WO2023234979A2 - High field magnetometry with hyperpolarized nuclear spins - Google Patents

Abstract

A nuclear magnetic spectroscopy system includes a sensor containing hyperpolarized nuclei, and a detector to detect changes in precession of the sensor nuclei when they are exposed to an analyte. A method of nuclear magnetic spectroscopy includes exposing a sensor containing hyperpolarized nuclei to an analyte, using a detector to detect changes in precession of the hyperpolarized diamond nuclei caused by the analyte, and identifying the analyte by the changes.

Description

HIGH FIELD MAGNETOMETRY WITH HYPERPOLARIZED NUCLEAR SPINS

RELATED APPLICATION

[0001] This application claims priority to and the benefit of US Provisional Patent Application No. 63/291,880, filed December 20, 2021, which is incorporated herein in its entirety.

STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH AND DEVELOPMENT

[0002] This invention was made with Government support under award number DE- AC0205CH11231, awarded by the Department of Energy, and award number N00014-20-1, Office of Naval Research. The Government has certain rights in this invention.

TECHNICAL FIELD

[0003] This disclosure relates to nuclear magnetic resonance (NMR) spectroscopy, more particularly to NMR spectroscopy in high fields.

BACKGROUND

[0004] The discrimination of chemical analytes with sub-micron scale spatial resolution is an important frontier in Nuclear Magnetic Resonance (NMR) spectroscopy. Quantum sensing methods have attracted attention as a pathway to accomplish these goals. Such sensors predominantly operate a low magnetic fields. Instead, however, for high-resolution spectroscopy, the high-field regime is naturally advantageous because it allows high absolute chemical shift discrimination.

[0005] These are t pified by sensors constructed from the Nitrogen Vacancy (NV) defect center in diamond - electrons that can be optically initialized and interrogated, and made to report on nuclear spins in their environment. However, NV sensors are still primarily restricted to bulk crystals and operation at low magnetic fields Bo < 0.05 T). Instead, for applications in NMR, high fields are naturally advantageous because the chemical shift dispersion is larger, and analyte nuclei carry higher polarization.

[0006] The challenge of accessing this regime arises from the rapidly scaling electronic magnetogyric ratio, y_e, which makes electronic control difficult at high fields.

Simultaneously, precise field alignment is required to obtain viable NV spin-readout contrast. The latter has also made nanodiamond (particulate) magnetometers challenging at high fields. If viable, such sensors could yield avenues for "targetable" NMR detectors that are sensitive to analyte chemical shifts, and provide a sub-micron scale spatial resolution determined by particle size.

BRIEF DESCRIPTION OF THE DRAWINGS

[0007] FIGs. 1 A-B shows graphical representations of diamond lattice with hyperpolarized nuclei.

[0008] FIGs. 2A-C show graphs related to a hyperpolarized sensing sequences.

[0009] FIGs. 3A-D show graphs related to signals in a NMR spectroscopy system.

[0010] FIGs. 4A-G show graphs related to enhanced signal delay with an applied AC field.

[0011] FIGs. 5A-H show graphs related to high-field sensor magnetometry

[0012] FIGs. 6A-C show graphs related to the frequency response of a ¹³C magnetometer. [0013] FIGs. 7A-D show graphs related to tracking magnetometer signals in the time domain. [0014] FIGs. 8A-C show views of an embodiment of an NMR system and probe.

DETAILED DESCRIPTION OF THE EMBODIMENTS

[0015] The embodiments here use an alternative approach towards overcoming challenges. The embodiments use a magnetometer constructed out of hyperpolarized nuclear spins, a representation of which is shown in FIGs. 1A-B. FIG. 1 A shows a diamond lattice with dipolar (dashed lines) coupled ¹³C nuclei, hyperpolarized by optically pumped nitrogen vacancy (NV) center defects, shown as blue arrows. FIG. IB Hyperpolarized ¹³C nuclei are driven into the x axis (red arrow) via spin-locking. When the AC field (blue) is applied, the spins undergo secondary precessions shown by the arrows. ¹³C nuclei in the diamond serve as the primary magnetic field sensors, while NV centers instead play a supporting role in optically initializing them.

