A Gallium-based Chiral Solvating Agent Enables the Use of 1H NMR Spectroscopy to Differentiate Chiral Alcohols.

In situ, direct 1H NMR chiral analysis by using chiral solvating agents is a convenient and efficient analytical technique. Here we developed a Ga-based chiral anionic metal complex for 1H NMR chiral analysis of alcohols. Utilizing the optimal pKa value, the Ga complex was able to differentiate 1H NMR signals of each (R)- and (S)-enantiomer of alcohols, measured at room temperature. This direct 1H NMR chiral analysis of alcohols was used to rapidly determine enantiomeric excess and conversion in a kinetic resolution and an asymmetric synthesis.

Introduction

The chirality of molecules is an exceptionally important property, and a large community in organic chemistry is dedicated to preparing and characterizing chiral molecules for a variety of applications (Chen et al., 2015, You et al., 2015, Leung et al., 2012, Hembury et al., 2008). Among the standard methods of analysis, nuclear magnetic resonance (NMR) spectroscopy enjoys a special status, as it is convenient, versatile, and routinely available. Unfortunately, the NMR signatures of enantiomers are by definition identical and it is therefore not possible to use standard NMR techniques to differentiate enantiomers from each other. Thus, chromatographic techniques, such as high-performance liquid chromatography (HPLC) or gas chromatography (GC), that can separate enantiomers are commonly used for characterization and reaction monitoring (Han, 1997, Schurig and Nowotny, 1990). One possible way of enabling NMR to differentiate enantiomers is to employ chiral solvating agents that form diastereomeric adducts with the substrate in question and deliver distinctive NMR signatures for each diastereomer (Figure 1) (Wenzel, 2007, Wenzel, 2017, Lacour and Moraleda, 2009, Seco et al., 2004, Schneider et al., 1998, Parker, 1991).

Figure 1

1H NMR Chiral Analysis by Using a Chiral Solvating Agent

This useful technique was successfully demonstrated for a variety of substrates with chiral amines and carboxylic acids being the most commonly targeted substrates (Benedict et al., 2018, Chen et al., 2018, Ema et al., 2018, Khanvilkar and Bedekar, 2018, Merino et al., 2018, Liu et al., 2011, Chinchilla et al., 1995). The emphasis on these species is not surprising, because the majority of chiral solvating agents employ non-covalent interactions such as hydrogen bonds and electrostatic attractions for structural recognition (Benedict et al., 2018, Chen et al., 2018, Ema et al., 2018, Khanvilkar and Bedekar, 2018, Merino et al., 2018, Liu et al., 2011, Chinchilla et al., 1995). Substrates that form relatively weak hydrogen bonds and are less strongly coordinating such as alcohols are more challenging to study. Nonetheless, several reagents have been shown to be effective at stereospecifically binding to chiral alcohols and allowing for separation of the NMR signals of the enantiomeric alcohols.

To overcome the intrinsic difficulty of relatively weak non-covalent interactions that are common for alcohols, several multifunctional analytes such as β-hydroxy esters (Uccello-Barretta et al., 1995), 1,2-diols (Pal et al., 2014, Ema et al., 2007, Wilen and Qi, 1991, Sweeting, 1987), cyanohydrins (Moon et al., 2009, Moon et al., 2010), and α-hydroxy carboxylic acids (Bai et al., 2019, Pal et al., 2015) or amides (Wolf et al., 2014) have been used to enhance interactions with chiral solvating agents, as shown in Figure 1A. For alcohols without any other polar functional groups, bis(seleno)urea (Bian et al., 2016) and cobalt(III) trication (Luu et al., 2018) were reported to be effective, but the scope of the substrates tested was relatively narrow. In 2018, we demonstrated that the sodium salt of a negatively charged aluminum complex (CASA-Na) can be used for the solvation and resolution of chiral alcohols (Figure 2) (Seo et al., 2018). Various substrates were successfully analyzed including primary, secondary, and tertiary alcohols with alkyl and aryl substituents. Unfortunately, low temperatures in a range of 0°C to −40°C were required owing to the small binding constants. Thus, one desirable enhancement strategy is to strengthen the intermolecular interactions between the anionic metal complex and the analyte. We discovered that significant improvement can be achieved by replacing aluminum with gallium, which is not only environmentally benign, non-toxic, and inexpensive, but also allows for differentiating the chiral alcohols with standard NMR methods at room temperature. In addition, in situ chiral analysis of an alcohol was practically demonstrated in a chiral resolution and an asymmetric synthesis.

