Biomedical Engineering Reference
In-Depth Information
in-line optics for control of fraction collection. These systems are designed to minimize
allowable tolerances for buffer compositions at any given stage of a process by adaptively
controlling mixing rates.
Natarajan and Purdom [40] designed a system to monitor the formation of a salt
gradient used in an ion exchange process step by measuring the conductivity of the
effluent solution. They demonstrated that the shape of the resultant gradient was
nonlinear and resulted from valve dynamics and the nonlinearity of the mixing of
solutions of different densities. They developed a feed-forward control strategy based on
a mathematical description of the dependence of the effluent conductivity on the mixing
ratio. Their PLC logic employed a quadratic fit model based on variation of effluent
conductivity at various mix ratio values along with the desired conductivity at a given
time to monitor the accuracy of the control. With this system they demonstrated an
improvement in the accuracy of the gradient.
12.3.2 Online Monitoring of Downstream Processes
Although online HPLC used to control downstreamprocesses is receivingmore attention
nowadays, the technology has been successfully employed in downstream bioprocess
manufacturing since the early 1980s [39, 42]. It has been demonstrated to provide
benefits such as increased operating efficiency, cycle time reduction, step coupling to
produce semicontinuous operations, reducing the possibility of processing errors,
eliminating manual handling of fractions and samples, reducing the opportunities for
product contamination, minimizing variability, enabling feedback control of critical
process parameters, and increasing throughput capabilities.
Cooley et al. [42] used online HPLC to reduce the variability of eluate collection
caused by column loading, the purity of the starting material, the affinity of various
components in the process stream for the stationary phase, and the column's operating
parameters such as the generation of the gradient used to elute the column. This
variability generally leads to conservative collection set points that may reduce product
yield.
In Cooley's system, the online HPLC sends the product purity value, derived from
the online HPLC data, to the distributed control system (DCS). The product purity value
generated by the onlineHPLC is comparedwith the product purity set point in theDCS. If
the product purity value is greater than or equal to the set point, the eluate is sent to the
cooling tank. This allows a more consistent process output based on product quality
decisions. The key to any such implementation is that the speed with which the analytical
results are generated must be amenable to the timing of the process step. Because the
process steps that the scientist and engineers were monitoring allowed “relevant time”
feedback, they were able to leverage this strategy at a later point. They designed a more
complex system to couple two process ion chromatography steps by the introduction of
the pure fraction from the first column to be loaded, via a switching valve, onto a second
column.
As is more often the case, other scientist and engineers tried similar techniques at
development scale, but business decisions were made not to proceed forward to
production scale. There are quite a few examples of implementing online HPLC in a
