Amgen Chemical and Process Development Event

With about 80 participants, the Best Practices in Chemical and Process Development Information Sharing Event organized by METTLER TOLEDO and kindly hosted by Amgen in Cambridge, MA on April 10, was clearly a success. It featured best practices presentations around the theme of chemical development in the local pharma industry from

  • Rob Milburn (Amgen)
  • Theo Martinot (Infinity)
  • Steven Ferguson (Biogen)
  • Jon Lorenz and Nizar Haddad (Boehringer Ingelheim).

Rob described the use of Raman spectroscopy as a Process Analytical Technology (PAT) to better understand, and control, an oxidation step of a sulfide into a sulfoxide with hydrogen peroxide. The investigation involved a team of chemists for screening and optimization of reaction conditions, as well as chemical engineers for safety evaluation, and kinetic modeling developed using quantitative modeling from PAT analysis. This great presentation provided the opportunity to better understand the use of kinetic data to predict and control a process, how PAT methodology can be developed from lab scale to plant scale, and the use of lab safety studies to ensure process safety across scale-up. The process was eventually implemented in-house at the 10 kg scale, and further developed with an outside partner.

Theo Martinot gave an impressive account on the use of real time in situ ATR-FTIR, chemometrics tools, and 25-100 mL scale calorimetry to control a double deprotonation reaction of a benzyl derivative using a strong alkyl lithium base. In addition to being able to monitor the reaction in real time, when chromatography cannot, the goal of PAT (both ATR-FTIR and calorimetry) here is also to be able to better control the reaction, and therefore product quality, as the reaction is progressively scaled-up. We learned that the first deprotonation step is much more exothermic than the second one and, through 3-dimension FTIR surface views, understood the reaction progression as a function of time and process operations. Theo emphasized the use of reaction heat, quantitatively measured using calorimetry, as a PAT and made the point, confirmed by Nizar Haddad and Jon Lorenz from Boehringer Ingelheim, that it can also be used at plant scale by monitoring the temperature differential between the reaction mixture and the jacket. Once the resulting double anion molecule is formed, it can react with oxygen resulting from improper sparging. This leads to the formation of a benzyl alcohol side-product which formation should be minimized. A better knowledge of the kinetics and thermodynamics of its formation helped Theo and his team make sure it would not represent a major safety and quality risk. I would just like to add this fascinating article from Eli Lilly on the topic of control oxygen level in a manufacturing environment: Org. Process Res. Dev. 2014, 18, 246−256.

The presentation from Steven Ferguson (currently with Biogen) was based on his PhD work with the Glennon Research Group (SSPC) at University College Dublin (UCD). He laid out for us the specifics, and relative benefits of particle crystallization using the various following methodologies:

  • Batch
  • Plug Flow
  • Mixed-Suspension, Mixed-Product-Removal (MSMPR).

Being a non-expert myself, like most of the audience, but familiar with the use of real time in situ particle chord length measurement, I found Steven’s presentation to be very educational. Producing small particles to avoid milling using a batch process is challenging because the solid is likely to crash out before a high enough supersaturation is reached. The plug flow approach helps as it allows to directly reaching the right level of concentration, hence supersaturation. Steven presented a case study where the anti-solvent was defined through screening, the optimum plug flow residence time determined to be 30 s (steady state), and de-supersaturation monitored using real time in situ ATR-FTIR coupled with ParticleTrack using Focused Beam Reflectance Measurement (FBRM) technology. It was followed by an investigation of anti-solvent / feed ratio as well as feed concentration. The one-stage mixed-suspension, mixed-product-removal (MSMPR) approach studied at UCD used a 250 – 1000 ml synthesis workstation platform with integrated PAT. This approach can yield a large variety of morphology by tweaking reaction conditions, and produce material on a large scale. For more details view the Webinar: PAT Based Design of Continuous Crystallizations.

Jon Lorenz and Nizar Haddad, both involved in kilo lab and pilot plant manufacturing at Boehringer Ingelheim, gave us a captivating perspective of larger scale use of PAT, e.g. ATR-FTIR. Jon described a multi-step conversion of a carboxylic acid into a nitrile (SO2Cl2, then ammonia) that required further lab investigation after facing scale-up issues at the 100 Gal scale. One of the issues was related to emissions as the large excess of ammonia used in the second step (5-8 equiv.) made proper off-gas scrubbing challenging with existing plant equipment. Jon’s investigation stands out by the use of plant-scale FTIR-ATR that required classified area certification (flammable environment). Nizar explained the results of a two-step approach to the development of dipeptide intermediate Faldaprevir (Org. Process Res. Dev., Articles ASAP, DOI: 10.1021/op400285y, Dec. 24, 2013). The investigation combined screening reaction conditions to find an optimum synthetic route (1. TsCl, NMM, MeCN, 2. Hydroxyproline, MeCN) with the use of ATR-FTIR to characterize its kinetics and thermodynamics properties. ATR-FTIR was used to detect the formation and consumption rates of key intermediates using unique bands (carboxylic acid, amide, oxazolone bonds). The data was next used to generate an empirical kinetic model (Donna Blackmond’s approach as described in Angew. Chem., Int. Ed. 2005, 44, 4302– 4320) applied to determine rate constant and reactant orders, and, beyond, help understand process robustness, and define optimum operation conditions for pilot plant scale-up. Calorimetry results (heat of reaction, adiabatic temperature rise) helped the team ensure process safety and consistency across development from lab through plant scale.

The meeting finished with a walk across the street to enjoy food and drinks. The mid-afternoon break and this post-seminar get-together were undoubtedly the best moments of the day to exchange ideas, share passions, or just get to know each other better.

There will be another similar event in the Cambridge area, focused around biochemistry, in the fall. Plus additional events around the US: Chemical & Process Development in Kissimmee, FL on May 21, Process Development & Scale-up in Woodbridge, NJ on June 19 and Polymerization R&D in Houston, TX in July. I welcome you all at these events!