Characterization of photonic band resonators for DNP NMR of thin film samples at 7 T magnetic field
Graphical abstract
Introduction
In recent years, Dynamic Nuclear Polarization (DNP) has been experiencing a continued growth including new experimental methods and instrumentation developments [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16] motivated by diverse applications of DNP spanning across materials science, chemistry, and, importantly, structural biology fields [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28]. Several polarization transfer schemes can be implemented in DNP including: saturation of the electron spin transitions in the classical Overhauser experiment, polarization transfer by the forbidden transitions in the solid- and cross-effect DNP, and, notably, by polarization transfer in the rotating frame as in the nuclear orientation via electron spin locking (NOVEL) and its recent variants [29], [30], [31]. Regardless of the polarization transfer scheme, all these DNP experiments require manipulation of the electronic spins at magnetic fields, at which NMR experiments are performed - typically 7 T or higher. The technical implementation of such experiments has proven to be challenging because of the requirement of generating sufficiently high B1e fields over the sample at the resonance frequencies of the electron spins that fall within the mm-wave (mmW) range for these magnetic fields.
Currently, the high B1e requirements at mmW frequencies are only met by high power gyrotrons, which mainly operate in the continuous-wave (CW) mode [11], [32], [33], [34], [35]. For example, a gyrotron with ca. 20 W power output for liquid sample DNP at 260 GHz has been reported [36]. However, even at such high output powers it is still necessary to minimize the mmW transmission losses and to optimally focus the mmW beam onto a DNP sample placed inside an NMR coil [37], [38]. Another attractive way to increase and focus B1e at the sample is to employ resonator cavities – the technology that has a long development history in EPR instrumentation.
One of the earlier examples of developing specialized mmW resonators for DNP was provided by an elegant design based on a cylindrical TE011 mode cavity formed by a conductive metallic cylinder slotted in a helical pattern to act simultaneously as a solenoid radiofrequency (rf) coil [39]. A sample tube with o.d. = 0.7 mm and sample volume of only 0.3 μl was placed through the center of the cavity where mmW B1e and rf B1 fields are parallel. It was shown that such a 139.5 GHz resonator operating at 5 T magnetic field when used with a solid-state source generating just ≈4 mW of mmW power at the cavity yielded B1e≈1.25–1.5 G at the sample, thus resulting in a significantly higher 1H signal enhancement vs. a configuration without the cavity. Another example of employing a cylindrical single-mode EPR cavity for DNP was described by Liu et al. [40] who repurposed a commercial Bruker W-band (94 GHz) ENDOR probehead with matched ENDOR coils for detecting 13C NMR signal at ≈36 MHz. However, above 95–140 GHz single mode resonators and matching EPR sample tubes become too small and impractical, and Fabry-Perot resonators can be used instead. For example, Denysenkov et al. described a DNP probe based on a Fabry-Perot resonator operating at 263 GHz and integrated with a flat stripline rf coil [41]. Such a probehead was suited for DNP of liquid aqueous samples which were loaded into a thin flat holder positioned within one of the maxima of the B1e field [41], [42]. We note that a similar approach was implemented decades ago to optimize mmW EPR of aqueous samples [43], [44].
It is worthwhile to mention here that most of the resonator-based DNP probeheads described so far, which were based on EPR designs, are capable of accommodating only very small sample volumes – typically, from several tens to hundreds of nl. While such configurations are beneficial for enhancing NMR signals from limited-volume samples (e.g., small crystals, thin films, etc.), several tens to hundreds of microliters can be readily accommodated by the standard NMR probes, resulting in similar if not greater signal intensities simply due to the increased sample volume. From these considerations, in order to fully realize the advantages of DNP signal gains in NMR, one has to develop high-Q mmW resonators for sample volumes approaching those of standard NMR coils. Finally but not lastly, the resonator should be integrated with rf-efficient NMR coils tunable to at least two channels to enable modern two- and three-dimensional solid-state NMR experiments.
Recently, we have reported on a multifrequency probehead consisting of one-dimensional (1D) photonic band-gap (PBG) resonator placed inside a saddle coil tuned to two rf channels [45]. The PBG resonator was formed by a stack of ceramic discs with alternating dielectric constants (high, ε2 and low, ε1) and thicknesses corresponding to (2 m + 1)λ/4, where m = 0, 1, 2… and λ is the mmW wavelength of the dielectric material. A defect in such a 1D photonic crystal, e.g., realized as a mirror with a sample positioned at about λ/4 distance from the dielectric stack, creates a resonant structure by confining the electromagnetic waves [46]. The resonant frequency of such a structure is determined by the width and dielectric permittivity of the defect and the former can be fine-tuned by adjusting the mirror position. The electric E = 0 node, corresponding to the maximum of the magnetic field (antinode), is located at the mirror surface where the sample is placed. The practical sample volume is then determined by the product of the sample thickness (<λ/4) and the square of the disc diameter, which for DNP probes is mainly limited by the ability of integrating it with an efficient NMR coil (typically, up to 6–8 mm in diameter). Thus, as compared to cylindrical TE01n-type resonators, the DNP probeheads based on 1D PBG structures can readily accommodate samples with much larger volumes - up to several μl at 200 GHz [45].
