Reaction-Controlled Silylation

In this scheme, a functionality that does not react with a silylating reagent is converted to a reactive group to incorporate Si selectively in the exposed area, or conversely, a reactive functionality is rendered inert by a radiation-induced process so that Si is selectively introduced in the unexposed region. 

---

Scheme 2.2 illustrates several examples of the Mukaiyama aldol reaction. Entries 1 to 3 are cases of addition reactions with silyl enol ethers as the nucleophile and TiCl4 as the Lewis acid. Entry 2 demonstrates steric approach control with respect to the silyl enol ether, but in this case the relative configuration of the hydroxyl group was not assigned. Entry 4 shows a fully substituted silyl enol ether. The favored product places the larger C(2) substituent syn to the hydroxy group. Entry 5 uses a silyl ketene thioacetal. This reaction proceeds through an open TS and favors the anti product. 

There are, however, serious problems that must be overcome in the application of this reaction to synthesis. The product is a new carbocation that can react further. Repetitive addition to alkene molecules leads to polymerization. Indeed, this is the mechanism of acid-catalyzed polymerization of alkenes. There is also the possibility of rearrangement. A key requirement for adapting the reaction of carbocations with alkenes to the synthesis of small molecules is control of the reactivity of the newly formed carbocation intermediate. Synthetically useful carbocation-alkene reactions require a suitable termination step. We have already encountered one successful strategy in the reaction of alkenyl and allylic silanes and stannanes with electrophilic carbon (see Chapter 9). In those reactions, the silyl or stannyl substituent is eliminated and a stable alkene is formed. The increased reactivity of the silyl- and stannyl-substituted alkenes is also favorable to the synthetic utility of carbocation-alkene reactions because the reactants are more nucleophilic than the product alkenes. 

Recently, silicon-tethered diastereoselective ISOC reactions have been reported, in which effective control of remote acyclic asymmetry can be achieved (Eq. 8.91).144 Whereas ISOC occur stereoselectively, INOC proceeds with significantly lower levels of diastereoselection. The reaction pathways presented in Scheme 8.28 suggest a plausible hypo thesis for the observed difference of stereocontrol. The enhanced selectivity in reactions of silyl nitronates may he due to 1,3-allylie strain. The near-linear geometry of nitrile oxides precludes such differentiating elements (Scheme 8.28). 

Although in the recent years the stereochemical control of aldol condensations has reached a level of efficiency which allows enantioselective syntheses of very complex compounds containing many asymmetric centres, the situation is still far from what one would consider "ideal". In the first place, the requirement of a substituent at the a-position of the enolate in order to achieve good stereoselection is a limitation which, however, can be overcome by using temporary bulky groups (such as alkylthio ethers, for instance). On the other hand, the ( )-enolates, which are necessary for the preparation of 2,3-anti aldols, are not so easily prepared as the (Z)-enolates and furthermore, they do not show selectivities as good as in the case of the (Z)-enolates. Finally, although elements other than boron -such as zirconium [30] and titanium [31]- have been also used succesfully much work remains to be done in the area of catalysis. In this context, the work of Mukaiyama and Kobayashi [32a,b,c] on asymmetric aldol reactions of silyl enol ethers with aldehydes promoted by tributyltin fluoride and a chiral diamine coordinated to tin(II) triflate... 

On the other hand, the method of Mukaiyama can be succesfully applied to silyl enol ethers of acetic and propionic acid derivatives. For example, perfect stereochemical control is attained in the reaction of silyl enol ether of 5-ethyl propanethioate with several aldehydes including aromatic, aliphatic and a,j5-unsaturated aldehydes, with syir.anti ratios of 100 0 and an ee >98%, provided that a polar solvent, such as propionitrile, and the "slow addition procedure " are used. Thus, a typical experimental procedure is as follows [32e] to a solution of tin(II) triflate (0.08 mmol, 20 mol%) in propionitrile (1 ml) was added (5)-l-methyl-2-[(iV-l-naphthylamino)methyl]pyrrolidine (97b. 0.088 mmol) in propionitrile (1 ml). The mixture was cooled at -78 °C, then a mixture of silyl enol ether of 5-ethyl propanethioate (99, 0.44 mmol) and an aldehyde (0.4 mmol) was slowly added to this solution over a period of 3 h, and the mixture stirred for a further 2 h. After work-up the aldol adduct was isolated as the corresponding trimethylsilyl ether. Most probably the catalytic cycle is that shown in Scheme 9.30. 

The asymmetric catalytic aldol reaction of silyl allenolates ICH=C=CR2OSiMe3 with aldehydes R CHO has been achieved by Li et al. by using N-C3F7CO oxazaborolidine as the catalyst [43], The fluoroacyl group of the catalyst was found to be crucial for control of enantioselectivity. The reaction provides the first enantioselective approach to / -halo Baylis-Hillman-type adducts. 

