Optically induced polarization magnetic resonance spectroscopy at mesoscale
with a spin quantum sensor
We demonstrate an optically induced polarization detection (OIPD) technique for mesoscopic magnetic resonance spectroscopy and imaging. Our method employs a single spin in highly purified diamond as the magnetic quantum sensor, allowing us to detect the spectra of polarized electron spin magnetization of a pentacene crystal with the size of tens of micrometers. We detected the magnetic resonance spectra of polarized electron spins, measured its relaxation time and observed the electron spin polarization. This is the first application of NV-based magnetic resonance to sense polarized electron spins. Compared to thermal distribution, the polarization of these electron spins is improved a thousandfold by optical pumping. The method can be extended to sense polarized nuclear spin magnetization at mesoscale with ultrahigh polarization by employing dynamic nuclear polarization.
As one of the most important techniques, magnetic resonance spectroscopy finds broad applications in chemistry, biology, material and life sciences. Nanoscale magnetic resonance based on optical detection of electron spin resonance of nitrogen-vacancy (NV) centers in diamond have recently received broad attention in the context of quantum sensing. Magnetic resonance spectroscopy with nanoscale organic samples 1 (); 2 (); wrachtrup2017science () and single molecules 3 (); 4 () have been realized. Until now, the majority of nanoscale experiments measured a statistical fluctuation magnetization of spins which is much stronger than the mean thermal magnetization () with nano-detection volume under the ambient condition with the magnetic field of several hundred gauss. However, the fluctuation signal reduces dramatically with increased distance between the NV sensor and the sample. For the mesoscale quantum sensing, e.g. cellular-sized magnetic resonance, the thermal polarization magnetization is stronger than the fluctuations. Additionally, higher polarization can be achieved via hyperpolarization approaches such as optical-induced polarization op (), dynamic nuclear polarization (DNP) dnp (); dnp1 (); dnp2 (), and quantum-rotor-induced polarization qrip (); qrip1 (). The polarization signal can be dominant once the spin polarization is reasonably high (normally, for electron spins and for nuclear spins) even for the nanoscale sensing.
Here we report an optically induced polarization detection (OIPD) technique with a single NV quantum sensor for sensitive mesoscale quantum sensing. The polarization of the detected electron spins is improved a thousandfold and hence results in three orders of magnitude sensitivity enhancement. With this technique, we can detect the magnetic resonance spectra and measure its spin relaxation of a pentacene crystal with the size of a few tens of micrometers. The techniques demonstrated here pave the way for mesoscopic quantum sensing in chemistry, biology and material science with ambient conditions.
A schematic of a non-interacting spin system probed via the NV sensor is shown in Fig. 1(a). The signal being probed originates from magnetic dipolar interaction between the NV center and the sample spins. It can be viewed as a varying magnetic field in the vicinity of the NV center. Whereas the fluctuation signal (FS) governed by the spin-noise variance of the local field is in proportion to the square of the dipolar interaction (i.e. ), the polarization signal (PS) governed by the mean value of the varying magnetic field is linear to the dipolar interaction (i.e. ). Consequently, the calculated radius of PS detection volume, , is about an order of magnitude larger than that of FS volume, . After the integration of a sample volume larger than detection one, the PS is independent of the NV depth while the FS decreases cubically with increased NV depth. By projecting the magnetic field onto the NV symmetry axis, we obtain the signals that NV can detect as shown below,
where is the spin density of the sample, is a constant with the value for electron spins, and is the NV depth below the diamond surface. From these equitations we can compare the PS and FS detection method in Fig. 1(b).
In our experiment, we used the -oriented NV center in a diamond chip, optically detected by a confocal microscope with a 532-nm laser excitation, shown in Fig. 2(a). The NV sensor was a few microns below the diamond surface. The sample for detection was a single crystal of p-terphenyl doped with pentacene-, 0.05 mol, where the long axis of the pentacene molecule was placed to align with the -NV axis. The crystal thickness was 15 m. Another 520nm-laser with the beam intensity of W/m was applied to the crystal to optically induce the electron spin polarization. The energy diagram is shown in Fig. 2(b). Between the diamond and the crystal, there were a 150nm-Agentum layer and a 100nm-PMMA layer isolating the 532-nm and 520-nm laser beams as well as fluorescence generated from NV centers and the sample crystal. The copper wire on the top generated microwave fields to coherently control both pentacene and NV.
