Journal of Magnetic Resonance 238 (2014) 20–25

Contents lists available at ScienceDirect

Journal of Magnetic Resonance journal homepage: www.elsevier.com/locate/jmr

Communication

Magic-angle spinning solid-state multinuclear NMR on low-field instrumentation Morten K. Sørensen a, Oleg Bakharev a, Ole Jensen b, Hans J. Jakobsen c, Jørgen Skibsted c, Niels Chr. Nielsen a,⇑ a Center for Insoluble Protein Structures (inSPIN), Interdisciplinary Nanoscience Center (iNANO) and Department of Chemistry, Aarhus University, Gustav Wieds Vej 14, DK-8000 Aarhus C, Denmark b Nanonord A/S, Skjernvej 4A, DK-9220 Aalborg Ø, Denmark c Instrument Centre for Solid-State NMR Spectroscopy, Interdisciplinary Nanoscience Center (iNANO) and Department of Chemistry, Aarhus University, Langelandsgade 140, DK-8000 Aarhus C, Denmark

a r t i c l e

i n f o

Article history: Received 25 August 2013 Revised 18 October 2013 Available online 1 November 2013 Keywords: Low-field NMR Magic-angle spinning 23 Na second-order quadrupolar coupling 31 P–19F dipole–dipole coupling Sodium nitrate Potassium monofluorophosphate

a b s t r a c t Mobile and cost-effective NMR spectroscopy exploiting low-field permanent magnets is a field of tremendous development with obvious applications for arrayed large scale analysis, field work, and industrial screening. So far such demonstrations have concentrated on relaxation measurements and lately highresolution liquid-state NMR applications. With high-resolution solid-state NMR spectroscopy being increasingly important in a broad variety of applications, we here introduce low-field magic-angle spinning (MAS) solid-state multinuclear NMR based on a commercial ACT 0.45 T 62 mm bore Halbach magnet along with a homebuilt FPGA digital NMR console, amplifiers, and a modified standard 45 mm wide MAS probe for 7 mm rotors. To illustrate the performance of the instrument and address cases where the low magnetic field may offer complementarity to high-field NMR experiments, we demonstrate applications for 23Na MAS NMR with enhanced second-order quadrupolar coupling effects and 31P MAS NMR where reduced influence from chemical shift anisotropy at low field may facilitate determination of heteronuclear dipole–dipole couplings. Ó 2013 Elsevier Inc. All rights reserved.

1. Introduction Along with maintained strong focus on high-field NMR spectroscopy, the past few years have witnessed a considerable interest in the development of low-field NMR instrumentation for low-cost, mobile, and screening NMR applications [1,2]. A variety of lowfield NMR instruments have recently been developed for relaxation measurements [3–7], magnetic resonance imaging [8–11], and lately also high-resolution liquid-state NMR spectroscopy [1,11–14]. The instrumentation involves many different sources to the fundamental magnetic field, including earth field devices [15–17], low-field permanent magnets in single-sided setups [4,12,18–20], arrays of cubic magnets [21,22], and cylindrical Halbach [11,14,23–25] geometries. In combination with advanced digital FPGA (Field Programmable Gate Array) based rf consoles [26–29], such instrumentation has found applications in diverse areas such as geology, mining, and cement chemistry [5,30–34], food science and control [35–38], chemical synthesis [39,40], as well as clinical-oriented disease screening [7,41–43]. The obvious advantage of low-field NMR is instrumentation cost, mobility and ⇑ Corresponding author. Fax: +45 8 6196199. E-mail address: [email protected] (N.Chr. Nielsen). 1090-7807/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jmr.2013.10.015

robustness, while spectral resolution/dispersion and sensitivity are remaining challenges. The sensitivity challenge has recently been addressed through combination with dynamic nuclear polarization [44] and parahydrogen [45] hyperpolarization technology. In this paper, we demonstrate that a simple combination of a commercial low-field Halbach magnet with a homebuilt FPGA digital rf console, amplifiers, and a standard MAS probe offers the possibility to perform low-field high-resolution solid-state multinuclear MAS NMR spectroscopy. Such instrumentation may have interesting industrial applications, as well as scientific applications where different scaling of field-dependent nuclear spin interactions may offer complementary insight to that obtained using standard instrumentation at higher magnetic field strengths. The latter applies to chemical shift (isotropic and anisotropic) interactions, which scale linearly with the magnetic field, and to second-order quadrupolar couplings which scale inversely to the field, while dipole–dipole couplings and quadrupolar coupling interactions to first order remain invariant with respect to the external field. Such aspects are here illustrated by determination of the 23Na quadrupolar coupling parameters for NaNO3 from the second-order MAS NMR powder pattern for the central transition (CT) and by determination of the 31P–19F dipole–dipole coupling

Communication / Journal of Magnetic Resonance 238 (2014) 20–25

for K2PO3F using 31P MAS NMR, taking advantage of the reduced influence from the 31P chemical shift anisotropy (CSA) at low field.

