Skip to content

Probing GHz SpinDynamics across Magnetic Phase Transitionsin CrCl3 Nanoflakes Using Nitrogen-Vacancy Microscopy

FieldValue
JournalACS Nano
Published2026-07-13
AuthorsBenjamin Hammons, Jitender Kumar, Sehrish Iqbal, Prem Bahadur Karki, Karishma Prasad
DOI10.1021/acsnano.6c07232

The ferromagnetic state of CrCl3 nanoflakes increased the NV spin-lattice relaxation rate by about two orders of magnitude. Broadband FMR on bulk CrCl3 placed the spin dynamics at 4 to 15 GHz and reached a linewidth of about 24 mT at 16 to 18 K. The two measurements identify a dense GHz magnetic-noise bath in the ferromagnetic phase.

Magnetic phases and the measurement target

Section titled “Magnetic phases and the measurement target”

CrCl3 orders ferromagnetically within each layer and antiferromagnetically between adjacent layers. Its paramagnetic-to-ferromagnetic transition occurs near 18 K. A ferromagnetic-to-antiferromagnetic transition follows below about 13 K. The authors measured CrCl3 nanoflakes from 4.3 to 45 K, crossing all three regimes. Their target was noncoherent GHz fluctuations in thin flakes. Bulk SQUID, neutron scattering, and FMR do not readily resolve that signal across a broad frequency range at the nanoscale. Earlier NV T2 spectroscopy examined MHz magnon sound in CrCl3. Earlier ODMR with nanodiamonds detected coherent antiferromagnetic resonance in bulk CrCl3.

Bulk CrCl3 was grown by self-transport from 0.6 g of 99.9% powder in a 20 cm evacuated quartz ampoule. The furnace reached 973 K over 24 h and held that temperature for another 24 h. A 723 to 823 K hot-to-cold gradient was then maintained for several days. Millimeter-sized crystals were mechanically exfoliated. Atomic force microscopy identified flakes of 60 ± 1 nm and 42 ± 0.5 nm thickness. Both were transferred onto the sensing diamond.

The sensing diamond was electronic grade, 2 × 1 × 0.1 mm³, and cut and polished along the (100) plane. A shallow 6 nm NV layer formed through nitrogen-15 ion implantation. The sample then received 1373 K annealing for 2 h at 2 × 10⁻⁶ Torr. A boiling 473 K tri-acid mixture removed graphite residues. The CrCl3 flake sat directly above the NV layer.

A closed-cycle cryostat held the diamond on a 25 × 25 mm sapphire carrier. VGE varnish improved thermal contact with the cold finger. Au microwave striplines supplied the drive. A 532 nm laser excited NVs over about 23 × 23 μm². An avalanche photodiode and sCMOS camera collected 650 to 710 nm fluorescence. This bandpass rejected strong CrCl3 autofluorescence above 720 nm.

ODMR used about 50 mW microwave power and 36 mW laser power at 1.45 mT. Rabi used a 10 W source at 2.5% duty cycle, corresponding to about 250 mW. $T_1$ used a 0.4% duty cycle over 20 μs, corresponding to about 40 mW. Rabi heating shifted the apparent temperature by about 2 K, and the authors included that correction.

ODMR and Rabi oscillations measured coherent NV response. $T_1$ relaxometry measured GHz fluctuations because it is sensitive to noise near the 2.87 GHz NV zero-field splitting.

Four ODMR peaks appeared for the four NV spin classes at 1.45 mT. Lorentzian-fitted $f^-$ spectra separated into three temperature bands matching the AFM, FM, and PM states. Contrast and area fell sharply in the 13 to 18 K FM window. Both recovered when the flake entered the AFM state below about 12 K. The signal-to-noise ratio was about 41.66 for 1 s averaging. The authors interpret the contrast loss as greater magnon-mediated microwave absorption and magnetic noise.

Rabi oscillations nearly vanished in the FM state. Their amplitude and fitted $T_{\mathrm{Rabi}}$ both fell, then recovered in the AFM and PM states. The microwave drive $B_1$ did not change in the FM regime. This supports decoherence from CrCl3 rather than a drive-amplitude change. Off-flake Rabi data remained featureless across all phases. These controls link the decoherence to CrCl3 spin noise.

At 30 K and 1.45 mT, the off-flake $T_1$ was 1.19 ± 0.05 ms. With the 60 nm flake present, PM-phase $T_1$ was 25.9 ± 1.6 μs. This was a roughly 45-fold reduction. The authors attribute it to PM Cr spin noise. For comparison, prior α-RuCl3 work reported a 35-fold PM reduction.

The $\Gamma_1$ curve peaked near 18 to 19 K at the PM-to-FM transition. It showed a shoulder near 12 to 13 K at the FM-to-AFM transition. For the 42 nm flake at 27.77 mT and 18 K, $\Gamma_1$ increased by about 30 times. At fields from 13.95 to 40.38 mT, the raised $\Gamma_1$ region extended from about 13 K to 24 K. The authors interpret this as field-stabilized short-range FM correlations.

The 42 nm flake sat about 20 μm from the microwave line. Its Rabi signal did not fully collapse, which the authors relate to weaker local microwave excitation.

