Computational Approaches in the Design of Synthetic Receptors - Designing Receptors for the Next Generation of Biosensors - page 152

The design of experiments follows a mathematical framework for changing several
selected factors simultaneously to predict the optimum conditions, thus reducing
the number of experiments necessary [ 178 ]. In effect the goal is to plan and perform
experiments in order to extract the maximum amount of information in the fewest
number of trials. Chemometrics has various applications in science, such as opti-
mization of experimental parameters, design of experiments, data retrieval and
statistical analysis, analysis of structure-property relationship estimations, signal
processing, pattern recognition, and modeling.

2.10 Examples of Chemometrics Methods Used in MIP Design

Steinke et al. [ 179 ] used the chemometrics approach to investigate the effect of
variables such as type and quantity of monomers, cross-linker, porogens, initiator,
type of initiation (UV or thermal), polymerization pressure, temperature, reaction
time, and reaction vial dimensions have on the properties of synthesized polymers.
Davies et al. [ 74 ] have used chemometrics design expert software to optimize the
composition of a MIP for sulfonamide, in particular the template:monomer:cross-
linker ratio. MAA was selected as functional monomer and EGDMA as cross-
linker. They selected the amounts of template, monomer, and cross-linker (T:M:X)
in the MIP using a three-level full factorial design. The chemometrics design
required the synthesis of a small number of MIPs with different T:M:X ratios and
used the results of their testing to determine the polymer composition with optimum
recognition characteristics. The polymer with the optimized ratio of 1:10:55 was
compared to MIPs, prepared with more commonly used molar ratios such as 1:4:20
[ 180 ] and 1:8:40 [ 181 ]. The experiment proved that the predicted ratio generated a
polymer with superior binding properties.
In another interesting study, Kempe et al. [ 123 ] employed multivariate data
analysis (Modde 6.0 software, Umetrics, Umea, Sweden) for the optimization of
monomer and cross-linker ratios in the design of polymers specific for propranolol.
Mijangos et al. [ 182 ] used chemometrics (MODDE 6.0 software, Umetrics,
Sweden) to optimize several parameters such as concentration of initiator
(1,1 0 -azobis(cyclohexane-1-carbonitrile) or 2,2-dimethoxy-2-phenylacetophenone)
and polymerization time, required for the synthesis of high performance MIPs for
ephedrine. A small set of (

) ephedrine imprinted polymers was synthesized and
tested by HPLC for their ability to interact with (+) and ( ) ephedrine. This

chemometric study provided evidence that in order to achieve high performance
in chiral separation MIPs should be synthesized for long reaction times using low
concentrations of initiator and at low temperatures.

2.11 Examples of Neural Network Methods Used in MIP Design

Prachayasittikul's group used data from the literature, such as monomer composi-
tion, retention, and imprinting factors of MIP and NIP, mobile phase composition,