Lecture: P1_Wk1_L5

Inter-Molecular Forces: Keesom Force

Ron Reifenberger
Birck Nanotechnology Center
Purdue University
2012

Last Lecture: Electrostatic Intermolecular Interactions

ion

ion

(1)

fixed angle

polar molecule

angle averaged

non-polar molecule

polarization

induced dipole
P1_Wk1_L5

(2)

(3)

(4)
2

This Lecture: Electrostatic Intermolecular Interactions

fixed angle
interacts
with

(5)

polar molecule
angle averaged

(6)

Keesom

polar molecule
interacts
with

non-polar molecule

polarization

(7)
induced
dipole
Debye

non-polar molecule

non-polar molecule

interacts
with
P1_Wk1_L5

Keesom:Debye:London100:10:1

fluctuating
induced
dipoles

(8)
London
3

5. Interaction between polar molecules: fixed orientation

already four parameters: separation distance (z) + 3 angles (1,2,)
+q1

p1

relative twist
angle

+q2

p2 E1
U (z) =

-q1

=
E1 E=
1, dipole

p2

3 ( p1 z ) z p1

-q2

4 o
z3

p1 z )( p2 z )
(
p1 p2
U (z)=
3
z3
z3
or

U (1, 2 ,, z ) =
=

P1_Wk1_L5

p1p2 [ 2 cos(1 )cos( 2 ) sin(1 )sin( 2 )cos( )] 1

4 0
z3

p1p2 1
f (1, 2 , ) ;
4 0 z 3

2 f (1, 2 , ) < 2

2p p 1
0 U (0,0,, z ) =
2 =
1 2 3
lowest energy when 1 =
4 0 z

Dipole-Dipole Interactions
The magnitude and sign of the dipole-dipole energy
depends strongly on orientation
z
p

p
z

p
z

p
P1_Wk1_L5

U(z) is
minimum

U(z) is
maximum

U(z) is
zero
5

Representative plots of f(1,2,)

f(1,2,)
color bar:
2.0

360o

0.0

360o

-2.0

0o
0o

360o

=0; in plane configuration

U (1, 2 ,, z ) =

0o 0o

360o

=90o; perpendicular configuration

p1p2 1
f (1, 2 , )
4 0 z 3

f (1, 2 , ) =
[ 2cos(1 )cos( 2 ) sin(1 )sin( 2 )cos( )]
P1_Wk1_L5

Interaction potential energy between

in-plane (=0o) permanent dipoles
diagram for
d/s 0.75

d Fixed, NO
rotation

3.0
d/s=0.75

+
q1

+ x q3
q4
s

Symbol

Actual
Charge

q1

+q

q2

-q

q3

+q

q4

-q

d/s=0.50

1.5

Utotal (in 10-18 J)

q2

|q|=1x10-18 C; s=10 nm

d/s=0.30

0.0

-1.5

-3.0
-180

-120

-60
0
60
(degrees)

120

180

Note: |p1|=|p2|
P1_Wk1_L5

6. Interaction between polar molecules: angle averaged

dipole-dipole interaction must be properly angle-averaged by allowing
the orientation of dipoles p1 and p2 to fluctuate. Must calculate
U ( z ) U ( z,1, 2 , ) e

kBT

p1p2 [ 2cos 1 cos 2 sin1 sin 2 cos ] 1

U ( z,1, 2 , ) =
4 0
z3

where
here

U ( z,1, 2 , )

.... represents the proper thermal average over all angles

The complete derivation is given in Appendix A :

2

CK
2 1 p1p2 1
UKeesom ( z ) =

6 =
3 kBT 4 0 z
z
(also known as the Keesom interaction; W.H. Keesom, Physik. Z. 22, 129 (1921) and
22, 643 (1921))

P1_Wk1_L5

The Keesom interaction between two ethanol molecules

from P1_Wk1_L3

Material
Ethanol

Chemical Formula

,
dielectric
constant1

p,
dipole
moment
(in Debye)2,3

o/4o
polarizability
volume
(in 10-30 m3)

(CH3)CH2OH

24

1.7

5.1

p1 = 1.7 D

<1>

relative twist
angle

p2 = 1.7 D

<>

<2>

z=2 nm
2

3.33 10 30 Cm
(1.7 D )
2

1
D

p1p2
2 1
1
2
1

UKeesom ( z ) =

6 =

2
3 kBT 4 0 z
3 1.38 10 23 J / K ( 300K ) 4 (24) 8.85 10 12 C
2
Nm

( 2 10

5.66 10 30 2

2
1
1

4
=
=
5.38 1019 1.44 10 100 1.56 1052 J
Nm

20
9
53
6
6.40 10 m
3 1.24 0 J 2.67 10

1eV

=
12.08 10 29 J
7.55 10 10 eV
=
19
1.6 10 J
9
P1_Wk1_L5

)(

)(

7. Interaction potential energy between a

polar and non-polar molecule
polar
molecule

+q1

specify location
of non-polar
molecule

p1, o,1

o,2

P2

(aligned with Edipole)

non-polar
molecule

-q1

Edipole 1( z , )

p1

1
2 cos z + sin
3
4 o z

p1 1
2
2
2

sin
=
+
4
cos
Edipole 1( z , )

