Lecture 11: Reactions in Liquids

We have discussed the theories of chemical kinetics for reactions occurring in the gas
phase. Now, we will turn our attention to reactions in solution (in the liquid phase).
There are substantial, qualitative differences in the dynamics of molecules in the liquid
phase compared to the gas phase, and liquid phase kinetics are not understood at the same
level as gas phase kinetics. Models of reactions in solution will therefore be simpler and
more approximate than those used in the gas phase.

In the gas phase, collisions between reactants are rare, and the interval between collisions
is random. The collision process can be considered Markovian, meaning that the
occurrence of a collision does not affect the likelihood of having another collision. In
contrast, molecules in solution interact continuously with the solvent, and in order to
move long distances they must squeeze past closely-spaced solvent molecules. As a
result, when two solute molecules (reactants) collide, they are likely to undergo multiple
collisions while they are held together by the surrounding solvent molecules, before
diffusing apart. In the approach we will take here, each of these sets of collisions is called
an encounter between the reactants, and while the reactants are trapped close together, we
refer to them as an encounter pair or an encounter complex. This difference in collision
distributions between the gas and liquid phase is illustrated schematically below.

The gas phase system and solution system shown above have the same overall collision
rate between reactants, but the distribution of those collisions in time is dramatically
different, and this changes the way in which reactions occur. In addition to changing the
pattern of collisions, the solvent also can change the energetics of the reaction. For
+
example, formation of ions is often stabilized by solvation, and as a result, NaCl → Na +
-
Cl is highly endothermic in the gas phase, but exothermic in solution. A third important
effect of the solvent is rapid energy transfer to reactants and products, so that thermal
equilibrium is maintained.

We will now consider a simple phenomenological model of reaction in solution. In order

to react, the molecules must first diffuse together, then react before they diffuse back
apart. We can think of this as

Where A and B are the reactants, and {AB} is the encounter complex, the pair of
reactants trapped close together and undergoing multiple collisions. k is the rate constant
1
for the reactants to diffuse together and form the encounter complex, k is the rate
-1
constant for breakup of the encounter complex (rate constant for the reactants to diffuse
apart), and k is the rate constant for the reaction of the encounter complex to form
2
products. We may apply the pseudo-steady-state approximation to the collision complex
{AB}. Doing so gives

You should note the similarity between this analysis and the Lindemann-Hinshelwood
analysis of pressure dependent reactions. It is similar, except that the first step is not the
process of a reactant getting enough energy to react, but the process of a pair of reactants
coming close enough to each other to react.

Before looking at the general case, we can consider two limiting cases of this expression.
If reaction is much faster than diffusion (k >> k ) then the reaction will be diffusion
2 -1
limited, and the effective rate constant will be k = k . If diffusion is much faster than
eff 1
reaction (k >> k ), then the process will be reaction limited, and the effective rate
-1 2
constant will be k = k k /k .
eff 2 1 -1

When the reaction is diffusion limited, the rate of reaction is given by the rate at which
the reactants diffuse together. If we consider a single A molecule, fixed in space (or use
a reference frame that moves with the A molecule) we can derive an expression for the
rate at which B molecules collide with it (or encounter it). Of course, when an individual
A and B molecule meet, they react and both disappear (forming products). However, this
picture of B molecules diffusing toward A still gives a correct picture of the average
behavior of a large number of equivalent A and B molecules. If we consider a
distribution of A and B molecules, the B molecules that are close to A molecules will
react first, and then are removed from consideration so that, on average, the distribution
becomes depleted near the A molecules. The flux of B molecules, relative to a fixed A
molecule, is given by Fick’s first law:

where J is the total flux of B toward the A molecule, [B] is the B concentration as a
r
function of the distance from the A molecule, and D is the diffusion coefficient of B
AB
relative to A. Note that the sign convention used here is opposite the normal sign
convention, since we are treating flux of B toward A (at r = 0) as positive. From a shell
balance over an infinitesimally thin spherical shell at radius r, we obtain

Integrating this once gives

J=C
1

Where C is a constant of integration. Combining this with Fick’s first law gives
1

Integrating a second time gives

Far from the A molecule, the B concentration goes to its bulk value [B] = [B]. If the A
r
and B molecules react instantaneously when they get within a distance r = r of each
AB
other, then at that distance the B concentration will be zero ([B] (r=r )=0). If we use
r AB
these boundary conditions to evaluate the two constants of integration, we get
This is the average flux of B molecules to an individual (average) A molecule. To get the
total reaction rate, we multiply this by the total concentration of A. This gives

A similar result can be derived for the related case of diffusion-limited reaction between
ions. In this case, there is an additional contribution to the flux because of the attraction
or repulsion between the reactants. If V(r) is the interaction potential between the
reactants, then the flux of B molecules toward the fixed A molecule is given by

For ions, the interaction potential is given by

where z and z are the charges of the reactants, e is the elementary charge, ε is the
A B o
permittivity of vacuum, ε is the relative permittivity of the solution, and r is the Onsager
r c
escape distance, defined as

The absolute value of r is the distance at which |V(r )| = kT. That is, the distance where
c c
the Coulomb interaction potential is approximately equal to the average thermal energy
of the molecules. Note that r is negative for oppositely charged reactants and positive
c
for reactants with the same charge. Repeating the above derivation for neutral molecules
using this flux that has been modified to include electrostatic attraction or repulsion
between the reactants leads to
where

This is identical to what we obtained before, but with the distance at which the molecules
react (r ) replaced by the effective distance defined above. For oppositely charged
AB
reactants, r ≈ r > r . For reactants with the same charge, r < r . Note that the
eff c AB eff AB
electrostatic interaction between the molecules does not change the distance at which
they react. Rather, the random motion of the particles is biased by their interaction,
increasing or decreasing the probability of them coming close enough together to react.

