Skip to main content
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2021 May 6;59(9-10):975–990. doi: 10.1002/mrc.5141

Looking into the dynamics of molecular crystals of ibuprofen and terephthalic acid using 17O and 2H nuclear magnetic resonance analyses

Chia‐Hsin Chen 1, Ieva Goldberga 1, Philippe Gaveau 1, Sébastien Mittelette 1, Jessica Špačková 1, Chuck Mullen 2, Ivan Petit 3, Thomas‐Xavier Métro 4, Bruno Alonso 1, Christel Gervais 3, Danielle Laurencin 1,
PMCID: PMC8518726  PMID: 33615550

Abstract

Oxygen‐17 and deuterium are two quadrupolar nuclei that are of interest for studying the structure and dynamics of materials by solid‐state nuclear magnetic resonance (NMR). Here, 17O and 2H NMR analyses of crystalline ibuprofen and terephthalic acid are reported. First, improved 17O‐labelling protocols of these molecules are described using mechanochemistry. Then, dynamics occurring around the carboxylic groups of ibuprofen are studied considering variable temperature 17O and 2H NMR data, as well as computational modelling (including molecular dynamics simulations). More specifically, motions related to the concerted double proton jump and the 180° flip of the H‐bonded (–COOH)2 unit in the crystal structure were looked into, and it was found that the merging of the C=O and C–OH 17O resonances at high temperatures cannot be explained by the sole presence of one of these motions. Lastly, preliminary experiments were performed with a 2H–17O diplexer connected to the probe. Such configurations can allow, among others, 2H and 17O NMR spectra to be recorded at different temperatures without needing to tune or to change probe configurations. Overall, this work offers a few leads which could be of use in future studies of other materials using 17O and 2H NMR.

Keywords: solid state NMR, deuterium, molecular crystals, oxygen‐17, dynamics, GIPAW, diplexer, molecular dynamics, tautomerism, hydrogen bonding


17O and 2H nuclear magnetic resonance (NMR) analyses of crystalline ibuprofen and terephthalic acid are reported. First, improved 17O‐labelling protocols of these molecules are described using mechanochemistry. Then, dynamics occurring around the carboxylic groups of ibuprofen are studied considering variable temperature 17O and 2H NMR data, and computational modelling. Lastly, preliminary experiments were performed with a 2H–17O diplexer connected to the probe. Overall, this work offers leads for future studies of other materials using 17O and 2H NMR.

graphic file with name MRC-59-975-g003.jpg

1. INTRODUCTION

The potential of solid‐state nuclear magnetic resonance (NMR) techniques for helping study the structure and reactivity of complex (bio)molecular and materials systems has significantly increased in recent years, thanks to numerous developments made in terms of instrumentation (e.g., ultrahigh magnetic fields and ultrafast magic angle spinning—MAS—probes),[ 1 , 2 , 3 ] pulse sequence developments (e.g., 1H‐detected sequences, ultrawide line, and broadband acquisition methods),[ 2 , 4 , 5 , 6 , 7 , 8 ] more efficient and/or selective isotopic enrichment approaches,[ 9 , 10 , 11 , 12 ] and Dynamic Nuclear Polarization (DNP).[ 13 , 14 ] Studies on challenging quadrupolar nuclei of low receptivity are increasingly being reported, including for some of the most « exotic » ones,[ 15 , 16 ] like 43Ca, 61Ni, 87Sr and 90Zr, just to name a few. Moreover, investigations aiming at understanding the dynamics occurring at different timescales within (bio)molecular and materials systems are seen as highly important, not only because of the impact they can have on the NMR spectra and their acquisition conditions, but also and more importantly because of the insight they can provide on the properties of a given (bio)molecule or material.[ 17 , 18 , 19 , 20 , 21 ]

When it comes to studying dynamics in materials involving small organic molecules, 2H and 17O are both attractive quadrupolar nuclei.[ 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 ] This is notably true in crystals composed of molecules with carboxylic acid groups associated as dimers. Indeed, the dynamics of protons « hopping » between the H‐bonded C=O and C–OH groups have been the object of much research, including using 2H and 17O magnetic resonance techniques.[ 24 , 30 , 31 , 32 ] Among recent 2H studies, Schmidt and Sebastiani's work showed by ab initio calculations that the 2H quadrupolar parameters in strongly H‐bonded systems could not be directly transposed into structural data such as bond lengths, due to the importance of non‐linear effects and collective motions.[ 32 ] Regarding 17O, Wu and co‐workers' investigations are worth highlighting: they demonstrated by using variable temperature 17O NMR analyses that it is possible to study the concerted double‐proton jumping, and estimate the energy asymmetry between the two H‐bonded tautomers experimentally.[ 24 ] More generally speaking, when looking at the available literature, it appears that the information arising from 2H and 17O NMR analyses can potentially provide complementary information on the dynamic processes taking place within carboxylic acid dimers.

Considering the above context, the purpose of this article is to expand the possibilities for studying dynamics within molecular crystals by 2H and 17O NMR. Our studies were performed on two molecules: terephthalic acid and ibuprofen. Because of the very low natural abundance of both isotopes (Table 1), analyses were performed on enriched samples. First, improved 17O‐labelling protocols based on mechanochemistry[ 33 ] are described for both molecules. They were used to produce two doubly labelled compounds (enriched in 2H and 17O), referred to as D‐TA* and D‐IBU*, which were isolated as molecular crystals (Scheme 1). Second, a variable temperature 17O MAS NMR study of D‐IBU* is presented. The 17O NMR spectra are discussed using complementary computational modelling studies, including molecular dynamics simulations and ab initio calculations of NMR parameters. Lastly, preliminary NMR investigations involving a 2H–17O diplexer connected to the NMR probe are presented. This system was used in the case of D‐TA*, enabling some of the 2H and 17O NMR experiments to be performed at different temperatures in a back‐to‐back fashion with no need to retune the probe. This configuration is not standard considering the very similar Larmor frequencies of both nuclei (Table 1). Overall, the results presented in this manuscript will be of interest for future investigations of structure and dynamics using 17O and/or 2H NMR.

TABLE 1.

Nuclear spin properties of 1H, 2H and 17O

1H 2H 17O
Spin 1/2 1 5/2
Natural abundance 99.98% 0.015% 0.037%
Larmor frequency ν0 (MHz) at 14.1 T 600.1 92.1 81.4
Quadrupole moment Q (×10−30 m2)[ 34 ] / 2.86 −25.58

SCHEME 1.

SCHEME 1

Structures of the doubly labelled 17O and 2H molecules: terephthalic acid (left) and ibuprofen (right). The 17O‐labelling procedure by mechanochemistry leads to the predominant labelling of one oxygen per carboxylic group, but with both O atoms having the same probability to be enriched (which is why they are both highlighted in red)

2. EXPERIMENTAL SECTION

2.1. Isotopic labelling procedures for terephthalic acid and ibuprofen

The following reagents were used as received: ibuprofen (C13H18O2, Sigma Aldrich, ≥98% purity, racemic form, noted here IBU), terephthalic acid (C8H6O4, Janssen Chemicals, 98% purity, noted here TA), and deuterated terephthalic acid (C8D4H2O4, with full deuteration on the aromatic cycle, 98% purity, Cambridge Isotope Laboratories, noted here D‐TA) 1,1′‐carbonyldiimidazole (C7H6N4O, TCI, >97% purity, noted here CDI).

18O‐labelled water was purchased from Eurisotop (isotopic composition indicated in the certificate of analysis: 97.1% 18O, 1.1% 17O and 1.8% 16O) or CortecNet (isotopic composition indicated in the certificate of analysis: 99.3% 18O, 0.2% 17O and 0.5% 16O), and used for the optimisation of the enrichment protocols of terephthalic acid and ibuprofen, respectively, as detailed below.

17O‐labelled water was purchased from CortecNet. The isotopic composition indicated in the certificate of analysis was 8.6% 18O, 90.4% 17O and 1.0% 16O for the ~90% 17O‐enriched H2O, and 43% 18O, 41% 17O and 16% 16O for the ~ 40% 17O H2O.

D2O (> 99.96%, CAS 7789‐20‐0) was purchased from Sigma‐Aldrich.

Reagent grade solvents were used in all purification protocols.

