Sei sulla pagina 1di 9

Chemistry 3820 Lecture Notes Dr. M.

Gerken Page195

16. Organometallics
We now move on to consider transition-metal organometallic compounds. Organometallic compounds are compounds that
contain a metal-carbon bond. Organic derivatives of the main-group metals were well known by the turn of the century.
Examples include alkyl lead and zinc compounds and, of course, Grignard reagents. All attempts to make analogous σ-bonded
alkyl derivatives with transition metals failed miserably. We now know why they failed, and today there are such derivatives, but
they are not the most common organometallics.

16.1 Carbonyl complexes


One important class of organometallics which were known, but whose structure and bonding were a mystery, were metal
carbonyls. Homoleptic carbonyls have the general formula: M(CO)n
These are also called binary metal carbonyls. There are also many mixed carbonyl compounds, but the binary compounds were
known first. Of these, the first known one was Ni(CO)4, discovered by Mond in 1890. A fascinating aspect of carbonyl chemistry
is that none of the main-group elements form metal carbonyls! So there are no zinc or magnesium derivatives. This started a
hypothesis that the set of metals which formed alkyl and carbonyl complexes were mutually exclusive. Nowadays we know this
to be untrue.
16.1.1 Binary carbonyls and the 18-electron rule
Stable organometallic complexes almost always obey the 18-electron rule. There are some important exception to this
statement, some of which were will cover in this course. But on the whole it is a good rule of thumb. In this way, organometallics
are more like organic compounds, which are also almost always closed-shell species, and less like classical coordination
complexes. Another important aspect of the most common transition-metal organometallic compounds is that the metal tend to be
in low oxidation states. This was once thought to be an absolute must, but recently important exceptions to this generalization
have been discovered. An important consequence of this which we will soon see is that the energy of the metal and ligand orbitals
are much better matched than for classical complexes. The stabilization of the low oxidation state, i.e., electron-rich metal centre,
is based on the π-acid character of CO. The bonding in carbonyl complexes is described as backbonding and synergistic bonding,
since σ-electron density is donated to while π-electron density is removed from the metal, resulting in strong bonds.
Counting electrons:
Up till now, we have often assigned
a charge to a metal to obtain its dn
configuration. This conceptualization is
not as important for organometallics, and
since there compounds are much more
covalent, it becomes more and more
arbitrary where to assign the electrons.
For this reason, it is the universal
practice of organometallic researchers to
use neutral atoms and ligands for
electron counting.
The neutral CO ligand is still a 2-
electron donor ligand, i.e., the lone pair
on the C atom is used to form a bond to
the metal.
Let’s do some examples of M(CO)n
structures and see how the rule applies.
16.1.2. Bonding modes of CO
In all complexes that we have considered
so far, carbonyl ligands exhibited a
terminal bonding mode. However, CO
can also act as a bridging ligand between
to metal centres coordinating the
carbonyl lone pair to both metal centres.
Chemistry 3820 Lecture Notes Dr. M. Gerken Page196

Carbonyl ligands can also cap a trigonal face formed by three metal centres, hence bridging three metal atoms. This time the lone
pair is shared by three metal centres.

16.1.3. 18-electron rule and polynuclear metal complexes


Some carbonyl complexes cannot attain an 18-electron count for neutral metal centres. Such metals form dinuclear
complexes (or complexes of higher nuclearity) with metal-metal bonds, e.g., Mn2(CO)10. Pentacarbonyl iron(0) can lose carbonyl
ligands upon formation of di- and even trinuclear complexes:


2 Fe(CO)5 → CO + Fe2(CO)9
HOAc
orange solid 90% yield,
m.p. 100 °C

Average Bond Lengths


Complex M-C/pm M-M/pm 10.1.4. Metric criteria for backbonding
Ni(CO)4 138.8 Occupation of the antibonding π2 orbital in CO should lengthen
Fe(CO)5 181.4 (ax) the bond, since the overall bond-order decreases. What is the experimental
181.2 (eq) evidence for that? Coordinated CO typically has a C-O bond length of 115
V(CO)6 200.8 pm. This is longer than in free carbon monoxide, 112.8 ppm. But it is not a
Cr(CO)6 191.3 very large difference. Note that in ketene, H2C=C=O, the CO bond is 117
Mo(CO)6 206 pm. This is a better reference than a ketone, because the latter involves a
Mn2(CO)10 179 (ax) 292.3 hybridization change at carbon as well. The small differences observed are
183 (eq) consistent with the generally smaller reduction sin bond length as bond
Fe2(CO)9 183.8 (term) 252.3 order increases. Consideration of the M-C bond lengths is even less
201.6 (br) useful, since no reliable standard exists. What should a Cr-C single bond
Co2(CO)8 180 (term) 252.2 length be taken as?
190 (br)
Ru3(CO)12 194.2 (ax) 285.2 10.1.5. Infrared stretching vibrations
192.1 (eq) The small difference in bond length between ketene and CO
Os3(CO)12 194.6 (ax) 287.7 nevertheless hides a significant change in bond-strength, as evidenced be
191.2 (eq) infrared spectroscopy, where free C≡O has ν = 2143 cm-1, while most
Ir4(CO)12 187 269 C=O bonds fall in around 1700-1800 cm-1. Metal carbonyls have infrared
Rh6(CO)16 186.4 (term) 277.6 bands at:
216.8 (tri. br) Organic ν(CO), cm-1 Metal complexes ν(CO), cm-1
The abbreviations are: ax=axial, eq=equatorial, CO
term=terminal, br=bridging, and tri.br. = triply CO(g) 2143 terminal 2200-1850
bridging Acetone 1750 Doubly bridging 1850-1750
(CH3)2O 1000 Triply bridging 1730-1620

Bond strength is given simply be Hooke’s law: ν(CO) = 1/(2πc) (kCO/µCO)1/2, where kCO is the force constant for the bond
and µCO is the reduced mass. This law fits nicely the range of organic CO bond types, and also can be used to advantage in metal
carbonyls. For example, we might expect the extent of synergistic bonding to increase with increasing charge on the metal. This
is seen in the series:
Mn(CO)6+ 2090 cm-1; Cr(CO)6 1990 cm-1; V(CO)6- 1860 cm-1
If we look back at the table of example binary carbonyls, we see that there are some CO’s which bridge between two metal
centres. In this bonding mode, the ligand acts as a 2-electron donor, and is formally a formaldehyde diradical. This means that
each metal receives only one electron. In some other complexes, CO exist as a triply bridging ligand, capping a trigonal face of
metal atoms, as in [Cp4Fe4(CO)4], in which the faces of the tetrahedron of iron atoms, each connected to a terminal Cp ligand, is
capped with CO ligands. This makes CO a derivative with an sp3 hybridized carbon.
Clearly the shift in the ν(CO) follow the predicted bond-order changes using such very simplistic arguments. There is a large
range within each group, which is readily explained by the wide variety of metal centres and charges involved.

16.1.6. Carbon-13 NMR spectroscopy


The 13C NMR spectra of metal carbonyls have been extensively examined. For diamagnetic complexes, they are typically found
in the ragne of 200 to 300 ppm. This usually allows carbonyls to be distinguished from other organic carbon in a mixed-ligand
metal complex, since these usually fall below 200 ppm. The exception is metal akylidenes, 250 to 350 ppm, which overlap the
region.
Chemistry 3820 Lecture Notes Dr. M. Gerken Page197
Chemistry 3820 Lecture Notes Dr. M. Gerken Page198

16.1.7. Reactions of metal carbonyls


Substitution reactions:
Cr(CO)6 → Cr(CO)5 + L → Cr(CO)5L
Such a substitution reaction is dissociative and can be multistep. It is initiated either thermally (as done in the lab) or
photochemically using Pyrex-filtered UV light. For example, the Cr(CO)3(CH3CN)3 complex is formed by refluxing the reaction
mixture in CH3CN solvent, while purging with N2 to remove any CO that comes off, and hence prevent any revere reaction. Most
multi-step substitutions go cis- or fac, so that the CO ligand can have as extensive backbonding as possible.
Reduction to carbonyl anions:
2Na + (OC)5Mn-Mn(CO)5 → 2Na[Mn(CO)5]
Such anions are synthetically useful reagents, and can be protonated.
Mn(CO)5- + HCl → [HMn(CO)5] + Cl-
The protonated carbonyl complexes are acidic and the pKa’s vary greatly from complex to complex, and decrease down the
periodic table. The 1H NMR spectra of these so-called transition metal hydride complexes are found somewhere in the range of -8
to -60 ppm.