4 o z 3

1
p1 1
2
2

3cos
1
=
+

4 o z 3

P1_Wk1_L5

When a polar molecule is in

close proximity to a non-polar
molecule, the polarizing electric
field is generated by the fixed
dipole (p1) of the polar molecule.
This induces a dipole in the
electron cloud of the non-polar
molecule.
The
resulting
interaction energy is called the
Debye interaction energy.

U Debye ( z ) =2

p12 o ,2

1
(4 0 ) 2 z 6

10

The induced dipole moment is a function of E

pinduced ( E ) = o ,2 E
In this case, E is due to a dipole

2
U Debye ( z , ) =
pinduced ( E ) E =
( o,2 Edipole
)
2

3cos 2 + 1 1 2
2
2

3cos
+ 1
p

p
o
,2
1

1
=
o ,2
=

3
2
4 o

z
z6
( 4 o )

1
(See Appendix 1, P1_Wk1_L4)
if an angle average over is performed , cos 2 =
3
p12 o ,2 1
U Debye ( z ) =2
2
6
( 4 o ) z

P1_Wk1_L5

11

If the non-polar molecule is replaced by a polar one, then

the second molecule could induce a dipole in the first:
+q2
polar molecule 1

+q1

p2, o,2

p1, o,1

polar molecule 2

-q2

-q1
The net effect is additive, so we must have

p12 o ,2 + p22 o ,1 1
U Debye ( z ) =
(4 0 ) 2 z 6

If the two polar molecules are identical, then

2 p 2 o 1
U Debye ( z ) =
(4 0 ) 2 z 6

P1_Wk1_L5

12

Appendix A: Derivation of the angle-averaged Keesom interaction

The derivation of the angle-averaged Keesom interaction is
usually not given in most textbooks. Rather, the final result
is often written down after some statement like After a
lengthy calculation, we find..
In what follows the derivation is given using the same logic
followed when discussing the ion-diople interaction in the
previous lecture (C1_Wk1_L3).
The starting point is the dipole-dipole interaction energy for
two dipoles p1 and p2 separated by a distance z

p1p2 [ 2cos 1 cos 2 sin1 sin 2 cos ] 1

U ( z,1, 2 , ) =
4 0
z3

P1_Wk1_L5

13

At first, you might be tempted to perform a straightforward angle

average of U(z, 1 , 2, ) as follows:
U ( z,1....) =

p1p2 [ 2cos 1 cos 2 sin1 sin 2 cos ] 1

4 0
z3

U ( z,1...) =
Uo ( z ) [ 2cos 1 cos 2 sin1 sin 2 cos ] =
Uo ( z )f ( )

Let

Uo ( z )

where
2

p1p2 1
4 0 z 3

and

f () =2cos 1 cos 2 + sin1 sin 2 cos

U ( z,1,....)sin1 d1 sin 2 d 2

f ()sin1 d1 sin 2 d 2

I
U (z) =
U=
Uo ( z ) A
o (z)

2
2
IB
d sin1 d1 sin2 d2
d sin1 d1 sin2 d2
=
1 0

IA =

2 =0

=
1 0=
2 0

=
1 0=
2 0

[ 2cos

=
1 0=
2 0

cos 2 + sin1 sin 2 cos ] sin1 d1 sin 2 d 2 d

=
0 =1 0=
2 0

2

2

2
0
=
2 d sin1 cos 1d1 sin 2 cos 2 d 2 + cos d sin 1d1 sin2 2 d 2 =
0=
2 0
=
2 0

=
=
=
=
1 0
0
1 0
2

IB

d sin1 d1 sin 2 d 2 8
=
=
1 0=
2 0

I
U ( z )= Uo ( z ) A= 0
IB

The result is zero because the angle average is performed for a free dipole in
which all angles are equally likely. Instead, a weighted angle average must be
calculated that preferentially weights those angles having lowest energy. This
involves a correct treatment of the Boltzmann weighting factor and requires some
14
understanding of statistical thermodynamics.