The OIPD pulse sequence is shown in Fig. 2(c) and comprises the following steps. (1) The NV spin state was initialized into with a 1.5- 532-nm laser pulse. (2) By introducing a pulse the NV spin was brought into the superposition state , which evolved with a phase in the following double-electron-electron resonance (DEER) pulse sequence. Synchronous with the -pulse in the DEER sequence, a 520-nm laser pulse for time was applied to generate spin polarizations of pentacene, defined as . Note that for the inhomogeneity of the field the effective polarization (in the maximum mixing state of and ) contributed to the magnetic field . (3) After phase accumulation, a second pulse converted this phase into a measurable population that was read out by the final laser pulse. The pulse sequence was typically repeated for two million times to accumulate sufficient statistics, ensuring that the electron PS of pentacene with flipping frequency was probed.
We first recorded a electron spin resonance spectrum of pentacene with an NV of depth m at the field 512 G. The OIPD sequence was repeated with a sequential scan of , in which the laser irradiation time 1.5 s, spin echo time 20 s. Fig. 3(a) presents the results obtained from applying two different phases of the second pulse. For pulse, a strong peak (red circle) at 821.6 MHz was observed while the peak is absent for pulse. Such a difference can be explained by noting that pulse leads to a signal contrast and pulse leads to a signal contrast . The latter is negligible for small .
The microwave frequency of the electron resonance peak scales linearly with the external magnetic field with a slope of MHz/G, shown in Fig. 3(a), inset. As shown in ref. 8 (), the zero-field splitting (ZFS) parameters of pentacene along the long molecular axis are MHz and MHz. Due to the strong electric dipolar coupling of the component in the ZFS (non-commute with the Zeeman interaction), it associates with strong spin mixing of the triplet states, and , at the magnetic field of G. Consequently, instead of the gyromagnetic ratio of electron, 2.8 MHz/G, the slope is calculated to be 2.55 MHz/G (Fig. 3, inset, green line) agreeing well with our measurements. It indicates that the electron signal does indeed originate from the polarised triplet states of pentacene and not from electron contamination on the surface of or within the diamond.
The dynamical property, such as the lifetimes of the pentacene manifold and , has been investigated in an adaptation to the OIPD sequence. A relaxation interval was inserted between the microwave irradiation for pentacene (MW1) and the second spin-echo pulse, as shown in the upper inset of Fig. 3(b). During this interval the population of could relax, which can be readily measured using the NV-based magnetic resonance. While the population of is not accessible to direct measurement, it can be detected by driving the population to state with a microwave pulse . As a result, the relaxation duration was inserted between the laser and microwave pulses, as shown in the lower inset of Fig. 3(b). By monitoring the amplitude of ESR peak as a function of and , a measure of -manifold lifetimes, and , was obtained and shown in Fig. 3(b). Fitting to exponential curves, we obtained s and s at G which are comparable with that in ref. 7 (). We can ignore the spin lattice relaxation of the electron spin as it is much longer than the lifetime of the manifold 10 ().
The optically induced polarization is an important property for the remote quantum sensing. To quantitatively analyze the electron polarizaiton of pentacene molecules, we compare the OIPD peaks of three NV depths with a model that includes a polarized cylindrical sample volume with the thickness = 15 m and the electron spin density nm. Experimentally, the sample was illuminated for 1.5 s by a Gaussian beam with the waist radius of 35 m. After integration of the sample volume, we can fit the data to a calculated decay and obtain the polarization , shown in Fig. 4(a) (red line). The signal contrast decays with increasing the depth of NV owing to limited sample volume, unlike the case of infinite one in Fig. 1(b). Larger polarzation can be achieved by increasing the 520nm-laser pulse duration. The PS signal is saturated with 4 s, indicating that the triplet manifold reaches equilibrium with the siglet state. A typical time s corresponding to the spin polarization of is obtained by curve fitting, as shown in Fig. 4(b). The saturated polarization at 4 s is which is comparable with the data in ref. 6 ().
These investigations were devised to develop techniques that enable the polarization signal to be probed with long distance NV sensing. Using the dipolar interactions of the NV center and a bulk magnetization, this approach has been achieved. With this long-distance sensing protocol, we have studied the dynamical properties of pentacene molecule in a single crystal indicating broad applications in chemistry, biology and material science. Boltzmann-polarised spin magnetisation can also be probed when the spin-lattice relaxation time of the sample is longer than the coherence time of NV sensor with picoliter-detection volume at high magnetic field. With the improvement of polarization via hyperpolarization method, our approach can enable the applications of mesoscopic nuclear magnetic resonance spectroscopy and imaging at ambient conditions.
