Researchers have demonstrated that spin waves oscillating at terahertz frequencies inside antiferromagnetic crystals can generate electrical voltage signals in adjacent metal layers, a result that bridges ultrafast magnetic phenomena with conventional electronics. The work, built on experiments using chromium oxide and heavy-metal detectors, offers a practical route to reading out data encoded in spin dynamics at speeds far beyond what current semiconductor technology allows. A separate German-Japanese collaboration has since extended the principle by converting terahertz spin waves into electrical signals through an intermediate light step, confirming the approach works across different material platforms.
How Chromium Oxide Produces Electrical Signals From Spin Waves
The core experiment that established this capability used 0.240 THz radiation to drive antiferromagnetic resonance in chromium oxide (Cr2O3), an insulating material whose atomic magnetic moments point in alternating directions. When the radiation frequency matched the natural precession rate of those moments, the crystal’s spin system absorbed energy and began oscillating coherently. That collective oscillation, a spin wave, carried angular momentum to the crystal’s surface and injected a pure spin current into thin films of platinum (Pt) and beta-tantalum deposited on top. The resonance occurred at applied magnetic fields of approximately 2.7 T, a value consistent with known properties of Cr2O3’s antiferromagnetic order.
Inside the heavy-metal layers, a quantum-mechanical process called the inverse spin Hall effect deflected spin-polarized electrons sideways, producing a measurable open-circuit voltage. Platinum and beta-tantalum have opposite spin Hall angles, so the voltage peaks appeared with reversed polarity in the two metals, serving as a built-in consistency check. The experiment was conducted at UC Santa Barbara’s Institute for Terahertz Science and Technology, where the spin current produced in Cr2O3 was converted into a DC voltage readable by standard electronics.
Why Electrical Detection Matters at Terahertz Frequencies
Terahertz radiation sits between microwave and infrared on the electromagnetic spectrum, and measuring it has long been difficult. Traditional detection relies on bulky optical setups or cryogenic bolometers that are impractical for integration into chips. By converting the magnetic excitation directly into voltage, the chromium oxide experiments sidestep those constraints entirely. A team of scientists reported an electrical detection method for terahertz electromagnetic waves that could, in principle, be implemented on microchips and enhance sensitivity to weak signals.
The practical payoff is speed. Antiferromagnets resonate at frequencies hundreds to thousands of times higher than the ferromagnets used in today’s hard drives and magnetic memory. If those resonances can be written and read electrically, data processing could in principle operate at terahertz clock rates rather than the gigahertz ceiling that silicon transistors face. That prospect has drawn attention from researchers working on spintronic memory and logic devices, where information is stored in magnetic orientation rather than electric charge.
Corroborating Evidence From Independent Groups
Several independent experiments reinforce the central finding. A preprint study demonstrated that sub-terahertz spin pumping from an insulating antiferromagnet into a Pt layer produced an inverse spin Hall voltage, and that circularly polarized sub-THz irradiation could select which chirality of spin-wave mode was excited. That chirality control is significant because it means researchers can address individual magnetic modes selectively, a prerequisite for encoding binary information.
Separately, work on magnetic heterostructures showed that ultrafast spin currents can be generated and detected via inverse spin Hall conversion into terahertz transients, confirming that spin-to-charge conversion can be read out at THz timescales. And research on the ultrafast spin Seebeck effect demonstrated that spin currents driven by THz fields at approximately 0.5 THz could be converted into electrical currents in Pt and Ta layers through the same inverse spin Hall mechanism. Together, these studies indicate that the Cr2O3 results are not an isolated curiosity but part of a broader pattern linking antiferromagnetic dynamics to electrical signals.
Separating Spin Signals From Thermal Noise
One persistent challenge in this field is distinguishing genuine spin-current signals from thermoelectric artifacts. Intense radiation can heat the sample unevenly, and temperature gradients across a metal film produce voltages that mimic spin Hall signals. Research on magnetic-insulator and heavy-metal bilayers showed that competing thermoelectric contributions can be separated from spin Seebeck signals using terahertz emission and detection methods. That separation is essential for trusting that the voltages measured in the Cr2O3 experiments genuinely reflect spin transport rather than simple heating.
