Si‐Doped Nanocrystalline Diamond as Highly Sensitive Luminescence Thermometer for the Physiological Temperature Range
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | Advanced Functional Materials |
| Published | 2026-07-24 |
| Authors | María Gragera, Iván Carrillo‐Berdugo, Josue Millán, Paulius Pobedinskas, Antonio Freire |
| DOI | 10.1002/adfm.77324 |
Silicon-vacancy luminescence in Si-doped nanocrystalline diamond reached a maximum relative thermal sensitivity of 2.3% K⁻¹ over 300 to 325 K. The value came from multiple linear regression of the SiV⁻ zero phonon line intensity and full width at half maximum, measured under 633 nm excitation. It was nearly twice the best single-parameter result and establishes a proof of concept for thin-film temperature sensing in the physiological range.
Growth and film selection
Section titled “Growth and film selection”The authors grew the films by microwave plasma-enhanced chemical vapor deposition in an ASTeX 6500 system at 2.5 kW. They diced silicon wafers with the ⟨100⟩ orientation into 5 by 5 mm² pieces and seeded them with about 10¹¹ nanodiamond seeds cm⁻². Nonseeded silicon pieces surrounded the samples as sacrificial silicon sources.
The authors varied pressure from 40 to 80 Torr at 1.0% CH₄. They then varied CH₄ among 0.5%, 1.0%, and 1.5% at 40 Torr. The 40 Torr condition gave the strongest SiV⁻ photoluminescence in the pressure series. The authors identified it as the practical lower growth boundary for their clam-shell reactor. Lower pressure improved silicon radical transport but also caused a severe growth-rate drop. They avoided compensating with more CH₄ because that would alter the sp³/sp² phase balance.
At 40 Torr, the 1.5% CH₄ film produced more SiV⁻ light than the 1.0% film. The authors selected 1.0% CH₄ as a compromise between emission and structure. It combined defined diamond diffraction peaks, moderate surface disorder, high sp³ content, and intense SiV⁻ emission.
| Pressure (Torr) | CH₄ (%) | Growth time | Temperature (°C) | Thickness (nm) |
|---|---|---|---|---|
| 40 | 1.0 | 3 h 37 min | 740 | 1215 |
| 45 | 1.0 | 2 h 43 min | 775 | 1080 |
| 55 | 1.0 | 2 h 56 min | 790 | 1125 |
| 65 | 1.0 | 3 h 15 min | 820 | 1170 |
| 80 | 1.0 | 3 h 17 min | 880 | 1080 |
| 40 | 0.5 | 5 h 15 min | 735 | 1080 |
| 40 | 1.5 | 2 h 59 min | 750 | 1395 |
Structure and dopant incorporation
Section titled “Structure and dopant incorporation”X-ray diffraction showed the diamond (111) and (220) peaks at 2θ values of 43.8° and 75.2° for the pressure series. The authors detected no amorphous carbon or graphite peaks. Raman spectroscopy showed the crystalline diamond peak near 1333 cm⁻¹. The graphite G band at 1583 cm⁻¹ and trans-polyacetylene feature near 1490 cm⁻¹ became stronger at lower pressure. A weak disorder-induced D shoulder appeared near 1370 cm⁻¹.
XPS found an sp³ fraction between 80% and 90% across the films. The selected 40 Torr film had an sp³ fraction of 83%. Its Si 2p signal occurred near 102 eV, consistent with Si-O-C bonding rather than elemental silicon from the substrate at 99 eV. The Si signal was undetectable in films grown at 65 and 80 Torr.
The authors attributed the lower-pressure non-diamond carbon to reduced atomic hydrogen flux and weaker etching of sp² carbon. Higher pressure improved crystallinity and sp³ content but reduced silicon incorporation. This explains why the higher-quality films produced weaker SiV⁻ photoluminescence.
Optical defect model
Section titled “Optical defect model”Room-temperature photoluminescence was recorded with a 633 nm laser, a 7 s acquisition time, seven accumulations, and a 10% neutral-density filter. After baseline subtraction, the SiV⁻ zero phonon line appeared at 739.4 nm. Its phonon sideband appeared at 755.6 nm. Cathodoluminescence at 83 K confirmed the same SiV⁻ features under electron excitation.
First-principles calculations assigned a doublet excited state to SiV⁻ and a triplet excited state to SiV⁰. The calculated SiV⁻ zero phonon line was 1.70 eV, corresponding to 729 nm. Its energy was close to the experimental 1.68 eV, although the calculated wavelength differed from the measured 739.4 nm line. The calculated SiV⁰ transition was 1.52 eV, or 816 nm. This was about 0.2 eV above the 1.31 eV experimental value from prior literature.
The calculated Huang-Rhys factors were 0.167 for SiV⁻ and 0.170 for SiV⁰. The authors interpreted these values as weak electron-phonon coupling and a strong zero phonon line with a weak sideband. Only SiV⁻ emission was detected under 633 nm excitation. They consider the framework suitable for relative defect screening, but state that absolute predictions require methods such as hybrid functionals or GW.