2. Methods The low-field solid-state NMR spectrometer (illustrated in Fig. 1a) was assembled using the components listed in (i)–(vi) below. (i) A commercial 0.45 T (19 MHz 1H frequency) tomograph Halbach magnet (Fig. 1a) was purchased from ACT (Aachen, Germany). The magnet material SmCo has a temperature coefficient of 300 ppm/K (see also Fig. 1f) and is arranged in a cylindrical Halbach type design with a horizontal bore of 62 mm, and the direction of the magnetic field perpendicular to the bore axis. In the present applications the magnet was operated without additional shims beyond the native configuration of the magnet. In this configuration, the MAS probe with its 7 mm o.d./5 mm i.d. rotor (sample volume of 220 ll) produces a line width (Full Width at Half Maximum) of FWHM = 1 ppm for a single-scan 1H spectrum of H2O (i.e., FWHM = 19 Hz) in a rotating sample. (ii) A Nanonord (Aalborg, Denmark) broadband, single channel, digital rf console with FPGA and Digital Signal Processing (DSP) units on a 10  16 cm2 board, which operates at a sampling frequency of 25 MHz. (iii) A high-power amplifier (400 W, 10  22  4.5 cm3) and a preamplifier have both been designed and built in-house and the latter implemented on the console board. (iv) Crossed diodes and quarter lambda stubs for simple passive transmit/receive switching are also homebuilt. (v) A PC with interface for data processing and user-friendly control of the spectrometer has been programmed in MATLABÒ for these purposes. (vi) An old prototype 7 mm Varian 400 MHz narrow-bore (45 mm) MAS probe, which was modified years ago in our lab with an improved homebuilt 7 mm ceramic MAS stator and a double-tuned rf coil for high-field CP/MAS NMR, has been brought into use again. To achieve the lowfield operation of 19 MHz for 1H NMR and additional lower-field X-nuclei operations, the probe has now been further modified by introducing a new double-tuned rf coil wounded on a PEEK (PolyEther Ether Ketone) support with a 0.5 mm thread pitch and 20 turns of 0.25 mm copper wire (Fig. 1b). With this setup quality factors of 39 and 52 are obtained for 31P (7.7 MHz) and 1H (19 MHz) on the low- and high-field rf channel, respectively. The probe is tunable for other frequencies, e.g. 79Br (4.77 MHz), by inserting chip capacitors. Because the direction of the magnetic field is orthogonal to the bore direction, the standard magic-angle adjustment is supplemented with an additional device for angle adjustment, which controls the rotation of the probe (see Fig. 1c and d). Rotation of the probe is important because the range of the original angle-setting mechanism is insufficient for the probe and its rotor sample to be aligned in a plane, which includes the directions of both the magnet bore and the magnetic field axes. Thus, adjustment of the exact magic-angle is performed by rotation of the original MAS-angle setting for the probe by 45°. The MAS probe employed 7 mm PSZ (Partially Stabilized Zirconia) rotors, as mentioned above with a wall thickness of 1 mm for the present applications, and spinning frequencies (mr) within the range mr = 4.3–4.8 kHz (stable within ±20 Hz without rotor control) using air pressures of 2.6 and 1.25–1.5 bar for the air-bearing and -drive, respectively. The magic angle has been set employing 79Br MAS NMR for a sample of KBr [46], primarily by a 45° rotational adjustment of the probe, followed by minor touches of the ordinary, homebuilt precise MASangle adjustment device in the probe. This procedure results in a FWHM = 157 Hz (incl. 20 Hz of Lorentzian line broadening) for the centerband in the 79Br spectrum shown in Fig. 1e and f. Considering the ‘‘additional non-shimmed’’ B0 inhomogeneity of the ACT Halbach permanent magnet, this value agrees well with the typical

