Method for the hyperpolarisation of nuclear spin in a diamond via a long-range interaction

Abstract

The invention concerns a method for the hyperpolarisation of .sup.13C nuclear spin in a diamond, comprising an optical pumping step, in which colour centre electron spins in the diamond are optically pumped. The method further comprises a transfer step in which the polarisation of a long-lived state of the colour centre electron spins is transferred to .sup.13C nuclear spins in the diamond via a long-range interaction.

Claims

1. Method for hyperpolarisation of .sup.13C nuclear spin in a diamond, comprising an optical pumping step, in which colour centre electron spins in the diamond are optically pumped, wherein in a transfer step, the polarisation of a long-lived state of the colour centre electron spins is transferred to .sup.13C nuclear spins in the diamond via a long-range interaction.

2. The method according to claim 1, wherein the transfer step involves an application of a microwave field or a radio frequency (RF) field.

3. The method according to claim 2, wherein the microwave field or the RF field is pulsed.

4. The method according to claim 1, wherein the colour centre electron spin polarisation is transferred to the .sup.13C nuclear spins in the diamond via a long-range interaction while fulfilling the Harman-Hahn condition.

5. The method according to claim 4, wherein the Hartmann-Hahn condition is achieved by a microwave field, the intensity of which field is chosen to match an energy difference between dressed colour centre electronic spin eigenstates and the .sup.13C nuclear spins in an external magnetic field.

6. The method according to claim 5, wherein a magnetic flux density of the external magnetic field is smaller than 3 T.

7. The method according to claim 1, wherein the transfer step is performed by an excitation of selective transitions caused by a dipolar interaction of .sup.13C nuclear spin states with the colour centre spin.

8. The method according to claim 1, wherein the transfer step is performed by interaction comprising at least two colour centre spins and a nuclear spin.

9. The method according to claim 1, wherein the colour centre is a nitrogen vacancy centre.

10. The method according to claim 1, wherein the optical pumping step and the transfer step are repeated cyclically.

11. The method according to claim 10, wherein after each cycle, a pause allows for the nuclear spin polarisations to spread throughout the diamond.

12. The method according to claim 1, wherein the method is carried out for less than 10 minutes.

13. The method according to claim 1, wherein the optical pumping step is performed with at least one laser pulse.

14. The method according to claim 1, wherein the diamond is a synthetic diamond.

15. The method according to claim 1, wherein the diamond has volume of between 50 nm.sup.3 and 1 cm.sup.3.

16. The method according to claim 1, wherein the diamond is coated with a non-diamond material.

17. A method for the nuclear spin hyperpolarisation of nuclear spins in a molecule, wherein the molecule is brought near or into contact with a diamond and prior to, during or after that, the diamond is hyperpolarised in a method according to claim 1.

18. Method for producing an imaging agent, wherein a diamond is coupled to a molecule and prior to, during or after the coupling, the diamond is hyperpolarised in the method according to claim 1.

19. Use of a hyperpolarised diamond according to claim 1 in medical or cell based imaging, in a quantum information processor or a quantum sensor based on spin degrees of freedom.

20. Diamond with a volume of above 1 nm.sup.3, in which the .sup.13C nuclear spin polarisation is greater than 0.001%.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 shows the experimental setup in two conceptual representations.

(2) FIG. 2 shows at the top an optical microscope picture of the fabricated structure on glass used in magnetic resonance experiments. At the bottom, a picture of the holder with the strip line structure is displayed.

(3) FIG. 3 shows a confocal map of single NV centres adjusted to a microwave stripline.

(4) FIG. 4 shows a photo of the magnet stage with a cylindrical magnet attached.

(5) FIG. 5 shows a graphical representation of the polarisation transfer protocol using the solid state effect.

(6) FIG. 6 shows the difference in nuclear polarisation build-up depending in the ESR line width compared to the Larmor frequency.

(7) FIG. 7 shows the process of microwave driven polarisation transfer based on the cross effect is shown.

(8) FIG. 8 displays the pulse sequence that was used to polarise nuclear spins on the left and on the right, it shows the low frequency components in the spin-locking signal on the x-axis for various microwave driving fields with the corresponding Rabi frequency on the y-axis.

