Electron Counting in Organometallic Chemistry

1. The 18-Electron Rule; definition & rationalisation

The constitution and structure of main group element complexes can be

predicted and rationalised by a combination of the "Octet Rule" and VSEPR.

e.g. SnMe4, BH3.NH3

In transition metal (TM) complexes, the "18 Electron Rule" generally applies.

"A stable TM complex is formed when the sum of d-electrons, donated

ligand electrons and charge on complex = 18."

(In a d-block metal the 9 valence orbitals (the five d-orbitals, the three p-orbitals and
the single s-orbital) can combine with ligand orbitals to make 9 bonding combinations
and 9 antibonding combinations. Thus filling all the bonding orbitals with electron
pairs gives 18 bonding electrons).

Classically, the number of valence electrons (VE) in complexes are counted by

using one of two conventions e.g. in the familiar example of ferrocene;

ionic covalent

[C5H5]
-
6 VE [C5H5] 5 VE

Fe [Fe]2+ 6 VE [Fe]0 8 VE

[C5H5]
-
6 VE [C5H5] 5 VE

18 VE 18 VE

The ionic counting system is misleading; organometallics usually have a large amount
of covalent character. The covalent counting system is similarly misleading, but more
particularly it does not help us to describe the electronic structure of the molecule. A
more instructional "classification" of organometallic compounds has been suggested
by Green. We will describe this in some detail later.

1
 Why don't classical coordination compounds obey the 18e rule?
In order to answer that we have to look more closely at what makes the 18e
rule work in the first place. Firstly, we can classify all complexes into the following
groups;

Class1
number of valence electrons or 18,
First row TM, Remember?
small, i.e. weak field ligands like F-, Cl-,
H2O, NH3 Hence, t2g non-bonding and eg eg
weakly antibonding and so it's not so important

Energy
how many electrons are in these orbitals.

NB, electron count could be accidentally 18!
t2g
[TiF6]2- 12 VE
[Mn(H2O)6]2+ 17 VE
[Co(NH3)6]2+ 19 VE
[Zn(en)3]2+ 22 VE

What determines the number of VE in Class 1 complexes?

Class 2
number of valence electrons 18,
Second/third row TM,
large
Hence, t2g non-bonding (0-6e allowed) and eg antibonding (0e preferred).

[WCl6] 12 VE
[WCl6]2- 14 VE
[W(CN)8]3- 17 VE
[PtF6]2- 18 VE

Class 3
number of valence electrons = 18,
Organometallics,
large i.e. strong field organometallic ligands like CO, C2H4, C5H5 (-acceptors
which stabilise t2g orbitals- see Coordination Chemistry Handout page 27!)
Hence, t2g now bonding (fully occupied, 6e) and eg antibonding (unoccupied).

2
Exceptions to the 18e Rule
There are quite a few examples of organometallics which have 16 VE. As with
all chemistry, the excuse is either electronic or steric (or both).

(i) Electronic effects

Late TM with d8 electron configurations e.g. Rh(I), Ir(I), Pd(II), Pt(II)
have a strong tendency to form square planar 16 VE complexes. Similarly, d10
complexes tend to form trigonal 16 VE complexes. As the atomic number Z
increases, the d-shell is stabilised (lowers in energy). The occupied dz2 orbital
(perpendicular to the plane) is no longer involved in ligand bonding.

occupied dz 2

PPh3 PPh3
Cl Pt PPh3 Ph3P Pt
Cl PPh3

d8 square planar d10 trigonal

or...

(ii) Steric Effects

Early TM have fewer d-electrons to start with than the middle and late
TM, so they must achieve their 18e count by coordination of a larger number of
ligands. If the ligands involved are too bulky, then low-electron count complexes
are formed.

Cl
Zr The 16e "Zirconocene dichloride."
Cl

4s is ALWAYS above 3d in complexes, hence although the electronic configuration of e.g.

Ti atoms is usually given as [Ar]3d24s2, in Ti complexes it is always [Ar]3d4. That also
makes Rh(I) = d8.

3
2. The Green Guide To Electron Counting
The basic principle of this method of counting electrons in organometallic compounds
is that all ligands and the metal centre are counted as neutral. Ligands are divided into
two main classes:

L ligands which contribute 2 electrons

X ligands which contribute 1 electron

L Ligands
These are the conventional neutral ligands which bond via lone pair donation
or donation from a -system, e.g. PR3, N2, C2H4, H2 (molecular hydrogen), CO,
RNC, C2H2.

