Tải bản đầy đủ - 0 (trang)
METAL ALKYLS, ARYLS, AND HYDRIDES AND RELATED σ -BONDED LIGANDS

METAL ALKYLS, ARYLS, AND HYDRIDES AND RELATED σ -BONDED LIGANDS

Tải bản đầy đủ - 0trang

54



METAL ALKYLS, ARYLS, AND HYDRIDES AND RELATED σ -BONDED LIGANDS



considered a founder of organometallic chemistry. It was only with Victor Grig­

nard’s discovery (1900) of the alkylmagnesium halide reagents, RMgX, however,

that organometallic chemistry began to make a major impact through its appli­

cation in organic synthesis. The later development of organolithium reagents is

associated with Schlenk (1914) and Ziegler (1930). Ziegler was also instrumental

in showing the utility of organoaluminum reagents.

Metal Alkyls as Stabilized Carbanions

Grignard reagents, RMgX, provided the first source of nucleophilic alkyl groups,

Rδ− , to complement the electrophilic alkyl groups, Rδ+ , available from the alkyl

halides. Metal alkyls result from combining an alkyl anion with a metal cation.

In combining, the alkyl anion is stabilized to a different extent depending on the

electronegativity of the metal concerned. Alkyls of the electropositive elements

of groups 1–2, as well as Al and Zn, are sometimes called polar organometallics

because the alkyl anion is only weakly stabilized and retains much of the strongly

nucleophilic and basic character of the free anion. Polar alkyls all react with

traces of humidity to hydrolyze the M−C bond to form M−OH and release RH.

Air oxidation also occurs very readily, and so polar organometallics must be

protected from both air and water. Alkyls of the early transition metals, such

as Ti or Zr, can also be very air and water sensitive, but as we move to less

electropositive metals by moving to the right and down the periodic table, the

compounds become much less reactive, until we reach Hg, where the Hg−C bond

is so stable that [Me−Hg]+ cation is indefinitely stable in aqueous sulfuric acid

in air. As we go from the essentially ionic NaCH3 , to the highly polar covalent

Li and Mg species, to the essentially covalent late metal alkyls, the reactivity

falls steadily along the series, showing the effect of changing metal (Fig. 3.1).

The inherent stability of the R fragment plays a role, too. As an sp3 ion, CH3 −

is the most reactive. As we move to sp2 C6 H5 − and even more to sp RC≡C− ,

where the lone pair of the anion is stabilized by being in an orbital with more s

character, the intrinsic reactivity falls off. The same trend makes the acidity of the

hydrocarbons increase as we go from CH4 (pKa = ∼50) to C6 H6 (pKa = ∼43)

and to RC≡CH (pKa = ∼25), so the anion from the latter is the most stable and

least reactive.

Following the successful syntheses of main-group alkyls, many attempts were

made to form transition metal alkyls. Pope and Peachey’s Me3 PtI (1909) was

an early but isolated example of a d-block metal alkyl. Attempts during the

1920s through 1940s to make further examples of d-block alkyls all failed. This

was especially puzzling because by then almost every nontransition element had

been shown to form stable alkyls. These failures led to the view that transi­

tion metal–carbon bonds were unusually weak; for a long time after that, few

serious attempts were made to look for them. In fact, we now know that such

M−C bonds are strong (30–65 kcal/mol is typical). It is the existence of several

easy decomposition pathways that makes many transition metal alkyls unstable.

Kinetics, not thermodynamics, was to blame for the synthetic failures. This is



55



TRANSITION METAL ALKYLS AND ARYLS



Nucleophilic

reactivity





M



CH3



M C6H5



M



Na



CCR



Mg



Ti



Cu



Au



2







1

Electronegativity



Pt



FIGURE 3.1 Schematic diagram showing qualitatively how the nucleophilic reactivity

of main-group and transition metal alkyls to protons or air oxidation depends on the alkyl

itself and the electronegativity of the metal.



fortunate because it is easier to manipulate the system to block decomposition

pathways than it is to increase the bond strength. In order to be able to design

stable alkyls, we must look at some of these pathways to see how they can be

inhibited. This example of the historical evolution of our ideas implies that just

as some of the early assumptions in this area proved to be wrong, some of our

ideas today will probably turn out to be wrong, too—the problem is we do not

know which ones!

