4 Compounds of Group 15 (As, Sb, Bi) and Silicon Compounds
Ian Fleming, , Science of Synthesis, (2002) 4, 1.
The order in which the sections in this volume are presented, both the order of the four elements (As, Sb, Bi, Si), and the order of their compounds within the sections devoted to each element, is determined by the rules used throughout the series Science of Synthesis. Although admirably logical in itself, it is not designed to reveal underlying patterns in the chemistry of the four elements. The synthesis of the different classes of compounds, which is the main focus of the volume, is rarely an end in itself, although there are organoarsenic and organosilicon compounds in commercial production. More often than not, the main reason for making the compounds is their application in synthesis, for which an overview may be helpful.
The Organic Chemistry of Arsenic, Antimony, and Bismuth
The first three sections in this volume (Sections 4.1–4.3) record the major methods by which the compounds of the three group 15 elements (As, Sb, Bi) are synthesized, and each section includes a section or sections on their applications in organic synthesis. They are not, in fact, all that often used in organic synthesis, rarely having any advantage over the elements with which they can best be compared, and having the disadvantage of frequently being more expensive, toxic, and smelly.
Arsenic, Antimony, and Bismuth Compared with Phosphorus
A useful comparison for each of the three elements arsenic, antimony, and bismuth is with the more familiar chemistry of the element immediately above them in the periodic table, phosphorus. Each of these elements allows Wittig-like chemistry to be carried out, with subtle shifts in selectivity from that seen with phosphorus. Thus, stabilized ylides of arsenic and antimony, on the one hand, give Wittig-like reactions with carbonyl compounds,[1,2] but unstabilized ylides based on arsenic react like sulfonium ylides and undergo the Corey–Chaykovsky reaction with carbonyl compounds (Scheme 1).[3] In between, partly stabilized ylides can be fine-tuned to react either way. These trends are of great fundamental interest, but it is rare in synthesis for there to be any advantage over the phosphorus- and sulfur-based reactions.
Scheme 1 Wittig-like and Corey–Chaykovsky-like Reactions of Arsenic and Antimony Ylides[1–3]
On the other hand, the triorgano derivatives of arsenic(III), and to a lesser extent antimony and bismuth, do have useful differences in their capacities as ligands for transition metals, allowing fine tuning in the stability and reactivity of catalysts derived from their coordination compounds.
Arsenic, Antimony, and Bismuth Compared with Metals
There is an orderly decrease in electronegativity and an increase in metallic character down the series arsenic, antimony, bismuth, and a corresponding increase in effectiveness as substitutes for metals. Like silicon they all have rather feeble metallic properties, but that can be a virtue in controlling reactivity.
Alkylantimony(V) compounds react directly with acid chlorides,[4] and allylantimonium salts react directly with aldehydes in the same way as alkyl- and allylmetal compounds in general (Scheme 2).[5] Similarly, allylbismuth reagents, prepared in situ, react with aldehydes like other allylmetal compounds.[6] In these reactions the alkyl and allyl groups are nucleophilic, and the antimony or bismuth is behaving as a mild electrofugal group.
Scheme 2 Alkylantimony, Allylantimony, and Allylbismuth Reagents as Carbon Nucleophiles[4–6]
Triorganoarsenic(III) compounds react with bromine to give dibromotriorganoarsenic(V) products, which readily undergo reductive elimination (in transition-metal terminology) to give organic bromides and the bromodiorganoarsenic(III) compound (Scheme 3).[7] Sometimes, as when one of the organic groups is a vinyl or substituted vinyl group, the intermediate is not observed. Again, the arsenic is behaving overall as an electrofugal group.
Scheme 3 Bromodealkylation of Triorganoarsenic Compounds[7]
Arsenic, Antimony, and Bismuth in Redox Chemistry
All three elements arsenic, antimony, and bismuth have a rich redox chemistry. In the first place, they have several subclasses with one or more bonds to oxygen, in addition to the (III) and (V) coordination levels, leading, with arsenic, for example, to families of arsanes, arsinous acids, arsonous acids, arsinic acids, and arsonic acids. These, and their counterparts with antimony and bismuth, are handled separately in the various sections devoted to each of the three group 15 elements.
