@article{epps_delgado-licona_yang_kim_volk_han_jun_abolhasani_2023, title={Accelerated Multi-Stage Synthesis of Indium Phosphide Quantum Dots in Modular Flow Reactors}, volume={1}, ISSN={["2365-709X"]}, url={https://doi.org/10.1002/admt.202201845}, DOI={10.1002/admt.202201845}, abstractNote={Development and scalable nanomanufacturing of high‐quality heavy metal‐free quantum dots (QDs) with high‐dimensional experimental design spaces still remain a challenge. In this work, a universal flow chemistry framework for accelerated fundamental and applied studies of heavy metal‐free QDs with multi‐stage chemistries is presented. By introducing flexible time‐ and temperature‐to‐distance transformation using modular fluidic blocks, an in‐flow synthetic route of InP QDs with the highest reported first excitonic absorption peak to valley ratio is unveiled with a reaction time one order of magnitude faster than batch reactors. The flexible time‐ and temperature‐to‐distance transformation as an enabling factor for generalization of flow reactors toward the accelerated discovery, development, and nanomanufacturing of high‐quality emerging nanomaterials for next‐generation energy, display, and chemical technologies is discussed.}, journal={ADVANCED MATERIALS TECHNOLOGIES}, author={Epps, Robert W. and Delgado-Licona, Fernando and Yang, Hyeyeon and Kim, Taekhoon and Volk, Amanda A. and Han, Suyong and Jun, Shinae and Abolhasani, Milad}, year={2023}, month={Jan} }
 @article{volk_epps_yonemoto_masters_castellano_reyes_abolhasani_2023, title={AlphaFlow: autonomous discovery and optimization of multi-step chemistry using a self-driven fluidic lab guided by reinforcement learning}, volume={14}, ISSN={["2041-1723"]}, url={https://doi.org/10.1038/s41467-023-37139-y}, DOI={10.1038/s41467-023-37139-y}, abstractNote={Abstract Closed-loop, autonomous experimentation enables accelerated and material-efficient exploration of large reaction spaces without the need for user intervention. However, autonomous exploration of advanced materials with complex, multi-step processes and data sparse environments remains a challenge. In this work, we present AlphaFlow, a self-driven fluidic lab capable of autonomous discovery of complex multi-step chemistries. AlphaFlow uses reinforcement learning integrated with a modular microdroplet reactor capable of performing reaction steps with variable sequence, phase separation, washing, and continuous in-situ spectral monitoring. To demonstrate the power of reinforcement learning toward high dimensionality multi-step chemistries, we use AlphaFlow to discover and optimize synthetic routes for shell-growth of core-shell semiconductor nanoparticles, inspired by colloidal atomic layer deposition (cALD). Without prior knowledge of conventional cALD parameters, AlphaFlow successfully identified and optimized a novel multi-step reaction route, with up to 40 parameters, that outperformed conventional sequences. Through this work, we demonstrate the capabilities of closed-loop, reinforcement learning-guided systems in exploring and solving challenges in multi-step nanoparticle syntheses, while relying solely on in-house generated data from a miniaturized microfluidic platform. Further application of AlphaFlow in multi-step chemistries beyond cALD can lead to accelerated fundamental knowledge generation as well as synthetic route discoveries and optimization.}, number={1}, journal={NATURE COMMUNICATIONS}, author={Volk, Amanda A. and Epps, Robert W. and Yonemoto, Daniel T. and Masters, Benjamin S. and Castellano, Felix N. and Reyes, Kristofer G. and Abolhasani, Milad}, year={2023}, month={Mar} }
 @article{wang_li_ardekani_serrano-lujan_wang_ramezani_wilmington_chauhan_epps_darabi_et al._2023, title={Sustainable materials acceleration platform reveals stable and efficient wide-bandgap metal halide perovskite alloys}, volume={6}, ISSN={["2590-2385"]}, DOI={10.1016/j.matt.2023.06.040}, abstractNote={The vast chemical space of emerging semiconductors, like metal halide perovskites, and their varied requirements for semiconductor applications have rendered trial-and-error environmentally unsustainable. In this work, we demonstrate RoboMapper, a materials acceleration platform (MAP), that achieves 10-fold research acceleration by formulating and palletizing semiconductors on a chip, thereby allowing high-throughput (HT) measurements to generate quantitative structure-property relationships (QSPRs) considerably more efficiently and sustainably. We leverage the RoboMapper to construct QSPR maps for the mixed ion FA1−yCsyPb(I1−xBrx)3 halide perovskite in terms of structure, bandgap, and photostability with respect to its composition. We identify wide-bandgap alloys suitable for perovskite-Si hybrid tandem solar cells exhibiting a pure cubic perovskite phase with favorable defect chemistry while achieving superior stability at the target bandgap of ∼1.7 eV. RoboMapper’s palletization strategy reduces environmental impacts of data generation in materials research by more than an order of magnitude, paving the way for sustainable data-driven materials research.}, number={9}, journal={MATTER}, author={Wang, Tonghui and Li, Ruipeng and Ardekani, Hossein and Serrano-Lujan, Lucia and Wang, Jiantao and Ramezani, Mahdi and Wilmington, Ryan and Chauhan, Mihirsinh and Epps, Robert W. and Darabi, Kasra and et al.}, year={2023}, month={Sep}, pages={2963–2986} }
