The MRL is an integral part of I-FIM’s grand goal of building functional intelligent materials (FIM), and it will continuously be developed and upgraded over time. The main task for the lab is the systemic design, synthesis and optimization (via Dynamic-Machine Learning) of FIM building blocks, and their integration into functional stimuli-response systems. It will focus on machine-learning optimized smart synthesis and characterization of inorganic materials, wet synthesis of organic molecules and polymer design, synthesis of biological systems, as well as the integration of components in a dynamic, kinetically or thermodynamically controlled environment to achieve particular functionality. In order to do so, the lab will be equipped with high throughput machinery that will collect data to aid learning, both in a dynamic environment to study evolution, as well as static snapshots of the energy landscape of the smart materials. I-FIM is closely collaborating with computers scientists and AI specialists to hasten the development of this central facility.  

Of the capabilities built to date, the membrane permeation setup and reverse osmosis setup (please refer to ‘Wet labs and Analytical lab’) are some prime examples of such big data acquisition hardware. These 2 systems can execute unlimited number of pre-set processes with varying experimental parameters to characterize the membrane samples, and will potentially be integrated with dynamic ML for real time optimization. A tool for ML optimization of optical properties is also being designed, comprising of a nanoparticle assembler connected to a 3-stage spectrograph via microfluidic channels. The 3-stage spectrograph is a versatile spectroscopy platform allowing simultaneous measurement of Raman, photoluminescence, and fluorescence utilizing user-coupled excitation laser and optical elements. This integrated system will allow automated and precise control over the size and motion of single core-shell droplets generated by the nanoassembler, which will then be subjected to the diverse in-line imaging modes available in the spectrograph. A road map for building more such machinery is being drawn out, currently.