AUTOMATED INORGANIC MATERIAL SYNTHESIS
Summer 2019
Boston University Grinstaff Group
PROJECT DESCRIPTION
Lithium ion batteries (LiBs) were first invented in the 1970s and have since dominated the technologic market, penetrating all aspects of modern life. Every smartphone, tablet, laptop and electric vehicle on the market contains some version of the original LiB, as the high energy density and low cost are suitable to power all of these technologies[1].
However, as useful as LiBs are, they have several shortcomings. The energy contained in a LiB quickly approaches its theoretical maximum and relies on expensive and rare cobalt, an ingredient that is largely mined in unsavory conditions, chiefly in Congo. Additionally, the liquid electrolyte used in LiBs is highly flammable and is responsible for the well-documented battery fires often seen in media.
New battery technologies have not reached the market yet because researchers are still trying to find exactly the right combination of materials to make up the three essential battery components: anodes, cathodes, and
electrolytes that need to work in perfect harmony[2].
The purpose of this project is to create a fully autonomous inorganic synthesis machine that is capable of running thousands of synthesis iterations, in twice the time it would take manually, to aid in the discovery of a
solid-state electrolyte for lithium ion batteries.
INTRODUCTION
The material synthesis portion of this project will focus on the development of solid oxide electrolyte materials for use in solid state batteries. By replacing the flammable liquid electrolyte in LiBs with a non-flammable solid, a safer and longer lasting battery can be realized[3].
Computational material sciences can predict new materials with superior properties[4], but the time- consuming step is actually synthesizing it. It is a lengthy process to test out one material with one process, observe the results, change a variable, observe the response, and so on.
Using my mechanical engineering background and passion for robotics, my role on this team was to program and integrate the Franka Emika Panda Research robotic arm into the workflow procedure.
By combining modern machine learning techniques and robotics into this chemical synthesis process, we aim to provide a safer and less-intensive method for material synthesis development. My work in the automation of the material synthesis is exceptionally useful to the team as it allows more experiments to by run in conditions that are normally unsafe for human involvement; allowing for those hours to be used instead to design new experiments and analyze results. This project will revolutionize the ongoing
development of a new lithium ion battery.
METHODOLOGY
The automation of the material synthesis process involves four separate machines the Franka Emika Panda Research robotic arm must interact with; which are the liquid handler, furnace, x-ray machine, and a sampler handler.
To complete the full synthesis cycle, the robotic arm must first pick up the rectangular sample holder from the liquid handler, move the sample to the furnace and place it inside, wait for the furnace to complete its heating and cooling processes, pick up the sample from inside the furnace, and bring it to the sample handler. The arm must place the sample holder onto the handler wherein the sample handler processes the samples and returns a circular sample holder that the arm must then place inside the x-ray machine, wait for the x-ray machine to complete its radioactive cycle, pick up the circular sample holder, and deliver it back to the sample handler as its final step.
In order to communicate with the robotic arm, I learned the programming languages Python and Robotic Operating System (ROS) to code each “task” for the robotic arm to complete. Each task required time to program, troubleshoot, loop with another task, and then more troubleshooting. To assist with visualizing the sequence of actions I was programming, I utilized the package Rviz within ROS to create virtual poses of whatever I was programming at the moment. In the results section, I recreate the synthesis process with my poses.
CONCLUSION
Using the programming languages Python and Robotic Operating System (ROS), I was able to successfully program a main sequence of steps that the Franka Emika Panda Research robotic arm is able to replicate and repeat easily. There are five major steps currently in the material synthesis process, and many intermediary poses in-between each primary task not shown. Each complete
cycle takes approximately 10 hours to finish.
While the five main positions and steps in this inorganic material synthesis process are fully programmed and trouble-shot, there is still much to do on this project. For the sake of simplicity and space, many of the in-between steps it takes to complete each step were left out of this presentation; but there are many intermediary tasks that must be programmed and trouble-shot individually.
The scientists from Samsung’s Advanced Institute of Technology and I are currently continuing our work on this project for the foreseeable future. Future steps include fine-tuning each step in this automation process as well as
a continuation of the material synthesis process search.
REFERENCES AND ACKNOWLEDGMENTS
Thank you to my project manager Lincoln Miara for allowing me the opportunity to collaborate on this amazing research project. Thank you to Sam Cross and the rest of the team from Samsung’s Advanced Institute of Technology who patiently guided and supported me while I worked on this project; I learned so much and had a great time working with such a dynamic group. Thank you to Professor Grinstaff and the Grinstaff Group for being so welcoming and helpful. Lastly, a huge thank you to the Undergraduate Research Opportunities Program for funding my work on this project and supplementing my research with interactive workshops throughout the summer. I enjoyed the topics covered and the opportunity to interact with my fellow researchers.
M. M. Thackeray et al., “Electrical energy storage for transportation—approaching the limits of, and going beyond, lithium-ion batteries,” Energy Environ. Sci., vol. 5, no. 7, pp. 7854–7863, 2012.
J. B. Goodenough and Y. Kim, “Challenges for Rechargeable Li Batteries †,” Chem. Mater., vol. 22, no. 3, pp. 587–603, Feb. 2010.
N. J. Dudney and B. J. Neudecker, “Solid state thin-film lithium battery systems,” vol. 4, no. 1999, pp. 479–482,
2008.