Unraveling combustion chemistry of dimethyldiethoxysilane. I. A comprehensive pyrolysis investigation with insight into ethanol formation mechanism in combustion of ethoxysilane flame synthesis precursors

Ethoxysilanes are a family of precursors widely used in flame synthesis of silica nanoparticles. The existence of a silicon atom greatly amplifies the complexity of ethoxysilane combustion reactions, especially the detection of silicon-containing products and exploration of the specific reaction pathways, which hinders the unambiguous understanding of the combustion chemistry of ethoxysilanes. As the first part of a serial work on the combustion of dimethyldiethoxysilane (DMDEOS), a representative ethoxysilane precursor, reports a theoretical, experimental, and kinetic modeling investigation on its pyrolysis. The potential energy surface and rate constants show that the four-membered ethylene elimination dominates the decomposition of DMDEOS. Pyrolysis products in the micro-flow reactor pyrolysis of DMDEOS are detected using synchrotron vacuum ultraviolet photoionization mass spectrometry (SVUV-PIMS), including six silicon-containing products and an abundant of hydrocarbon molecules and radicals. Novel insight is provided into the unclear ethanol formation mechanism in previous pyrolysis investigations of ethoxysilanes. Previously proposed one-step mechanisms are found to be less efficient based on theoretical exploration and experimental evidence. A new multi-step mechanism initiated from the ethanol elimination of HOSi(CH₃)₂OC₂H₅ is concluded to be energy-favorable, which is supported by the identification of relevant products in the micro-flow reactor pyrolysis experiment. Based on the product information detected by SVUV-PIMS and the exploration of ethanol formation mechanism, a kinetic model of DMDEOS pyrolysis is constructed and validated against the new data that flow reactor pyrolysis of DMDEOS at 1.05 atm and 849–1113 K using gas chromatography. The model can effectively reproduce the formation of observed products and successfully address the substantial underprediction of ethanol caused by previously proposed one-step mechanisms. Modeling analyses, including rate of production analysis and sensitivity analysis, are performed to provide insight into the key pathways of DMDEOS decomposition and product formation.