 @article{bateni_epps_antami_dargis_bennett_reyes_abolhasani_2022, title={Autonomous Nanocrystal Doping by Self-Driving Fluidic Micro-Processors}, volume={3}, ISSN={["2640-4567"]}, url={https://doi.org/10.1002/aisy.202200017}, DOI={10.1002/aisy.202200017}, abstractNote={Lead halide perovskite (LHP) nanocrystals (NCs) are considered an emerging class of advanced functional materials with numerous outstanding optoelectronic characteristics. Despite their success in the field, their precision synthesis and fundamental mechanistic studies remain a challenge. The vast colloidal synthesis and processing parameters of LHP NCs in combination with the batch‐to‐batch and lab‐to‐lab variation problems further complicate their progress. In response, a self‐driving fluidic micro‐processor is presented for accelerated navigation through the complex synthesis and processing parameter space of NCs with multistage chemistries. The capability of the developed autonomous experimentation strategy is demonstrated for a time‐, material‐, and labor‐efficient search through the sequential halide exchange and cation doping reactions of LHP NCs. Next, a machine learning model of the modular fluidic micro‐processors is autonomously built for accelerated fundamental studies of the in‐flow metal cation doping of LHP NCs. The surrogate model of the sequential halide exchange and cation doping reactions of LHP NCs is then utilized for five closed‐loop synthesis campaigns with different target NC doping levels. The precise and intelligent NC synthesis and processing strategy, presented herein, can be further applied toward the autonomous discovery and development of novel impurity‐doped NCs with applications in next‐generation energy technologies.}, number={5}, journal={ADVANCED INTELLIGENT SYSTEMS}, publisher={Wiley}, author={Bateni, Fazel and Epps, Robert W. and Antami, Kameel and Dargis, Rokas and Bennett, Jeffery A. and Reyes, Kristofer G. and Abolhasani, Milad}, year={2022}, month={Mar} }
 @article{epps_volk_reyes_abolhasani_2021, title={Accelerated AI development for autonomous materials synthesis in flow}, volume={12}, ISSN={["2041-6539"]}, url={https://doi.org/10.1039/D0SC06463G}, DOI={10.1039/D0SC06463G}, abstractNote={Autonomous robotic experimentation strategies are rapidly rising in use because, without the need for user intervention, they can efficiently and precisely converge onto optimal intrinsic and extrinsic synthesis conditions for a wide range of emerging materials. However, as the material syntheses become more complex, the meta-decisions of artificial intelligence (AI)-guided decision-making algorithms used in autonomous platforms become more important. In this work, a surrogate model is developed using data from over 1000 in-house conducted syntheses of metal halide perovskite quantum dots in a self-driven modular microfluidic material synthesizer. The model is designed to represent the global failure rate, unfeasible regions of the synthesis space, synthesis ground truth, and sampling noise of a real robotic material synthesis system with multiple output parameters (peak emission, emission linewidth, and quantum yield). With this model, over 150 AI-guided decision-making strategies within a single-period horizon reinforcement learning framework are automatically explored across more than 600 000 simulated experiments – the equivalent of 7.5 years of continuous robotic operation and 400 L of reagents – to identify the most effective methods for accelerated materials development with multiple objectives. Specifically, the structure and meta-decisions of an ensemble neural network-based material development strategy are investigated, which offers a favorable technique for intelligently and efficiently navigating a complex material synthesis space with multiple targets. The developed ensemble neural network-based decision-making algorithm enables more efficient material formulation optimization in a no prior information environment than well-established algorithms.}, number={17}, journal={CHEMICAL SCIENCE}, publisher={Royal Society of Chemistry (RSC)}, author={Epps, Robert W. and Volk, Amanda A. and Reyes, Kristofer G. and Abolhasani, Milad}, year={2021}, month={May}, pages={6025–6036} }
 @article{volk_epps_abolhasani_2021, title={Accelerated Development of Colloidal Nanomaterials Enabled by Modular Microfluidic Reactors: Toward Autonomous Robotic Experimentation}, volume={33}, ISSN={["1521-4095"]}, url={https://doi.org/10.1002/adma.202004495}, DOI={10.1002/adma.202004495}, abstractNote={In recent years, microfluidic technologies have emerged as a powerful approach for the advanced synthesis and rapid optimization of various solution‐processed nanomaterials, including semiconductor quantum dots and nanoplatelets, and metal plasmonic and reticular framework nanoparticles. These fluidic systems offer access to previously unattainable measurements and synthesis conditions at unparalleled efficiencies and sampling rates. Despite these advantages, microfluidic systems have yet to be extensively adopted by the colloidal nanomaterial community. To help bridge the gap, this progress report details the basic principles of microfluidic reactor design and performance, as well as the current state of online diagnostics and autonomous robotic experimentation strategies, toward the size, shape, and composition‐controlled synthesis of various colloidal nanomaterials. By discussing the application of fluidic platforms in recent high‐priority colloidal nanomaterial studies and their potential for integration with rapidly emerging artificial intelligence‐based decision‐making strategies, this report seeks to encourage interdisciplinary collaborations between microfluidic reactor engineers and colloidal nanomaterial chemists. Full convergence of these two research efforts offers significantly expedited and enhanced nanomaterial discovery, optimization, and manufacturing.}, number={4}, journal={ADVANCED MATERIALS}, publisher={Wiley}, author={Volk, Amanda A. and Epps, Robert W. and Abolhasani, Milad}, year={2021}, month={Jan} }
 @article{sitapure_epps_abolhasani_sang-il kwon_2021, title={CFD-Based Computational Studies of Quantum Dot Size Control in Slug Flow Crystallizers: Handling Slug-to-Slug Variation}, volume={60}, url={https://doi.org/10.1021/acs.iecr.0c06323}, DOI={10.1021/acs.iecr.0c06323}, abstractNote={Recently, slug-flow crystallizers (SFCs) have been proposed for continuous manufacturing of colloidal quantum dots (QDs). Despite the intriguing advantages of SFCs for controlled manufacturing of Q...}, number={13}, journal={Industrial & Engineering Chemistry Research}, author={Sitapure, Niranjan and Epps, Robert W. and Abolhasani, Milad and Sang-Il Kwon, Joseph}, year={2021}, pages={4930–4941} }