Figure 2

Binding Models and Synthetic Procedure

(A) Binding models of CASA.

(B) Synthetic procedure of metal complexes with L1-L3. See also Figures S1–S7 for 1H & 13C NMR spectra.

Results and Discussion

Preparation and Physical Properties of Charged Metal Complexes

Our chiral aluminum complex (CASA) is a unique class of 1H NMR chiral solvating agent based on a negatively charged “ate” complex with metal-centered chirality. Experimental data such as crystal structures and job plots supported 1:1 binding models between CASA and analytes, as shown in Figure 2A. Accordingly, two stable ionic salts with H+ and Na+, CASA-H and CASA-Na, were demonstrated to be effective for 1H NMR chiral solvation of amines and carboxylic acids, respectively (Figure 2A) (Seo and Kim, 2015). CASA was shown to be an excellent chiral solvating agent working for all types of amines and carboxylic acids with a general analyte scope and a solvent compatibility. The utility of CASA-Na was recently expanded to cover chiral alcohols (Seo et al., 2018). Although the analyte scope of CASA-Na for chiral alcohols was sufficiently wide, low temperature measurement (0°C ~ −40°C) was required for baseline peak separation owing to weak intermolecular interaction between CASA-Na and an alcohol, which may limit the practical applications. According to our experiments, the binding constants for an alcohol are about 80–90 times weaker than those for the corresponding carboxylic acid (Seo et al., 2018). To achieve chiral solvation of alcohols at room temperature, we intended to further modulate the anionic octahedral metal complex. Our synthetic strategy is (1) to modulate the ligand's electronic property or (2) to change the metal(III) center. Although an increase of intermolecular interaction is desirable for better chiral solvation ability, the synthetic direction and the resulting physical properties are not simply predicted in the anionic octahedral complex.

In this study, six chiral metal complexes were prepared, as shown in Figure 2B (see also Transparent Methods). A base-mediated coupling between L1 and M(III) precursors such as AlCl3, Ga(acac)3, Sc(OTf)3, and In(acac)3 gave [Al-L1]Na, [Ga-L1]Na, [Sc-L1]Na, and [In-L1]Na with excellent yield (95%–98%) and stereoselectivity (>99%) (Figure 2B and see also Figures S1–S7). In addition, bromo-substituted ligands L2 and L3 prepared by bromination of 2,2′-dihydroxybenzophenone with N-bromosuccinimide were used to form [Al-L2]Na and [Al-L3]Na, respectively, with 99% yields (Figure 2B and see also Transparent Methods). In all cases, the metal-centered chirality was completely controlled to provide a single diastereomeric metal complex.

We then evaluated the physical properties of six metal complexes by measuring pKa values in dimethyl sulfoxide (DMSO) (Figure 3 and Table 1, see also Transparent Methods) (Christ et al., 2011). [Al-L1]H was previously measured to show a pKa value of 5.02 (Seo and Kim, 2015). To develop better chiral solvating agents for alcohols than [Al-L1]Na, a more basic metal complex with a higher pKa value may be necessary to establish stronger intermolecular interactions. Complexes with brominated ligands, [Al-L2]H and [Al-L3]H, were found to be more acidic, with pKa values of 4.48 and 3.33, respectively. The inductive effect by electronegative Br atoms can stabilize the conjugate bases [Al-L2]Na and [Al-L3]Na, resulting in increased acidity of [Al-L2]H and [Al-L3]H. On the other hand, complexes with different central metals, [Ga-L1]H, [Sc-L1]H, and [In-L1]H, were found to be more basic, with pKa values of 5.73, 6.35, and 6.77, respectively. These pKa values are in a linear relationship with the effective ionic radii of Al3+, Ga3+, Sc3+, and In3+, which are 53.5, 62, 74.5, and 80 pm, respectively.

Figure 3

Anionic Octahedral Metal Complexes of Al, Ga, Sc, and In

Table 1

Measured pKa Values and Metal to Oxygen Bond Lengths Obtained from Density Functional Theory Calculation

	Al-L3	Al-L2	Al-L1	Ga-L1	Sc-L1	In-L1
Average M-O (Å)	1.87	1.87	1.88	1.91	2.04	2.05
pKa (DMSO)	3.33	4.48	5.02	5.73	6.32	6.77

See also Tables S1–S11 for Cartesian coordinates.

