Extreme Scale-Down Approaches for Rapid Chromatography Column Design and Scale-Up During Bioprocess Development - Measurement, Monitoring, Modelling, and Control of Bioprocesses - page 126

4.5.3 Fraction Collection
When collecting fractions using the robotic shuttle set-up, the exact droplet vol-
ume can vary within and between studies depending upon the rate of liquid flow,
geometry of the outlet piece on the column and buffer properties such as interfacial
liquid tension [ 34 ]. Well-to-well variations will cause considerable noise, and so
liquid sensing on the robot is useful to determine the real well volumes and so
correct data values such as absorbance measurements for the actual liquid path
length. Furthermore, if one seeks to quantify the droplet concentration by spec-
trophotometry and depending on the number of droplets collected per fraction,
different types of collection plates may be needed to increase the height of the
liquid column to obtain an adequate signal.
4.5.4 Hold-Up and Wall Effects
It is important to note that dead volumes upstream and downstream are particularly
significant for very small beds such as miniature columns, since they can dominate
peak retention and dispersion. Similarly, one other potentially important factor for
an accurate miniature column (and pipette tip) separation that seeks to be poten-
tially representative of a larger-scale column is the impact of the wall effect.
4.5.5 Miniature Column Examples
Wiendahl et al. [ 34 ] used a robot combined with 200-lL miniature columns to
generate breakthrough and elution curves [ 35 ]. Breakthrough curves were gener-
ated and fractions were collected in a 96-well plate. Miniature columns were also
operated on the ÄKTA chromatography pumping system, and scale-up runs were
performed using a 5-mm-diameter, 10-cm-long column on the same system. The
linear gradients on the ÄKTA used the same length, slope and initial concentra-
tions as the step gradients on the robot. Generally, dynamic capacities reflect the
quantity of a target molecule bound to a resin at a specific percentage on the
breakthrough curve, and for those molecules which are highly valuable such as
those found in the biopharmaceutical arena, a typical value quoted is at 10 %
breakthrough or lower. Breakthrough curves were generated on miniature columns
for bovine serum albumin when applied to a range of ion exchangers. The
breakthrough curve method was shown to be acceptably reproducible
Elution studies involved separating three different protein mixtures. The first
was BSA and Lipolase, where an identical sample load was used at scale-down
and scale-up. After loading the protein solution, which contained 2 mg/mL BSA
and 2 mg/mL Lipolase, a complex elution gradient was used consisting of three
phases, each with different buffer concentrations and numbers of column volumes.
The elution time of the BSA elution peak maxima in the miniature column mat-
ched closely with that of a larger column, and the use of successive incremental