 @article{volk_epps_yonemoto_castellano_abolhasani_2021, title={Continuous biphasic chemical processes in a four-phase segmented flow reactor}, volume={6}, ISSN={["2058-9883"]}, url={https://doi.org/10.1039/D1RE00247C}, DOI={10.1039/D1RE00247C}, abstractNote={A four-phase segmented flow regime for continuous biphasic reaction processes is introduced, characterized over 1500 automatically conducted experiments, and used for biphasic ligand exchange of CdSe quantum dots.}, number={8}, journal={REACTION CHEMISTRY & ENGINEERING}, publisher={Royal Society of Chemistry (RSC)}, author={Volk, Amanda A. and Epps, Robert W. and Yonemoto, Daniel and Castellano, Felix N. and Abolhasani, Milad}, year={2021}, month={Jul} }
 @article{han_ramezani_tomhon_abdel-latif_epps_theis_abolhasani_2021, title={Intensified continuous extraction of switchable hydrophilicity solvents triggered by carbon dioxide}, volume={23}, ISSN={["1463-9270"]}, url={https://doi.org/10.1039/D1GC00811K}, DOI={10.1039/D1GC00811K}, abstractNote={Green solvent utilization and recovery enabled by switchable hydrophilicity solvents (SHSs), using carbon dioxide as the switching trigger, offer intriguing advantages in sustainable chemistry. To further elevate SHSs, an intensified continuous flow strategy is presented, providing an accurate in situ reaction monitoring and a scalable green solvent extraction route.}, number={8}, journal={GREEN CHEMISTRY}, publisher={Royal Society of Chemistry (RSC)}, author={Han, Suyong and Ramezani, Mahdi and TomHon, Patrick and Abdel-Latif, Kameel and Epps, Robert W. and Theis, Thomas and Abolhasani, Milad}, year={2021}, month={Apr}, pages={2900–2906} }
 @misc{epps_abolhasani_2021, title={Modern nanoscience: Convergence of AI, robotics, and colloidal synthesis}, volume={8}, ISSN={["1931-9401"]}, url={https://doi.org/10.1063/5.0061799}, DOI={10.1063/5.0061799}, abstractNote={Autonomous experimentation and chemical discovery strategies are rapidly rising across multiple fields of science. However, closed-loop material development approaches have not been widely employed in colloidal nanoscience mainly due to the challenges in synthesis space size, sensitivity to reaction conditions, and the complexity of monitoring multiple synthesis outputs. Recent advancements in automated reactor designs for controlled and reproducible nanocrystal synthesis and intelligent experiment selection algorithms are leading to wider propagation of artificial intelligence-guided autonomous experimentation techniques in colloidal nanoscience. This review will cover the current literature on closed-loop, autonomous platforms for accelerated development of colloidal nanomaterials and discuss the critical features and strategies for developing autonomous robotic experimentation systems suitable to problems in colloidal nanoscience, while providing the context, effectiveness, and prospects of each technique. Then, we will discuss some immediate opportunities in the field for more rapid technological advancement and colloidal nanomaterial discovery.}, number={4}, journal={APPLIED PHYSICS REVIEWS}, publisher={AIP Publishing}, author={Epps, Robert W. and Abolhasani, Milad}, year={2021}, month={Dec} }
 @article{sitapure_epps_abolhasani_kwon_2021, title={Multiscale modeling and optimal operation of millifluidic synthesis of perovskite quantum dots: Towards size-controlled continuous manufacturing}, volume={413}, url={https://www.sciencedirect.com/science/article/pii/S1385894720340249}, DOI={https://doi.org/10.1016/j.cej.2020.127905}, abstractNote={Inorganic lead halide perovskite quantum dots (QDs) have emerged as a promising semiconducting nanomaterial candidate for widespread applications, including next-generation solar cells, displays, and photocatalysts. The optoelectronic properties of colloidal QDs are majorly dictated by their bandgap energy (related to their size). Thus, it is important to fine-tune the size while having fast and continuous production of QDs. However, the mass and heat transfer limitations of batch reactors with batch-to-batch variations have hindered precise control over the size-dependent optoelectronic properties of QDs. Thus, to address this knowledge gap, we propose a multiscale model for continuous flow manufacturing of colloidal perovskite QDs. Specifically, a first-principled kinetic Monte Carlo model is integrated with a continuum model to describe a plug-flow crystallizer (PFC). The PFC has two manipulated inputs, precursor concentration and superficial flow velocity, to fine-tune the size of QDs. Furthermore, a neural network based surrogate model is designed to identify an optimal input trajectory which will ensure that the desired QD size is achieved, thereby taking a step towards controlled and reliable nanomanufacturing of QDs.}, journal={Chemical Engineering Journal}, author={Sitapure, Niranjan and Epps, Robert and Abolhasani, Milad and Kwon, Joseph Sang-Il}, year={2021}, pages={127905} }
 @article{sitapure_epps_abolhasani_kwon_2021, title={Multiscale modeling and optimal operation of millifluidic synthesis of perovskite quantum dots: Towards size-controlled continuous manufacturing}, volume={413}, ISSN={["1873-3212"]}, DOI={10.1016/j.cej.2020.127905}, abstractNote={Inorganic lead halide perovskite quantum dots (QDs) have emerged as a promising semiconducting nanomaterial candidate for widespread applications, including next-generation solar cells, displays, and photocatalysts. The optoelectronic properties of colloidal QDs are majorly dictated by their bandgap energy (related to their size). Thus, it is important to fine-tune the size while having fast and continuous production of QDs. However, the mass and heat transfer limitations of batch reactors with batch-to-batch variations have hindered precise control over the size-dependent optoelectronic properties of QDs. Thus, to address this knowledge gap, we propose a multiscale model for continuous flow manufacturing of colloidal perovskite QDs. Specifically, a first-principled kinetic Monte Carlo model is integrated with a continuum model to describe a plug-flow crystallizer (PFC). The PFC has two manipulated inputs, precursor concentration and superficial flow velocity, to fine-tune the size of QDs. Furthermore, a neural network based surrogate model is designed to identify an optimal input trajectory which will ensure that the desired QD size is achieved, thereby taking a step towards controlled and reliable nanomanufacturing of QDs.}, journal={CHEMICAL ENGINEERING JOURNAL}, author={Sitapure, Niranjan and Epps, Robert and Abolhasani, Milad and Kwon, Joseph Sang-Il}, year={2021}, month={Jun} }
