The structures of cyclic aminals from aldehydes and “roofed” mono-N-sulfonyl cis-diamines

ABSTRACT

Racemic Aldehydes were converted into aminals by using “roofed” mono-N-sulfonyl cis-diamines. Diastereomeric aminals could be separated by silica-gel TLC. The whole configuration of each aminal could be determined by NMR analyses. The chiral reagents were prepared from aminals of racemic reagents and a chiral aldehyde. The reverse may be also true. Epimerization of some aminal remains to be solved.

Abbreviations: RDA-Ms: N-(12-amino-9,10-dihydro-9,10-ethanoanthracen-11-yl)-methylsulfonamide; RDA-Ts: N-(12-amino-9,10-dihydro-9,10-ethanoanthracen-11-yl)-4-methylbenzenesulfonamide; RDA-1Nps: N-(12-amino-9,10-dihydro-9,10-ethanoanthracen-11-yl)-1-　naphthylsulfonamide; RDA-2Nps: N-(12-amino-9,10-dihydro-9,10-ethanoanthracen-11-yl)-2-naphthylsulfonamide

KEYWORDS:

• Aldehyde
• “roofed” mono-N-sulfonyl cis-diamine
• cyclic aminal
• absolute configuration

Identification of the chirality of aldehydes at α-position has much attention [1]. One of the determination methods is derivatizations into cyclic derivatives, for example, 2-substituted thiazoline [2,3], 2-substituted oxazoline [4], and imidazolidine [5,6] to analyze the chirality. It has been reported that the enantiomeric purity of aldehydes could be determined by NMR or chromatographic methods of imidazolidine derivatives with N,N-dimethyl-1,2-diphenyl ethylene diamine (DMPEDA) and target aldehydes [5]. We imagined what would happen if the parent nucleus of 1,2-diamine reagents like DMPEDA was changed to “roofed” skeleton (rel-(11 R, 12S)-9,10-dihydro-9,10-ethanoanthracene-11,12-diamine, 1) as shown in Figure 1. The roofed cyclic system enforces the two amino groups to locate syn-periplaner in 1. In addition, the “roofed” side is blocked. It would be expected that the diastereomeric aminals could be separated easier due to their fixed structures.

Figure 1. Structures of “roofed” cis-diamine 1 and its aminal derivatives

Figure 2. Structures of the derivatives of “roofed” diamines

Accordingly we would expect readily separation of the diastereomeric aminals. Thus, we aimed to determine the chemical and physical properties of cyclic aminals with 1, including the absolute configuration of aldehydes by NMR analyses. We first planned to analyze the cyclic aminals from achiral 1 with isobutyraldehyde. Several racemic aldehydes including lilial 6 would be subjected to the derivatization and the structures of some animals would be determined through 2D NMR analyses. Furthermore, chiral 2–5 would be obtained by acidic methanolysis [7] of the corresponding aminals with (R)-7.

We first investigated the cyclic aminal with symmetric “roofed” cis-diamine 1 [8] and isobutyraldehyde. The condensation proceeded smoothly to form two diastereomeric aminals 1a (endo, major) and 1b (exo, minor) in a ratio 4:1 in the ¹H NMR spectrum in CDCl₃ (Figure 1). The geminal methyl protons in 1a and 1b showed an upfield shift at 0.64 and 0.75 ppm compared to those (δ 1.06) of isobutyraldehyde, respectively. C1-H signal of 1a appeared at 2.56 ppm while that of 1b was 3.52 ppm, suggesting C1-H of 1b was magnetically shielded by π-electrons of the benzene ring, which enabled to elucidate the configurations of 1a and 1b as endo- and exo-configurations, respectively. These aminals were not separated on a silica-gel TLC plate by several solvent systems. However, these experiments revealed that ”roofed” diamine 1 effectively distinguishes aminal diastereomers in ¹H NMR spectroscopic manner.

We, therefore, introduced sulfonyl groups to 1 to improve the stereoselectivity in aminalization reaction and to add chirality. “Roofed” mono-N-sulfonyl cis-diamines (2 [9], 3 [9,10], 4 [9], 5 [11]) were prepared from 1 and corresponding sulfonyl chlorides, which were reacted with isobutyraldehyde to form stereoselectively endo-type cyclic aminals (imidazolidines) 2a-5a as shown in Figure 1. The physicochemical data for 2, 4 and 5 are reported because those were not found in literature. Hereinafter we will explain the structure of 2a as shown in Figure 1. (Some data is summarized in Supporting information.)

