Weak Intermolecular CH…N Hydrogen Bonding: Determination of 13 CH- 15 N Hydrogen-bond Mediated J couplings by Solid-state NMR Spectroscopy and First-principles Calculations

Weak hydrogen bonds are increasingly hypothesised to play key roles in a wide range of chemistry from catalysis to gelation to polymer structure. Here, 15 N/ 13 C spin-echo magic-angle spinning (MAS) solid-state NMR experiments are applied to “view” intermolecular CH…N hydrogen bonding in two selectively labelled organic compounds, 4-[ 15 N] cyano-4’-[ 13 C 2 ] ethynylbiphenyl ( 1 ) and [ 15 N 3 , 13 C 6 ]-2,4,6-triethynyl-1,3,5-triazine ( 2 ). The synthesis of 2- 15 N 3 , 13 C 6 is reported here for the first time via a Hz ( 2 )) are compared with density functional theory (DFT) gauge-including projector augmented wave (GIPAW) calculations, whereby species independent coupling values 2h K CN (29.0 ( 1 ), 27.9 ( 2 ) × 10 19 kg m − 2 s − 2 A − 2 ) quantitatively demonstrate the J couplings for these “weak” CH…N hydrogen bonds to be of a similar magnitude to those for conventionally observed NH…O hydrogen-bonding interactions in uracil ( 2h K NO : 28.1 and 36.8 × 10 19 kg m − 2 s − 2 A − 2 ). Moreover, the GIPAW calculations show a clear correlation between increasing 2h J CN (and 3h J CN ) coupling and reducing C(H)…N and H…N hydrogen bonding distances, with the Fermi Contact term accounting for at least 98% of the isotropic 2h J CN coupling. quantifying intermolecular of the 1 H chemical shift a probe of weak CH…A hydrogen using an approach that combines experimental solid-state magic-angle spinning (MAS) NMR with gauge-including projector augmented wave (GIPAW) density-functional theory (DFT)-based calculations, 24-27 significant (up to 2 molecule to changes in the 1 H chemical shift due to intermolecular CH…A hydrogen hybridized CH


Introduction
Hydrogen bonds are known to play a crucial role in directing the assembly of organic molecules into supramolecular structures and feature highly throughout structural chemistry and biology. 1,2 There is increasing interest in so-called 'weak' CH…A (where A = N or O) hydrogen bonds [3][4][5] whereby key roles in, for example, catalysis through the stabilisation of transition states, [6][7][8][9][10][11][12] or in promoting supramolecular gelation 13 or helical folding of peptides 14 , or in the structure of cellulose polymorphs, 15,16 a high-mobility donor-acceptor copolymer 17 or poly(-caprolactone) 18 have been proposed. In this context, note that Jeffrey makes a distinction (see Tables 2.1 and 2.2 in Ref. 19 ) between moderate (including NH…N and NH…O) and weak (with a CH donor) hydrogen bonds, with a D…A hydrogen bonding distance cut-off between the moderate and weak categories of 3.2 Å. While CH…A proximities can be identified in crystal structures obtained by diffraction, this approach alone cannot conclusively answer the question as to whether a close contact between a potential donor and acceptor is a bonding interaction that contributes to stabilizing the structure or is rather the result of other packing constraints. 4,[20][21][22][23] NMR probes the local structure around a nucleus and is thus a powerful method for identifying and quantifying intermolecular interactions. Previous papers have demonstrated the sensitivity of the 1 H chemical shift as a probe of weak CH…A hydrogen bonding. Specifically, Brown and coworkers have demonstrated, using an approach that combines experimental solid-state magicangle spinning (MAS) NMR with gauge-including projector augmented wave (GIPAW) densityfunctional theory (DFT)-based calculations, [24][25][26][27] significant (up to 2 ppm) molecule to crystal changes in the 1 H chemical shift due to intermolecular CH…A hydrogen bonding for sp 3 , sp 2 and sp hybridized CH donors in maltose anomers, uracil and 4-cyano-4'-ethynylbiphenyl, respectively. 28,29 For maltose, the largest 1 H chemical shift changes are associated with a short H…O distance (< 2.7 Å) and a CHO bond angle greater than 130°. In uracil, the chemical shift changes due to CH…O hydrogen bonding of 2.0 and 2.2 ppm are ~40% of those due to conventional NH…O hydrogen bonding (5.1 and 5.4 ppm). In solution-state NMR, Solà et al. have observed a shift to low ppm of 2.1 ppm for a CH 1 H resonance between the syn and anti conformers of an organometallic complex, with intramolecular CH…O hydrogen bonding only being possible for the syn conformer. 30 Hydrogen-bond mediated J couplings constitute a direct means of identifying a specific pair of hydrogen-bonded nuclei as well as quantifying the strength of a particular hydrogen bond. 31 Following their observation by solution-state NMR, 32 the adoption of specific NH…N intraand intermolecular hydrogen bonding arrangements has been demonstrated in the solid state by the observation in two-dimensional refocused INADEQUATE 33,34 MAS NMR spectra of cross peaks corresponding to 2h JNN couplings. [35][36][37] Moreover, the measurement of J couplings using a spinecho MAS NMR experiment 38 allows the quantitative determination of these 2h JNN couplings, and hence the strength of the corresponding NH…N hydrogen bonds. 