 @article{abdel-latif_epps_bateni_han_reyes_abolhasani_2021, title={Self-Driven Multistep Quantum Dot Synthesis Enabled by Autonomous Robotic Experimentation in Flow}, volume={3}, url={https://onlinelibrary.wiley.com/doi/abs/10.1002/aisy.202000245}, DOI={https://doi.org/10.1002/aisy.202000245}, abstractNote={Identifying the optimal formulation of emerging inorganic lead halide perovskite quantum dots (LHP QDs) with their vast colloidal synthesis universe and multiple synthesis/postsynthesis processing parameters is a challenging undertaking for material‐ and time‐intensive, batch synthesis strategies. Herein, a modular microfluidic synthesis strategy, integrated with an artificial intelligence (AI)‐guided decision‐making agent for intelligent navigation through the complex colloidal synthesis universe of LHP QDs with 10 individually controlled synthesis parameters and an accessible parameter space exceeding 2 × 10 7 , is introduced. Utilizing the developed autonomous microfluidic experimentation strategy within a global learning framework, the optimal formulation of LHP QDs is rapidly identified through a two‐step colloidal synthesis and postsynthesis halide exchange reaction, for 10 different emission colors in less than 40 min per desired peak emission energy. Using two in‐series microfluidic reactors enables continuous bandgap engineering of LHP QDs via in‐line halide exchange reactions without the need for an intermediate washing step. Using an inert gas within a three‐phase flow format enables successful, self‐synchronized continuous delivery of halide salt precursor into moving droplets containing LHP QDs, resulting in accelerated closed‐loop formulation optimization and end‐to‐end continuous manufacturing of LHP QDs with desired optoelectronic properties.}, number={2}, journal={Advanced Intelligent Systems}, author={Abdel-Latif, Kameel and Epps, Robert W. and Bateni, Fazel and Han, Suyong and Reyes, Kristofer G. and Abolhasani, Milad}, year={2021}, pages={2000245} }
 @article{abdel-latif_epps_bateni_han_reyes_abolhasani_2021, title={Self-Driven Multistep Quantum Dot Synthesis Enabled by Autonomous Robotic Experimentation in Flow}, volume={3}, ISSN={["2640-4567"]}, url={https://doi.org/10.1002/aisy.202000245}, DOI={10.1002/aisy.202000245}, abstractNote={Identifying the optimal formulation of emerging inorganic lead halide perovskite quantum dots (LHP QDs) with their vast colloidal synthesis universe and multiple synthesis/postsynthesis processing parameters is a challenging undertaking for material‐ and time‐intensive, batch synthesis strategies. Herein, a modular microfluidic synthesis strategy, integrated with an artificial intelligence (AI)‐guided decision‐making agent for intelligent navigation through the complex colloidal synthesis universe of LHP QDs with 10 individually controlled synthesis parameters and an accessible parameter space exceeding 2 × 107, is introduced. Utilizing the developed autonomous microfluidic experimentation strategy within a global learning framework, the optimal formulation of LHP QDs is rapidly identified through a two‐step colloidal synthesis and postsynthesis halide exchange reaction, for 10 different emission colors in less than 40 min per desired peak emission energy. Using two in‐series microfluidic reactors enables continuous bandgap engineering of LHP QDs via in‐line halide exchange reactions without the need for an intermediate washing step. Using an inert gas within a three‐phase flow format enables successful, self‐synchronized continuous delivery of halide salt precursor into moving droplets containing LHP QDs, resulting in accelerated closed‐loop formulation optimization and end‐to‐end continuous manufacturing of LHP QDs with desired optoelectronic properties.}, number={2}, journal={ADVANCED INTELLIGENT SYSTEMS}, publisher={Wiley}, author={Abdel-Latif, Kameel and Epps, Robert W. and Bateni, Fazel and Han, Suyong and Reyes, Kristofer G. and Abolhasani, Milad}, year={2021}, month={Feb} }
 @article{bateni_epps_abdel-latif_dargis_han_volk_ramezani_cai_chen_abolhasani_2021, title={Ultrafast cation doping of perovskite quantum dots in flow}, volume={4}, ISSN={["2590-2385"]}, url={https://doi.org/10.1016/j.matt.2021.04.025}, DOI={10.1016/j.matt.2021.04.025}, abstractNote={Among all-inorganic metal halide perovskite quantum dots (PQDs), cesium lead chloride (CsPbCl3) with its large band-gap energy is an excellent candidate for enhancement of PQD radiative pathways through incorporation of additional internal energy transfer within its exciton band gap. In this study, we introduce a post-synthetic chemistry for ultrafast metal cation doping of CsPbCl3 QDs with a high degree of tunability, using a model transition metal impurity dopant, manganese. Due to the fast nature of the post-synthetic metal cation-doping reaction, an engineered time-to-space transformation strategy is employed to unravel the kinetics and fundamental mechanism of the doping process. Using a modular microfluidic platform equipped with a translational in situ absorption and photoluminescence spectroscopy probe, we propose a heterogeneous surface-doping mechanism through a vacancy-assisted metal cation migration. The developed in-flow doping strategy can open new avenues for on-demand optoelectronic properties tuning and scalable precision synthesis of high-quality metal cation-doped PQDs.}, number={7}, journal={MATTER}, publisher={Elsevier BV}, author={Bateni, Fazel and Epps, Robert W. and Abdel-latif, Kameel and Dargis, Rokas and Han, Suyong and Volk, Amanda A. and Ramezani, Mahdi and Cai, Tong and Chen, Ou and Abolhasani, Milad}, year={2021}, month={Jul}, pages={2429–2447} }
