Skip to content

Diagnosing high-pressure superconductors using diamond quantum sensors - a review

FieldValue
Journalnpj Advanced Manufacturing
Published2026-05-25
AuthorsKin On Ho, Sen Yang
DOI10.1038/s44334-026-00088-7

Diamond NV magnetometry detected about 50% local magnetic-field suppression in CeH9 at 118 GPa, while Hao et al. kept NV functionality to 241 GPa. The review shows that implanted NV centers can map Meissner screening, trapped flux, stress, and sample inhomogeneity inside diamond anvil cells. These measurements can test whether a resistive transition matches a superconducting response in the sampled region.

Superhydrides have produced conflicting claims because high pressure restricts sample volume, electrical access, and spatial resolution. The review links unreliable data quality to concerns about experimental signatures and several retractions. Conventional probes also face poor spatial resolution and large background noise.

Pressure changes a material without altering its chemical composition, making it a systematic test of superconductivity. H3S reached a critical temperature, Tc, of about 203 K at 155 GPa. LaH10-δ reached about 150 K at around 200 GPa, and YH9 reached about 243 K at 201 GPa. These values placed hydrides among the main high-pressure candidates but did not resolve sample quality or magnetic response.

A diamond anvil cell, or DAC, compresses a pressure-transmitting medium inside a gasket between opposed diamond anvils. Pressure scales inversely with culet size, and a 100 micrometer diameter culet can reach pressures well beyond 100 GPa. The chamber is tiny, numerical aperture is low, and electrical access can fail under load.

Superhydride synthesis adds heat to pressure. LaH10-δ forms above 175 GPa and 1000 to 1500 K, while YH9 forms at 160 to 175 GPa and about 1500 K. Pulsed laser heating supplies the high temperature.

Ammonia borane or trapped hydrogen gas provides the hydrogen source. Ammonia borane is easy to handle but leaves residues and byproducts, and the solid source does not provide high hydrostaticity. Hydrogen offers a cleaner source, yet it diffuses into surrounding parts and can cause hydrogen embrittlement. A gold layer can reduce hydrogen diffusion.

NV centers combine optical initialization and readout with microwave spin control. Their optically detected magnetic resonance, or ODMR, spectrum responds to longitudinal and transverse magnetic fields, while pressure shifts the zero-field splitting D. The ground-state splitting is 2870 MHz, and the gyromagnetic ratio is 28 MHz per mT.

Three formats are covered: bulk slabs, implanted NV centers in [111] or [100] diamond, and fluorescent nanodiamonds. Slabs can perturb the pressure environment. Nanodiamonds down to about 30 nm can beat the optical diffraction limit, but surface charges limit pulsed performance and unknown axes complicate analysis.

Stress quenching isolates the [111] axis and removes three near-in-plane backgrounds. The selected axis reached about 30% ODMR contrast at 130 GPa. Separate defect engineering kept NV functionality to 241 GPa.

The [100] orientation retains vector information from four NV axes. A 7 micrometer pillar with a 2 micrometer deep trench restored sensitive ODMR to at least 130 GPa. The review reports possible [111] breakage under pressure, while a hydrogen-loaded pillar can fail through hydrogen embrittlement.

Interpreting contrast changes under stress

Section titled “Interpreting contrast changes under stress”

Nonhydrostatic stress can reverse ODMR contrast from a negative dip to a positive peak. The reviewed mechanism attributes this to symmetry-breaking stress, which reorders intersystem crossing rates and reverses ground-state spin polarization. The microwave transition then drives the system from a dark state to a bright state.

The review reports that the n minus state becomes dominant at about 65 GPa in one system. The n plus population remains unchanged to 100 GPa. The inversion can occur within a region narrower than 120 nm near chamber boundaries. A strong magnetic field can suppress stress-induced spin mixing.

These effects make the pressure-transmitting medium part of the measurement. Contrast changes must be separated from field expulsion caused by superconductivity.

Bhattacharyya et al. used a [111] diamond anvil with implanted NV centers, a 100 micrometer culet, and nitrogen concentration no greater than 200 ppm. A 30-keV carbon ion implant at 5 × 10^12 cm⁻² was followed by annealing above 850 °C for 12 hours. The NV layer lay about 50 nm beneath the culet surface. They then performed continuous-wave and pulsed ODMR to about 140 GPa.

