Technologies such as batteries, biomaterials and heterogeneous catalysts have functions that are defined by mixtures of molecular and mesoscale components. As yet, this multi-length-scale complexity cannot be fully captured by atomistic simulations, and the design of such materials from first principles is still rare1,2,3,4,5. Likewise, experimental complexity scales exponentially with the number of variables, restricting most searches to narrow areas of materials space. Robots can assist in experimental searches6,7,8,9,10,11,12,13,14 but their widespread adoption in materials research is challenging because of the diversity of sample types, operations, instruments and measurements required. Here we use a mobile robot to search for improved photocatalysts for hydrogen production from water15. The robot operated autonomously over eight days, performing 688 experiments within a ten-variable experimental space, driven by a batched Bayesian search algorithm16,17,18. This autonomous search identified photocatalyst mixtures that were six times more active than the initial formulations, selecting beneficial components and deselecting negative ones. Our strategy uses a dexterous19,20 free-roaming robot21,22,23,24, automating the researcher rather than the instruments. This modular approach could be deployed in conventional laboratories for a range of research problems beyond photocatalysis.
This is a preview of subscription content, access via your institution
Open Access articles citing this article.
Nature Communications Open Access 03 November 2023
npj Computational Materials Open Access 10 August 2023
Data-driven design of new chiral carboxylic acid for construction of indoles with C-central and C–N axial chirality via cobalt catalysis
Nature Communications Open Access 31 May 2023
Access Nature and 54 other Nature Portfolio journals
Get Nature+, our best-value online-access subscription
$29.99 / 30 days
cancel any time
Subscribe to this journal
Receive 51 print issues and online access
$199.00 per year
only $3.90 per issue
Rent or buy this article
Prices vary by article type
Prices may be subject to local taxes which are calculated during checkout
The implementation of the liquid-dispensing station, photolysis station and the workflow, along with three-dimensional designs for labware developed in the project, are available at https://bitbucket.org/ben_burger/kuka_workflow, the code for the robot at and the Bayesian optimizer is available at https://github.com/Taurnist/kuka_workflow_tantalus and https://github.com/CooperComputationalCaucus/kuka_optimizer. Additional design details can be obtained from the authors upon request.
Kang, K., Meng, Y. S., Bréger, J., Grey, C. P. & Ceder, G. Electrodes with high power and high capacity for rechargeable lithium batteries. Science 311, 977–980 (2006).
Woodley, S. M. & Catlow, R. Crystal structure prediction from first principles. Nat. Mater. 7, 937–946 (2008).
Gómez-Bombarelli, R. et al. Design of efficient molecular organic light-emitting diodes by a high-throughput virtual screening and experimental approach. Nat. Mater. 15, 1120–1127 (2016).
Collins, C. et al. Accelerated discovery of two crystal structure types in a complex inorganic phase field. Nature 546, 280–284 (2017).
Davies, D. W. et al. Computer-aided design of metal chalcohalide semiconductors: from chemical composition to crystal structure. Chem. Sci. 9, 1022–1030 (2018).
King, R. D. Rise of the robo scientists. Sci. Am. 304, 72–77 (2011).
Li, J. et al. Synthesis of many different types of organic small molecules using one automated process. Science 347, 1221–1226 (2015).
Dragone, V., Sans, V., Henson, A. B., Granda, J. M. & Cronin, L. An autonomous organic reaction search engine for chemical reactivity. Nat. Commun. 8, 15733 (2017).
Bédard, A.-C. et al. Reconfigurable system for automated optimization of diverse chemical reactions. Science 361, 1220–1225 (2018).
Granda, J. M., Donina, L., Dragone, V., Long, D.-L. & Cronin, L. Controlling an organic synthesis robot with machine learning to search for new reactivity. Nature 559, 377–381 (2018).
Tabor, D. P. et al. Accelerating the discovery of materials for clean energy in the era of smart automation. Nat. Rev. Mater. 3, 5–20 (2018).
Langner, S. et al. Beyond ternary OPV: high-throughput experimentation and self-driving laboratories optimize multi-component systems. Preprint at https://arxiv.org/abs/1909.03511 (2019).
MacLeod, B. P. et al. Self-driving laboratory for accelerated discovery of thin-film materials. Preprint at https://arxiv.org/abs/1906.05398 (2019).
Steiner, S. et al. Organic synthesis in a modular robotic system driven by a chemical programming language. Science 363, eaav2211 (2019).
Wang, Z., Li, C. & Domen, K. Recent developments in heterogeneous photocatalysts for solar-driven overall water splitting. Chem. Soc. Rev. 48, 2109–2125 (2019).
Shahriari, B., Swersky, K., Wang, Z., Adams, R. P. & Freitas, N. D. Taking the human out of the loop: a review of Bayesian optimization. Proc. IEEE 104, 148–175 (2016).
