Lapkin Lab

We've heard from many people that they'd like to use historical data to speed up reaction optimization. Out now in @ACSCentSci, we demonstrate a first step in this direction using multitask Bayesian optimization (MTBO)! pubs.acs.org/doi/10.1021/acs…

Trying to track down that extra impurity messing up your reactions? Check out our new paper on predicting reaction impurities in @ChemMethods! chemistry-europe.onlinelibra… #chemtwitter

Our new paper creates a workflow for predicting protodeboronation rates from a SMILES string! Amazing work from @DanielSWigh! #chemtwitter pubs.acs.org/doi/10.1021/acs…

Want to understand the latest in chemical reaction optimization? Check out our new review in @ACSChemRev! pubs.acs.org/doi/10.1021/acs…

Congrats to @alexanderpomb on his new paper on active learning driven pH adjustment. On average, less than 4 experiments were required to adjust the pH of a wide range of buffer systems! A collaboration with @BASF doi.org/10.1016/j.cej.2022.1…

Abstract submissions for 4th annual ML in (Bio)Chemical Engineering Conference due this Sunday 30 May docs.google.com/forms/d/e/1F… #chemtwitter #realtimechem @cebcambridge

We're excited to to host the 4th annual ML in (Bio)Chemical Engineering Conference this summer on 6-7 July! Registration and keynotes here: mabc-cambridge.ai/ We're also accepting abstracts for talks/posters (deadline 30 May): docs.google.com/forms/d/1r8S…

We made better shampoo! We used robotics + machine learning to improve formulations: cell.com/cell-reports-physic… TL;DR: Our algorithm significantly outperforms traditional DoE but with ~60% of the experiments! @cebcambridge @CellPressNews @CroninLab youtube.com/watch?v=ekNNnfVV…

Our new paper introduces the first integration of a CO2 electroreduction (a carbon capture technique) into an existing bulk chemical process (ethylene oxide production)! arxiv.org/abs/2012.03198 #Realtimechem #ClimateChange

Our recent work on using reinforcement learning for chemical process development! sciencedirect.com/science/ar… #MachineLearning #sustainability