Ce metal and ammonia borane were compressed above 100 GPa. A 1070-nm laser focused to about 3 micrometers heated local sites to form CeH9. Confocal ODMR tracked zero-field cooling, field cooling, and flux trapping, while integrated electrodes measured resistance.

At 118 GPa, the local ratio ΔBz/ΔHz was 0.52 on the sample. Resistance fell sharply toward zero, and Tc decreased as the applied field increased to 8 T. The map separated two disconnected superconducting regions, linking the uneven signal to synthesis inhomogeneity.

At 137 GPa, CeH9 became superconducting at 91 K in sample S2. Local magnetic cooling and resistance gave the same Tc. Several positions showed field reduction, while an edge spot showed enhancement from edge Meissner currents.

After field cooling at 103 G to 66 K and 81 K, the applied field was reduced to zero. A 34 G remanent field remained and was assigned to trapped flux. Hysteresis showed a temperature-dependent slope near 81 K but none at 66 K. A separate LaH9.6 measurement reported a Tc of about 220 K at about 150 GPa.

Widefield NV imaging covers a larger sample area than confocal microscopy. Dailledouze et al. applied it to HgBa2Ca2Cu3O8+δ, which has a Tc of about 140 K at 4 GPa. Their [100] diamond system used a 300 micrometer culet and a field of view about 100 micrometers across.

The measured Tc map varied by about 4 K, with ΔTc/Tc below 3%. Meissner screening extended beyond the physical sample boundary. A defect retained about 50% of the applied field after flux trapping, consistent with enhanced flux pinning. The widefield analysis used ODMR spectral correlation for Tc and frequency splitting and spread for field maps.

A separate BSCCO study found significant transition differences between KBr and cubic BN pressure media. The result shows that nonhydrostatic stress can shift the apparent superconducting transition.

For bulk nickelates, three 2025 studies reported diamagnetic responses and flux trapping. Mandyam et al. mapped La3Ni2O7 at 20 K under an optimal mean stress of 21 GPa. They found normal stress stabilized superconductivity, while shear stress suppressed it. The measured phase diagram reached a maximum Tc of about 80 K and linked a local field hole to the shear-stress vector. The analysis also correlated inferred stoichiometry variation with the Meissner response.

Reduced optical contrast at high pressure remains a major obstacle to wider use. It lowers sensitivity and signal-to-noise ratio, so reliable operation beyond the 241 GPa record will require improved diamond engineering, optimized NV creation, and advanced readout.

Pulsed ODMR and quantitative field reconstruction under pressure are already demonstrated. The reconstructed maps yielded optically extracted critical current density. Extending calibrated T1 and T2 measurements to vortex dynamics remains prospective. The same applies to penetration depth and coherence length measurements at megabar pressure.

The review reports sensitivity on the femtotesla per square root hertz scale for [111] NV centers at megabar pressures. It reports orders-of-magnitude gains over other NV formats and SiC color centers. The map omits explicit corrections for pressure medium and stress-induced positive contrast, which limits direct platform comparisons.

Alternative defects in SiC operate to tens of GPa, while negatively charged boron vacancies in hBN support stress mapping and magnetic imaging. Engineers should treat the anvil cut, implant layer, pressure medium, and microwave geometry as coupled design variables.

The high-pressure quantum sensor needs single-crystal diamond with shallow implanted NV centers, a defined [111] or [100] cut, and machining for a DAC anvil. The matching listed family includes MPCVD-grown single-crystal diamond in 6CCVD’s diamond materials catalog, supplied as plates and wafers up to 125 mm, with thickness from 0.1 to 500 micrometers and substrates up to 10 mm. The paper-specific implantation, annealing, culet, and readout requirements should be reviewed through engineering support.

View Original Abstract

Understanding superconductivity under extreme conditions is crucial for the search for room-temperature superconductors, yet conventional high-pressure probes often lack sufficient sensitivity and spatial resolution. This review highlights recent advances in nitrogen-vacancy (NV) center magnetometry in diamond anvil cells for studying high-pressure superconductivity. We discuss superhydride synthesis, sample inhomogeneity, and spatially resolved magnetic imaging at megabar pressures, emphasizing NV magnetometry’s ability to quantitatively visualize superconducting phenomena under extreme conditions.