DESCRIPTION OF SPECIFIC EMBODIMENTS OF THE INVENTION

(9) Experimental Setup

(10) FIG. 1a conceptually details the experimental setup with a diamond 1 placed in the magnetic field of a permanent magnet 2. The diamond 1 contains a colour centre 3, which is an NV centre. A laser 4 serves to excite a colour centre 3 electron. In order to move the diamond 1 into the focus of the laser 4, the diamond 1 is mounted on a piezo stage (not shown). The magnet 2 is mounted on rotation/translation stages 5 (not shown) to be able to align the magnetic field with the crystallographic axis of the colour centre 3. A microwave source 6 allows facilitation of polarisation transfer from electron to .sup.13C nucleus. In the conceptual representation of the experimental setup in FIG. 1b, the diamond 1 is placed between the permanent magnet 2 and the microwave source 6, which allows for the Hartmann-Hahn double resonance to be generated in the diamond 1. The optical path 8 of the laser 4 (not shown) is directed through a glass coverslip 7 and focussed into the diamond 1 in order to polarise the electrons in an NV centre.

(11) Diamond Material

(12) The following experiments were performed in a synthetic diamond layer formed by CVD doped with NV centres during growth. The sample used in these experiments possesses two layers with different properties, the substrate and a CVD grown layer. The substrate is a type IIa diamond 1 with a (111) cut and a natural abundance of .sup.13C. The CVD grown layer is also a (111) cut with a natural abundance of .sup.13C and a 1 ppm concentration of phosphorus donors. The donors were added to stabilise the charge state of the NV centre. For some of the dynamic nuclear polarisation protocols a different donor concentration is preferable, and will be mentioned in the description.

(13) For the direct polarisation of external spins via NV centres, ultra-small nanodiamonds, i.e. diamonds with volume between 1 nm.sup.3 and 1000 nm.sup.3, are preferable. Polarisation transfer will be enabled by dipolar interactions between NV centre spins and external nuclear spins. In addition, it is possible to use other electron spins as mediators for spin polarisation. Alternatively, nitrogen (P1 centre) present in 100 ppm or higher concentration in synthetic high pressure high temperature diamond can be used for this purpose.

(14) Confocal Microscopy of Single NV Centres

(15) Single NV centres were detected using a confocal microscopy technique. A laser beam diode pumped solid state laser 4 operating at 532 nm was focussed onto a diffraction limited spot using a high numerical aperture microscope objective (Olympus UPLAPO 60). The sample was scanned using a piezo driven stage (nPoint, Inc.). Fluorescence was collected by the same microscope objective and focussed on avalanche photodiodes with single photon sensitivity (SPCM-AQRH, Excelitas). By observation of photon-antibunching, it could be detected that an individual NV centre was in focus. Fluorescence detection of magnetic resonance on single electron spin is based on optical contrast of spin states associated with NV centres.

(16) Microwave Excitation

(17) In order to excite microwave transitions of single colour centres 3 in diamond 1, the sample was placed on a home built microwave strip line providing efficient excitation of the diamond 1. At the top in FIG. 2, an optical microscopic picture of the structure is shown, which was fabricated on a glass cover slip by conventional photolithography and was used in the magnetic resonance experiments. The width and gap of each microstrip is 20 m. At the bottom in FIG. 2, a picture of the holder with the strip line structure can be seen. The signal is applied via coaxial cables connected to SMA connectors and matched to the two coplanar microstrips.

(18) A commercial microwave source 6 (Anritsu MG 37020A) was used in the experiments. In order to achieve Rabi frequencies of a few MHz, the source was amplified using a commercial high power microwave amplifier (10 W, Gigatronics GT 1000A). Phase control of microwave fields was achieved using commercially available phase shifters (Narda, Inc.). Microwave pulses were formed using commercial microwave switches (General Microwave, F9914). The strength of the microwave drive was controlled by the output level of the microwave source 6.

(19) In FIG. 3, the fluorescence image of a diamond 1 sample on top of the 4-strip microstructure is shown. On the top and the bottom of the image, one strip is displayed each. Between the strips, the diamond area can be seen. Bright spots correspond to the fluorescence emissions of NV centres.