[0016] In one embodiment, the sensor contains hyperpolarized nuclei. While the cunent discussion focuses on ¹³C nuclei, any hyperpolarized nuclei can be used. The sensor is exposed to the analyte nuclei by being brought into proximity to analyte. The detection occurs via time-varying fields produced by the analyte on the probe nuclei. The detector may be optical or radio frequency and detects the changes in the precession of the hyperpolarized nuclei caused by the analyte. These changes allow for identification of the analyte.

[0017] The advantages of ¹³C nuclei as sensors stem from their attractive properties. Their low y_n ~ y_e /3000, enables control and interrogation at fields in the range of (Bo> 0.05 T). The field may comprise a magnetic field of at least 0.05 T, but may have a range of at least 0.1 T, at least 0.3 T, or a range of 0.5 T. Spin-1/2 ¹³C sensors can be housed in randomly oriented particles, as they are agnostic to crystallite orientation. They have long rotating frame lifetimes T₂\ orders of magnitude greater than their NV center counterparts. Similarly, their longitudinal lifetimes Ti > 10 min are long even at modest fields. This can allow a physical separation betw een field regions corresponding to ¹³C initialization and sensing, and for the ¹³C sensors to be transported between them. ¹³C nuclei can be non-destructively readout via

RF techniques, allowing continuous sensor interrogation without reinitialization. This portends magnetic field tracking for extended periods. RF ¹³C readout is also background-free and immune to optical scattering, increasing amenability to deployment in real-world media. [0018] While these properties appear attractive at first glance, nuclear spins are often considered ineffective as quantum sensors. The low y_n, while ideal for high field operation, would be expected to yield low sensitivity, and poor state purity. The thermal polarization is ~ 10'⁵ even at 7 T. Strong (~ 1 kHz) dipolar coupling between ¹³C nuclei makes Ramsey-like sensing protocols untenable. This results in a rapid free induction decay (FID) T₂ < 2 ms, and limits sensor integration time.

[0019] The embodiments here demonstrate that these shortcomings can be mitigated. They exploit optical hyperpolarization of ¹³C nuclei, as shown in FIG. 1A and spin-lock readout scheme that suppresses evolution under dipolar interactions. The resulting greater than 10,000-fold extension in ¹³C lifetimes, from T₂ T₂“, provides the basis for expanding sensor readout to minute long periods. These long ¹³C rotating-frame lifetimes can at least partially offset sensitivity⁷ losses arising from the low yn. The sensing strategy' is described in FIG. IB. Hyperpolarized ¹³C nuclei are placed along the transverse axis x (red arrow) on the

Bloch sphere at high field, where they are preserved for multiple-second long T2 periods. Any subsequent deviation of the spin state from x-y plane can be continuously monitored and constitutes the magnetometer signal. In the presence of the target magnetic field, Bac(t) =BAC COS(2TI/AC¹)Z at a frequency /AC, the nuclei undergo a secondary precession in the y-z plane that carries an imprint of / C. Long T₂ yields high spectral resolution.

[0020] Experiments were conducted on a single-crystal diamond, but can be extended to powders and materials. These may include xenon, pentacene, anthracene, and porphyrin. The sample has ~1 ppm NV center concentration and natural abundance ¹³C. Hyperpolarization occurs through a method previously described at 38 mT. Examples of techniques that may be used to hyperpolarize the nuclei are discussed in US Patent Publications 202102216, and 20210364583, incorporated by reference herein in their entirety. FIG. 2 shows the subsequently applied ¹³C magnetometry protocol at 7 T. It entails a train of 0 pulses, spinlocking the nuclear spins along x, as shown in FIG. 2A. ¹³C nuclei remain in quasi- equilibrium along x for several seconds. Flip angle 0 can be arbitrarily chosen, except for