21

value of  140 Hz observed at higher fields of 7–14 T in our laboratories. To account for the temperature dependence of the field drift for the permanent magnet, scans are averaged with correction for the drift. An algorithm that performs two separated post-processing schemes of the same data is applied to the averaged data, collected in blocks of 1–10 min, where the time period depends on the stability of the local temperature and the intrinsic line width of the resonances. First, a drift curve, representing the drift for the time of each block, is obtained through peak positions of the Fourier transform of an average of several blocks around the block and with a large line broadening (100–500 Hz for the cases presented here) to achieve a sufficient SNR. The obtained drift curve from this moving average is then smoothened. Second, the original blocks are Fourier transformed with low (or no) line broadening and summed employing the shifts from the drift curve, whereby the final spectrum is achieved. To account for a severe probe ringing with a ring-down time (Tringdown) of about 100 ls (as determined from an experimental FID without a sample according to the expression Aexp( t/Tringdown)) caused by the low resonance frequencies, we applied either: (i) a rotor-synchronized Hahn echo [47] experiment with suppression of ringing from both rf pulses by a 4-step phase cycle (phase of first pulse: x, x, x, x, second pulse: x, x, y, y, and receiver: x, x, x, x) or (ii) a single-pulse experiment with subtraction of the first 960 ls of the FID for a data set obtained without a sample in order to eliminate effects of the ring-down in the spectral FIDs. We note that there is still visible ringing in the FID (without a sample) after about 500 ls, which is comparable in size to the low-intensity NMR signal. All simulations were performed using the open-source simulation program SIMPSON [48,49], where also the definition of the internal interaction parameters (CQ, gQ, daniso, gCS, b/2p) used in this work can be found. We note, using this approach, that the internal nuclear spin interactions are formulated in the high-field approximation which is valid here considering the field and the spin systems studied. For the 31P simulations we used powder averaging of 30–168 sets of REPULSION [50] a,b crystal angles and 8–16 c angles, whereas 232 ZCW [48] angle pairs and up to 128 c angles were used for the 23Na simulations.

3. Results and discussion Fig. 2a shows the low-field 5.03 MHz 23Na MAS spectrum for mr = 4.5 kHz of the CT for a powder sample of NaNO3 which exhibits a clearly visible second-order quadrupolar line shape. Fig. 2b illustrates a simulation of the 23Na second-order line shape for the CT using the earlier reported parameters, CQ = 337 kHz, gQ = 0.0, determined from a satellite transition (ST) spectrum [51]. An excellent agreement between the experimental and simulated spectrum is observed. Fig. 2c shows an expansion for the CT, as obtained from a simulation of the complete 23Na MAS spectrum (CT and STs) at 105.8 MHz (9.4 T), and it is noted that the second-order line shape is not resolved at high field strength. The inset in Fig. 2d shows a simulation of the complete 23Na MAS spectrum (with a cut-off of the CT). It is evident that low-field applications are particularly interesting in cases where the line shapes or spinning sideband (ssb) patterns are influenced by other interactions such as chemical shift anisotropy (CSA) or as a supplement to high-field measurements through exploitation of the field-dependent scaling of different interactions. As an example of the reduced influence on the CSA interaction at low field, Fig. 3 shows experimental and simulated low/highfield 31P MAS NMR spectra of Brushite (calcium hydrogen phosphate dihydrate, CaHPO4
2H2O) considering the CSA parameters

22

Communication / Journal of Magnetic Resonance 238 (2014) 20–25

(a)

(b)

(c)

(d)

(f)

Drift (Hz)

400

200

0

(e)

0

5

10

15

Time (h)

30

10

-10

-30

kHz

Fig. 1. (a) Photo of the low-field NMR spectrometer, which includes a 0.45 T Halbach permanent magnet along with a modified MAS probe to the left. The rf circuit with the digital console and power amplifier as well as the controlling PC are seen on the table as well. (b) The homebuilt MAS stator includes a 7.7 mm i.d. double-tuned rf coil wound on a PEEK support (see text). Part (c) shows the precision-screw arrangement for supplementary adjustment of the MAS angle by rotation of the probe (see text). (d) Schematic drawing of the angle-adjustment system. The magic angle (MAnew) of 54.7° between the rotor axis and the magnetic-field direction is achieved by a probe rotation of u  45° relative to the rotor axis being positioned almost exactly in the vertical plane defined by the directions of the magnet-bore and -field axes, while the original MAS angle (MA0) adjustment of the probe in a standard high-field magnet (54.7°) has been preserved. (e) 79Br MAS spectrum of KBr at the magic angle, obtained using a Hahn echo for mr = 4.5 kHz, 250 blocks of 1024 scans (256,000 scans total, 14.2 h), 200 ms repetition delay, and pulse widths of 5.5 ls and 11 ls for the 90° and 180° pulses, respectively. (f) Graph illustrating the drift (in Hz) for the overnight 79Br experiment of the centerband in KBr. This shows that after an initial negative drift of about 100 Hz, the resonance drifts over an interval of about 540 Hz (i.e. 113 ppm) in about 13 h, which using the temperature coefficient of 300 ppm/K for the magnet material (see text) corresponds to a decrease in temperature of 0.4 °C in 13 h. Note: A positive drift corresponds to a decrease in temperature.