To understand the linear correlation between the ionic radii and pKa values of the metal complexes, DFT computation was performed to optimize the complex geometries and to calculate M-O bond lengths (see also Tables S1–S6). For Al, Ga, Sc, and In complexes with L1, there is a good linear relationship between the calculated average M-O bond lengths and the measured pKa values with R2 = 0.896 (Table 1). As the M-O bond increases in the anionic octahedral metal complex, the complex becomes more basic. A linear correlation between gas phase bond lengths and experimental pKa values was well demonstrated in organic compounds, such as carboxylic acids and phenols (Caine et al., 2018, Harding and Popelier, 2011), and is applicable to anionic octahedral metal complexes. However, for Al complexes with L1-L3, almost the same M-O bond lengths of 1.87–1.88 were obtained despite the large difference in pKa values (ΔpKa = 1.7). It appears that the relationship between M-O bond length and pKa value is not applicable to the inductive effect.

A linear relationship between bond length and pKa value was then confirmed by solid-state crystal structures of Al, Ga, Sc, and In complexes. As shown in Figure 4, crystal structures of [Al-L1], [Ga-L1], [Sc-L4] (Seo and Kim, 2015), and [In-L1] were obtained (L4 was prepared from diaminocyclohexane instead of diphenylethylenediamine) (see also Figures S8–S10). Interestingly, the average M-O bond lengths of 1.86, 1.94, 2.04, and 2.12 Å for Al, Ga, Sc, and In complexes, respectively, are in an excellent linear relationship with the measured pKa values (R2 = 0.988) (Table 2). Thus, both the computed and solid-state structures can be used to correlate the basicity of the negatively charged octahedral metal complexes. Accordingly, we prepared chiral anionic metal complexes with a pKa range of 3.33–6.77.

Figure 4

Crystal Structures (Thermal Ellipsoids at 50% Probability: All Hydrogens and Sodium Cations Are Omitted for Clarity)

(A) [Al-L1]Na.

(B) [Ga-L1]Na.

(C) [Sc-L4]Na.

(D) [In-L1]Na.

See also Figures S8–S10.

Table 2

Metal to Oxygen Bond Distances Obtained from the Crystal Structures

	Al-L1	Ga-L1	Sc-L4	In-L1
M-O1 (Å)	1.83	1.89	2.03	2.09
M-O2 (Å)	1.83	1.90	2.04	2.11
M-O3 (Å)	1.89	1.96	2.04	2.11
M-O4 (Å)	1.89	2.00	2.06	2.21
Average M-O (Å)	1.86	1.94	2.04	2.12

Chiral Solvation of Alcohols with [Ga-L1]Na at Room Temperature

The chiral solvation ability of six anionic metal complexes was investigated. In CD3CN (0.5 mL), an equimolecular amount of the metal complex (0.01 mmol) and rac-1-phenyl-1-butanol (1, 0.01 mmol) was mixed and 1H NMR spectra were then recorded at room temperature. As found in the proposed binding model in Figure 2A, the hydroxyl group of the alcohol directly interacts with the metal complex mediated by the sodium cation and the peak separation of the hydroxyl group is mainly observed in this system. Compared with a default complex [Al-L1]Na providing ΔΔδ of 0.020 ppm, [Al-L2]Na and [Al-L3]Na showed poor peak resolution owing to the decreased basicity with the bromo substituents (Figures 5, see also Figures S11 and S12). When a more basic metal complex [Ga-L1]Na was tested, a significant enhancement of the peak resolution (ΔΔδ = 0.065 ppm) was observed with [Ga-L1]Na. However, when two more basic metal complexes [Sc-L1]Na and [In-L1]Na were used, the hydroxyl signal of the alcohol disappeared and merged with the water peak (Figure 5). Because of the observed intermolecular proton transfer between the hydroxyl analyte and the residual water, [Sc-L1]Na and [In-L1]Na are found to be too basic to observe the analyte's hydroxyl signals. Accordingly, [Ga-L1]Na exhibits an optimal basicity in terms of being used for chiral solvation of rac-1-phenyl-1-butanol (1).

