BINOL as a Chiral Solvating Agent for Sulfiniminoboronic Acids

1H NMR spectroscopic studies using BINOL as a chiral solvating agent (CSA) for a scalemic sulfiniminoboronic acid (SIBA) have revealed concentration- and enantiopurity-dependent variations in the chemical shifts of diagnostic imine protons used to determine enantiopurity levels. 11B/15N NMR spectroscopic studies and X-ray structural investigations revealed that unlike other iminoboronate species, BINOL–SIBA assemblies do not contain N–B coordination bonds, with 1H NMR NOESY experiments indicating that intermolecular H-bonding networks between BINOL and the SIBA analyte are responsible for these variations. These effects can lead to diastereomeric signal overlap at certain er values that could potentially lead to enantiopurity/configuration misassignments. Consequently, it is recommended that hydrogen-bonding-CSA-based 1H NMR protocols should be repeated using both CSA enantiomers to ensure that any concentration- and/or er-dependent variations in diagnostic chemical shifts are accounted for when determining the enantiopurity of a scalemic analyte.

−12 In a representative protocol, treatment of a scalemic amine (e.g., α-methylbenzylamine 1) with achiral 2-formylphenyl boronic acid (2-FPBA) and an enantiopure diol (e.g., BINOL) produces a mixture of diastereomeric iminoboronate esters (IBEs, e.g., 2a and 2b), whose diastereomeric ratio (dr) can then be determined by integration of baseline-resolved 1 H NMR imine resonances.Since no kinetic resolution occurs in this derivatization process, these dr values are an accurate reflection of the enantiomeric ratio (er) of the parent scalemic amine analyte (Figure 1a). 1,13,14−18 We recently reported a modified version of this threecomponent CDA method for determining the enantiopurities of S-chiral sulfinamides using 1 H/ 19 F NMR spectroscopy. 19In this case, high-yielding formation of SulfinIminoBoronate Esters (SIBEs) required development of a modified two-step "one-pot" protocol to overcome the lower nucleophilicity of the sulfinamide amino group.This new method involved prior reaction of a scalemic sulfinamide (e.g., Ellman's sulfinamide 3) with 2-FPBA in CDCl 3 to generate a scalemic SulfinIminoBoronic Acid (SIBA, e.g., 4) that was then reacted with enantiopure pinanediol to afford a mixture of SIBE diastereomers (e.g., 5a and 5b) (Figure 1b).These SIBE diastereomers display wellresolved 1 H NMR imine resonances, the integral ratios of which could be used to accurately determine dr values.Substituting 3fluoro-2-FPBA for 2-FPBA in this complexation reaction allows the use of both 1 H and 19 F NMR spectroscopic analysis to accurately determine SIBE dr values using two independent measurements (Figure 1b). 19ptimization studies aimed at identifying the best chiral reporter diol to derivatize (R S )-SIBA 4 produced pairs of SIBE diastereomers that generally displayed well-resolved imine peaks in their 1 H NMR spectra (Δδ H = 0.010−0.085ppm, Table S1).Pinanediol gave the best result, producing tert-butyl SIBE diastereomers 5a and 5b that displayed sharp 1 H NMR imine peaks with a large Δδ H value of 0.085 ppm.This led to pinanediol being chosen as the optimal chiral diol to derivatize other chiral sulfinamide analytes, whose SIBE diastereomers all displayed well-resolved 1 H NMR imine peaks.(Table S2). 19owever, derivatization of SIBA (R S )-4 with (R)-BINOL gave Figure 1.(a) Three-component complexation of scalemic amine (α-R)-1 (90:10 er), 2-FPBA and enantiopure (R)-BINOL produces diastereomeric IBEs (α-R,R)-2a and (α-S,R)-2b, whose drs are determined by 1 H NMR spectroscopic analysis. 13(b) Stepwise three-component complexation of scalemic sulfinamide (R S )-3 (90:10 er), 2-FPBA, and enantiopure pinanediol produces diastereomeric SIBEs (R S ,R)-5a and (S S ,R)-5b, whose drs can be measured by 1 H/ 19 F NMR spectroscopic analysis. 19 inferior results, producing 1 H NMR spectra exhibiting broader partially overlapped imine resonances.This result was surprising since previous three-component complexation reactions of (rac)-amines, 2-FPBA, and BINOL had produced diastereomeric IBEs displaying well-resolved 1 H NMR imine resonances (e.g., Δδ H = 0.17 ppm for 2a/2b). 13hese anomalous results prompted us to investigate the SIBA-BINOL complexation reaction further, with this study now reporting that reaction of SIBA 4 with BINOL does not in fact produce SIBE complexes like other diols.Instead, conformationally demanding BINOL (restricted rotation around biaryl bond) acts as a chiral solvating agent (CSA) to generate hydrogen-bonded BINOL−SIBA assemblies that display less well-resolved 1 H NMR imine resonances (Figure 1c).Furthermore, investigations into the use of BINOL as a CSA for determining the enantiopurities of scalemic SIBA 4 samples revealed that intermolecular hydrogen-bonding effects produce concentration-and er-dependent variations in the diagnostic 1 H NMR imine chemical shift values used to determine er values.These chemical shift variations can result in partially (or fully) overlapped 1 H NMR imine resonances at certain analyte concentration and er levels that prevent er from being determined.This signal overlap issue can be resolved by repeating the process the opposite enantiomer of BINOL for derivatization to produce a new 1 H NMR spectrum that displays fully resolved imine resonances.
