🔬 materials science

Questioning van der Waals Epitaxy of Non-Layered Materials on Mica: The Case of ScN

This study demonstrates that ScN films grown on mica follow conventional epitaxy with strain buildup and dislocation generation, thereby refuting prior claims of van der Waals epitaxy for non-layered materials on mica and urging that conventional epitaxy be the default assumption unless specific conditions are proven.

Original authors: Susmita Chowdhury, Faezeh Alijan Farzad Lahiji, Mikael Ottoson, Olivier Donzel-Gargand, Robert J. W. Frost, Martin Magnuson, Ganpati Ramanath, Arnaud le Febvrier, Per Eklund

Published 2026-02-19

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Susmita Chowdhury, Faezeh Alijan Farzad Lahiji, Mikael Ottoson, Olivier Donzel-Gargand, Robert J. W. Frost, Martin Magnuson, Ganpati Ramanath, Arnaud le Febvrier, Per Eklund

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Question: Is it "Glue" or "Velcro"?

Imagine you are trying to build a house of cards (the film) on top of a smooth, slippery table (the substrate).

For years, scientists have been fascinated by a special way of building these houses called Van der Waals Epitaxy (vdWE). Think of this as building with Velcro.

How it works: The cards stick to the table just enough to stay put, but they don't glue themselves down.
The Magic: Because they aren't glued, the house of cards can be incredibly thick without cracking. If you pull the table away, the whole house lifts off cleanly, ready to be stuck onto something else. This is the "holy grail" for making flexible electronics (like bendable screens).

However, there's a catch. This "Velcro" method usually only works if the table itself is made of layers (like a stack of paper). But recently, some scientists claimed they could use this "Velcro" method to grow Scandium Nitride (ScN)—a hard, 3D material—on top of Mica (a layered mineral). They said the ScN was growing via vdWE, meaning it was loosely attached and stress-free.

This paper says: "Wait a minute. That's not Velcro. That's Super Glue."

The Investigation: Finding the Stress

The authors of this paper decided to test this claim by growing ScN films of different thicknesses on Mica and looking very closely at what was happening.

1. The "Rubber Band" Test (Strain)
Imagine stretching a rubber band. If you stretch it, it gets tighter.

The Theory (vdWE): If the film is attached by "Velcro," it shouldn't feel any tension. No matter how thick the film gets, the atoms inside should stay relaxed, like a loose shirt.
The Reality (Conventional Epitaxy): If the film is attached by "Super Glue" (strong chemical bonds), the atoms get squeezed as the film gets thicker. They build up stress, like a rubber band being pulled tighter and tighter.

What they found:
As the ScN film got thicker, the scientists measured the distance between the atoms. They saw that the atoms were getting squeezed closer together. This proved that stress was building up. A "Velcro" connection cannot hold enough stress to squeeze atoms; only a strong "Glue" connection can.

2. The "Crack" Test (Dislocations)
When you stretch a rubber band too far, it eventually snaps or develops a kink to relieve the tension.

The Theory (vdWE): Because there is no stress, the film grows perfectly smooth and thick without any defects.
The Reality (Conventional Epitaxy): Because the "Glue" is holding the film tight, the stress gets so high that the film has to "break" internally to let go. The scientists used a super-powerful microscope (STEM) to look at the edge where the film meets the mica. They saw dislocations—tiny defects where the atomic layers were twisted and broken.

The Analogy:
Think of the film as a heavy blanket being laid over a mattress.

vdWE (Velcro): The blanket floats slightly above the mattress. You can pull the mattress out, and the blanket comes with it, perfectly smooth.
Conventional Epitaxy (Glue): The blanket is glued to the mattress. As you try to make the blanket thicker (adding more layers), the glue pulls it tight. Eventually, the fabric rips or bunches up (dislocations) because it can't handle the tension. That is exactly what the scientists saw.

The Verdict

The paper concludes that the ScN film is not growing via Van der Waals Epitaxy. Instead, it is growing via Conventional Epitaxy.

The Bond: It is a strong chemical bond (like Super Glue), not a weak physical bond (like Velcro).
The Result: The film is under stress, and that stress causes defects as it gets thicker.
The Lesson: Just because a material grows on a layered substrate (like Mica), it doesn't automatically mean it's "Van der Waals." Unless you can prove the film is stress-free and independent of thickness, you should assume it's the "glued" version.

Why Does This Matter?

If scientists think they have a "Velcro" film when they actually have a "Glued" one, they might try to peel it off to make a flexible device, only to find it shatters or breaks because it was never actually loose.

This paper is a "reality check." It tells the scientific community: Don't assume it's magic (vdWE) just because it looks like it. Check the stress. If there's stress, it's not vdWE.

Summary in One Sentence

The researchers proved that Scandium Nitride films on Mica are tightly "glued" and under stress, debunking the idea that they are loosely "Velcroed" together, which means we need to be much more careful before claiming we have achieved this special type of growth for non-layered materials.

1. Problem Statement

Van der Waals epitaxy (vdWE) is a growth technique where films are deposited on layered substrates (like mica) via weak intermolecular forces rather than strong chemical bonds. This method is highly desirable for creating stress-free, flexible, and transferable optoelectronic devices. However, a significant controversy exists regarding non-layered (3D) materials (such as transition metal nitrides) grown on mica.