Figure 5

Partial 1H NMR Spectra of Hydroxyl Peaks of rac-1 with the Anionic Metal Complexes in CD3CN

See also Figures S11–S13 for full 1H NMR spectra

To verify the assumption that a more basic anionic metal complex can strongly bind to an alcohol analyte, the binding constants of [Al-L1]Na and [Ga-L1]Na were compared. (R)- and (S)-1-phenylethanol (2) were used as an analyte, and the binding constants were measured by a 1H NMR titration experiment (see also Transparent Methods). The binding constants of [Al-L1]Na with (R)-2 and (S)-2 were measured to be 0.382 and 0.136 M−1, respectively, whereas those of [Ga-L1]Na with (R)-2 and (S)-2 were measured to be 1.01 and 0.493 M−1, respectively (Figure 6). These data indicate that the more basic [Ga-L1]Na binds about 2.5–3.5 times stronger to the chiral alcohol 2 than [Al-L1]Na, consistent with our idea to enhance the intermolecular interactions with more basic anionic complexes. We also found that the binding constants of both Al and Ga complexes for (R)-2 are about two times greater than those for (S)-2. The energy difference for the formation of diastereomeric mixtures may be further utilized in the separation of racemic analytes (Mittal et al., 2015, Fogassy et al., 2006). Indeed, the increased binding constants of [Ga-L1]Na resulted in a better peak resolution, as shown in Figure 6. Although [Al-L1]Na gave only partial peak separation of rac-2 in CD3CN at room temperature, [Ga-L1]Na provided cleanly separated 1H NMR signals with a ΔΔδ value of 0.041 ppm.

Figure 6

Binding Constants with 1-Phenylethanol (1)

(A) [Al-L1]Na.

(B) [Ga-L1]Na.

In addition, the effect of counter cation was investigated. Among [Ga-L1] complexes with counter cations Li+, Na+, K+, or Cs+, [Ga-L1]Na was found to be the most efficient chiral solvating agent for 2 (see also Figure S13). Because the crystal structures showed various intermolecular interactions between the sodium cation and solvent molecules, we proposed a binding model as shown in Figure 2A, where the sodium cation plays a critical role to mediate the anionic complex and the alcohol analyte.

The analytical scope of [Ga-L1]Na is summarized in Figure 7 (see also Figures S14–S17). When various chiral alcohols 1-12 were mixed with a stoichiometric amount of [Ga-L1]Na at room temperature in CD3CN (20 mM), the 1H NMR spectra showed full baseline peak separation for efficient and reliable chiral analysis. In many cases, the hydroxyl proton signals were split upon the addition of [Ga-L1]Na, supporting our proposed intermolecular interactions among [Ga-L1]−, Na+, and the analyte alcohol (Figure 2A). Secondary chiral alcohols 1-6 with phenyl substituents showed the same peak pattern where (S)-enantiomer is more downfield shifted. This splitting pattern can be used to determine the absolute chirality of secondary alcohols, but the opposite splitting pattern was found for phenyl-substituted alcohols 7, 8, and 9 with cyclopropyl, acetyl, and ester groups, respectively. Thus, statistical analysis as well as theoretical investigation is further required to use this chiral solvating agent for the determination of absolute chirality.

Figure 7

Partial 1H NMR Spectra of Chiral Alcohols with [Ga-L1]Na at Room Temperature in CD3CN. 1H NMR Signals of Best Peak Separation Are Shown

(A) 1-Phenyl-1-ethanol.

(B) 1-Phenyl-1-butanol.

(C) 1-Phenyl-1-dodecanol.

(D) 1-Phenyl-4-buten-1-ol.

(E) α-(Trifluoromethyl)benzyl alcohol.

(F) 1-Indanol.

(G) α-Cyclopropylbenzyl alcohol.

(H) 1-Phenyl-2-propyn-1-ol.

(I) Ethyl mandelate.

(J) Menthol.

(K) 1-(2-Chlorophenyl)-1-phenylmethanol.

(L) 1,2-Diphenylethane-1,2-diol.

See also Figures S14–S17 for full 1H NMR spectra and Figures S18–S20 for comparison with reported chiral solvating agents.

In addition, other secondary alcohols including dialkyl or diaryl-substituted alcohols (10 or 11) and diol 12 were successfully used for 1H NMR baseline peak separation of enantiomers with [Ga-L1]Na. However, 2-butanol (13) and mevalonolactone (14) gave poor peak resolution when a stoichiometric amount of [Ga-L1]Na was used at room temperature. Instead of low temperature measurement, we sought other convenient operational protocols. We found that an increased amount of [Ga-L1]Na improved the peak separation of enantiomers and six equivalents of [Ga-L1]Na gave a full baseline peak separation for 13 and 14 (Figure 8, see also Figures S21–S23). Given the general analyte scope shown in Figures 6 and 7, 1H NMR chiral analysis of alcohols by using [Ga-L1]Na can be an efficient and convenient procedure at room temperature. To compare the chiral solvation ability of [Ga-L1]Na with other chiral solvating agents, we tested 1H NMR chiral analysis of four alcohols 1, 4, 10, and 15 with previously reported chiral solvating agents, including Pirkle's alcohol (Wenzel, 2017, Wenzel and Chisholm, 2011), chiral crown ether (Lacour and Moraleda, 2009), β-cyclodextrin (Schneider et al., 1998), and europium shift agent (Sweeting, 1987), in CD3CN (see also Figures S18–S20). Among 14 chiral solvating agents, [Ga-L1]Na was the only one to give the full baseline peak separation of the four analytes, whereas other chiral solvating agents showed no or partial peak separation. Moreover, ligand L1 was not effective for the chiral analysis of alcohols, supporting the importance of chiral octahedral anionic metal complex for chiral solvation of alcohols.