37,39,40 It has also been shown that 2h JNO couplings due to intermolecular NH…O hydrogen bonds in uracil can be measured using a heteronuclear 15 N- 17 O MAS spin-echo experiment. 41 Schanda et al. have further detected small 3h JNC couplings (average value 0.5 Hz) across C=O…HN hydrogen bonds in the protein ubiquitin using a three-dimensional HNCO MAS experiment. 42 In solution-state NMR, very small 3h JCC couplings of 0.2 to 0.3 Hz have been measured across CH…O=C hydrogen bonds in -sheet regions of a small protein. 43 Moreover, cross peaks due to J couplings for CH… interactions have been observed between methyl and aromatic side chains or carbonyl groups in proteins by Plevin et al. 44 and Perras et al, 45 where DFT calculation predicts J coupling magnitudes of ~0.1 Hz. This paper investigates weak intermolecular CH…N hydrogen bonding in the solid-state structures adopted by 4-cyano-4'-ethynylbiphenyl, 1, and 2,4,6-triethynyl-1,3,5-triazine, 2, for which diffraction structures are available in the Cambridge Structural Database (CSD). 46,47 In both cases, the hydrogen-bond donor is a sp hybridized alkyne moiety, while the acceptor is a nitrogen atom in a nitrile group in 1 or in a triazine ring in 2. In order to focus on the intermolecular CH…N hydrogen bonds, the molecules 1 and 2 were synthesized with selective isotopic labelling: specifically, both carbons in the alkyne group were 13 C labelled, while all nitrogens were 15 N labelled. Note that the synthesis of 2-15 N 3 , 13 C 6 is reported here for the first time. Scientists are striving for the ability to view directly hydrogen bonds: 48,49 in this work, we apply a combined experimental and calculation approach to observe and measure, for what we believe to be the first time in the solid state, J couplings across weak intermolecular CH…A hydrogen bonds.  Solid-state NMR spectroscopy Experiments were performed on a Bruker AVANCE III NMR spectrometer, operating at 1 H, 13 C and 15 N Larmor frequencies of 500, 125 and 50 MHz, respectively, using a Bruker 3.2 mm triple-resonance MAS probe and a spinning frequency of 10 kHz. 15 N/ 13 C heteronuclear spin-echo experiments were performed using the pulse sequences shown in Figure 1. An 8-step phase cycle was used: φ1 = -x, +x; φ2 = +y, +y, -x, -x, -y, -y, +x, +x; receiver = +x, -x, -x, +x, +x, -x, -x, +x to select p = 1 and 2 on the 1 H π/2 (φ1) and 15 N π pulses (φ2), respectively, where p is the coherence order. 15   For non-selective heteronuclear spin-echo experiments,  pulses on 13 C (top option in Figure 1) were applied with the 13 C carrier frequency in the centre of the spectrum, whereas for selective 13 C inversion, a Gaussian shaped pulse (bell shape, bottom option in Figure 1) of length, P, (2) 4.0 ms or (10) 2.0 ms was used, whereby the 13 C carrier frequency was centred on the carbon resonance of interest. Thus, for selective heteronuclear spin-echo experiments, the initial  = 0 case corresponds to P. Following the procedure described in ref 52 , the data obtained using selective 13 C pulses was fit to equations (see below) containing a shifted echo term, ( − sh), to account for the evolution which occurs during the length of the selective pulse, where sh = (2) 0.96 ms or (10) 0.48 ms. 15 N homonuclear spin-echo experiments were performed by omitting the  pulse on the non-observe, 13 C channel. Experiments were interleaved such that each set of homonuclear and heteronuclear spectra were recorded consecutively for each evolution interval, , and all echo delays were rotor synchronised. (1) 768, (2) 320 and (10) 32 transients were co-added for each  increment for (1, 2) (16 s recycle delay) and 10 (12 s recycle delay), respectively.  53 ) is liquid ammonia at −50 C, it is necessary to add 379.5 ppm to the given values. 54 The 13 C chemical shifts were referenced to tetramethylsilane (TMS, Si(CH3)4) by using powdered L-alanine as an external secondary reference (δiso = 177.9, 51.0, 20.5 ppm)this is equivalent to using adamantane at 38.5 ppm that corresponds to TMS at 0 ppm. 55 S HET () and S HOM () integrals were taken over the respective resolved peaks after Fourier transformation with respect to t 2 and quotients, S Q () = S HET ()/S HOM (), determined. S Q () was normalized such that S Q ( = 0) = 1.00. Uncertainties, HET and HOM, were estimated for S HET and S HOM data points, respectively, based on the signal to noise ratios observed for heteronuclear 15 N/ 13 C spin-echo spectra obtained at  = 0 and uncertainties,   Q , in S Q were calculated using eq 1.

