🔬 materials science

Influence of Charge Density Waves on the Hall coefficient in NiTi

This paper presents a mean-field charge density wave theory for NiTi, demonstrating that biaxial charge density waves dominated by Ni d-orbital "hot spots" are required to accurately reproduce the experimental Hall coefficient and other transport properties, whereas uniaxial waves and standard Boltzmann transport theory fail to do so.

Original authors: Adrian Braun, Henrik Dick, Timon Sieweke, Alexander Kunzmann, Klara Lünser, Gabi Schierning, Thomas Dahm

Published 2026-01-22

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Adrian Braun, Henrik Dick, Timon Sieweke, Alexander Kunzmann, Klara Lünser, Gabi Schierning, Thomas Dahm

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

Imagine a piece of metal called NiTi (Nickel-Titanium), famous for its "shape memory." If you bend it, it snaps back to its original shape when heated. This happens because the atoms inside rearrange themselves, like a crowd of people suddenly shifting from standing in a perfect square grid to a slanted, diamond-like formation.

Scientists have long known how the atoms move, but they've been puzzled by why the electrons inside behave the way they do during this shift. Specifically, they were trying to understand a measurement called the Hall coefficient, which acts like a "traffic report" for electrons moving through the metal. It tells us how many electrons are moving and how easily they flow.

Here is the story of what the researchers found, explained simply:

1. The Mystery of the "Traffic Report"

The researchers took a sample of NiTi and measured how its electrons behaved as they cooled it down.

The Expectation: They used a standard computer model (like a GPS for electrons) to predict what the "traffic report" should look like.
The Reality: The computer model failed miserably. It predicted the electrons would flow one way, but the experiment showed they were flowing the opposite way. It was like a GPS telling you to turn left when you were actually driving right.

2. The "Hot Spots" on the Map

To figure out why the GPS was wrong, the team looked closer at the "map" of the electrons (called the Fermi surface). They found that the electrons weren't all behaving the same way.

Most of the electrons were just cruising along, doing nothing special.
However, there were a few specific "hot spots" on the map where the electrons were very active.
The Key Discovery: These hot spots were mostly made of Nickel atoms, not Titanium. The behavior of these specific Nickel electrons was the main reason the "traffic report" looked so strange.

3. The Missing Ingredient: The "Charge Density Wave"

The standard computer model assumed the electrons were just a smooth, calm sea. But the researchers suspected the electrons were actually forming a pattern, like ripples on a pond. In physics, this is called a Charge Density Wave (CDW).

They tested three different types of these "ripples":

Type A (Uniaxial): Ripples going in one direction (like stripes on a zebra).
Type C (Uniaxial): Another stripe pattern.
Type B (Biaxial): Ripples going in two directions at once, creating a checkerboard pattern.

The Result:

The "stripe" patterns (Types A and C) made the traffic report even worse. They couldn't explain the data at all.
The "checkerboard" pattern (Type B) was the magic key. When the researchers added this specific ripple pattern to their model, the "traffic report" suddenly matched the real-world experiment perfectly!

4. The Temperature Twist

The researchers also looked at how this changed with temperature:

In the Hot Phase (Austenite): The metal is in its square-grid shape. The researchers found that a tiny, weak version of the "checkerboard ripple" might already be starting to form here, like a faint echo before a loud sound.
In the Cold Phase (Martensite): As the metal cools and the atoms shift to the slanted shape, this "checkerboard ripple" gets much stronger and louder.

5. The "Heat" Connection

Finally, they checked if this ripple theory made sense with how much heat the metal holds (specific heat).

Usually, when electrons form a pattern (like a CDW), you might expect them to have less energy available to hold heat.
Surprisingly, their model showed that the "checkerboard ripple" actually increased the amount of energy the electrons could hold.
When they compared this to real heat measurements, the numbers matched perfectly. This confirmed that the "checkerboard ripple" theory was likely correct.

The Bottom Line

The paper concludes that the strange behavior of electrons in NiTi isn't just random noise. It's caused by a specific, invisible pattern of electrons (a biaxial Charge Density Wave) that acts like a checkerboard. This pattern is weak when the metal is hot but gets strong when it gets cold, and it is the reason the "traffic report" (Hall coefficient) looks the way it does. Without accounting for this pattern, standard physics models simply cannot explain what is happening inside this shape-memory metal.

Technical Summary: Influence of Charge Density Waves on the Hall Coefficient in NiTi

Problem Statement
The electronic and structural properties of Nickel-Titanium (NiTi), particularly regarding its reversible martensitic phase transition between the high-temperature cubic austenite (B2) and low-temperature monoclinic martensite (B19') phases, remain incompletely understood. While molecular dynamics simulations have addressed structural changes, they rely on the Born-Oppenheimer approximation, neglecting non-adiabatic effects such as retarded electron-phonon interactions crucial for instabilities like Charge Density Waves (CDWs). Although experimental evidence (superlattice reflections, optical conductivity gaps, Fermi surface nesting) has long suggested CDW formation in NiTi, a quantitative theoretical study linking CDW formation to electronic transport properties has been lacking. Recent experiments revealed a strong decrease in charge carrier density and a significant increase in mobility in the martensite phase, alongside a large electronic entropy contribution, hinting at a strong electronic subsystem involvement that standard band theory has failed to reproduce accurately.