Redox reactions with a change, one way or the other, between the element(III) and element(V) oxidation states are easy with these three elements, as can be seen in the large number of methods for the synthesis of the various subclasses of compounds which involve a change from (III) to (V), or the reverse. The compounds of arsenic, antimony, or bismuth are rarely used simply as reducing or oxidizing agents. Two exceptions might be the use of bismuth(V) reagents as selective oxidants for converting acyloins into 1,2-diketones,[8] and for converting alcohols in general into ketones in the presence of such easily oxidizable groups as thiols and indoles.[9]
A more obvious application using the easy change of oxidation state, in addition to the reductive elimination mentioned above, is to take advantage of the capacity of a bismuth(V) compound to deliver one of its five ligands, effectively as an electrophile, with concomitant lowering of the oxidations state from bismuth(V) to bismuth(III) (Scheme 4).[10] In this property, bismuth resembles lead(IV) and thallium(III), with the metal center being nucleofugal, characteristic of elements low in the periodic table in high oxidation states.
Scheme 4 An Arylbismuth(V) Compound as an Electrophilic Arylating Agent[10]
The Organic Chemistry of Silicon
The organic chemistry of silicon lacks the redox capacities of the three elements arsenic, antimony, and bismuth, but is considerably more diverse in the influence it has on the chemistry of the functional groups present in silicon-containing molecules, with substantial chemistry from the many different arrangements by which the silicon atom may be connected to the organic functionality. These principles are best revealed by successively comparing silicon with carbon, the obvious first choice, with metals in general, since silicon is a metalloid, and finally, and in some ways most tellingly, with hydrogen.
Silicon Compared with Carbon
In one sense the chemistry of silicon, relative to that of its vertical neighbor in the periodic table, carbon,[11] is simple. Silicon chemistry is dominated by the chemistry of single bonds to silicon, whereas a high proportion of the chemistry of carbon is based on its functional groups having double bonds. Kipping's disappointment that the silicon equivalent of acetone is an unreactive polymer is well-known (Scheme 5).[12]
Scheme 5 A Ketone and Silicone Compared[12]
The chemistry of double bonds to silicon has comparatively little application in organic synthesis, but this topic is included briefly in this volume (Sections 4.4.1 and 4.4.2), because it is interesting, because it illustrates the difficulty in making and handling compounds with double bonds, and because there are potential applications. Similarly, Section 4.4.3 on the chemistry of silylenes, that is, silicon(II), is brief. These sections concentrate on the preparation of the stable silenes and silylenes, but there is a substantial chemistry hardly mentioned, mostly high-temperature, gas-phase chemistry, in which they are reactive intermediates. This, too, might have substantial importance for organic synthesis, but this possibility, largely studied so far to unravel mechanistic complexities, has had little development as a synthetic method as yet.
Silicon Compared with a Metal
In another sense, the chemistry of silicon is not so simple, since, as an element more electropositive than carbon, and as a second-row element, it can easily accept more than the usual four ligands. This capacity to be a Lewis acid gives it a chemistry quite unlike that of carbon. Thus tetrachlorosilane (Section 4.4.12) is a moderately strong, and selectively oxophilic Lewis acid, allowing it to be used to make trichlorosilyl enol ethers, and function as a bridge in Mukaiyama-like aldol reactions (Scheme 6). The silicon in the bridge is able to accept yet another ligand, which gives an opportunity to use a chiral Lewis base with which to induce high levels of enantiocontrol.[13]
Scheme 6 The Trichlorosilyl Group as a Lewis Acid in an Aldol Reaction
Trimethylsilyl trifluoromethanesulfonate (Section 4.4.14) is a rather weaker but still useful Lewis acid, and there is much scope for tailoring the Lewis acidity to achieve intermediate levels of reactivity by changing ligands. The Lewis acidity of silicon compounds is usually noticeable only when the silicon atom carries one or more electronegative ligands, but the ease with which a β-oxy anion and a trialkylsilyl group undergo syn elimination, sometimes known as the Peterson elimination (Section 4.4.37), is another consequence of Lewis acidity detectable even without the presence of electronegative substituents.[14]
As well as being oxophilic, silyl groups are halophilic, and, most unlike carbon, fluorophilic, making fluoride ion a highly selective nucleophile for silicon, a property that stems in part from the extraordinarily strong bond fluorine makes to silicon.