 @article{epps_volk_ibrahim_abolhasani_2021, title={Universal self-driving laboratory for accelerated discovery of materials and molecules}, volume={9}, url={http://dx.doi.org/10.1016/j.chempr.2021.09.004}, DOI={10.1016/j.chempr.2021.09.004}, abstractNote={Self-driving laboratories are quickly growing in capability, making research in the exploration of advanced functional materials and molecules on the edge of a new era of productivity. As researchers near the widespread adoption of these powerful tools, we must assess their trajectory and the impact of their future developments. Self-driving laboratories are quickly growing in capability, making research in the exploration of advanced functional materials and molecules on the edge of a new era of productivity. As researchers near the widespread adoption of these powerful tools, we must assess their trajectory and the impact of their future developments. For as long as scientific methods have existed, researchers have desired faster and more efficient methods of experimentation and discovery. With the rapid rise of robotics and artificial intelligence (AI)-guided research, we are on the horizon of a renaissance in chemistry that fulfills these aspirations. Tasks that previously required extensive time, labor, and reagents can now be automatically conducted with greater precision, efficiency, and scope. As a result, researchers may focus on defining the next big scientific problem, employ more creative exploration techniques, and gain access to otherwise unreachable regions of the chemical universe. A self-driving laboratory is comprised of two components: (1) the hardware that automatically prepares the precursors, conducts the experiment, and measures the outcome, and (2) the AI brain (i.e., the data-driven modeling/decision-making strategy), which analyzes the data and autonomously selects the next experiment based on the pre-set objective by a human researcher. The self-driving laboratory serves as an assistant to scientists, who define its objectives as well as initial hypotheses and chemical and physical boundaries. In existing work focusing on advanced functional materials and molecules, the hardware of self-driving laboratories takes a variety of forms. These systems include robotic platforms spanning entire labs1Burger B. Maffettone P.M. Gusev V.V. Aitchison C.M. Bai Y. Wang X. Li X. Alston B.M. Li B. Clowes R. et al.A mobile robotic chemist.Nature. 2020; 583: 237-241Crossref PubMed Scopus (238) Google Scholar down to compact workstations for precursor preparation and sample handling.2Chan E.M. Xu C. Mao A.W. Han G. Owen J.S. Cohen B.E. Milliron D.J. Reproducible, high-throughput synthesis of colloidal nanocrystals for optimization in multidimensional parameter space.Nano Lett. 2010; 10: 1874-1885Crossref PubMed Scopus (169) Google Scholar Such workstations can be integrated with batch or flow reactors for automatically conducting reactions in series or parallel.3Steiner S. Wolf J. Glatzel S. Andreou A. Granda J.M. Keenan G. Hinkley T. Aragon-Camarasa G. Kitson P.J. Angelone D. et al.Organic synthesis in a modular robotic system driven by a chemical programming language.Science. 2019; 363: eaav2211Crossref PubMed Scopus (164) Google Scholar,4Bédard A.C. Adamo A. Aroh K.C. Russell M.G. Bedermann A.A. Torosian J. Yue B. Jensen K.F. Jamison T.F. Reconfigurable system for automated optimization of diverse chemical reactions.Science. 2018; 361: 1220-1225Crossref PubMed Scopus (214) Google Scholar Regarding the AI brain of self-driving laboratories, experiment selection algorithms depend primarily on the nature of the research being conducted. For example, pharmaceutical research has mostly used cheminformatic-based strategies that employ a combination of physical models and literature data to select high probability candidate molecules and reaction synthesis routes.5Coley C.W. Thomas D.A. Lummiss J.A.M. Jaworski J.N. Breen C.P. Schultz V. Hart T. Fishman J.S. Rogers L. Gao H. et al.A robotic platform for flow synthesis of organic compounds informed by AI planning.Science. 2019; 365: eaax1566Crossref PubMed Scopus (256) Google Scholar Conversely, the highly sensitive and multidimensional nature of nanomaterial syntheses has made lab-to-lab and batch-to-batch consistency difficult to achieve, hindering the broad adoption of informed AI methods to nanoscience research. Therefore, self-driving laboratories for nanoscience studies have typically relied on AI algorithms that excel without prior knowledge, such as Bayesian optimization, reinforcement learning, or evolutionary algorithms.6Epps R.W. Bowen M.S. Volk A.A. Abdel-Latif K. Han S. Reyes K.G. Amassian A. Abolhasani M. Artificial Chemist: An Autonomous Quantum Dot Synthesis Bot.Adv. Mater. 2020; 32: e2001626Crossref PubMed Scopus (84) Google Scholar Self-driving experimentation platforms have achieved notable success in both academic research and industry, but application of these technologies by a non-specialized researcher comes with several challenges. First, determining the ideal hardware for experimentation is not straightforward and depends on the material or molecules of interest. Robotic systems integrated with batch reactors are the more versatile approach and correlate directly to most methods found in literature with respect to heat and mass transfer rates. They also have access to most characterization methods a human operator may use. However, these systems have slow sampling rates, consume large quantities of reagents per condition (milliliters to liters), and generally cannot operate for extended periods without user intervention. Parallelized batch reactors and more specialized combinatorial screening systems can be significantly more time and material efficient, with reagent volumes down to a nanoliter scale and sampling rates on the order of thousands per day,7Buitrago Santanilla A. Regalado E.L. Pereira T. Shevlin M. Bateman K. Campeau L.