Chapter 5 – Bonding Models in Inorganic Chemistry: 2 The Covalent Bond

Lewis Structures

Do this (pp 138-9) as a review.

Valance Bond (VB) Theory

VB theory begins with the assumption that the atomic orbitals of two bound atoms overlap to

form a bond. Mathematically, atomic orbitals overlap by multiplying the wave functions.

Thus for H2: ψH2 = ψA(1)ψB(2)

where ψA(1) refers to the first hydrogen atom and its electron, and

where ψB(2) refers to the second hydrogen atom and its electron.
The next thing that is done is to correct for inaccuracies in this equation. For example, in the

equation it is assumed that each electron remains closely associated with the nucleus to which it was

initially assigned. In fact, there is no reason why the electrons cannot switch positions. Likewise,

while not very energetically favorable, there is no reason why both electrons cannot be on the same

nucleus for brief periods. The result can be thought of as an ionic correction (i.e. H+ H-):

ψ = ψA(1)ψB(2) + ψA(2)ψB(1) + λψA(1)ψA(2) + λψB(1)ψB(2)

where λ << l

A third correction is for the electrons shielding each other. These corrections yield about 85% of the

experimental energy and come within 0.008 Å of the experimental bond length. Other, more

involved corrections, provide more accurate predictions. The best wave function to date has over

100 terms and is accurate to within 0.002%. It’s worth noting that, unlike the sum you’ve been

shown here, not all of those one hundred terms have physically explainable meanings.

Resonance occurs when more than one energetically reasonable structure can be drawn for a

molecule. There are two possible types of resonance structures. One takes charge separation into

account. For example, for the molecule HCl there will be a significant ionic contribution from

H+Cl-. There will also be a very small contribution by H-Cl+.

ψ = aψcov + bψH+Cl- + cψH-Cl+ where a > b >> c

The second resonance type is the more familiar type. The book uses a classic example, the
 2

carbonate ion. Here three equivalent structures can be drawn.

.. ..
:O: :O: :O: :O:
2 2
3 3
.. C .. C .. C C
..
:O
.. O:
.. :O
.. O:
.. :O
.. O:
.. :O
.. 2
3
O:
..

I II III average

ψ = dψI + eψII + fψIII where d = e = f since the structures are equivalent

Here the double bond is delocalized over 4 atoms. In the case of carbonate all three structures

are equivalent so each contributes equally to the actual structure. As we will see next, this is not

necessarily so. When structures contribute different amounts, the relative contributions must be

determined.

General rules to yield resonance structures (in order of importance) include:

1) The number of bonds should be maximized consistent with other structure drawing rules

(e.g. no pentavalent carbons)

2) The atoms must always occupy the same relative positions.

3) Formal charges on atoms should be minimized and should be placed reasonably according to

atom electronegativities. (i.e. negative charges on more electronegative atoms).

O O -
+
O P O O P O
O O

Opposite charges should reside as close to each other as possible.

4) The number of unpaired electrons should be the same (usually zero).

Formal Charges

These are charges assigned to atoms as if each atom possessed half of the electrons in bonds it
makes.

Q = NAE - NLPE - ½NBPE NAE = number of atomic electrons

NLPE = number of lone pair electrons
NBPE = number of bonding pair electrons

Example: CH4 QC = 4 - 0 - ½(4 x 2) = 0

QH = 1 - 0 - ½ (1 x 2) = 0
 3

Remember that formal charges are just that, they should not be taken literally. What formal

charges really tell us is something about the charge distribution within a molecule. Typically, you

can do this by simply looking at electronegativities, but that doesn’t always work. For example, the

formal charges on carbon monoxide are reversed from the electronegativities (p. 148) resulting in a

nearly non-polar molecule. Thus, if the formal charges are reversed (based on electronegativity)

bond polarity will be reduced. Conversely, if they align as expected the bond will be more polar

than expected. Nonetheless, while formal charges may provide you information about where

nucleophilic or electrophilic attack may occur, these sites should not be viewed as ionic. On p. 148

the book shows a way to calculate formal charges that is interesting, but not required.
Hybridization

This is a central feature of VB theory. We will first address how, then why.

