Controlling quantum spin coherence using a microscopic metallic structure
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | Institutional Repositories DataBase (IRDB) |
| Published | 2026-03-01 |
| Authors | Nomura Shintaro, Watanabe Hideyuki, Sasaki Susumu, Kashiwaya Satoshi |
A 10 μm Ti/Au wire with a shorted 305 mm cable produced a slow Hahn-echo T2 component of 230 μs at 36 mT. A diamond chip without the wire gave 24.3 μs under the same Hahn-echo conditions.
Control principle and experimental architecture
Section titled “Control principle and experimental architecture”The experiment controlled the electromagnetic environment around shallow nitrogen-vacancy centers after sample preparation. A Ti/Au wire tapered to 10 μm served as an impedance-matched microstrip structure on silicon. The wire connected through printed-circuit-board microstrip lines and a 305 mm RG174/U coaxial cable. Its distal end was left open or shorted, depending on the experiment. The other microstrip end used a 50 Ω termination.
The cable had a 67% wavelength reduction factor and a 109 mm effective wavelength at 1.85 GHz. Thus, its 305 mm length corresponded to approximately 2.8 effective wavelengths. The open termination produced a 0 phase shift, whereas the short termination produced π. These choices changed interference between directly induced and reflected currents near the wire. The authors used this arrangement to control microwave and radio-frequency (RF) fields around the NV centers.
Diamond preparation and optical readout
Section titled “Diamond preparation and optical readout”A 2.0 × 2.0 mm² (100)-oriented ultrapure electronic-grade diamond chip served as the substrate. A 15 nm 15N-doped, isotopically purified 12C layer with 99.999% purity was grown by MPCVD. He+ ions were implanted to promote NV center formation. The chip was annealed in Ar/H2 at 1000 °C for 1.5 h, then in air at 450 °C for 24 h. A diamond chip without the Ti/Au structure had an estimated NV density of 0.6 ppm.
Optical readout used 520 nm laser pulses, a ×50 objective with NA 0.55, and a scientific CMOS camera. Photoluminescence from the NV centers was integrated over selected micrometer regions. Imaging was required because the wire caused spatial variation in the local microwave field. A pair of permanent magnets applied a bias field along the [111] direction. Rabi and Hahn-echo measurements used 36 mT. NV electron spins were driven by a planar ring antenna and microwave pulses from an arbitrary waveform generator.
Termination-dependent microwave control
Section titled “Termination-dependent microwave control”Spatially resolved Rabi oscillations showed that cable termination set the microwave amplitude near the wire. The integration region was 1.3 × 4.0 μm² near the wire edge.
| Cable condition and region | Rabi frequency at 36 mT | Hahn-echo T2 components at 36 mT | Result |
|---|---|---|---|
| No wire structure | 3.5 MHz | 24.3 μs | Reference |
| Short cable, under wire edge | 16.2 MHz | 32 μs and 230 μs | Slow component extended |
| Open cable, under wire edge | 30.4 MHz | 14 μs and 26 μs | No clear extension |
The termination dependence was attributed to near-field current interference and reflected-current phase. Stronger local fields also damped the Rabi oscillations more strongly. The authors linked this added damping mainly to increased microwave noise. Small changes in baseline and contrast were attributed to conversion between negative and neutral NV charge states. The reported microwave field amplification factor was 4.6 to 8.7 relative to the 3.5 MHz reference.
Hahn-echo coherence and spatial reach
Section titled “Hahn-echo coherence and spatial reach”At 36 mT, Hahn-echo decay became biexponential near the wire with the short termination. The open termination also produced two components, but the authors found no clear extension over the no-wire reference. Prior studies cited by the authors assigned the initial fast decay to extrinsic factors and the slow component to intrinsic decoherence. Both terminations reduced the initial echo signal relative to the reference. The authors attributed this reduction probably to increased high-frequency RF noise from extrinsic sources. With the short termination, the echo beneath the wire exceeded the reference after τ > 70 μs. This long-delay gain suggests that low-frequency RF noise was suppressed near the wire.
