»
ROUNDTABLE
» CC:
Recent developments in Transmission Raman Spectroscopy (TRS) have made this technique a potential alternative to Near
Infrared (NIR) for the quantification of APIs and excipients in tablets and capsules. Several approaches to quantify APIs using TRS in combination with Multivariate Data Analysis have been published (Anal Bioanal Chem (2013) 405:3367-3379 and literature herein). The recent advent of very robust instrumentation with optimized laser excitation and collection optics allows for a larger volume of a tablet to be sampled with minimal sub-sampling effects. TRS suppresses interfering surface fluorescence and Raman signals from capsules and coatings. Other advantages are rapid analysis times (seconds or less per sample), high accuracy, and quantification of low concentration components with the use of enhancement optics. An increasing number of scientists have adopted this technique and it is now a common tool (more than 10 publications in the last two years). Although the effects of physical properties (hardness, density, tablet thickness, API properties and excipient particle size) in the TRS spectra are better understood, more research on the topic is needed and expected in the near future. This technology seems well-suited to real-time online measurements that require high speed testing.
RC:
One of the most promising trends in Raman spectroscopy can be summed up in one word, “speed”. Raman is typically
thought of as lumbering, requiring long integration times (several minutes per point) in order to acquire enough photons to pull the signal out of the noise. Nowadays, due to a convergence of various technologies, the technique is capable of high speed point sampling as well as rapid hyperspectral imaging.
linear (perhaps even non-linear) process variables must be identified and studied to determine the relationship among these variables and find out how they may correlate to the process composition in order to predict and control final product quality attributes. This is the essence of what it means to move beyond traditional laboratory analysis. Essentially, you cannot conduct PAT very effectively using current classical linear models as they do not allow the user to predict multiple input functions simultaneously or use those inputs to affect a feed-forward or feed- back control strategy in real time during manufacturing. Multivariate Analysis (MVA), strictly speaking, chemometrics, offers an approach for understanding process and product throughout the product lifecycle, from early development to market monitoring, and simultaneously measuring and controlling the final product quality attributes during manufacturing processing. Other mathematical systems certainly exist (Bayesian Statistics and Monte-Carlo Simulations are popular), however, most MVA techniques work well with smaller populations and generally fewer factors are required for modeling. The models are generally robust (if the correct factors and the correct number of factors are chosen), accurate and precise.
MS: This is due to the maturation
of spectrum-stabilized diode lasers that can now direct hundreds of milliwatts into a submicron spot allowing for high speed confocal imaging or, contrastively, into multi-watt line scanners for use on conveyor systems.
Not all of the speed is coming from higher power though. Taking advantage of the compact size and lower power consumption of free- space single mode sources to build smaller handheld and portable systems is an emerging trend. These systems use the Raman effect’s dependency on power density (not absolute power) to acquire data in a fraction of the time of traditional portable Raman spectrometers. These improvements in testing throughput allow for improved efficacy in drug development, quality control, and production, furthering the already rapid adoption of Raman spectroscopy in the pharmaceutical Industry.
Discuss how chemometrics contributes to a successful PAT platform.
GR:
While chemometrics is not a requirement for a successful PAT implementation, (there are other non-spectroscopic systems
such as rapid microbial enumeration and viability methods that do not require chemometrics), in order to “achieve process understanding”, meaning a mechanistic (or a mathematical) understanding of the manufacturing process’ impact on the process composition, the PAT Guidance suggests that users “Develop mathematical relationships between product quality attributes and measurements of critical material and process attributes”. This is broadly accepted to mean that multiple
24 | | November/December 2013
Chemometric tools are an essential part of spectroscopy and PAT for classification, quantitative prediction and
process modeling. They provide the capability to measure and ultimately control process variance. For example, chemometrics permits the rapid mapping of batch process trajectories, as well as the quantitative measurement of critical quality attributes (CQAs) of raw materials, intermediates and finished products. In fact, multivariate tools provide the relationships between process variables and CQAs that lead to process understanding. Chemometric tools are essential for routine troubleshooting because they aid in detecting analyzer faults in addition to distinguishing unexpected material and process equipment changes. Chemometric tools provide continuous validation to detect sample outliers that are outside the experience of the established methods.
BM: CC:
From my perspective chemometrics is a driver for effective implementation of accurate and reproducible sensors and
analyzers. Without efficient data analysis protocols and models, it would be nearly impossible to handle the volume of data being produced. Chemometric routines provide the ability to quickly reduce data sets to pertinent information that can be used for further understanding or control.
Chemometrics is one of the key enablers of PAT. It plays an important role during PAT method development, specifically
the creation of models based on latent variable techniques such as Principal Component Analysis and Partial Least Squares Regression. Timely and robust predictions based on robust models are a requirement in the successful implementation of the PAT platform. Through the use of chemometrics, which encompasses DOE, multivariate calibration and multivariate data analysis, it is increasingly common to look at more complex approaches to support development of a formulation, process or product. Multi-factorial analyses, which encompass the interpretation of large amounts of data from multiple PAT probes and sources, are replacing One-Factor-At-a-Time approaches. New software functionality enables the industry to apply chemometrics broadly and integrate results into the ICH Q8 concepts (risks assessment and design
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68 |
Page 69 |
Page 70 |
Page 71 |
Page 72 |
Page 73 |
Page 74 |
Page 75 |
Page 76 |
Page 77 |
Page 78 |
Page 79 |
Page 80 |
Page 81 |
Page 82 |
Page 83 |
Page 84 |
Page 85 |
Page 86 |
Page 87 |
Page 88 |
Page 89 |
Page 90 |
Page 91 |
Page 92