23

Communication / Journal of Magnetic Resonance 238 (2014) 20–25

(a)

(a)

(b) (b)

(c) (c)

(d)

200

0.5

0

-0.5

100

0

-1.0

-100

-1.5

-200

kHz

kHz

Fig. 2. 23Na MAS spectra of a powder sample of NaNO3 for mr = 4.5 kHz. (a) Experimental spectrum of the CT at 0.45 T (5.03 MHz) resulting from single-pulse excitation (pulse width 1 ls, 33 kHz rf field strength) and averaged over 65,536 scans with a recycle delay of 200 ms. Temperature drift compensation was accomplished for blocks of 256 scans based on a moving average of 5376 scans. (b) Corresponding simulation of the second-order line shape for the CT using CQ = 337 kHz, gQ = 0 [51], and 70 Hz line broadening. (c) Corresponding simulation under high-field (9.4 T, 105.8 MHz) condition. (d) Simulation at high-field (9.4 T, 105.8 MHz) of the complete spectrum (CT and STs) employing the above parameters.

previously reported [52]. Whereas the high-field spectrum in Fig. 3d is dominated heavily by the CSA pattern, only low-intensity first-order spinning sidebands are observed in the experimental low-field spectrum in Fig. 3a. These spinning sidebands may originate from a 31P–1H dipole–dipole coupling as justified by the simulated spectrum in Fig. 3c which considers the 31P CSA interaction and a 31P–1H dipole–dipole coupling constant of bPH/ 2p = 4.6 kHz, corresponding to a 31P–1H distance of 2.2 Å obtained from the crystal structure [53]. For comparison, Fig. 3b shows a simulation where the 31P–1H dipole–dipole coupling is not taken into account, thereby illustrating that the small spinning sidebands do not arise from the 31P CSA. However, most likely magnetic field inhomogeneity also contributes to (and may even dominate) the intensity of the first-order spinning sidebands observed in the experimental spectrum. Such first-order spinning

(d)

10

5

0

-5

-10

kHz

Fig. 3. Experimental and simulated low/high-field 31P MAS spectra of CaHPO4
H2O (Brushite) for mr = 4.5 kHz. (a) Experimental spectrum at 0.45 T (7.7 MHz) obtained using single-pulse excitation (8 ls pulse length, 31 kHz rf field strength), a 20 s repetition delay, 9663 blocks of a single scan, and apodization using 100 Hz line broadening. (b–d) Simulated spectra using anisotropic chemical shift parameters of daniso = 65.1 ppm, gCS = 0.78 under conditions of external fields of (b and c) 0.45 T and (d) 9.4 T. Parts (c and d) include a dipole–dipole coupling of bPH/2p = 4.6 kHz in the simulations.

sidebands, caused by magnetic field inhomogeneity, are generally observed in high-field multinuclear MAS NMR studies of various nuclei (e.g., 13C, 27Al, 29Si, 31P, etc.), where the different line broadening interactions are absent or have been eliminated experimentally. It is noted, that with the availability of a two-channel console, based on two replicas of the prototype homebuilt single-channel console used in this preliminary low-field solid-state multinuclear MAS NMR study, 31P–{1H}decoupling experiments would allow determination of the contribution that the 1H–31P dipolar coupling makes to the intensity of the 1st order sidebands in Fig. 3a. The low- and high-field 31P MAS NMR spectra of potassium monofluorophosphate (K2PO3F) illustrated in Fig. 4 show that while the high-field (7.05 T) spectra in Fig. 4c and d are likewise heavily dominated by the CSA contribution, the 31P–19F dipole–dipole coupling (bPF/2p) combined with the isotropic one-bond 31 P–19F J coupling constant (1JPF) dominate the low-field spectrum in Fig. 4a. From the high-field 31P MAS NMR spectrum, recorded

