arXiv:2610.24815
Low-Field J-Coupling Spectroscopy Measured with a Diamond Nitrogen Vacancy Spectrometer
Stephen J. DeVience, John W. Blanchard, Camille G. McDonnell, Ronald L. Walsworth
Abstract
Nitrogen vacancy centers in diamond (NVs) are point defects well suited to quantum sensing of magnetic fields. NV spin states can be initialized and read out optically with high spatial resolution, enabling detection of pL-scale volumes of nuclear spins. Using Coherently Averaged Synchronized Readout (CASR), standard NMR spectroscopy techniques such as free decay, spin echo, and PGSE have been performed with Hz-scale spectral resolution using NV-diamond sensors. However, using NV detection for conventional high-resolution NMR at multi-tesla fields is challenging due to the need to manipulate NV electronic spins at resonance frequencies 100 GHz or higher. Low-field J-coupling spectroscopy provides an alternative path to chemical resolution in a regime more easily accessible to NVs. Here, we demonstrate homonuclear and heteronuclear J-coupling spectroscopy using an NV-NMR spectrometer based on permanent magnets operating at $\sim$890 kHz proton Larmor frequency (bias magnetic field $\sim$21 mT). This field strength enables chemical characterization based on J-coupling using sequences such as Synchronized Echo (SyncE) and Spin-Echo Double Resonance (SEDOR), which do not require the strong chemical shift dispersion and pristine field homogeneity of high-field NMR. The frequency range is easily detected by NVs using standard XY dynamical decoupling sequences and is well-suited for applying Overhauser DNP, which can boost proton signal by two orders of magnitude. We acquire initial NV-NMR measurements of T$_1$, T$_2$, and Overhauser DNP in glycerol before advancing to SEDOR spectroscopy of methanol-$^{13}$C and a methanol-$^{13}$C/acetonitrile-2-$^{13}$C mixture at sample volumes $\sim$50 pL, in which we demonstrate measurement of $^1J_\text{CH}$ couplings and resolution between species. We also acquire a homonuclear J-coupling spectrum of a natural-abundance propylene carbonate sample.
AI Summary
The paper investigates whether low-field NV-diamond NMR can obtain chemical specificity from J-couplings rather than chemical shifts, avoiding high-field microwave-control requirements. It combines permanent magnets at approximately $21\,\mathrm{mT}$ ($890\,\mathrm{kHz}$ proton Larmor frequency), coherently averaged synchronized readout (CASR), Overhauser DNP, and established spin-echo spectroscopy sequences. Initial glycerol measurements establish thermal-signal detection, $T_2=28\pm3\,\mathrm{ms}$, and DNP enhancements exceeding 100 with TEMPOL. Heteronuclear spectroscopy uses SEDOR and multi-echo CPMG-SEDOR on carbon-13-enriched samples. Methanol yields $^1J_{\mathrm{CH}}=139\pm1\,\mathrm{Hz}$ with SEDOR and $140\pm1\,\mathrm{Hz}$ with CPMG-SEDOR. A two-dimensional Fourier analysis gives a J-dimension linewidth of $4.6\pm0.8\,\mathrm{Hz}$ despite a direct-dimension linewidth of $480\pm30\,\mathrm{Hz}$. A methanol/acetonitrile mixture produces distinguishable but incompletely resolved features at $139\pm1$ and $135\pm1\,\mathrm{Hz}$. Homonuclear synchronized-echo spectroscopy of natural-abundance propylene carbonate shows dips at $12.8\pm0.2$ and $20.9\pm0.7\,\mathrm{Hz}$, consistent with spin-dynamics simulations. For sealed-tube measurements, approximately $50\,\mathrm{pL}$ contributes half the detected signal; this is not the total sample volume. The approach operates without shimming or active magnet stabilization. Stated limitations include relaxation-limited J resolution, a trade-off between DNP enhancement and radical-induced relaxation, and cumulative pulse errors. Tests at different echo spacings yield apparent methanol couplings from $136.8$ to $148.4\,\mathrm{Hz}$; simulations attribute these shifts to proton pulse errors and identify less-sensitive spacings. The homonuclear dip intensities exceed simulations and require empirical baseline correction. Single-cell metabolic imaging remains a proposed application rather than a demonstrated capability.
Error found: The approximately 50 pL volume quoted for capillary spectroscopy is a modeled 50%-signal sensing volume rather than the demonstrated total sample volume, with approximately 8 nL needed to supply 95% of the signal. The main CPMG-SEDOR coupling uncertainties do not demonstrably account for the echo-spacing-dependent pulse-error shifts documented in Appendix C, leaving absolute coupling accuracy and the quantitative interpretation of the mixture's 4 Hz separation insufficiently established.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.