0=7C. Pulse duty cycle is high (19-54%) and interpulse spacing r < 100 ps (FIG. 2A). The nuclei are inductively interrogated in tacq windows between pulses. The points in FIG. 2B show typical raw data. For each window, the magnitude of the heterodyned Larmor precession, (FIG. 2C, here at 20 MHz) — effectively the rotating frame magnetization component — constitutes the magnetometer signal. These are the blue points in FIG. 2D. Every pulse, therefore, provides one such point. In typical experiments, the process applies N

200k pulses. In the absence of an external field, successive points show very little decay. However, with an AC field at fs.c. there is an oscillatory response, shown by the blue line FIG. 2D). It is strongest at the “resonance condition ”,

Corresponding to when the AC field periodicity is matched to the time to complete 2TI rotation. FIG. 2A describes the resonant situation for 0 = n/2.

[0021] FIG. 3 shows the magnetometer signal over 20 s (275 k pulses) with no applied field and /AC = 2.760 kHz (close to resonance). In the former case, there is a slow decay T2 = 31 s shown by the red line in FIG. 3A. The AC field however causes a comparatively rapid ¹³C decay shown by the blue line in FIG. 3A along with magnetization oscillations, shown in FIG. 3C. The process separates these contributions, decomposing the signal (dashed box) as S = Sd + So, where Sd is the (slow) decay component as shown in FIG. 3B and So is the oscillatory part with zero mean, shown in FIG. 3C, and zoomed in FIG. 3D. In practice, Sd is obtained via a 73 ms moving average filter applied to S in FIG. 2, and the oscillatory component isolated as So - S - Sd.

[0022] The process first focuses attention to the decay component Sd. In FIG. 4, the process varies / C, and study the change in integrated value of Sd over a 5 s period, shown in FIG. 3B. Here the phase of the applied AC field is random. FIG. 4A reveals that the ¹³C decay is unaffected for a wide range of frequencies, except for a sharp decay response, referred to here as the dip, centered at resonance fies shown by the dashed vertical line. With a Gaussian fit, the line width is estimated ~ 223 Hz. The dip matches intuition developed from average Hamiltonian theory (AHT), shown in FIG. 4B-C, and discussed below. It can be considered to be an extension of dynamical decoupling (DD) sensing for arbitrary 9. FIG. 4D shows the scaling of /AC with pulse width t_P, and agrees well with theory, assuming t_P =x for 0 = n/2.

[0023] FIGs. 4E-F, display individual decays in FIG. 4A on a linear scale and logarithmic scale against Points far from resonance (e g. (i) DC and (iv) 5 kHz) exhibit a stretched exponential decay oc exp(-t^1/2) characteristic of interactions with the Pl center spin bath. These manifest as the straight lines in FIG. 4F. On the other hand, for points within the /res dip in FIG. 4A, one can observe a potentially chaotic response evidenced by sharp jumps, shown in FIG. 4E-F. These features belie a simple explanation from AHT alone. The jumps occur for strong Rabi frequencies, and when /AC approaches resonance, evident in the comparison between /AC=5 kHz and/AC=2.885 kHz in FIG. 4E-F, shown by the blue and yellow lines. Elucidating the route to chaos here is beyond the scope of this manuscript and will be considered elsewhere. FIG. 4G describes the linewidth dependence of the obtained resonance dip as a function of the number of pulses employed. Contrary to DD sensing, the linewidth does not fall with increasing number of pulses, suggesting it is dominated by ¹³C dipolar couplings. [0024] Despite this relatively broad linewidth, high-resolution magnetometry can be extracted from the oscillatory component, So, shown in FIG. 3C. FIG. 5A shows So over 20 s for /AC=2.760 kHz. FIG. 5B zooms into a representative 22 ms window. Strong ¹³C oscillations are evident here. Taking a Fourier transform, one observes four sharp peaks, shown in FIG. 5C. The process identifies the two strongest peaks as being exactly at JAC and 2 fxc, as show n in FIG. 5C with the labels 1 and 2. This discussion refers to them as primary and secondary harmonics respectively. They are zoomed for clarity in FIGs. 5D-E, along with the noise level in FIG. 5F, from where the process extracts the AC magnetometry linewidths as 92 mHz and 96 mHz respectively. Two other smaller peaks in FIG. 5C are aliased versions of the third and fourth harmonics, marked 3 and 4, at frequencies fs = 2B - 3JAC = 5.418 kHz and ft = 2®- 4 /AC = 2.658 kHz. The bandwidth ® = 1/(2T) here is determined by the interpulse interval in FIG. 2A. For clarity, FIG. 5G show s the data in FIG. 5C in a logarithmic scale, with the harmonics marked.