The main advantage of employing 1D PBG resonators for DNP is in providing significant gains in the average mmW power over the sample vs. non-resonant structures. Specifically, it was previously demonstrated that for the same mmW input power the average gain in could reach ca. 12-fold or 11 dB by properly selecting the dielectric constants ε1 and ε2 of the PBG alternating layers [45]. These initial results motivated us to systematically compare various 1D PBG resonator configurations within the practical limitations imposed by the availability of different dielectric materials and their suitability for fabricating such resonators. Here we consider several combinations of the dielectric materials to further increase the effective resonator finesse and, consequently, the gains in at the sample. We also introduce a mixed-pair PBG configuration, which achieves even higher relative redistribution of towards the sample by terminating the PBG dielectric stack with a higher ε2′ plate and forming an additional λ/4 air gap adjacent to the sample. By employing an odd number of dielectric layers with a larger ratio of the dielectric constants and using a mixed-pair PBG configuration, up to 50-fold increase in the average mmW power at the sample is obtainable with the existing DNP PBG probe. These results are rationalized by theoretical considerations in an analytical form and by a numerical modeling of mmW fields for a realistic configuration of the PBG DNP probehead.
Section snippets
Instrumentation and methods
DNP Spectrometer. The 200 GHz mmW bridge assembled at NCSU consists of solid state and quasioptical mmW components. A broadband voltage-controlled W-band (90–100 GHz) solid-state synthesizer source (Tx 248, Virginia Diodes, Inc., Charlottesville, VA) is frequency-locked to within ± 100 Hz by EIP 578B counter (Phase Matrix, Inc., Santa Clara, CA). The source output is attenuated by a high-precision direct reading attenuator (Model 45726H-1000, Hughes Aircraft Company, Torrance, CA) and then
Theoretical considerations
All the resonators including EPR cavities and NMR rf coils are characterized by the Q-factor defined as:
For the PGB resonators described herein, the energy dissipation may originate from dielectric losses in all the materials including the dielectric discs themselves, the sample, resistive losses at the mirror, and additional losses due to scattering. The latter are expected to worsen with the imperfections in the dielectric layers and any misalignments
Experimental results
Based on the theoretical considerations and simulations presented above, several new 1D PBG resonators were constructed and tested. The first resonator was assembled from LiTaO3 and optical quartz, both being representative of hard crystalline materials with low impurity levels. To ascertain whether or not the resonator performance is affected by the tapered transition from the corrugated waveguide to the DNP probehead, the curves were measured for both the mock resonator (Fig. 3A) and the
Conclusions
A detailed analysis of several configurations for one dimensional PBG resonators aimed at increasing the electromagnetic energy density at the sample for DNP NMR at 300 MHz/200 GHz have been presented. The analysis included an analytical model, finite-element modeling of the electromagnetic field, and comparative 13C DNP NMR experiments performed with a thin film sample consisting of a commercial polyester-based 3 mil lapping film containing microdiamonds. It was demonstrated that while
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
Supported by NIH 5R01GM130821 to A.A.N and A.I.S.
References (52)
- et al.
Dynamic nuclear polarization at 9T using a novel 250 GHz gyrotron microwave source
J. Magn. Reson.
(2003) - et al.
Dynamic nuclear polarization at 700 MHz/460 GHz
J. Magn. Reson.
(2012) - et al.
Pushing NMR sensitivity limits using dynamic nuclear polarization with closed-loop cryogenic helium sample spinning
Chem. Sci.
(2015) - et al.
Frequency swept microwaves for hyperfine decoupling and time domain dynamic nuclear polarization
Solid State Nucl. Magn. Reson.
(2015) - et al.
Low-temperature dynamic nuclear polarization with helium-cooled samples and nitrogen-driven magic-angle spinning
J. Magn. Reson.
(2016) - et al.
Efficient cross-effect dynamic nuclear polarization without depolarization in high-resolution MAS NMR
Chem. Sci.
(2017) - et al.
A ferromagnetic shim insert for NMR magnets - Towards an integrated gyrotron for DNP-NMR spectroscopy
J. Magn. Reson.
(2017) - et al.
DNP enhanced frequency-selective TEDOR experiments in bacteriorhodopsin
J. Magn. Reson.
(2010) - et al.
Nuclear-spin orientation via electron-spin locking novel
J. Magn. Reson.
(1988) - et al.
A 250 GHz gyrotron with a 3 GHz tuning bandwidth for dynamic nuclear polarization
J. Magn. Reson.
(2012)