Methodology
The authors employ a multi-stage theoretical approach grounded in Density Functional Theory (DFT) and mean-field CDW theory:

Band Structure Calculation: Using QuantumEspresso with the PBEsol functional, the authors calculated the ground state energy and band structures for the B2 and B19' phases. They utilized a machine-learning-inspired tight-binding method to generate accurate models from DFT data, avoiding the oscillatory artifacts common in Fourier interpolation near band crossings.
Boltzmann Transport Analysis: The Hall coefficient ( $R_H$ ) and resistivity were calculated using Boltzmann transport theory within the constant relaxation time approximation. The authors analyzed momentum-resolved contributions to $R_H$ and projected Fermi surface states onto atomic orbitals.
Mean-Field CDW Theory: To address discrepancies between standard theory and experiment, the authors introduced a mean-field CDW Hamiltonian. They modeled three specific commensurate CDW scenarios based on observed phonon softening and nesting vectors:
- Model A: Uniaxial CDW with wave vector $Q_A = \frac{2}{3}(\pi, \pi, 0)$ .
- Model B: Biaxial CDW with wave vectors $Q_{B1} = \frac{2}{3}(\pi, \pi, 0)$ and $Q_{B2} = \frac{2}{3}(\pi, -\pi, 0)$ .
- Model C: Uniaxial CDW with wave vector $Q_C = (\pi, \pi, 0)$ .
  The order parameter $W$ was determined self-consistently, and the resulting subbands were used to recalculate transport properties.
Experimental Comparison: Theoretical results were compared against experimental data from a Ni $_{50.6}$ Ti $_{49.4}$ sample grown via micro-pulling-down, measuring resistivity, Hall coefficient (via classical and van der Pauw methods), and specific heat.

Key Results

Failure of Standard Boltzmann Theory: Without CDWs, the constant relaxation time approximation fails to reproduce experimental data. In the B2 phase, the theory predicts a negative Hall coefficient, whereas experiments show a positive one. In the B19' phase, while the magnitude is roughly correct, the temperature dependence is weak, failing to capture the experimentally observed strong increase in $R_H$ as temperature decreases.
Dominance of "Hot Spots": Analysis of the Fermi surface reveals that the Hall coefficient is dominated by specific "hot spots" rather than the entire surface. These regions are primarily composed of Ni d-orbitals.
Success of Biaxial CDW (Model B): The introduction of a biaxial CDW (Model B) successfully reproduces the experimental Hall coefficient.
- In the B19' phase, a biaxial CDW with a negative order parameter ( $W < 0$ ) yields a Hall coefficient that increases with decreasing temperature, matching the experimental trend and magnitude ( $\sim 3 \times 10^{-10}$ m $^3$ /C).
- In the B2 phase, a biaxial CDW with a small amplitude ( $W \approx -0.04$ eV) corrects the sign of the Hall coefficient from negative to positive, aligning with experimental observations.
Uniaxial Models: Both uniaxial models (A and C) fail to reproduce the experimental data, as they result in a decreasing Hall coefficient with increasing CDW order parameter magnitude and cannot achieve the necessary positive values in the B2 phase.
Density of States and Specific Heat: Contrary to naive expectations where CDW formation opens a gap and reduces the density of states (DOS), the biaxial CDW in the B19' phase leads to an increase in the DOS at the Fermi level due to the splitting of bands. This increased DOS ( $D_{CDW} \approx 1.006$ states/eV/atom vs. $D_0 \approx 0.618$ states/eV/atom) quantitatively matches the experimental electronic specific heat coefficient ( $\gamma$ ), whereas the standard DFT result without CDW underestimates it.
Resistivity: While the resistivity in the martensite phase can be fitted using standard electron-phonon scattering models, the austenite phase requires either an unrealistically high elastic scattering rate or a significantly enhanced electron-phonon coupling strength to match data, suggesting complex scattering mechanisms not fully captured by the current model.

Significance and Claims
The paper claims that the anomalous electronic transport properties of NiTi, specifically the sign and temperature dependence of the Hall coefficient, cannot be explained by standard DFT-based Boltzmann transport theory. Instead, the authors demonstrate that a biaxial Charge Density Wave provides a consistent theoretical framework.

The study suggests a physical picture where a small-amplitude, precursor biaxial CDW may already exist in the austenite (B2) phase. Upon transitioning to the martensite (B19') phase, this CDW is "boosted" to a larger amplitude. This mechanism not only corrects the Hall coefficient sign and temperature dependence but also explains the enhanced electronic specific heat observed experimentally. The work provides the first quantitative theoretical study linking CDW formation in NiTi to macroscopic transport and thermodynamic properties, resolving long-standing discrepancies between theory and experiment regarding the electronic subsystem's role in the martensitic transition.