CH2210 – Experimental Kinetics

Kinetics can be used within industrial settings in many ways, it can be used to determine the
economic feasibility of a reaction as well as potential risks (e.g. a sudden change in heat/pressure
associated with one stage of the mechanism) and in turn reactants can be altered and the reactors
themselves better engineered without ever risking the safety of those involved.

Kinetics allows for the determination of the steps involved in a reaction, as well as the mechanism
itself, the typical kinetics experiment consists of 4 steps;

1. Mixing reactants
2. Initiating reactants (often using heat or light)
3. Measuring temperature or pressure to prevent a runaway reaction
4. Measure reactant/product concentrations

The methods used to measure concentration commonly employed are the use of a pH meter,
measuring conductivity or photometry. Which method is used is often based on the half-life, t1/2, of
the reaction and the mixing time which is, preferably, as fast as possible. To determine the rate of a
reaction relative to a given species the concentration of that species over time can be measured, a
plot of Conc vs. Time will give a straight horizontal line for order 0, a plot of Ln(Conc) vs. Time will
give a straight line for order 1 and 1/(Conc) vs. Time will give a straight line for order 2, the
straightness of the line can be computationally analysed using R 2 values, the closer to 1 the more
ideal the line.
Another method for determining order is to use the quarter-life, t1/4. t1/4 is independent of the initial
pressure of the reagents for a first-order reaction, but it is proportional for second-order reactions,
by running an experiment at the same temperature but differing pressures and comparing the t 1/4
values the order can be assessed.
There is also the initial rates method, since;

( ) ( ( )) ( )

A plot of ln(-R0) vs. ln(PR(t=0)) should be linear with a slope equal to the order, n. The initial rate can be
obtained from the slope of the tangent at t=0 or more accurately by fitting a cubic polynomial to the
data.

From this data the rate constant, k, can be determined and an Arrhenius plot can then be used to
determine the activation energy, EAct, by plotting ln(k) vs. 1/T with the slope equal to –EA/R.

For reactions with a very short t1/2 (1msec/1sec) a continuous flow method can be used;
This method is very wasteful of reactants however and there are difficulties in using moving
detectors.

A less wasteful technique is the stopped flow method;

For even faster reactions (t1/2 <1msec) relaxation methods are used. For a system at equilibrium the
system can be perturbed (i.e. removing some of the reactant or product) and the time taken to re-
equilibrate.

The perturbation should occur in 10-6/10-7 seconds and may be initiated by;
-pressure jump
-temperature jump
-high electric field pulse
-ultrasonic vibrations
-light pulse (flash photolysis) in conjunction with ultrafast spectroscopy (able to detect over
timescales of femto-pico seconds).

In flash photolysis a cell is subjected to a light

pulse producing atoms, radicals and excited
states whose concentration is followed with
time. It is advantageous as it has no mixing time,
the time-scale of the reaction is only limited by
the pulse duration (less than 10-12 seconds using
laser sources) and the species are formed in the
centre of the cell, and so wall-reactions may be
ignored.
These methods all consider systems in which the reactants have a thermal distribution of energies
(according to the Boltzmann principle), to study the elementary reactions ideally it is preferred that
the molecules possess a specific energy and only undergo one collision during the reaction process.
This is achieved using molecular beams. Through collimation and rapid expansion a beam of
molecules can be formed that is intense, linear and importantly possesses minimal vibrational
excitation. This energy can be tunes by varying temperature or seeding the beam with inert
elements (e.g. He) however this method is neither cheap nor simple to achieve.

Heterogeneous Catalysis
Heterogeneous catalysis is a branch of surface chemistry and analysis worth $10 trillion dollars, in
particular due to reactions involving NH3 and CH3OH and C2H4 oxidation (using metallocenes). 90% of
global chemicals rely on heterogeneous catalyzed processes. It is also an important aspect in
environmental preservation, ozone is depleted by aerosols in Polar Stratospheric Clouds (PSC’s),
pollution can be controlled used catalytic converters in engines (using Pt/Rh/Al 2O3 surfaces), also by
generating cleaner ways to synthesise chemicals at lower temperatures with more benign solvents,
and forming more efficient methods of generating power, all of which are examples of
heterogeneous catalysis.

Surface kinetics can be split into two areas, the gas phase above a surface and the kinetics at the
surface itself.