24

Communication / Journal of Magnetic Resonance 238 (2014) 20–25

(a)

(b)

10

5

0

-5

-10

kHz

(c)

with 19F decoupling (Fig. 4d), the following CSA parameters have been extracted: daniso = 88.8 ± 2.2 ppm, gCS = 0.02 ± 0.14, and diso = 6.3 ± 0.2 ppm. Using these CSA parameters, along with an assumed relative orientation for the CSA and 31P–19F dipole–dipole coupling tensors set to a = b = c = 0, a fit to the low-field experimental spectrum in Fig. 4a gives bPF/2p = 9.1 ± 1.2 kHz and 1JPF = 900 ± 50 Hz. As shown elsewhere [54], the line shapes of the doublet splittings observed in the experimental spectrum (Fig. 4a) depend on the relative sign for (bPF/2p)apparent and 1JPF being the same or opposite. The negative sign reported here for 1 JPF follows from two simulated spectra of which the spectrum in Fig. 4b corresponds to similar sign for the two couplings. The optimized fit to the experimental spectrum in Fig. 4a is shown in Fig. 4b. The 31P CSA and 1JPF parameters are in good agreement with those reported by Grimmer et al. [55] (1JPF = 850 Hz, diso = 5.7 ppm, daniso = 104 ppm, and assuming gCS = 0) from analysis of a static 31P NMR spectrum acquired at 6.35 T. The determined value bPF/2p = 9.1 ± 1.2 kHz, using the above assumptions, corresponds to a 31P–19F distance of 1.71 Å (1.64–1.80 Å) if an isolated spin–pair system is assumed. This value is fairly similar to the distance of 1.61 Å determined from the crystal structure study reported by Payen et al. [56]. We note that the obtained 31P–19F distance of 1.71 Å (1.64–1.80 Å), determined from the dipole– dipole coupling bPF/2p = 9.1 ± 1.2 kHz, can be larger/smaller than the actual distance, because of a possible contribution from the anisotropic J coupling (Janiso) i.e., (bPF/2p)apparent = bPF/2p + Janiso. However, generally the value for Janiso is smaller than the isotropic part 1JPF. Thus, a possible additional contribution of Janiso to bPF/2p has been neglected here, because Janiso is within the error limits of the value determined for bPF/2p. In this Communication, we have demonstrated the applicability of low-field, low-cost MAS solid-state NMR spectroscopy as a complement to standard high-field MAS NMR. The focus has been exploitation of different spectral appearances arising from nuclear spin interactions with different field dependency. We anticipate that low-field MAS solid-state NMR has potential for being widely applied in on-site characterization within industry and science. Of particular importance are studies of compounds with large CSA’s such as, for example, vanadium compounds [57,58], phosphates (as studied here), heavy-metal compounds of e.g. platinum, mercury or lead for which anisotropic chemical shifts may exceed ten thousands of ppm [59,60]. Acknowledgments We thank Erik E. Pedersen and Eigil G. Hald for the modifications made to the old prototype Varian probe and construction of the angle-adjustment device, and Nicholas Balsgart for initial high-field results on K2PO3F. The project was financially supported by the Lauritzen Foundation and the Danish National Research Foundation (DNRF59).

(d)

References 20

15

10

5

0

-5

-10

-15 kHz

Fig. 4. Low- and high-field 31P MAS NMR spectra of K2PO3F for mr = 4.3 kHz. (a) Experimental low-field NMR spectrum at 0.45 T (7.7 MHz) obtained using a Hahn echo (echo delay s2 = 234 ls), 7 ls pulse widths, 36 and 71 kHz rf field strength (for the 90° and 180° pulses, respectively), 385 blocks of 128 scans (49,280 scans in total), 5 s repetition delay, and 20 Hz line broadening. (b) Simulation (optimized fit) of the low-field spectrum in (a) for the parameters shown in the text and as described therein. (c and d) High-field (7.05 T, 121.4 MHz) 31P MAS NMR spectra acquired on a Varian Unity INOVA-300 spectrometer using single-pulse excitation (1.5 ls pulse width and an rf field strength of cB1/2p = 50 kHz) and a 30 s repetition delay. Part (c) is obtained without 19F decoupling using a 7 mm CP/MAS probe and the same sample as for the low-field experiment (352 scans) while part (d) is acquired with 19F decoupling using a 5 mm CP/MAS probe, the WALTZ decoupling scheme and a 19F rf field strength of 105 kHz (64 scans).