The concentration-and er-dependent variations in diagnostic 1 H NMR chemical shifts described in this study could potentially occur in other hydrogen-bonding-CSA-based 1 H NMR protocols used to determine er values of other types of analyte.Therefore, it is recommended that duplicate 1 H NMR spectra of scalemic analytes complexed to both CSA enantiomers should be acquired to determine whether erdependent variations in diagnostic 1 H NMR chemical shift values need to be accounted for when determining er values.
■ EXPERIMENTAL SECTION General Experimental Details.Reagents and solvents were obtained from commercial suppliers and used without further purification.Reactions were performed without air exclusion or drying, at room temperature, and with magnetic stirring, unless otherwise stated.Anhydrous MgSO 4 or Na 2 SO 4 was used as a drying agent for organic solvents.Thin-layer chromatography (TLC) was carried out on Macherey−Nagel aluminum-backed plates that were precoated with silica.Compounds were visualized by either quenching of ultraviolet (UV) fluorescence at 254 nm or by dip-staining (KMnO 4 , PMA, curcumin, 20 I 2 ) followed by gentle heating.Purification by flash column chromatography was performed using high-purity-grade silica gel (60 Å pore size, 40−75 μm particle size).PE refers to petroleum ether 40−60 °C.Capillary melting points were determined using a Stuart digital SMP10 melting point apparatus and are reported uncorrected to the nearest °C.Optical rotations were measured using an Optical Activity Ltd.AA-10 Series Automatic Polarimeter, with a path length of 1 dm and with concentration (c) quoted in g/100 mL.Nuclear Magnetic Resonance (NMR) spectroscopy experiments were performed in deuterated solvents at 298 K (unless otherwise stated) on either a Bruker 500 MHz spectrometer or an Agilent ProPulse 500 MHz spectrometer. 1 H, 13 C, 11 B, and 19 F NMR chemical shifts (δ) are quoted in parts per million (ppm) and are referenced to either the residual solvent peak or tetramethylsilane (TMS) when possible. 21 11B NMR spectra were referenced directly, using the lock signal, to external BF 3 .Et 2 O (0 ppm).Coupling constants (J) are given in Hz.In those cases where 13 C signals could not be observed by one-dimensional (1D) NMR spectroscopy due to low solubility, adjacent quadrupolar nuclei, or lack of adjacent 1 H nuclei, then chemical shifts were measured indirectly from two-dimensional (2D) 1 H− 13 C HMBC experiments. 19 15N NMR chemical shifts were measured indirectly from 2D 1 H− 15 N HMBC spectra using a Bruker Avance 500 MHz spectrometer. 22 15N NMR experiments were carried out at 50 mM sample concentrations. 15N NMR spectroscopy was carried out in CDCl 3 containing 50 mM nitromethane internal standard.The MeNO 2 15 N resonance was used as the reference for 15 N chemical shifts, setting δ N (MeNO 2 ) = 0.00 ppm. 23nfrared (IR) spectra were recorded using a PerkinElmer Spectrum 100 Fourier-transform infrared (FTIR) spectrometer fitted with a Universal ATR FTIR accessory, with samples run neat and selected absorbances quoted as ν in cm −1 .Highresolution mass spectrometry (HRMS) results were acquired on an externally calibrated Bruker Daltonics maXis HD ultrahighresolution-time of flight (UHR-TOF) mass spectrometer coupled to an electrospray source (ESI-TOF) or an Agilent QTOF 6545 with Jetstream ESI.In most cases, molecular ions were detected either in positive mode as their protonated, sodiated, or ammonium adduct forms, or in negative mode as deprotonated or acetate adduct species.