Another manifestation of its Lewis acidity is the easy nucleophilic substitution that takes place at silicon (Scheme 7), much easier than at carbon, a property it shares with all the second-row elements.[15]
Scheme 7 Comparison of an SN2 Reaction at Silicon with Carbon
Thus, at one extreme, a silyl triflate (Section 4.4.14) is an aggressive silylating agent, and, at the other extreme, even carbon groups can sometimes be displaced as anions from silicon, even if they have no anion-stabilizing groups attached. Naturally, this is much easier if they do have anion-stabilizing groups, with the result that α-silyl carbonyl compounds (Section 4.4.35), and cyano- (Section 4.4.24), ethynyl- (Section 4.4.30), aryl- (Section 4.4.33), vinyl- (Section 4.4.34), benzyl- (Section 4.4.39), and allylsilanes (Section 4.4.40), can often be persuaded to react as carbon nucleophiles when fluoride ion is used to displace the silyl group (Scheme 8).
Scheme 8 Fluoride Ion as a Powerful Nucleophile for Silicon
Similarly, fluoride ion attack on a trimethylsilylmethyl group attached to an iminium ion (Section 4.4.28) is a source of azomethine ylides for 1,3-dipolar cycloadditions. Another application is the Brook rearrangement (Scheme 9) found in the chemistry of α-silyl alcohols (Section 4.4.28) and acylsilanes (Section 4.4.25), where an α-alkoxide attacks the silyl group, rendering the α-carbon nucleophilic.[16] Such behavior in the carbon series is almost unimaginable, just as the Peterson elimination has no carbon counterpart.
Scheme 9 The Brook Rearrangement[16]
As electropositive elements go, however, silicon is one of the least metallic. This weakness as a metal allows it to be joined to a more metallic metal comparatively easily, and then used as a nucleophile, with considerable potential to tailor the reactivity by the choice of metal. There is significant silicon-anion chemistry found for its compounds with lithium, copper, zinc, aluminum, boron, and tin (Sections 4.4.6–4.4.11), and even to another silicon atom (Section 4.4.54.4.5).
A high proportion of its properties used in organic synthesis can be identified with silicon being the least metallic of metals. In this context, the comparison of silicon with carbon is less illuminating than the comparison with fully metallic elements. Thus silyl enol ethers (Section 4.4.16), both those derived from esters and those derived from ketones and aldehydes, are used as carbon nucleophiles in many of the same reactions as metal enolates [reactions such as the Mukaiyama aldol (Scheme 10),[17] Mannich, and Michael reactions, and similar manifestations of d2 reactivity] except that the silyl enol ethers need more powerful electrophiles than metal enolates, a condition usually achieved by coordinating the electrophile to a Lewis acid.[18] Similarly, the O-silyl ethers of secondary amides, and their aromatic counterparts such as 2-pyridone, are also useful nitrogen nucleophiles (Section 4.4.15), conspicuously in nucleoside synthesis.[19]
Scheme 10 Silyl Enol Ethers as d2 Synthons
A silicon hydride (Section 4.4.4) is a hydride source less powerful than a metal hydride, but more powerful than elemental hydrogen. Silicon hydrides do not reduce carbonyl groups, unless activated by fluoride ion, but they do reduce such cationic electrophiles as carbocations and protonated carbonyl groups (Scheme 11). Unlike the hydrides of more metallic elements, silicon hydrides react relatively slowly with protic acids, allowing acids to be used to generate cations in the presence of the silicon hydride reducing agent.[20] Silicon hydrides are also less effective as hydrogen-atom donors than most metal hydrides, but this, too, can be used when it is important to slow down the hydride-transfer process in radical cyclizations.[21,22]
Scheme 11 Silicon Hydrides as Reducing Agents
The comparatively metallic properties of a silyl group also show up in the way a silyl group stabilizes a β-carbocation by hyperconjugation, the well-known β-effect (Scheme 12),[23] which is also manifest in such thermodynamically stabilized compounds as α-silyl carbonyl compounds (Section 4.4.35) and silylketenes (Section 4.4.31).
