🔬 mesoscale physics

Second Harmonic Generation Spectroscopy of the Surface Charge Density Wave in the Weyl Semimetal CoSi

Using temperature-dependent rotational anisotropy second harmonic generation, researchers identified a purely surface-driven charge density wave transition on CoSi (001) at 90 K that exhibits a critical exponent consistent with the 3D XY universality class, a result attributed to the coupling between surface Fermi-arc states and bulk topology.

Original authors: Awadhesh K. Das, Wesley E. Deeg, Sujan Subedi, Chandra Shekhar, Claudia Felser, Darius H. Torchinsky

Published 2026-08-19

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Awadhesh K. Das, Wesley E. Deeg, Sujan Subedi, Chandra Shekhar, Claudia Felser, Darius H. Torchinsky

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 surface of a solid is rarely just a boundary; it is often a place where the rules of the material change. In a special class of crystals known as Weyl semimetals, the connection between the inside and the outside is particularly profound. Deep within the crystal, electrons move in a way that creates a unique topological structure, a kind of mathematical knot that cannot be untied. Because this knot cannot simply end at the surface, the electrons are forced to form open, two-dimensional paths that run along the skin of the material. These paths are called Fermi arcs. While the bulk of the crystal remains stable, these surface paths are delicate and can be easily disturbed by the interactions between electrons. When electrons on these paths begin to organize themselves into a repeating pattern, a phenomenon known as a charge density wave, it creates a new state of matter. Understanding how this happens on the surface, and whether the surface acts as an isolated two-dimensional layer or remains tied to the three-dimensional bulk, is a key question in modern physics.

A team of researchers at Temple University and the Max Planck Institute for Chemical Physics of Solids has now provided a clear answer for a specific crystal called CoSi. By using a technique that involves bouncing light off the crystal's surface, they discovered that the electrons on the (001) face of CoSi organize themselves into a charge density wave at a temperature of 90.0 ± 0.8 Kelvin. This transition, where the electrons shift from a chaotic state to an ordered one, was detected by measuring how the crystal reflects light of a specific color. When the researchers shone a beam of infrared light onto the crystal, the surface emitted light at exactly half the wavelength, a process called second harmonic generation. This effect is extremely sensitive to the symmetry of the surface and the arrangement of electrons, acting as a precise gauge for the order of the material.

The researchers rotated the polarization of the incoming light and measured the intensity of the reflected signal as they cooled the crystal from 115 Kelvin down to 50 Kelvin. They found that as the temperature dropped below 90 Kelvin, the signal grew significantly stronger, indicating that the electrons had settled into a new, ordered pattern. By analyzing how this signal grew with temperature, they calculated a critical exponent, a number that describes the nature of the transition. They found this number to be 0.30 ± 0.03. In the language of physics, this specific value points to a three-dimensional system, known as the 3D XY universality class. This result was surprising because the charge density wave exists on the surface, which is geometrically a two-dimensional layer. In a strictly two-dimensional world, such a system would be expected to behave differently, often failing to maintain long-range order due to thermal fluctuations, or undergoing a different kind of transition entirely.

To ensure that this ordering was truly a surface phenomenon and not a bulk effect, the team performed additional measurements. They angled the light beam to probe the interior of the crystal and also tested a different face of the crystal, the (111) surface, which is sensitive to the bulk properties. In both cases, they found no sign of the transition. The bulk of the crystal remained unchanged, confirming that the ordering was confined entirely to the (001) surface. This ruled out the possibility that the entire crystal was undergoing a phase change, isolating the event to the very top atomic layers.

The fact that a surface layer behaves like a three-dimensional system requires an explanation. The researchers propose that the answer lies in how the surface electrons are connected to the bulk. Previous studies using a scanning tunneling microscope had shown that the charge density wave on the surface does not sit flat on a single layer of atoms. Instead, the pattern shifts by a half-step between adjacent sub-layers within the crystal's unit cell. This creates a strong link between the surface and the layers immediately beneath it. Because the surface electrons are tied to the bulk's topological structure, they cannot be treated as an isolated two-dimensional sheet. The coupling between the surface Fermi arcs and the bulk topology provides enough stability, or "phase stiffness," to allow the order to extend coherently in three dimensions. This connection allows the surface to bypass the limitations that usually prevent two-dimensional systems from sustaining long-range order.

The study confirms that the instability driving this ordering is electronic and originates from the surface Fermi arcs, rather than a structural change in the crystal lattice. The transition is continuous, meaning the order grows smoothly as the temperature drops, rather than appearing suddenly. The researchers attribute the specific three-dimensional behavior to the intrinsic link between the surface states and the bulk topology. While the exact microscopic details of how this coupling supplies the necessary stability remain an open question, the evidence points to a system where the surface is not an independent entity but a manifestation of the bulk's deeper structure. This work suggests that similar instabilities could exist in other Weyl semimetals, offering a new platform to study how topology and electron interactions combine to create emergent phases of matter.