β Elimination

The major decomposition pathway for alkyls is β elimination3 (Eq. 3.1), which

converts a metal alkyl into a hydridometal alkene complex. We study it in detail

in Section 7.4. For the moment we need only note that this very common mech­

anistic type can occur whenever

1. The β carbon of the alkyl bears a hydrogen substituent.

2. The M−C−C−H unit can take up a roughly coplanar conformation,3b

which brings the β hydrogen close to the metal.

3. There is a vacant site on the metal, symbolized here as �, cis to the alkyl.

4. The reaction is much more rapid for d 2 and higher metals than for d 0 and

main-group alkyls.



56



METAL ALKYLS, ARYLS, AND HYDRIDES AND RELATED σ -BONDED LIGANDS



Requirements 1 and 3 arise because it is the β hydrogen of the alkyl that is

transferred to the metal to give the product hydride. The geometry of the situa­

tion means that a cis site is required on the metal and a coplanar M−C−C−H

arrangement in the ligand. The elimination is believed to be concerted; that is,

C−H bond breaking and M−C and M−H bond making happen at the same time.

H2C



CH2



LnM



H



H2C

b elimination



CH2



LnM



LnM



H + H2C



CH2



(3.1)



H

3.1



The term “vacant site” of requirement 3 needs some clarification. It does not

simply mean that there should be a gap in the coordination sphere large enough

to accommodate the incoming ligand. There must also be an empty orbital ready

to accept the β-H, or more exactly, the pair of electrons that constitutes the

β-C−H bond. The electron count of the product alkene hydride is 2e more than

that of the alkyl starting material. An 18e alkyl is much more reluctant to β­

eliminate via a 20e intermediate than is a 16e alkyl, which can go via an 18e

alkene hydride. Even if the alkene subsequently dissociates, which is often the

case, we still have to stabilize the transition state leading to the alkene hydride

intermediate if we want the reaction to be fast. An 18e alkyl, on the other hand,

is said to be coordinatively saturated. By this we mean that an empty orbital is

not available. Some 18e alkyls do β-eliminate, but detailed mechanistic study

often shows that the prior dissociation of some ligand is required in the ratedetermining step.

Main-group alkyls can also β-eliminate (e.g., Eq. 3.2), but this usually happens

much more slowly. The reason for this difference is believed to be the greater

ability of d-block metals to stabilize the transition states involved that resemble

agostic alkyl complexes such as 3.7.

reflux



[(EtMeCH)3 Al]2 −−−→ [(EtMeCH)2 Al(µ-H)]2 + butene



(3.2)



Stable Alkyls

To have a kinetically stable alkyl, we must block the β-elimination pathway for

decomposition. This can happen for

1. Alkyls that have no β hydrogen:

WMe6



Ti(CH2 Ph)4



C2 F5 Mn(CO)5



W(CH2 SiMe3 )6

LAuCF2 CF2 Me



TaCl2 (CH2 CMe3 )3

Pt(C≡CCF3 )2 L2



Pt(CH2 COMe)Cl(NH3 )2



57



TRANSITION METAL ALKYLS AND ARYLS



2. Alkyls for which the β hydrogen is unable to approach the metal as a result

of the geometry or bulk of the ligand:

PtH(C≡CH)L2



PdPh2 L2



Cr(CMe3 )4



Cr(CHMe2 )4



CpL3 MoCH=CHCMe3

3. Alkyls in which the M−C−C−H unit cannot become syn-coplanar:3c



Ti



Cr

[Cr(1-adamantyl)4]



L2Pt



4



4



Ti(6-norbornyl)4



L2Pt(CH2)3



The first two would give “forbidden” anti-Bredt olefins if they were to

β-eliminate.