-C. Schneeweis J. Berritt S. Shi Z.-C. Nantermet P. et al.Organic chemistry. Nanomole-scale high-throughput chemistry for the synthesis of complex molecules.Science. 2015; 347: 49-53Crossref PubMed Scopus (285) Google Scholar but they have limited control of the reaction environment and access to precise online characterization methods. Flow reactors are highly efficient (microliter reagent consumption and sampling rates rivaling combinatorial screening) and can combine a large library of online characterization techniques together with precise control over reaction parameters. However, the benefit of flow reactors is a double-edged sword. Their high heat and mass-transfer rates make them an ideal choice for process intensification, but the different heat and mass-transfer dynamics of flow reactors compared with batch make it more difficult for researchers to adopt literature protocols, specifically for nanomaterials, developed through batch reactions. Most critically, flow reactors struggle to accommodate solid reagents, products, and byproducts. Because of the shortcomings of each strategy outlined, completely different automated experimentation strategies are often necessary at different stages of reaction exploration. Therefore, many examples of self-driving labs have been restricted to isolated reaction stages instead of covering the full experimentally accessible parameter space of a specific class of materials or molecules. Beyond these difficulties in optimal platform selection in multi-stage systems, researchers looking to build an autonomous experimental platform must also navigate an absence of readily available equipment. Navigating hardware availability is a problem beyond complex multi-stage reactions. Designing and building a self-driving experimental system from the ground up is costly and requires a considerable time investment. This barrier is not a significant issue for researchers specialized in platform development because the design itself is the end goal of the work, but for a chemist or material scientist aiming to improve a synthesis without advancing an experimentation platform, this limitation creates a large barrier. Furthermore, without consistency in reaction environments, identical input conditions can likely result in different reaction products between two different self-driving platforms equipped with different size reactors. As the capabilities of autonomous systems grow further toward general application, it will be critical for the field to emphasize the development of systems built from accessible and standardized components that produce consistent results. However, historically, widespread adoption of standards in unregulated communities has been driven by either extreme necessity or convenience. In self-driven experimental systems, development efficiency in future studies will likely rely on the latter. One already occurring example of convenience-driven standardization has been the use of tubing-based flow reactors.8Volk A.A. Epps R.W. Abolhasani M. Accelerated Development of Colloidal Nanomaterials Enabled by Modular Microfluidic Reactors: Toward Autonomous Robotic Experimentation.Adv. Mater. 2021; 33: e2004495Crossref PubMed Scopus (27) Google Scholar In these systems, reactions are conducted in commercially available micro scale junctions and tubing channels, typically composed of chemically resistant materials (Teflon or stainless steel), and the dimensions of these channels are manufactured with high precision under standardized dimensions. Consequently, research in developing these tubular flow reactors has formed a library of readily available, high-efficiency experimentation devices with directly transferable heat- and mass-transfer characteristics between systems. Similarly, many flask-based automated systems rely on custom 3D-printed modules coupled with commercial components, both of which may be quickly reproduced and applied in new applications.9Salley D. Keenan G. Grizou J. Sharma A. Martín S. Cronin L. A nanomaterials discovery robot for the Darwinian evolution of shape programmable gold nanoparticles.Nat. Commun. 2020; 11: 2771Crossref PubMed Scopus (28) Google Scholar It is the onus of academic research to emphasize these more accessible variants of automated experimental tools and highlight the importance of accessibility. The next critical steps in autonomous robotic experimentation toward achieving a universal self-driving lab will be (1) the introduction of greater transparency in system design, (2) the modularization and standardization of the hardware, and (3) the creation of open-access datasets for benchmarking and selecting suitable AI modeling and decision-making algorithms. AI, computational, and machine-learning communities have long valued transparency and open resources in academic publications. Equivalent community standards would be hugely advantageous in the field of autonomous robotic experimentation. A self-driving lab capable of autonomous planning and conduction of reactions may be significant in its abilities, but it possesses little functional application to the broader scientific community if it cannot be reconstructed in a different environment. Field standards for reporting of novel autonomous platforms should, therefore, allow for the complete reproduction of the system by an uninformed, skilled scientist and include all associated control and analysis software and relevant component models. Furthermore, widespread publication of all generated experimental data with comprehensive demonstrations of sampling precision would allow for the rapid development of cheminformatic and materials informatics strategies for different classes of materials and molecules. The various biases of published data skew the effectiveness of literature-driven algorithms toward high-performing regions of the chemical