classes of non-conventional nanowires discussed in this review, with the precursor routes, seed materials, and one-dimensional morphologies demonstrated for each.

The enabling advance across all material classes is framed as developing a seed selection framework that extends beyond binary eutectics to alloyed and multicomponent droplets. p. 4
Where the paper says it

5 hierarchical structures [108,109]; surfaceguided planar wires [90]; alloyed wires [103] 108, 109 Dichalcogenide monolayer ribbons MoS2, MoSe2, WS2 Salt-generated volatile intermediates (sodium molybdate) [78,82] Molten alkali salt droplets (NaCl, KCl, hydroxides) [78,82,83] Monolayer and few-layer ribbons [82,83]; core-shell architectures [81] 78, 81- 83 Transition metal trichalcogenides NbS3, NbSe3, TiS3 (TaS3, ZrS3 as bulk) Salt-generated volatile intermediates [58]; CVT for bulk crystals [130] Multicomponent Au-alkali halide alloy droplets [58] Wires and ribbons with distinct growth orientations, lateral dimension set by droplet size [58] 58, 130 Emerging quasi-1D vdW and quantum materials, nonseeded to date Ta2Pd3Se8, Ta2Ni3Se8, Nb2Se9, Bi4I4, Sb2S3, GaSeI, InSeI, SnIP, SnTe, Cd3As2, TaAs, WP, BaTiS3, BaZrS3 CVT [140]; liquidprecursor CVD [141]; solid-state growth [136]; uncatalyzed vaporphase crystallization [138,139] None established; in Bi4I4, Au particles act as VS nucleation sites or, at specific compositions, as VLS seeds [138] Wires, ribbons, and quasi-2D sheets selected by growth parameters or seed composition [138,139, 146, 149, 150]; helical and twisted morphologies [137] 136- 142, 146, 149, 150 1. Precursors 1.

Ultimately, the challenge lies in understanding the particle at the tip.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improve predictive modeling for seed selection by developing a framework that extends beyond binary eutectics to include multicomponent droplets, as suggested by The compositions of the multicomponent droplets that enable trichalcogenide growth correspond to no measured phase diagram.

  2. Enhance in situ characterization capabilities by developing platforms capable of delivering controlled fluxes of species hostile to conventional instruments, such as semi-automated, hybrid chemical vapor deposition reactors that integrate inorganic and metal-organic precursor delivery, independent temperature control, and in situ precursor switching.

  3. Develop data-driven models for growth optimization by creating a searchable materials database that integrates synthesis parameters with structural and chemical characterization to allow for quantitative identification of the key parameters governing growth outcomes.

  4. Improve understanding of seed dynamics by developing methods to resolve the physical state of the seed during growth, addressing the gap where the physical state of the seed during oxide growth remains contested in even the most studied systems.

  5. Expand heterostructure synthesis capabilities by focusing research on mechanisms that control axial heterostructures with abrupt interfaces, specifically investigating how to overcome limitations in precursor delivery and seed behavior for non-conventional materials.

Abstract

Vapor-liquid-solid (VLS) growth has delivered a level of synthetic control in group IV and III-V semiconductor nanowires that remains unmatched in any other class of one-dimensional materials. For Si, Ge, and GaAs, decades of mechanistic study have made diameter, growth direction, crystal phase, and axial composition into design parameters. For oxides, carbides, and chalcogenides, including emerging van der Waals materials, which we refer to collectively as non-conventional nanowires, one-dimensional structures have been reported in abundance, but comparable deterministic control has not been achieved despite their predicted functional properties and broad application potential. This review examines where and why the two groups diverge. We survey and categorize the literature on VLS and VLS-related synthesis of non-conventional nanowires, following the sequence of the growth process itself, namely precursor creation and delivery, seed particle formation and evolution, and nucleation and growth of one-dimensional structures. Each step is compared against the group IV and III-V baseline to locate shared behavior and genuine differences. From this comparison we identify seed particles, spanning the choice of seed material, the chemical composition of the seed during and after growth, and its dynamics, as the principal bottleneck restricting deterministic synthesis, with precursor creation and delivery acting as an upstream constraint that narrows the accessible seed design space. We discuss potential solutions, including new precursor chemistries, improved reactor designs, and in situ characterization of growth, that would deepen mechanistic understanding of nanowires already demonstrated by VLS and extend seeded growth to emerging quantum and functional materials, ultimately paving the way for deterministic synthesis and integration of complex one-dimensional nanomaterials.

Related papers