[1] B. Luy, Towards portable high-resolution NMR spectroscopy, Angew. Chem., Int. Ed. 50 (2011) 354–356. [2] B. Blumich, F. Casanova, S. Appelt, NMR at low magnetic fields, Chem. Phys. Lett. 477 (2009) 231–240. [3] D.D. Griffin, R.L. Kleinberg, M. Fukuhara, Low-frequency NMR spectrometer, Meas. Sci. Technol. 4 (1993) 968–975. [4] G. Eidmann, R. Savelsberg, P. Blumler, B. Blumich, The NMR MOUSE, a mobile universal surface explorer, J. Magn. Reson. 122A (1996) 104–109. [5] G.R. Coates, L. Xiao, M.G. Prammer, NMR Logging – Principles and Applications, Halliburton Energy Services, 1999. [6] N. Sun, T.J. Yoon, H. Lee, W. Andress, R. Weissleder, D. Ham, Palm NMR and 1chip NMR, IEEE J. Solid-State Circ. 46 (2011) 342–352. [7] W.K. Peng, L. Chen, J. Han, Development of miniaturized, portable magnetic resonance relaxometry system for point-of-care medical diagnosis, Rev. Sci. Instrum. 83 (2012).

Communication / Journal of Magnetic Resonance 238 (2014) 20–25 [8] A. Constantinesco, P. Choquet, G. Cauffet, J.M. Fournier, S. Ravier, J.M. Drillon, G. Aubert, Low-field dedicated and desktop magnetic resonance imaging systems for agricultural and food applications, Magn. Reson. Chem. 35 (1997) S69–S75. [9] M. Rokitta, E. Rommel, U. Zimmermann, A. Haase, Portable nuclear magnetic resonance imaging system, Rev. Sci. Instrum. 71 (2000) 4257–4262. [10] S.M. Wright, D.G. Brown, J.R. Porter, D.C. Spence, E. Esparza, D.C. Cole, F.R. Huson, A desktop magnetic resonance imaging system, Magn. Reson. Mater. Phys., Biol. Med. 13 (2002) 177–185. [11] E. Danieli, J. Mauler, J. Perlo, B. Blumich, F. Casanova, Mobile sensor for high resolution NMR spectroscopy and imaging, J. Magn. Reson. 198 (2009) 80–87. [12] J. Perlo, V. Demas, F. Casanova, C.A. Meriles, J. Reimer, A. Pines, B. Blumich, High-resolution NMR spectroscopy with a portable single-sided sensor, Science 308 (2005) 1279. [13] A. McDowell, E. Fukushima, Ultracompact NMR: 1H spectroscopy in a subkilogram magnet, Appl. Magn. Reson. 35 (2008) 185–195. [14] E. Danieli, J. Perlo, B. Blumich, F. Casanova, Small magnets for portable NMR spectrometers, Angew. Chem., Int. Ed. 49 (2010) 4133–4135. [15] P.T. Callaghan, M. Legros, Nuclear spins in the Earth’s magnetic-field, Am. J. Phys. 50 (1982) 709–713. [16] J.N. Robinson, A. Coy, R. Dykstra, C.D. Eccles, M.W. Hunter, P.T. Callaghan, Twodimensional NMR spectroscopy in Earth’s magnetic field, J. Magn. Reson. 182 (2006) 343–347. [17] C.A. Michal, A low-cost spectrometer for NMR measurements in the Earth’s magnetic field, Meas. Sci. Technol. 21 (2010). [18] B. Blumich, P. Blumler, G. Eidmann, A. Guthausen, R. Haken, U. Schmitz, K. Saito, G. Zimmer, The NMR-mouse: construction, excitation, and applications, Magn. Reson. Imaging 16 (1998) 479–484. [19] B. Manz, A. Coy, R. Dykstra, C.D. Eccles, M.W. Hunter, B.J. Parkinson, P.T. Callaghan, A mobile one-sided NMR sensor with a homogeneous magnetic field: The NMR-MOLE, J. Magn. Reson. 