2.1.1. 17O‐enrichment of terephthalic acid and deuterated terephthalic acid by mechanochemistry

The enrichment procedure was first optimised using 18O‐enriched water, due to its lower cost compared to 17O‐enriched water. Terephthalic acid (50 mg, 0.30 mmol and 1.0 eq) and CDI (107 mg, 0.66 mmol and 2.2 eq) were introduced in a stainless‐steel grinding jar (10 ml inner volume) with two stainless steel balls (10 mm diameter). The jar was closed and subjected to grinding for 60 min in the MM400 mixer mill operating at 25 Hz. 18O‐labelled water (16 μl, 0.88 mmol and ~ 3 eq) was then introduced in the jar, and the mixture was subjected to further grinding for 60 min at 25 Hz. To help recover the product, non‐labelled water (1 ml) was introduced in the jar, and the medium was subjected to grinding for 2 min at 25 Hz. Then, it was transferred to an Erlenmeyer flask (together with 2 × 1 ml of non‐labelled water, used here to rinse the grinding jar). The medium was acidified to pH ~ 1 by adding a few drops of concentrated HCl (6 mol.L−1 aqueous solution, ~15 drops). The white precipitate was immediately filtered on a glass frit, washed with 4 × 0.5 ml of a 1 mol.L−1 aqueous solution of HCl, and then with 0.5 ml of ultrapure water. The recovered solid was dried overnight under vacuum. Average yield (n = 3, n representing the number of independent labelling experiments performed): 43 mg, 86%. Average 18O‐enrichment level per oxygen, as determined by MS (n = 3) = 46 ± 1%.

For the preparation of the 17O‐ and 2H‐enriched phase (noted D‐TA*), D4‐terephthalic acid was used as a precursor, and the 17O‐labelling step was performed following the optimised 18O‐enrichment protocol described above using ~ 40% 17O‐labelled water for the hydrolysis step instead (yield: 42 mg; average 17O‐enrichment level per oxygen, as determined by MS = 20 ± 1%). The 1H and 13C solution NMR spectra, ESI‐MS spectrum and X‐ray powder diffraction (XRD) powder pattern of D‐TA* are shown in supporting information (Figures S1–S3).

2.1.2. 17O‐enrichment of ibuprofen by mechanochemistry

The enrichment procedure was first optimised using 18O‐enriched water, due to its lower cost compared to 17O‐enriched water. Ibuprofen (120 mg, 0.58 mmol and 1.0 eq) and CDI (103 mg, 0.64 mmol and 1.1 eq) were introduced into a Retsch MM400 stainless steel grinding jar (10 ml inner volume) containing two stainless steel balls (10 mm diameter). The jar was closed and subjected to grinding for 30 min in the MM400 mixer mill operating at 25 Hz. 18O‐labelled water (99.3%, 16 μl, 0.87 mmol and 1.5 eq) was then added into the jar, and the mixture was subjected to further grinding for 30 min at 25 Hz. To help recover the product, non‐labelled water (2 ml) was added into the jar, and the content was subjected to grinding for 2 min at 25 Hz. Then, the medium was transferred to a beaker (together with a sufficient amount of non‐labelled water [8 ml] used here to rinse the jar). The medium was then acidified to pH ~1 with an aqueous solution of HCl (6 mol.L−1, ~ 18 drops). The white precipitate was filtered on a glass frit, and 2 × 5 ml of non‐labelled water was used to help recover the rest of the product from the beaker. The product was washed with a 1 mol.L−1 aqueous solution of HCl and ultrapure water (3 × 1 ml of each), and then dried under vacuum. Average synthetic yield (n = 3): 89 mg, 75%. Average 18O‐enrichment level per oxygen, as determined by MS (n = 3): 45 ± 2%.

For the preparation of the 17O‐enriched phase (noted IBU*), the 17O‐labelling step was first performed following the optimised 18O‐enrichment protocol described above, but using ~ 90% 17O‐labelled water (16 μl and 1.5 eq.) for the hydrolysis step. Synthetic yield: 96 mg, 79%. Average 17O‐enrichment level per oxygen, as determined by MS = 36 ± 2% (average of five measurements). The 1H and 13C solution NMR spectra, ESI‐MS spectrum, and XRD powder pattern of IBU* are shown in supporting information (Figures S4–S6).

2.1.3. 2H‐enrichment of 17O‐labelled ibuprofen

The deuteration step was performed by mixing 59.4 mg of 17O‐labelled ibuprofen (IBU*) with 0.5 ml of D2O (~95 eq) in a 1.0‐ml Eppendorf tube. The sample was sonicated for 1 min to ensure good mixing of the reagents. The mixture was then left for 3 days on a laboratory rocker shaker (Stuart SSL4—Rocker), with 70 oscillations per minute at room temperature. The mixture was spun down at 20,000 rpm for 15 min, and the excess water was discarded. The sample was then freeze‐dried under vacuum for 6 h to remove the rest of the water. Synthetic yield: 56.4 mg, 95%. The IR and XRD powder patterns of the doubly labelled product D‐IBU* can be found in supporting information (Figures S4 and S8). A similar 2H‐labelling protocol was also applied to non‐labelled ibuprofen, on which complementary 2H MAS NMR experiments were performed.

2.2. General characterisation protocols of the enriched compounds

Infrared (IR) spectra were recorded on a Perkin Elmer Spectrum 2 FT‐IR instrument. The attenuated total reflectance (ATR) measurement mode was used (diamond crystal), and measurements were performed in the 400 to 4000 cm−1 range.

Powder XRD analyses were carried out on an X'Pert MPD diffractometer using Cu Kα1 radiation (λ = 1.5406 Å) with the operation voltage and current maintained at 40 kV and 25 mA, respectively. Diffractograms were recorded between 5° and 50° (or 60°) in 2θ, with a step size of 0.017°, and a time per step of 20 to 40 s.

Mass spectrometry (MS) analyses were performed on a Waters Synapt G2‐S apparatus, using electrospray ionisation in negative mode in a range of 50–1500 Da. Capillary and cone voltages were set to 3000 and 30 V, respectively. The source temperature was 100°C, and the desolvation temperature was set to 250°C. Data were processed by MassLynxV4.1 software. For each product, a solution was prepared (in ethanol or DMSO, depending on the solubility), which was analysed five times by ESI‐MS.

1H and 13C solution NMR spectra were recorded on an Avance III Bruker 600 MHz spectrometer equipped with a TCI Prodigy cryoprobe, using DMSO‐d 6 as a solvent. Chemical shifts were referenced to the residual solvent peaks at 2.50 ppm (1H NMR spectra) and 39.52 ppm (13C NMR spectra).

The melting points of ibuprofen (IBU) and its enriched counterparts (IBU* and D‐IBU*) were measured on a Büchi B‐540 Melting Point Apparatus. Samples were heated up to 70°C, and then with a ramp of 1°C min−1 up to 80°C. All measurements were done in triplicate (n = 3) and are reported in Table S1. The 2H and 17O NMR spectra of melted D‐IBU* are shown in Figure S9.

2.3. Solid‐state NMR experiments

Solid‐state NMR experiments were performed on a Varian VNMRS 600 MHz (14.1 T) spectrometer, using, in the vast majority of cases, a PhoenixNMR HXY probe equipped with a 3.2 mm probe head. The PhoenixNMR probe was tuned to 1H (599.82 MHz), 2H (92.07 MHz) and 17O (81.33 MHz), using a 2H–17O diplexer. Conversion from multiple to single port tuning required a ‘special’ tuning plug‐in and diplexer built specifically for 2H and 17O. Radio‐frequency (RF) for 2H and 17O from the NMR console goes through a PhoenixNMR diplexer that provides >100 dBc isolation and then into the probe on a single channel. The PhoenixNMR probe is adapted to provide an over‐coupled double resonant structure for the 2H and 17O frequencies, which in conjunction with the diplexer can allow the observation of one channel while irradiating on the other. Spectra were recorded under static or MAS conditions, with spinning speeds ranging from 5 to 18 kHz, depending on the sample. In the case of the 2H‐enriched ibuprofen (not labelled in 17O), complementary 2H MAS NMR experiments were performed on a Varian T3 HX probe, tuned to 2H and 1H, and spinning at 5 kHz.

17O NMR experiments were recorded using DFS (double frequency sweep) pulse sequence with a rotor‐synchronised echo of one rotor period. The parameters were as follows: DFS pulse of 500 μs (νRF[17O] ~ 7.4 kHz), with a sweep between 70 and 200 kHz, followed by a 17O excitation pulse of 2 μs, echo delay of 55.55 μs, and a π pulse of 4 μs. All experiments were performed with a MAS frequency of 18 kHz and recycle delay of 0.5 s, and the number of transients acquired was 2400 for D‐TA* and 4000 for D‐IBU*. 17O NMR chemical shifts were referenced to D2O at −2.7 ppm (which corresponds to tap water at 0 ppm). The 17O nutation experiment was also recorded using D2O at room temperature (with a natural abundance of 17O).

