Photoswitchable Radicals as Reporter Spins for Quantum Sensing with Spin Defects in Diamond
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | Journal of the American Chemical Society |
| Published | 2026-06-17 |
| Authors | Lakshmy Priya Ajayakumar, David J. Durden, Aksshay Nandakumar Regeni, Mingcai Xie, Surya Gopidas |
| DOI | 10.1021/jacs.6c09023 |
| Citations | 1 |
Photoactivated Alexa Fluor 488 radical anions on the diamond surface produced an inferred areal spin density about one order of magnitude higher than clean-diamond controls in single-NV DEER measurements. Bulk EPR showed that a significant fraction survived longer than an hour. Single-NV data also showed coherent radical Rabi oscillations and strong site-to-site variation. The work couples reporter ensembles, not a single radical, to a single NV.
Reporter spins mediate near-surface sensing
Section titled “Reporter spins mediate near-surface sensing”The NV response to a nearby magnetic source weakens rapidly with separation, so nanoscale sensing favors NVs only a few nanometers below the surface. That placement also increases surface-induced decoherence and charge instability. The proposed architecture uses surface-localized radical anions as intermediaries between an external target and a shallow NV readout. Earlier reporter-spin work used surface-bound electronic spins, called dark spins, which were unstable and lacked optical control. The new reporters are optically identifiable and photoswitchable.
Alexa Fluor 488 reaches its triplet state through intersystem crossing. MEA quenches that state and produces an S = 1/2 radical anion while fluorescence turns off. The electron donor must exceed dissolved oxygen, about 200 µM, to favor radical formation. Oxygen or 405 nm light can return the dye to its fluorescent state.
Bulk EPR and ensemble relaxometry verify radical coupling
Section titled “Bulk EPR and ensemble relaxometry verify radical coupling”Initial validation used about 400 µM Alexa Fluor 488 in 100 mM MEA at pH 9.6. The solution was unsealed and held at room temperature. Irradiation at 488 nm for 5 minutes produced radical signals in continuous-wave EPR. A series of recordings showed that a significant fraction survived beyond an hour.
The authors then measured T1 relaxation of a dense, shallow NV ensemble. The control solution contained 100 mM MEA at pH 9.6 without dye. The test solution added 400 µM Alexa Fluor 488 and received 5 minutes of continuous 488 nm irradiation. The radical-containing sample showed faster NV relaxation despite an existing sub-ms component. The authors associated that component tentatively with charge fluctuations in dense NV ensembles.
Single-NV DEER detects the radical resonance
Section titled “Single-NV DEER detects the radical resonance”The single-NV sample was a 4 mm by 4 mm by 0.5 mm electronic-grade diamond chip. One half received nitrogen-15 ions at 7 keV and 7° tilt at 1 × 10⁹ cm⁻² fluence to form isolated substitutional nitrogen atoms. The other half received 3 × 10¹³ cm⁻² to form ensembles. The chip was vacuum annealed at 800 °C, washed in piranha solution, oxygen annealed, and washed again.
For single-NV experiments, Alexa Fluor 488 was covalently attached through carbodiimide crosslinking. The mean NV depth was about 12 nm. A 50 mM MEA solution at pH 8 flowed over the chip. The authors avoided higher pH to limit pH-induced NV charge instability. They generated surface radicals with 488 nm irradiation for 30 to 120 seconds at about 5 × 10⁴ W/cm².
DEER used a 532 nm pulse to initialize the NV, a Hahn echo at the NV transition, and synchronized RF pulses at the radical resonance. Radical pulses in both echo halves reversed the reporter spins and disrupted the NV echo. Most microscopy used a DC field near 200 G along the NV axis.
For the reported spectrum at 233 G, echo delay τ was 900 ns and radical pulse duration Tₛ was 100 ns. The coherence dip occurred near 652 MHz and had a linewidth of about 20 MHz. That width matched the scale of radical hyperfine splittings seen in bulk EPR.
Coupled radicals likely explain site-to-site variability
Section titled “Coupled radicals likely explain site-to-site variability”Several NVs produced clear Rabi oscillations of the radical spins that decayed as Tₛ increased. Other sites produced heavily damped curves that plateaued above zero coherence. The finite Tₛ had to enter the model because it was comparable to the 900 ns echo delay. An instantaneous drive approximation was therefore invalid.
Independent quantum spins and classical magnetization without relaxation did not match the plateau shapes, although classical magnetization with short T2 could. Simulations of an NV coupled to a small, mutually dipole-coupled spin ensemble also reproduced both oscillations and plateaus for random spin positions. Random configurations from the same distribution covered the observed range. The authors therefore identify varying inter-radical couplings as the likely source of the heterogeneity.