183 (2006) 25–31. [20] B. Blumich, J. Perlo, F. Casanova, Mobile single-sided NMR, Prog. Nucl. Magn. Reson. Spectrosc. 52 (2008) 197–269. [21] C. Hugon, F. D’Amico, G. Aubert, D. Sakellariou, Design of arbitrarily homogeneous permanent magnet systems for NMR and MRI: theory and experimental developments of a simple portable magnet, J. Magn. Reson. 205 (2010) 75–85. [22] C. Hugon, P.M. Aguiar, G. Aubert, D. Sakellariou, Design, fabrication and evaluation of a low-cost homogeneous portable permanent magnet for NMR and MRI, C. R. Chim. 13 (2010) 388–393. [23] B.P. Hills, K.M. Wright, D.G. Gillies, A low-field, low-cost Halbach magnet array for open-access NMR, J. Magn. Reson. 175 (2005) 336–339. [24] R.C. Jachmann, D.R. Trease, L.S. Bouchard, D. Sakellariou, R.W. Martin, R.D. Schlueter, T.F. Budinger, A. Pines, Multipole shimming of permanent magnets using harmonic corrector rings, Rev. Sci. Instrum. 78 (2007). [25] H. Soltner, P. Blumler, Dipolar Halbach magnet stacks made from identically shaped permanent magnets for magnetic resonance, Concepts Magn. Reson. 36A (2010) 211–222. [26] S.B. Belmonte, R.S. Sarthour, I.S. Oliveira, A.P. Guimaraes, A fieldprogrammable gate-array-based high-resolution pulse programmer, Meas. Sci. Technol. 14 (2003) N1–N4. [27] K. Takeda, A highly integrated FPGA-based nuclear magnetic resonance spectrometer, Rev. Sci. Instrum. 78 (2007). [28] K. Takeda, OPENCORE NMR: open-source core modules for implementing an integrated FPGA-based NMR spectrometer, J. Magn. Reson. 192 (2008) 218– 229. [29] W.N. Tang, W.M. Wang, A single-board NMR spectrometer based on a software defined radio architecture, Meas. Sci. Technol. 22 (2011). [30] R.L. Kleinberg, A. Sezginer, D.D. Griffin, M. Fukuhara, Novel NMR apparatus for investigating an external sample, J. Magn. Reson. 97 (1992) 466–485. [31] S. Anferova, V. Anferov, D.G. Rata, B. Blumich, J. Arnold, C. Clauser, P. Blumler, H. Raich, A mobile NMR device for measurements of porosity and pore size distributions of drilled core samples, Concepts Magn. Reson. 23B (2004) 26–32. [32] G.H. Sorland, H.W. Anthonsen, J.G. Seland, F. Antonsen, H.C. Wideroe, J. Krane, Exploring the separate NMR responses from crude oil and water in rock cores, Appl. Magn. Reson. 26 (2004) 417–425. [33] M. Megawati, M.V. Madland, A. Hiorth, Probing pore characteristics of deformed chalk by NMR relaxation, J. Pet. Sci. Eng. 100 (2012) 123–130. [34] P.J. McDonald, J. Mitchell, M. Mulheron, P.S. Aptaker, J.P. Korb, L. Monteilhet, Two-dimensional correlation relaxometry studies of cement pastes performed using a new one-sided NMR magnet, Cem. Concr. Res. 37 (2007) 303–309.