The 2H NMR experiments performed on the PhoenixNMR probe were carried out using a solid echo pulse sequence with a π/2 pulse length of 3.7 μs. 2H MAS NMR experiments on D‐TA* were performed with a MAS frequency of 5 kHz and rotor‐synchronised echo delay of one rotor period (200 μs), and spectra were acquired using 14,000 transients and recycle delay of 0.5 s. Static experiments were performed on D‐IBU*, using an echo delay of 30 μs, and acquiring 122,500 transients with a recycle delay of 0.5 s. 2H chemical shifts were referenced to pure D2O at 4.6 ppm. The 2H nutation experiment was also recorded using D2O at room temperature.

The 1H nutation experiment was recorded using adamantane at room temperature with a MAS frequency of 10 kHz.

Variable temperature studies were carried out under static or magic‐angle spinning conditions (spinning at 5 or 18 kHz). In each situation, the temperature was calibrated using Pb(NO3)2.[ 35 ] When working under MAS conditions, careful attention was paid to the setting of the magic angle using KBr (79Br resonance). The accuracy of the magic‐angle at each temperature of analysis could be verified on D‐TA*, by the absence of symmetric splitting of the 2H resonance and sidebands under magic angle spinning (see Figure S10). If judged necessary, the magic angle was carefully reset at the temperature of interest using KBr.

2.4. Computational details

The NMR chemical shift calculations were performed within the density functional theory (DFT formalism using the QUANTUM‐ESPRESSO (QE))[ 36 ] software. The PBE generalised gradient approximation[ 37 ] was used, and the valence electrons were described by norm‐conserving pseudopotentials[ 38 ] in the Kleinman–Bylander form.[ 39 ] The wave functions were expanded on a plane wave basis set with kinetic energy cutoff of 80 Ry. The shielding tensor was computed using the Gauge Including Projector Augmented Wave (GIPAW)[ 40 ] approach, which permits the reproduction of the results of a fully converged all‐electron calculation. Absolute shielding tensors are obtained. The isotropic chemical shift δiso is defined as δiso = −[σ − σ ref], where σ is the isotropic shielding and σ ref is the isotropic shielding of the same nucleus in a reference system as previously described.[ 11 ] The principal components V xx, V yy and V zz of the electric field gradient (EFG) tensor are obtained by diagonalisation of the tensor. The quadrupolar interaction can then be characterised by the quadrupolar coupling constant C Q and the asymmetry parameter η Q, which are defined as C Q = eQV zz/h and η Q = (V yy − V xx)/V zz. The experimental values of the quadrupole moments of 17O (Q = −25.58 × 10−30 m2) and 2H (Q = +2.86 × 10−30 m2)[ 41 ] were used to calculate C Q.

Two crystallographic structures of racemic ibuprofen (CCDC 1041383 and 128796) were tested as starting points. The geometry optimisation of all atomic positions (keeping cell parameters fixed to experimental values) was performed for both low‐ and high‐energy tautomers using the VASP code[ 42 ] and a Monkhorst−Pack k‐space grid size of 2 × 3 × 2. NMR parameters were then calculated keeping the relaxed atomic positions.

For the evaluation of the rotational barrier energy of the H‐bonded (–COOH)2 dimeric unit in ibuprofen, the crystalline structure # CCDC 128796 was used as a starting point (because the ibuprofen dimer is at the centre of the unit cell, thereby facilitating the application of a geometrical torsion). The dihedral angle Car–C(H)–C–O was varied and constrained, while the geometry of the rest of the molecule was optimised, and the energy subsequently calculated.

AIMD (Ab initio Molecular Dynamics) simulations were carried out with the CP2K code[ 43 ] consisting in the Born–Oppenheimer MD (BOMD) with PBE electronic representation, including the Grimme (D3) correction for dispersion,[ 44 ] GTH pseudopotentials,[ 45 ] combined plane wave, and triple‐zeta valence with polarization (TZVP) basis sets.[ 46 ] The BOMD was performed using the NVT ensemble. The Nose–Hoover thermostat was used to control the average temperature at 300 K. Trajectories were accumulated over ~ 20 ps with a time step of 0.5 fs. 2H and 17O NMR calculations were performed with QE every 400 steps, that is, every 200 fs.

3. RESULTS

3.1. Improved 17O‐enrichment protocols and preparation of doubly labelled molecules

Due to the very low natural abundance of 17O and 2H (Table 1), isotopic labelling is needed in order to be able to perform variable temperature solid‐state NMR analyses in a reasonable time. While the deuteration of the carboxyl group can be easily achieved by exposure of the molecules to an excess of D2O, the 17O‐labelling of the carboxylic oxygen atoms is not as straightforward, notably due to the high cost of 17O‐enriched water (1 ml of 90% 17O‐enriched water can cost up to ~2000 €). A previous study showed that terephthalic acid and ibuprofen could be enriched in 17O using mechanochemistry in a cost‐efficient and user‐friendly way.[ 11 ] However, our enrichment levels only averaged to ~ 3%–8% per carboxylic oxygen for these molecules, which is ~ 2.5 to 7 times less than the maximum enrichment we could have expected based on the reactions and precursors involved. Hence, as part of this work, we first re‐optimised the 17O‐enrichment protocols for these two molecules.

The 17O‐labelling procedure we have developed involves two ball‐milling steps, which are performed back‐to‐back and followed by IR spectroscopy[ 11 , 47 ]: an activation of the carboxylic function using 1,1′‐carbonyldiimidazole (CDI), followed by the hydrolysis of the acyl‐imidazole intermediate using 17O‐enriched water. If either of these steps is incomplete, the final enrichment level decreases. We retested the reaction conditions for TA and IBU using 18O‐enriched water (due to its lower cost compared to 17O‐enriched water) and found that longer milling times (30–60 min, instead of 5–10 min) were needed to ensure a better mixing of the reagents and full completion of activation and hydrolysis (see Figure S7 for illustrations in the case of IBU). We verified the reproducibility of the newly optimised enrichment protocols and then used them for the 17O‐labelling of terephthalic acid and ibuprofen. In the case of terephthalic acid, the optimised 17O‐enrichment protocol was applied to a deuterated form of the molecule (D‐TA, fully deuterated on the aromatic cycle), while for ibuprofen, it was applied to the non‐labelled form of the molecule. The average 17O‐enrichment level per carboxylic oxygen achieved was ~ 20% for D‐TA* (when using ~ 40% 17O‐labelled water for the hydrolysis) and ~ 36% for IBU* (when using ~ 90% 17O‐labelled water for the hydrolysis). The mass spectra of the 17O‐labelled molecules can be found in the supporting information, together with other analyses that demonstrate the purity and crystallinity of the isolated molecules (1H NMR, 13C NMR and powder X‐ray diffraction) (see Figures S1–S6).

For ibuprofen, the 17O‐enriched compound was then suspended in the presence of an excess of D2O in order to exchange the carboxylic O–H group by O–D and form the doubly‐labelled compound D‐IBU*. This reaction was followed by IR spectroscopy by looking at the O–H vibration modes' replacement by O–D ones (see Figure S8). It is worth noting that after completion of the reaction, the NMR rotor was immediately packed and stored under vacuum in the freezer to avoid back exchange of the carboxylic deuterium upon exposure to atmospheric humidity.

3.2. Variable temperature 17O MAS NMR of ibuprofen: Experiments and computational modelling

The 17O MAS NMR spectra of D‐IBU* were recorded at 14.1 T, while regulating the temperature between 0°C and +60°C (Figure 1). The 17O spectra show two second‐order quadrupolar lineshapes at 0°C, which correspond to the C=O and C–OH groups. These signals progressively merge as the temperature is increased, leading to a single resonance with a characteristic second‐order quadrupolar lineshape at 60°C. A similar observation had been made in our previous 17O NMR study of non‐deuterated ibuprofen.[ 11 ] Fits of the spectra recorded at 0°C and 60°C can be found in supporting information (Figure S11). The fitted parameters at 0°C (i.e., δ iso = 308 ± 2 ppm and |C Q| = 7.9 ± 0.1 MHz for the C=O, and with δ iso = 179 ± 2 ppm and |C Q| = 7.3 ± 0.1 MHz for the C–OH) were in line with those reported previously for non‐deuterated ibuprofen, and with results from GIPAW‐DFT calculations on the crystal structure.[ 11 ] A tentative fit of the data recorded at 60°C was also performed, resulting in intermediate δ iso and |C Q| values (δ iso = 244 ± 2 ppm and |C Q| = 7.6 ± 0.1 MHz).