Inferred spin density exceeds the dark-spin background
Section titled “Inferred spin density exceeds the dark-spin background”The line shape required a coupled-network model, so the authors used a simplified areal-density estimator rather than claiming an exact local density. The estimate uses the mean NV depth instead of each sensor’s measured depth. It underestimates density when target T2 is short and overestimates density when the signal lies below the noise floor.
Controls used nonfunctionalized diamond cleaned with a tri-acid mixture and piranha solution. They were tested in pure water and the same MEA buffer at pH 8.
| Sample set | Inferred areal density |
|---|---|
| Dye-coated, Figure 3A through Figure 3E | 0.31, 0.16, 0.14, 0.14, and 0.10 spins per nm² |
| Clean controls, Figure 3F | 0.016, 0.0098, 0.016, 0.022, 0.020, 0.036, and 0.020 spins per nm² |
On average, the coated values were about one order of magnitude above the controls. All coated values exceeded 0.05 spins per nm², the prior upper bound for dark-spin density on similarly prepared oxygen-terminated diamond. Every control remained below that bound. The control estimates also agreed with previous surface dark-spin concentrations.
The authors could not exclude an anomalously shallow NV coupled to a stable dark spin for any one trace. However, several observations favored dye-derived spins on balance.
Time dependence reveals competing photochemical effects
Section titled “Time dependence reveals competing photochemical effects”Time-resolved DEER showed that the reporter signal did not remain fixed after activation. In one measurement, data from the first 400 minutes after photoactivation gave an estimate of 0.13 spins per nm². The next 400 minutes gave 0.08 spins per nm². The decline was qualitatively consistent with conversion of radicals back to fluorescent dye. Surface holes also recovered fluorescence after a few hours in the dark.
A second measurement moved from 0.12 to 0.19 spins per nm², while the absolute contrast fell by more than half. The authors suggest that repeated 532 nm exposure may have produced additional radicals. Alexa Fluor 488 absorbs about 5% of its maximum at 532 nm, and the sequence ran for many hours. Local concentrations and laser intensities may determine whether radical formation increases or decreases.
Engineering implications and unresolved steps
Section titled “Engineering implications and unresolved steps”The main limitation is reporter density. The current interface contains ensembles on the order of tens of spins. Inter-radical coupling prevents accurate extraction of a local radical density. The authors aim to connect one NV with one radical. At lower dye density, they say optical colocalization will be needed to separate reporter signals from dark-spin background.
Surface termination, covalent dye coupling, MEA concentration, pH, and activation dose must be controlled when reproducing the reported signal. The chosen pH 8 supports radical generation while limiting pH-induced charge instability. The authors propose a thin dielectric coating deposited by atomic layer deposition to suppress dark spins and protect the NV from elevated pH. Alternative mild reducing agents could extend the operating pH range.
The demonstrated readout still requires a nearby NV, so the radicals are not yet standalone magnetic sensors. Correlative super-resolution fluorescence and nanoscale magnetic resonance imaging are proposed future applications.
Materials and support
Section titled “Materials and support”For reporter-spin magnetometry, the material requirement is an NV-enabled diamond chip with shallow, individually resolvable defects and a surface prepared for covalent dye attachment. 6CCVD’s diamond materials catalog offers MPCVD-grown single-crystal diamond plates and wafers up to 125 mm, but the offering does not state that the paper’s nitrogen implantation, 800 °C vacuum anneal, oxygen treatment, and dye coupling are included. Those device-specific steps should be reviewed through engineering support before material selection.
View Original Abstract
The rapid decay of target signal strength with distance from the sensor presents a key challenge in nanoscale magnetic sensing with the nitrogen-vacancy (NV) center in diamond, limiting the scope of accessible information as well as the sensitivity and spatial resolution with which that information can be recovered. Here, we introduce a strategy to overcome these limitations by leveraging radical anions formed from rhodamine-derived organic dyes localized to the diamond surface. These radicals, generated through photoreduction, are optically identifiable and persist on time scales exceeding an hour. We experimentally demonstrate their coherent manipulation and detection using single, shallow NV centers for readout. We observe heterogeneity in the local magnetic environments of the photoactivated spins from site to site, likely due to variations in inter-radical dipolar couplings across our measurements. Looking forward, our approach enables correlative nanoscale magnetic/optical imaging and opens new pathways for single-molecule magnetic resonance spectroscopy and quantum many-body simulations in strongly interacting dipolar-coupled spin ensembles.
High Concentration Quantum Grade Single Crystal Diamond
View Specs →