Scheme 12 The β-Effect[23]
Nevertheless, all these properties are muted relative to those found for metals such as lithium, magnesium, and even zinc, allowing most silicon-containing organic compounds to be handled with ease, without special precautions for avoiding protic solvents or oxygen, and to be carried through many steps in an organic synthesis without the silicon—carbon bond being disturbed.[24] Silicon appears to be comparatively benign in its environmental impact, and ways of using silicon in organometallic chemistry, as an alternative to tin and other more troublesome metals, are constantly being sought and found, as in the palladium-catalyzed cross coupling of organic halides and triflates with ethynylsilanes (Section 4.4.30) and vinylsilanes (Section 4.4.34) (Scheme 13).[25] This reaction, the Hiyama coupling, has been improved recently by the discovery that aryl- and vinylsilanes which are also silanols are excellent substrates.[26,52]
Scheme 13 Electrofugal Silyl Groups in Cross-Coupling Reactions[25]
A less obvious way in which silicon can be compared with a metal is in its capacity to be a π-electron-withdrawing group, stabilizing α-anions (Section 4.4.23). This is not a property most organic chemists readily associate with an electropositive element, but in fact most metals, when they have an empty p orbital, have this capacity. In the case of silicon, it is not an empty p orbital that provides the stabilization, but the bonds from the silicon to its other ligands, since they are almost always polarized away from silicon towards ligands less electropositive than silicon (Scheme 14).[27]
Scheme 14 Stabilization of an α-Anion by a Silyl Group
This polarization achieves, by negative hyperconjugation, the same effect as an empty p orbital, but to a reduced extent, and an α-silyl group is therefore mildly anion-stabilizing. This property shows up in the comparative stability of diazo(trimethylsilyl)methane (Section 4.4.26), an apparently benign substitute for diazomethane itself,[28] in the capacity of vinylsilanes (Section 4.4.34) to be attacked by nucleophiles leading to addition, in the increased acidity of silanols,[29] and in the reduced Lewis basicity of silyl ethers and silylamines relative to free alcohols, ethers and amines.[30] Silyl groups on oxygen and nitrogen reduce, but do not remove, the nucleophilicity of the oxygen and nitrogen atom, and silyl ethers (Section 4.4.17) are less nucleophilic than either the free alcohol or its metal alkoxide. The residual nucleophilicity allows silyl ethers to be used as oxygen nucleophiles in the absence of protons, which can have considerable advantages, and comparable chemistry is found for silylamines (Section 4.4.21), silyl azides (Section 4.4.20), silyl sulfides and selenides (Section 4.4.19), and silylphosphorus compounds (Section 4.4.22).
Silicon Compared with Hydrogen
Perhaps the most telling of all comparisons is between silicon and hydrogen, where having a silyl group in place of a hydrogen atom provides chemo-, regio-, and stereocontrol that the presence of a hydrogen atom does not allow. There are two contrasting ways in which this property shows up: one when the silicon is on an electronegative atom, and another when it is on carbon.
In general, a silyl group on oxygen can be thought of as a large, greasy, but rather feeble proton.[15] Silyl ethers, for example, are commonly used as protected versions of the corresponding alcohols, because the silyl group is more difficult to remove from an oxygen atom than a proton is (Scheme 15).
Scheme 15 The Silyl Group as a Feeble Proton on Oxygen
The silyl group does not indulge so significantly in bonding analogous to hydrogen bonding; therefore, silylated alcohols are more volatile than the free alcohols. By varying the degree of steric hindrance around the silicon atom, a large family of protecting groups, well classified with respect to their acid and base stability, has been developed (Section 4.4.17).[31] It is also easy to assemble two different groups attached by oxygen atoms to silicon, allowing a silicon diether (Section 4.4.13) to be an easily disassembled bridge between reacting partners.[32] Silyl ethers are important not only for alcohols but also for enols, as already seen with silyl enol ethers. The comparison with metal enolates is only one way of thinking about silyl enol ethers; another equally valid comparison is with enols themselves. The nucleophilicity of a silyl enol ether in its d2 reactions is close to that of an enol. The difference is that it is not, in general, possible to have specific enols, and so the reactions of enols are not regiocontrolled in the way that metal enolate reactions can be.[33] The only easily handled specific enols are those made from 1,3-dicarbonyl compounds, and even they are not always present in high concentration. In contrast, it is relatively easy to form specific silyl enol ethers; they do not need extra functional groups, and they retain the regiochemistry with which they were generated. Furthermore, silyl enol ethers can react with a number of electrophiles such as tertiary alkyl halides, and acetals in the presence of Lewis acids, these being electrophiles incompatible with other d2 synthons.[34,35]
A silyl group on carbon is still large and greasy, but, provided that it is not much more hindered than a trimethylsilyl group, it can be thought of, in contrast to an oxygen-bonded silyl group, as a super-proton.[15] If the nucleophile is a halogen- or oxygen-based nucleophile, a trimethylsilyl group is more easily removed from carbon than a proton with the same functional relationship (Scheme 16).