Previous studies have claimed vdWE for non-layered materials like ScN, GaN, and MoO₂ on mica, suggesting these films grow without strain buildup and can be easily delaminated. The authors argue that these claims lack rigorous experimental evidence. Specifically, they question whether the weak bonding required for true vdWE can exist for materials with strong covalent/ionic bonding (like ScN), or if the observed growth is actually conventional epitaxy characterized by strong interface bonding, strain accumulation, and dislocation generation.

2. Methodology

The authors investigated the growth of rock-salt Scandium Nitride (ScN) films on fluorophlogopite mica (f-mica) substrates using a comprehensive multi-technique approach:

Film Growth: ScN films with varying thicknesses (8 nm to 138 nm) were grown via direct-current magnetron sputter deposition at ~800°C on freshly exfoliated f-mica (001) surfaces.
Structural Characterization:
- X-ray Diffraction (XRD): Used symmetric $\omega-2\theta$ scans, high-resolution scans, and $\omega$ -rocking curves to analyze interplanar spacings ( $d$ -spacing), crystallinity, and mosaicity.
- Pole Figures & $\phi$ -scans: Performed to determine the epitaxial relationship and crystal orientation between the film and substrate.
- Scanning Transmission Electron Microscopy (STEM): High-resolution bright-field imaging of cross-sections to visualize the interface, lattice distortion, and defect structures at the atomic scale.
Compositional Analysis:
- ToF-ERDA (Time-of-Flight Elastic Recoil Detection Analysis): Used to determine elemental stoichiometry and depth profiles, confirming the N/Sc ratio and identifying adventitious impurities (O, C, H).
- 4D-STEM & EDX: Used to map elemental distribution and extract diffraction patterns from specific regions (film, interface, substrate).

3. Key Results

The study provides compelling evidence that ScN growth on f-mica follows conventional epitaxy, not vdWE:

Epitaxial Relationship: The films exhibit a single-domain epitaxial relationship defined as 1̅01ScN || 010mica. The lattice mismatch is calculated to be approximately -3.95% and -3.78%, driving the formation of defects.
Strain Evolution (Thickness Dependence):
- A critical finding is the monotonic decrease in the (111) and (222) interplanar spacings ( $d_{111}$ and $d_{222}$ ) as film thickness increases.
- Thinner films (8 nm) show larger $d$ -spacings (indicating tensile strain or less relaxation), while thicker films (up to 138 nm) show $d$ -spacings approaching the bulk value but still exhibiting significant compressive strain relative to the substrate.
- Significance: In true vdWE, strain should be independent of thickness due to weak bonding. The observed thickness-dependent strain relaxation is a hallmark of conventional epitaxy.
Defect Generation:
- STEM Imaging: Revealed a sharp interface with atomic registry but also showed edge dislocations (indicated by extra half-planes) and lattice distortion extending a few atomic planes into the mica substrate.
- Rocking Curves: The Full-Width-at-Half-Maximum (FWHM) of the ScN peaks decreased with increasing thickness (from 4.70° to 3.39°), indicating domain coalescence and defect annihilation, but the values remain high, suggesting a high defect density consistent with strain relaxation mechanisms.
Bonding Nature: The presence of strain buildup and the formation of dislocations to relieve this strain refute the existence of weak vdW bonding. The authors argue that the resilience of these films against delamination (observed in prior literature) is due to strong covalent/ionic bonds (1–10 eV/atom) at the interface, not weak vdW forces (~1–100 meV/atom).

4. Key Contributions

Refutation of vdWE Claims: The paper definitively demonstrates that ScN, a non-layered material, grows via conventional epitaxy on mica, contradicting previous claims of vdWE for ScN and similar nitrides.
Establishment of a Diagnostic Criterion: The authors propose that monitoring the thickness-dependence of strain evolution is a necessary condition to validate vdWE claims. If interplanar spacing changes with thickness, the growth is conventional, not vdW.
Atomic Registry Visualization: The study provides the first clear atomic-scale visualization of the ScN/mica interface, showing the specific lattice mismatch and the resulting dislocation network.
Correction of Assumptions: It challenges the community to stop assuming vdWE for non-layered materials on mica unless explicit evidence of strain independence and weak bonding is provided.

5. Significance

This work has profound implications for the field of flexible electronics and 2D heterostructures:

Scientific Rigor: It prevents the misinterpretation of growth mechanisms, ensuring that future device designs are based on accurate physical models of film-substrate interactions.
Device Engineering: Understanding that ScN grows via conventional epitaxy means that these films will likely contain strain and defects that affect their electronic and optical properties, unlike the "stress-free" ideal of vdWE. This impacts the design of flexible plasmonic and thermoelectric devices.
General Applicability: The findings suggest that other non-layered transition metal nitrides and oxides (e.g., TiN, NiO) claimed to grow via vdWE on mica should be re-evaluated using the same strain-thickness diagnostic criteria.

In conclusion, the authors assert that conventional epitaxy should be the default assumption for non-layered materials on mica until compelling evidence of strain independence and weak interface bonding is explicitly demonstrated.