Over the past three decades, pharmaceutical manufacturing has undergone a fundamental transformation as chemical engineering became more deeply integrated into process development. In turn, it has changed how manufacturers develop, scale, control, and produce increasingly complex medicines.

That shift in process development has laid the foundation for many of today’s manufacturing advances, including process control, automation, and emerging catalytic technologies such as biocatalysis. While chemistry is central to drug development, it is increasingly the integration of engineering principles that is shaping how modern medicines are manufactured.

Manufacturing doesn’t end when the reaction does

Prior to the last 30 years, process development was largely driven from the perspective of synthetic chemists, says Thomas D. Roper, PhD, co-director of pharmaceutical engineering at Virginia Commonwealth University’s Center for Pharmaceutical Engineering and Sciences and graduate program director for chemical and life science engineering.

With more than 30 years of experience in small molecule manufacturing — including 22 years at GSK Pharmaceuticals — Roper said process chemists were primarily focused on designing and optimizing the chemical reaction itself. Once a successful reaction had been identified, comparatively less attention was devoted to the downstream operations that helped determine product quality, manufacturability, and process sustainability.

“The rest of the unit operations, like all the separations, the filtrations, crystallizations, and drying steps were not worked out particularly well, because most chemists would lose interest after the chemistry is over,” Roper said.

Chemical engineers, however, have played a central role in strengthening many of those downstream operations — a finding of the research article The Role of Chemical Engineers in the Pharmaceutical Industry, published in the Annual Review of Chemical and Biomolecular Engineering.

“Given the pharmaceutical industry’s stringent requirements for high-purity drug substances and detailed impurity profiles, it is no surprise that significant innovation has occurred in separation technologies such as crystallization, filtration, and chromatography,” the authors wrote.

Roper added that the industry’s chemist approach often viewed scale-up as reproducing laboratory reactions at larger volumes rather than redesigning processes around the realities of manufacturing.

“It kind of undervalued the expertise that engineering could bring,” he said. “It also brought about some problems when it came to tech transfer, because manufacturing doesn’t really look like a laboratory on a massive scale. It looks like an engineering facility.”

As manufacturing processes became more complex and companies scaled commercial production throughout the 1990s, organizations like GSK expanded the role of chemical engineers within process development teams.

“The engineers helped chemists see what a partnership with chemical engineers looks like and how that could improve the quality of what we did in process development,” Roper said. “The addition of engineering was a very important step.”

Engineering transforms process development

Although chemical engineers have supported pharmaceutical manufacturing for more than a century, Roper said the 1990s represented a turning point in how process development organizations were structured.

Beyond improving unit operations such as separations, chemical engineers also expanded their influence through advances in reaction engineering, process control, and continuous manufacturing. During the 1990s, they pioneered the use of reaction calorimetry to guide scale-up, ushering in data-rich experimentation and process analytical technologies (PAT), while making engineering-led reaction development integral to process safety and operability.

Through these advances, chemical engineering has accelerated the industry’s adoption of quality by design, where quality is intentionally built into a manufacturing process rather than relying primarily on end-product testing.

That evolution continues today through the industry’s growing focus on quality by control, which seeks to move beyond understanding manufacturing processes toward actively controlling critical process parameters through automated systems, according to Roper.

“That’s the idea of quality by control,” Roper said. “If your product is within specification, then you’re automatically collecting it. If your product is not within specification, then there’s some diversion protocol. I think these are very important things that are happening now within the industry.”

While the concept is gaining traction, Roper believes the automation infrastructure needed to fully implement quality by control is still developing.

“I think quality by control is probably just getting off the ground as far as really being able to control all of your critical process parameters so that you're arriving at your product which meets your critical quality attributes,” he observed.

As manufacturers continue investing in automation, digital technologies, and advanced process control strategies, Roper sees pharmaceutical manufacturing becoming increasingly data-driven, with systems capable of monitoring and optimizing production in real time.

That increasingly data-rich manufacturing environment is also creating opportunities for machine learning, which Roper sees as the most practical near-term application of artificial intelligence (AI) in pharmaceutical manufacturing. Because manufacturers already generate enormous volumes of process data, he believes machine learning is well suited to support process optimization and other data-intensive applications across the research and manufacturing continuum.

While today’s models are typically developed for specific datasets or manufacturing challenges rather than serving as general-purpose AI systems, he expects machine learning to see broad adoption because “it can be applied anywhere you’ve got some data,” he said.

Those engineering advances are also creating opportunities for entirely new manufacturing approaches — including catalytic technologies such as biocatalysis — that rely on deeper process understanding, reaction engineering, and scalable manufacturing.

Biocatalysis expands the manufacturing toolbox

Among the technologies Roper sees benefiting from this engineering-driven evolution is biocatalysis, which uses enzymes to carry out chemical transformations that can be difficult to achieve through conventional synthetic chemistry.

“It’s a very important capability that has been evolving over the past 10, 20 years and continues to be impressive,” Roper said, pointing to advances by companies including GSK and Merck as examples of how biocatalysis is becoming an increasingly practical manufacturing technology.

One example of those advances was recently demonstrated by Merck researchers, who described in a Science paper a new enzyme-driven manufacturing process for enlicitide decanoate, an investigational oral macrocyclic peptide designed to lower LDL cholesterol. Rather than relying solely on traditional solid-phase peptide synthesis, the team developed a biocatalytic manufacturing route that significantly reduced the number of process steps while improving scalability.

David Thaisrivongs, executive director, head of biocatalysis at Merck and one of the paper's authors, discussed this research in-depth during an episode of Pharma Manufacturing’s Off Script podcast, and detailed how the company recognized early in development that conventional peptide synthesis would struggle to support the manufacturing volumes needed for a global commercial product.

“While we could make it using conventional synthetic approaches, if that was the best we could do, we would not be able to achieve our aspirations of bringing this important molecule to patients worldwide,” Thaisrivongs said.

Rather than simply improving an existing manufacturing process, the team reimagined the manufacturing route itself. Because enzymes can selectively perform reactions without relying on the extensive use of protecting groups required in traditional peptide chemistry, manufacturers can eliminate unnecessary synthetic steps, simplify downstream processing, and develop routes that would be difficult to achieve through conventional chemistry alone.

The Merck team also reduced reliance on chromatography through crystallization strategies that improved purification efficiency and helped reduce both manufacturing time and cost.

Looking beyond traditional small molecules

The implications of that manufacturing evolution extend well beyond a single technology or therapeutic class. As pharmaceutical pipelines continue shifting toward increasingly complex modalities, both Roper and Thaisrivongs see manufacturing innovation becoming just as important as molecular innovation.

“When I started at Merck, the world of pharmaceuticals was very binary, at least in my mind,” Thaisrivongs said. “There were the small molecules — that’s sort of my training as an organic chemist — and there were the large molecules, the antibodies, and not so much in between.”

That distinction has blurred.

“Now we have peptides, we have oligonucleotides of various lengths, we have conjugates between small molecules and large molecules,” Roper said. “At one time you could just look at it as large and small molecules. Now, honestly, it’s just a molecule.”

For manufacturers, that evolution places even greater emphasis on developing flexible process technologies capable of handling increasingly diverse molecular architectures.

Roper believes the industry’s growing reliance on engineering disciplines — from reaction engineering and advanced process control to automation and biocatalysis — has fundamentally changed how pharmaceutical manufacturing approaches that challenge.

As manufacturers continue adopting automation, AI, advanced process analytics, and enzyme-enabled manufacturing routes, competitive advantage will increasingly come not only from discovering new medicines but from developing smarter, more efficient ways to manufacture them.