This account outlines our studies on organic synthesis inspired by the relevance to carbohydrates in various ways. The contents include, (1) my personal encounter with carbohydrate synthesis, (2) cationic hafnocene species in glycoside synthesis, (3) aryl C-glycoside antibiotics, (4) O→ C-glycoside rearrangement, (5) [2+2+2] approach to phenylnaphthalene motif in polyaromatic natural products, (6) regioselective O-glycosylation of pseudo-C2 symmetric phenanthrenediols in the benanomicin-pradimicin antibiotics and TAN-1085, and (7) oligo (polyphenol) synthesis based on the analogy to carbohydrate chemistry.
Rh (I)-catalyzed hydroacylation of 4, 6-dienals proceeded smoothly, giving cycloheptenone derivatives. It was found that the size of the ring formed in this hydroacylation was dependent on the presence of a substituent at C 7 position and the geometry of the olefin. During ongoing investigation of this reaction, we found that Rh (I) -catalyzed cycloisomerization of 1, 3-dienes with alkenes in a tether gave cyclopentene derivatives in good yields. It was interesting that the existence of a heteroatom between a 1, 3-diene moiety and alkene in the tether affected the reaction course and that only a [4+2] cycloaddition product was produced. These two reactions proceeded using the same cationic Rh (I) catalyst under almost the same reaction conditions. We therefore investigated a new cascade reaction by a combination of these reactions. It was found that this reaction gave the bicyclo [5.3.0] decenone derivative in a stereoselective manner, and the Thorpe-Ingold effect played an important role in the second cycloisomerization step of this cascade cyclization. By the use of this cascade cyclization as a key step, the synthesis of (±) -epiglobulol has been accomplished.
In 1978, Nozaki, Oshima, and Takai reported that CH2X2-Zn-TiCl4 reagent is effective for Wittig-type methylenation of carbonyl compounds. We investigated this reagent system in detail and concluded that CH2 (ZnI) 2 and β-TiCl3 are real reactive species. Chemoselective methylention using dizinc reagent is also shown.
In recent years, DNA has attracted much attention as a conductive biopolymer. Recent efforts to elucidate the mechanism of long-range hole transport in DNA has encouraged us in the view that DNA can be a good mediator for hole transport by the selection of an appropriate sequence. However, when natural DNA is used as a molecular wire, serious unavoidable oxidative degradation of G bases occurs. In addition, the hole transport in natural DNA is strongly influenced by the sequence and the transport distance. Of great importance in the realization of a real DNA wire is the molecular design of an artificial nucleobase that can effectively mediate hole transport and, at the same time, is not oxidatively decomposed. We now report on a protocol for designing an artifi-cial nucleobase that can act as an effective mediator for long-range hole transport without subsequent decomposition. Additionally, we also describe the applications of the artificial hole-transporting DNA to (i) combinational logic gates, (ii) drug-releasing systems, and (iii) genotyping biofilms on a gold electrode.
On the bases of the concept that a coordination of 1, 2-diols 1 with an appropriate metal ion (Mn+) might form an activated five-membered intermediate in which the OH group is much activated than those of non-activated 1, we examined a monobenzoylation of 1 by benzoyl chloride using tin and copper ions as Mn+. As the results, the monobenzoylation of 1 was achieved with high selectivity even in the presence of monools. The monobenzoylation reactions were then applied to kinetic resolution of dl-1 and asymmetric desymmetrization of meso-1 using a chiral organotin or BOX (bisoxazoline) -CuCl2 as a catalyst. In particular, the asymmetric outcome using (R, R) -Ph-BOX-CuCl2 catalyst was promising and attractive. That is, the kinetic resolution of dl-1 (with up to >99.9% ee for monobenzoylation of hydrobenzoin), its application to a synthesis of D-inositol-1-phosphate (100% ee), the asymmetric desymmetrization of meso-1 (with up to 94% ee for monobenzoylation of meso-1, 2-hydrobenzoin and 93% ee for monocarbamoylation of meso-1, 2-cyclooctanediol), and the asymmetric desymmetrization of meso-1, 3-diols (with up to 55% ee for the monobenzoylation of ethyl 2-ethyl-2-hydroxymethyl-3-hydroxypropanoate) were achieved. Also, the method was applied to electrochemical oxidation of 1 and its asymmetric version (38% ee for the asymmetric desymmetrization of meso-hydrobenzoin and 72% ee for the kinetic resolution of dl-hydrobenzoin).
The Ohrui-Akasaka method is the only one methodology to solve the intrinsic problem of diastereomer method. The Ohrui-Akasaka ester discriminates an asymmetric center remote from up to 24 carbons. The gauche effect combined with CH-π interaction of the ester sets anthracene ring close to acyl side chain. The distinctive conformation provides a chiral field to a remote asymmetric center. This review describes the examples of the determination of the absolute configuration of the natural products, leustroducsin B, S 365 A and plakoside A, by using the Ohrui-Akasaka method.