Figure 8

Partial 1H NMR Spectra of Chiral Alcohols with Different Equivalents of [Ga-L1]Na Complex in CD3CN

(A) 2-Butanol.

(B) Mevalonolactone.

See also Figures S21–S23 for full 1H NMR spectra.

Direct 1H NMR Analysis of Chiral Alcohol with [Ga-L1]Na in Kinetic Resolution and Asymmetric Synthesis

Chiral alcohols can be synthesized via (1) multistep transformations from a chiral pool such as terpenes, amino acids, and carbohydrates (Fan et al., 2019, Hung et al., 2019, Brill et al., 2017) or (2) kinetic resolution of racemic alcohols or their derivatives (Liu et al., 2019, Selier et al., 2019, Zhang and Ma, 2018). In addition, asymmetric synthesis such as reduction or nucleophilic addition of prochiral carbonyls recently has been developed (Neves-Garcia et al., 2018, Tsai et al., 2018; Bieszczad and Gilheany, 2017, Nakamura et al., 2017). Although HPLC and GC with chiral stationary phases are commonly used for chiral analysis of alcohols as a reliable direct analytical technique, derivatization of alcohols such as introducing chromophores, polar functional groups, or hydrophobic units is often necessary for the desirable analytical results. However, the chiral analysis of alcohols by 1H NMR spectroscopy can be a direct, efficient, and convenient analytical technique that is practically applicable for the synthesis of chiral alcohols by resolution or asymmetric synthesis.

To demonstrate the utility of 1H NMR chiral analysis of alcohols with [Ga-L1]Na, a challenging analyte, 1-penten-3-ol (15), was selected in this study. In the previous reports, the direct chiral analysis of 1-penten-3-ol (15) was not achieved by HPLC or GC analysis, likely due to the lack of aryl chromophores or minor structural differences between ethyl and ethylene groups (Fernandez-Perez et al., 2016, Ryan and Jaimison, 2006, Paquette and Sweeney, 1990). Accordingly, further synthetic efforts such as benzoylation are required for the HPLC analysis, which takes an additional 18 h together with purification by silica gel column chromatography (Fernandez-Perez et al., 2016). In contrast, we demonstrated that [Ga-L1]Na can be used for a direct chiral analysis of 1-penten-3-ol within minutes by 1H NMR spectroscopy.

Kinetic resolution is a protocol to separate enantiomers by a kinetically controlled reaction. Because the reaction theoretically is quenched at 50% conversion, direct reaction monitoring is highly desirable to achieve high yields and enantiopurities. The Sharpless epoxidation of 1-penten-3-ol (15) was reported for kinetic resolution of 1-penten-3-ol (15), but the enantiopurity of the product was determined only by measuring optical rotation and the detailed conditions were not provided (Ryan and Jaimison, 2006, Paquette and Sweeney, 1990). As we can directly analyze the enantiopurity of 1-penten-3-ol (15), the Sharpless epoxidation was monitored by 1H NMR spectroscopy. As shown in Figure 9, the enantiomeric excess was found to be 66% after 7 h and >99% ee after 14 h (see also Figure S24). Thus, the kinetic resolution process can be readily optimized and this protocol can be used for other kinetic resolutions of alcohols.

Figure 9

Kinetic Resolution by Sharpless Asymmetric Epoxidation of 1-peten-3-ol (15)

(A) Reaction scheme.

(B) Partial 1H NMR spectra of 1-penten-3-ol (15) with [Ga-L1]Na during kinetic resolution.

See also Figure S24.