Methods
Errors on fitted parameters were determined using the covariance method as described in Ref 40 .
The analysis assumes that there is no deviation from the magic angle that would lead to a residual dipolar coupling. 56 In this respect, note that the dipolar coupling constant corresponding to a 13 C and 15 N spin pair at a distance of 3.25 Å is less than 100 Hz: as shown in section S6 of Ref. 56 , a deviation of more than 2° from the magic angle would be required to see a noticeable change in the spin-echo evolution for such a small dipolar coupling. This work used the carboxylate 13 C linewidth of L-alanine in a 13 C CP MAS experiment to check the setting of the magic angle, which, as discussed in section S5 of Ref. 56 , ensures a setting to within a small fraction of a degree from the magic angle. The geometry optimized crystal structure of 1 (all atoms relaxed) as described in ref 29 , as generated starting from the crystal structure with CSD reference code and number JOQSEN and 112101, 46 was used for J-coupling calculations. For 2, a geometry optimization was performed upon the crystal structure with CSD reference code and number HULSEM and 193181. 47 The calculations used Ultrasoft pseudopotentials, 57 planewaves up to a maximum energy of 600 eV and a Brillouin Zone sampling of 0.2 × 2 Å −1 . An initial partial geometry optimisation was performed allowing only the hydrogen atoms to move. However, the forces on the carbon atoms were as large as 4.5 eV/ Å, thus, a full optimisation was carried out allowing all atoms to move. Note that while the published crystal structure of 2 has Z' = 1, symmetry was not imposed during the DFT calculations, and the resulting geometry-optimised structure (allowing all atoms to move) has Z' = 2 (see Supporting Information). However, note that the differences between the two molecules are very small, for example there are no differences (to two decimal places for distances in Å and to one decimal place for the angles in degrees) in the intermolecular CH…N hydrogen bonding distances discussed below.
For 10, geometry optimization was performed upon the crystal structure with CSD reference code HISTCM12. 61 A partial geometry optimisation (only hydrogen atom positions) was performed, since the forces remaining on the non-hydrogen atoms after partial optimisation were all less than

Results and Discussion
The structure of this paper is to describe experimental results first followed by DFT GIPAW calculations. Within the experimental section, there is first a description of the synthesis of the selectively labelled compounds followed by a description of the hydrogen bonding in the crystal structures, before the experimental MAS NMR data is presented.

Synthesis
The synthesis of [ 13 C≡ 13 CH, 15    geometry optimization), namely the C(H)…N and H…N distances and the CHN angle, for the four distinct molecules in 1 (see also Figure 3 of Ref. 29 ). The average C(H)…N and H…N distances are 3.28 Å and 2.21 Å, respectively, with an average CHN angle of 173º.
