Page 91 - The engineering of chemical reactions

Reaction-Rate Data 75

4. Kinetics measurements. When detailed literature data are not available and when one
badly needs accurate kinetics, then the only recourse is to obtain kinetic data in the
laboratory.

Reaction chemistry

First one tries to find out what reactions should be expected with the given reactants and
process conditions. Here one must know enough chemistry to decide what reactions can
occur. If one is not very experienced in the chemistry of the process in question, one
must examine the relevant literature or ask experts (organic chemists for organic reactions,
polymer chemists for polymer reactions, material scientists for solids reactions, etc.).
Once one has formulated a list of possible reactions, one should then look up the
relevant free energies AG: in order to calculate the equilibrium composition for the desired
feed and temperature. If AC: is sufficiently negative, that reaction should be expected
to be irreversible. If the equilibrium constant is sufficiently small so that equilibrium
product concentrations are small for the conditions used, then that reaction can be neglected
compared to others.

Batch-reactor data

Reaction kinetics are most easily and inexpensively obtained in a small batch reactor. With
liquids this is frequently just a mixture of liquids in a beaker or flask placed on a hot plate
or in a thermostatted water or sand bath. With gases the experiment would involve filling
a container with gases and heating appropriately. One starts the process at t = 0 with Cj,
and records Ci (t) .
The most difficult aspect of these experiments is finding a suitable method of analyzing
the composition of the reactor versus time. Gas and liquid chromatography are by far the
most used techniques for analyzing chemical composition. Spectroscopic methods (IR,
visible, UV, NMR, ESR, etc.) can be used in some situations, and mass spectrometry is
a versatile but difficult technique. For reactions in gases with a mole number change in
a constant-volume batch reactor, the conversion can be determined by simply measuring
the pressure change as the reaction occurs. All these techniques require calibration of the
instruments under the conditions and the composition range of the experiment.
One measures Cj (t, T) for given Cj, and then finds a suitable method of analyzing
these data to find a suitable rate expression that will fit them. For liquid solutions the
typical method is to obtain isothermal batch-reactor data with different CjoS and continues
to gather these data as a function of temperature to find a complete rate expression. For a
simple irreversible reaction we expect that the rate should be describable as

r(Cj, T) = k(T) fi I$?
j=l
Thus we expect the rate to be given by a power dependence on the concentrations
and an exponential dependence on temperature kr (T) = kfOe-E’RT. This form of the
rate expression is not always accurate, especially for catalytic and enzyme reactions
for which Langmuir-Hinshelwood and Michaelis-Menten expressions are required to fit
experimental data.