Possibility of ferro-octupolar order in Ba$_2$CaOsO$_6$ assessed by X-ray magnetic dichroism measurements

This study utilizes X-ray absorption and magnetic circular dichroism spectroscopy on Ba$_2$CaOsO$_6$ to provide experimental evidence supporting the existence of ferro-octupolar order in its hidden-ordered state, revealing a strong nearest-neighbor octupole exchange interaction of approximately 1.5 meV and a residual cubic crystal field splitting of about 18 meV.

The Gist

Imagine you have a tiny, invisible dance floor inside a crystal, and on this floor, electrons are dancing. Usually, we think of electrons as tiny spinning tops. But in certain special materials, like the one studied in this paper (Ba₂CaOsO₆), these electrons are so heavy and move so fast that their "spin" gets tangled up with their "orbit" (how they move around the atom). This creates a weird, complex dance that scientists call strong spin-orbit coupling.

Here is the story of what the scientists found, explained without the heavy jargon.

1. The Mystery of the "Hidden Order"

The scientists were studying a crystal containing Osmium (Os) ions. These ions have two electrons dancing in a specific way (called a $5d^2$ state).

When they cooled the crystal down to about -223°C (50 Kelvin), something strange happened. The material changed its state, like water turning to ice. But here's the catch: nothing looked different.

• The crystal structure didn't change shape (no "Jahn-Teller distortion").
• Neutron beams (which usually see magnetic patterns) didn't see any new magnetic lines.

It was a "Hidden Order." The electrons had organized themselves, but they were hiding their secret from normal eyes. The question was: What kind of secret dance were they doing?

2. The "Octupole" vs. The "Quadrupole"

To understand the secret, imagine the electrons aren't just spinning tops, but they have shapes.

• Electric Quadrupole: Imagine a dumbbell shape. If the electrons arrange themselves like dumbbells pointing in different directions, that's a quadrupole order. This usually makes the crystal squish or stretch (change shape). But our crystal didn't change shape, so this wasn't it.
• Magnetic Octupole: Now, imagine a much more complex, 3D shape that looks like a twisted knot or a specific pattern of magnetic fields. This is an octupole. It's like a "magnetic flower" that doesn't change the shape of the crystal but creates a very specific, hidden magnetic pattern.

The scientists suspected the electrons had formed a Ferro-Octupolar Order. This means all the "magnetic flowers" on the dance floor were pointing in the same direction, creating a unified, hidden state.

3. The Detective Work: X-ray Flashlights

How do you see something invisible? The scientists used X-rays as super-powered flashlights.

• They shot X-rays at the crystal and measured how much light was absorbed (XAS).
• Then, they used a special trick: they spun the X-rays like a corkscrew (circular polarization) and applied a strong magnetic field. This created a "magnetic echo" called XMCD.

Think of it like this: If you shout into a cave, the echo tells you about the shape of the cave. If you shout with a specific spin (left-handed vs. right-handed), the echo changes depending on what's inside. The way the X-rays echoed back told the scientists exactly how the electrons were arranged.

4. The "Ghost" Magnetic Field

When they analyzed the echoes, they found something surprising. The electrons were reacting to a magnetic field much stronger than the one the scientists applied in the lab.

• The scientists applied a field of 7 Tesla (very strong, like a giant MRI machine).
• But the electrons acted as if they were in a field of 19 Tesla (7 + 12 "ghost" Tesla).

This "ghost field" was actually the collective whisper of all the neighboring electrons talking to each other. It proved that the electrons were strongly interacting and organizing themselves.

5. The Final Verdict: A Strong Bond

By using a computer model (like a video game simulation of the atoms), the scientists tried to figure out how strong the "hand-holding" (exchange interaction) between the electrons needed to be to create this hidden order.

They calculated that the electrons were holding hands with a force of about 1.5 meV.

• To put that in perspective: This is a very specific, strong bond that is just right to keep the "magnetic flowers" (octupoles) aligned without breaking the crystal's shape.
• This number matched perfectly with what other scientists had predicted using different math methods.

The Big Picture Analogy

Imagine a crowd of people in a stadium.

• Normal Magnetism: Everyone stands up and waves their arms in the same direction. You can see the movement easily.
• The Hidden Order (Octupole): Everyone stays seated, but they all tilt their heads in a very specific, complex pattern (like a wave of "tilts"). From a distance, the crowd looks still (no shape change), and if you just look at the crowd, you don't see the pattern. But if you use a special camera (the X-ray), you can see that everyone is perfectly synchronized in this weird, hidden tilt.

Why Does This Matter?

This paper confirms that magnetic octupoles are real and can organize themselves in crystals. This is exciting because:

1. New Physics: It proves that nature has more ways to organize matter than we thought (beyond just simple magnets).
2. Future Tech: These "hidden" states might be useful for making super-fast, super-efficient computers or new types of sensors that don't rely on standard magnets.

In short, the scientists used X-ray flashlights to catch a crystal "hiding" a complex magnetic dance, proving that the electrons were performing a synchronized "octupole" routine that keeps the crystal's shape perfectly intact.

Technical Summary

1. Problem Statement

The study focuses on the nature of the "hidden order" phase transition observed in the cubic double perovskite Ba$_2$CaOsO$_6$ at a critical temperature $T^* \approx 50$ K.

• The Mystery: Below $T^*$, the material exhibits a cusp-like anomaly in magnetic susceptibility, yet neutron diffraction reveals no magnetic Bragg peaks (ruling out conventional dipolar magnetic order). Muon-spin rotation ($\mu$-SR) detects small local and staggered magnetic moments, while X-ray diffraction confirms the crystal structure remains cubic down to low temperatures (ruling out electric quadrupolar order which would induce Jahn-Teller distortion).
• The Hypothesis: Theoretical predictions suggest that localized $5d^2$ electrons (specifically Os$^{6+}$ ions) in a cubic crystal field can form magnetic octupolar orders. The authors aim to experimentally verify if the hidden order in Ba$_2$CaOsO$_6$ is indeed a ferro-octupolar state and to quantify the exchange interactions driving this order.