Additional details of materials, reagents, and instrumentation are provided in the Supporting Information.Additionally, full synthetic procedures for the preparation of all compounds reported herein, along with full characterization data including NMR spectra, can also be found in the Supporting Information.
Scalemic and racemic samples of Ellman's sulfinamide 3 were prepared by combining different amounts of enantiopure solutions of (R)-and (S)-sulfinamide 3 in CDCl 3 .Samples required for concentration screening experiments were prepared directly from 100 mM solutions by dilution, as required. 1H− 15 N HMBC spectra were acquired by using 50 mM solutions of the desired compound in CDCl 3 containing 50 mM MeNO 2 as an internal standard.
These (R)-BINOL concentration-dependent increases in imine Δδ H values were initially unexpected since the 1 H NMR spectra of other SIBA-diol complexes did not display concentration-dependent imine chemical shift variations. 1,19his led us to conclude that treatment of SIBA 4 with BINOL does not produce IBEs like with other diols.Instead, a change in reaction manifold occurs whereby BINOL serves as a hydrogenbonding CSA to differentiate the SIBA enantiomers (Figure 1c).Further evidence for BINOL acting as a hydrogen-bonding CSA for 4 is also evident from the gradual downfield shift of the BINOL−OH resonance as the concentration of BINOL increases, and the broader imine resonances produced in the 1 H NMR spectra of these BINOL systems compared to typical SIBE/IBE CDA assemblies, also typical of a hydrogen-bonding system.
−32 In order to confirm that BINOL was functioning as a CSA in our case, we next carried out dilution experiments on solutions of (rac)-SIBA 4 and 1 equiv of (R)-BINOL in CDCl 3 , which revealed significant variation in the chemical shifts of the diagnostic imine as the sample was diluted (Figure 2b and Table S4). 33,34 1H NMR spectroscopic analysis at concentrations decreasing from 100 to 50 mM gave partially resolved imine peaks centered at δ H 9.056 and 9.037 ppm, respectively (Figure 2b, Entries 1 and 2), while samples recorded at concentrations of ≤25 mM displayed only a single unresolved imine singlet at δ H ≤ 8.973 ppm (Figure 2b, Entries 3−7).Corresponding upfield chemical shift drift of the imine peaks of enantiopure (S S )/(R S )-SIBA 4 complexed to (R)-BINOL was also observed on sample dilution (see Figure S1 and Table S5).The concentration-dependent changes in chemical shift values and imine peak resolution in these complexation reactions are all consistent with greater noncovalent interactions occurring between the SIBA enantiomers and BINOL at greater BINOL loadings and higher sample concentrations.Further evidence that BINOL was acting as a CSA in CDCl 3 was obtained by treating SIBA 4 (varying er) 35 with 100 mol % BINOL in CD 3 CN.These reactions produced concentrationindependent 1 H NMR spectra displaying a single unresolved imine peak at δ H 8.73 ppm, which is consistent with the more polar coordinating CD 3 CN solvent preventing H-bonding interactions from forming between the SIBA enantiomers and BINOL, thus not inducing enantiodiscrimination (see Figure S4).
To investigate the structural reasons for this change in complexation manifold, 11 B and 15 N NMR spectroscopic analyses on samples of SIBA (R S )-4, SIBE 5a, [SIBA (R S )-4• (R)-BINOL], and BINOL-derived IBE 2a in CDCl 3 were then carried out to determine which systems contained N→B coordination bonds (Figure 3, see Table S10 for full data set, including other diastereomers).δ N chemical shift values were determined indirectly through analysis of 1 H− 15 N HMBC spectra (Tables S10 and S11). 22,36SIBA (R S )-4 in CDCl 3 exhibited 11 B and 15 N NMR resonances at δ B = 28.6 ppm and δ N = −67.0ppm that are consistent with (R S )-4 containing a planar sp 2 -hybridized boron atom and a noncoordinated imine nitrogen atom.The absence of an N→B coordination bond in (R S )-4 in the solid state was confirmed by X-ray crystallographic analysis (see Figures 4a and S61 and S59). 1,37,38he absence of a stabilizing intramolecular N→B coordination bond in pinanediol-derived SIBE 5a is likely due to steric congestion and/or the electron-withdrawing N-sulfinyl group preventing electron density from being donated from its imine lone pair into the empty p orbital of the boron atom.This contrasts with BINOL-derived (α-R,R)-IBE 2a whose structure contains an intramolecular N→B coordination bond that generates sp 3 -hybridized boron and nitrogen atoms.