[0025] The embodiments emphasize differences with respect to DD quantum sensing. Oscillations here are at the absolute AC frequency JAC, as opposed to the difference frequency from T'¹ in the DD case. Second, sensing can be obtained with an arbitrary flip angle 0 (except 0), allowing for greater robustness compared to DD sequences that require precisely calibrated 0 = rr pulses. Pulse error affects AC magnetometry peak intensity but not their position. FIG. 5H shows the scaling of the harmonic intensities with |BAC|. One can observe a linear and quadratic dependence of the primary and secondary harmonic intensities, as shown in FIG. 5H), differing from DD sensing.

[0026] The inventors performed experiments to determine the frequency response of the sensor, unraveling the sensitivity' profile at different frequencies, shown in FIG. 6. The experiments seek to determine how it relates to the /res dip in FIG. 4. The experiments applied a chirped AC fiel kHz, in a AB =1-4 kHz window in FIG. 6A as the shaded

region. The AC field mimics the effect of the analyte on the nuclei.

[0027] The sweep is slow, (7=20 s), and occurs only once during the full sequence, and does not start synchronously with it. If the frequency response was independent of frequency, one would expect an approximately box-like Fourier signal intensity' over AB. Instead, the experiments obtain a narrow response with a central cusp in a small portion of the AB band, shown as zoomed in FIG. 6B. This response is strongest resonance (Eq. (1)). Simultaneously, the inventors observed a strong Gaussian response in the region outside AB, shown in FIG. 6C. The process identifies these features as the primary and secondary' harmonic response respectively (labeled 1 and 2 as before). Indeed, their frequency centroids are in the ratio 1:2 as cam be seen in FIG. 5C./res arises where the cusp dips to its lowest point, show n as the dashed line in FIG. 6B. It is believed that the cusp indicates an effect from ¹³C dipolar interactions.

[0028] The discussion now turns to a description of ¹³C sensor operation, and outlines two complementary' viewpoints to explain the observations in FIG. 4 and FIG. 5. A first picture equivalent to DD quantum sensing, and second using a “rotating-frame” NMR experiment analogue. Consider first that the ¹³C Hamiltonian in the lab-frame is, 3~C = %_z + %dd + %AC, where H=wiL is the Zeeman Hamiltonian, 9-C_dd=

is the interspin dipolar interaction and JCAC = y_n BAC cos(2a fpd + (po)L is the applied AC field. Here I refers to spin-1/2 Pauli matrices, WL is the nuclear Larmor frequency, (po is the initial (arbitrary) phase of the AC field, and then the median dipolar coupling is estimated as J = <bki> ~ 663 Hz. The spins are prepared initially along x in a state p(0) ~ sA. where e ~ 0.2% is the hyperpolarization level. [0029] The sequence in FIG. 2A can be conveniently treated in the rotating frame by average Hamiltonian theory (AHT). After N pulses, its action can be described by the unitary, U(Nr) = |exp(/Wv) e.\p(/Jfr)] ^v. where one assumes 5-pulses for simplicity. The signal obtained following the procedure in FIG. 2 then simply corresponds to the measurement of |<A>² + <I_y>2 ]^1/2. or equivalently the spin survival probability in the x-y plane. This evolution can be expressed as U (AT) =

are toggling frame Hamiltonians after every' pulse, Jf(j) = exp(ij6Ix)J£exp(-iJ0Ix). For time t=Ni, this can be recast as, U(t) = expCMi Ni) where

captures the effective system dynamics under the pulses. To leading order in parameter £ = 2TT.IT in a Magnus expansion, and assuming ( «1, the dynamics can be captured by an average Hamiltonian,

■

[0030] Since non-commuting effects do not affect the leading order term, one can consider the effect of each of the dipolar part and the AC field separately, the former, there is the equation below,

with the flip-flop Hamiltonian, J-Ca = Iplkz + Ip Iky. The initial state p(0) is conserved under sinc 0. This leads to a quasi-equilibration of spins along x, with a

lifetime that scales with a power law of the pulsing frequency T¹. This is the red signal in FIG. 3A, and here . One should note that for sufficiently smal s valid for

arbitrary flip-angle 0, except for certain special values (0=7t, 2JI).

