🔬 mesoscale physics

Observation of magnetic quantum phase crossovers in a semiconductor spin ladder

This paper demonstrates the realization of a programmable Heisenberg spin ladder in a germanium quantum dot array, where tunable exchange interactions and Hamiltonian-learning protocols enable the quantitative mapping of magnetic quantum phase crossovers and the detection of higher-order spin correlations to establish the platform as a controllable system for studying quantum magnetism and unconventional superconductivity.

Original authors: Elizaveta Morozova, Xin Zhang, Utso Bhattacharya, Pablo Cova Fariña, Daniel Jirovec, Alexander Nico-Katz, Stefan D. Oosterhout, Sougato Bose, Giordano Scappucci, Menno Veldhorst, Eugene Demler, Lieven

Published 2026-08-19

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Elizaveta Morozova, Xin Zhang, Utso Bhattacharya, Pablo Cova Fariña, Daniel Jirovec, Alexander Nico-Katz, Stefan D. Oosterhout, Sougato Bose, Giordano Scappucci, Menno Veldhorst, Eugene Demler, Lieven M. K. Vandersypen

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 world built not from atoms, but from tiny islands of electricity, where the rules of everyday physics give way to the strange, interconnected logic of the quantum realm. In this microscopic landscape, scientists study how particles that carry a magnetic property called spin interact with one another. When these spins are arranged in specific patterns, they can form collective states that behave like a single, unified system, much like a crowd of people moving in perfect unison. One of the most fascinating arrangements is the spin ladder, a structure resembling a ladder where the rungs and sides are made of these magnetic particles. Theorists have long predicted that by adjusting the strength of the connections between these particles and applying a magnetic field, the entire system should snap between different fundamental states: one where the particles pair up and cancel each other out, another where they align in a tilted, cooperative order, and a third where they all point in the same direction. Understanding these transitions is crucial because they are believed to be the secret behind how certain materials become superconductors, conducting electricity with zero resistance, a property that could revolutionize energy transmission and computing.

For decades, these ideas remained largely theoretical, tested only in bulky, solid materials where scientists could not easily tweak the individual connections between particles or see what was happening at the level of a single pair. The disorder inherent in real-world materials often obscured the clean physics researchers hoped to observe. Now, a team of researchers has brought this abstract concept to life in a controlled laboratory setting, using a chip made of germanium to create a miniature, programmable spin ladder. Instead of relying on the rigid structure of a crystal, they engineered a chain of eight tiny quantum dots—nanoscale traps for electrons—arranged in two parallel rows. By applying precise voltages to metal gates above the chip, they could tune the strength of the magnetic interaction between neighboring dots, effectively turning the "rungs" and "legs" of their ladder on and off or making them stronger and weaker at will. This level of control allowed them to map out the behavior of the system as they swept through different magnetic conditions, watching the quantum state evolve in real time.

The researchers set out to see if they could observe the predicted transitions between the three distinct phases. They began by holding the magnetic field steady and gradually changing the strength of the connections across the rungs of the ladder. As they did this, they measured how many of the particle pairs were in a "triplet" state, a configuration where the spins are aligned, versus a "singlet" state, where they are paired and opposed. They found that the system did not jump abruptly from one state to another, as it would in an infinitely large material, but rather flowed smoothly through a crossover region. In the first phase, the particles formed tight pairs across the rungs, canceling out their magnetic effects. As the connections were weakened, the system entered a middle ground where the particles began to interact more freely, forming a state where they were partially aligned but still fluctuating. Finally, as the magnetic influence took over, all the spins aligned in the same direction. This smooth transition, rather than a sharp break, is exactly what one expects when working with a small, finite system, and the team's measurements matched the theoretical predictions for this behavior with remarkable precision.

To confirm that they were truly seeing the underlying physics and not just random noise, the team looked deeper than simple measurements of alignment. They calculated how the spins on different parts of the ladder influenced one another, specifically looking at complex patterns involving four particles at once. These higher-order correlations are incredibly difficult to detect in traditional experiments, which usually only see how two particles relate. By measuring these intricate relationships, the researchers could distinguish between the different phases with high confidence. They observed that in the middle phase, the particles exhibited a specific type of anti-alignment, where neighbors tended to point in opposite directions, a signature of the predicted "canted antiferromagnetic" state. This state is unique because it combines a general magnetic tilt with a hidden, long-range order that is invisible to simpler probes. The data showed that as they strengthened the connections along the sides of the ladder, this middle phase grew wider, exactly as the theory suggested it should.

The study also revealed how the material's internal structure, specifically a property called spin-orbit interaction, affects these quantum states. In germanium, the motion of the particles is tightly linked to their spin, which can cause the magnetic axis to wobble slightly. The team found that this effect smoothed out the sharp boundaries between phases, turning what would be a sudden jump in a perfect system into a gradual crossover. By using advanced computational techniques to learn the exact parameters of their system from the data, they were able to reproduce the observed behavior in simulations, confirming that their model captured the essential physics. This work demonstrates that semiconductor chips can serve as powerful simulators for complex quantum magnetism, offering a clean, tunable platform to explore phenomena that are otherwise hidden in messy, real-world materials. It opens the door to future experiments where researchers can add or remove particles from these ladders to study how doping affects magnetic order, potentially shedding light on the mechanisms behind high-temperature superconductivity and other exotic states of matter.