2. Methodology

The researchers employed a combination of advanced spectroscopic techniques and theoretical modeling:

• Experimental Techniques:
  • X-ray Absorption Spectroscopy (XAS): Measured at the Os $L_{2,3}$ edges to probe the electronic structure and ligand-field splitting.
  • X-ray Magnetic Circular Dichroism (XMCD): Measured under magnetic fields ($B = \pm 7$ T) at temperatures of 10 K and 60 K. This technique is sensitive to orbital and spin magnetic moments.
  • Sample: High-quality polycrystalline Ba$_2$CaOsO$_6$ pellets.
• Theoretical Analysis:
  • Ligand-Field (LF) Multiplet Calculations: Performed using the XTLS 8.5 package to simulate the XAS and XMCD spectra.
  • Parameters: The model included the cubic ligand-field splitting ($\Delta_{LF}$), spin-orbit coupling (SOC), Hund's coupling, and Slater integrals (reduced to account for hybridization).
  • Molecular Field Approximation: To reproduce the experimental XMCD intensity, the authors introduced an effective "molecular field" in the calculation, simulating the internal exchange field generated by neighboring Os ions.

3. Key Results

A. Electronic Structure and Ligand Field

• Ligand-Field Splitting: XAS spectra revealed a double-peak structure at the $L_3$ and $L_2$ edges, corresponding to the transition from Os $2p$ to $5d$ states. The analysis deduced a $t_{2g}$-$e_g$ ligand-field splitting of $\Delta_{LF} \approx 4$ eV.
• Ground State: The $5d^2$ configuration under strong SOC forms a $J_{eff}=2$ multiplet. The residual cubic crystal field splits this into a non-Kramers $E_g$ doublet ground state and a $T_{2g}$ triplet excited state.
• Energy Gap: The temperature dependence of the XMCD spectra indicated a residual cubic splitting ($\Delta_c$) of $\sim 18$ meV between the $E_g$ ground state and the $T_{2g}$ excited state. This value is consistent with previous neutron scattering data.

B. Magnetic Moments and Anisotropy

• XMCD Sum Rules: Analysis yielded an orbital magnetic moment ($M_{orb}$) of $-0.016 \pm 0.002 \mu_B$/Os and an effective spin moment ($M_{spin}^{eff}$) of $0.058 \pm 0.007 \mu_B$/Os at 10 K.
• Orbital Contribution: The relatively large unquenched orbital moment confirms the dominant role of strong spin-orbit coupling in the electronic structure.
• Magnetic Dipole Term: The analysis separated the effective spin moment into a pure spin component and a magnetic dipole term ($T$). The large $T$ value reflects significant anisotropic spin distribution induced by strong SOC.

C. Evidence for Ferro-Octupolar Order

• Multiplet Calculation under Magnetic Fields: Theoretical calculations showed that a magnetic field along the $\langle 111 \rangle$ direction splits the non-Kramers $E_g$ doublet into purely magnetic octupolar eigenstates ($|\psi_{g,\pm}\rangle$).
• Exchange Interaction Estimation:
  • The experimental XMCD data required an effective internal molecular field of $B_{mol} \approx 12 \pm 2$ T (in addition to the external 7 T) to be reproduced.
  • Extrapolating this to the transition temperature ($T^* \approx 50$ K), the authors estimated that the internal exchange field required to split the octupolar doublet by $\sim 4$ meV (where $k_B T^* \approx 4$ meV) would be approximately 300 T.
  • Assuming a face-centered cubic (fcc) lattice with 12 nearest neighbors, this field corresponds to an exchange interaction strength of $J \approx 1.5$ meV.

4. Key Contributions

1. Experimental Validation: Provided the first direct spectroscopic evidence (via XMCD temperature dependence) supporting the existence of a non-Kramers $E_g$ ground state separated from the $T_{2g}$ state by $\sim 18$ meV in Ba$_2$CaOsO$_6$.
2. Quantification of Exchange Interaction: Derived a quantitative estimate for the nearest-neighbor exchange interaction ($J \sim 1.5$ meV) required to stabilize the ferro-octupolar order. This value aligns well with previous theoretical predictions (DFT and model calculations) of $\sim 1$ meV.
3. Mechanism Elucidation: Demonstrated that the hidden order is consistent with a ferro-octupolar state rather than an electric quadrupolar state, as the latter would require a structural distortion (Jahn-Teller) which is absent, and the octupolar scenario fits the magnetic field splitting behavior observed in the multiplet calculations.

5. Significance

• Fundamental Physics: This work confirms that strong spin-orbit coupling in $5d^2$ systems can stabilize exotic multipolar orders (octupoles) that are invisible to conventional magnetic probes (neutron diffraction).
• Material Design: It establishes Ba$_2$CaOsO$_6$ as a prototype material for studying multipolar physics, providing a benchmark for theoretical models of correlated electron systems.
• Future Directions: The authors suggest that while XMCD supports the octupolar hypothesis, direct proof could be achieved via impurity doping (to break symmetry and induce detectable dipoles) or large-angle neutron scattering. The findings also highlight the potential of non-linear Hall effects as a probe for such hidden orders.

In conclusion, the study successfully links the macroscopic "hidden order" transition in Ba$_2$CaOsO$_6$ to a microscopic ferro-octupolar ordering of Os$^{6+}$ ions, driven by strong spin-orbit coupling and a specific exchange interaction strength of $\sim 1.5$ meV.