A versatile strategy for efficient intramolecular oc-arylation of ketones was achieved by the reaction of silyle enol ethers with PET-generated arene radical cations. This strategy involved one-electron transfer from the excited methoxy-substituted arenes to ground-state DCN [42]. Pandey et al. reported the construction of five- to eight-membered benzannulated as well as benzospiroannulated compounds using this approach (Sch. 20) [42a]. The course of the reaction can be controlled via the silyl enol ether obtained... 

The nitrido complex was applied to the direct asymmetric animation with a silyl enol ether as a substrate. Although several examples for achiral aminations of silyl enol ethers have been reported [32], an asymmetric version of reagent-controlled reaction has not appeared except for the one recent example [33] and the diastereoselective reactions with silyl enol ethers having a chiral auxiliary [34], The amination, which is presumed to take place via an aziridine intermediate [5g, lid,32], proceeded smoothly to give the A-tosylated a-aminoketone in 76% yield with 48% ee. When the same silyl enol ether was treated with complex 15 under Carreira s condition, the TV-trifluoroacetylated a-aminoketone was obtained in 58 % yield with 79 % ee (Scheme 24). 

Reaction control experiments show that the syn-isomer is the kinetic product of the addition. The results achieved are in accordance with Houk s52 analysis of electrophilic attack on alkenes bearing an adjacent stereogenic center. It predicts that the kinetic product is the syn-isomer and the anti-isomer 3 A should be kinetically favored in the addition to silyl ether 1 (R1 = TBDMS R2 = CHj),... 

The utility of BF3-OEt2, a monodentate Lewis acid, for acyclic stereocontrol in the Mukaiyama aldol reaction has been demonstrated by Evans et al. (Scheme 10.3) [27, 28]. The BF3-OEt2-mediated reaction of silyl enol ethers (SEE, ketone silyl enolates) with a-unsubstituted, /falkoxy aldehydes affords good 1,3-anti induction in the absence of internal aldehyde chelation. The 1,3-asymmetric induction can be reasonably explained by consideration of energetically favorable conformation 5 minimizing internal electrostatic and steric repulsion between the aldehyde carbonyl moiety and the /i-substituents. In the reaction with anti-substituted a-methyl-/ -alkoxy aldehydes, the additional stereocontrol (Felkin control) imparted by the a-substituent achieves uniformly high levels of 1,3-anti-diastereofacial selectivity. 

Since the middle of the 198O s remarkable progress has been achieved in the development of asymmetric aldol reactions of silyl enolates. In the beginning of this evolution, chiral auxiliary-controlled reactions were extensively studied for this challenging subject [106]. As new efficient catalysts and catalytic systems for the aldol reactions were developed, much attention focused on catalytic enantiocontrol using chiral Lewis acids and transition metal complexes. Thus, a number of chiral catalysts realizing high levels of enantioselectivity have been reported in the last decade. 

Chiral 3,5,6-trihydroxyheptanoic acids are potentially useful intermediates for the synthesis of natural products. The backbone can be constructed by a titanium-mediated aldol reaction of silyl enol ether 536 with 929. The syn adduct 949 is formed exclusively, as predicted by the chelation-controlled Cram cyclic model. 

The aldol reaction and related processes have been of considerable importance in organic synthesis. The control of syn/anti diastereoselectivity, enantioselectivity and chemoselectivity has now reached impressive levels. The use of catalysts is a relatively recent addition to the story of the aldol reaction. One of the most common approaches to the development of a catalytic asymmetric aldol reaction is based on the use of enantiomerically pure Lewis acids in the reaction of silyl enol ethers with aldehydes and ketones (the Mukaiyama reaction) and variants of this process have been developed for the synthesis of both syn and anti aldol adducts. A typical catalytic cycle is represented in Figure 7.1, where aldehyde (7.01) coordinates to the catalytic Lewis acid, which encourages addition of the silyl enol ether (7.02). Release of the Lewis acid affords the aldol product, often as the silyl ether (7.03). 

Photochemical [2 + 2] ring closures have been used to synthesize cyclophanes and related polycyclic ring systems containing silicon in the ring using triplet sensitizers such as benzophenone or dicyanonaphthalene as shown in Scheme 67. The photochemistry of simpler allylsilanes was also investigated, as shown in the scheme. In these reactions the silyl group and its substituents acted as tethers which controlled the orientation of the two carbon-carbon double bonds involved in the photocyclization. 