Problem and Context
The interaction between strong electronic correlations and topological band structures remains an open frontier in condensed matter physics. While bulk Weyl semimetals host protected surface states known as Fermi arcs, the consequences of symmetry-breaking order parameters acting specifically on these boundary arcs are not well understood. Recent scanning tunneling microscopy (STM) measurements identified an incommensurate, unidirectional charge density wave (CDW) on the (001) surface of the Weyl semimetal CoSi, locked to the crystal's structural chirality. However, the fundamental nature of this surface transition—specifically its universality class, effective dimensionality, and microscopic coupling—remained undetermined. A key theoretical question is whether a strictly two-dimensional surface layer with a continuous order parameter would undergo a Berezinskii–Kosterlitz–Thouless (BKT) transition, as predicted by the Mermin-Wagner theorem, or if coupling to the bulk topology could alter this behavior.

Methodology
The authors employed temperature-dependent rotational anisotropy second harmonic generation (RA-SHG) spectroscopy to probe the CoSi (001) surface. SHG is a sensitive, non-contact probe of symmetry-breaking order parameters, particularly effective for surface-confined effects due to electric-dipole selection rules.

Experimental Setup: Experiments were conducted on polished CoSi (001) single crystals in a closed-cycle cryostat (50–115 K). A 1500 nm fundamental beam was used, with polarization controlled via a rotating linear polarizer and quarter-wave plate.
Geometry and Selection Rules: Measurements were performed at near-normal incidence to isolate the surface response. The (001) surface belongs to the $C_1$ point group, allowing all six independent surface tensor elements of the second-order nonlinear susceptibility ( $\chi^{(2)}$ ) to contribute. Crucially, the bulk contribution (allowed only for $\chi^{(2)}_{xyz}$ in this space group) is suppressed at normal incidence because the incident field lacks a $z$ -component and the resulting bulk polarization does not radiate along the propagation axis.
Data Analysis: The SHG intensity was measured across four polarization geometries ( $I_H, I_V, I_\parallel, I_\perp$ ). Global fits were used to extract the six surface tensor elements. To isolate the CDW contribution, a high-temperature background was subtracted at the tensor-element level before reconstructing the intensity. This approach preserves the linear coupling between the susceptibility and the order parameter, avoiding the distortion caused by subtracting intensities directly.
Bulk Verification: To confirm the surface-specific nature of the transition, the authors performed RA-SHG on the (111) face (which probes the bulk-allowed $\chi^{(2)}_{xyz}$ ) and on the (001) face at an off-normal angle (~30°).

Key Results

Surface Phase Transition: The RA-SHG data on the (001) surface reveals a clear second-order phase transition with an onset temperature of $T_{CDW} = 90.0 \pm 0.8$ K. Below this temperature, the SHG intensity increases by nearly a factor of two, while the angular anisotropy (shape of the polar plots) remains unchanged, consistent with an incommensurate CDW that does not alter the $C_1$ symmetry.
Critical Exponent and Universality Class: The background-subtracted SHG intensity, which scales as the square of the order parameter ( $\Delta^2$ ), follows a power law $I \propto (1 - T/T_C)^{2\beta}$ . The extracted critical exponent is $\beta = 0.30 \pm 0.03$ . This value is consistent with the 3D XY universality class ( $\beta_{3DXY} \approx 0.3485$ ) and inconsistent with the mean-field limit ( $\beta=0.5$ ) or the infinite-order BKT transition expected for a strictly 2D system with a continuous order parameter.
Surface Confinement: Measurements on the (111) face and off-normal incidence on the (001) face show no comparable anomaly in the bulk response. The absence of a transition signal in the bulk-allowed tensor elements confirms that the CDW is confined to the (001) surface.
Tensor Element Behavior: The temperature dependence is dominated by two tensor elements, $\chi^{(2)}_{xxy}$ and $\chi^{(2)}_{yxx}$ . The ratio of these two dominant elements remains constant across the transition, indicating they track a common underlying factor. In contrast, the ratios of the weaker tensor elements fluctuate significantly above $T_{CDW}$ , signaling a loss of long-range phase coherence.

Significance and Claims
The paper establishes that the surface CDW in CoSi represents a purely surface-driven phase transition that defies the standard expectation for a two-dimensional system.

Resolution of the Dimensionality Paradox: Although the CDW is spatially confined to the surface layer, the observed 3D XY critical exponent contradicts the Mermin-Wagner theorem's prediction of a BKT transition for a strictly 2D system. The authors attribute this to the intrinsic three-dimensional structure of the order parameter.
Mechanism: The authors propose that the coupling between the surface Fermi-arc states and the bulk topology they are tied to provides the necessary phase stiffness. This is grounded in previous STM observations of a nearly $\pi$ phase shift between adjacent sublayers within the unit cell across step edges. This intra-unit-cell phase relationship links inequivalent surface terminations, effectively extending the order parameter coherently along the [001] direction and preventing the fluctuation-driven suppression typical of 2D systems.
Topological Correlations: The work demonstrates that topological surface states can host correlated electronic order as an intrinsic, surface-confined phenomenon. The specific scaling behavior suggests that the "two-dimensionality" of topological arcs is distinct from that of isolated 2D electron gases, as the arcs lack independent existence apart from the 3D bulk topology.

The study provides the first determination of the universality class for a surface CDW in a Weyl semimetal, highlighting a novel interplay where bulk topology dictates the critical behavior of a surface instability.

More like this