Synthesis of Sulfonated Phenylsilsesquioxanes Guided by Machine Learning.

Area of Science:

• Materials Science
• Organic Chemistry
• Computational Chemistry

Background:

• Sulfonated octaphenylsilsesquioxane (SPOSS) possesses unique properties due to its sulfonic acid (-SO3H) groups, leading to diverse applications.
• Achieving high -SO3H functionality in SPOSS synthesis has been a significant challenge.

Purpose of the Study:

• To develop a novel machine learning-guided approach for synthesizing SPOSS with enhanced -SO3H functionality.
• To explore the synthesis of SPOSS variants with 2, 4, 6, and 8 -SO3H groups, specifically targeting SPOSS-8.

Main Methods:

• Utilized machine learning algorithms to optimize SPOSS synthesis recipes.
• Synthesized SPOSS with varying degrees of sulfonation (SPOSS-2, SPOSS-4, SPOSS-6, SPOSS-8).
• Employed Fourier transform infrared (FTIR) spectroscopy, nuclear magnetic resonance (NMR) spectroscopy, and MALDI-TOF MS for structural characterization.

Main Results:

• Successfully synthesized SPOSS-8 using machine learning predictions based on 21 low-functionality SPOSS synthesis attempts.
• Identified potassium sulfate (K2SO4) as a crucial additive for improving SPOSS functionality.
• Proposed and validated a synthetic mechanism where K2SO4 generates highly reactive sulfur trioxide (SO3).
• Demonstrated superior thermal stability of SPOSS compared to octaphenylsilsesquioxane (OPS) via TGA and TG-FTIR.

Conclusions:

• Machine learning provides an effective strategy for optimizing the synthesis of highly functionalized SPOSS.
• Potassium sulfate plays a key role in the sulfonation process, likely by generating SO3.
• Synthesized SPOSS exhibits enhanced thermal stability, broadening its potential applications.