Emil W. Ciurczak, DoraMaxx Consulting11.09.21
For the past several columns, I have been concentrating on the “view from 35,000 feet.” That is, I was concentrating on the supply chain crisis due to the pandemic and other such factors and not the little details or render any suggestions for corrections. Now, it is time to be (more or less) true to the column’s name and discuss some modern analysis equipment for process and R&D work. But, one last supply chain factoid before hardware.
The cost of goods sold (COGS) and, therefore, profits are based on many factors: 1. Actual price of raw materials; 2. Storage unit costs (heat, lights, personnel); 3. Shipping costs of materials; 4. Production costs (labor, machinery, HVAC, taxes) 5. Losses (materials lost in-process, OOS lots, recalls); and 6. Shipping costs of products.
Why, you might ask did I include points three and six as major points? Quite simply, the world-wide supply chain has been an anchor on our economy for over a year. Renting a typical shipping container pre-COVID cost about $2,000. Now, if you want something delivered quickly (relative to the new normal), it could cost $25,000 per container. And that could still take 3-4 times longer than pre-2020. Even “normal,” non-premium deliveries can take 10-12 weeks longer than pre-2020.
Since these have sadly become fixed costs and shortages of petroleum (Texas freeze and Louisiana hurricane damage) have increased electricity and heating costs, short of drastically increasing prices, some way must be found to streamline production costs. Ordering more raw materials than immediately needed for “emergency use” has added even more strain on the system, delaying deliveries even more. So, what’s left for us to do?
How about bringing your processing out of the 1980s? Yes, I am referring to PAT/QbD and, gasp, even continuous manufacturing. In addition to dosage forms becoming more complex (layered, osmotic pumps), it seems that biopharmaceuticals will be more and more common—not just vaccines and antibiotics, but gene therapies, etc., as well. These newer products just scream for better process control: real-time control, because old-fashioned, lab-based batch (GMP) analyses and off-line “controls” will not do the trick.
Over the next several columns, I will discuss the instrumentation and applications of “new and improved” analytical instruments. Some, previously only used for development, will clearly be useful for process control, enhancing quality and reducing costs. They will also become commonplace in the CMO/generics world where dosage forms become complex and biotech becomes more and more important.
I recently attended my first conference in a year and a half (Scix 2021 in Rhode Island), so I got to see some of the things the instrument companies have been building/improving during the pandemic. One of the technologies that seems to have made the biggest leap forward is Raman (you thought I was going to say NIRS?). When I was introduced to Raman (c. 1970), it used a pretty strong LASER, primitive optical systems, and (gasp!) no computer assistance. In fact, when we looked at the early unit at Rutgers (Newark) the intense LASER literally blew up our sample!
Several improvements have been made in modern Raman instruments (aside from a plethora of hand-held models): LASER diodes are replacing older “normal” LASERs and computers (on chips, even) are much more powerful. The biggest downside to using a Raman instrument has always been the fact that, for the strongest Raman spectrum, the shorter the wavelength (higher power), the stronger the Raman effect. The downside? The shorter the wavelength, the stronger the fluorescence background. Using longer wavelength LASERs reduces the fluorescence, but the Raman signal diminishes by the fourth power as the wavelength of the excitation light goes to longer wavelengths. So, while you get rid of most of the background interference, you lose most of your Raman spectrum, too. Figure 1 (above in image slider) shows the effects of different wavelengths.
How do the newer models overcome the physics of the technique? Well, several instrument manufacturers have come up with “work-arounds.” They include:
If we do not already routinely use NIRS for blend uniformity and are forced to take discreet samples, Raman may be used on the powders, themselves, in lieu of dissolving the sample for UV or HPLC analysis. While not necessarily a true PAT method, it is quite a time-saver, not to mention saving on chemical costs (purchase, storing, disposal) and not even requiring a hood for analysis (another safety fact).
However, where Raman truly shines in is aqueous solutions. Without giving a lecture on selection rules, suffice it to say that water, an extremely strong infrared (mid and near) absorber, with its permanent dipole, is an extremely weak Raman scattering molecule. The result is that, while solid matrices generate interfering fluorescence, water is a mild inconvenience. This makes Raman far superior to IR and NIR for solutions (liquid dosage forms).
Raman can be used in production to measure product homogeneity and API(s) level in real time. With its ability to be focused beyond a containers’ wall, it may be used as a quality check on vials and bags of liquid pharmaceutical products in real time. This ability to vary the focal point would allow the unit to measure either the contents or the composition of a plastic bag.
As we will see in subsequent columns, instrumentation that was considered only for R&D has been modified and redesigned for portability or installation in process streams. Techniques such as chromatography, terahertz spectroscopy, and many others have been reimagined for “in the field” use and will be highlighted in later columns.
Emil W. Ciurczak
DoraMaxx Consulting
Emil W. Ciurczak has worked in the pharmaceutical industry since 1970 for companies that include Ciba-Geigy, Sandoz, Berlex, Merck, and Purdue Pharma, where he specialized in performing method development on most types of analytical equipment. In 1983, he introduced NIR spectroscopy to pharmaceutical applications, and is generally credited as one of the first to use process analytical technologies (PAT) in drug manufacturing and development.
The cost of goods sold (COGS) and, therefore, profits are based on many factors: 1. Actual price of raw materials; 2. Storage unit costs (heat, lights, personnel); 3. Shipping costs of materials; 4. Production costs (labor, machinery, HVAC, taxes) 5. Losses (materials lost in-process, OOS lots, recalls); and 6. Shipping costs of products.