FIGURE 1.

FIGURE 1

Variable temperature 17O MAS NMR spectra of D‐IBU*. All 17O NMR spectra shown here were recorded using the same acquisition conditions (including the same number of scans). The temperatures indicated correspond to the sample temperature inside the rotor at each spinning speed, as determined from calibrations using Pb(NO3)2. Tentative fits of the 0°C and 60°C spectra can be found in supporting information (Figure S11). « * » symbols correspond to spinning sidebands

Although several polymorphs of racemic ibuprofen have been reported,[ 48 ] the most stable form was obtained here, for which no polymorphic change is expected over the temperature range investigated by NMR. The changes in 17O solid‐state NMR spectra around room temperature could, however, be due to (i) variations in the relative populations of the two tautomeric forms undergoing double proton jumps (Scheme 2a),[ 24 ] and/or (ii) 180° flips of the (‐COOH)2 units (Scheme 2b),[ 49 ] the latter motion having been suggested to exist for ibuprofen by Geppi and co‐workers.[ 50 ] Both of these motions were further analysed, using additional computational simulations to help rationalise the observations made by 17O NMR.

SCHEME 2.

SCHEME 2

Illustration of dynamics that can occur between H‐bonded carboxylic acids, leading to interconversion between the two tautomeric forms: (a) concerted double proton jump and (b) concerted 180° flips of the H‐bonded (–COOH)2 units

First, the possible influence of concerted double proton jumps between the two tautomeric forms of ibuprofen on the 17O MAS NMR data was looked into. Based on Wu and co‐workers' recent work on a series of molecular crystals involving H‐bonded carboxylic dimers, these motions can result in changes in 17O NMR spectra with temperature, depending on the relative population of the tautomeric forms.[ 24 ] Such proton exchanges occur on the ps timescale and can participate in the averaging of 17O NMR parameters between the two tautomeric forms. Here, a molecular dynamics simulation of racemic ibuprofen at 300 K was performed over a duration of 20 ps. This simulation brings evidence of the concerted double proton jumps and shows the evolution of the calculated average 17O NMR parameters of each oxygen (Figure 2), which appear to all progressively evolve towards averaged values. Nevertheless, a full averaging of the 17O NMR parameters could not be reached within the 20 ps timescale studied here.

FIGURE 2.

FIGURE 2

Molecular dynamics simulation of the structure of racemic ibuprofen, performed at 300 K, for a duration of 20 ps, with steps of 0.5 fs, with a focus on calculated 17O NMR data. (a) Evolution of the H⋯O bond distances (in Å) in the dimer of the unit cell for which a concerted double 1H jump was observed over the timescale of the calculation performed here (for the other dimer, no jump was observed). (b) Evolution of the DFT‐calculated averaged 17O NMR parameters for each oxygen site in this dimer (NMR parameters were calculated every 400 steps, i.e., every 200 fs; and for each new point calculated along the MD, the new value was averaged with the previous ones)

For concerted double‐proton jumps, the extent of averaging of the NMR parameters between both tautomers depends on the energy difference between the two forms and the analysis temperature.[ 24 ] In the case of racemic ibuprofen, Kolesov et al. reported an energy difference between both tautomers ~ 7.7 kJ mol−1, based on Raman spectroscopy measurements.[ 51 ] In our case, starting from the crystalline structure of racemic ibuprofen,[ 52 ] the energy difference between the tautomeric forms was calculated by DFT, yielding a value ~ 8.8 kJ mol−1, which is of the same order of magnitude as the experimental value of Kolesov. A similar energy difference was found when performing calculations on structures in which the initial OH positions before geometry relaxation were varied, when starting from different cristallographic files (corresponding to neutron or X‐ray data), and when modifying the dispersion energy parameters in VASP. Using these energy differences and the DFT‐calculated 17O NMR parameters of each tautomer (Figure 3a), an approach similar to the one recently described by Wu and co‐workers was then applied,[ 24 ] in order to see how the 17O NMR parameters may vary with temperature (Figure 3b). The most significant variations were observed for the isotropic chemical shifts. Yet, based on these calculations, a change of less than 10 ppm would have been expected for the « C=O » and « C‐OH »‐like resonances over the temperature range studied here (green‐shaded region, Figure 3b), regardless of the energy difference chosen (8.8 or 7.7 kJ mol−1). Similar conclusions would also have been expected for deuterated ibuprofen, for which an intermediate energy difference ~ 6.7 kJ mol−1 between the two tautomeric forms has been derived from Raman spectroscopy analyses.[ 51 ] Overall, this implies that no merging of the two 17O signals as observed in Figure 1 would have been expected to arise from concerted double 1H‐jumps only, on the basis of the calculated 17O NMR parameters of the two tautomers and their energy difference.

FIGURE 3.

FIGURE 3

(a) Representation of the two tautomeric forms of ibuprofen and their DFT‐calculated 17O NMR parameters (for the (–COOH)2 dimer) (not all atoms in the unit cell are shown here for clarity). (b) Temperature dependency of the 17O isotropic chemical shifts of the two tautomeric forms, for ΔE values of 8.8 kJ mol−1 (as calculated here by DFT for protonated ibuprofen) and of 7.7 and 6.7 kJ mol−1 (as determined experimentally by Raman for protonated and deuterated ibuprofen, respectively).[ 51 ] The region shaded in green corresponds to the temperature range studied here by 17O NMR (i.e., between 0°C and 60°C). The equations used for the plot are recalled in the grey‐shaded box[ 24 ] (in which the DFT‐calculated values of δC=O and δC‐OH of the two tautomeric forms given just above (Figure 3a) were used)

The motion related to a concerted rotation of the (–COOH)2 group was then looked into. As mentioned above, the presence of this 180° flip had already been proposed in racemic ibuprofen on the basis of comprehensive 1H and 13C NMR analyses.[ 50 ] Moreover, it is worth noting that the observations of the evolution of the variable temperature 17O MAS NMR spectra of ibuprofen recall those reported by Wu and co‐workers for nicotinic acid, in which the 180° rotation of an H‐bonded –COOH unit was studied between ~ −20°C and 100°C (although in the latter case, it is a single –COOH moiety H‐bonded to N which undergoes the 180° flip).[ 22 ] Here, in contrast to the double proton jumps, the 180° flip motions in ibuprofen could not be made evident by molecular dynamics simulation due to the « short » timescale of our MD calculations. Nevertheless, the activation energy associated with this motion was evaluated by DFT, using an approach similar to the one proposed by Wu and co‐workers for nicotinic acid.[ 22 ] More specifically, starting from the published structure of racemic ibuprofen, the (–COOH)2 dimeric fragment was rotated (keeping all other atomic positions constant), and the energy difference for each torsion angle was then calculated after the relaxation of all atomic positions (except for those defining the dihedral angle). Some of the calculated geometries are shown in Figure 4. In doing so, it was found that the energy barrier is ~ 75–100 kJ mol−1 (the value depending on the direction chosen for the rotation of the (–COOH)2 unit). These values are very different from the one reported by Geppi on the basis of 1H and 13C NMR analyses, which had estimated the barrier to ~13 kJ mol−1. However, they are still consistent with the energy barrier reported for (–COOH)2 180° flips in other carboxylic dimers, for which values ranging from 50 to 85 kJ mol−1 have been reported.[ 30 , 53 ] Part of the difference with the value reported by Geppi may come from the fact that computationally, we are probing the rotation of the carboxylic group by « freezing » the rest of the structure, while in the experimental study, other motions are occurring within the crystal structure,[ 50 , 54 , 55 ] which may indirectly facilitate the rotation of the carboxylic dimer. Overall, although the experimental and computational data shown here does not yet achieve a complete picture of the (–COOH)2 180° flip and double proton exchange processes occurring in racemic ibuprofen, it nevertheless demonstrates the added value of performing high‐resolution 17O NMR analyses and computational simulations to try to help understand these movements.

FIGURE 4.

FIGURE 4

DFT evaluation of the activation energy for the rotation of the (–COOH)2 dimeric unit in crystalline racemic ibuprofen (using the CCDC crystal structure # 128796). Illustrations of some of the torsion angles are shown on the right (with C, O and H in grey, red and white, respectively; not all atoms in the unit cell are displayed here for clarity)

3.3. Variable temperature analyses using a 1H–2H–17O tuning configuration

To complement the 17O MAS NMR study of the H‐bonded carboxylic groups dynamics in D‐IBU*, 2H NMR experiments were carried out. Here, the possibility of performing 2H and 17O NMR analyses in a back‐to‐back fashion was investigated, using a diplexer connected to the probe for 1H–2H–17O tuning. Indeed, being able to analyse 2H and 17O local environments at any given temperature without changing the NMR probe configuration can be attractive, especially to avoid hysteresis effects that can occur upon heating or cooling some crystalline phases.

