APR Nov/Dec 2022 - 34

» SPECTROSCOPY
»
an intuitive user interface that allows non-experts to benefit from
everything that Raman technology has to offer. Manufacturers can
now easily integrate this technology into their production process,
increasing efficiency and product quality.
Detailed Process Information
Raman spectroscopy can handle samples in different forms,
including solid, liquid, gas, powder, aqueous solutions or slurry, and
each peak in the spectrum provides detailed information about the
substances that are present. This flexibility allows testing at various
points in the production process to monitor the analyte of interest,
which is especially beneficial when dealing with a bioreactor that
often contains many types of molecules. Quantifying the amount
of a certain substance from a Raman spectrum is straightforward as
there is a linear relationship between the intensity of a peak and the
concentration of the corresponding molecule. It is therefore easy to
build quantitative models that accurately predict the concentration
even when dealing with a relatively small sample set. These useful
features open up a range of possibilities and can be used throughout
the entire biopharma manufacturing process to verify the integrity
of raw materials, monitor bioreactor processes in real-time and
evaluate the end product. Raman spectroscopy can answer questions
such as 'Are the cells supplied with the right amount of glucose?',
'Are too many secondary metabolites building up?', 'Have the cells
begun to produce the desired product?' and 'How much product has
been produced, and does it have the right characteristics?'. As all the
answers are provided in real time, it is possible to make continuous
adjustments to optimize the processes.
Monitoring Across the Entire
Production Chain
Process analysis involves several types of measurements that can
be divided into four primary classes, defined by their location and
whether the sample needs to be removed from the production line
for testing:1
In-line measurement
During in-line measurements, a probe or sampling interface is placed
either inside or in line with the process or product flow. This usually
means inserting a probe directly into a flow system or bioreactor,
to continuously monitor the product. Using Raman spectroscopy at
this stage allows evaluations at several different locations in parallel
to determine product consistency throughout the process. This is
possible because neither probe nor sample needs to be removed
during the measurement process.
On-line measurement
On-line measurement is similar to in-line monitoring, but with some
differences; although the sample can still be measured without being
34 |
| November/December 2022
Figure 1. Preprocessed bioreactor data collected from three
different units.
removed, a part of the product is redirected for analysis. This means
that measurements are performed on just a portion of the product,
and the diverted sample can be re-introduced to the process stream
or diverted to waste, depending on the application.
At-line and off-line measurement
Contrary to in-line and on-line measurements that do not require
the sample to be removed for analysis, at-line and off-line
measurements involve checks away from the production line. In
case of at-line measurements, the tests are run in close proximity
to the production facility, while for off-line measurements the
sample is transported to a remote laboratory. Compact Raman
analyzers - for example miniaturized handheld Raman analyzers
with quantitative analysis capabilities - are ideal for effective
measurements in either of these environments.
Optimizing Glucose Levels
Raman spectroscopy has proven to be useful for many different
applications in biopharmaceutical production, and one example is
modeling of the glucose content of a bioprocess. Glucose is required
for cell reproduction and is subsequently added through a feeding
cycle to keep a constant rate of cell production. Keeping glucose at the
right level is of utmost importance for most processes as it is directly
connected to the yield, and modeling its content can therefore help
gain precise control of the production rate.
Modeling requires data which, in this example, has been collected
from a bioprocess performed at varying global locations,2
using the
same instrument set-up at each site (see Figure 1). The stability and
accuracy of the analyzer ensured consistency of results between the
different locations. Additionally, fundamental preprocessing methods
were used to target and amplify the relevant signals within the Raman
data. This means that, although the individual data sets were small,
the gathered information could be combined, resulting in a precise
APR Nov/Dec 2022

Table of Contents for the Digital Edition of APR Nov/Dec 2022
