Presentations & Speakers
|
|
|
| Developing an Academic High Throughput Synthesis Facility – Towards Data Driven Synthetic Methodology |
| |
|
|
|
Brandon Jolly, PhD
Research and Development Engineer
University of California, Los Angeles (UCLA),
High Throughput Synthesis and Catalysis Facility,
Molecular Instrumentation Center
|
|
 |
Abstract:
The Molecular Instrumentation Center at UCLA’s Department of Chemistry and Biochemistry recently opened a new High Throughput Synthesis and Catalysis Facility (HTSC) in the Summer of 2025 driven by Professor Abigail Doyle. The facility aims to provide automation capabilities to the department’s research with its two Juniors designed for reaction screening and optimization comprised of solid and liquid handling, OSR for high pressure chemistry, photo- and electrochemistry, and heating/cooling/stirring. Particularly the HTSC’s Juniors are apt for collecting reliable, high quality reaction screening datasets for the training of Machine Learning/Bayesian Optimization/Artificial Intelligence models. Leveraging these capabilities the HTSC supports a breadth of ongoing research in subfields such as organic chemistry, inorganic chemistry, chemical biology, and materials, with a central theme of data driven reaction discovery and optimization. |
| |
|
|
Biography:
Brandon Jolly is a research & development engineer in the Department of Chemistry & Biochemistry at UCLA specializing in high throughput experimentation and chromatography. He received his PhD in Inorganic Chemistry at UCLA under Prof. Chong Liu working on spatial control for multi-step organometallic catalysis, where he first gained exposure to data driven experimentation. Upon graduating, he began working with Prof. Abigail Doyle to develop the High Throughput Synthesis & Catalysis Facility (HTSC) at UCLA. Equipped with two Unchained Labs Juniors for reaction screening, the HTSC aims to leverage high throughput automation capabilities to support the diverse research projects ongoing in the department in making data-driven decisions. |
| |
|
|
| |
|
|
|
Self-Driving HTE Labs for Process Development: Closing the Loop with an Autonomous Mobile Robot
|
|
|
|
Dimitrios Chatzinikolis
Robotics Engineer IV
Takeda Pharmaceuticals Company Limited,
Synthetic Molecule Process Development (SMPD),
Pharmaceutical Sciences
|
|
 |
|
Abstract:
Self-driving laboratories (SDLs) capable of executing self-optimizing high-throughput experimentation (HTE) workflows represent a transformative opportunity for synthetic molecule process development, where the exploration of wide parameter spaces across diverse instrumentation is routine. Realizing this vision, however, requires more than algorithm-guided experiment selection — it demands the seamless physical integration of synthesis platforms and analytical instruments into a unified, closed-loop pipeline. The transport of samples between platforms and the manual loading and unloading of equipment remain persistent bottlenecks that limit the throughput and consistency of iterative HTE workflows. Established automation approaches — from gantry systems and track-mounted robotic arms to configurable benchtop automation such as the Unchained Labs Junior — have proven effective for high-throughput screening and workflow coordination. An AMR complements these solutions by bridging instruments across the wider laboratory, integrating into existing footprints without structural changes and scaling flexibly as laboratories expand or repurpose space to accommodate new scientific priorities.
Beyond physical sample transport and instrument loading, we are developing an orchestration layer that leverages API access to programmatically trigger and coordinate instruments across the workflow. Here, we present a closed-loop workflow in which an AMR connects HTE synthesis platforms with analytical characterization, while a machine learning–assisted decision engine suggests subsequent experiments with human-in-the-loop oversight. We discuss our integration approach, early results, and lessons learned in advancing toward fully self-driving HTE labs.
|
| |
|
Biography:
Dimitrios Chatzinikolis is a robotics engineer specializing in lab automation and high-throughput experimentation. A graduate of MIT with a dual masters degree in EECS and Design Computation, he sits at the intersection of physical science and computer science. Dimitrios is driven by the conviction that the future of discovery isn't just about faster robots, but about better-designed experimental architectures. He is currently focused on building the autonomous labs that will define the next generation of scientific breakthroughs as part of Takeda's Synthetic Molecule Process Development (SMPD) department.
|
| |
|
|
| |
|
|
| Toward the Automation of Drug Discovery Chemistry |
| |
|
|
|
Kaori Sasaki
Scientist
Daiichi Sankyo
|
|
 |
Abstract:
We have developed an end-to-end automated synthesis platform to accelerate drug discovery chemistry, integrating synthesis, workup, purification, analysis, and concentration into a unified workflow. Central to this system is the Unchained Labs Big Kahuna, which acts as an integration hub, enabling seamless coordination with third-party instruments across the workflow. In this presentation, we will describe the design of this platform and its role in enabling a data-driven DMTA cycle. Through real project examples, we highlight the automation of complex steps such as purification and analysis, as well as the translation of human decision-making into automated processes. This integrated approach increases experimental throughput and allows researchers to focus more on molecular design and interpretation. This talk is intended for a broad audience, from researchers considering automation to those actively developing and scaling automated laboratory platforms. |
| |
Biography:
Kaori Sasaki is a Scientist at Daiichi Sankyo, where she applies lab automation to accelerate drug discovery using the Unchained Labs Big Kahuna platform. She holds a Master's degree in Organic Synthesis from Tokai University Graduate School. Her foundation in synthetic chemistry continues to inform her approach to automated experimentation. Prior to joining Daiichi Sankyo, Kaori spent approximately five years at Unchained Labs as a Field Automation Scientist, where she gained deep expertise in automation workflows and hands-on experience supporting scientists across a wide range of applications. She is passionate about bridging the gap between traditional synthetic chemistry and modern automated platforms to drive more efficient and innovative drug discovery.
|
| |
|
|
| |
|
|
|
| Synthesis of Highly Crystalline Covalent Organic Frameworks Using Large Language Models and High-throughput Robotics |
| |
|
|
|
Kaiyu Wang, PhD
Postdoctoral Scholar
University of California, Berkeley,
Department of Chemistry
|
|
 |
|
Abstract:
Materials discovery is often limited by slow, heuristic, manual synthesis optimization, where each new structure requires extensive trial and error. Here, we present the Large Language Model For Accelerated Synthesis Technique (LFAST), a workflow that integrates large language model reasoning, robotic synthesis, and high-throughput characterization to accelerate the search for optimal reaction conditions.
Unchained Labs' Junior plays a central role by enabling many parallel reactions to be prepared reproducibly and reliably with minimal manual intervention. Such automation transforms synthesis from a one-at-a-time process into a data-generating platform, allowing rapid feedback between experimental results and AI-guided decision-making.
Using this approach, we significantly improved crystallinity of a benchmark Covalent Organic Framework (COF) and discovered a previously unreported COF within weeks rather than months or years. More broadly, LFAST demonstrates how robotic platforms can serve as the experimental engine for AI-guided materials discovery, making synthesis faster and more reproducible.
|
| |
|
Biography:
Kaiyu Wang is a postdoctoral scholar in the Department of Chemistry at the University of California, Berkeley working with Prof. Omar M. Yaghi. He received his Ph.D. in Chemistry from UC Berkeley, where his dissertation, Data-Driven Reticular Chemistry in the Era of Artificial Intelligence, focused on the intersection of AI, laboratory automation, and reticular materials. His work develops and studies covalent organic frameworks (COFs), combining synthesis and characterization with data-driven approaches to understand and optimize structure–property relationships in porous materials. He is also interested in leveraging large language models to accelerate synthesis design and improve crystallinity outcomes in COF discovery. Kaiyu has co-authored papers in journals including Science, Nature, and JACS.
|
| |
|
|
| |
|
|