A schematic illustration of the mode of connection of the diplexer to the NMR probe is shown in Figure 5a, with further details being provided in the experimental section. The nutations obtained upon calibration of RF pulses using D2O for 2H and 17O, and adamantane for 1H, are shown in Figure 5b. The loss in sensitivity of this triple resonance mode compared to the double resonance configuration was found to be reasonable (see Figure S12).

FIGURE 5.

FIGURE 5

(a) Schematic representation of the connection mode of the diplexer to the NMR probe and (b) nutations achieved on the 1H, 2H and 17O channels in triple resonance mode. The 810°/90° intensity ratio determined here for 2H is 0.51

This probe configuration was first tested on the doubly labelled D‐TA* phase due to its higher weight percentage in both 17O and 2H. The 2H and 17O MAS NMR spectra were recorded between −30°C and +60°C (Figure 6). No significant changes were observed in the 2H NMR. Although 2H relaxation measurements may have led to observable differences, these were not performed at this stage.

FIGURE 6.

FIGURE 6

Variable temperature 17O and 2H NMR spectra of D‐TA*, recorded under magic angle spinning conditions. The temperatures indicated correspond to the sample temperature inside the rotor, as determined from calibrations at different MAS speeds using Pb(NO3)2

In the case of 17O NMR, only subtle variations with temperature were observed. A tentative deconvolution of the spectra is shown in Figure S13 (supporting information), considering the presence of 2 resonances corresponding to « C=O » and « C–OH » like environments, in agreement with the previous 17O NMR work reported for non‐deuterated terephthalic acid.[ 11 ] Due to the overlap of both resonances, MQMAS experiments at each temperature of analysis (or variable temperature analyses at a second magnetic field) would have been needed to confirm the fitted parameters. Nevertheless, the small variations observed for D‐TA* over this temperature range may reflect effects of H‐bonding tautomerism and/or polymorphic changes of terephthalic acid.[ 56 , 57 , 58 , 59 , 60 , 61 , 62 ] Indeed, on the one hand, previous studies in the literature have shown that the H‐bonded carboxylic protons of terephthalic acid are dynamically disordered at room temperature, undergoing concerted double‐proton jumps (Scheme 2a).[ 56 , 61 , 63 ] The fact that the free enthalpy difference between both tautomeric forms has been estimated experimentally to only ~ 2 kJ mol−1 for terephthalic acid[ 61 ] implies that small changes should be observable over the temperature range studied here (because this value is of the same order of magnitude as those reported for cinnamic acid and aspirin).[ 24 ]

On the other hand, the variations in the 17O MAS NMR spectra may also reflect the existence of several polymorphic forms of terephthalic acid. Out of the three polymorphs reported to date for this molecule, triclinic forms (I) and (II) have been the object of much attention.[ 57 , 58 , 59 , 60 , 62 , 64 ] Both consist of chains of terephthalic acid molecules, in which each molecule is hydrogen‐bonded to two others through carboxylic acid dimer motifs. These chains are further assembled into 2D layers, which pack differently from one polymorph to the other. Moreover, when looking at the terephthalic acid motif itself, slightly different dihedral angles are found between the carboxylic function and the aromatic cycle when both polymorphs are compared.[ 62 ] Although both forms can be observed under ambient temperature and pressure, form (I) appears to be the least stable. Its transformation into form (II) has been studied both experimentally and computationally.[ 57 , 58 , 60 ] More specifically, this transformation was found to occur above 70°C, to be sensitive to pressure and to be caused by a « surface‐mediated nucleation » type of process, triggered by the movement of the supramolecular chains of terephthalic acid at the surface of the crystallites.[ 58 , 59 , 60 ] In our case, given that D‐TA* was obtained here as polymorph (I) (see Figure S1), it is possible that the subtle modifications in the 17O NMR spectra upon heating are indicative of the onset of transformation into form (II), caused by the increase in temperature and possibly also pressure (due to the spinning). More specifically, 17O NMR may indicate that changes in the local environment of the carboxylic functions precede more significant movements of the supramolecular chains. Although the observations made in 17O NMR for D‐TA* were not further investigated at this stage, they point to the interest of analysing molecular crystals by this technique, notably to help understand polymorphic transformations.

Using this 2H–17O probe configuration, preliminary 2H NMR studies were also performed in the case of D‐IBU*. The static 2H NMR spectra revealed the presence of two main resonances (Figure 7a): one broad signal with a characteristic deuterium quadrupolar lineshape and a much sharper signal at the centre of the spectrum, the relative intensity of which was found to increase with temperature, under the measurement conditions used here (Figure S14). The fitting of the broad 2H NMR signal at each temperature was performed, yielding similar quadrupolar parameters between 0°C and 60°C, with |C Q| ~ 170 kHz and η Q ~ 0.1 (Figure S14). These values are close to those reported previously for supercooled and glassy states of deuterated ibuprofen,[ 65 ] as well as for other crystal structures of organic molecules involving H‐bonded carboxylic dimers.[ 66 ] Hence, this 2H resonance is characteristic of (–COOD)2 dimeric structures in crystalline D‐IBU*, with the 2H NMR parameters being averaged between the two interconverting tautomeric forms of ibuprofen. The MD simulations and DFT calculations of 2H NMR parameters associated with these double proton jumps are provided in Figure 7b,c. In agreement with previous observations made by Schmidt and Sebastiani[ 32 ] for H‐bonded carboxylic acids, these calculations show that the « instantaneous » C Q values tend to follow the evolution of the longer O⋯H distance (Figure 7b, dashed vertical lines).

FIGURE 7.

FIGURE 7

(a) Fit of the experimental static 2H NMR spectrum of D‐IBU* at 30°C, after symmetrisation of the 2H lineshape (experimental spectrum in black, and its fit in red, considering two contributions, with in blue the H‐bonded –(COOD)2 dimeric unit of D‐IBU*, and in green the sharp mobile species. (b) Molecular dynamics simulation of the structure of racemic ibuprofen, performed at 300 K, with steps of 0.5 fs (shown here over duration of ~15 ps), with a focus on calculated 2H NMR data. Evolution of the H⋯O bond distances (in Å) in the dimer of the unit cell for which a concerted double proton jump was observed over the timescale of the calculation, and evolution of the « instantaneous » DFT‐calculated 2H quadrupolar coupling constant C Q for this dimer (NMR parameters calculations were performed as a first approximation from the MD structures of protonated ibuprofen, every 200 steps, up to step # 14000, and then every 400 steps). (c) Evolution of the average calculated |C Q| values along with the MD timescale, showing that the 2H |C Q| converges towards an average value

The sharp central 2H NMR resonance, on the other hand, is very weak at 0°C, where it is centred at 5.2 ppm. At 60°C, it becomes increasingly sharp and intense and shifts to 9.4 ppm. Based on the observed chemical shifts and the 2H NMR measurements also performed on melted D‐IBU* (see Figure S9), this resonance appears to arise from different species: it is consistent with a small amount of residual liquid (mobile) water at ~ 0°C, and to the predominant presence of highly mobile ibuprofen at 60°C (in a melt‐like state, the melting temperature of ibuprofen being ~ 75°C). It is worth noting that in comparison to previous variable temperature 2H NMR studies on supercooled and glassy ibuprofen,[ 65 ] the fully motionally narrowed 2H NMR signal is observed here at a higher temperature (80°C in Figure S9, vs. 36°C in the work by Bauer et al.[ 65 ]). Overall, it is clear that 2H and 17O NMR spectra provide complementary information regarding the dynamics occurring around the carboxylic groups in racemic ibuprofen. It can be hypothesised that the progressive increase in mobility within ibuprofen crystals with temperature (especially around the carboxylic function) allows some molecules to behave as in a melt‐like state, which may favour dynamics like the 180° (–COOH)2 flip to occur with energy barriers lower than the ones calculated by DFT, thereby explaining the observations made by 17O NMR. However, more extensive investigations would be needed to confirm this.