Mikami reported that BINOL derived titanium complex efficiently catalyzed the aldol reaction of silyl enol ether with excellent control of both absolute and relative stereochemistry [106] (Scheme 14.37). The reaction was proposed to proceed via a prototropic ene reaction pathway that is different from that of Mukaiyama aldol condensation. A cyclic antiperiplanar transition-state model was proposed to explain the pref erential formation of the syn diastereomer from either (E)- or (Z)-silyl enol ethers [106]. Further modifications of the catalyst system include the use of perfluorophenols and other activating additives [107], or performing the reaction in supercritical fluids [108]. Furthermore, the nucleophile could be extended to enoxysilacyclobutane derivatives [109]. 

Desilylative coupling of cinnamyltrimethylsilane results in 3,6-diphenyl-1,5-hexadiene as shown in Scheme 2.60. The cross-coupling reaction of silyl enol ethers and allylic silanes proceeds chemoselectively to give Y,8-unsaturated ketones, in which the oxovanadium(V) oxidatively desilylates the more readily oxidizable organosilicon compound [126], Their redox potentials determine whether they will act as a radical generator or acceptor. These redox potentials can be predicted from calculated ionization potentials. VO(OR)Cl2 is a versatile oxidant, which can induce chemoselective coupling via the oxidative desilylation of a variety of organosilicon compounds under controlled conditions, as shown in Scheme 2.61. 

The reaction of silyl and tin enolates with nitrosobenzene, the so-called nitroso aldol reaction, was studied by Yamamoto and coworkers aiming at an overall enantioselective hydroxylation [254, 255]. This approach faces, however, the problem that in a noncatalyzed reaction, the nucleophilic silyl and stannyl enol ethers 514 attack the nitrogen atom of the ambident electrophile nitrosobenzene 515 so that the formation of hydroxyamino ketones 516 results [254a]. Fortunately, the authors developed suitable procedures wherein, under catalysis by chiral silver-bisphosphane complexes, aminooxy ketones 517 result in high ambidoselectivity. Alternatively, a controlled attack at nitrogen under formation of hydroxyamino ketones also became feasible by tuning of the catalytic system [254b,c] (Scheme 5.127). 

Paterson et al. [98] in their attempt used a similar disconnection for rhizopodin as described by Menche (fragments 144 and 149) (Scheme 2.151). However, unlike, Menche, they used silyl ketene acetal 16 in an asynunetric VMAR for the addition to ( )-iodoacrolein (142) to obtain dioxinone 143 in 94% ee. Methanolysis removed the aceto-nide, and the subsequent Narasaka reduction [99] provided the syn-diol 144 in 80% yield and a 10 1 selectivity for the desired isomer. The synthesis of segment 149 started with aldehyde 145, which was ultimately derived from Roche ester. Carbon chain extension was achieved through a chelation-controlled Mukaiyama aldol reaction with silyl ketene acetal 146, which installed the new chiral center with excellent stereocontrol (20 1 dr). For the installation of the third secondary alcohol, six-membered lactone 148 was obtained by treatment with K COj in methanol. Subsequent borane reduction provided stereospecifically the desired alcohol, which was then further transformed to the desired acid (149). 

Abe, M. and Nojima, M., The Paterno-Biichi photocydoaddition reactions of silyl ketene acetals controlling factors on the regio- and stereoselectivity in the oxetane formation,/. Synth. Org. Chem. Jpn., 59, 855, 2001. 

Several books and other publications on the stabilization of polymers have been mentioned in this chapter. In addition, two patents are noteworthy. The one by Hamilton discusses protection of polymer blends against transesteri-fication. For example,blends of PC with semi-crystalline PEST (e.g., PET) finds application in the automotive industry. Since the resistance to solvents depends on the PET crystallinity, which decreases with advancing transreactions, its control is essential. The patent specifies addition of a sUyl phosphate compound for inhibiting the ester-carbonate interchange. The stabilization is achieved by deactivation of the residual metaUic catalyst in the reaction with silyl phosphate stabilizers or their mixture. 

The ketone is added to a large excess of a strong base at low temperature, usually LDA in THF at -78 °C. The more acidic and less sterically hindered proton is removed in a kineti-cally controlled reaction. The equilibrium with a thermodynamically more stable enolate (generally the one which is more stabilized by substituents) is only reached very slowly (H.O. House, 1977), and the kinetic enolates may be trapped and isolated as silyl enol ethers (J.K. Rasmussen, 1977 H.O. House, 1969). If, on the other hand, a weak acid is added to the solution, e.g. an excess of the non-ionized ketone or a non-nucleophilic alcohol such as cert-butanol, then the tautomeric enolate is preferentially formed (stabilized mostly by hyperconjugation effects). The rate of approach to equilibrium is particularly slow with lithium as the counterion and much faster with potassium or sodium. 

---