4. An 18-electron species with firmly bound ligands, which will not dissociate

to generate a vacant site:

Cp(CO)2 FeCH2 CH3 ,

3.2



O

N





H



O



N



Co

N

O

H



Cp(CO)3 MoCH2 CH3 ,

3.3



Cp(CO)IrPrH,



H 2O N



[Cr(H2O)5Et]2+



3.5



3.6



O

3.4



5. Some d 0 alkyls:

Cl

CH2



CH2



PMe2

Ti

PMe2



H

Cl



Cl

3.7



Some of these cases call for special comment. WMe6 , like WH6 , has a trigonal

prismatic structure 3.8,4a not the octahedral structure usually found for ML6



58



METAL ALKYLS, ARYLS, AND HYDRIDES AND RELATED σ -BONDED LIGANDS



species. Albright and Eisenstein4b had previously predicted that d 0 MX6 species

would be trigonal prismatic where X is not a π donor. Methyl compounds are

especially numerous, and the small size of this ligand allows the formation of

polyalkyls. Often, substitution with electron-withdrawing or bulky groups (e.g.,

−CH2 Ph, −CH2 SiMe3 ) also gives stable alkyls. The vinyl and phenyl groups

both have β hydrogens, but they do not β-eliminate easily. One reason may be

that the β hydrogens are further from the metal in these sp2 -hybridized systems

with 120◦ angles at carbon, less favorable for delivery of the β-H than in the

sp3 ethyl group (109◦ ). In addition, as is the case for other electronegative alkyl

groups, the phenyl and vinyl groups have stronger M−C bonds than does the

ethyl group.



W





3.8



The iso-propyl and tert-butyl chromium complexes are unusual. Presumably,

their steric bulk prevents the β-C−H bond from reaching the metal. These

structures seem to be sterically saturated. The examples containing noncopla­

nar M−C−C−H groups mainly involve cyclic alkyls, in which the rigidity of

the ring system holds the M−C−C−H dihedral angle near 60◦ and far away

from the value of 0◦ required for β elimination. The fourth group includes those

systems with no vacant site (3.2, 3.3, and 3.5) and others that have such a site,

but not cis to the alkyl (3.4, assuming that the aqua ligand can dissociate). Com­

pound 3.6 is not an 18e species, but as a d 3 Cr(III) complex it is coordination

inert (see Section 1.4).

Agostic Alkyls

Rarer are those species in which all the criteria appear to be favorable but in

which β elimination still does not occur. In some of these (e.g., 3.7) the β-C−H

bond is bound to the metal in a way that suggests that the alkyl is beginning

the approach to the transition state for β elimination, but the reaction has been

arrested along the way. These agostic alkyls can be detected by X-ray or neutron

crystal structural work and by the high-field shift of the agostic H in the proton

NMR. The lowering of the J (C,H) and ν(CH) in the NMR and IR spectra,

respectively, on binding is symptomatic of the reduced C−H bond order in the

agostic system.5 The reason that β elimination does not occur in 3.7 is that the

d 0 Ti has no electron density to back donate into the σ ∗ orbital of the C−H bond.

This back donation breaks the C−H bond in the β-elimination reaction, much

as happens in oxidative addition (see Eq. 1.5). Agostic binding of C−H bonds

also provides a way to stabilize coordinatively unsaturated species. They are also



59



TRANSITION METAL ALKYLS AND ARYLS



found in transition states for reactions such as alkene insertion/β elimination

either by experiment (see Fig. 12.4) or in theoretical work.6

We saw earlier that we need a 2e vacant site (an empty d orbital) on the

metal for β elimination. Now we see that we also need an available electron pair

(a filled d orbital) for breaking the C−H bond. There is a very close analogy

between these requirements and those for binding a soft ligand such as CO. Both

processes require a metal that is both σ acidic and π basic. In the case of CO,

binding leads to a reduction in the CO bond order. In the case of the β-C−H

bond of an alkyl group, this binding can reduce the C−H bond order to zero,

by cleavage to give the alkene hydride complex. Alternatively, if the metal is a

good σ acid but a poor π base, an agostic system may be the result, and the

C−H bond is only weakened, not completely broken. Many of the characteristic

reactions of organometallic chemistry require both σ -acid and π-base bifunctional

character. This is why transition metals, with their partly filled d orbitals, give

these reactions so readily.