4. CONCLUSION

In this article, three different aspects related to the structural analysis of molecular crystals containing carboxylic functions have been looked into. First, improved 17O labelling protocols based on mechanochemistry have been developed. This has allowed average enrichment levels exceeding 20% in 17O per carboxylic oxygen to be reached for two key molecules: (i) ibuprofen, a non‐steoridal anti‐inflammatory drug, which is seen as a « golden standard » for numerous investigations in pharmaceutical sciences aiming at improving drug formulations, and (ii) terephthalic acid, which is one of the most commonly used ligands for the design of metal–organic frameworks (MOFs). In the latter case, the 17O labelling was performed here on a deuterated version of the terephthalic acid precursor (with deuteration on the aromatic ring), thereby leading to the formation of a doubly labelled molecule, which would be of key interest for studying the structure and reactivity or a variety of MOFs, including the most complex ones. Such ligands should enable studying both the dynamics related to the 180° flips of the aromatic ring (using 2H NMR) and the binding properties to metal cations in a given system (using 17O NMR).

Second, the experimental and computational study of 17O and 2H enriched ibuprofen has shed light on the interest in looking at both of these nuclei when studying the dynamics occurring around the carboxylic groups. Although further investigations would be needed to fully understand the experimental observations, this work nevertheless complements previous studies on racemic ibuprofen crystals involving 1H and 13C NMR, by underscoring the complexity of the molecular motions around the carboxylic functions. More generally, this study shows how the NMR study of quadrupolar nuclei like 17O and 2H may provide new opportunities for investigating polymorphic transitions of ibuprofen,[ 67 ] as well as its confined or « supercooled » states.[ 65 , 68 , 69 ] Indeed, the polarity and H‐bonding capability of carboxylic functions imply that these are often key to interactions between molecules and with materials surfaces, meaning that direct insight into the local environment of the carboxylic atoms (among which oxygen) is important for detailed structure characterization purposes.

Lastly, we have presented some of the possibilities provided by combining 2H–17O diplexers to NMR probes for studying by NMR the 2H and 17O nuclei at any given temperature, without having to change probe configurations. The RF performance of the probe used along with the diplexer was shown to be suitable for such applications, as illustrated for both terephthalic acid and ibuprofen. While the experiments shown here were mainly performed in a back‐to‐back fashion, we will look into recording them using « double receiver » setups in future studies. Moreover, beyond these possibilities, the next step will consist of using such hardware configurations to perform multi‐channel 2H and 17O correlation experiments. This pair of nuclear spins has not yet been studied due to the very similar Larmor frequencies of the two isotopes but could offer new opportunities for helping understand further the structure of a variety of molecular and materials systems, considering that H and O are two atoms, which are very often present. This is a point we will be looking into in the near future. Such instrumental developments will add to the ongoing effort to broaden the scope of heteronuclear correlations accessible to NMR spectroscopists with spin pairs of similar Larmor frequencies (e.g., 13C–27Al,[ 70 , 71 ] 13C–45Sc,[ 71 ] 13C–81Br …[ 72 ]).

Supporting information

Figure S1. Powder X‐ray diffraction patterns before and after 17O‐labelling. These correspond to the reported « form (I) » of terephthalic acid (CCDC 1269122).

Figure S2. 13C and 1H solution NMR spectra before and after 17O labelling (DMSO‐d 6 , 600 MHz; solvent peaks are crossed out).

Figure S3. MS analyses of the non‐17O‐labelled precursor in comparison to the 17O‐enriched product.

Figure S4. Powder X‐ray diffraction patterns of the starting ibuprofen material, and the 17O and 17O‐2H enriched forms. These correspond to the polymorph shown on the left (also known as phase « I »).

Figure S5. 13C and 1H solution NMR spectra (DMSO‐d 6 , 600 MHz; solvent peaks marked with a star).

Figure S6 . MS analyses of the non‐17O‐labelled precursor in comparison to the 17O‐enriched product.

Figure S7. ATR‐IR analysis of the starting material, final product, and reaction intermediates, during the 18O‐labelling of ibuprofen by mechanochemistry. The dashed line shows that the C=O stretching frequency of the 18O‐enriched product is shifted to lower wavenumbers in comparison with the non‐labelled precursor.

Figure S8. ATR‐IR analysis of non‐labelled ibuprofen (IBU), and 17O‐labelled ibuprofen, before (IBU*) and after deuteration (D‐IBU*). The replacement of OH vibration modes by OD ones in D‐IBU* confirms the 2H isotopic enrichment (see, for example, O‐D stretching modes in the 1900–2400 cm−1 region highlighted in light green colour).

Figure S9. 2H (left) and 17O (right) NMR spectra of melted D‐IBU*, recorded at 80°C.

Figure S10. a/2H MAS NMR spectra recorded on highly crystalline D‐TA* at −30°C, with the magic angle properly set (top), and after a small mis‐adjustment of the angle (bottom). The splitting of the resonances (with a symmetric splitting over the sidebands) attests to the mis‐setting of the angle. The zoomed section of the central part of the spectrum (dashed grey box) is shown on the right.

Figure S11. Experimental 17O NMR spectra of D‐IBU*, recorded at 0°C and 60°C, under MAS conditions, together with their fits. Parameters used for the fits are reported in the table to the top‐right. The bottom‐right table recalls the parameters which had been reported previously for non‐deuterated ibuprofen.

Figure S12 . Spectra recorded on D‐TA* to compare the sensitivity in different probe configurations for a/2H NMR (acquisition conditions: B0 = 14.1 T, T = −30°C, 5 kHz MAS, solid echo with 1 rotor period echo delay, D1 = 0.5 s, NS = 6,000), and b/17O NMR (acquisition conditions: B0 = 14.1 T, 8 kHz MAS, DFS‐echo with 1 rotor period echo delay, D1 = 0.4 s, NS = 1,000) (It is worth noting that the 17O MAS NMR experiments shown here were performed with 1H decoupling in double resonance mode, and without 1H decoupling in triple resonance mode. However, it was verified that the lack of 1H decoupling in the latter case did not affect the overall signal intensity, and thus cannot explain the differences observed between double and triple resonance configurations.).

Figure S13 . 17O MAS NMR spectra of D‐TA* were recorded at different temperatures and their tentative fit considering the presence of 2 oxygen sites shown. The green and purple lines correspond to the fits, and the dashed orange line the sum of the two contributions.

Figure S14. a/Experimental 2H NMR spectra of D‐IBU*, recorded at 0, 30 and 60°C under static conditions; b/Zoom onto the sharp component of the static 2H NMR spectra of D‐IBU*, showing the progressive shift in peak maximum position with temperature (see Figure S9 for the signature of melted ibuprofen). c/Fit of the experimental 2H static NMR data at 30°C, after symmetrisation of the 2H lineshape (experimental spectrum in black, and its fit in red, considering two contributions, as detailed in the main text). d/Variation with the temperature of the 2H NMR parameters of the broader quadrupolar component (blue spectrum in the fit), as measured from the 2H spectra recorded under static conditions for doubly labelled D‐IBU*, and from the 2H spectra recorded at 5 kHz MAS for deuterated ibuprofen (non‐17O labelled).

Table S1. Melting temperatures of ibuprofen and its isotopically enriched counterparts. The ranges reported here are the average over three independent measurements.

ACKNOWLEDGEMENTS

This project has received funding from the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation programme (grant agreement No 772204; 2017 ERC‐COG, MISOTOP project). NMR spectroscopic calculations were performed using HPC resources from GENCI‐IDRIS (Grant 097535). Powder X‐ray diffraction, mass spectrometry and solution NMR characterizations were performed with the support of the local Balard Plateforme d'Analyses et de Caractérisation (PAC Balard).

Chen C‐H, Goldberga I, Gaveau P, et al. Looking into the dynamics of molecular crystals of ibuprofen and terephthalic acid using 17O and 2H nuclear magnetic resonance analyses. Magn Reson Chem. 2021;59:975–990. 10.1002/mrc.5141

Chia‐Hsin Chen and Ieva Goldberga contributed equally to this work.