Reactions of the CO ligand:


When the metal is relatively electron-poor,
the C atom of the metal carbonyl is
susceptible to nucleophilic attack. The
product complex is called a Fischer
carbine, or more correctly an alkylidene
complex. It contains a true metal-carbon
double bond. Later we will look at the
bonding in this type of complexes.

When the metal is quite electron-rich, as


indicated by a low value of ν(CO), the
oxygen end of the molecule becomes quite
negative and can react with electrophiles.
Typical examples include bridging
carbonyls, or those in anionic metal
complexes. And example is the following
reaction:
Fe2Cp2(CO)4 + 2AlEt3 → Fe2Cp2(CO)4· 2AlEt3
Chemistry 3820 Lecture Notes Dr. M. Gerken Page199

16.2 Metallocenes and other polyene complexes


In organic chemistry, the aromatic hydrocarbons are an important an special class of compounds. They are characterized by
planarity, a set of pπ orbitals which allow for electron delocalization and the formation of ring currents. This special property of
the aromatic hydrocarbons also leads to special types of organometallic derivatives. However, because electron count becomes
more flexible in the presence of a metal atom which can accept or donate electron density as required, complexes include both,
aromatic (4N+2) and anti-aromatic (4N) rings, and these are named together as cyclic polyenes.
Thus we will now look at the important class of organometallic complexes known as the metal polyenes, and especially at
the most important class of these, the metallocenes, known also as sandwich complexes.
16.1 Polyenes which form metal complexes
The following aromatic polyenes commonly form metal complexes:

The aromatic 6π systems are the most common, and cyclopentadienyl complexes are by far the most abundant of these. However,
transition-metal complexes of anti-aromatic rings are also well known, and in fact cylcobutadiene was first isolated as a
coordination complex.
10.2 Hapticity
An organic ligand can coordinate to one metal centre
via one, two, three or more carbon atoms; such ligands are
characterized by their hapticity, which gives the number of
atoms of one ligand in approximately the same distance
from the transition metal. Such ligands are labelled as
monohapto, dihapto, trihapto, etc. and its hapticity is
denoted using η (eta, superscript: number of coordinated
ligand atoms) in front of a ligand: η1, η2, η3, η4, η5, etc.
Cyclopentadienyl (C5H5- = Cp) acts as a monohapto,
trihapto, and pentahapto ligand:

16.3 Cyclopentadienyl complexes


The cyclopentadienyl ligand is usually abbreviated with Cp (+ the hapticity). These are prepared from cyclopentadiene,
C6H5, by deprotonation, generating the cylcopentadienide anion:

This may occur in situ by converting a metal diene complex into a Cp complex, but more usually is done beforehand using an
active metal. Thus NaC5H5 and TlC5H5 are common synthetic sources of Cp-. Examples are:
2NaCp + FeCl2 → Cp2Fe (ferrocene) + 2NaCl
TlCp + Mn(CO)5Cl → CpMn(CO)3 + TlCl + 2CO
MoCl5 + 3NaCp + 2NaBH4 → Cp2MoH2 + 5NaCl + B2H6 + C5H5-
In the lab (exp #10), the cyclopentadienide anion is generated from cylcopentadiene and KOH. There are some rare examples of
other synthetic routes, including nucleophilic displacement:

This example contains an η1-Cp and the more common η5-Cp coordination modes.
Chemistry 3820 Lecture Notes Dr. M. Gerken Page200

Structure and bonding in metallocenes:


By metallocenes, I mean the M(η5-C5H5)2 complexes, which are known for V to Ni in the first row as well as Ru and Os. These
complexes are unusually for organometallic
complexes, in that only the group-8 Electronic configuration and M-C bond lengths in M(η5-C5H5)2 complexes
complexes with Fe, Ru, and Os are 18- Complex Valence electrons Electron R(M-C)/pm
electron complexes. The range of the electron configuration
count is thus 15 to 20. All these complexes V(C5H5)2 15 (e2’)2 (a1’)1 228
have a planar sandwich strcutre. Let us Cr(C5H5)2 16 (e2’)3 (a1’)1 217
examine the electronic structure to see why Mn(C5H4CH3)2* 17 (e2’)3 (a1’)2 211
this structure is adopted and why such a large Fe(C5H5)2 18 (e2’)4 (a1’)2 206
variation in electron count is permissible. Co(C5H5)2 19 (e2’)4 (e1”)1 (a1’)2 212
We start with the SAO’s for D5h for Ni(C5H5)2 20 (e2’)2 (e1”)2 (a1’)2 220
planar Cp rings. We will use only the pπ * Data are quoted for this complex because Mn(C5H5)2 has a high-spin
orbitals! These will interact with the metal to configuration: (a1’)2 (e2’)2 (e1”)2 and, hence an anomalously long M-C bond
form a bunch of bonds, which are neither σ length (238 pm) .
nor π bonds. The bonds within the Cp rings

and to the hydrogens are irrelevant, and are omitted from this
discussion.
Step 1: Form the Cp SAO’s into sets of SAO’s for the Cp2
arrangement, in eclipsed geometry.
a”2 → a1’ + a2”
e1’ → e1” + e1’
e2” → e2’ + e2”
Step 2: Order these by the number of nodes they have.
Step 3: Allow them to interact with metal orbitals of suitable
symmetry. (from the character tables or from visual inspection).

Consequence of this structure:


The MO diagram is given above. The orbitals in the blue box
are the frontier orbitals. They are neither strongly bonding, nor strongly antibonding, and in fact very strongly resemble the
frontier orbitals of an octahedral transition-metal complex. The e2’ and a1’ together are closely allied to the Oh t2g set, while the
e1” directly corresponds to the eg set. This accounts for the variable electron-count and allows for a variety of electron
configurations, including high- and low-spin variants of MnCp2.
This also rationalizes the chemical behaviour of the various metallocenes, those with access electrons are easily oxidized, the
others are easily reduced. Only the 18-electron examples are air-stable solids.
Chemistry 3820 Lecture Notes Dr. M. Gerken Page201

Reactions:
Ferrocene has a vast organic chemistry because it is sufficiently stable to survive the reaction conditions required for
electrophilic aromatic substitution. A typical reaction involving the aromatic ligand in ferrocene is the Friedel-Crafts acylation
(experiment #11 from the lab): AlCl3
CH3COCl + Fe(η5-C5H5)2 → Fe(η5-C5H5)(η5-C5H4COCH3) + HCl

16.4 Benzene and other arene complexes


Neutral benzene is a 6-electron donor, and bis benzene chromium(0) is the classic example of a relatively stable derivative. It
is a sandwich complex with the Cr atom centred between two flat benzene rings. On the whole, benzene complexes are less stable
than Cp complexes.
CrCl3 + Al + AlCl3 + C6H6 → [Cr(C6H6)2]+ → [Cr(C6H6)2]

Cylcooctatetraene complexes (COT)


This ligand has quite a rich variety of coordination modes. It only forms true polyene complexes in its aromatic form, where it is
the 10π dianion. The classic example is uranocene:
UCl4 + COT2- → U(COT)2
2
U(COT) is a genuine sandwich compound, with the metal centred between two flat COT rings, which are bonded in an
octahapto fashion.

Fluxional behaviour of metallocenes


Metallocenes are prone to fluxionality, e.g., Cp rings can rotate fast around the M-Cp
axis. The Ru(η4-C8H8)(CO)3 complex also exhibits fluxional behaviour on the NMR-
time scale. X-ray crystallography shows that the C8H8 ligand is tetrahapto, however,
the 1H NMR spectrum consists of one single line at room temperature. Lowering the
temperature produces four lines that are expected for the tetrahapto coordination
mode.
Chemistry 3820 Lecture Notes Dr. M. Gerken Page202

16.3 Metal complexes of alkenes and related ligands


Compounds containing carbon-carbon double bonds can form metal complexes where the double bond itself is the ligand. In
fact the earliest known organometallic compounds (albeit not recognized as such for over 150 years) is this type of compound.
1. Classical ethylene complexes
Aqueous solutions of PtCl42- react with ethylene to give the products “PtCl2·C2H4” and “PtCl3·C2H4-”. These have the following
structures:
-
Cl
Cl Cl
Pt Pt Cl Pt
Cl Cl Cl