Halide Elimination

β Elimination of halide can also occur. Early transition metals, such as Ti, the

lanthanides, and the actinides do not tend to form stable fluoroalkyls because

the very high M−F bond strengths of these elements encourages β elimination

of the halide. The late transition metals have weaker M−F bonds and do form

stable fluoroalkyls. Not only do these ligands lack β-Hs, but the M−C bond

strengths are very high, as is also true for other alkyls MCH2 X, where X is any

electronegative group. CF3 , like PF3 , can also act as a π acceptor via the σ ∗

orbitals of the C−F bond (see Section 4.2), which also makes the M−C bond

stronger for the π-basic late metals. The C6 F5 group forms extremely stable

aryls with the late transition metals in which an aryl π ∗ orbital acts as electron

acceptor.7

Reductive Elimination

A second very common decomposition pathway for metal alkyls is reductive

elimination (“red. elim.” in Eq. 3.3).8 This leads to a decrease by two units in

both the electron count and the formal oxidation state. (This is why the reaction is

labeled “reductive.”) We study it in detail in Chapter 6. In principle, it is available

to all complexes, even if they are d 0 or 18e, provided a stable oxidation state

exists two units more reduced than the oxidation state in the starting alkyl. In

fact, in many instances reductive elimination is not observed, for example, if X in

3.9 is a halogen. The reason is that for alkyl halides, the position of equilibrium

for Eq. 3.3 usually lies well over to the side of 3.9; in other words, 3.9 is usually

more stable thermodynamically. Some examples of the loss of alkyl halide are

known, however.

red. elim.



Ln M(Me)X −−−−→ Ln M + MeX

3.9, 18e



3.10, 16e



(3.3)



60



METAL ALKYLS, ARYLS, AND HYDRIDES AND RELATED σ -BONDED LIGANDS



On the other hand, when X = H, the reaction is usually both kinetically facile

and thermodynamically favorable, so isolable alkyl hydrides are rare. Where

X = CH3 , the thermodynamics still favor elimination, but the reaction is generally

much slower kinetically. It is often the case that reactions involving a hydrogen

atom are much faster than those involving any other element; this is because H

carries no electrons other than bonding electrons, and these are in a 1s orbital,

which is capable of making and breaking bonds in any direction in the transition

state. The sp3 orbital of the CH3 fragment is directed in space, and so there can

often be poorer orbital overlap in the transition state.

Stability from Bulky Substituents

Bulky ligands provide a general strategy for stabilizing many different classes

of organometallic complex. Associative decomposition pathways for alkyls, such

as by reaction with the solvent or with another molecule of the complex, can

also be important, especially for 16e metals. These can often be suppressed with

bulky coligands. For example, square planar Ni(II) alkyls are vulnerable to attack

along the z direction perpendicular to the plane. The o-tolyl complex 3.11, in

which this approach is blocked, is more stable than the analogous diphenyl,

3.12, for example. This steric factor has made the use of bulky alkyl groups,

such as neopentyl (CH2 CMe3 ) or trimethylsilylmethyl (CH2 SiMe3 ) common in

organometallic chemistry.



L



L

Ni

L



Ni

L



3.11



3.12



Where β elimination cannot occur for the reasons discussed above, α elimina­

tion sometimes takes over. This leads to the formation of species called carbenes,

which have M=C double bonds. The first step in the thermal decomposition

of Ti(CH2 t-Bu)4 is known to be α elimination to Ti(=CHt-Bu)(CH2 t-Bu)2 .

Similarly, attempts to prepare Ta(CH2 t-Bu)5 led to formation of the carbene

complex, t-BuCH=Ta(CH2 t-Bu)3 . Carbenes and α elimination are discussed in

Sections 11.1 and 7.4.