universe, and comprehensive reporting, including failed reactions, would fill many of the information gaps not currently covered in the literature. For black-box algorithms, one of the current challenges in many fields is the selection of suitable algorithms for specific scenarios. Off-the-shelf decision-making algorithms often cannot be directly applied to a focused application without further tuning of the meta-decision structure and algorithm parameters. From the perspective of an experimentalist, required algorithm tuning can slow research and even defeat the purpose of applying the algorithm to begin with. Data availability would enable researchers to test new algorithms on multiple experiment-based benchmark surrogate systems, leading to expedited implementation of higher performing algorithms. Although autonomous robotic experimentation will most likely not converge onto a single optimal design, researchers today can make significant gains in shifting the field from its current state of diverse, isolated platforms toward a single unified system of modularized, self-driving labs. No single field in autonomous experimentation possesses all the tools necessary to explore the complexities of the chemical world, but a combined approach could bring the scientific community much closer. Modularization of experimental systems enables cross-disciplinary application of otherwise inaccessible devices and tools. A publicly available library of accessible platform designs with corresponding control and experiment selection algorithms, illustrated in Figure 1, would improve the rate of hybrid system development and lower the barrier for entry among researchers. Furthermore, this modular approach to platform development would provide a direct precursor to autonomous device fabrication and optimization.10MacLeod B.P. Parlane F.G.L. Morrissey T.D. Häse F. Roch L.M. Dettelbach K.E. Moreira R. Yunker L.P.E. Rooney M.B. Deeth J.R. et al.Self-driving laboratory for accelerated discovery of thin-film materials.Sci. Adv. 2020; 6: eaaz8867Crossref PubMed Scopus (128) Google Scholar Sharing of datasets generated with these unified platforms would then provide large quantities of transferable experimental information, leading to higher performing AI models and algorithms and greater mastery of the chemical world. Unifying the direction of research and shifting the current standards of the field will require a concerted effort across academia and industry. One promising step in this direction is the recently established Acceleration Consortium hosted at the University of Toronto. The output of this association and similar future establishments will help reveal the efficiencies of self-driven experimentation platforms to a larger audience. While industry will always have strong incentives to maintain proprietary information about novel materials and molecules, they can benefit from collaborations with academic researchers to develop an accessible modular experimentation core from which the self-driving platforms may be built, applied, and expanded upon. The authors gratefully acknowledge the financial support provided by the National Science Foundation (Award # 1940959) and the UNC Research Opportunities Initiative (UNC-ROI) grant. The lead contact is a Global Member of the Acceleration Consortium.}, journal={Chem}, publisher={Elsevier BV}, author={Epps, Robert W. and Volk, Amanda A. and Ibrahim, Malek Y.S. and Abolhasani, Milad}, year={2021}, month={Sep} }
 @article{epps_volk_abdel-latif_abolhasani_2020, title={An automated flow chemistry platform to decouple mixing and reaction times}, url={https://doi.org/10.1039/D0RE00129E}, DOI={10.1039/D0RE00129E}, abstractNote={Although a vital parameter in many colloidal nanomaterial syntheses, precursor mixing rates are typically inconsistent in batch processes and difficult to separate from reaction time in continuous flow systems. Here, we present a flow chemistry platform that decouples early-stage precursor mixing rates from reaction time (residence time) using solely off-the-shelf, commercially available, and standard dimension components. We then utilize the developed flow chemistry platform towards time- and material-efficient studies of the mass transfer-controlled synthesis of cesium lead bromide perovskite quantum dots.}, journal={Reaction Chemistry & Engineering}, publisher={Royal Society of Chemistry (RSC)}, author={Epps, Robert W. and Volk, Amanda A. and Abdel-Latif, Kameel and Abolhasani, Milad}, year={2020} }
 @article{epps_bowen_volk_abdel‐latif_han_reyes_amassian_abolhasani_2020, title={Artificial Chemist: An Autonomous Quantum Dot Synthesis Bot}, url={http://dx.doi.org/10.1002/adma.202001626}, DOI={10.1002/adma.202001626}, abstractNote={The optimal synthesis of advanced nanomaterials with numerous reaction parameters, stages, and routes, poses one of the most complex challenges of modern colloidal science, and current strategies often fail to meet the demands of these combinatorially large systems. In response, an Artificial Chemist is presented: the integration of machine‐learning‐based experiment selection and high‐efficiency autonomous flow chemistry. With the self‐driving Artificial Chemist, made‐to‐measure inorganic perovskite quantum dots (QDs) in flow are autonomously synthesized, and their quantum yield and composition polydispersity at target bandgaps, spanning 1.9 to 2.9 eV, are simultaneously tuned. Utilizing the Artificial Chemist, eleven precision‐tailored QD synthesis compositions are obtained without any prior knowledge, within 30 h, using less than 210 mL of total starting QD solutions, and without user selection of experiments. Using the knowledge generated from these studies, the Artificial Chemist is pre‐trained to use a new batch of precursors and further accelerate the synthetic path discovery of QD compositions, by at least twofold. The knowledge‐transfer strategy further enhances the optoelectronic properties of the in‐flow synthesized QDs (within the same resources as the no‐prior‐knowledge experiments) and mitigates the issues of batch‐to‐batch precursor variability, resulting in QDs averaging within 1 meV from their target peak emission energy.}, journal={Advanced Materials}, author={Epps, Robert W. and Bowen, Michael S. and Volk, Amanda A. and Abdel‐Latif, Kameel and Han, Suyong and Reyes, Kristofer G. and Amassian, Aram and Abolhasani, Milad}, year={2020}, month={Jul} }