Technical Summary: Observation of Magnetic Quantum Phase Crossovers in a Semiconductor Spin Ladder

Problem Statement
Understanding collective phases of strongly correlated quantum magnets typically relies on theoretical models that are difficult to realize experimentally with precise microscopic control. Antiferromagnetic Heisenberg spin ladders serve as a paradigmatic theoretical setting for studying quantum phase transitions (QPT) between gapped and gapless phases, as well as unconventional pairing tendencies upon doping. However, experimental studies in bulk materials (e.g., via neutron scattering or Raman spectroscopy) face significant limitations: they lack continuous tunability of exchange interactions, suffer from unresolved disorder effects, and are restricted to measuring two-point correlation functions. Furthermore, sharp phase transitions in finite, symmetry-breaking systems are replaced by crossovers, requiring detailed analysis of equilibrium properties and excitation spectra to understand fundamental many-body behavior. There is a need for a programmable platform that allows site-resolved control of spin interactions and access to higher-order correlators to probe the underlying phase structure.

Methodology
The authors realized a programmable Heisenberg spin ladder using a $2 \times 4$ array of germanium (Ge) quantum dots embedded in a Ge/SiGe heterostructure. The system operates in the half-filled regime, where each site hosts a single hole spin.

Device Control: Electrostatic plunger gates define the quantum dots, while barrier gates tune the exchange couplings between neighboring sites. The rung coupling ( $J_\perp$ ) and leg coupling ( $J_\parallel$ ) are independently controlled.
Hamiltonian Learning: To quantitatively characterize the system, the authors employed Hamiltonian-learning protocols. By combining equilibrium measurements (ground state properties) and dynamical measurements, they extracted microscopic parameters, including on-site interaction energy, hopping terms, local g-factors, and spin-orbit interaction (SOI) terms.
State Preparation and Readout: The ground state was prepared via adiabatic evolution from a deep charge configuration $(0,2)/(2,0)$ to the operation point in the $(1,1)$ charge configuration. Readout was performed using Pauli spin blockade (PSB) and charge sensing. To access all four rungs in a single shot, a SWAP operation was utilized to transfer the state of the third rung to the second sensor, enabling the extraction of four bits of information per experimental shot.
Data Processing: Readout errors were mitigated using a State Preparation and Measurement (SPAM) matrix and Maximum Likelihood Estimation (MLE) to reconstruct probability distributions of triplet occupations and higher-order correlators.
Theoretical Framework: The experimental data was compared against an effective one-dimensional XXZ Heisenberg model derived from a spin-orbit coupled Fermi-Hubbard model. This model incorporates longitudinal and transverse effective magnetic fields and Dzyaloshinskii-Moriya (DM) interactions arising from SOI.

Key Contributions and Results

Realization of Tunable Spin Ladder: The team demonstrated a semiconductor platform capable of continuously tuning the ratio of rung to leg exchange interactions ( $J_\perp/J_\parallel$ ) and the Zeeman energy ( $h_z$ ), mapping the system across three distinct magnetic regimes.
Observation of Phase Crossovers: Under a fixed magnetic field, the authors varied the couplings to observe the evolution from a rung-singlet (RS) phase to a canted antiferromagnetic (CAFM) phase, and finally to a fully polarized (FP) phase. Due to the finite system size and SOI, sharp quantum phase transitions were observed as smooth crossovers.
- The RS phase (low $h_z - J_\perp$ ) is characterized by a low triplet density.
- The FP phase (high $h_z - J_\perp$ ) exhibits a high triplet density.
- The intermediate CAFM phase is a gapless regime where the triplet number variance peaks, indicating enhanced fluctuations and the presence of mobile triplons.
Quantitative Agreement with Theory: The experimental phase boundaries, determined by the slopes of the transition lines in the $(h_z - J_\perp)$ vs. $J_\parallel$ plane, agreed with theoretical predictions. Specifically, the slope of the CAFM-to-FP boundary was found to be approximately twice that of the RS-to-CAFM boundary (ratio $\approx 2$ ), consistent with mean-field theory for the critical points.
Higher-Order Correlators: The study successfully measured four-point spin correlators (specifically connected longitudinal correlators and joint probability distributions of rung configurations). These measurements, inaccessible to conventional bulk probes, revealed signatures of the phase structure:
- The RS and FP phases showed minimal inter-site correlations.
- The CAFM phase exhibited anti-correlations between neighboring rungs, which strengthened with increasing $J_\parallel$ , consistent with the theoretical prediction of quasi-long-range antiferromagnetic order.
Role of Spin-Orbit Interaction: The inclusion of SOI in the theoretical model was essential to reproduce the observed crossover behavior and the smearing of finite-size plateaus. The SOI pins the ordering direction, resulting in a weakly gapped pseudo-Goldstone mode in the CAFM phase.

Significance and Claims
The paper establishes germanium quantum dot arrays as a controllable platform for investigating quantum magnetism. The authors claim that their results provide a direct experimental realization of the phase diagram of a Heisenberg spin ladder, validating theoretical models of quantum criticality in finite systems. By demonstrating the ability to measure higher-order spin correlators and reconstruct the ground state properties with high fidelity, the work opens routes to investigate more complex phenomena, such as:

Unconventional Superconductivity: The platform is positioned to study doped ladders, where pairing tendencies reminiscent of $d$ -wave pairing in two-dimensional systems are expected.
Bose-Einstein Condensation (BEC): The phase crossovers can be mapped to the BEC of hard-core bosons (triplons), offering a solid-state realization of this theoretical proposal.
Hamiltonian Learning: The demonstrated approach to accurately determine Hamiltonian parameters, including SOI, is highlighted as crucial for developing algorithms to solve practical problems, such as simulating nuclear spin dynamics for NMR inference.

The authors maintain a modest tone, noting that while the system size is small ( $L=4$ ), the results show good agreement with thermodynamic limit predictions when finite-size effects and SOI are accounted for. They emphasize that the platform's ability to access higher-order correlations and its tunability make it a compelling candidate for future studies of frustrated magnetism and exotic phases of matter.

More like this