(20) Magnetic Field Control

(21) Experiments were performed in a magnetic field on the order of 0.4 T generated by a permanent magnet 2 (magnets4you GmbH) located about 100 m from the diamond face. In order to align the magnetic field with the crystallographic axis (z-axis) of the NV defect, the magnet 2 was moved using rotation and translation stages 5 (Micos GmbH), as shown in FIG. 4. For ensemble experiments aiming to polarise large samples, magnetic field need to be homogeneous enough to fulfil resonance conditions for the whole sample. Permanent magnet arrangements or electromagnets can be used for this purpose.

(22) Time Resolved Measurements

(23) Optical pulses for optical spin polarisation and time resolved detection of magnetic resonance were produced using acousto-optical modulators (Crystal Technology). Microwave, optical pulses, sample scanning and data acquisition were synchronised by a computer controlled pulse generator (Tektronix, DTG) connected to drivers of acousto-optical modulators, microwave switches and a fast photon counter (FastComtec, P7998).

(24) The optical detection of magnetic resonance was carried out in accordance with the scientific publications Jelezko, F. et al., Single defect centres in diamond: A review. Physica Status Solidi (a) Applications and Materials Science, 2006. 203(13): pages 3207 to 3225, Jelezko, F. et al., Read-out of single spins by optical spectroscopy., Journal of Physics-Condensed Matter, 2004. 16(30): pages R1089 to R1104 and Jelezko, F., et al., Observation of coherent oscillations in a single electron spin, Physical Review Letters, 2004. 92(7), the relevant portions of which are incorporated into the present disclosure by way of reference.

(25) Polarisation of Electron Spin

(26) Electron spins associated with NV centres were polarised by the application of a short (300 ns) laser 4 pulse. Optical pumping was achieved by excitation of the NV centre into an excited electronic state. The decay of this state occurs predominantly into one of the spin sublevels of the ground state.

(27) Dynamical Polarisation Transfer from Electron Spin to Nuclear Spin

(28) Exchange of polarisation between optically pumped electron spin of NV centre and nuclear spins can be performed using several established dynamic nuclear polarisation protocols, e.g. the solid effect, the cross effect, thermal mixing, the NOVEL sequence and more. Most of these protocols either involve interactions between electron spins or are based on two underlying physical mechanisms: fulfilling the Hartmann-Hahn condition and excitation of selective transitions. The DNP protocols differ in the configurations for achieving these conditions and by the usage of pulses or continuous waves.

(29) For the above DNP protocols, the experimental setup is similar, with the difference in the microwave frequency, pulse sequence and/or magnetic field strength. We used the same equipment for all three protocol examples detailed below, as all three protocols are in the regime of our equipment.

(30) The solid effect (excitation of forbidden transition involving double, electron nuclear spin flips using microwave driving) followed by electron spin relaxation is known to induce efficient polarisation transfer. Notably, the weak electron spin relaxation process can be significantly enhanced by optical pumping of NV centre.

(31) A rigorous theoretical treatment of the solid effect has been performed in numerous papers, e.g. Abragam A, Goldman M. Rep Prog Phys 1978; 41:395, W. T. Wenckebach Applied Magnetic Resonance 2008, 34, 227-235. A graphical representation of the polarisation transfer protocol using the solid state effect is shown in FIG. 5.

(32) At the first stage the laser polarises the NV centre by optical pumping, as described above. Next, via the forbidden transition a microwave excitation excites simultaneously the NV spin and the nuclear spin which results in nuclear polarisation. The NV spin is then re-polarised via optical pumping.