[0031] A similar AHT analysis can be carried out for the 7f.\c term. Consider first a DC field (/AC=0) and 0=7i/2. FIG. 2C shows the toggling frame Hamiltonians

, which consists only of single body terms and hence can be plotted in a phasor representation. In a cycle consisting of four pulses (required to complete a 2n rotation), the average Hamiltonian

, evident from the symmetrically distributed phasors in FIG. 4(i). Hence the DC field is decoupled. Alternately, consider the resonant AC case (/AC =fns). The analysis here is simplest to carry out assuming a square-wave, as opposed to sinusoidal, field. In this situation, the phasor diagram is asymmetrical and the average Hamiltonian after four pulses, H_AC oc — ly. When stroboscopically observed, the spins rotate away from the x-y plane. This yields the dip in the integrated signal in FIG. 4.

[0032] An illuminating alternate viewpoint is obtained by noting that under spin-locking the spins are requantized in the rotating frame with an effective field. Qeff = fl where the

factor in brackets is the pulsing duty cycle. The resonance frequency identified above is then exactly fies= Qeff. Therefore, the experiment can cast as a rotating-frame analogue of a conventional (lab-frame) NMR experiment. The pins are quantized along x, Qeff sen es analogous to the Larmor frequency, and at the resonance condition (JAC =fres), BAC is the effective Rabi frequency. Spins initially prepared along x, are then constantly tipped away from this axis by /AC. The dip in FIG. 4 reflects this tilt away from the x-y plane. For the resonant case, the trajectory of the spins in the rotating frame can be simply written as:

In effect, the spins are undergoing a “secondary" precession in the rotating frame around x at frequency Q_cff. For each point on this motion, they are also precessing in the lab frame at ML.

The latter yields the inductive signal measured in the raw data in FIG. 2B. Upon taking a

Fourier transform, as shown in FIG. 2C, one can extract the magnitude of the spin vector in the x-y plane, which has the form, S(t) = \cos²y_nBAct) + sin² (ynBAd)sin²(2'n:t Act)] ^1/2 . This is the signal measured in FIG. 3, and the oscillations here at Ac and 2 Ac, corresponding to the observed first and second harmonics respectively. While this analysis was for the resonant case, it is simple to extend it to off-resonant AC fields. Once again, the oscillations can be demonstrated to be at exact harmonics of Ac. For sufficiently small fields BAC « JAC, the amplitude signal oscillations So x B^_c. Alternatively, if the FT phase (instead of magnitude) was taken in FIG. 2C, measuring now⁷ the phase of the spins in the x-y plane in the rotating frame, Eq. (3) indicates there will once again be oscillations. How ever, these will scale oc B%_c. and can provide higher sensitivity. Extracting these oscillations comes with complication related to unwrapping the phase every 2n, and accounting for phase accrual during the t_P pulse periods.

[0033] Overall therefore, Eq. (3) illustrates that the oscillations observed in FIG. 5 are equivalent to w atching the Larmor precession of the spins in the rotating frame. Since the sequence suppresses static I-field inhomogeneity, the lifetime of the oscillations can extend up to T^, as evidenced in FIG. 5A. The growing oscillation strength in FIG. 5A indicates the spins tipping further away from the x axis. However, projections of the spin vector away from x do undergo dipolar decay, and the true amplitude of the oscillations observed depend on an interplay betwee and Ac. Finally, one should note that Eq. (3)

demonstrates that FIG. 7 is analogous to the AC field driving a rapid adiabatic passage in the rotating frame.

