Real-Time Detection of Contaminants in Biopharmaceuticals by AI-Enabled Quantum Coherence Spectroscopy
Abstract
The proposed research seeks to greatly accelerate the state-of-the-art (SOA) in the detection and discrimination of biomolecular mixtures such as those found in drug formulations in support of the fundamental science behind DTRA’s mission of effective chemical and biological (CB) defense. We will develop a reconfigurable approach to interrogate and identify closely related biomolecules in complex mixtures not distinguishable by other non-destructive optical means. The proposed spectroscopy is highly nonlinear with respect to the strength of the optical fields, offering the ability to generate a unique fingerprint for each molecular component in isolation or as part of a mixture. Critically, this method detects the optical response of selected targets across the entire terahertz region of the electromagnetic spectrum (0.1 – 100 THz), but uses a novel photothermal detection scheme with sensitivity approaching the single molecule limit. The high sensitivity and fast acquisition (>100 samples per second) of this method, in combination with additive manufacturing of drug products, will allow the creation of large (>100,000 formulations) spectral databases that encompass all known active pharmaceutical ingredients (APIs), excipients, and contaminants, even those that exist in trace quantities (less than 1% by volume). This large training dataset may then be used within a machine learning context to identify the components of complex mixtures in the field or to provide real-time feedback during drug manufacturing to ensure a safe, reliable, and effective product.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jun 14, 2022
- Source ID
- HDTRA12110026
Entities
People
- Elad Harel
Organizations
- Defense Threat Reduction Agency
- Michigan State University