Häse, F., Roch, L. M., Kreisbeck, C. & Aspuru-Guzik, A. Phoenics: a Bayesian optimizer for chemistry. ACS Cent. Sci. 4, 1134–1145 (2018).
Roch, L. M. et al. ChemOS: orchestrating autonomous experimentation. Sci. Robot. 3, eaat5559 (2018).
Chen, C.-L., Chen, T.-R., Chiu, S.-H. & Urban, P. L. Dual robotic arm “production line” mass spectrometry assay guided by multiple Arduino-type microcontrollers. Sens. Actuat. B 239, 608–616 (2017).
Fleischer, H. et al. Analytical measurements and efficient process generation using a dual-arm robot equipped with electronic pipettes. Energies 11, 2567 (2018).
Liu, H., Stoll, N., Junginger, S. & Thurow, K. Mobile robot for life science automation. Int. J. Adv. Robot. Syst. 10, 288 (2013).
Liu, H., Stoll, N., Junginger, S. & Thurow, K. A fast approach to arm blind grasping and placing for mobile robot transportation in laboratories. Int. J. Adv. Robot. Syst. 11, 43 (2014).
Abdulla, A. A., Liu, H., Stoll, N. & Thurow, K. A new robust method for mobile robot multifloor navigation in distributed life science laboratories. J. Contrib. Sci. Eng. 2016, 3589395 (2016).
Dömel, A. et al. Toward fully autonomous mobile manipulation for industrial environments. Int. J. Adv. Robot. Syst. 14, https://doi.org/10.1177/1729881417718588 (2017).
Schweidtmann, A. M. et al. Machine learning meets continuous flow chemistry: automated optimization towards the Pareto front of multiple objectives. Chem. Eng. J. 352, 277–282 (2018).
Zhi, L. et al. Robot-accelerated perovskite investigation and discovery (RAPID): 1. Inverse temperature crystallization. Preprint at https://doi.org/10.26434/chemrxiv.10013090.v1 (2019).
Matsuoka, S. et al. Photocatalysis of oligo (p-phenylenes): photoreductive production of hydrogen and ethanol in aqueous triethylamine. J. Phys. Chem. 95, 5802–5808 (1991).
Shu, G., Li, Y., Wang, Z., Jiang, J.-X. & Wang, F. Poly(dibenzothiophene-S,S-dioxide) with visible light-induced hydrogen evolution rate up to 44.2 mmol h−1 g−1 promoted by K2HPO4. Appl. Catal. B 261, 118230 (2020).
Pellegrin, Y. & Odobel, F. Sacrificial electron donor reagents for solar fuel production. C. R. Chim. 20, 283–295 (2017).
Sakimoto, K. K., Zhang, S. J. & Yang, P. Cysteine–cystine photoregeneration for oxygenic photosynthesis of acetic acid from CO2 by a tandem inorganic–biological hybrid system. Nano Lett. 16, 5883–5887 (2016).
Wang, X. et al. Sulfone-containing covalent organic frameworks for photocatalytic hydrogen evolution from water. Nat. Chem. 10, 1180–1189 (2018).
Schwarze, M. et al. Quantification of photocatalytic hydrogen evolution. Phys. Chem. Chem. Phys. 15, 3466–3472 (2013).
Bai, Y. et al. Accelerated discovery of organic polymer photocatalysts for hydrogen evolution from water through the integration of experiment and theory. J. Am. Chem. Soc. 141, 9063–9071 (2019).
Zhang, J. et al. H-bonding effect of oxyanions enhanced photocatalytic degradation of sulfonamides by g-C3N4 in aqueous solution. J. Hazard. Mater. 366, 259–267 (2019).
Hutter, F., Hoos, H. H. & Leyton-Brown, K. Parallel Algorithm Configuration 55–70 (Springer, 2012).
Mynatt, C. R., Doherty, M. E. & Tweney, R. D. Confirmation bias in a simulated research environment: an experimental study of scientific inference. Q. J. Exp. Psychol. 29, 85–95 (1977).
King, R. D. et al. Functional genomic hypothesis generation and experimentation by a robot scientist. Nature 427, 247–252 (2004).
Pulido, A. et al. Functional materials discovery using energy–structure–function maps. Nature 543, 657–664 (2017).
Campbell, J. E., Yang, J. & Day, G. M. Predicted energy–structure–function maps for the evaluation of small molecule organic semiconductors. J. Mater. Chem. C 5, 7574–7584 (2017).
Fuentes-Pacheco, J., Ruiz-Ascencio, J. & Rendón-Mancha, J. M. Visual simultaneous localization and mapping: a survey. Artif. Intell. Rev. 43, 55–81 (2015).
Rasmussen, C. E. & Williams, C. K. I. Gaussian Processes for Machine Learning (MIT Press, 2006).