[0034] The discussion can now⁷ elucidate parameters of ¹³C sensor performance and discuss some of its special features. The emphasis of the embodiments is not to optimize sensitivity, but from FIG. 5E, one can estimate its value to be ~330 nT^Hz Bo=7 T. This neglects hyperpolarization time because the sensor interrogation time can be exceedingly long. While the sensitivity' here is lower than NV quantum sensors at low field, this value is still competitive with respect to other high-field sensors. Sensitivity can be significantly enhanced. Sample filling factor, presently r| ~ 10%, and the hyperpolarization level (~ 0.3%) can both be boosted by close to an order of magnitude. Employing a 10% ¹³C enriched sample would provide a further ten-fold gain in inductive signal strength. Measurement of FT phase in FIG. 2, as opposed to magnitude, will also yield a significant sensitivity' gain. From this, one can estimate a sensitivity approaching InT /HZ is feasible in a (10pm)³ volume, sufficient to measure fields from precessing ¹H nuclei in the same volume at high field.

[0035] ¹³C sensor resolution is 5 ~ 1/NT. Currently, FIG. 5 demonstrates a resolution better than 100 mHz. However, finite memory limitations restricted capturing the ¹³C Larmor precession here to t < 30s as shown in FIG. 5A. Overcoming these memory limits can allow acquisition of the entire spin-lock decay, lasting over 573 s. Under these conditions, an estimate a resolution of 2.2 mHz is feasible. This would correspond to a field precision of 3 ppt at a 7 T bias field, more than sufficient precision to discriminate chemical shifts. On the other hand, sensor bandwidth ®=l/2r is determined by the minimum interpulse delay from FIG. 4C. Current Rabi frequencies limit bandwidth to ~20 kHz. One can estimate that improving filling-factor q and RF coil Q factor (presently 50), could increase ® further to ~500 kHz. The sensor strategy lends itself to a wide operating field range. While the experiments were carried out at 7 T, the slow scaling ¹³C magnetogyric ratio makes sensing easily viable even for fields greater than 24 T. This greatly expands the field range for spin sensors, where the operating field is predominantly less than 0.3 T. [0036] One should note the robustness of the sensing method to pulse error. It can be operated with any flip angle 0 n. Moreover, yies has a relatively wide profile -230 Hz, such as shown in FIG. 4A, meaning that flip-angle (RF) inhomogeneity has little impact. These features are responsible for the greater than 275 k pulses applied to the ¹³C spins in experiments in this work.

[0037] Finally, one should note some special features of the magnetometry protocol of the embodiments. Compared to previous work using hyperpolarized gaseous ¹²⁹Xe nuclei as sensors, the experiments employed hyperpolarized nuclear spins in solids. This provides natural advantages due to the ability for in-situ replenishment of hyperpolarization at the sensing site. Moreover, multiple AC fields can be discerned in a Fourier reconstruction in a single-shot, as opposed to point-by-point.

[0038] Fundamentally, experiments here illustrate the feasibility of quantum sensing in the coupled sensor limit, making the spins sensitive to external fields while negating the effect of intersensor interactions. Sensor operation exploits “Floquet prethermalization” - quasiequilibrium nuclear states under periodic driving. As such, this provides a compelling demonstration of exploiting stable non-equilibrium phases for sensing applications.

[0039] From a technological perspective, RF interrogated sensing, as described here, presents advantages in scattering environments. All data here were carried out with the diamond immersed in ~4mL water, over 2000-fold the volume of the sample. Traditional NV sensors are ineffective in this regime due to scattering losses and concomitant fluorescence fluctuations. Similarly, optically hyperpolarized sensors present advantages because majority of the sensor volume can be illuminated by the impinging lasers, with no geometrical constraints from requirements of collection optics. The experiments employed an array of low-cost laser diode sources for hyperpolarization, allowing recruiting a large volume of spins for sensing with a low overhead. Extension to powder samples could be advantageous for optimally packing a sensor volume.