The salt of the latter, K[(η2-C2H4)PtCl3]·H2O, is known as Zeise’s salt. Zeise first prepared these interesting and prototypical
complexes in 1827, making these the oldest known transition-metal organometallic compounds! However, he originally
formulated them as KCl·PtCl2·C2H5OH and the true nature of this complex was discovered only in the 1950’s.
The orientation of the alkene fragment in these and similar complexes suggests that the π HOMO of the alkene acts as a σ-
donor to the metal. However, it has since been shown that alkenes also bond lower-valent metals, for example Pt and Ni in the
zero-valent state bond to four double bonds of two cyclic diene ligands, i.e., COD – cylcooctadiene:

Note that in this complex the centres of the double bonds form a tetrahedral arrangement, consistent with d10, rather than the
square-planar geometry of Zeise’s salt, which contais a d8-transition-metal centre. Such complexes of zero-valent metals suggest
that alkenes can have π-acceptor as well as σ-donor properties. This can be rationalized using the MO description of ethene,
which has both, a π bonding MO and a π* antibonding MO (see below).
Another important class of zero-valent metal complexes of alkenes contains the Ni, Pd, and Pt(0) complexes of the type:

Structural parameters of alkene complexes


In general, coordination of alkenes lengthens the C=C bond somewhat, while the R
groups bend back out to of the double-bond plane. The angle between the R-C-R plane
and the C-C vector is called α, the “bend-back angle”, as shown above. In Zeise’s salt,
the value for α is about 16° and the C-C bond length is 137.5 pm, versus 134 pm in free
ethene. In the Pt(0) alkene complexes, the bond lengths are about 144 pm.
For R=CN: C-C = 149 pm, α = 32°; for R=F, α=40°.

Bonding in metal-alkene complexes


Like the metal polyenes, alkene ligands have no lone pairs. The most available
electron pair on an alkene turns out to be the π-HOMO and the bonding model used
involves synergistic bonding involving both the HOMO and LUMO of the alkene. This
is usually called the Dewar-Chatt-Duncanson model, after the British chemists who
first proposed it. The alkene π orbital is the HOMO and acts as a σ donor to the metal,
since this orbital is polarized into the region between the Ca toms, but above and below
the plane. Upon coordination, only on face of the planar alkene is attached, thus the
Chemistry 3820 Lecture Notes Dr. M. Gerken Page203

faces are distinguished, which can be of importance in catalytic sequences involving optical activity introduced during the
reaction involving an alkene. The alkene π* orbital is polarized away from the region between the C atoms, making it perfectly
adapted to interact with the filled metal d orbital. It thus acts as a π acceptor.
We can explain the structural parameters discussed above in terms of this model. With Pt(II), the metal is relatively electron
poor, so that ethylene coordinates mostly as an electron donor, and the extent of back-bonding is not so significant. Since back-
bonding fills the π* orbital, lengthening of the C-C bond is expected. The bend-back of the R groups, in this case H, is explained
by partial rehybridization towards sp3 at C.
In Pt(0) complexes, the metal is much more electron-rich causing much more extensive π back-bonding. Thus the C-C bond
is lengthened more and the rehybridization is greater.
When a strong electron-withdrawing group is used on the alkene, the π-acceptor character of the ligand is increased, and the
extent of back-bonding becomes extreme, with very long C-C bonds and extensive bend-back of the R groups. In the limit, this
forms a metallacyclopropane complex:

The limit of back-bonding: metallacyclopropane: here the C-C bond is a single bond and the C atoms
are sp3 hybridized.
M

Molecular orbital diagram for Zeise’s salt


We now put all this together for a specific example: the MO energy levels of the Zeise-anion type of complex. The symmetry
is C2v (see assignment #5 for a complete diagram of a square-planar complex – D4h - without π interaction). The resulting
molecule is a 16-electron complex: 6 electrons are involved in Pt-Cl bonds, which are lower-lying in this diagram (involving the
s, p, and dx2-y2 orbitals). The following MO energy-level diagram contains only the d orbitals. The 2a1 (dx2-y2) orbital of the PtCl3-
fragment is the antibonding MO – the bonding MO forms the σ bond to the ligands.

Potrebbero piacerti anche