How accurate are physical property estimation programs for organosilicon compounds?

Area of Science:

• Environmental Chemistry
• Computational Chemistry
• Chemical Engineering

Background:

• Organosilicon compounds are prevalent in commerce, yet crucial physical property data for environmental assessment is scarce.
• The accuracy of predictive software for these properties remains largely unverified.
• The US Environmental Protection Agency (USEPA) processes numerous new chemical notifications containing silicon (Si).

Purpose of the Study:

• To create a comprehensive database of measured physical property values for organosilicon compounds.
• To assess the performance of freely available estimation programs for key environmental properties.
• To identify the most accurate predictive tools for organosilicon compounds.

Main Methods:

• Compiled an extensive database of measured physical properties for organosilicon compounds.
• Evaluated the accuracy of multiple no-cost estimation programs, including SPARC, EPI Suite, and TEST.
• Assessed properties such as melting point (mp), boiling point (bp), vapor pressure (vp), water solubility, n-octanol/water partition coefficient (log Kow), and Henry's law constant.

Main Results:

• SPARC, MPBPWIN, and TEST demonstrated similar accuracy for boiling point and vapor pressure.
• Molinspiration and WSKOWWIN were the top performers for estimating n-octanol/water partition coefficient and water solubility, respectively.
• TEST's melting point estimation method significantly outperformed MPBPWIN.

Conclusions:

• The evaluated estimation programs provide sufficient accuracy for the initial screening of organosilicon compounds.
• The study identifies specific areas where predictive models for organosilicon properties require further development and refinement.
• Accurate property estimation is vital for the environmental assessment of emerging chemical substances.