[0040] The work presented here can be extended in several promising directions. First, the high resolution-to-bandwidth ratio (6/' ®~l 0^fi possible via ¹³C sensing at high fields, as in FIG. 5, suggests possibilities for detecting chemical shifts from precessing analyte nuclei external to the diamond. Seminal work by Warren et al. (W. S. Warren, W. Richter, A. H. Andreotti, and B. T. Farmer, “Generation of impossible cross-peaks betw een bulk w ater and biomolecules in solution nmr,” Science 262, 2005 (1993), and W. Richter, S. Lee, W. S. Warren, and Q. He, “Imaging with intermolecular multiple-quantum coherences in solution nuclear magnetic resonance,” Science 267, 654 (1995)) and Bowiell (R. Bowtell, “Indirect detection via the dipolar demagnetizing field,” Journal of Magnetic Resonance (1969) 100, 1 (1992)) showed that nuclear spins of one species (“sensor”) could be used to indirectly probe NMR information of other physically separated (“analyte”) nuclei.

[0041] This indirect NMR detection strategy is appealing as the absolute dimensions diminish to sub-micron length scales. Building on these ideas, one can envision the possibility of micro-scale NMR detectors using hyperpolarized ¹³C nuclei in nanodiamond (ND) particles. These experiments require control applied to the target nuclei in order to heterodyne chemical shifts to within the ¹³C sensor bandwidth.

[0042] Second, while the current experiments exploit a low-field DNP mechanism for the initialization of the ¹³C sensors, interesting opportunities arise from employing complementary all-optical DNP techniques that operates directly at high field. This will allow in-situ sensors at high field without the need for sample shuttling.

[0043] More broadly, this work suggests an interesting applications of DNP for quantum sensing. It is intriguing to consider sensor platforms constructed in optically-active molecular systems where abundant, long-lived nuclear spins, can be initialized via interactions with electronic spin centers. Finally, one can envision technological applications of the ¹³C sensors described here for bulk magnetometry, sensors underwater and in scattering media, and for spin gyroscopes.

[0044] The embodiments here have proposed and demonstrated a high-field magnetometry approach with hyperpolarized ¹³C nuclear spins in diamond. Sensing leveraged long transverse spin ¹³C lifetimes and their ability to be continuously interrogated, while mitigating effects due to interspin interaction. The embodiments demonstrated magnetometry⁷ with high- resolution (. lOOmHz) and at high-field (7T), yielding advantages over counterpart NV sensors in this regime. This work opens avenues for NMR sensors at high fields, and suggests interesting possibilities for employing dynamic nuclear polarization for quantum sensing.

[0045] FIG. 8 shows a probe set up used in the above experiments that gives on overview of the basic components of a system. FIG. 8A shows an embodiment of internal probe, or sensor or detector, components showing a RF coil used for ¹³C NMR and a z-coil by which the test AC is applied. The probe 10 has the RF coil 12 and the z-coil 14 into the centers of which the sensor is shuttled. The probe may include capacitors 16 to manage the electrical signals in the detector.

[0046] FIG. 8B shows a zoomed out view of the sensor. In one embodiment, the probe comprises oxygen-free high (thermal) conductivity (OFHC) coil, and an OFHC shield 18. In one embodiment, the OFHC shield is 54 nm in diameter. The funnel 20 provides a port through which the hyperpolarized nuclei can be inserted into the probe after exposure to the analyte. [0047] Both coils connect to independent rigid coaxial cables 36 and 58, but share a common ground. The diamond sensor may be held under water in a test tube and then shuttled into the center of the RF coil.

[0048] FIG. 8C displays an embodiment of a circuit. The upper block 30 shows an embodiment of a circuit used for the AC field application. The upper block 30 is only used as an estimation of an analyte. Analytes generate a time-varying field, referred here as an AC field, that affects the hyperpolarized nuclei, so the block 30 generates a low-level AC field to mimic the effect of exposure to an analyte.