P1_Wk1_L5

Weighting the angles using the Boltzmann factor gives

2

U ( z ) U ( z,1,...) e

U ( z,1 ,...)

kBT

U ( z,1 ,...)

kBT

sin1 d1 sin 2 d 2

=
1 0=
2 0

d
0

let

U ( z,1,....) e

U ( z,1 ,...)

kBT

sin1 d1 sin 2 d 2

=
1 0=
2 0

Uo ( z )
, then
kBT
2

U ( z,1,....) e

U ( z,1 ,....)

kBT

= Uo ( z )

f ( ) e f ( ) sin1 d1 sin 2 d 2

=
1 0=
2 0

d
0

e f ( ) sin1 d1 sin 2 d 2

=
1 0=
2 0

Evaluating such a quantity is simplified by recognizing that the numerator and

denominator are related by a derivative. Using the well known properties of
the natural logarithm:
U ( z,1,....) e

since

U ( z,1 ,....)

kBT

d
= Uo ( z )
ln d e f ( ) sin1 d1 sin 2 d 2
d 0 =
1 0=
2 0

d
ln d e f ( ) sin1 d1 sin 2 d 2
d 0 =
1 0=
2 0

2
d f ( )e f ( ) sin1 d1 sin 2 d 2
2

0 =
1 0=
2 0
f ()
sin1 d1 sin 2 d 2
d e

1 0=
2 0
0 =

P1_Wk1_L5

15

This means we must evaluate the integral

if f ( ) << 1, we can write

2

e f ( ) sin1 d1 sin 2 d 2

=
1 0=
2 0

2 2
sin1 d1 sin 2 d 2 d 1 + f () +
f ( ) + ... sin1 d1 sin 2 d 2
2

=
0
1 0=
2 0

f ()

=
1 0=
2 0

d [1] sin

d1 sin 2 d 2

=
1 0=
2 0

d [ f ()] sin

d1 sin 2 d 2

=
1 0=
2 0

2 2

+ d f ( ) sin1 d1 sin 2 d 2 + ...

=
0
1 0=
2 0 2
= I1 + I2 + I3 + ...
I=
IB= 8
1

I3=

2
=

2
2

I=
0
2

d [ 2cos

=
1 0=
2 0

=
1 0=
2 0

2
2

= Ia + I b + Ic

4 cos2 1 cos2 2 sin1 d1 sin 2 d 2

d [ 4 cos

cos 2 sin1 sin 2 cos ] sin1 d1 sin 2 d 2

=
1 0=
2 0

P1_Wk1_L5

cos 2 + sin1 sin 2 cos ] sin1 d1 sin 2 d 2

2
2

d
0

=
1 0=
2 0

sin2 1 sin2 2 cos2 sin1 d1 sin 2 d 2

16

Ia= 4

cos2 1 cos2 2 sin1 d1 sin 2 d 2

=
1 0=
2 0

=
1 =
2

1
= 4
[2 ] cos3 1
2
3
1
=

Ib =4

=4
=4
Ic =
=

2
2

2
2

d [cos

= 2

[sin ] =0

d
0

=
1 0=
2 0

=
1 0=
2 0

cos
0

1
4 16
3
3
= 4
[2 ] ( 1) (1) = 4 2 = 2
2
9
9
3

cos 2 sin1 sin 2 cos ] sin1 d1 sin 2 d 2

=
1 0=
2 0

cos d

cos

3
2
0=

cos 1 sin2 1 cos 2 sin2 2 d1 d 2

=
1 =
2

1 3
3 sin 1

1
=

1 3
3 sin 2

2
0=

=0
0

sin2 1 sin2 2 cos2 sin1 d1 sin 2 d 2

=
1 0=
2 0

sin3 1 sin3 2 d1 d 2

=
2=

1 =
2
2

1
=
( + sin cos )

2 2

=
2

12 ( cos31 9cos 1 )

1
0=
2

12 ( cos3 2 9cos 2 )

2
0=
2

1
1
8
1
8
8
2 4
2 =
+

2
=
)
[ ] 1 9( 1) (1 9(1)=

2
2
2
9
12

12 12
3

P1_Wk1_L5

17

I3 = Ia + I b + Ic =

So

16
8
24
8
2 +0 + 2 =
2 = 2
9
9
9
3

2
+ ...
I1 + I2 + I3 + ... = 8 1 +
3

U ( z ) U ( z, ) e
= Uo ( z )

U ( z, )

d
ln d e f ( ) sin1 d1 sin 2 d 2
Uo ( z )
=
d 0 =
1 0=
2 0
2

kBT

d
ln ( I1 + I2 + I3 + ...)
d

2
d
d
ln 8 1=
= Uo ( z )
+
+ ... Uo ( z )
d
3
d

ln
8
ln
1
...
+
+
+
( )

1
2
0 + Uo ( z )
=

2 3
1+
3
2
Uo ( z )
3
P1_Wk1_L5

18

now since

U (z)
kBT

=
o
and Uo ( z )

p1p2 1
,
3
4 0 z

we finally obtain
U ( z ) U ( z,1,...) e

U ( z,1 ,...)

kBT

2
=
Uo ( z )
3

pp 1 2 1
pp 1
= 1 2 3
1 2 3
4 0 z 3 kBT 4 0 z
2

2 p1p2 1
=

3kBT 4 0 z 6

P1_Wk1_L5

19