Matthews, A. G. G., Rowland, M., Hron, J., Turner, R. E. & Ghahramani, Z. Gaussian process behaviour in wide deep neural networks. Preprint at https://arxiv.org/abs/1804.11271 (2018).
Millman, K. J. & Aivazis, M. Python for scientists and engineers. Comput. Sci. Eng. 13, 9–12 (2011).
Sachs, M. et al. Understanding structure-activity relationships in linear polymer photocatalysts for hydrogen evolution. Nat. Commun. 9, 4968 (2018).
We acknowledge financial support from the Leverhulme Trust via the Leverhulme Research Centre for Functional Materials Design, the Engineering and Physical Sciences Research Council (EPSRC) (grant number EP/N004884/1), the Newton Fund (grant number EP/R003580/1), and CSols Ltd. X.W. and Y.B. thank the China Scholarship Council for a PhD studentship. We thank KUKA Robotics for help with gripper design and the initial implementation of the robot.
The authors declare no competing interests.
Peer review information Nature thanks Volker Krueger, Tyler McQuade and Magda Titirici for their contribution to the peer review of this work.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data figures and tables
The key locations in the workflow are labelled. Other than the black location cubes that are fixed to the benches to allow positioning (see also Extended Data Fig. 1), the laboratory is otherwise unmodified.
a, Photograph showing the robot at the solid dispensing / cartridge station. The two cartridge hotels can hold up to 20 different solids; here, four cartridges are located in the hotel on the left. The door of the Quantos dispenser is opened using custom workflow software that interfaces with the command software that is supplied with the instrument before loading the correct solid dispensing cartridge into the instrument (Supplementary Video 3). Since the KUKA Mobile Robot is free-roaming and has an 820 mm reach, it would be simple to extend this modular approach to hundreds or even thousands of different solids given sufficient laboratory space. b, Photograph showing the KUKA Mobile Robot at the photolysis station (see also Supplementary Videos 3, 6). c, Photograph showing the KUKA Mobile Robot at the combined liquid handling/capping station. The robot can reach both the liquid stations and the Liverpool Inertization Capper-Crimper (LICC) station after six-point positioning, such that liquid addition, headspace inertization and capping can be carried out in a single coordinated process (see Supplementary Videos 3, 5), without any position recalibration. d, Photograph of the KUKA Mobile Robot parked at the headspace gas chromatography (GC) station. The gas chromatography instrument is a standard commercial instrument and was unmodified in this workflow.
Results of a robotic screen for sacrificial hole scavengers using the mobile robot workflow. Of the 30 bioderived molecules trialed, only cysteine was found to compete with the petrochemical amine, triethanolamine. Scavengers are labelled with the concentration of the stock solution that was used (5 ml volume; 5 mg P10). The error bars show the standard deviation.
The gripper is shown grasping various objects. a, The empty gripper; b, gripper holding a capped sample vial (top grasp); c, gripper holding an uncapped sample vial (side grasp); d, gripper holding a solid-dispensing cartridge; and e, gripper holding a full sample rack using an outwards grasp that locks into recesses in the rack. The same gripper was also used to activate the gas chromatography instrument using a physical button press (see Supplementary Video 3; 1 min 52 s).
Average timescales for the various steps in the workflow (sample preparation, photolysis and analysis) for a batch of 16 experiments. These averages were calculated over 46 separate batches. These average times include the time taken for the loading and unloading steps (for example, the photolysis time itself was 60 min; loading and unloading takes an average of 28 min per batch). The slowest step in the workflow is the gas chromatography analysis.
This file contains Supplementary Methods and Supplementary Notes. This file presents the technical specifications of the robot, the experimental stations, workflow benchmarking, the sacrificial hole scavenger screen, control experiments, in silico benchmarking of the search algorithm, experimental robustness tests, and 24/7 monitoring of the workflow.
This file contains the data that was obtained during the autonomous search. This includes the masses and volumes suggested by the optimizer, mass and volumes measured during the autonomous experiment, and the GC measurements (amounts of hydrogen evolved).
This video shows the autonomous system from a bird's eye view running over 48 hours with a speed up factor of 2,880.
This video shows the autonomous system from a bird's eye view running in the dark; speed up factor = 360.
This video shows a close-up of all steps in the workflow at various speeds (20x – 100x).
This video shows a liquid module dispensing 1 mL of water using PID control (double speed).
This video shows the cap crimping process (double speed).
This video shows the vibratory mixing used in the photolysis station (double speed).
About this article
Cite this article
Burger, B., Maffettone, P.M., Gusev, V.V. et al. A mobile robotic chemist. Nature 583, 237–241 (2020). https://doi.org/10.1038/s41586-020-2442-2
This article is cited by
Nature Computational Science (2023)
Nature Reviews Methods Primers (2023)
Nature Synthesis (2023)
Nature Reviews Materials (2023)
Nature Communications (2023)