This absence of an intramolecular N→B coordination bond in pinanediol-derived SIBE 5a also provides an explanation for the inability of BINOL to form SIBE complexes.BINOL-SIBE formation would require incorporation of the conformationally restricted diol unit of BINOL (hindered rotation around the  biaryl bond) into a strained seven-membered cyclic boronate ester ring.Furthermore, the absence of a stabilizing N→B bond means that this ring system would also need to accommodate a planar sp 2 -hybridized boron atom, thus introducing further strain into an already strained ring system.−31 Further evidence that BINOL does not readily react with aryl boronic acids to produce cyclic boronate esters was acquired by carrying out complexation reactions of m-tolylboronic acid and 2-FPBA with pinanediol and BINOL (1.50 equiv) in CDCl 3 . 1−41 Crystals of BINOL−SIBA complexes could not be obtained, however, X-ray crystal analysis of (R S )-SIBA 4 revealed the presence of strong intermolecular hydrogen bonds between the sulfinyl oxygen atom of one SIBA unit and the boronic acid protons of another to produce well-defined 4-membered Ar− B(OH) 2 ••• − O-+ S-R hydrogen-bonding assemblies (Figure 4a).−44 Further evidence for strong intermolecular H-bonding interactions between BINOL and SIBA (R)-4 in CDCl 3 was obtained from selective 1D 1 H NOE spectroscopic studies on equimolar BINOL−SIBA mixtures in CDCl 3 .Selective excitation of the tert-butyl protons of SIBA (R)-4 revealed significant NOE interactions with all six 1 H NMR aromatic environments of (R)-BINOL.These NOE interactions are consistent with a hydrogen bond between the sulfinyl oxygen of a SIBA unit (acceptor) and a phenolic group of a BINOL unit (donor) serving to hold the SIBA tert-butyl group and the BINOL aromatic protons in close proximity (Figure 4c).
Having established that BINOL was acting as a hydrogenbonding CSA, we next explored its potential for determining the ers of scalemic samples of SIBA 4. Scalemic samples of SIBA 4 (10% er increments from (S S )-4 to (R S )-4) were treated with 1.50 equiv (maximum solubility level) 45 of (R)-BINOL at a 100 mM concentration in CDCl 3 .Examination of these 1 H NMR spectra revealed er-dependent variations in the chemical shifts of the diagnostic imine resonances of these scalemic samples (Figure 5a and Table S6).Treatment of (S S )-4 and (R S )-4 with (R)-BINOL gave 1 H NMR spectra displaying imine signals with essentially identical chemical shift values at δ H = 8.999 and δ H = 9.000 ppm, respectively (Figure 5a, entries 1 and 11).This is in contrast to the chemical shift values observed for (rac)-SIBA 4 and (R)-BINOL, which displayed resolved imine peaks at higher δ H = 9.015 and δ H = 9.030 ppm values, respectively (Figure 5a, entry 6).Increasing the er of the SIBA analyte from 60 (S) :40 (R) to 90 (S) :10 (R) er resulted in the chemical shift of the minor (R)imine resonance shifting incrementally downfield from δ H 9.035 to 9.045 ppm, and the major (S)-imine resonance moving incrementally upfield from δ H 9.017 to 9.009 ppm (Figure 5a, entries 5−2).These er-dependent changes in chemical shift produce a corresponding increase in Δδ H value from −0.018 ppm for a racemic sample to −0.036 ppm for a 90 (S) :10 (R) er sample (Figure 5a, cf.entries 6 and 2).However, regardless of their overall Δδ H values, integration of fully resolved major and minor imine peaks of each of these scalemic samples enabled their er values to be accurately determined.