Funding information GENCI‐IDRIS, Grant/Award Number: 097535; Horizon 2020, Grant/Award Number: 772204

REFERENCES

  • 1. Gan Z., Hung I., Wang X., Paulino J., Wu G., Litvak I. M., Gor'kov P. L., Brey W. W., Lendi P., Schiano J. L., Bird M. D., Dixon I. R., Toth J., Boebinger G. S., Cross T. A., J. Magn. Reson. 2017, 284, 125. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2. Ishii Y., Wickramasinghe A., Matsuda I., Endo Y., Ishii Y., Nishiyama Y., Nemoto T., Kamihara T., J. Magn. Reson. 2018, 286, 99. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Penzel S., Oss A., Org M.‐L., Samoson A., Böckmann A., Ernst M., Meier B. H., J. Biomol. NMR 2019, 73(1), 19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Wijesekara A. V., Venkatesh A., Lampkin B. J., VanVeller B., Lubach J. W., Nagapudi K., Hung I., Gor'kov P. L., Gan Z., Rossini A. J., Chem‐Eur J. 2020, 26(35), 7881. [DOI] [PubMed] [Google Scholar]
  • 5. Zhang R., Mroue K. H., Ramamoorthy A., Acc. Chem. Res. 2017, 50(4), 1105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Sanders K. J., Pell A. J., Wegner S., Grey C. P., Pintacuda G., Chem. Phys. Lett. 2018, 697, 29. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Perras F. A., Venkatesh A., Hanrahan M. P., Goh T. W., Huang W., Rossini A. J., Pruski M., J. Magn. Reson. 2017, 276, 95. [DOI] [PubMed] [Google Scholar]
  • 8. O'Dell L. A., Ultra‐wideline Solid‐State NMR: Developments and Applications of the WCPMG Experiment, in Modern Magnetic Resonance, (Ed: Webb G. A.), Springer International Publishing, Cham: 2017, pp. 1–22. [Google Scholar]
  • 9. Gupta S., Tycko R., J. Biomol. NMR 2018, 70(2), 103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Wong V. W. C., Reid D. G., Chow W. Y., Rajan R., Green M., Brooks R. A., Duer M. J., J. Biomol. NMR 2015, 63(2), 119. [DOI] [PubMed] [Google Scholar]
  • 11. Métro T.‐X., Gervais C., Martinez A., Bonhomme C., Laurencin D., Angew. Chem. Int. Ed. Engl. 2017, 56(24), 6803. [DOI] [PubMed] [Google Scholar]
  • 12. Hope M. A., Zhang B., Zhu B., Halat D. M., MacManus‐Driscoll J. L., Grey C. P., Chem. Mater. 2020, 32, 7921. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Lilly Thankamony A. S., Wittmann J. J., Kaushik M., Corzilius B., Prog. Nucl. Magn. Reson. Spectrosc. 2017, 102‐103, 120. [DOI] [PubMed] [Google Scholar]
  • 14. Hooper R. W., Klein B. A., Michaelis V. K., Chem. Mater. 2020, 32(11), 4425. [Google Scholar]
  • 15. Leroy C., Bryce D. L., Prog. Nucl. Magn. Reson. Spectrosc. 2018, 109, 160. [DOI] [PubMed] [Google Scholar]
  • 16. Bryce D. L., Dalton Trans. 2019, 48(23), 8014. [DOI] [PubMed] [Google Scholar]
  • 17. Matlahov I., van der Wel P. C. A., Methods 2018, 148, 123. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Mandala V. S., Williams J. K., Hong M., Annu. Rev. Biophys. 2018, 47(1), 201. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Milić J. V., Im J.‐H., Kubicki D. J., Ummadisingu A., Seo J.‐Y., Li Y., Ruiz‐Preciado M. A., Dar M. I., Zakeeruddin S. M., Emsley L., Grätzel M., Adv. En. Mater. 2019, 9(20), 1900284. [Google Scholar]
  • 20. Dunstan M. T., Halat D. M., Tate M. L., Evans I. R., Grey C. P., Chem. Mater. 2019, 31(5), 1704. [Google Scholar]
  • 21. Xiang Y.‐X., Zheng G., Zhong G., Wang D., Fu R., Yang Y., Solid State Ion. 2018, 318, 19. [Google Scholar]
  • 22. Lu J., Hung I., Brinkmann A., Gan Z., Kong X., Wu G., Angew. Chem. Int. Ed. Engl. 2017, 56(22), 6166. [DOI] [PubMed] [Google Scholar]
  • 23. Wu G., Prog. Nucl. Magn. Reson. Spectrosc. 2019, 114‐115, 135. [DOI] [PubMed] [Google Scholar]
  • 24. Wu G., Hung I., Gan Z., Terskikh V., Kong X., J. Phys. Chem. A 2019, 123(38), 8243. [DOI] [PubMed] [Google Scholar]
  • 25. Catalano J., Murphy A., Yao Y., Zumbulyadis N., Centeno S. A., Dybowski C., Solid State Nucl. Magn. Reson. 2018, 89, 21. [DOI] [PubMed] [Google Scholar]
  • 26. Kolokolov D. I., Jobic H., Stepanov A. G., Guillerm V., Devic T., Serre C., Férey G., Angew. Chem. Int. Ed. Engl. 2010, 49(28), 4791. [DOI] [PubMed] [Google Scholar]
  • 27. Khudozhitkov A. E., Kolokolov D. I., Stepanov A. G., J. Phys. Chem. C 2018, 122(24), 12956. [Google Scholar]
  • 28. Jiang X., Duan H.‐B., Jellen M. J., Chen Y., Chung T. S., Liang Y., Garcia‐Garibay M. A., J. Am. Chem. Soc. 2019, 141(42), 16802. [DOI] [PubMed] [Google Scholar]
  • 29. Dai Y., Terskikh V., Brinkmann A., Wu G., Cryst. Growth Des. 2020, 20(11), 7484. [Google Scholar]
  • 30. Scheubel W., Zimmermann H., Haeberlen U., J. Magn. Reson. 1988, 80(3), 401. [Google Scholar]
  • 31. Kong X., Shan M., Terskikh V., Hung I., Gan Z., Wu G., J. Phys. Chem. B. 2013, 117(33), 9643. [DOI] [PubMed] [Google Scholar]
  • 32. Schmidt J., Sebastiani D., J. Chem. Phys. 2005, 123(7), 074501. [DOI] [PubMed] [Google Scholar]
  • 33. Friščić T., Mottillo C., Titi H. M., Angew. Chem. Int. Ed. Engl. 2019, 59, 1018. [DOI] [PubMed] [Google Scholar]
  • 34. Pyykkö P., Mol. Phys. 2018, 116(10), 1328. [Google Scholar]
  • 35. Bielecki A., Burum D. P., J. Magn. Reson., Ser. A 1995, 116(2), 215. [Google Scholar]
  • 36. Giannozzi P., Baroni S., Bonini N., Calandra M., Car R., Cavazzoni C., Ceresoli D., Chiarotti G. L., Cococcioni M., Dabo I., Dal Corso A., de Gironcoli S., Fabris S., Fratesi G., Gebauer R., Gerstmann U., Gougoussis C., Kokalj A., Lazzeri M., Martin‐Samos L., Marzari N., Mauri F., Mazzarello R., Paolini S., Pasquarello A., Paulatto L., Sbraccia C., Scandolo S., Sclauzero G., Seitsonen A. P., Smogunov A., Umari P., Wentzcovitch R. M., J. Phys. Condens. Matter 2009, 21(39), 395502. [DOI] [PubMed] [Google Scholar]
  • 37. Perdew J. P., Burke K., Ernzerhof M., Phys. Rev. Lett. 1996, 77(18), 3865. [DOI] [PubMed] [Google Scholar]
  • 38. Troullier N., Martins J. L., Phys. Rev. B 1991, 43(3), 1993. [DOI] [PubMed] [Google Scholar]
  • 39. Kleinman L., Bylander D. M., Phys. Rev. Lett. 1982, 48(20), 1425. [Google Scholar]
  • 40. Pickard C. J., Mauri F., Phys. Rev. B 2001, 63(24), 245101. [Google Scholar]
  • 41. Pyykkö P., Mol. Phys. 2008, 106(16–18), 1965. [Google Scholar]
  • 42. Hafner J., J. Comput. Chem. 2008, 29(13), 2044. [DOI] [PubMed] [Google Scholar]
  • 43. VandeVondele J., Krack M., Mohamed F., Parrinello M., Chassaing T., Hutter J., Comput. Phys. Commun. 2005, 167(2), 103. [Google Scholar]
  • 44. Grimme S., Antony J., Ehrlich S., Krieg H., J. Chem. Phys. 2010, 132(15), 154104. [DOI] [PubMed] [Google Scholar]
  • 45. Goedecker S., Teter M., Hutter J., Phys. Rev. B 1996, 54(3), 1703. [DOI] [PubMed] [Google Scholar]
  • 46. VandeVondele J., Hutter J., J. Chem. Phys. 2007, 127(11), 114105. [DOI] [PubMed] [Google Scholar]
  • 47. Špačková J., Fabra C., Mittelette S., Gaillard E., Chen C.‐H., Cazals G., Lebrun A., Sene S., Berthomieu D., Chen K., Gan Z., Gervais C., Métro T.‐X., Laurencin D., J. Am. Chem. Soc. 2020, 142(50), 21068. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48. Dudognon E., Correia N. T., Danède F., Descamps M., Pharm. Res. 2013, 30(1), 81. [DOI] [PubMed] [Google Scholar]
  • 49. Filsinger B., Zimmermann H., Haeberlen U., Mol. Phys. 1992, 76(1), 157. [Google Scholar]
  • 50. Carignani E., Borsacchi S., Geppi M., ChemPhysChem 2011, 12(5), 974. [DOI] [PubMed] [Google Scholar]
  • 51. Demkin A. G., Kolesov B. A., J. Phys. Chem. A 2019, 123(26), 5537. [DOI] [PubMed] [Google Scholar]
  • 52. Ostrowska K., Kropidłowska M., Katrusiak A., Cryst. Growth Des. 2015, 15(3), 1512. [Google Scholar]
  • 53. Idziak S., Mol. Phys. 1989, 68(6), 1335. [Google Scholar]
  • 54. Carignani E., Borsacchi S., Marini A., Mennucci B., Geppi M., J. Phys. Chem. C 2011, 115(50), 25023. [DOI] [PubMed] [Google Scholar]
  • 55. Geppi M., Guccione S., Mollica G., Pignatello R., Veracini C. A., Pharm. Res. 2005, 22(9), 1544. [DOI] [PubMed] [Google Scholar]
  • 56. Kolesov B. A., J. Phys. Chem. Solids 2020, 138, 109288. [Google Scholar]
  • 57. Goossens D. J., Chan E. J., Acta Crystallogr. Sect. B 2017, 73(1), 112. [Google Scholar]
  • 58. Beckham G. T., Peters B., Starbuck C., Variankaval N., Trout B. L., J. Am. Chem. Soc. 2007, 129(15), 4714. [DOI] [PubMed] [Google Scholar]
  • 59. Śledź M., Janczak J., Kubiak R., J. Mol. Struct. 2001, 595(1), 77. [Google Scholar]
  • 60. Davey R. J., Maginn S. J., Andrews S. J., Black S. N., Buckley A. M., Cottier D., Dempsey P., Plowman R., Rout J. E., Stanley D. R., Taylor A., J. Chem. Soc. Faraday Trans. 1994, 90(7), 1003. [Google Scholar]
  • 61. Meier B. H., Ernst R. R., J. Solid State Chem. 1986, 61(1), 126. [Google Scholar]
  • 62. Bailey M., Brown C. J., Acta Crystallogr. 1967, 22(3), 387. [Google Scholar]
  • 63. Pritchina E. A., Kolesov B. A., Spectroc. Acta A 2018, 202, 319. [DOI] [PubMed] [Google Scholar]
  • 64. McKinnon J. J., Fabbiani F. P. A., Spackman M. A., Cryst. Growth Des. 2007, 7(4), 755. [Google Scholar]
  • 65. Bauer S., Storek M., Gainaru C., Zimmermann H., Böhmer R., J. Phys. Chem. B. 2015, 119(15), 5087. [DOI] [PubMed] [Google Scholar]
  • 66. Poplett I. J. F., Smith J. A. S., J. Chem. Soc., Faraday Trans. 2 1981, 77(8), 1473. [Google Scholar]
  • 67. Dudognon E., Danède F., Descamps M., Correia N. T., Pharm. Res. 2008, 25(12), 2853. [DOI] [PubMed] [Google Scholar]
  • 68. Adrjanowicz K., Kaminski K., Dulski M., Wlodarczyk P., Bartkowiak G., Popenda L., Jurga S., Kujawski J., Kruk J., Bernard M. K., Paluch M., J. Chem. Phys. 2013, 139(11), 111103. [DOI] [PubMed] [Google Scholar]
  • 69. Tielens F., Folliet N., Bondaz L., Etemovic S., Babonneau F., Gervais C., Azaïs T., J. Phys. Chem. C 2017, 121(32), 17339. [Google Scholar]
  • 70. Pourpoint F., Trébosc J., Gauvin R. M., Wang Q., Lafon O., Deng F., Amoureux J.‐P., ChemPhysChem 2012, 13(16), 3605. [DOI] [PubMed] [Google Scholar]
  • 71. van Wüllen L., Koller H., Kalwei M., Phys. Chem. Chem. Phys. 2002, 4(9), 1665. [Google Scholar]
  • 72. Makrinich M., Sambol M., Goldbourt A., Phys. Chem. Chem. Phys. 2020, 22, 21022. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Figure S1. Powder X‐ray diffraction patterns before and after 17O‐labelling. These correspond to the reported « form (I) » of terephthalic acid (CCDC 1269122).