For 2 (CSD reference code: HULSEM 47 ), each molecule has three hydrogen-bonding donor CH groups and three hydrogen-bonding acceptor nitrogens, leading to a hexagonal structure as shown in Figure 2 (bottom). Note that there is a single mirror plane through one of the ring nitrogen atoms (N A in Figure 2), such that there are two distinct ring nitrogen atoms (N B and N C are equivalent in Figure 2) in the asymmetric unit (and hence, also, two distinct ring carbon atoms and two distinct alkyne groups). Table 1   For 2, one separate resonance is observed in Figure 3c for the 15 N spectrum, while two separate resonances in the 13 C spectrum (Figure 3) are observed that correspond to the two alkyne atoms, labelled C2 and C3 in Figure 2. The CP efficiency is different for the protonated C3 CH moiety as compared to the non-protonated C2 carbon, hence explaining the different relative intensity in Figure 3d. As noted above, the mirror symmetry in the CSD crystal structure of 2 (see the thick vertical yellow line in Figure 2) means that two distinct nitrogen resonances as well as two distinct C2 and two distinct C3 resonances would be expected in Figure 3c and 3d, respectively. This may be the origin of the unusual lineshapes in these spectra. However, for such labelled compounds, unusual lineshapes (such as those observed in Figure 3d) can also arise from the cross correlation of chemical shift anisotropy and dipolar couplings as well as J couplings, as noted previously for cis-azobenzene-15 N2 dioxide 68 and fully 13 C-labelled L-alanine. 69 For comparison, 13 C CP MAS NMR spectra for samples 1 and 2 at natural isotopic abundance are presented in Figure S3 in the Supporting Information. or to C3 to be observed. However, there is again still the complication that, for the CSD structure, two distinct nitrogen resonances (labelled NA and NB in Figure 2) as well as two distinct C2 and two distinct C3 resonances are expected. These are not resolved in Figure 3c and 3d, such that there will be a superposition of the (HOM and HET) spin-echo curves for the distinct sites.
Using the method presented in Ref. 41 Note in eq 3 and 4 that a time-shift, ( − sh), takes into account the length of the selective pulse. 52 It is evident that good fits are obtained in all three cases, with the fit parameters being listed in Table 2 (see also the listing of correlation coefficients in Table S1 in the Supporting Information).
As noted above, the fit functions in eq (2) to (4) include an exponential term, exp(−/ T2'), to take into account the difference in dephasing time for a homonuclear and heteronuclear spin echo.
However, for 1, this term was set equal to one (otherwise, correlation coefficients with magnitude above 0.9 are observed), while the large errors bars on the fitted T2' parameters for 2 are indicative of large correlation coefficients, between 0.8 and 0.9 (see Table S1 in the Supporting Information), in all cases, showing that the small difference in dephasing time for homonuclear and heteronuclear spin echo cannot be reliably determined for these spin-echo curves.
For 1, a zero crossing is observed for SQ() in Figure 4b, allowing the reliable determination of the largest 13 C-15 N heteronuclear J coupling. As shown in Figure S4 in the Supporting Information, a better fit is obtained when an additional modulation to a second J coupling is included, i.e., the best fit of SQ() in Figure 4b is for two J couplings of 4.7 and 2.9 Hz. The two SQ() curves in Figure 4d corresponding to selective spin echoes for the C2 (black) and C3 (grey) resonances of 2 are best fit to a cos 2 ( J ), eq 3, and a single cos( J ), eq 4, modulation, with fitted J couplings of 4.7 and 4.1 Hz, respectively. The change in shape of the SQ() curve for the case of modulation due to two equal J couplings as compared to a single J coupling is evident in Figure 4d (see also Figure S4 in the Supporting Information).   Table 1). e Fit of the SQ() data in Figure 4b to eq 2. f ( ) was set to 1 (otherwise correlation coefficients greater than 0.9 were obtained). g The calculated JCN values correspond to the two distinct nitrogen atoms, see Table 1 and S5. h Fit of the SQ() data in Figure 4d (solid and grey black line to eq 3 and 4, respectively), where sh = 0.96 ms.