Where a heteroatom such as N or O is present to activate the adjacent C−H

bonds for reaction, double C−H bond cleavage can occur at the same carbon.9 In

Eq. 3.4, the first cleavage, an oxidative addition, and the second, an α elimination,

can be observed stepwise for R = H. Even for the ArNEt2 analog (R = CH3 ),

where there is a choice between α elimination and β elimination in the second

step, the product still comes exclusively from α elimination. In Eq. 3.5, the



61



TRANSITION METAL ALKYLS AND ARYLS



carbene is again formed but the hydrogen produced is now trapped by half of

the Ru(II) as a dihydrogen complex.

CH2R



CH2R

N



[IrH2(Me2CO)2(PPh3)2]+



N



+



CH2R



O



−H2



L



N



N

L



Ir



CHR



H



(3.4)

CH2R

a

elimination



+



H



L



N



N

L



Ir



CR



H



(L = PPh3; R = H or CH3)

H

[RuHCl(PiPr3)2]2



O



[RuH(H2)Cl(PiPr3)2]



+



Cl



PR3 O



Ru



R 3P



(3.5)

The α heteroatom may stabilize the alkyl by allowing back donation into the

C-X σ * to be discussed in Chapter 4 (Fig. 4.3), while the carbene is additionally

stabilized by X to C(pπ ) donation (see 11.1).

Preparation of Metal Alkyls

The chief methods for the synthesis of alkyls involve (1) an R− reagent, (2) an

R+ reagent, (3) oxidative addition, and (4) insertion. Typical examples of these

are shown in Eqs. 3.6–3.15:

1. From an R− reagent (nucleophilic attack on the metal):

LiMe



WCl6 −−−→ WMe6 + LiCl

ZnMe2



NbCl5 −−−→ NbMe2 Cl3 + ZnCl2



(3.6)

(3.7)



2. From an R+ reagent (electrophilic attack on the metal):

MeI



Mn(CO)5 − −−−→ MeMn(CO)5 + I−



(3.8)



62



METAL ALKYLS, ARYLS, AND HYDRIDES AND RELATED σ -BONDED LIGANDS

Ph2 I+



CpFe(CO)2 − −−−→ Cp2 Fe(CO)2 Ph + PhI



(3.9)



−CO



CF3 COCl



[Mn(CO)5 ]− −−−−→ CF3 COMn(CO)5 −−−→ CF3 Mn(CO)5 (3.10)



3. By oxidative addition:

MeI



IrCl(CO)L2 −−−→ MeIrICl(CO)L2

MeI



PtL4 −−−→ MePtIL2



(3.11)

(3.12)



(L = PPh3 )

MeI



Cr2+ (aq) −−−→ CrMe(H2 O)5 2+ + CrI(H2 O)5 2+



(3.13)



PtHCl(PEt3 )2 + C2 H4 −−−→ PtEtCl(PEt3 )2



(3.14)



4. By insertion:



CH2 N2



Cp(CO)3 MoH −−−→ Cp(CO)3 MoCH3



(3.15)



A Grignard or organolithium reagent usually reacts with a metal halide or a

cationic metal complex to give an alkyl, often by nucleophilic attack on the metal.

Alternatively (case 2), a sufficiently nucleophilic metal can undergo electrophilic

attack. Both these pathways have direct analogies in reactions that make bonds to

carbon or nitrogen in organic chemistry (e.g., the reaction of MeLi with Me2 CO

or of NMe3 with MeI). Transfer of an alkyl group from one metal, such as Zn,

Mg, or Li, to another, such as a transition metal, is called transmetalation. In

Eq. 3.10, we use the fact that acyl complexes can often be persuaded to lose CO

(Section 7.1). This is very convenient in this case because reagents that donate

CF3 + are not readily available; CF3 I, for example, has a δ − CF3 group and a δ + I.