In the gas phase the controlling parameters are;

 The concentration of the gas

 The collision rate, zAA
 The collision cross-section, σA

At the surface the rate of molecules colliding with the surface (the flux)=n.ċ (speed) where;

The Hertz-Knudsen equation can then be used to determine the rate;

( )

The surface chemistry is phenomenally fast due to the higher collision probability offered by a large
surface, in other words there is a vast collisional cross section.

When considering surfaces the two important terms are substrate (adsorbent), the solid surface
where adsorption occurs, and the adsorbate, the atomic/molecular species adsorbed on the
substrate.
When a molecule collides with a surface either nothing will happen, the collision is elastic and no
energy is transferred, or atoms/molecules feel an attractive potential and stick to the surface, this is
adsorption. The sticking probability, s, is then defined by the no. of molecules that stick/no. of
molecules that impact and 0≤s≤1.
This arises from the van der Waals forces between a species and the surface. As the species
approaches the van Der Waals potential, U(d), decreases (a favourable attraction) until the
internuclear repulsion, V(d), becomes the primary interaction and the species is repulsed. If this
repulsion is overcome by the dipole formation adsorption occurs.

Since a species adsorbing to a surface loses a degree of freedom (3 becomes 2) and so entropy is
decreasing. According to Gibbs-Helmholtz, ΔG=ΔH-TΔS, since entropy is decreasing the final term
becomes positive, however a decrease in ΔG is preferred and must be less than 0 for a process to
occur spontaneously, this means ΔH must also be less than 0 and so adsorption is always
exothermic.

For dissociative adsorption, in which adsorption occurs leading to a decomposition of the adsorbed
species, the X-X bond must be broken (with dissociation energy D XX) and a two X-M bonds are
formed (with bond strength EX-M).
For adsorption to remain exothermic then EX-M>DXX/2. Any species with a higher EX-M will therefore
be more ‘specific’ when considering catalysis as it can only achieve adsorption to metals capable of
forming strong enough bonds.

Types of Absorption - Chemisorption and Physisorption

Chemisorption Physisorption
ΔH>50 kJ mol-1 ΔH<50 kJ mol-1
Chemical adsorption, the formation of chemical Physical adsorption, van der Waals (permanent
bonds or induced dipoles)
Confined to a single monolayer Multilayer adsorption
Adsorption may be activated Unactivated adsorption
Dissociation possible if EX-M is large enough Enhanced at low temperature (le Chatelier
principle)
Adsorbed atoms can still diffuse across the Non-dissociative
surface
Simple equilibrium between gas and adsorbed
molecules gives the Frenkel equation;
Surface lifetime, τ=τ0eΔH/RT

Chemisorption, involving the formation of chemical bonds, is an electron-transfer process, as a result

bonds within the adsorbate are weakened, this is fundamental to catalysis.
Chemisorption may be associative (in which atoms/molecules bind at the surface) or dissociative (in
which the adsorbed molecule dissociates in some manner). Depending on the desired process the
catalyst must be finely tuned to adsorb the species associatively/dissociatively, allow for diffusion
and reaction of adsorbates and allow desorption (often achieved as one of the diffusion reaction
products has a weak X-M bond).
For dissociative adsorption the most common energy profile involves activation;

At the point at which the two profiles cross

the molecule will undergo dissociation and
approach the surface further, performing
adsorption. This requires an input of energy
to overcome the initial repulsion, it then
becomes favourable for the dissociated
atoms to bind to the surface and so the
repulsion is negligible until a much shorter
distance from the surface.

If no energy is required to overcome this

repulsion then spontaneous dissociation will occur, this is non-activated dissociative chemisorption,
for example CO on Ta.

Measuring Adsorption and the Surface Area of a Catalytic Surface

Using the kinetic theory of gases a simple volumetric method for measuring adsorption can be
achieved, since P1V1=P2(V1+V2), by using two chambers, one larger than the other, and opening a
valve between the two to allow for equilibration, then performing the same experiment with a
surface present in the second chamber, the comparative pressure will decrease (due to adsorption
at the surface) and so by measuring pressure the number of molecules adsorbed can be determined.

Using this data alongside adsorption isotherms allows for the calculation of the surface area of a
given surface. By plotting the number of molecules adsorbed (or the surface coverage, θ, where θ=0
means an empty surface and θ=1 a full surface) against pressure the monolayer uptake – the
number of molecules/atoms/ions which completely cover the surface, can be determined.

