Invention of Synthetic Reactions Based on Σ−Bond Activation |

No.166 No.166 ResearchResearch ArticleArticle Invention of Synthetic Reactions Based on σ−Bond Activation Tamejiro Hiyama RDI Professor Research and Development Initiative, Chuo University 1. Introduction 2. Organosilicon Reagent/Fluoride Ion During my studies with Professor Hitosi Nozaki, we It would be fascinating if both the cationic and anionic had weekly discussions about the present and future of our species were generated by the cleavage of element–element chemistry. It may have been in the summer of 1980 that bonds. I pictured a reaction scheme in which disilane and a Professor Nozaki said one day “When I gave a lecture in fluoride ion produce a silyl anion and fluorosilane, with the Europe, I asked the audience ‘Do you know a reaction system fluorosilane behaving similar to an electrophilic silyl cation. I that involves both carbanionic and carbocationic species?’ No found that when hexamethyldisilane and tetrabutylammonium one gave me the correct answer.” He then suggested the reaction fluoride (TBAF) are mixed in hexamethylphosophric triamide of allylic acetate with trialkylaluminum, R3Al, whereby an (HMPA) in the presence of an aldehyde, carbonyl attack of the allylic cation and alanate are produced and the alanate delivers trimethylsilyl anion occurs, followed by O-trimethylsilylation. an alkyl anion to the allylic cationic center. The audience then After aqueous workup, we obtained an α-silyl alcohol, understood his meaning. The reaction of organoaluminums is demonstrating that the designed reaction2 works well. well-documented, but Professor Nozaki’s view is quite unique Heterolysis of an element–element bond is conceptually and straightforward. I am convinced this new point of view, rewarding and is closely related to the recently established or philosophy, is essential for devising and developing novel activation of molecular hydrogen by a frustrated base pair synthetic reactions. We have observed that the reaction of (Scheme 2). trimethylaluminum with diastereometric cyclopropylmethyl- type acetates provides the same product, endo-methylated norcarane, and have shown that the reaction1 proceeds via a common cyclopropylmethyl-type carbocationic intermediate (Scheme 1). OCOCH3 CH3 OCOCH3 Al(CH3)3 Me Al Me Al O CH3 3 3 H Al H H3C O CH3 + CH3 Scheme 1. TBAF H H (5 mol%) H3O Si Si + n-C10H21CHO n-C10H21 C O n-C10H21 C OH HMPA Me3Si SiMe3 Me3Si 67% Scheme 2. 2 No.166 During this research, Dr. Kiyosi Kondo, who later became and trifluorovinyllithium are both unstable organometallic the president of the Sagami Chemical Research Center (SCRC), species and were thought only to be generated at extremely often visited Kyoto. Dr. Kondo repeatedly and earnestly invited low temperatures. Thus, the idea that removal of a metal me to move to Kyoto. I finally agreed, and in 1981 started counter-ion from carbenoids (i.e., organometals with a leaving my own research group at SCRC at the age of 34. Dr. Michio group on the α-carbon) to generate naked carbanionic species Obayashi, one of my first PhD coworkers, soon discovered that with a leaving group on the α-carbon led to the discovery of hexamethyldisilane/TBAF reductively disilylates butadienes. useful synthetic reactions that can be conducted at ambient For example, 2,3-dimethyl-1,3-butadiene is converted to temperature. We further extended the concept of organosilicon (E)-1,4-bis(trimethylsilyl)-2,3-dimethyl-2-butene.2 Product reagents/fluoride ion-based reaction to hydrosilane/fluoride ion formation is believed to involve a trimethylsilyl anion which in the hope of generating a naked hydride. Dr. Makoto Fujita, undergoes single electron transfer (SET) to butadiene to give a currently a professor at the University of Tokyo, carried out trimethylsilyl radical and a butadiene radical anion. Coupling of reactions of α-chiral ketones with dimethylphenylsilane/fluoride 3 these at the terminal carbon should give a TMSCH2-substituted ion and observed threo-selective reduction. We expected that allylic anion, which attacks hexamethyldisilane to give the a pentacoordinate hydrosilicate should be sufficiently bulky product and a silyl anion. In HMPA, SET should readily occur to stereoselectively reduce ketone carbonyls according to the from a naked (i.e., in the absence of a metal counter-ion) hard Felkin-Ahn model transition state. silyl anion. Regardless, you may consider the net transformation as an insertion of butadiene between the Si-Si σ-bond of It is reasonable to assume that hydrido and electronegative hexamethyldisilanes (Scheme 3). fluoro groups in trigonal bipyramidal pentacoordinate silicate should be both apical, and organic substituents locating at We set about generalizing the basic concept of equatorial positions. Since the bond angles of H–Si–organic organosilicon compound/fluoride ion-based reaction to groups in the trigonal bipyramide are 90°, the pentacoordiante generate metal-free (naked) anionic