GIPAW J coupling calculations. The calculation using density functional theory of NMR J couplings represents a valuable resource that complements experimental measurement. [80][81][82][83] This section compares the experimentally determined J couplings to those calculated using the GIPAW method 58,59 for the geometry-optimised crystal structures in the CSD (reference codes: JOQSEN 46 for 1 and HULSEM 47 for 2). The application of GIPAW calculation to complement experimental solid-state NMR measurements has been previously employed for 2h JNN and 2h JNO hydrogen-bond mediated couplings, 41,84,85 as well as 29 Si-31 P, 11 B-11 B, 31 P-77 Se and 31 P-31 P J couplings in inorganic materials. [86][87][88][89] Before comparing the experimental and calculated J couplings in Table 2, it is important to consider powder X-ray diffraction (PXRD) analysis for the two solid-state samples of 1 and 2 for which experimental solid-state NMR results are presented in this paper. Figure S1 and S2 (for 1 and 2, respectively) in the Supporting Information compare the experimental PXRD patterns with predicted patterns for the CSD structures (reference codes: JOQSEN 46 for 1 and HULSEM 47 for 2). For 1, there is excellent agreement. However, for 2, while there are experimental peaks in the PXRD pattern corresponding to those predicted for the HULSEM 47 CSD structure, there are a number of peaks in the experimental PXRD pattern in Figure S2 that are not in the predicted pattern for the CSD structure. Since solution-state NMR has confirmed the synthesis of the correct molecular structures (see experimental details above), the PXRD pattern in Figure S2 shows that our sample of 2 is likely a mixture of a new solid-state form (or forms) and the CSD solid-state structure. Efforts are ongoing to identify the new solid-state form(s) of 2 by structure determination from this PXRD pattern. Table 1 above lists the GIPAW calculated 2h JCN and 3h JCN couplings corresponding to the distinct intermolecular CH…N hydrogen bonds for the geometry-optimised crystal structures in the CSD, reference codes: JOQSEN 46 for 1 and HULSEM 47 for 2. For 1, the experimental SQ() data in Figure 4b corresponds to the superposition of spin-echo modulation for the four distinct, but not resolved, 15 N resonances (N1 to N4 in Table 1). For each 15 N resonance, Figure 1 and Table 1 both suggest that there should be two cos( J ) modulations corresponding to the 2h JCN and 3h JCN couplings, noting that there is no resolution of the resonances in the 13 C spectrum in Figure 3b for the two chemically distinct carbons in the alkyne bond. As shown in section S3 in the Supporting Information, the consequence of the superposition of the spin-echo modulation for the four distinct 15 N resonances is that an average 2h JCN and an average 3h JCN coupling is extracted from fitting the experimental SQ() data in Figure 4b to eq 2. While there is not perfect agreement between experiment and calculation in Table 2 in that the calculated values are larger than the experimental ones by approximately a factor of two, the calculations confirm that it is to be expected that experiment observes J couplings across the weak intermolecular CH…N hydrogen bonds.
Moreover, it is observed in both experiment and calculation that, in 1, the J coupling across the intermolecular hydrogen bond is bigger for the nearer carbon ( 2h JCN > 3h JCN). Note that the spinecho experiments cannot determine the sign of the J couplings.
For 2, the GIPAW calculation for the CSD structure HULSEM 47 reveals that there are intramolecular 2 JCN and 3 JCN couplings between the ring 15 N nuclei and the labelled 13 C nuclei of the alkyne groups that are of similar magnitude to the intermolecular 2h JCN and 3h JCN couplings, respectively (see Tables 1 and 3). For 2, distinct resonances are resolved in the 13 C spectrum for C2 and C3 (see Figure 3d), thus separate SQ() data is presented in Figure 4d corresponding to selective spin echoes applied to the C2 and C3 resonances. The GIPAW calculation for the CSD structure HULSEM 47 predicts that, for the C2 resonance, the dominant spin-echo modulation is cos 2 ( J ) for the two approximately equal intramolecular 2 JCN couplings (see Table 3 and Figure   2), while, for the C3 resonance, the dominant spin-echo modulation is cos( J ) for the intermolecular 2h JCN couplings (see Table 1 and Figure 2). There are also three-bond intramolecular 3 JCN (see Table 3) and intermolecular 3h JCN couplings (see Table 1) that are calculated to be significantly smaller than the 2 JCN and 2h JCN couplings.  Tables S4,   S5 and S9, respectively, in the Supporting Information. b These J couplings are not observed in this work because of the selective 13 C labelling of only the alkyne carbons. c In this case, the Para term is the largest contribution to the total isotropic J coupling.