The Cr2O3 work addressed this concern by comparing signals in metals with opposite spin Hall angles, monitoring the dependence on magnetic field direction, and checking that the voltage vanished when the THz frequency was tuned away from resonance. These cross-checks reduce the likelihood that the observed voltages are dominated by trivial thermal gradients. Even so, the community continues to refine experimental geometries and analysis methods to suppress or subtract residual thermoelectric backgrounds, particularly as researchers push toward smaller devices where heat management is more difficult.
Most coverage of these results has treated the voltage readout as a clean, solved problem. That framing glosses over the fact that no group has yet published conversion efficiency metrics (meaning the ratio of electrical output power to THz input power) for the Cr2O3 system in a peer-reviewed setting. Without those numbers, it is difficult to judge whether the technique can scale from a laboratory demonstration to a practical device. The signal levels reported so far are small enough that noise rejection and amplification remain open engineering problems.
Expanding the Material Palette
The approach is not limited to chromium oxide. A recent peer-reviewed study demonstrated terahertz spin-to-charge conversion in a different antiferromagnetic insulator, using a similar geometry of an insulating magnetic layer capped with a heavy metal. By exciting antiferromagnetic resonance with ultrafast electromagnetic pulses and measuring the resulting voltage across the metal film, the researchers confirmed that terahertz spin pumping and inverse spin Hall detection are broadly applicable across multiple compounds.
Earlier work on multiferroic oxides such as bismuth ferrite established that antiferromagnetic spin dynamics can couple strongly to electric polarization, enabling optical and electrical manipulation of spin waves at gigahertz and terahertz frequencies. Those studies did not directly measure inverse spin Hall voltages, but they mapped out how crystal symmetry, magnetic anisotropy and electric fields shape the resonance spectrum. That knowledge now guides the search for antiferromagnets with resonances that are both electrically accessible and compatible with thin-film growth on semiconductor substrates.
Expanding the material palette matters for several reasons. Different antiferromagnets offer a wide range of resonance frequencies, damping rates and thermal stabilities. Some exhibit room-temperature order and robust signals; others can be tuned with modest electric fields or strain. Heavy metals beyond Pt and Ta, including alloys and topological materials, may provide larger spin Hall angles and therefore stronger voltages for a given spin current. Systematically exploring these combinations will determine whether terahertz spintronic detectors can be engineered to meet practical requirements on sensitivity, bandwidth and power consumption.
From Laboratory Demonstration to Device Concept
Translating these findings into technology will require progress on several fronts. On the materials side, researchers need thin antiferromagnetic films with low damping and well-controlled interfaces to heavy metals. Interface roughness, intermixing or unwanted magnetic phases can all reduce the efficiency of spin pumping and spin-to-charge conversion. Epitaxial growth, interface engineering and careful characterization will be essential to optimize performance.
On the device side, engineers must design structures that couple efficiently to incoming terahertz radiation while remaining compatible with planar fabrication. The Cr2O3 experiments used free-space beams focused onto millimeter-scale crystals, but practical detectors would likely rely on on-chip antennas or waveguides to deliver THz fields into micron-scale active regions. Integrating such antennas with antiferromagnetic/heavy-metal bilayers, while minimizing parasitic capacitances and resistances, is a nontrivial design challenge.
Signal readout also demands attention. Because the inverse spin Hall voltages are typically in the microvolt to millivolt range, low-noise amplification and filtering will be required, especially if detectors are to operate at room temperature and in electrically noisy environments. Circuit-level solutions could include differential measurement schemes that exploit the opposite spin Hall angles of two metals, lock-in detection synchronized to pulsed THz sources, or on-chip amplification stages positioned close to the spintronic element.
Despite these hurdles, the conceptual advantages remain compelling. Antiferromagnetic spintronic detectors promise ultrafast response, inherent immunity to stray magnetic fields and potential compatibility with existing semiconductor processing. The Cr2O3 experiments, together with corroborating studies in other antiferromagnets and heterostructures, show that terahertz spin waves can indeed be converted into electrical signals that conventional electronics can read. The next steps will determine whether that principle can be engineered into robust components for sensing, communication or information processing at frequencies where traditional electronics struggle to keep up.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