Temperature-dependent measurement
Section titled “Temperature-dependent measurement”The selected 40 Torr, 1.0% CH₄ film was measured from 300 to 325 K in a confocal backscattering system under vacuum. The optical setup used a 50x objective with a numerical aperture of 0.50 and 633 nm excitation. The estimated laser focus was about 2 µm.
At each target temperature, the authors performed three independent measurements. Each measurement comprised a scan of 100 spectra, and the displayed spectrum was their average. Baseline subtraction isolated the SiV⁻ emission. Gaussian fitting of the zero phonon line then extracted peak intensity, integrated area, and full width at half maximum.
Peak intensity and integrated area decreased as temperature rose, while full width at half maximum increased. The authors associated these trends with thermal quenching and the expected population of higher vibrational states.
Sensitivity and regression
Section titled “Sensitivity and regression”The single-parameter relative thermal sensitivities were 0.64 to 0.76% K⁻¹ for integrated area, 0.94 to 1.23% K⁻¹ for peak intensity, and 0.35 to 0.39% K⁻¹ for full width at half maximum. Peak intensity therefore gave the strongest individual response. The area sensitivity fell between the other two values because the fitted Gaussian area depends on both amplitude and linewidth.
The multiple linear regression used peak intensity and full width at half maximum. The authors excluded area to reduce interdependence between the model inputs. Its weighted temperature estimates showed only minor deviations from the experimental temperatures across the full 300 to 325 K range.
The combined model reached a maximum relative sensitivity of 2.3% K⁻¹. This was nearly twice the 1.23% K⁻¹ upper value obtained from intensity alone. The authors describe this multiparametric approach as self-calibrated because it can reduce the influence of fluctuations in probe concentration, excitation power, and detection conditions.
Limits for biological use
Section titled “Limits for biological use”The experiments measured the material response under controlled laboratory conditions. The authors state that biological media could alter the result through ionic strength, dynamic pH, and protein coronas. These conditions may cause surface band bending and affect SiV⁻ charge-state stability. Validation in physiological buffers and cellular microenvironments remains necessary before claiming in vivo performance. Tests under surface functionalization and nanostructured probe geometries also remain outstanding.
The measured signal was an ensemble response from multiple SiV-containing nanograins within the excitation volume. It was not isolated single-particle emission. Free-floating nanodiamonds can migrate within fluids, whereas a fixed thin film can remain attached to a probe. The authors propose coatings on optical fiber tips or implantable microneedles, with spatial resolution determined by probe geometry. This device architecture was suggested but not demonstrated.
Earlier SiV thermometry relied mainly on bulk single-crystal diamond or top-down nanostructures, which the authors cite as limited by scalability, cost, and geometry. Earlier N and Si co-doped nanocrystalline diamond also produced broad NV phonon sidebands. Removing nitrogen co-doping gave this study a cleaner SiV⁻ spectral response.
Materials and support
Section titled “Materials and support”The paper requires Si-doped nanocrystalline diamond with optically active SiV⁻ centers for a proposed thin-film temperature probe on optical fiber tips or implantable microneedles. 6CCVD’s diamond materials catalog includes single-crystal and polycrystalline diamond grown by MPCVD, the closest listed family for that probe architecture. The listed facts do not specify silicon doping, so the required SiV⁻ formulation and 40 Torr process should be reviewed through engineering support.
View Original Abstract
ABSTRACT Si‐doped nanodiamonds are emerging as highly promising materials for luminescent thermometry in the physiological temperature range (300-325 K) due to their excellent biocompatibility, chemical inertness, and optical transparency, as well as their ability to host negative charged silicon‐vacancy (SiV − ) centers with sharp, photostable emission within the biological transparency windows. In this work, Si‐doped nanocrystalline diamond thin films were synthesized on silicon substrates via microwave plasma‐enhanced chemical vapor deposition (MPECVD) under various growth pressures and methane flow conditions, aiming to obtain the practical boundary limit of the photoluminescence intensity. The temperature dependence of key photoluminescence parameters, including intensity, full width at half maximum (FWHM), and area of the zero‐phonon line emission, was investigated in the physiological range (300-325 K). The experimental photoluminescence results were interpreted with the support of first‐principles density functional theory calculations, linking the defect electronic structure and electron-phonon coupling to the observed optical response. By employing multiparametric analysis, combining these thermally sensitive luminescence features in a single model, a self‐calibrated approach for robust thermometry was demonstrated, optimizing the thermal sensitivity of the SiV − centers‐based emission of the MPECVD diamond. A maximum relative thermal sensitivity of 2.3% K −1 was achieved, highlighting the potential of Si‐doped nanodiamonds for accurate and minimally invasive temperature monitoring in complex biological environments.
Free Standing Boron Doped Diamond Pcd Film
View Specs →
Boron Doped Single Crystal Diamond Plate
View Specs →