Magnetic resonance imaging, with its ability to provide three-dimensional, elementally selective imaging without radiation damage, has had a revolutionary impact in many fields, especially medicine and the neurosciences. Although challenging, its extension to the nanometre scale could provide a powerful new tool for the nanosciences, especially if it can provide a means for non-destructively visualizing the full three-dimensional morphology of complex nanostructures, including biomolecules1. To achieve this potential, innovative new detection strategies are required to overcome the severe sensitivity limitations of conventional inductive detection techniques2. One successful example is magnetic resonance force microscopy3, 4, which has demonstrated three-dimensional imaging of proton NMR with resolution on the order of 10 nm, but with the requirement of operating at cryogenic temperatures5, 6. Nitrogen–vacancy (NV) centres in diamond offer an alternative detection strategy for nanoscale magnetic resonance imaging that is operable at room temperature7. Here, we demonstrate two-dimensional imaging of 1H NMR from a polymer test sample using a single NV centre in diamond as the sensor. The NV centre detects the oscillating magnetic field from precessing protons as the sample is scanned past the NV centre. A spatial resolution of ∼12 nm is shown, limited primarily by the scan resolution.
Figure 1: Basic elements of the NMR imaging experiment.
a, A PMMA polymer sample attached to a quartz tuning fork is brought into contact with a diamond substrate containing a near-surface NV centre. The precessing protons in the sample are detected by the NV centre as the sample is scanned past it. The magnetic state of the NV centre is read out optically via spin-dependent fluorescence. b, Optical microscope image of the polymer sample. c, The sequence of microwave pulses manipulates the NV centre so that its precession phase is selectively perturbed by the oscillating proton magnetic field when the pulse periodicity matches one-half of the period of the proton field oscillations.
Figure 2: NV coherence in the presence of nearby protons.
A clear dip in coherence is observed when τ matches half the proton precession period. The decoherence is observed even when the polymer sample is retracted away from the NV centre by several micrometres, indicating the permanent presence of a proton contamination layer on the surface of the diamond, possibly adsorbed water or hydrocarbons. Analysis of the coherence dips suggests that the NV is ∼6.8 nm deep and the contamination layer is ∼1.6 nm thick.
Figure 3: Representative line scans showing the NMR signal as a function of position.
Each scan was taken at a different nominal y position, as indicated. A stepwise increase in NMR signal is seen as the sample is moved past the NV centre. The step height corresponds to a proton mean-square field of ∼(300 nT)2. The sharpest steps have transition widths of 12 nm, limited by the scan step size. Each vertical division represents a 25% change in NMR signal. For clarity, successive traces are displaced vertically by two divisions. Some variation in the x scan step size is evident due to adjustments made based on optical position measurements (see Methods).
Figure 4: Two-dimensional NMR images.
Results are from two different PMMA samples with two different NV centres. The height shown is proportional to the NMR signal s(x,y). Apparent roughness is primarily due to photon shot noise. a, Image based on 17 lines of scan data with 60 points in each scan line, including the data subset shown in Fig. 3. b, Image taken using both a different sample and a different NV centre (13 scan lines with 39 points per line).