Hybridization can be viewed as a three step process.

i) Promote each electron into its own atomic orbital

ii) Randomize the spins

iii) Mix the orbitals

Pictorially for carbon:

↑↓ ↑ ↑ (i) ↑ ↑ ↑ ↓
2s 2p 2s 2p
(ii)

↑ ↑ ↑ ↑ (iii) ↑ ↑ ↑ ↑
sp3 2s 2p
But why does this occur or, more specifically, why is it energetically favorable? First look at
the shape of the hybrid on p. 150. It is formed by adding the wave functions of an s and a p orbital
(to yield an sp hybrid).
s orbital

s+p

p orbital
 4

Where the signs of the wave functions are the same, the waves add (constructive interference)
and a large lobe is obtained. Where the signs are opposite they cancel (destructive interference), and
a small lobe is generated. The resulting hybrid has a huge lobe for overlap and consequently larger
bond energies are obtained. An interesting feature of these orbitals is their shape (p. 150). As you
can see, they are blunter than p orbitals, which accounts for their improved overlap. See Table 5-3
on p. 153: hybrid-H bonds 410-500 kJ/mol vs. p-H at 335 kJ/mol. Overlap order sp > sp2 > sp3 >>
p, due to shape. There are 5 sets of hybrid orbitals.
sp - linear (2) sp3d - trigonal bipyriamidal (5)
sp2 - trigonal (3) sp3d2 - octahedral (6)
sp3 - tetrahedral (4)
All except sp3d yield a set of equivalent orbitals, sp3d can be thought of a combination of 3 sp2
and 2 dp hybrids. Finally, like atomic orbital subshells, overlaying all of the orbitals of a given
hybrid set generates a spherical distribution of electron density. (See the Orbitron link on the class
webpage.)
Read the rest of this section on your own.
Molecular Orbital (MO) Theory

There is a major conceptual difference between VB and MO theory. In VB theory we talk of

overlapping orbitals to form a bond and this seems to be intuitively reasonable. In contrast, while
MO theory uses the atomic orbitals of its constituent atoms, the product MOs are thought of as
completely new entities, although the MOs frequently resemble the atomic orbitals. This is
reasonable since core electrons are not involved in bonding and even valence electrons are not going
to be completely separated from the parent atomic orbital. One thing to remember is that each
atomic orbital will give rise to one MO. The Orbitron website has nice images of both atomic and
molecular orbitals. It is linked on the 448 webpage.
The linear combination of atomic orbitals method of generating MO will be discussed now. It
is one of many methods. We will use H2 as an example with ψA and ψB representing the electrons
on the respective hydrogen atoms. There are two possible linear combinations
 5

ψb = ψA + ψB (ψb = bonding MO)

ψa = ψA - ψB (ψa = antibonding MO)
For H2+ (1 electron system) ψH2 = ψb = ψA + ψB
For H2 (2 electron system) ψH2 = ψb2 = (ψA + ψB)2
= ψA(1)ψA(2) + ψA(1)ψB(2) + ψA(2)ψB(1) + ψB(1)ψB(2)
Note how similar this is to the VB equation. Here the ionic contribution is weighted too heavily, but
this can be corrected for.
ψA and ψB can be represented pictorially (p.155) by:

ψ
•A •B
r

ψ
•A •B •A •B

ψ2
•A •B •A •B
ψA + ψB ψA - ψB

At the center points the values of ψ2 can be easily calculated:

(ψA + ψB)2 = ψA2 + 2ψAψB + ψB2 = 4ψA2 (since ψA = ψB for H2)
(ψA - ψB)2 = ψA2 - 2ψAψB + ψB2 = 0
Molecular orbitals can be divided into three categories:
bonding MOs - the signs of the interacting wave functions are the same, the interaction is a net
attraction.
antibonding MOs - the signs of interacting wave functions are opposed, so the interaction is
repulsive. That is, the atoms are pushed apart. This is because electron
density is forced from between the nuclei and the nuclear charges are not
 6

screened from each other.

nonbonding MOs - occurs for lone pairs and when half of interacting wave functions have the
same sign and half are opposed (e.g. ⊕ ⊕ – vs. ⊕ – –). If the signs of the
end functions are fixed and opposed the middle is irrelevant. Nonbonding
is no net interaction or overlap.
Skip normalization through end of section (pp. 156-157 (top)).
Symmetry and Overlap - Read on your own.

Symmetry of MOs
Sigma (σ) bonds possess no nodes that include the internuclear axis. Pi (π) bonds possess 1
node that includes the internuclear axis. Delta (δ) bonds possess 2 nodes that include the
internuclear axis. Antibonding orbitals are designated with an asterisk *. Note the symmetry of the
MO is the same as the atomic orbital of analogous designation: i.e. σ ≈ s = gerade, π≈ p =
ungerade.
MOs in Homonuclear Diatomics