The Langmuir Adsorption Isotherm

By assuming certain characteristics of adsorption Langmuir derived a simple equation that can
describe the adsorption isotherm of any species, enabling predictions of adsorbate coverage, θ, this
in turn allows for the calculation of reaction rates as well as providing data with which to optimize
the reaction conditions (T,P) to enhance the reaction.

The isotherm is based on the idea that chemical equilibria exist during all reactions;

The stability of the adsorbate compared to the gas/liquid species is defined by temperature (of both
the surface and the reaction media) and the pressure/liquid concentration above the catalyst. The
equilibrium exists between the species in solution, M, and the empty surface sides, S, as well as the
adsorbates.

[M] is proportional to the pressure/liquid concentration and S* is proportional to the vacancies on

the surface, and so proportional to (1-θ), this leads to the first assumption that there are a fixed
number of identical, localized surface sites. The second assumption is then that adsorption is
immobile, surface diffusion can be ignored, and finally the third assumption is that each adsorption
site is occupied by only one adsorbate molecule (there is only monolayer adsorption).

The equilibrium constant, b, then equals, according to the above assumptions and proportional
relationships;

[ ]
[ ] ( )

In terms of θ this gives;

Since b is proportional to the sticking probability, s, it is dependent on ΔH Ads. This leads to the fourth
assumption that ΔHads and thus b are temperature and pressure independent.
The value of b is important in catalysis, the higher the value the lower the pressure must be to
achieve saturation, understanding this can allow industrial processes to achieve the ideal pressure
from a very early stage, saving time and money in development.

By rearranging the Langmuir isotherm and relating it to experimentally measurable variables,

pressure and the number of molecules, the isotherm can be applied to a wider range of
experiments.

( )

A plot of P/n vs P should provide a straight-line with a slope equal to 1/nmonolayer. From nmonolayer and
the cross-sectional area of the adsorbate the surface area of the solid can be calculated. Instead of
the number of molecules the volume of gas adsorbed can also be used.

The Langmuir adsorption isotherm only works under the assumption only a monolayer is formed,
due to physisorption there is very often a second stage of adsorption in which a multilayer is formed,
to account for this a second isotherm is necessary, the BET adsorption isotherm.

The BET Adsorption Isotherm

Brunauer, Emmett and Teller developed a model that allowed for multilayer adsorption but also
accounted for the different enthalpy of adsorption of multilayers and monolayers (lower for
multilayers) according to;

( )
( )

ΔHDo is the enthalpy of desorption, taking into account the strength of the adsorbate-surface
interaction and ΔHVapo is the enthalpy of vapourisation and takes into account the strength of the
adsorbate-adsorbate interaction.
A plot of P/V(P-P0) vs P/P0 should give a straight line with a gradient equal to [(c-1)/Vmc] and an
intercept equal to 1/Vmc. The isotherm is only applicable for the range over which it is linear.
Determining the Heat of Adsorption
For a gas at pressure, P, in equilibrium with the condensed phase at temperature, T;

( )

( ) [ ]

Where ΔHads is the isoteric heat of adsorption.

By measuring the adsorption isotherms at different temperatures and plotting θ against Pressure,
then taking the pressure value at the same θ for several temperatures and plotting lnP against 1/T,
the slope will provide –ΔHads/R.

Applying the Langmuir Isotherm to Reactions

Unimolecular Decomposition
Following the mechanism;

A(g)⇌A(ads)→Products

It is assumed that decomposition occurs uniformly across surface sites (it is not restricted to
particular sites), that products are weakly bound to the surface and once found rapidly desorb and
that the rate controlling step is the surface decomposition step. The surface concentration of A can
then be calculated using the Langmuir isotherm. The rate will be equal to kθ (since θ, surface
coverage, is relative to the surface concentration) and so, expanded to give the rate in terms of the
gas pressure above the surface;

There are two limiting cases in this example;

 Limit 1; b.P<<1 and so 1+bP≈1, Rate≈kbP

Here the reaction is a 1st order reaction, with respect to A, with a 1st order rate constant.
This is a low pressure (weak binding) limit in which a steady state is achieved and θ is very
low.
 Limit 2; b.P>>1 and so 1+bP≈bP and Rate≈k
Here the reaction is 0 order with respect to A.
This is a high pressure (strong binding) limit in which a steady state at the surface is achieved
as θ=100%

Rate therefore shows the same pressure variation as θ.