The coupling constant between the vicinal methine protons (C1-H and C2-H) of 2a is 3.1 Hz, which indicates that the two protons are gauche form. As a result, one methyl group would get close to one benzene ring of the reagent. This methyl appears at – 0.01 ppm, is supposed to be affected by the strong magnetic anisotropy of the faced benzene ring as shown in Figure 1. As a result, the chemical shift difference of the geminal methyl protons was very huge (δ – 0.01 and δ 0.82, Δδ = 0.83). The NOESY data support the above conformation. The detail is disclosed in the supporting information. We considered that, because of the steric hindrance of Ms group of RDA-Ms 2, endo-oriented aminalization occurred stereoselectively and the C₂-methine proton of aldehyde moiety faces to the Ms group. Moreover, the NMR spectral data of 3a-5a also supported the same conformation of them. Additionally, the aromatic rings of sulfonyl moieties seem to be hardly affected the chemical shifts of proton signals in aldehyde moieties. The chemical shift differences between the geminal dimethyl protons and carbons for 3a-5a (Figure 1) were 0.76–0.83 ppm (¹H)and 8.13–8.31 ppm (¹³C) , respectively. As a whole, the chemical shifts of any aldehyde moiety of prepared RDA-aminals in this study changed drastically, and it was suggested that the movement of them depends on the position of each proton in the derivative.

The racemic cis-diamines 2–5 were reacted with four racemic aldehydes (2-methylbutanal, 2-methylpentanal, 2-phenylpropanal, and lilial 6) and chiral (1 R)-myrtenal 7. The resulting 20 diastereomeric pairs were all separated on silica-gel TLC plates. This finding is very promising for the optical resolution of racemic aldehydes.

Racemic 2 was reacted with (R)-myrtenal 7 (1.3 equivalent amount of 2) to afford only two diastereomeric aminals 2b and 2c in Figure 2 in spite four diastereomers were possible. In particular, it will be worth mentioning that these could readily be separated on silica-gel TLC. Both NOESY spectra of 2b and 2c showed NOEs between (NH-CH-N/C11ʹ-H, C12ʹ-H) indicated both of them are the endo-aminals. The NOESY spectrum of 2c also showed NOEs between (CH₃SO₂/C1-H, C3-H, C7-H), (NH/C1-H, 9-CH₃), (C1-H/C3-H), and (8-CH₃/C2ʹ-H, C3ʹ-H). These data indicate that compound 2c is the aminal with (12S)-2. Interestingly, 2b enubilated in imine (Fig. 2) in CDCl₃ (see Supporting information). The following acidic hydrolysis of 2b and 2c gave each enantiomer of 2. In the similar manner, we succeeded in obtaining both enantiomers of 3–5 (see Experimental of Support information). These results suggested that the optical resolution of “roofed” mono-N-sulfonyl cis-diamines was possible by derivatization into the corresponding cyclic aminals with chiral aldehyde 7 and subsequent separation on silica-gel followed by acidic hydrolysis.

With the establishment of practical method of the optical resolution of ”roofed” mono-N-sulfonyl cis-diamines, we next moved to the examination of the optical resolution of racemic aldehyde by use of the reagents. So far we examined, most aminals prepared from 2–5 were stable under the chromatographic conditions and in CDCl₃. However, it was found that the aminal 5e prepared from 5 and racemic lilial 6 were readily isomerized under mild conditions (5% AcOH in CHCl₃ or reflux in toluene) as shown in Fig. 2. We presume the imine formation from the cyclic aminal resulted in this epimerization. Development of the effective hydrolysis of 5e and 5d without racemization would become a practical preparation protocol for resolution of racemic lilial 6.

In conclusion, the four “roofed” mono-N-sulfonyl cis-diamines (2–5) and the corresponding cyclic aminals with several aldehydes were synthesized. The detailed structures of each cyclic aminals were thoroughly analyzed utilizing 1D- and 2D-NMR. Each diastereomers of cyclic aminals derived from chiral aldehyde 7 and racemic “roofed” mono-N-sulfonyl cis-diamines (2–5) was detected as separated spots on the TLC and the NMR spectra of them were totally distinct each other. We expect that these findings obtained from this study are suggesting the possibility of development of a new method for elucidation of the absolute configuration of α chiral aldehydes and optical resolution of racemic aldehydes, if the appropriate procedure for hydrolysis of cyclic aminal is developed without racemization. After finding the above processes, the optical resolution of racemic aldehyde at α-position will be successfully performed by using the present method. The findings through this study might be seeds of deracemization reagents and new ligands.

Author contributions

Y. F designed the research plan, L.A. A, S. F. and Y. F. performed the experiments and analyzed data, and Y.F. supervised the research. L.A.A wrote the manuscript with assistance from Y.F.

Acknowledgments

We are grateful to Mr. Yusuke Takata and Dr. Eri Fukushi of the GC-MS and NMR Laboratory, Faculty of Agriculture, Hokkaido University for their skill and assistance in analyzing the NMR and MS data. A scholarship to L.A.A from JASSO is gratefully acknowledged.

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary material

Supplemental data for this article can be accessed here.

Additional information

Funding

L.A.A receives a scholarship from Japan Student Services Organization (JASSO).