Gaowei
Li
a,
Jiangming
Cao
a,
Wen
Zong
b,
Xinxiang
Lei
*b and
Renxiang
Tan
*a
aInstitute of Functional Biomolecules, State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University, Nanjing 210093, P. R. China. E-mail: rxtan@nju.edu.cn
bCollege of Chemistry & Materials Engineering, Wenzhou University, Wenzhou 325035, P. R. China. E-mail: xinxianglei@gmail.com
First published on 26th November 2015
Enantiopure diphenylprolinols were synthesized from a commercially available starting material. The utility of 1H NMR spectroscopy for the differentiation of enantiomers using these chiral compounds as CSAs is stated, and their capacity acting as receptors for various carboxylic acids via hydrogen bonding is exploited. A linear relationship has been observed between the experimental and observed values of ee indicating the possible use of these compounds for quick and reliable analysis of enantiomerically enriched samples of various mandelic acids. From the experiments performed a preliminary conclusion indicated that the diphenylprolinol 1b with the free NH and OH is most effective in the chiral discrimination of carboxylic acids in 1H NMR.
Chiral carboxylic acids and their derivatives are organic molecules involved in a wide variety of biological processes and the investigation of the chiral recognition of carboxylic acids by artificial receptors was of critical importance in the preparation, separation, and analysis of enantiomerically pure carboxylic acids and disclosing the mechanism of interaction of the carboxylic acids with biological systems.7 The growing use of optically pure carboxylic acids has given rise to the need for the development of fast and accurate methodologies for the determination of the enantiomeric composition of chiral carboxylic acids. In the past few decades, a variety of chiral NMR solvating agents has been the focus of extensive investigation to determine the enantiomeric purity and understand the basic mechanism of host–guest complexation, particularly for carboxylic acids, such as chiral amines,8 amino alcohols,9 macrocyclic amines and amides,10 calixarenes,11 crown or aza-crown ethers,12 BINOL and their derivatives,13 chiral thiourea moieties,14etc. To date, however, there are only a few examples of optically active proline-derived receptors for the enantiomeric recognition of carboxylic acids.15 Therefore, the development of structurally simple yet efficient receptors for the enantioselective recognition of carboxylic acids still remains a challenging goal.
Among the variety of commercially available amino alcohol derivatives well documented in the literature, proline-derived diphenylprolinols have usually been widely employed as chiral ligands in the field of asymmetric synthesis and catalysis over the past decade. Because of the existence of multiple H-bonding and rigid heterocyclic ring of diphenylprolinols, they have mostly been used to catalyze the enantioselective alkylation,16 organocatalytic Michael addition,17 direct Aldol reaction18 and acyloin condensation,19 as well as promising applications, especially in efficient construction of axially chiral allenes or allenols.20 However, to the best of our knowledge, there is no report on the use of structurally simple diphenylprolinols as CSAs for the analysis of chiral carboxylic acids. In recent years, we have paid continuing attention to develop new and effective CSAs for the determination of enantiomeric excess and the application of chiral discrimination.21 Therefore, in continuation of our work on the development of structurally simple and effective CSAs in this aspect, we herein report L-proline-based diphenylprolinols and their recognition ability for carboxylic acids by 1H NMR spectroscopy.
The chiral N-benzyl-(S)-diphenyl(pyrrolidin-2-yl)methanol 1a was synthesized starting from L-proline according to the literature procedure,22 and the subsequent hydrogenolysis of 1a gave (S)-diphenyl(pyrrolidin-2-yl)methanol 1b in good yield. Their structures are shown in Scheme 1.
