A laboratory study reports a measurable sideways heat-transport signal in diamond under a magnetic field, according to an arXiv preprint dated 26 August 2026. The reported transverse thermal conductivity, κxy—the measure used for sideways heat flow—reached 340 W m−1 K−1 at 10 T, a value the authors describe as the largest ever observed.
The experiment drove heat along x, measured the transverse temperature gradient along y and applied the magnetic field along z. That three-direction arrangement is the basic geometry behind the thermal Hall measurement studied here.
A demanding measurement
Diamond’s ordinary, lengthwise heat flow was itself unusually large: in the 30 µm sample, longitudinal thermal conductivity, κxx, peaked at 9200 W m−1 K−1 at 115 K. The supplied analysis says this high conductivity suppressed the temperature signals and left relatively large error bars in the transverse measurement.
The work used two plate-like single-crystal diamonds. Both were 3.5 mm long and 2 mm wide; one was 30 µm thick and the other 100 µm thick. Longitudinal and transverse measurements were performed in two steps, using a Cernox thermometer and field-calibrated type E differential thermocouples.
To isolate the field-related part of the reading, the analysis examined the field-odd component of the transverse temperature difference—the part that changed sign when the field was reversed. That component dominated a nearly linear signal.
The response changed with sample thickness
In both crystals, κxy/B—the transverse conductivity divided by magnetic field—had a pronounced maximum. The peak was 7.5 W m−1 K−1 T−1 in the 30 µm sample and 34 W m−1 K−1 T−1 in the 100 µm sample.
A separate field-normalized comparison of transverse and longitudinal heat transport gave peak ratios of 0.8 ×10−3 T−1 in the thinner crystal and 2.7 ×10−3 T−1 in the thicker one.
The authors also report nearly coincident maxima for κxy and κxx. At their common maximum, the field-normalized ratio was approximately 10−3 T−1.
Transverse thermal resistivity, W⊥, a measure of resistance to sideways heat flow, decreased as temperature increased in diamond. Its reported W⊥/B magnitude was in the range 10−7–10−8 K m W−1 T−1.
Across the compared insulators, the maximum field-normalized Hall angle was not proportional to maximum κxx, although the values showed a visible positive correlation. That comparison suggests a pattern, but it does not by itself establish a common cause.
The proposed explanation
The authors interpret the diamond signal as an intrinsic thermal Hall source that does not require exotic carriers or chiral phonons. They propose that phonon–phonon interactions—interactions among the lattice vibrations in the heat-transport picture—are crucial.
To compare the observed scales with that interpretation, they use order-of-magnitude estimates rather than a rigorous microscopic calculation. The estimates put the Hall-angle scale at 5.4 ×10−4 T−1 and the transverse-resistivity scale at 10−7 K m W−1 T−1, using ρU near 100 GPa; both were considered comparable to the observations.
Promising signal, limited reach
The evidence remains narrow. Only two crystals were measured, one at each thickness, so the study leaves open whether the response is reproducible across a broader set of diamond samples. The two-sample design also does not identify a unique microscopic cause.
The reported peaks come without formal uncertainty intervals. The measurements also carried relatively large transverse error bars because diamond’s very high longitudinal conductivity suppressed the temperature signals.
The cross-insulator comparison is descriptive rather than controlled and may reflect differences in samples, methods or literature sources. More independent measurements on additional diamond crystals and other nonmagnetic insulators, together with a rigorous microscopic theory, would be needed to test whether the response is generic and to pin down its cause.
For now, the preprint supports a finite transverse response in these diamond crystals; it does not establish that all crystalline insulators share the effect.
Paper data and sources
Original title: Observation of thermal Hall effect in diamond
Authors: Marcin Matusiak, Andrzej Ptok, Maria Szlawska, Kamran Behnia
Journal/Repository: arXiv
Status: Preprint, not yet peer-reviewed
First online: 2026-08-26
DOI: Not available
Original paper · Full text