[0049] The NMR block 40 shows an embodiment of an NMR excitation and detection circuit. The NMR block 40 includes a signal generator 44 to generate the signals at RF frequencies. A quarter wave plate or line 46 and the filter 48, one embodiment a bandpass filter, ensure a high signal-to-noise ratio of the RF signal. A transmit and receive switch 56 allows switching between the two modes of the circuit. In one embodiment, the RF field coil 12 comprises an RF saddle coil. On the detection side of the switch 52 is an amplifier 54 and a NMR spectrometer that can detect the changes in the spins for analysis and identification of the analyte.

[0050] As discuss above, the process of using this detector involves exposing hyperpolarized nuclei to an analyte. The hyperpolarization of the nuclei may occur in one of many ways, as mentioned above. The hyperpolarized nuclei, after exposure to the analyte, is interrogated by application of a radio frequency signal, typically in the form of a pulse sequence. In one embodiment, the nuclei is inserted into a probe structure that resides inside a magnet that generates a high-field magnetic field. The resulting responses of the material generate a series of points that can be processed and analyzed to allow identification of analyte. [0051] All features disclosed in the specification, including the claims, abstract, and drawings, and all the steps in any method or process disclosed, may be combined in any combination, except combinations where at least some of such features and/or steps are mutually exclusive. Each feature disclosed in the specification, including the claims, abstract, and drawings, can be replaced by alternative features serving the same, equivalent, or similar purpose, unless expressly stated otherwise.

[0052] It will be appreciated that variants of the above-disclosed and other features and functions, or alternatives thereof, may be combined into many other different systems or applications. Various presently unforeseen or unanticipated alternatives, modifications, variations, or improvements therein may be subsequently made by those skilled in the art which are also intended to be encompassed by the embodiments.

Claims

WHAT IS CLAIMED IS:

1. A nuclear magnetic spectroscopy system, comprising: a sensor containing hyperpolarized nuclei: and a detector to detect changes in precession of the sensor nuclei when they are exposed to an analyte.

2. The system as claimed in claim 1, wherein the hyperpolarized nuclei comprise hyperpolarized nuclei of one selected from the group consisting of: diamond, xenon, pentacene, anthracene, and porphyrins.

3. The system as claimed in claim 1, wherein the detector comprises a detector that produces an RF field to interrogate the hyperpolarized nuclei.

4. The system as claimed in claim 1, wherein the detector further comprises an RF coil connected to a signal generator to cause the RF coil to produce the RF field.

5. The system as claimed in claim 4, wherein the detector further comprises a quarter wave line between the signal generator and the RF coil.

6. The system as claimed in claim 4, further comprising a nuclear magnetic resonance spectrometer to detect changes in the hyperpolarized nuclei caused by exposure to the analyte.

7. The system as claimed in claim 1, further comprising a magnet surrounding the sensor.

8. The system as claimed in claim 6, wherein the magnet produces a magnetic field of at least 0.5 T.

9. A method of nuclear magnetic spectroscopy, comprising: exposing a sensor containing hyperpolarized nuclei to an analyte; using a detector to detect changes in precession of the hyperpolarized nuclei caused by the analyte; and identifying the analyte by the changes.

10. The method as claimed in claim 9, further comprising hyperpolarizing the nuclei separate from the detector.

11. The method as claimed in claim 10, wherein hyperpolarizing the nuclei comprises hyperpolarizing the nuclei using optical pumping.

12. The method as claimed in claim 9, wherein using a detector comprises: inserting the hyperpolarized nuclei into a probe structure contained inside a magnet; applying a radio frequency signal to the hyperpolarized nuclei; and measuring a resulting signal to generate a signal profile.

13. The method as claimed in claim 9, wherein exposing the sensor containing hyperpolarized nuclei, comprises exposing the sensor containing hyperpolarized diamond nuclei.

14. The method as claimed in claim 11, wherein inserting the hyperpolarized nuclei into a probe structure comprises inserting the hyperpolarized nuclei into a probe structure contained inside a magnet that generates a field of at least 0.05 T.

15. The method as claimed in 12, wherein applying a radio frequency signal to the hyperpolarized nuclei comprises applying a sequence of radio frequency pulses to the hyperpolarized nuclei.