The RF signals detected by Hahn echo lie within a broad range from several kHz to around 1 MHz. Surface spin noise may induce an RF current in the wire. Propagation phase along the cable is negligible across this band because the cable is much shorter than the wavelengths. The authors suggest that the termination sets the reflected RF phase and alters local noise suppression.
Measured 1.9 μm outside the wire, the fitted T2 was 23 μs with the short termination and 9 μs with the open termination. The long-delay increase was therefore short-ranged. Rabi frequencies inside and outside the edge were 16.2 MHz and 17.2 MHz, differing by less than 7%. The authors therefore link the echo contrast to wire proximity.
Selective RF signal switching
Section titled “Selective RF signal switching”At 27 mT, an XY4-6 sequence tested whether the termination could switch a low-frequency RF signal. The photoluminescence integration area was 1.3 × 26.2 μm² near the wire edge. A signal around τ = 0.44 μs appeared with the open termination and was strongly suppressed with the short termination. Its characteristic frequency was approximately 1/(2τ) = 1.1 MHz. The authors assigned the signal probably to proton NMR from oxonium ions adsorbed on the diamond surface. The XY family partially compensates pulse imperfections. That property suits this setup because near-field effects caused spatial variation in microwave pulse amplitude. The result agrees with standard electromagnetic modeling, but the authors say detailed theory is needed. Termination switching therefore controlled a detected RF signal, not only a coherence metric.
Limits and engineering implications
Section titled “Limits and engineering implications”Residual reflections arise from imperfections at the SMA connector and in wiring between the PCB and substrate pads. The total effective electrical length includes the cable, both microstrip sections, and connectors, but it was not fully controlled. Adhesion between the diamond chip and the metal structure on silicon was insufficient. The mounting method remained sensitive to substrate warping and particulate contamination. Direct lithographic fabrication of the metal wire on diamond is expected to improve control and reproducibility. The present structure had a minimum width of 10 μm. Finer structures are expected to increase coupling to the spin-noise source. The authors limit the demonstrated coherence-control range to the low-frequency regime. Extending control to higher frequencies may require improved metal fabrication. The detailed origin of the switched XY4-6 signal remains unresolved.
For device engineering, termination, wire geometry, and the mounting interface are design variables alongside diamond preparation. The short-circuit benefit was local, so placing the conductor near the shallow NV layer was essential. The authors present this as a route to high-sensitivity RF quantum sensing and improved spin-qubit fidelity.
Materials and support
Section titled “Materials and support”This quantum-sensing device requires an MPCVD-grown diamond layer with shallow NV centers, a Ti/Au conductor, and a (100) chip for optical readout. 6CCVD’s diamond materials catalog includes MPCVD single-crystal diamond and metallization with Au or Ti, matching the base diamond and conductor families. The engineering support page can assess needs for the 15 nm nitrogen-doped layer, ion implantation, and 10 μm wire, which the catalog does not specify.
View Original Abstract
Increasing the coherence time of qubits is essential for enhancing the sensitivity of quantum sensors. For diamond nitrogen-vacancy centers, spin coherence times T2 have primarily been extended through optimizing material preparation techniques such as crystal growth condition and surface treatment. Here, we demonstrate an alternative approach in which T2 is tailored by engineering the local electromagnetic environment. By adjusting the termination conditions of a micrometer-scale metal wire placed near the diamond, we successfully controlled the microwave Rabi frequency and T2. Our findings highlight the potential of micrometer-scale metallic structures as a means to externally engineer and enhance quantum coherence.
Electronic Grade Intrinsic Diamond Epitaxial Wafer H Terminated Superior 100 Mosfet Ready Substrate
View Specs →
High Concentration Quantum Grade Single Crystal Diamond
View Specs →
Boron Doped Single Crystal Diamond Plate
View Specs →Original source
Section titled “Original source”- OpenAlex: https://openalex.org/W7167403386