Bimolecular Reactions
Langmuir-Hirshelwood Type Reactions

A(g)⇌A(ads)
B(g)⇌B(ads)
A(ads)+B(ads)→AB(ads)→AB(g)
Here the rate determining step is assumed to be the surface reaction between the two adsorbed
species. If both molecules are mobile on the surface and intermix then the reaction rate can be given
by the following equation for the bimolecular surface reaction step;

Rate=kθAθB

Considering the competition between A and B for the same adsorption sites the Langmuir isotherm
becomes;

Substituting these into the rate expression gives;

Again considering two limiting scenarios;

 Limit 1; bAPA<<1 and bBPB<<1

In this limit θA and θB are both low and Rate≈k.bAPAbBPB≈k.PAPB, the reaction is first order
with respect to both reactants.
 Limit 2; bAPA<<1<<bBPB and so θA→0 while θB→1, so Rate≈k.bAPA/bBPB and so the reaction is
first order with relation to A and negative first order with respect to B.

By controlling the concentrations of both species and getting it so that b APA=bBPB the optimum rate
can be achieved.

Eley-Rideal Type Reactions

Eley-Rideal reactions follow a slightly different mechanism;

A(g)→A(ads)
A(ads)+B(g)→AB(ads)→AB(g)

Here the rate-determining step is the reaction between gaseous B and adsorbed A. The rate for
these types of reaction is;

Rate=kθA[B]

Without extra evidence however it can be difficult to prove the Eley-Rideal mechanism is occurring
since the last step could also be a composite of stages similar to the Langmuir-Hirshelwood
mechanism;

B(g)⇌B(ads)
A(ads)+B(ads)→AB(ads)

In this case it may just be that there is an incredibly small steady-state coverage of adsorbed B. To
test this the surface coverage of A can be varied (or [A]/[B]), if at higher values of surface coverage
of A the rate drops off that is because B requires open surface sites and so it is a Langmuir-
Hirshelwood mechanism, otherwise the reaction rate would increase upto θ A=1 since B requires no
access to the surface in order to react.

Crystal Structure
The two most common structures in transition metals, and so in catalytic metal surfaces, are face
centred cubic (fcc) and body centred cubic (bcc).

To describe surfaces Miller indices are often utilized, the intercepts of the plane with the primitive
vectors (a,b,c) are taken, reciprocated and then multiplied by the smallest number that gives 3
integers in (h,k,l) notation.

When considering the surface energy these arrangements become important, an atom in the bulk
will be stabilized by 12 nearest neighbours, by cleaving this metal crystal to achieve a (1,1,1) fcc
plane, this becomes 9 nearest neighbours. A lower coordination number is directly linked to an
increase in potential and so a higher energy surface, which is important for surface-based reactions.
By comparison a (1,0,0) fcc surface will have only 8 nearest neighbours and (1,1,0) fcc surfaces have
only 6 nearest neighbours. This surface energy is reduced by adsorption, enhancing the favourability
of the process.

Surface Sensitivity
Supported metal particles can expose different crystal faces, the vertices on a surface will have a
different coordination to facial sites. There are also defects and steps in each particle, these are low
coordination sites and therefore regions of high potential energy, they facilitate bond dissociation at
the surface and so are often vital in surface catalysed dissociation reactions. Structure sensitivity is
the reliance of a reaction on one of these specific sites. The number of steps reflects the metal
particle size and so changing the particle size will modify the rate.

The dispersion of available sites can be mathematically shown as equal to;

( )

For small particles (<20Å) dispersion approaches 100%, and so since activity is proportional to
surface area an increase in particle size will lead to a decrease in rate (per mass of catalyst) provided
the exposed surface atom arrangement remains unchanged.

To determine if a reaction is specific to a certain metal site the turnover numbers (TON, the number
of reactions a single site can achieve) can be compared, a plot of 1/T against TON will give a gradient
proportional to EAct and so if the gradient is the same for each surface the reaction is specific (since
otherwise a variety of EAct would be found due to the variety of available sites). Since there is also a
relationship between the number of certain sites and the size of the metal particles it can also be
possible to determine which site the reaction is specific too based on which will be most or least
effected by the number of defects and how the rate is effected by this.