Conversely, acquisition of 1 H NMR spectra of (R)-BINOL complexes of scalemic SIBA samples of 40 (S) :60 (R) and 30 (S) :70 (R) er revealed increasing overlap of the diastereomeric imine peaks, which coalesced into a single peak at ≤20 (S) :80 (R) er (Figure 5a, Entries 7−10), meaning that their er levels could not be determined.However, repeating the 1 H NMR study using (S)-BINOL (opposite CSA enantiomer) allowed this erdependent imine overlap issue to be resolved (Figure 5b and Table S7).Acquisition of a new set of 1 H NMR spectra using (S)-BINOL as the CSA produced a new series of 1 H NMR spectra that displayed er-dependent variations in the chemical shifts of the major and minor imine peaks that mirrored those produced previously using (R)-BINOL (cf. Figure 5a,5b).This means that (R)-BINOL could be used as a CSA to accurately determine the ers of SIBA samples ranging from 50 (S) :50 (R) to >99 (S) :1 (R) er, while (S)-BINOL could be used to accurately determine the ers of SIBA samples ranging from 50 (S) :50 (R) to <1 (S) :99 (R) er.Detection limits of this CSA system were determined by treating scalemic SIBA 4 samples ranging from 95 (S) :5 (R) to >99 (S) :1 (R) er with 1.50 equiv of (R)-BINOL, whose corresponding 1 H NMR spectra all gave accurate er values (up to 99.5 (S) :0.5 (R) er, i.e., a 99% ee detection limit, Figures 6 and  S6).Examination of the 1 H NMR spectra produced using (S)and (R)-BINOL revealed that they were not exactly mirrored, with very small variations likely due to minor changes in impurity profiles or moisture content subtly affecting the hydrogen-bonding networks that are responsible for enantiodiscrimination in this system.
It was subsequently found that reducing the amount of (R)-BINOL used in the er-dependent variation studies of SIBA 4 to one equivalent resulted in a remarkable crossover event occurring in the chemical shift values of the major and minor imine peaks as the er of the SIBA analyte was varied (Figure 5c and Table S8).As seen previously, variation in the chemical shifts of the 1 H NMR imine peaks was observed for er values between 90 (S) :10 (R) and 50 (S) :50 (R) er, with minor imine peaks of the (R S )-SIBA enantiomers appearing downfield of the major imine peaks of the (S S )-enantiomers (Figure 5c, Entries 2−6).Once again, incremental changes from 50 (S) :50 (R) to 30 (S) 70 (R) er resulted in increasing overlap of the major and minor imine peaks, which coalesced into a single broad imine peak at 30 (S) :70 (R) er (Figure 5c, Entry 8).However, derivatization of SIBA samples of ≤20 (S) :80 (R) er gave 1 H NMR spectra where the major imine peaks of the (R S )-enantiomers were now upfield to the minor imine peaks of their (S S )-enantiomers (Figure 5c, Entries 9 and 10, imine peaks partially overlapped).This means that an er-dependent switch in the Δδ H sign of the imine resonances occurs in this system, with (R)-BINOL producing a large negative Δδ H value of −0.030 ppm for a 90 (S) :10 (R) er sample (Figure 5c, entry 2), but a small opposing positive Δδ H value of +0.007 ppm value for a 10 (S) :90 (R) er value (Figure 5c, entry 10).The results of this er determination study were verified by carrying out complementary complexation studies using (S)-BINOL as the CSA, which gave the expected mirrored results (see Figure S5 and Table S9 for details).
The multicomponent intermolecular hydrogen-bonding networks that are proposed in these BINOL−SIBA complexes provide a simple explanation to understand the er-dependent variations in 1 H NMR imine chemical shift values that occur as the enantiopurity of the SIBA analyte is varied.The chemical shift values of the imine protons of each BINOL-coordinated SIBA enantiomer will be determined by the cumulative intramolecular and intermolecular shielding and deshielding effects that they experience.These intermolecular hydrogenbonding networks are likely formed from interactions between the phenol groups of BINOL units and the sulfinyl and boronic acid groups of the SIBA analyte.This means that different scalemic SIBA samples will contribute different relative amounts of (R)-and (S)-enantiomers to their respective hydrogen bond networks, resulting in variable intermolecular imine shielding/ deshielding effects being present in different scalemic samples.
We believe these concentration-and er-dependent variations in chemical shift values have important implications for how hydrogen-bonding-CSA-based NMR protocols should be bechmarked, with a lack of appreciation for these effects potentially leading to incorrect er values or absolute configurations being assigned.For example, the use of (R)-BINOL as a CSA for a SIBA sample of 10 (S) :90 (R) er produces a 1 H NMR spectrum containing a single imine peak that could be misinterpreted as evidence that this SIBA sample was enantiopure.In contrast, the 1 H NMR spectrum of the same SIBA sample 10 (S) :90 (R) er in the presence of (S)-BINOL displays a set of well-resolved major and minor imine peaks, whose imine integral ratio is an accurate reflection of the enantiopurity of the sample (cf.Entries 10 of Figure 5a,5b).Alternatively, comparison of the 1 H NMR spectra of SIBA samples of 90 (s) :10 (R) er and 10 (S) :90 (R) er in the presence of Figure 6.Expanded imine regions of the 1 H NMR spectra (500 MHz, CDCl 3 , 100 mM) of scalemic SIBA 4 with high 95 (S) :5 (R) -99.5 (S) :0.5 (R) er levels in the presence of 1.50 equiv (R)-BINOL.Chemical shifts referenced to TMS as an internal standard (∼6 mM).Measured integral values are within the accepted margins of error.See Figure S6 for enhanced spectra containing insets of minor imine signals.
