DARPA targets a new era of rapid, verifiable and repeatable small molecule discovery and scalable synthesis

Synthetic molecules are the foundation for many products critical to the Department of Defense’s mission—from active pharmaceutical ingredients found in a medic’s kit to materials in modern batteries and fuel cells. Current processes for designing and producing new synthetic molecules, however, are very slow and can take years between the initial design of a molecular solution and when it’s available for use in large quantities.

DARPA’s Make-It program seeks to overcome this challenge by developing an automated synthesizer that could transform simple raw materials into known or new molecules defined by the user. The goal of Make-It is to develop a fully automated chemical synthesizer that can produce, purify, characterize and scale a wide range of small molecules. Accelerating the rate of discovery and production of molecules could speed advances in a number of areas important to national security.

“Synthesis is a bottleneck to the discovery and production of new molecules,” said Tyler McQuade, program manager in DARPA’s Defense Sciences Office. “Automated synthesizers exist commercially for discrete systems such as natural biopolymers, but the technology is not applicable across all molecules and scale-up is difficult. The vision for Make-It is to create an automated synthesizer that produces a wide variety of complex small molecules at production scale in weeks instead of years.”

If successful, Make-It would widen access to synthetic chemistry beyond chemists, enabling non-chemists from other scientific disciplines to leverage the power of synthesis for new applications. Another benefit would be to revolutionize reproducibility. Currently, knowledge transfer is a challenge in synthetic chemistry, because conditions and reaction set-ups vary between lab locations worldwide. With a Make-It synthesizer, chemical routes for new molecules could easily be replicated by others as simply as by sending a text message. Additionally, Make-It would make synthesis safer and greener by enabling better control of process inputs such as starting materials and solvents, as well as conditions such as temperature and pressure.

“A system that can automate the design and testing of specific chemical routes, keeping track of both success and failures, will usher in the next phase of synthetic discovery,” McQuade said.

Make-It

The Make-It program in the Defense Advanced Research Projects Agency DARPA Defense Sciences Office (DSO) is soliciting innovative research proposals in the area of research related to the development of an automated chemical synthesizer that can produce, purify, characterize and scale a wide range of small molecules.

The Make-It program will specifically address the development of an automated end-to-end system that includes capabilities for
(i) computational analysis and design of synthetic pathways,
(ii) a reagent/starting material delivery mechanism, and
(iii) interconnected fluidic modules for synthesis and in-line monitoring, purification and formulation.

A successful Make-It system will demonstrate maximal synthetic flexibility using a minimal set of modules by synthesizing several organic small molecule active pharmaceutical ingredients (APIs), including in-line formulation of one API to yield a finished drug product, as well as synthesizing a series of DARPA-defined small molecule challenge targets. DARPA anticipates that Make-It will completely overhaul the current workflow for chemical design, synthesis and scale-up.

The power of automation in chemical synthesis was demonstrated through the development of peptide and nucleic acid synthesizers decades ago. These now-robust commercial systems run a limited set of chemistry in well-defined cycles, allowing specialists and non-specialists access to natural biopolymers. Attempts to perform more complex chemistry using automation have had mixed success. For example, bead-based synthesizers have been demonstrated to perform multistep oligosaccharide synthesis, and parallel vial-based reactors that can automate single unit operations are now established commercial products. Although these systems can yield small amounts of material, their bead-based or batch format make scale-up difficult. To revolutionize the field of automated synthetic chemistry and provide a single, scalable methodology that is capable of synthesizing all families of organic small molecules, DARPA has initiated the MakeIt program. Make-It challenges proposers to take a radical departure from past automated synthesis efforts by incorporating cross-disciplinary expertise such as computational design and optimization of reaction pathways (e.g., network theory, topology and machine learning) and continuous synthesis techniques (e.g., flow and microfluidic chemistry) to create an end-to-end system.

To provide a fully automated system that can be utilized by chemists and non-chemists alike; however, a Make-It platform will have to provide much more than hardware and control systems for automated, continuous synthesis and characterization. The ideal Make-It platform would begin with user input in the form of a single or Markush structure. The platform would generate a retrosynthetic analysis and propose both commercially available starting materials and preliminary synthetic routes. The platform would then test and optimize the routes in context of highest yield, lowest cost, safest route, etc.

Recent advances in computer-aided molecular design and flow chemistry have indicated that the Make-It vision is achievable. For example, developments in machine learning and use of graph theory have led to computational tools that can define possible retrosynthetic pathways. Advances have also been made in computational selection of optimal reaction parameters that could be coupled to the retrosynthetic analysis to enhance route selection. Parallel developments in experimental approaches have begun to address issues such as complex syntheses and scalability using continuous flow technologies, as well as in-line characterization and process control. A recent report, for example, describes a strategy to interchangeably link several flow reactor modules and synthesize different products depending on which modules are used and in what order. Scalability of continuous approaches has been demonstrated using droplet-based techniques by employing longer run times or parallel reactor modules. Finally, in-line characterization and feed-forward/feed-back control systems are emerging that should enable monitoring of continuous synthetic processes in real-time.

If realized, a complete Make-It platform would eliminate the Edisonian approach to chemical synthesis and usher in a new era of rapid, verifiable and repeatable small molecule discovery and scalable synthesis.

SOURCES – DARPA, FBO