The authors thank C.G. Zeng for their help on the sample preparation. This work is supported by the 973 Program (Grants No. 2013CB921800, No. 2016YFA0502400), the National Natural Science Foundation of China (Grants No. 11227901, No. 31470835, and No. 91636217), the CAS (Grants No. XDB01030400 and No. QYZDY-SSW-SLH004, No. YIPA2015370), the CEBioM, and the Fundamental Research Funds for the Central Universities (WK2340000064).
T. X, and F. S. contributed equally to this work.
- (1) H. J. Mamin, M. Kim, M. H. Sherwood, C. T. Rettner, K. Ohno, D. D. Awschalom, D. Rugar, Science 339, 557 (2013).
- (2) T. Staudacher, F. Shi, S. Pezzagna, J. Meijer, J. Du, C. A. Meriles, F. Reinhard, J. Wrachtrup, Science 339, 561 (2013).
- (3) N. Aslam, M. Pfender, P. Neumann, R. Reuter, A. Zappe, F. F. de Oliveira, A. Denisenko, H. Sumiya, S. Onoda, J. Isoya, J. Wrachtrup, Science 10, 1126 (2017).
- (4) F. Shi, Q. Zhang, P. Wang, H. Sun, J. Wang, X. Rong, M. Chen, C. Ju, F. Reinhard, H. Chen, J. Wrachtrup, J. Wang, J. Du, Science 347, 6226 (2015).
- (5) I. Lovchinsky, A. O. Sushkov, E. Urbach, N. P. de Leon, S. Choi, K. De Greve, R. Evans, R. Gertner, E. Bersin, C. Müller, L. McGuinness, F. Jelezko, R. L. Walsworth, H. Park, M. D. Lukin, Science 351, 6275 (2016).
- (6) K. Tateishi, M. Negoro, S. Nishida, A. Kagawa, Y. Morita, and M. Kitagawa, Proc. Natl. Acad. Sci. USA 111, 7527 (2014).
- (7) C. Griesinger, M. Bennati, H. M. Vieth, C. Luchinat, G. Parigi, P. Höfer, F. Engelke, S. J. Glaser, V. Denysenkov, T. F. Prisner, Prog. Nucl. Magn. Reson. Spectrosc. 64, 4 (2012).
- (8) J. H. Ardenkjaer-Larsen, B. Fridlund, A. Gram, G. Hansson, L. Hansson, M. H. Lerche, R. Servin, M. Thaning, and K. Golman., Proc. Natl. Acad. Sci. USA 100, 10158 (2003).
- (9) D. Gajan, A. Bornet, B. Vuichoud, J. Milani, R. Melzi, H. A. van Kalkeren, L. Veyre, C. Thieuleux, M. P. Conley, W. R. Grüning, M. Schwarzwälder, A. Lesage, C. Coperet, G. Bodenhausen, L. Emsley, and S. Jannin., Proc. Natl. Acad. Sci. USA 111, 14693 (2014).
- (10) B. Meier, J. N. Dumez, G. Stevanato, J. T. Hill-Cousins, S. S. Roy, P. Hakansson, S. Mamone, R. C. D. Brown, G. Pileio, M. H. Levitt., J. Am. Chem. Soc. 135, 18746 (2013).
- (11) S. S. Roy, J. N. Dumez, G. Stevanato, B. Meier, J. T. Hill-Cousins, R. C. D. Brown, G. Pileio, and M. H. Levitt., J. Magn. Reson. 250, 25 (2015).
- (12) F. G. Patterson, H. W. H. Lee, W. L. Wilson, and M. D. Fayer, Chem. Phys. 84, 51 (1984).
- (13) J. L. Ong, D. J. Sloop, T. S. Lin, Chem. Phys. Lett. 241, 540 (1995).
- (14) D. J. Sloop, H. Yu, T. Lin, S. I. Weissman, J. Chem. Phys. 75, 3746 (1981).
- (15) J. L. Ong, D. J. Sloop, T. S. Lin, J. Phys. Chem. 97, 7833 (1993).
- (16) T. Kawahara, S. Sakaguchi, K. Tateishi, T. L. Tang, T. Uesaka, J. Phys. Soc. Jpn. 84, 044005 (2015).