25

[35] H.T. Pedersen, S. Ablett, D.R. Martin, M.J.D. Mallett, S.B. Engelsen, Application of the NMR-MOUSE to food emulsions, J. Magn. Reson. 165 (2003) 49–58. [36] A. Guthausen, G. Guthausen, A. Kamlowski, H. Todt, W. Burk, D. Schmalbein, Measurement of fat content of food with single-sided NMR, J. Am. Oil Chem. Soc. 81 (2004) 727–731. [37] H.C. Bertram, S.B. Engelsen, H. Busk, A.H. Karlsson, H.J. Andersen, Water properties during cooking of pork studied by low-field NMR relaxation: effects of curing and the RN–gene, Meat Sci. 66 (2004) 437–446. [38] O.V. Petrov, J. Hay, I.V. Mastikhin, B.J. Balcom, Fat and moisture content determination with unilateral NMR, Food Res. Int. 41 (2008) 758–764. [39] S.K. Kuster, E. Danieli, B. Blumich, F. Casanova, High-resolution NMR spectroscopy under the fume hood, Phys. Chem. Chem. Phys. 13 (2011) 13172–13176. [40] F. Dalitz, M. Cudaj, M. Maiwald, G. Guthausen, Process and reaction monitoring by low-field NMR spectroscopy, Prog. Nucl. Magn. Reson. Spectrosc. 60 (2012) 52–70. [41] I. Koh, R. Hong, R. Weissleder, L. Josephson, Sensitive NMR sensors detect antibodies to influenza, Angew. Chem., Int. Ed. 47 (2008) 4119–4121. [42] G.H. Liang, H. Chen, S. Zhang, W.J. Wu, J.L. Kong, Magnetic nanosensors for highly sensitive and selective detection of bacillus Calmette-Guerin, Analyst 137 (2012) 675–679. [43] H.L. Shao, C. Min, D. Issadore, M. Liong, T.J. Yoon, R. Weissleder, H. Lee, Magnetic nanoparticles and microNMR for diagnostic applications, Theranostics 2 (2012) 55–65. [44] D.M. TonThat, M.P. Augustine, A. Pines, J. Clarke, Low magnetic field dynamic nuclear polarization using a single-coil two-channel probe, Rev. Sci. Instrum. 68 (1997) 1527–1531. [45] B.C. Hamans, A. Andreychenko, A. Heerschap, S.S. Wijmenga, M. Tessari, NMR at earth’s magnetic field using para-hydrogen induced polarization, J. Magn. Reson. 212 (2011) 224–228. [46] J.S. Frye, G.E. Maciel, Setting the magic angle using a quadrupolar nuclide, J. Magn. Reson. 48 (1982) 125–131. [47] E.L. Hahn, Spin echoes, Phys. Rev. 80 (1950) 580–594. [48] M. Bak, J.T. Rasmussen, N.C. Nielsen, SIMPSON: a general simulation program for solid-state NMR spectroscopy, J. Magn. Reson. 147 (2000) 296–330. [49] T. Vosegaard, A. Malmendal, N.C. Nielsen, The flexibility of SIMPSON and SIMMOL for numerical simulations in solid- and liquid-state NMR spectroscopy, Monatsh. Chem. 133 (2002) 1555–1574. [50] M. Bak, N.C. Nielsen, REPULSION, a novel approach to efficient powder averaging in solid-state NMR, J. Magn. Reson. 125 (1997) 132–139. [51] H.J. Jakobsen, J. Skibsted, H. Bildsøe, N.C. Nielsen, Magic-angle spinning NMRspectra of satellite transitions for quadrupolar nuclei in solids, J. Magn. Reson. 85 (1989) 173–180. [52] M. Bak, J.K. Thomsen, H.J. Jakobsen, S.E. Petersen, T.E. Petersen, N.C. Nielsen, Solid-state 13C and 31P NMR analysis of urinary stones, J. Urol. 164 (2000) 856– 863. [53] N.A. Curry, D.W. Jones, Crystal structure of brushite, calcium hydrogen orthophosphate dihydrate – neutron-diffraction investigation, J. Chem. Soc. A (1971) 3725–3729. [54] H.J. Jakobsen, H. Bildsøe, M. Brorson, Z. Gan, I. Hung, Direct observation of 17 O–185/187Re 1J-coupling in perrhenates by solid-state 17O VT MAS NMR: temperature and self-decoupling effects, J. Magn. Reson. 230 (2013) 98–110. [55] A.R. Grimmer, K.H. Jost, D. Muller, J. Neels, Kristallographische und hochauflösende festkörper-NMR-untersuchungen an K3F(PO3F), J. Fluorine Chem. 34 (1987) 347–360. [56] J.L. Payen, J. Durand, L. Cot, J.L. Galigne, Structural study of potassium monofluorophosphate K2PO3F, Can. J. Chem. 57 (1979) 886–889. [57] J. Skibsted, N.C. Nielsen, H. Bildsøe, H.J. Jakobsen, 51V MAS NMR spectroscopy: determination of quadrupole and anisotropic shielding tensors, including the relative orientation of their principal-axis systems, Chem. Phys. Lett. 188 (1992) 405–412. [58] J. Skibsted, C.J.H. Jacobsen, H.J. Jakobsen, 51V chemical shielding and quadrupole coupling in ortho- and metavanadates from 51V MAS NMR spectroscopy, Inorg. Chem. 37 (1998) 3083–3092. [59] S.W. Sparks, P.D. Ellis, 195Pt shielding tensors in potassium hexachloroplatinate(IV) and potassium tetrachloroplatinate(II), J. Am. Chem. Soc. 108 (1986) 3215–3218. [60] R. Siegel, T.T. Nakashima, R.E. Wasylishen, Application of multiple-pulse experiments to characterize broad NMR chemical-shift powder patterns from spin-1/2 nuclei in the solid state, J. Phys. Chem. 108B (2004) 2218–2226.