1.00 equiv BINOL (Figure 5c, cf.Entries 2 and 9) reveals that their minor imine peaks are more deshielded in both cases.This could easily be misconstrued to mean that the same SIBA enantiomer is present in excess in both samples, with this example clearly illustrating the potential pitfalls of using the sign of Δδ H values to assign the absolute configuration of chiral analytes in hydrogen-bonding CSA systems.
Given the findings described herein, we now recommend that new hydrogen-bonding-CSA-based NMR protocols used to determine the ers of chiral analytes should be repeated using both enantiomers of the CSA to generate duplicate sets of 1 H NMR spectra at a fixed concentration and stoichiometry.These complementary 1 H NMR spectra should then be compared to check for the presence of any er-dependent variations in diagnostic chemical shifts, thus ensuring that the correct CSA enantiomer is chosen to determine accurate er values or absolute configurations.Furthermore, the dramatic variations in chemical shift observed in this hydrogen-bonding CSA NMR study clearly illustrate the importance of reporting NMR sample concentration to ensure the accuracy and reproducibility of er determination protocols.−48

■ CONCLUSIONS
This study reports that the enantiomers of SIBA 4 coordinate to BINOL (acting as a CSA) in CDCl 3 to produce diastereomeric hydrogen-bonded BINOL−SIBA complexes whose imine protons are distinguishable in their 1 H NMR spectra.The inability of BINOL to produce a three-component SIBE complex (like other diols) is due to the difficulty of incorporating its conformationally restricted diol unit and a planar sp 2 boron atom into a strained seven-membered boronate ester ring.Additionally, SIBAs and SIBEs were found to lack the N→B bond typical of IBEs, which further disfavors the incorporation of BINOL.The use of BINOL as a CSA to determine the ers of scalemic SIBA 4 samples revealed concentration-and er-dependent variations in diagnostic imine 1 H NMR chemical shift values that can result in imine peak overlap for selected er values.This imine peak overlap issue can be resolved by repeating the protocol using the opposite enantiomer of BINOL as the CSA.This adaptation enables BINOL to be used to accurately determine the enantiopurity levels of scalemic samples of SIBA 4 (and hence Ellman's sulfinamide 3) up to 99.5:0.5 er (99% ee).
Due to the potential complications highlighted throughout this work, we recommend that hydrogen-bonding-based CSA protocols used to determine er values of scalemic analytes should be repeated using both CSA enantiomers to identify whether er-dependent variations in diagnostic chemical shifts need to be accounted for, particularly in those cases where single resonances are observed (risk of peak overlap) or absolute configuration is being determined (possible signal crossover).We also re-emphasize the importance of carefully reporting analyte concentrations and CSA stoichiometries in experimental protocols to ensure that optimal CSA conditions can be reproduced effectively, and any risk of misassignment minimized.

Figure 3 .
Figure 3. 11 B and 15 N NMR chemical shifts of selected SIBA, SIBE, and IBE complexes.See TablesS10 and S11for the full data set and benchmarking studies.δ N is referenced to MeNO 2 .

Figure 4 .
Figure 4. (a) X-ray crystal structure of SIBA (R S )-4 showing intermolecular hydrogen-bonded ladders formed by the boronic acid moiety of one SIBA unit with the sulfoxide oxygen of another SIBA unit.(b) Hydrogen-bonding networks present in the X-ray crystal structures of (R S )-m-tolyl methyl sulfoxide complexed to BINOL.(Reproduced from ref 43 with permission from The Japanese Chemical Society).(c) Evidence of hydrogen bonding between BINOL and (R S )-4 in solution from 1 H NMR NOE interactions between the protons of the tert-butyl group of SIBA 4 and the aromatic protons of BINOL (see Figure S57 for spectra).H-bonds are represented by broken lines.