 @article{kerr_epps_abolhasani_2019, title={A low-cost, non-invasive phase velocity and length meter and controller for multiphase lab-in-a-tube devices}, url={https://doi.org/10.1039/C9LC00296K}, DOI={10.1039/C9LC00296K}, abstractNote={The non-invasive, optical phase velocity and length meter/controller effectively measures phase length and velocity in real-time with two low-cost photodetectors.}, journal={Lab on a Chip}, publisher={Royal Society of Chemistry (RSC)}, author={Kerr, Corwin B. and Epps, Robert W. and Abolhasani, Milad}, year={2019} }
 @article{abdel‐latif_epps_kerr_papa_castellano_abolhasani_2019, title={Facile Room‐Temperature Anion Exchange Reactions of Inorganic Perovskite Quantum Dots Enabled by a Modular Microfluidic Platform}, volume={29}, ISSN={1616-301X 1616-3028}, url={http://dx.doi.org/10.1002/adfm.201900712}, DOI={10.1002/adfm.201900712}, abstractNote={Abstract}, number={23}, journal={Advanced Functional Materials}, publisher={Wiley}, author={Abdel‐Latif, Kameel and Epps, Robert W. and Kerr, Corwin B. and Papa, Christopher M. and Castellano, Felix N. and Abolhasani, Milad}, year={2019}, month={Mar}, pages={1900712} }
 @article{epps_felton_coley_abolhasani_2018, title={A Modular Microfluidic Technology for Systematic Studies of Colloidal Semiconductor Nanocrystals}, volume={5}, ISSN={["1940-087X"]}, url={http://dx.doi.org/10.3791/57666}, DOI={10.3791/57666}, abstractNote={Colloidal semiconductor nanocrystals, known as quantum dots (QDs), are a rapidly growing class of materials in commercial electronics, such as light emitting diodes (LEDs) and photovoltaics (PVs). Among this material group, inorganic/organic perovskites have demonstrated significant improvement and potential towards high-efficiency, low-cost PV fabrication due to their high charge carrier mobilities and lifetimes. Despite the opportunities for perovskite QDs in large-scale PV and LED applications, the lack of fundamental and comprehensive understanding of their growth pathways has inhibited their adaptation within continuous nanomanufacturing strategies. Traditional flask-based screening approaches are generally expensive, labor-intensive, and imprecise for effectively characterizing the broad parameter space and synthesis variety relevant to colloidal QD reactions. In this work, a fully autonomous microfluidic platform is developed to systematically study the large parameter space associated with the colloidal synthesis of nanocrystals in a continuous flow format. Through the application of a novel translating three-port flow cell and modular reactor extension units, the system may rapidly collect fluorescence and absorption spectra across reactor lengths ranging 3 - 196 cm. The adjustable reactor length not only decouples the residence time from the velocity-dependent mass transfer, it also substantially improves the sampling rates and chemical consumption due to the characterization of 40 unique spectra within a single equilibrated system. Sample rates may reach up to 30,000 unique spectra per day, and the conditions cover 4 orders of magnitude in residence times ranging 100 ms - 17 min. Further applications of this system would substantially improve the rate and precision of the material discovery and screening in future studies. Detailed within this report are the system materials and assembly protocols with a general description of the automated sampling software and offline data processing.}, number={135}, journal={JOVE-JOURNAL OF VISUALIZED EXPERIMENTS}, publisher={MyJoVE Corp}, author={Epps, Robert W. and Felton, Kobi C. and Coley, Connor W. and Abolhasani, Milad}, year={2018}, month={May} }
 @article{epps_felton_coley_abolhasani_2017, title={Automated microfluidic platform for systematic studies of colloidal perovskite nanocrystals: towards continuous nano-manufacturing}, volume={17}, ISSN={["1473-0189"]}, url={https://doi.org/10.1039/C7LC00884H}, DOI={10.1039/c7lc00884h}, abstractNote={An automated microfluidic platform enables systematic high-throughput studies of mixing enhancement on the emission band-gap of in-flow synthesized perovskite quantum dots, resulting in kinetically tunable nanocrystals.}, number={23}, journal={LAB ON A CHIP}, publisher={Royal Society of Chemistry (RSC)}, author={Epps, Robert W. and Felton, Kobi C. and Coley, Connor W. and Abolhasani, Milad}, year={2017}, month={Dec}, pages={4040–4047} }
 @article{branch_epps_kosson_2017, title={The impact of carbonation on bulk and ITZ porosity in microconcrete materials with fly ash replacement}, volume={10}, DOI={10.1016/j.cemconres.2017.10.012}, abstractNote={Scanning electron microscopy was used to evaluate the porosity patterns found in microconcrete materials to better understand the gaseous diffusion pathways and reaction of CO2 within the bulk cement paste and interfacial transition zone (ITZ) of microconcrete materials containing different fly ash replacement types. Image segmentation was applied to evaluate the porosity as a function of distance from an exposed surface in the bulk cement paste and from aggregate boundaries in the ITZ in both carbonated and non‑carbonated microconcrete materials. The ITZ region remained more porous compared to the bulk cement paste for all microconcrete types after carbonation despite a change in the porosity profile across the ITZ region. Results indicate that both the carbonation reaction capacity and porosity before carbonation impact the decrease in porosity observed as a result of carbonation in the ITZ and bulk cement paste regions.}, journal={Cement and Concrete Research}, publisher={Elsevier BV}, author={Branch, J.L. and Epps, R. and Kosson, D.S.}, year={2017}, month={Oct} }