Scheme 16 The Silyl Group as a Super-Proton on Carbon
Thus, any pathway that leads to a carbocation β to a trimethylsilyl group is controlled in its outcome by the loss of the silyl group, placing a double bond specifically between the carbon atom carrying the positive charge and the carbon atom carrying the silyl group. This kinetic instability (Scheme 16), taken with the thermodynamic stability of β-silyl ethyl cations (Scheme 12), gives rise to a large amount of highly controlled carbocation chemistry, especially in the electrophilic substitution reactions of aryl- (Section 4.4.33), vinyl- (Section 4.4.34), allyl- (Section 4.4.40), allenyl- (Section 4.4.32), and propargylsilanes (Section 4.4.38).[36–38] These compounds show enhanced levels of nucleophilicity relative to their counterparts with a hydrogen atom in place of the silyl group, they show a high tendency to give alkene products rather than those of nucleophilic capture, and they show a high level of regiocontrol in where the double bond is located in the product. Furthermore, some of these reactions show a high level of stereocontrol stemming from the size, and possibly the electropositive nature, of the silyl group. Thus, illustrating all these points, allylsilanes react with electrophiles, not only regioselectively because of the β-effect, but stereospecifically anti, and the result is substitution, because of the ease with which the silyl group is lost (Scheme 17).[39–41] There is comparable chemistry with allenylsilanes[42] and propargylsilanes.[43]
Scheme 17 The Stereochemistry of the SE2′ Reaction of Allylsilanes[39–41]
Similarly, vinylsilanes react regioselectively with most electrophiles, again because of the β-effect, with retention of configuration for the same reason, and with overall substitution, because of the ease with which the silyl group is lost (Scheme 18).[37] There is comparable chemistry, without the stereochemical component, with aryl- and ethynylsilanes.[36]
Scheme 18 The Stereochemistry of Electrophilic Substitution of Vinylsilanes[37]
Similarly, the enolates of β-silyl carbonyl compounds (Section 4.4.41) are alkylated anti with a high level of stereoselectivity.[44] Other examples of cationic chemistry controlled by silicon are the conversion of silylepoxides into aldehydes or ketones,[45] the Stork–Colvin reaction (Section 4.4.29), and the ease of alkene formation from β-silyl alkyl halides (Section 4.4.36), as well as the ease of controlling cationic rearrangements in γ-silyl alkyl halides and alcohols (Section 4.4.42).[46]
Silicon in Redox Chemistry
Finally, the capacity of a silyl group to remain bonded to carbon, allowing it to be carried through many steps of an organic synthesis, especially when the functionality is not such as to provide a pathway for the loss of the silyl group, can be put to powerful, and essentially unique, use by the oxidation of silyl groups carrying at least one electronegative substituent to give alcohols (Section 4.4.18) (Scheme 19).[47–49] Since a phenyl group, and a number of other groups, can be easily removed from the silicon by electrophilic substitution to create the electronegative substituent on silicon, a group such as dimethylphenylsilyl can be used as a masked hydroxy, despite its being, in almost every respect, the chemical opposite: It is large, electropositive, Lewis acidic (if anything), nonpolar in the sense that its being tetrahedral contributes almost no dipole, and is not involved in hydrogen bonding. In contrast, a hydroxy group is small, electronegative, Lewis basic, polar, and a hydrogen-bonding donor and acceptor.
Scheme 19 Silyl-to-Hydroxy Conversion: The Tamao–Fleming Reaction[47–49]
The organic chemistry of silicon is orderly, highly predictable, and remarkably versatile. It goes from strength to strength, showing little sign of falling from favor as a weapon in the organic chemist's armory.[50,51]
References
Related Information
Return to Top