Why, you might ask did I include points three and six as major points? Quite simply, the world-wide supply chain has been an anchor on our economy for over a year. Renting a typical shipping container pre-COVID cost about $2,000. Now, if you want something delivered quickly (relative to the new normal), it could cost $25,000 per container. And that could still take 3-4 times longer than pre-2020. Even “normal,” non-premium deliveries can take 10-12 weeks longer than pre-2020.
Since these have sadly become fixed costs and shortages of petroleum (Texas freeze and Louisiana hurricane damage) have increased electricity and heating costs, short of drastically increasing prices, some way must be found to streamline production costs. Ordering more raw materials than immediately needed for “emergency use” has added even more strain on the system, delaying deliveries even more. So, what’s left for us to do?
How about bringing your processing out of the 1980s? Yes, I am referring to PAT/QbD and, gasp, even continuous manufacturing. In addition to dosage forms becoming more complex (layered, osmotic pumps), it seems that biopharmaceuticals will be more and more common—not just vaccines and antibiotics, but gene therapies, etc., as well. These newer products just scream for better process control: real-time control, because old-fashioned, lab-based batch (GMP) analyses and off-line “controls” will not do the trick.
Over the next several columns, I will discuss the instrumentation and applications of “new and improved” analytical instruments. Some, previously only used for development, will clearly be useful for process control, enhancing quality and reducing costs. They will also become commonplace in the CMO/generics world where dosage forms become complex and biotech becomes more and more important.
I recently attended my first conference in a year and a half (Scix 2021 in Rhode Island), so I got to see some of the things the instrument companies have been building/improving during the pandemic. One of the technologies that seems to have made the biggest leap forward is Raman (you thought I was going to say NIRS?). When I was introduced to Raman (c. 1970), it used a pretty strong LASER, primitive optical systems, and (gasp!) no computer assistance. In fact, when we looked at the early unit at Rutgers (Newark) the intense LASER literally blew up our sample!
Several improvements have been made in modern Raman instruments (aside from a plethora of hand-held models): LASER diodes are replacing older “normal” LASERs and computers (on chips, even) are much more powerful. The biggest downside to using a Raman instrument has always been the fact that, for the strongest Raman spectrum, the shorter the wavelength (higher power), the stronger the Raman effect. The downside? The shorter the wavelength, the stronger the fluorescence background. Using longer wavelength LASERs reduces the fluorescence, but the Raman signal diminishes by the fourth power as the wavelength of the excitation light goes to longer wavelengths. So, while you get rid of most of the background interference, you lose most of your Raman spectrum, too. Figure 1 (above in image slider) shows the effects of different wavelengths.
How do the newer models overcome the physics of the technique? Well, several instrument manufacturers have come up with “work-arounds.” They include:
- “Gated” instruments. This means an extremely short pulse of LASER beam, which excites the sample, resulting in both Raman and fluorescence light. Since Raman is based on scattering (elastic and in-elastic) and fluorescence is a true absorbance phenomenon, the Raman radiation emerges (largely) before the unwanted background. A “light-gate” (or window that rapidly opacifies) allows the bulk of the Raman signal to reach the detectors, but “closes” in the face of the fluorescence. The difference is seen in Figure 1. The analyst can choose the best wavelength for the Raman spectrum expected without worry of background.
- Spatially Offset Raman. Another method for avoiding the dreaded fluorescence background is to simply illuminate the sample at one point and collect the radiation from a point slightly off set or away from the illuminated point. Raman has a larger sphere of illumination—it is not as easily reabsorbed by the sample matrix—while the fluorescent light (higher energy) is rapidly absorbed by the sample. Figure 2 illustrates this phenomenon.
- Mathematically manipulation. The combined spectrum is manipulated/massaged to enhance the Raman signal and suppress the background. Several companies have units with built-in AI (artificial intelligence) based algorithms and have been largely (to date) used for raw material and dosage form identification. The technology may be used for quantitative as well as qualitative analyses (Figure 3).
If we do not already routinely use NIRS for blend uniformity and are forced to take discreet samples, Raman may be used on the powders, themselves, in lieu of dissolving the sample for UV or HPLC analysis. While not necessarily a true PAT method, it is quite a time-saver, not to mention saving on chemical costs (purchase, storing, disposal) and not even requiring a hood for analysis (another safety fact).
However, where Raman truly shines in is aqueous solutions. Without giving a lecture on selection rules, suffice it to say that water, an extremely strong infrared (mid and near) absorber, with its permanent dipole, is an extremely weak Raman scattering molecule. The result is that, while solid matrices generate interfering fluorescence, water is a mild inconvenience. This makes Raman far superior to IR and NIR for solutions (liquid dosage forms).
Raman can be used in production to measure product homogeneity and API(s) level in real time. With its ability to be focused beyond a containers’ wall, it may be used as a quality check on vials and bags of liquid pharmaceutical products in real time. This ability to vary the focal point would allow the unit to measure either the contents or the composition of a plastic bag.
As we will see in subsequent columns, instrumentation that was considered only for R&D has been modified and redesigned for portability or installation in process streams. Techniques such as chromatography, terahertz spectroscopy, and many others have been reimagined for “in the field” use and will be highlighted in later columns.
Emil W. Ciurczak
DoraMaxx Consulting
Emil W. Ciurczak has worked in the pharmaceutical industry since 1970 for companies that include Ciba-Geigy, Sandoz, Berlex, Merck, and Purdue Pharma, where he specialized in performing method development on most types of analytical equipment. In 1983, he introduced NIR spectroscopy to pharmaceutical applications, and is generally credited as one of the first to use process analytical technologies (PAT) in drug manufacturing and development.