Figure S2. 13C and 1H solution NMR spectra before and after 17O labelling (DMSO‐d 6 , 600 MHz; solvent peaks are crossed out).

Figure S3. MS analyses of the non‐17O‐labelled precursor in comparison to the 17O‐enriched product.

Figure S4. Powder X‐ray diffraction patterns of the starting ibuprofen material, and the 17O and 17O‐2H enriched forms. These correspond to the polymorph shown on the left (also known as phase « I »).

Figure S5. 13C and 1H solution NMR spectra (DMSO‐d 6 , 600 MHz; solvent peaks marked with a star).

Figure S6 . MS analyses of the non‐17O‐labelled precursor in comparison to the 17O‐enriched product.

Figure S7. ATR‐IR analysis of the starting material, final product, and reaction intermediates, during the 18O‐labelling of ibuprofen by mechanochemistry. The dashed line shows that the C=O stretching frequency of the 18O‐enriched product is shifted to lower wavenumbers in comparison with the non‐labelled precursor.

Figure S8. ATR‐IR analysis of non‐labelled ibuprofen (IBU), and 17O‐labelled ibuprofen, before (IBU*) and after deuteration (D‐IBU*). The replacement of OH vibration modes by OD ones in D‐IBU* confirms the 2H isotopic enrichment (see, for example, O‐D stretching modes in the 1900–2400 cm−1 region highlighted in light green colour).

Figure S9. 2H (left) and 17O (right) NMR spectra of melted D‐IBU*, recorded at 80°C.

Figure S10. a/2H MAS NMR spectra recorded on highly crystalline D‐TA* at −30°C, with the magic angle properly set (top), and after a small mis‐adjustment of the angle (bottom). The splitting of the resonances (with a symmetric splitting over the sidebands) attests to the mis‐setting of the angle. The zoomed section of the central part of the spectrum (dashed grey box) is shown on the right.

Figure S11. Experimental 17O NMR spectra of D‐IBU*, recorded at 0°C and 60°C, under MAS conditions, together with their fits. Parameters used for the fits are reported in the table to the top‐right. The bottom‐right table recalls the parameters which had been reported previously for non‐deuterated ibuprofen.

Figure S12 . Spectra recorded on D‐TA* to compare the sensitivity in different probe configurations for a/2H NMR (acquisition conditions: B0 = 14.1 T, T = −30°C, 5 kHz MAS, solid echo with 1 rotor period echo delay, D1 = 0.5 s, NS = 6,000), and b/17O NMR (acquisition conditions: B0 = 14.1 T, 8 kHz MAS, DFS‐echo with 1 rotor period echo delay, D1 = 0.4 s, NS = 1,000) (It is worth noting that the 17O MAS NMR experiments shown here were performed with 1H decoupling in double resonance mode, and without 1H decoupling in triple resonance mode. However, it was verified that the lack of 1H decoupling in the latter case did not affect the overall signal intensity, and thus cannot explain the differences observed between double and triple resonance configurations.).

Figure S13 . 17O MAS NMR spectra of D‐TA* were recorded at different temperatures and their tentative fit considering the presence of 2 oxygen sites shown. The green and purple lines correspond to the fits, and the dashed orange line the sum of the two contributions.

Figure S14. a/Experimental 2H NMR spectra of D‐IBU*, recorded at 0, 30 and 60°C under static conditions; b/Zoom onto the sharp component of the static 2H NMR spectra of D‐IBU*, showing the progressive shift in peak maximum position with temperature (see Figure S9 for the signature of melted ibuprofen). c/Fit of the experimental 2H static NMR data at 30°C, after symmetrisation of the 2H lineshape (experimental spectrum in black, and its fit in red, considering two contributions, as detailed in the main text). d/Variation with the temperature of the 2H NMR parameters of the broader quadrupolar component (blue spectrum in the fit), as measured from the 2H spectra recorded under static conditions for doubly labelled D‐IBU*, and from the 2H spectra recorded at 5 kHz MAS for deuterated ibuprofen (non‐17O labelled).

Table S1. Melting temperatures of ibuprofen and its isotopically enriched counterparts. The ranges reported here are the average over three independent measurements.


Articles from Magnetic Resonance in Chemistry are provided here courtesy of Wiley

RESOURCES