Page 89 - The engineering of chemical reactions

“What Should I Do When I Don’t Have Reaction Rates?” 73
8:;: Example 2-11 The monomers for nylon are either the amino acid HzN-(CH&COOH,
G which forms the polymer called Nylon 6 (six carbon atoms in the backbone) or two
8:; monomers adipic acid HOOC-(CH&COOH and hexamethylene diamine H2N-
3~ (CH&NHz, which form the polymer called Nylon 66 (six carbon atoms in the backbone
iii of each of the monomers). It is possible to make the 6-carbon atom monomers very easily
$4 with high purity, but the production of nylons with five, seven, or other carbons would be
i@ very difficult. Why?
a’-*:
“s’ iQ, ~ aj Cyclohexane. It can be prepared with high purity by distillation of a mixture of
4 alkanes from petroleum refining or by reduction of benzene.
8n(
n_ Cyclohexane can be partially oxidized by just attacking one C-C bond to open
in IX
.:i the ring with functional groups on each end of the six-carbon chain to produce the
,$:” six-carbon amino acid or adipic acid. Sketch the intermediates to these products
‘,Z”
starting with cyclohexanone.

“WHAT SHOULD I DO WHEN I DON’T HAVE REACTION RATES?”

Unfortunately, there are no tables of chemical reaction rates in this book, and you won’t
find them in other books either. It would be very useful of we could construct tables of rates
such as

Reaction Preexponential Activation energy Orders Range of validity
and then list all the important reactions we may be interested in running. We would then just
look up the reaction, find the rate that is applicable for the conditions at which we want to
operate, insert these equations into the mass and energy balances, and solve them to predict
performance. While you can find useful data tables in any text on thermodynamics, heat
and mass transfer, or separations, reaction-rate data do not exist for most technologically
interesting processes.
In thermodynamics the properties of pure materials are given to five significant figures,
and correlations for mixtures are frequently accurate to better than 1%. Diffusivities of many
binary systems are known to within a few percent, and, even for complex mixtures, they
are probably available to within 5%. Mass and heat transfer texts give correlations that are
usually good to a few percent, depending on the mixture and conditions. Separation process
efficiencies can be computed to similar accuracies using empirical correlations and notions
of equilibrium stages.
However, in reaction engineering we cannot even begin to do this. If someone claims
to have a general correlation of reaction rates, the prudent engineer should be suspicious.
The major problem is that most interesting reaction systems involve multiple reactions, and
one would have to somehow list rates of forming many products from several reactants; so
we would need a lot of data in tables to cover a process. Second, most interesting reactions
are catalytic, either on a solid surface or by an enzyme. Different catalyst systems behave
quite differently with different catalyst formulations, and they are notoriously sensitive to
trace impurities, which can poison the catalyst or promote one rate over the others. Catalytic
reaction rates also do not usually obey simple power-law rate expressions (such as Y = kCA),
and one frequently finds that effective parameters such as orders and activation energies