These are the simplest molecules and will be discussed before heteronuclear diatomic
molecules and polyatomic molecules.
There are two criteria for formation of a bond:
1) There must be net positive interaction between the orbitals of interacting atoms. (i.e. the
signs of the wave functions must be the same)
2) The orbitals must have roughly equal energies. (i.e. 1s interacts with 1s, 2s with 2s, 2p
with 2p, etc.) We shall soon see that if necessary, any two valence atomic orbitals with the
proper symmetry will combine.
A bonding interaction is expressed as: σ1s = ψ1sA + ψ1sB or σ1s = 1sA + 1sB
where A & B are labels used to designate the atoms of a homonuclear diatomic molecule
An antibonding interaction is expressed as: σ1s* = ψ1sA - ψ1sB or σ1s* = 1sA - 1sB
These expressions are very similar in presentation to those in H2 and are identical
mathematically.
 7

The level of bonding interaction between 2 atoms is described by bond order, which is similar
to the number of bonds in a molecule.
B.O. = 1/2 (number of bonding electrons - number of antibonding electrons)
Examples of some diatomic molecules include:
H2: σ1s2 B.O. = ½(2 - 0) = 1
He2: σ1s2 σ1s*2 B.O. = ½(2 - 2) = 0 This “molecule” does not exist.
B2: KK σ2s2 σ2s*2 π2p2 B.O. = ½(4 - 2) = 1
C2: KK σ2s2 σ2s*2 π2p4 B.O. = ½(6 - 2) = 2
N2: KK σ2s2 σ2s*2 σ2p2 π2p4 B.O. = ½(8 - 2) = 3
O2: KK σ2s2 σ2s*2 σ2p2 π2p4 π2p*2 B.O. = ½(8 - 4) = 2
There are several points worthy of note.
1) Fractional bond orders are possible. e.g. Li2+: σ2s2 σ2s*1 B.O. = ½ (2 - 1) = ½. This ion
actually exists with a longer bond length than Li2. Valence bond theory does not have any
formal way of dealing with fractional bonds.
2) You would expect the B2 molecule would begin by filling a σ2p before the π2p but the
reverse is observed. This is because the energy gap between 2s and 2p is not large, and for
the σ2p bond the 2s can mix in. This has the effect of raising the energy of the σ2p orbital.
Note on Fig. 5.13 (p. 165) mixing with the σ*2s also occurs.
3) When filling MOs the same rules apply as for atomic orbitals. Thus, for a pair of orbitals,
the first electron goes in either orbital, the second in the other orbital. This suggests that C2
and O2 should be paramagnetic, which is experimentally observed.
Bond Lengths and Ionization Energies

As was mentioned for Li2, experimental evidence for the MO model of molecules comes from
bond lengths. Ionization energy also provides further supporting evidence.
An example from the book:
B.O. Bond Length
O2+ 2.5 112 pm
 8

O2 2 121 pm
O2- 1.5 126 pm
O22- 1 149 pm
The book shows you that the ionization energy of NO (894 kJ/mol) is much smaller than that of
either an isolated nitrogen atom (1402 kJ/mol) or an oxygen atom (1314 kJ/mol). On first
consideration, one would expect the ionization energy of NO to be slightly less than average of these
values. This is because the ionization of an electron from the molecule would spread the charge over
two centers. The ionization energies of the component atoms are comparable, so the charge would be
shared roughly equally and, so, the ionization of the molecule would be a little lower than the average
of the ionization energies. Yet it is about 35% less. Why? The electron is taken out of a high energy
antibonding orbital. (An aside, based on electronegativities, one would expect oxygen to have a
higher ionization energy than nitrogen. Why are the values reversed?)
Electron Density in Li2 - F2: Read this section on your own

MOs in Heteronuclear Diatomic Molecules

The most important difference between heteronuclear and homonuclear diatomic molecules is
that the bonding atoms in the former have different electron affinities and ionization energies.
Hence, they have differing tendencies to gain or lose electrons.
The component atoms in homonuclear diatomic molecules are the same so their IEs and EAs
are the identical. A result is the equal sharing of electron density. In heteronuclear bonds the
electron density is shared unequally and is measured in terms of electronegativity. Greater
electronegativity is a greater tendency for an atom to attract electrons from an atom to which it is
bound.
This is represented in equation form by
ψb = aψA + bψB and ψa = bψA - aψB
where a = b in a homonuclear diatomic molecule and b > a if B is more electronegative than A.
Note that atom B contributes more to the bonding MO than A, but the converse is true for the
antibonding MO. The result is that the greater the difference in electronegativity between two
 9

atoms, the more the product MO will be like the parent atomic orbital of the more electronegative
atom. (The logical extreme is a = 0 and b = 1 in a hypothetical purely ionic bond.)
This last point relates the relative covalency/ionicity of a bond to MO theory. If a = b the bond
is completely covalent. As the difference increases, ionicity increases. At small differences a bond
is polar covalent, at large differences a bond is ionic.
This can be seen pictorially as follows:

σ* σ* ↑ σ*
↑
↑ ↑ ↑
↑
↑↓ ↑↓ ↑↓
A σ B A σ B A σ B
pure covalent polar covalent ionic
a=b b>a b >> a
One more thing needs to be considered. What about orbitals not used in bonding? These
orbitals, particularly if they are in a region of space away form bonding, remain largely unchanged.
These are one of the two major types of nonbonding orbitals and correspond to lone pairs in valence
bond theory.
Molecular Orbitals in Triatomic Molecules and Ions

We’ll use the two triatomic species used by the book: BeH2 and NO2+.
The valence shell of beryllium consists of the 2s and 2p orbitals. Recall from earlier that their
energies are similar. Thus, if conditions are right both may be involved in bonding.
First, assume a linear geometry (its least crowded) with H-Be-H along the z-axis. Then px and
py cannot participate in bonding because they are perpendicular to the bonding axis (and s orbitals
can't π-bond).
There are two approaches to how the MOs can be constructed. One is an intuitive method
which you would probably use at this point. That is, select an orbital on beryllium and match it with
an appropriate hydrogen orbital. This turns out to be quite difficult with larger molecules and is not
generally used.
 10

A second approach is to treat all atoms bound to the central atom as a group. The difference
will seem trivial for BeH2, however it is important for larger molecules. In this approach, the
hydrogens can be in phase (ψH1 + ψH2) or out of phase (ψH1 - ψH2). The in-phase combination
interacts constructively with 2s. The out-of-phase with 2p. Antibonding orbitals are obtained by
reversing the signs on the hydrogen atoms.
Thus: ψg = aψ2s + b(ψH1 + ψH2)
ψu = cψ2p + d(ψH1 - ψH2)
ψg* = bψ2s - a(ψH1 + ψH2)
ψu* = dψ2p - c(ψH1 - ψH2)
The NO2- ion contains 2 types of bonding: σ and π. The sigma bonds will require the s and 1 p
orbital on the nitrogen. Since the nitrogen will have a lone pair, a second p orbital is needed. The net
effect is a set of orbitals which is essentially equivalent to sp2.
The π system can be treated similarly to the σ system in BeH2. That is, the oxygen p-orbitals
can be treated as a group: ψO1 +ψO2 or ψO1 - ψO2

ψb = aψN + b(ψO1 + ψO2) ψb = bψN - a(ψO1 + ψO2)

The third orbital is a little different because the same result is obtained whether ψN is added or
subtracted.

ψn = ψN + (ψO1 - ψO2) ψn = ψN - (ψO1 - ψO2) ψn = ψO1 - ψO2

(or ψn = ψO1 - ψO2 ± ψN as your book puts it.)
This is the second type of nonbonding orbital. Skip the symmetry discussion from the middle of p.
178 - 182 (top).
Electronegativity
This is the ability of an atom in a molecule to attract electron density to itself. The Pauling
 11

scale is by far and away the most common scale and for general purposes works as well as the
others. On page 187-190 five different scales are shown. The Pauling scale is based on a comparison
of the bond energy of an A-B bond vs. the average of A-A and B-B bond energies.
Mulliken – Jaffe Electronegativity Scale
This method has the advantage of taking into account the number of things bound the atom (i.e.
oxidation state) and which orbitals are involved in bonding. The values are relatively easy to
calculate.
χ = ½(IE + EA)
Therefore, the scale is based both on how well on atom adds an electron and how well it holds
onto its own electrons. Read through the end of paragraph 2 on p. 185, then skip through the text on
p. 191.
Variations in Electronegativity
As you can see in Table 5.6 the electronegativity of an atom depends on its hybidization. Why?
A relatively simple way to address this question involves examining hydrocarbon reactivity. The
pKa of CH4 is about 60. For ethylene pKa ≈ 44, and acetylene pKa ≈ 25. Since carbon and hydrogen
have similar electronegativities, this property cannot give rise to the acidity difference.
Go back to Table 5.6. Look at the various hybrids and you will see that as s orbital character
increases, so does electronegativity. The reason is the same as for the acidity difference. First, one
will have to know what they have in common:
EN: As s character increases, the atom is better able to remove electron density from a
neighboring atom.
acidity: As s character increases, the carbon is better able to stabilize a product negative charge
or better it causes a larger charge separation (polarization in the C-H bond).
The s orbital causes this because it penetrates closer to the nucleus than do p orbitals. Thus,
electrons in s orbitals experience greater nuclear attraction.
Read the rest of this section on own.
Pauling’s Electronegativity – Read this section on your own.
 12

Other Methods of Estimating Electronegativity – Skip from here to the end of the chapter.

October 20, 2004