(33) The rate of polarisation transfer is maximal for microwave frequencies corresponding to the energy levels of the forbidden transitions .sub.+.sub.NV.sub.I for positive nuclear polarisation and .sub..sub.NV+.sub.I for negative nuclear polarisation, where .sub.0S denotes the NV spin Rabi frequency and .sub.I the nuclear spin Larmor frequency in the lab frame. As the polarisation rate is a function of the NV centre spin ESR line shape, effective polarisation transfer is achieved when the ESR line is narrow compared with the nuclear spins Larmor frequency (or the longitudinal hyperfine component of the interaction with the NV centre spin). The difference in nuclear polarisation build-up depending in the ESR line width compared to the Larmor frequency is depicted in FIG. 6. FIG. 6(a) depicts the nuclear polarisation as a function of the microwave frequency for the case where the nuclear Larmor frequency (.sub.I) is larger than the NV centre ESR line. This case is known as the well resolved solid effect. FIG. 6(b) depicts the nuclear polarisation for the case where the ESR line is not narrow compared to the nuclear Larmor frequency, which is known as the differential solid effect. In this case, the effects for positive nuclear polarisation and negative nuclear polarisationdepicted in dashed lines in FIG. 6(b)overlap, reducing the overall polarisation reached (solid line).

(34) Larmor frequencies of .sup.13C nuclear spins were approximately 5 MHz for magnetic fields used in our experiments, though stronger magnetic fields can be used for larger Larmor frequencies. For narrow NV centre ESR lines, diamonds with a small concentration of P1 (Nitrogen) donors (less than 10 ppm) are preferable. For instance, CVD grown diamonds with 10 ppm P1 donor will result in NV centre ESR line width which is only limited by .sup.13C, thus enabling efficient polarisation. Polarisation transfer is then enabled by continuous laser 4 optical pumping combined by resonant microwave 6 irradiation.

(35) An alternative method for transferring the NV centre spin polarisation to the nuclear spins is the so-called cross polarisation effect, involving two electron spins and one nuclear spin. This effect is particularly interesting for samples having high concentration of NV centres with strongly dipolar coupled electron spins. The basis for the cross effect are two dipolar coupled electron spins under the condition that the resonance frequency the electrons is separated by the nuclear Larmor frequency. Thus, the cross-effect can only occur if the inhomogeneously broadened ESR lineshape has a linewidth broader than the nuclear Larmor frequency, contrary to the condition for effective polarisation via the solid effect. An additional condition for the cross effect is that the homogeneously broadened ESR linewidth is narrower than the nuclear Larmor frequency.

(36) The cross effect was first discovered in the 1960s by Kessenikh et al. In Kessenikh et al. Phys Solid State 1963; 5:321, and later by Wollan DS. Phys Rev B 1976; 13:3671. In the last few years, it has again aroused interest after experiments which have shown a large DNP enhancement to the NMR signal in high magnetic fields (e.g. Hall et al. Science 1997; 276:930, . Song et al. J Am Chem Soc 2006; 128:11385).

(37) The cross effect is based on a three spin interaction (two electron spins and a nuclear spin) satisfying the relation:
.sub.s2.sub.s1=.sub.I, (1)

(38) with .sub.s1(2) denoting the EPR frequency of electron 1(2) and .sub.I denoting again the Larmor frequency of the nuclear spin.

(39) For driving the polarisation transfer, a microwave irradiation is added of frequency .sub.s1(2), leading to a negative(positive) nuclear polarisation. The polarisation process is depicted in FIG. 7. FIG. 7(a) depicts the population distribution at thermal equilibrium for a general three spin system (two electron spins and a nuclear spin) in an external magnetic field. In FIG. 7(b and c), the energy level have been set such that condition 1 is met. FIG. 7(b) depicts the new population distribution under microwave irradiation of frequency .sub.s1, which leads to a saturation of the allowed EPR transitions. As can be seen, this corresponds with negative nuclear polarisation. Microwave irradiation of frequency .sub.s2 leads to positive nuclear polarisation, see FIG. 7(c).

(40) For a typical diamond with 100 ppm P1 donors, the homogeneus broadening could be 100 KHz, and the inhomogeneus broadening is typically in the MHz range, but can be made larger by growing the diamond with intrinsic strain along some axis, or by increasing the .sup.13C concentration in the diamond. Additionally, one could imagine using the P1 donors' electron spin as a pair for the dipolar coupling in the cross effect with the NV centre spins.