Examples:

[Mo(CO)6] = MoL6 6 + 6x2 = 18 VE

[Pt(C2H4)3] = PtL3 10 + 3x2 = 16 VE (see above)

[W(iPr3P)2(CO)3(H2)] = WL2L3 L (= WL6) 6 + 6x2 = 18 VE

X Ligands
These are ligands which, in real life, have charges. In this scheme we count
them as radicals (i.e. neutral) and therefore contributing one electron, e.g. H, X (X
= halide), R (alkyl), Ar (aryl), RCO, OR, NR2, SH. Note that these are only simple X
ligands when they are terminal - see later for bridging.

Oxidation State (OS)

In neutral compounds, the number of X ligands = the oxidation state of the
metal, but see later for anionic and cationic complexes.

Examples:

[Mn(CO)5Me] = MnL5X = 7 + 5x2 + 1 = 18 VE, OS = 1.

[Os(PR3)H8] = OsLX8 = 8 + 1x2 + 8x1 = 18 VE, OS = 8.

[Pt(CO)2Cl2] = PtL2X2 = 10 + 2x2 + 2x1 = 16 VE, OS = 2.

This avoids problems with charge assignment (Ph+ or Ph-?).

4
Composite Ligands
We can use combinations of L and/or X to classify more complicated ligands, still
treating them as neutral, e.g.:
1- allyl M = . = X ( 1 electron )

3- allyl M M =
. = LX ( 3 electrons )

Chelating diphosphine R2P PR2 = L2 ( 4 electrons )

M

4-butadiene = L2 ( 4 electrons )
M
.
1-cyclopentadienyl M = = X ( 1 electron )
H

5-cyclopentadienyl M M = = L2X ( 5 electrons )

5-pentadienyl = L2X ( 5 electrons )

M M = .

6-arene = L3 ( 6 electrons )
M =

The -nomenclature
he prefix n for ligands means:
The number of carbons (n) bound to the metal centre
N.B. with no superscript n means n is taken as the maximum for that ligand

5
 Coordination Number (CN)

This is the sum of L and X ligands, not the number of ligand atoms bonded to
the metal.

For the complex MLnXm, CN = n + m

Examples:

(L 2X)

(X)
Ph
Mo = Mo(L 2 X)2 (X)(X) = MoL4 X4 = 6+4x2+4x1
H = 18 VE, OS= 4, CN= 8
(X)
(L 2X)

(L 2X)

Rh = Rh(L2 X)(L2) = RhL 4 X = 9+4x2+1

= 18 VE, OS= 1, CN= 5

(L 2)

(L 3)

= W(L3)(LX)(L)(X) = WL5 X2 = 6+5x2+2x1

= 18 VE, OS= 2, CN= 7
W
Ph3P
(LX)
(L) Cl
(X)

6
Cationic And Anionic Complexes
These are counted in exactly the same way, except that in a cation we have to
subtract one from the electron count for each positive charge and add one to the
number of Xs to get the oxidation state.

Example:
(L2X) +

= Fe(L2 X)(L)3 + = FeL5 X+ = 8+5x2+1-1

Fe
CO = 18 VE, OS = 2, CN = 6
(L) CO
CO (L)
(L)
Similarly, in anionic complexes we have to add one to the electron count for each
negative charge and subtract one from the number of Xs to get the oxidation state.
Example:
(L2X) -

= Fe(L2 X)(L)2- = FeL4 X- = 8+4x2+1x1+1

Fe = 18 VE, OS = 0, CN = 5
CO CO
L L

Hence,
Oxidation state = (number of X ligands) + (charge on complex)

Bridging Ligands
Halides often bridge between two metal centres using a lone pair, i.e.

Cl
M M

so they count as LX, X for the bond and L for the donor bond.
Example:

(L3)
(LX)

(LX) = Mo2 (L3 )2 (LX)2 (LX)2 = Mo2 L10 X4 = 2x6+10x2+4x1

Cl = 36 VE i.e. 18 VE per Mo,
Mo Mo = 4 X ligands, i.e. each Mo has OS = 2,
Cl
= 14 L+X, i.e. each MO has CN = 7
(LX)

(LX) (L3)

7
 Bridging hydrogens are impossible to count conventionally, so we are not
even going to attempt to do so! In this scheme we count them as LX, i.e.

