= Frederick M. Lewis
{c}
\b[<Frederick M. Lewis>], with <Frank R. Mayo>, coauthored the 1944 paper establishing a quantitative basis for comparing <monomers> in <copolymerization>, with styrene and methyl methacrylate as its named chemical system.<Copolymerization. I. A Basis for Comparing the Behavior of Monomers in Copolymerization; The Copolymerization of Styrene and Methyl Methacrylate by Frank R. Mayo and Frederick M. Lewis>[Primary publisher title, full byline and publication record, JACS 66, 1594–1601]{ref} The published name is retained without expanding the middle initial or assigning individual experimental steps to either coauthor.
The scientific problem is that incorporated composition can differ from the reacting-mixture composition. In the <terminal model of copolymerization>, two <monomer reactivity ratios> describe relative self-addition and cross-addition preferences. Inferring both ratios requires observations that distinguish their effects, rather than one measurement that happens to match the feed.
An original identifiability example uses the <copolymer composition equation> at equimolar feed. It reduces to
$$
F_1=\frac{r_1+1}{r_1+r_2+2}.
$$
An observation $F_1=1/2$ establishes $r_1=r_2$ within this model, but leaves their common value unknown. Pairs $(0.2,0.2)$, $(1,1)$ and $(2,2)$ all satisfy it. They need not predict the same sequence correlations or behavior at another feed.
At a second feed $f_1=0.20$, the three candidate pairs instead predict $F_1\approx0.3684$, $0.2000$ and $0.1429$, respectively. This invented comparison explains the information gained by varying the monomer mixture. It is not a reconstruction of the original authors' measured constants.
Modern IUPAC recommendations likewise use several starting compositions together with conversion and product-composition observations.<Radical Copolymerization Reactivity Ratios by IUPAC>[Official recommendation announcement, measurement design]{ref} Lewis's verified contribution belongs to the experimental and theoretical development of that comparison problem. Composition data test the kinetic model and its parameters; they do not acquire a unique interpretation merely because one fitted curve passes through one point.
Back to article page