In order to explore the binding properties of chiral diphenylprolinols (S)-1a and (S)-1b, they were then screened to see their efficacy in binding with a test sample of DL-mandelic acid in 1H NMR analysis. The NMR experiments were performed with stoichiometric amounts of mandelic acid and CSA (1:
1) (10 mM, in CDCl3). The upfield change in the position of the signal of CαH proton of mandelic acid upon treatment with CSA was measured (Δδ) while the degree of splitting was measured by the differences in the separated peaks in terms of chemical shift non-equivalences as chemical shift change (ΔΔδ). The two entries of Table 1 indicate that the N-Bn, OH derivative CSA 1a showed poor recognition (entry 1), while the basic unit (S)-CSA 1b with free NH and OH groups showed very good recognition (entry 2). This observation perhaps indicates that the presence of the complex system is sufficient to offer multiple intermolecular hydrogen-bonding between the free NH and OH of the CSA and the test substrate. Therefore, the (S)-CSA 1b was determined as the most effective receptor, and conducted a further study.
To further test the ability for enantiomeric discriminating of chiral diphenylprolinol 1b as CSA for carboxylic acids, we first recorded 1H NMR of 10 mM (R)-, (S)- and racemic mandelic acid (MA). The addition of (S)-CSA 1b to a solution of racemic MA in CDCl3 caused CαH resonance in the 1H NMR spectrum of MA to shift upfield and, in most cases, to split into two equal-intensity singlets that drift downfield slightly. Representative spectra for the most effective (S)-CSA 1b are presented in Fig. 1a. The two singlets were assigned to the MA enantiomers, which become desymmetrized through their interaction with the diphenylprolinol.
The stoichiometry was determined according to the Job's method of continuous variation. The Job plots of X*Δδ versus the molar fraction (X) of (R)- or (S)-mandelic acid in the mixture were obtained, which showed a maximum at X = 0.5, and the best baseline resolution was both achieved with a 1:
1 ratio of (S)-1b/MA-2, as is shown in Fig. 1b. This indicated that (S)-1b and the acid bind in a 1
:
1 complex under these conditions.
To further explore the practical applications in the general ability of discrimination of (S)-CAS 1b for a variety of analytes by 1H NMR spectroscopy, the following carboxylic acids were chosen as guests (Fig. 2). These carboxylic acids vary from one another in their substituent group. The ΔΔδ values of α-H signals were appropriate to give a good baseline resolution for most of the tested analytes which ranges from 0.005 to 0.096 ppm, the results are summarized in Table 2. The p-substituted aromatic hydroxy acids (Table 2, entries 2–6) almost showed a more bigger ΔΔδ value compared with the o-substituted aromatic hydroxy acids (Table 2, entries 7–10). In particular, p-CF3-substituted aromatic carboxylic acid (±)-3 showed the biggest ΔΔδ value as 0.076 ppm (Table 2, entry 2), while the o-halogen-substituted aromatic hydroxy acid showed the lowest ΔΔδ value as 0.005 ppm (Table 2, entry 10). Similar results can be obtained for the m-substituted aromatic hydroxy acids (Table 2, entries 11 and 12), the above results indicated that the enantiodifferentiation of (S)-CAS 1b could be possibly weakened due to steric hindrance of the o-substituted group. It is also interesting to note that chiral discrimination was observed for the OMe signals of (±)-7 (Table 2, entry 6). The aromatic carboxylic acid with no α-OH chiral discrimination was also observed for the Me/OMe and methine proton (CαH) signals (Table 2, entries 13 and 14).