Oxidative Addition10

With the third general method of making alkyls, we encounter the very important

oxidative addition reaction, which we study in detail in Chapter 6. This term is

used any time we find that an X−Y bond has been broken by the insertion of

a metal fragment Ln M into the X−Y bond. X and Y can be any one of a large



63



TRANSITION METAL ALKYLS AND ARYLS



number of groups, some of which are shown in Eq. 3.16:



X



+



X



MLn



Y



Y

OS = 0

16e

CN = n



(XY = H2, R3C



MLn



OS = 2

18e

CN = n + 2



H, Cl



H, RCO



Cl, Cl



Cl, Me



I, R3Si



H)



(3.16)

Certain Ln M fragments are often considered carbenelike because there is

an analogy between their insertion into X−Y bonds and the insertion of an

organic carbene, such as CH2 , into a C−H, Si−H, or O−H bond (Eq. 3.17).

In Section 13.2, we will see how the isolobal principle allows us to understand

the orbital analogy between the two systems. There are several mechanisms for

oxidative addition (Chapter 6). For the moment we need only note that the overall

process fits a general pattern in which the oxidation state, the coordination num­

ber, and the electron count all rise by two units. This means that a metal fragment

of oxidation state n can normally give an oxidative addition only if it also has

a stable oxidation state of (n + 2), can tolerate an increase in its coordination

number by 2, and can accept two more electrons. This last condition requires

that the metal fragment be 16e or less. An 18e complex can still undergo the

reaction, provided at least one 2e ligand (e.g., PPh3 or Cl− ) is lost first. Oxidative

addition is simply the reverse of the reductive elimination reaction that we saw

in Section 3.1.



X



+









X

CH2



Y



Y

6e

CN = 2

(XY = R3C



H, R3Si



CH2



(3.17)



8e

CN = 4



H, RCO2



H, RO



H)



A special case of oxidative addition is cyclometalation, in which a C−H bond

in a ligand oxidatively adds to a metal to give a ring. Because of this ring

formation, the reaction can be highly selective, for example, only one of the nine

distinct CH bonds in benzoquinoline is cleaved when cyclometalation of Eq. 3.18

occurs. This kind of selectivity has been used in catalytic tritiation (Chapter 9)

and in the Murai reaction10 (page 428), in which aromatic ketones first undergo

selective cyclometalation and the resulting aryl group is then functionalized in



64



METAL ALKYLS, ARYLS, AND HYDRIDES AND RELATED σ -BONDED LIGANDS



subsequent steps.

+



N



N



(3.18)

PR3



[Ir(cod)(PR3)2]+



Ir



H2, 20°



R3 P



O H2

H



C2H4



L2(CO)Rh



H



L2(CO)Rh



CH2



H

CH2



(3.19)

C2F4



L2(CO)Rh



CF2



H

CF2



The third example of oxidative addition (Eq. 3.13) is a binuclear variant

appropriate to those metals (usually from the first row) that prefer to change

their oxidation state, coordination number, and electron count by one unit rather

than two.

Insertion

The fourth general route, insertion (studied in detail in Chapter 7), is particularly

important because it allows us to make an alkyl from an alkene and a metal

hydride. We shall see in Chapter 9 how this sequence can lead to a whole series

of catalytic transformations of alkenes, such as hydrogenation with H2 to give

alkanes, hydroformylation with H2 and CO to give aldehydes, and hydrocyanation

with HCN to give nitriles. Such catalytic reactions are among the most important

applications of organometallic chemistry. Olefin insertion is the reverse of the

β-elimination reaction of Section 3.1. Since we insisted earlier on the kinetic

instability of alkyls having β-H substituents, it might seem inconsistent that we

can make alkyls of this type in this way. In practice, it is not unusual to find that

only a small equilibrium concentration of the alkyl may be formed in such an

insertion. This is enough to enable a catalytic reaction to proceed if the alkyl is

rapidly trapped in some way. For example, in catalytic hydrogenation, the alkyl

is trapped by reductive elimination with a second hydride to give the product

alkane. On the other hand, if the alkene is a fluorocarbon, then the product of

insertion is a fluoroalkyl, and these are often very stable thermodynamically.11

Compare the reversibility of C2 H4 insertion with the irreversible formation of the



Tài liệu bạn tìm kiếm đã sẵn sàng tải về

METAL ALKYLS, ARYLS, AND HYDRIDES AND RELATED σ -BONDED LIGANDS

Tải bản đầy đủ ngay(0 tr)

×