As noted above, the SQ() data in Figure 4d for C2 (black) and C3 (grey) is fit (see Table 2 For 2, while there is again not perfect agreement between experiment (and noting the above discussion of the different solid-state forms of 2) and calculation in Table 1  Hydrogen bond mediated 2h JCN couplings can be visualised using coupling deformation density (CDD) maps of Malkina and Malkin, 97 as shown in Figure 5 for 2 (our implementation of CCD maps is described, for through-space Se-P J couplings, in Ref. 88 , see section S5). As for a 2h JCN coupling in a model NCH…NH3 interaction (see Figure 5 in Ref. 97 ), the directionality of the hydrogen bond mediated 2h JCN couplings in 2 is evident in Figure 5. The Supporting Information also compares experimentally determined JCN couplings (see section S5) to the nitrogen sites for [U-13 C, 15 N] L-histidine.HCl.H2O to those calculated using the GIPAW approach (see Table S9). Reasonable agreement, noting that the calculated one-bond J couplings are smaller than in experiment, is observed (see Table S10), noting also the experimental 13 Table S11 in the SI for conversion factors). b 4-cyano-4'-ethynylbiphenyl, 1 (this work, see Tables 1 and 2), where an average is taken over Z'= 4 molecules. c 2,4,6-triethynyl-1,3,5-triazine, 2 (this work, see Tables 1 and 2 Table 4 highlights the greater discrepancy with respect to solid-state NMR experiment for the GIPAW calculated 2h JCN couplings in this work as compared to previous reports of GIPAW calculated 2h JNN and 2h JNO couplings. The only other case where a similar approximately factor of two discrepancy between experiment and GIPAW calculation has been noted is for through-space intermolecular JPP couplings in two organochalcogen systems. 89 By comparison, good agreement between GIPAW calculation and experiment has been observed for similar through-space JPSe couplings in peri-substituted napthalenes 88 as well as for 2 JPSi couplings of P-O-Si connectivities in silicophospates and calcium phosphates. 86

Conclusions
The results presented in this work demonstrate that, using ( 15 N/ 13 C) heteronuclear spin-echo NMR experiments, we are able to measure the small intermolecular 2h JCN couplings that occur in 1 and 2, whereby the hydrogen-bond donor is a sp hybridized alkyne moiety, while the acceptor is a nitrogen atom in a nitrile group in 1 or in a triazine ring in 2. We believe this to be the first solid-state determination of NMR J (scalar) couplings across 'weak' CH···N hydrogen bonds. As well as adding a new category to the range of observed hydrogen-bond mediated J couplings, see for example Figure 7 of Ref. 98 , what is noteworthy is that the magnitude of the experimentally fitted intermolecular 2h JCN couplings (4.7 and 4.1 Hz for 1 and 2, respectively) is comparable to that for extensively studied 2h JNN couplings for NH…N hydrogen bonds adopted by nucleobases (between 6 and 11 Hz). As such, these are an order of magnitude larger than measurements by solution-state NMR of 3h JCC couplings of 0.2 to 0.3 Hz for CH…O=C hydrogen bonds in -sheet regions of a small protein 43 as well as for J couplings due to CH… interactions that have been observed between methyl and aromatic side chains or carbonyl groups in proteins by Plevin et al. 44 and Perras et al, 45 where DFT calculation predicts J coupling magnitudes of ~0.1 Hz. We further note that, following previous ab initio predictions, 99,100 JCC couplings between 0.2 and 0.5 Hz have recently been measured for through-space van der Waals interactions between aliphatic side groups of the GB3 protein. 101 In this work, J couplings have been calculated using the GIPAW method. 58,59,84 By comparison with previously presented calculations for conventional NH···N, NH···O hydrogen bonds, 41,84,85 it is observed that the reduced spin coupling constants (that are independent of the magnetogyric ratio of the coupled nuclear spins) for the so-called CH…N weak hydrogen bonds, 2h KCN, are of similar magnitude to those for conventional (NH…O and NH…N) intermolecular hydrogen bonds, 2h KNO and 2h KNN. As for 2h JNN couplings for NH…N hydrogen bonds adopted by nucleobases, [90][91][92][93] for both 1 and 2, the GIPAW calculations show that a larger 2h JCN (and also 3h JCN) coupling is correlated with a shorter C…N (and H…N) intermolecular hydrogen bonding distance. We hence argue that experimentally accessible 2h JCN couplings are well suited for "viewing" weak CH…N hydrogen-bonding interactions, noting the increasing recognition that close CH···N proximities can correspond to structure-determining interactions, playing important chemical roles, for example in the stabilisation of transition states in catalysis. 7,12 Noting that, as early as 1963, it has been shown that the formation of C-H…N bonds upon adding pyridine causes the C-H stretching frequency in the IR spectrum of 1,3,5-trichlorobenzene to be lowered by 35