When considering the hydrogenolysis and dehydrogenation of cyclohexane over a metal surface the
number of defects will increase the rate of the former, suggesting that reaction requires defect sites,
whereas the latter process occurs over a diverse range of surface arrangements (this is an easier
reaction and so less energy is required, the reaction can therefore occur without the additional
energy provided by defects in the metal structure).

Surface Analysis
Apparatus Used in Surface Analysis
Ultra-high vacuums (10-7 Torr) are required in analyzing the nature of a surface, this is to prevent
perturbation of the trajectories of the electrons/ions used and to keep the surface free of
contamination.
Two detectors used are the Retarding Field Analyser (RFA) and the Concentric Hemispherical
Analyser (CHA);

X-Ray Photoelectron Spectroscopy (XPS)

XPS uses a known photon source (often Mg Kα=1254 eV, or Al Kα=1487 eV) to release electrons from
an atom, by measuring the energy required to release the electron a great deal of information about
the sample can be obtained. The kinetic energy of the electron (equal to hν-BE, the binding energy)
can be assessed and from this the binding energy, which is specific to a given element can be
defined. By using a monochromator the resolution of an XPS spectrum can be enhanced (there will
be some background noise from Bremsstrahlung radiation).
Alternatively a synchrotron source may be used, this provides a continually variable source of
monochromatic X-rays for complete analysis of a sample.

Auger Electron Spectroscopy (AES)

AES also uses X-ray or electron excitation sources however the process is considerably different. By
providing enough energy to remove an electron from the ground state, forming an initial excited
state, the Auger process then occurs (in which an electron falls back into the ground state, providing
enough energy to release a further electron as a photoelectron) and it is the energy of the second
released electron that is considered. Since the energy of this electron will be dependent on the
energy different between the initial excited state and ground state, which is unaffected by the
energy of the incident X-ray, there is no need to control the input energy.

The energy of the emitted Auger electron involves the levels w, x and y;

Ewxy=Ew(Z)-Ex(Z)-Ey(Z)

Low-medium energy electron spectroscopies (50-500eV) are surface sensitive due to a ‘Universal
Escape Depth Curve’ characteristically found when plotting escape depth against kinetic energy. At
low kinetic energies the escape depth of the electrons is quite high, since these low-energy electrons
are unable to transfer their energy to other particles (they are unable to suffice the quanta
necessary for promotion of an electron in another atom), at high energies the electrons are able to
penetrate straight through the other atoms and so again there is a large escape depth, a balance of
these two behaviours creates an environment in which electrons have a very low escape depth, this
energy is chosen to ensure surface-sensitivity.

These techniques can be used to analyse growth modes of surfaces, for example some surfaces form
monolayers first before forming additional layers (Frank-Van Der Merwe growth) whereas others
develop local structures, areas of several layers, while others remain at different stages of growth
(Stranski-Krastanov growth), some develop layers completely randomly with multilayers appearing
in some places before a monolayer forms (Volmer-Webber growth) and others may form
simultaneous multilayers. By considering the response of the surface to XPS or AES over time it can
be understood which growth mode is occurring.

By accurately measuring the binding energy using XPS the functionality of a surface (e.g. in polymer
coatings) can be determined since the binding energies increase when electron withdrawing species
are present (and vice versa with electron donating groups) but also the oxidation state of the surface
can be determined.
Low Energy Electron Diffraction (LEED) Analysis
LEED analysis uses low energy electrons and their diffraction off a surface to analyse the crystal
structure of a surface, according to the Bragg equation dsinθ=nλ when considering the following
process;

The separation of the areas of constructive interference will be proportional to 1/d (the
atomic/molecular spacing).

Temperature Programmed Desorption

By increasing the temperature of a surface gradually and placing a mass spectrometer above it is
possible to determine the temperatures of desorption of the products of a surface reaction. This can
be useful in determining the mechanism of the reaction (for example something that requires a
higher temperature to desorb is often the species bound to the surface since this weakens other
bonds within the molecule allowing earlier release of any other products present).

The activation energy (EACT) can be calculated for simple 1st order reactions using the data provided
by the desorption peak temperature and placing it in the Redhead equation;

[ ( ) ]

Where β=the heating rate (typically 1-1000 K sec-1), Tp=the peak temperature and ν =the frequency
factor (1015 s-1).