Millimeter-sized CrCl3 crystals were mounted on a coplanar waveguide in a closed-cycle cryostat. The authors swept magnetic field at 5, 7, and 10 GHz from 13.9 to 35.6 K. Lock-in detection and derivative Lorentzian fits yielded resonance fields and linewidths. The linewidth reached about 24 mT at 16 to 18 K. The corresponding damping constant was about 1 × 10⁻³. The resonance field increased linearly in the FM phase and saturated near 209 mT in the PM phase. The two trends intersected at a Curie temperature of about 19 K, matching the NV $\Gamma_1$ peak. FM data followed the Kittel relation, while PM data were linear and ESR-like. These trends support the NV assignment of a broad magnon bath.

The relaxation model summed a common background with AFM, FM, and PM noise channels. The AFM channel used a gapped magnon dispersion. The FM channel used a quadratic Kittel-like dispersion. The PM channel used Curie-Weiss-like susceptibility and a Debye spectral factor. Smooth temperature windows controlled the crossover between channels. The windows were not microscopic phase fractions.

The FM channel reproduced the excess relaxation near the transition. AFM-associated fluctuations dominated at low temperature, while PM fluctuations described the high-temperature tail. All channels shared one background, which constrained separate offsets. The authors state that the fitted parameters are effective quantities rather than physical constants.

Four limits remain. The FMR measurements reached no lower than 13.9 K and did not enter the AFM state. NV measurements used nanoflakes, while FMR used bulk microcrystals. High-power Rabi measurements required a 2 K heating correction. The decomposition remained phenomenological despite its use of measured FMR linewidths.

The ferromagnetic phase presents the main coherence risk for nearby CrCl3 spin sensors. Magnetic fluctuations reduced ODMR contrast, erased Rabi oscillations, and shortened $T_1$ near the transition. Device thermal and field conditions must therefore account for the shifted phase boundary. A field of 13.95 to 40.38 mT extended correlated behavior into nominally paramagnetic temperatures.

For NV-assisted quantum-magnon systems, sensing-stack design must preserve optical access, uniform microwave delivery, and cryogenic thermal contact. CrCl3 autofluorescence also requires spectral separation from NV fluorescence. Phase-resolved noise measurements can guide CrCl3 selection for magnonic and hybrid quantum-magnon devices. The study also provides a nanoscale method for examining field- and strain-tuned phase diagrams.

Cryogenic NV magnetometry requires an NV-enabled (100)-oriented diamond substrate with a shallow 6 nm defect layer and Au striplines. The sensor also needs optical access and cryogenic thermal contact. The 6CCVD diamond materials catalog offers MPCVD-grown single-crystal and polycrystalline plates and wafers up to 125 mm with Au metallization. Its engineering support can help define integration needs, but the implanted 6 nm NV layer remains paper-specific.

View Original Abstract

CrCl3, a layered van der Waals (vdW) magnet, exhibits in-plane magnetic anisotropy and enhanced interlayer coupling upon stacking, making it an ideal platform to host exotic nanoscale magnetic phenomena such as magnon hydrodynamics and Meron-like topological spin defects. When interfaced with other vdW materials, its antiferromagnetic-to-ferromagnetic and ferromagnetic-to-paramagnetic phase transitions and magnetic anisotropy can be tuned by voltage, strain, and layer stacking. Understanding the spin dynamics of CrCl3 at its magnetic phase transitions is crucial to its applications in magnonics. Here, we investigate the spin dynamics of CrCl3 nanoflakes using cryogenic diamond quantum sensing microscopy, based on measuring optically detected magnetic resonance, Rabi oscillations, and spin-lattice relaxation time (T1) of shallow nitrogen vacancy (NV) centers in diamond. In the ferromagnetic regime, we observe a pronounced reduction in the NV spin resonance contrast, a collapse of the Rabi oscillations, and a strong enhancement by 2 orders of magnitude of the relaxation rate Γ1 = 1/T1. These observations indicate intensified spin fluctuations in the gigahertz range. Broadband ferromagnetic resonance spectroscopy on CrCl3 crystals reveals resonance frequencies in the 4-15 GHz range together with a line width of ∼24 mT, further supporting the NV measurements. A phenomenological model of magnetic-noise-induced NV relaxation reproduces the temperature dependence of Γ1 by combining antiferromagnetic, ferromagnetic, and paramagnetic fluctuation channels, indicating that magnetic noise is strongest in the ferromagnetic regime and evolves markedly across the phase transition. These results are crucial for using CrCl3 in 2D magnonics and hybrid quantum-magnon systems.

  1. 2017 - Discovery of Intrinsic Ferromagnetism in Two-Dimensional van Der Waals Crystals [Crossref]
  2. 2022 - The Magnetic Genome of Two-Dimensional van Der Waals Materials [Crossref]
  3. 2019 - Photonics with Hexagonal Boron Nitride [Crossref]
  4. 2025 - Progress and Challenges in the Synthesis of Two-Dimensional van Der Waals Ferroic Materials and Heterostructures [Crossref]
  5. 2019 - Direct Observation of van Der Waals Stacking-Dependent Interlayer Magnetism [Crossref]
  6. 2025 - Towards Multiferroicity in Two-Dimensional Van Der Waals Materials: Challenges and Opportunities [Crossref]
  7. 2018 - Giant Tunneling Magnetoresistance in Spin-Filter van Der Waals Heterostructures [Crossref]
  8. 2020 - Topological Superconductivity in a van Der Waals Heterostructure [Crossref]
  9. 2017 - Layer-Dependent Ferromagnetism in a van Der Waals Crystal down to the Monolayer Limit [Crossref]
  10. 1960 - On the Magnetic Properties of a CrBr3 Single Crystal [Crossref]