species, and found that hydride is likely bulkier than tetrahedral (109.5°) borohydrides. halomethyl(trimethyl)silanes and trifluorovinyl(trimethyl) Consequently, carbonyl groups are reduced with a high level silane undergo an addition reaction3 with aldehydes, even at of stereoselectivity. In any event, we failed to produce a naked room temperature, upon treatment with tris(dimethylamino) hydride (Scheme 4, 5).4 sulfur difluoro(trimethyl)silicate (TASF) in HMPA or THF. This finding is remarkable given that halomethyllithiums TBAF + Si Si Si HMPA Si 81%, 99 : 1 Scheme 3. OH O OH HSiMe Ph HSiMe Ph Z 2 Z 2 Z R F– R H+ R Me >95 : 5 Me >93 : 7 Me Z = OR', NR'2, CONR'2 Scheme 4. H Me PhSiMe2-H + F Ph Si H + PhSiMe F Me 2 F bulky reductant naked hydride? counter ion: Bu4N or (Et2N)3S Scheme 5. 3 No.166 No.166 3. Cross-Coupling Reaction with Organosilicon accessible compared to Suzuki-Miyaura coupling. Accordingly, Compounds while Professor Kyoko Nozaki was working with me in Kyoto, she showed that one organic group on tetraorganosilanes can We next envisaged what would happen if a transition be in situ converted to a halo- or alkoxy group by the action metal catalyst was present in the organosilicon/fluoride ion of a fluoride ion. We recently found that C–N coupling7 is system. Is the organic group on silicon successfully transferred conveniently achieved with trimethylsilylamines and allows to the transition metal, or does the fluoride ion simply attack easy synthesis of triarylamines which display a characteristic the transition metal center to give metal fluoride and lose reactivity profile different from that obtained using Hartwig– catalytic activity? To find out, Dr. Yasuo Hatanaka, who is Buchwald coupling (Scheme 7). currently a professor at Osaka City University, mixed Pd(II) complex, TASF as a fluoride source, trimethylvinylsilane and In 2004, Dr. Yoshiaki Nakao, who is currently a professor 1-iodonaphthalene in HMPA, and isolated 1-vinylnaphthalene at Kyoto University, suggested using HOMSi (dimethyl(o- in a nearly quantitative yield.5 His findings demonstrate that hydroxymethylphenyl)silane) tetraorgansilane reagents8, our working hypothesis was correct; this new silicon-based predicting that they would be converted to a pentacoordinate cross-coupling reaction was later named Hiyama Coupling. structure by intramolecular nucleophilic attack of the hydroxyl The TASF fluoride source may be replaced with TBAF or group. Protection of the hydroxyl group with a common KF. To our surprise, KOH could also activate the C–Si bond. protecting group significantly stabilizes the silicon reagents However, attempts to extend this approach to substituted under chromatographic separation and common reaction vinylsilanes were unsuccessful. This limitation was overcome conditions; upon deprotection, the reagents become coupling- by introducing a fluorine or alkoxy group on silicon: this active. After the reaction, the silicon residue is recovered as effectively extended the scope of the cross-coupling reaction a cyclic silyl ether, which is reconverted back to the HOMSi and demonstrated that a heteroatom on silicon assists formation reagent by reaction with organolithium or –magnesium of pentacoordiate silicates, an essential species for successful reagents. Recently, direct C–H silylation of (hetero)aromatics reaction. A theoretical study6 and our intuition both suggested and terminal alkenes has been demonstrated starting with a a four-membered cyclic transition state for the transmetalation THP-protected HOMSi hydride reagent and an iridium catalyst from Si to Pd. Fluoride is believed to play a key role in the (Scheme 8). transmetalation. The following scheme summarizes the silicon- based cross-coupling reaction (Scheme 6). Cross-coupling reaction with HOMSi reagents has recently been applied to polymer synthesis under mild conditions. The Halosilanes and alkoxysilanes are sensitive to moisture resulting π-conjugated polymers find important applications as and are thus incompatible with chromatographic purification, light-emitting and solar-cell materials (Scheme 9).9 so cross-coupling with these organosilicon reagents is less 1 2 Pd cat 1 2 R –SiYmRn + R –X R —R + SiXYmRn F 1 2 Pd(0) 1 R R X R2 SiYmRn: R : 2 1 R F R Si F SiMe3 vinyl, alkynyl, silyl, NR2 reductive X F oxidative Pd Y SiMe2F, SiMe2OR alkenyl, aryl elimination Si 1 Y addition 1 SiF2R aryl (R = Et, n-Pr, Cy) R Si Y R 1 2 Y Y allyl (R = Ph) R Pd R Y X Pd R2 F Y X = F etc SiF3 alkyl, allyl X Si Y transmetalation Y Scheme 6. Ar–Br Ar + Pd(dba)2 (1 mol%) XPhos (2 mol%) N SiMe3 CsF (1.1 eq) P N DMI, 100 °C Ar = Ph 90% p-Tolyl 97% p-Ph2N-C6H4 95% Xphos 2-Naph 99% Scheme 7. 4 No.166 O R MgX/H2O R R' Si or Me 2 LiAlH4, alkyne, Pt cat. PdCl cat. 2 K CO (CuI cat.) 2 3 DMSO ligand O HO Me I R' Si K R Si Me R, R' = alkenyl, aryl Me2 R Orthogonally THPO CH3COO t-BuPh2SiO protected HOMSi reagents R Si R Si R Si Me2 Me2 Me2 Deprotection PPTS/MeOH OH, HAI(i-Bu)2 F Scheme 8.

Load more