(41) Another proposed experimental realization of a DNP protocol for the polarisation transfer is achieved by establishing a Hartmann-Hahn condition between the electron and nuclear spin. This is achieved by driving the electron spin transitions between ms=0 and ms=1 state by means of a microwave field whose intensity is chosen to match the energy difference between dressed electronic spin eigenstates and the nuclear spins in an external magnetic field.

(42) The dynamics of the NV electronic spin and an additional nuclear spin, in the presence of a continuous driving microwave field have been theoretically analysed in Cai, J.M. et al., Diamond based single molecule magnetic resonance spectroscopy, New Journal of Physics, 2013, 15, 013020, http://arxiv.org/abs/arXiv: 1112.5502 and the article's supplementary information; the relevant portions of the publication and the supplementary information are incorporated into the present disclosure by way of reference. The Hamiltonian describing the NV centre electronic m.sub.s=0, 1 states and an additional .sup.13C nuclear spin, in the presence of an external magnetic field B and a resonant microwave field is
H=.sub.z1+.sub.Nl|B.sub.eff|.sub.z+.sub.NA.sub.hyp.sub.x(sin .sub.x+cos .sub.z) (1)

(43) where is the Rabi frequency of the driving field and are the spin-1/2 operators, defined in the microwave-dressed basis

(44) $.Math.)=\frac{1}{\sqrt{2}}\left(.Math.0).Math.-1)\right)$
for the electronic basis, and in the (|z.sup.r, |z.sup.r basis for the nuclear spins, where z is defined along the direction of B.sub.eff. B.sub.eff is an effective magnetic field and is given by B.sub.eff=B(1/2)A.sub.hyp, where A.sub.hyp is the hyperfine vector which characterises the coupling between the two spins. In equation (1), .sub.N is the gyromagnetic ratio of the nuclear spin and cos =.Math.{circumflex over (b)}, where and {circumflex over (b)} are the directions of the hyperfine vector A.sub.hyp and the effective magnetic field B.sub.eff, respectively. The first two terms in the Hamiltonian form the energy ladder of the system ( for the dressed NV spin, and .sub.N|B.sub.eff| for the Larmor frequency of the nuclear spin), whereas the last two terms are responsible for electron-nuclear spin interaction. Here, the former represents mutual spin-flips, or coherent evolution of the electron-nuclear pair, and the latter is the nuclear spin dephasing caused by electron flips. When the driving field is adjusted properly, an energy matching condition (known as the Hartmann-Hahn condition) given by
=.sub.N|B.sub.eff|=.sub.N|B(1/2)A.sub.hyp|, (2)

(45) can be engineered, equalising the first two terms in Hamiltonian (1). Then, the coupling term in the Hamiltonian becomes dominant, and the time evolution of the system is a coherent joint evolution of the electron nuclear pair. For instance, starting in the |+, state, the system evolves according to |=|+, cos(Jt) .sup.+|, sin (Jt), with J given by
J=1/4.sub.N|A.sub.hyp|sin . (3)

(46) Thus, at time t=/2J the two spins become maximally entangled, and after a t=/J a full population transfer occurs and the states of the two spins are in effect swapped.

(47) Larmor frequencies of .sup.13C nuclear spins were approximately 5 MHz for magnetic fields used in our experiments. In order to transfer the electron spin to the nuclei, we applied a sequence, in which a short laser 4 pulse (300 ns) is used for the polarisation of the electron spin in the ground state of the NV centre and for readout of the population via spin-dependent fluorescence. The microwave manipulation is the alternating spin locking sequence for 8 s as shown in FIG. 8 on the left. A sweep of the source power through the Hartmann-Hahn double resonance while counting all the photons yielded the trace shown in FIG. 8 on the right. The low frequency components in the spin-locking signal for various microwave driving fields is shown on the x-axis, the corresponding Rabi frequency is shown on the y-axis. The oscillations appearing in the spectrum at the Hartmann-Hahn condition (when the Rabi frequency of electron spin matches the nuclear spin Larmor frequency) indicate flip-flops between electron spins and nuclear spins.

REFERENCE NUMBER LIST

(48) 1 diamond

(49) 2 magnet

(50) 3 colour centre

(51) 4 laser

(52) 5 magnet stage

(53) 6 microwave source

(54) 7 glass coverslip

(55) 8 optical path