Entry | Carboxylic acids | ΔΔδb (ppm) | ΔΔδb (Hz) | Spectrumb |
---|---|---|---|---|
a Typical conditions: concentration of the guest and the (S)-CSA 1b is 10 mM (1![]() ![]() |
||||
1 | (±)-2 | 0.062 | 31 |
![]() |
2 | (±)-3 | 0.076 | 38 |
![]() |
3 | (±)-4 | 0.056 | 28 |
![]() |
4 | (±)-5 | 0.058 | 29 |
![]() |
5 | (±)-6 | 0.058 | 29 |
![]() |
6 | (±)-7 | 0.060 | 30 |
![]() |
0.003 | 1.5 |
![]() |
||
7 | (±)-8 | 0.022 | 11 |
![]() |
8 | (±)-9 | 0.008 | 4 |
![]() |
9 | (±)-10 | 0.012 | 6 |
![]() |
10 | (±)-11 | 0.005 | 2.5 |
![]() |
11 | (±)-12 | 0.044 | 22 |
![]() |
12 | (±)-13 | 0.050 | 25 |
![]() |
13 | (±)-14 | 0.023 | 11.5 |
![]() |
0.003 | 1.5 |
![]() |
||
14 | (±)-15 | 0.052 | 26 |
![]() |
0.014 | 7 |
![]() |
||
15 | (±)-16 | 0.096 | 48 |
![]() |
16 | (±)-17 | 0.057 | 28.5 |
![]() |
17 | (±)-18 | 0.061 | 30.5 |
![]() |
18 | (±)-19 | 0.007 | 3.5 |
![]() |
0.003 | 1.5 |
![]() |
||
19 | (±)-20 | 0.006 | 3 |
![]() |
0.014 | 7 |
![]() |
What's more, (S)-CAS 1b can even successfully discriminate the enantiomers of aliphatic carboxylic acids (Table 2, entries 15–19), especially for the pyromucic acid (±)-16 showed the biggest ΔΔδ value as 0.096 ppm (Table 2, entry 15), while the propionic acid derivatives (±)-19–20 gave poor results (Table 2, entries 18 and 19).
To explore the quantitative analysis ability of (S)-1b as a CSA for enantiomeric determination, the ee values of non-racemic mandelic acid 7 samples were determined by integration of the α-H signal of 7 in 1H NMR. The results shown in Fig. 3 were calculated based on the integrations of the NMR signals, and were within ±1% of the actual enantiopurity of the samples. The linear relationship between the NMR-determined values and those gravimetry determined values is excellent with R2 = 0.9995.
Having demonstrated enantiodiscrimination of the above carboxylic acids, we then tried to investigate the applicability of CSA 1b for the determination of absolute configuration of chiral carboxylic acids by a suitable method.14d,23 The NMR spectral behavior of a series of chiral carboxylic acids of known absolute configuration was studied using (S)-CSA 1b and its enantiomer (R)-CSA 1b to find the existence of any correlation between the absolute configuration and the NMR chemical shifts (Table 3). The results indicate that negative ΔδR,Sα-H correlates with (S)-carboxylic acids and positive ΔδR,Sα-H correlates with (R)-carboxylic acids.
Entry | Chiral carboxylic acids | ΔδR,Sα-H![]() |
---|---|---|
a ΔδR,Sα-H(ΔδR − ΔδS) values for α-H chiral carboxylic acids are shown. b Assigned configurations are labeled in parentheses. | ||
1 | (R)-Mandelic acid | +0.026(R) |
2 | (S)-Mandelic acid | −0.027(S) |
3 | (R)-4-Chloromandelic acid | +0.022(R) |
4 | (S)-4-Chloromandelic acid | −0.030(S) |
5 | (R)-4-Methoxymandelic acid | +0.027(R) |
6 | (S)-4-Methoxymandelic acid | −0.031(S) |
7 | (R)-3-Chloromandelic acid | +0.014(R) |
8 | (R)-2-Phenylpropionic acid | +0.006(R) |
9 | (S)-2-Tetrahydrofuroic acid | −0.163(S) |
10 | (S)-2-Hydroxy-3-methylbutanoic acid | −0.069(S) |
11 | (S)-2-Hydroxy-4-methylpentanoic acid | −0.070(S) |
12 | (S)-2-Hydroxy-3,3-dimethylbutyric acid | −0.149(S) |
We have developed structurally simple diphenylprolinols and tested their efficacy as CSAs in 1H NMR spectroscopy to discriminate carboxylic acids. A linear relationship has been observed between the experimental and observed values of ee indicating the possible use of these compounds for quick and reliable analysis of enantiomerically enriched samples of mandelic acid. From the experiments performed a preliminary conclusion indicated that the diphenylprolinol 1b with the free NH and OH is most effective in the chiral discrimination of carboxylic acids in 1H NMR.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5qo00264h |
This journal is © the Partner Organisations 2016 |