Now, we will generalize the above analysis (for neutral molecules) to the case where the
reaction is not diffusion controlled. Then, we cannot assume that the concentration of B
is zero at r = r . Instead, we will assume that there is a reactive flux at r = r that is
AB AB
given by

We will consider the rate constant for reaction (k ) in more detail below. This flux must
R
be equal to the diffusive flux at r = r (otherwise, B would accumulate at that distance).
AB
This gives us a new boundary condition at r = r .
AB

The boundary condition at r→∞ and the diffusion equation itself are unchanged, so we
again integrate it twice to get

and

Applying the new boundary condition at r = r gives

AB
So

And

As before, the flux is independent of distance. The total reaction rate is this flux
multiplied by the concentration of A.

and

We recognize 4πr D as the diffusion limited rate coefficient that we derived before
AB AB
(k = 4πr D ) so
D AB AB

At the beginning of this section, using the mechanism

We derived
Comparing this to the expression for k derived from the reaction-diffusion equation
eff
shows that

, and k = k (= 4πr D )
D 1 AB AB

So, k is the rate constant that would be observed if diffusion were fast enough that an
R
equilibrium concentration of encounter pairs ({AB}) was maintained. That is

A rough estimate for K can be obtained by assuming that the A and B molecules are
AB
randomly distributed in the solvent. Then the concentration of encounter pairs will be
given by

where n is an average coordination number in the solution (how many total molecules are
adjacent to an A molecule), and [B]/[S] is the fraction of molecules that are [B]
molecules. It is assumed that [B]<<[S]. From this analysis, we see that an estimate of
K is given by
AB

This will be a reasonable estimate if the molecules form an ideal mixture (they obey
Raoult’s law; all bimolecular interactions (A-A, A-B, A-S, B-B, B-S, S-S) are similar).

In principle, we can apply transition state theory to determine k . In practice, however,

2
this is usually not possible because it is much more difficult to apply the relevant
statistical mechanics in the liquid phase. For example, the partition functions can no
longer be separated into different modes, and the partition functions of a species can
depend on its surroundings. So, we will not attempt to apply the statistical mechanical
version of transition state theory to reactions in solution. However, we can get some
insights by applying the thermodynamic formulation of transition state theory. In this
case, we can write k as
2
 ‡
just as we did in the gas phase. In this case, ∆G is the change in free energy in going
from the encounter complex to the transition state. The equilibrium constant for
formation of the complex can be written as

⎛ −∆GAB
0
⎞
K AB = exp ⎜ ⎟
⎝ RT ⎠
where ∆G AB
0
is the change in free energy in going from separated A and B molecules to
the {AB} encounter pair. So, the overall rate constant could be written as

*
Where ∆G is the change in free energy in going from the separated A and B molecules
to the transition state. In a non-ideal solution, the equilibrium constants are normally
written in terms of the species activities rather than their concentrations, and the free-
energy changes are evaluated for a reference state rather than for the actual conditions in
solution. If that is done, then the above expression for the rate constant must be modified
by the appropriate activity coefficients.

Where ∆G0* is now evaluated at standard-state conditions. Note that both ∆G0* and the
activity coefficients must refer to the same standard state (usually infinite dilution for
compounds at low concentrations in solution).

For some reactions, comparisons can be made between rates in the gas phase and rates in
solution, or between rates in different solvents. In the gas phase, the free energy of
activation is evaluated with the pure species as ideal gases as the reference state. In
solution, this reference state is likely to be the species at infinite dilution in the solvent
under consideration. So, the reference state, and therefore the free energy of activation,
depends on the solvent. If the transition state is stabilized by solvation more than the
reactants are, then the reaction will be faster in solution than in the gas phase. If the
reactants are more stabilized by solvation, then the reaction will be slower in solution.
These are only rough generalizations, but they give some insight into how reaction rates
are affected by the solvent. There are additional effects on the dynamics of reactions due
to the constant collisions between solvent and reactants. These can, for example, induce
re-crossing of the transition state. However, detailed discussion of these effects would go
well beyond the scope of this course, so we will stick with the phenomenological
descriptions presented above.
 After completing your study of these lecture notes you should be able to:
(1) Describe the differences between the dynamics of reactions in the gas phase and in
solution
(2) Use expressions developed here to analyze diffusion limited reactions between neutral
species and between ions of like or unlike charge
(3) Be able to identify and model the change from reaction-rate-controlled to diffusion-
controlled conditions for a reaction
(4) Understand the application of the thermodynamic formulation of transition state
theory to reactions in the liquid phase, and how the activity coefficients of the
reactant(s) and transition state play a role in determining the reaction rate