H X
M L M
Examples:
(X) (LX) (X)
= B2 (LX)2 X4 = B2 L2 X6 = 2x3+2x2+6x1
H H H
= 16 VE, i.e.8 VE per B,
B B
H H = 6 X ligands, i.e. each B has OS = 3,
H
(X) (LX) (X) = 8 L+X ligands, i.e. each B has CN = 4.
(L) (L)
(L) CO (LX) CO = Cr2 (L)10 LX- = Cr2 L11 X - = 2x6+11x2+1x1+1
CO CO (L)
H = 36 VE, i.e. 18 VE per Cr
(L) CO Cr Cr CO (L) = 1 X ligand + (-1 charge), i.e. each Cr has OS= 0
CO CO = 12 L+X ligands, i.e. each Cr has CN= 6
CO CO (L)
(L)
(L) (L)

An advanced rule:
You may be able to see that X- = L, thus Cr2L11X- = Cr2L12

We count bridging alkyl groups in exactly the same way, as LX. i.e.

C X
M L M

Agostic hydrogens are also bridging ligands in a sense and are classed as L,
except that we use the half-arrow convention to indicate the agostic bond, i.e.

M C
L X
H
Example:
(LX)

CH2

+ H L
Mo = Mo(L2 )(L)2 (LX)(L)+ = MoL6 X+ = 6+6x2+1x1-1
PMe3 = 18 VE, OS = 2, CN = 7.
(L)
PMe3
(L2) (L)

8
Alkylidenes and Alkylidynes
There are two classes of alkylidenes: Fischer alkylidenes in which the
alkylidene carbon is electrophilic, which are classed as L ligands, e.g.

CO
CO (L) OMe
CO Cr :C = CrL6 = 18 VE, OS = 0, CN = 6
CO Ph
CO

and Schrock alkylidene in which the alkylidene carbon is nucleophilic, which are
classed as X2 ligands, e.g.

t
Bu
t
Bu
Ta CH = Ta(X)3 (X2) = TaX5 = 5+5x1
t
Bu (X2) = 10 VE, OS= 5, CN = 5

t
Bu

Similarly, Fischer alkylidynes are classed as LX and Schrock alkylidynes as X3.

Metal-Metal Bonds
Here we assign one of the electrons of the electron pair in a single metal-metal
bond to each of the metal centres e.g.

1e 1e
(CO)5Mn Mn(CO) 5
5x2+7+1=18 5x2+7+1=18

Metal-metal double, triple and quadruple bonds are treated in an analagous

fashion by assigning 2, 3 or 4 electrons respectively to each metal.

9
Nitrosyls
Remember that the neutral NO molecule is a radical with one unpaired
electron, and can bind to a metal in a bent or linear fashion.
In a bent nitrosyl the NO is acting as a one electron X ligand and the lone pair
on nitrogen is not being used in bonding (as a consequence of the bent geometry) i.e.

M N

O
(X)
Example:

Cl
Ph3P (X) = Ir(L)2 (X)2 (X) = IrL2 X3 = 9+2x2+3x1
Ir N
= 16 VE, OS = 3, CN = 5
Ph3P O
Cl

In linear nitrosyl complexes the lone pair on nitrogen can now donate to the metal
(into an empty d orbital) as a result of the linear geometry, so that a linear nitrosyl
counts as LX, i.e.

M N O

(LX)
Example:
CO

CO Co N=O = Co(L)3 (LX) = CoL4 X = 9+4x2+1x1

(LX) = 18 VE, OS = 1, CN = 5
CO

10
 FGNC PROBLEM CLASS 1
Assuming the eighteen electron rule is obeyed, what are the values of t-z
and the oxidation state of the metal in (a)-(g)?

(a) [Mn(CO) t(COMe)] t=

O.S. =

(b)
u=
O.S. =
Mo
Me (PMe 3)u
Me

(c) [CoH(CO)v] v=
O.S. =

(d) [OsHw(PPh3)] w=
O.S. =
+
x=
(e) Mo (H)x
O.S. =

CH 3 -
(f)
I y=
Rh (CO)y
I O.S. =
I

(g)
z=
Fe
O.S. =
(CO) z

11