The second practical example of using [Ga-L1]Na is a rapid chiral analysis of 1-penten-3-ol (15) in Rh-catalyzed hydrogenative desymmetrization of 1,4-pentadien-3-ol (17) (Figure 10) (Fernandez-Perez et al., 2016). This reaction was reported by Vidal-Ferran and co-workers and a chiral phosphite ligand was found to be efficient (>99% yield and 84% ee). To demonstrate the direct assessment of %ee by 1H NMR, a mixture of 1,4-pentadien-3-ol (17) and rac-1-penten-3-ol (15) was mixed with [Ga-L1]Na in CD3CN. 1H NMR spectra showed clean baseline separation of all hydroxyl peaks, allowing us to calculate conversion and enantiomeric excess together by a single measurement (Figure 10A, see also Figures S25 and S26). Using this rapid analytical protocol within minutes employing 1H NMR spectroscopy, we tested 14 commercially available chiral phosphorus ligands and the resulting %conversion and %ee are summarized in Figure 10C (see also Figures S27 and S28). In general reaction conditions, a solution of [Rh(nbd)2]BF4, phosphine ligand, and 1,4-pentadien-3-ol (17) in CD2Cl2 (0.2 mmol) was stirred for 23 h at −40°C with a H2 balloon. For the determination of %ee and %conversion, 50 μL of crude mixture was mixed with 0.04 mmol of [Ga-L1]Na in 450 μL of CD3CN. The enantiomeric excess ranging from 3% to 88% was successfully measured, even when the yield was as low as 11%. As the enantiomeric excess can be measured at low conversion, the 1H NMR chiral analysis by using [Ga-L1]Na can be used to monitor asymmetric reactions.

Figure 10

Rh-Catalyzed Hydrogenative Desymmetrization of 1,4-Pentadiene-3-ol

(A) Partial 1H NMR spectrum of 1:1:2 mixture of 1,4-pentadiene-3-ol (17) and 1-penten-3-ol (15) with [Ga-L1]Na. See also Figures S25–S27.

(B) Reaction scheme and structures of ligand P1-14.

(C) Direct analysis of %conversion and %enantiomeric excess with various phosphorus ligands P1-P14 using [Ga-L1]Na.

See also Figure S28.

Conclusion

In summary, we have developed a Ga-based chiral metal complex as a general, efficient, and practical 1H NMR chiral solvating agent for various alcohols. 1H NMR chiral analysis of chiral alcohols without any other functional groups has been challenging owing to weak binding interactions with chiral solvating agents. In this study, several anionic chiral metal complexes were synthesized and we found that the pKa values are in a good linear relationship with M-O bonds in the octahedral geometries. The Ga-based chiral metal complex had an optimal pKa value of 5.7 in DMSO, providing sufficient baseline peak separations of chiral alcohols when [Ga-L1]Na was simply added to CD3CN. Various chiral alcohols, including primary, secondary, and tertiary alcohols, with alkyl and aryl substituents were tested, and all enantiomers were cleanly resolved in 1H NMR spectra measured at room temperature with [Ga-L1]Na. To develop a practical direct technique for chiral analysis of alcohols, we demonstrated that 1H NMR chiral analysis of 1-penten-3-ol can be efficiently used for a kinetic resolution by the Sharpless epoxidation and Rh-catalyzed hydrogenative desymmetrization of 1,4-pentadien-3-ol, where both %ee and %conversion were determined by a single measurement. As a complementary analytical technique, 1H NMR spectroscopy will be useful for rapid and reliable chiral analysis of alcohols under operationally simpler and more practical conditions with a general chiral solvating agent.

Experimental Procedures

General Procedure for 1H NMR Chiral Analysis of Alcohols

In a 5-mm NMR tube, 1.5 μL of rac-1-phenyl-1-butanol was dissolved in 0.5 mL of CD3CN. 1H NMR spectra (400 MHz) was taken of the solution (data-1). To the solution, 7.0 mg of [Ga-L1]Na was added and shaken until the solution became clear. 1H NMR spectra (400 MHz) was taken of the solution (data-2). By comparing two spectra of data-1 and data-2, resolved peaks of the hydroxy group at 3.20–3.35 ppm were analyzed. For the determination of R/S configuration, enantiopure or enantioriched samples can be analyzed.

Limitations of the Study

1H NMR chiral analysis of alcohols with [Ga-L1]Na could not be achieved in polar solvents such as DMSO-d6, Acetone-d6, and methanol-d4. Moreover, anhydrous CD3CN dried over 4Å molecular sieves is necessary because residual water in the NMR solvent decreases the amount of peak separation and increases line broadening of hydroxyl proton signals.

Methods

All methods can be found in the accompanying Transparent Methods supplemental file.