在本研究當中,吾等藉由結合Pitzer-Debye-Huckel模型和COSMO-SAC模型發展出一具有預測性的活性係數模型。此模型成功地預測了許多電解質水溶液之離子活性係數、滲透係數以及蒸氣壓。 在交互作用力的描述上,首先將電解質溶液中之物質分為四大類:單原子離子、離子團、可形成氫鍵類和非氫鍵類。再分別以Pitzer-Debye-Huckel模型和COSMO-SAC模型針對長距離和短距離之作用力做描述。 本模型之特色在於不需任何的交互作用參數。此外,除了離子半徑之外,亦不需要任何針對不同物質的特性參數。所有的交互作用力以十三個通用參數來描述,如此一來,對其預測能力有很大的提升。 利用此模型,吾等以三大類的電解質系統來檢測其預測能力,分別為單一鹽類之水溶液(100個系統於25℃,32個系統於不同溫度狀態)、多種鹽類之水溶液 (8個系統)以及單一鹽類之醇-水混和溶液(2個系統)。結果呈現出此模型對於完全解離之電解質溶液有最佳的預測能力而針對第一類之100個常溫下的系統,其預測之結果(%ARD=9.76)和一以PC-SAFT為基礎之文獻(%ARD=9.12)有相近的準確性。
In this work, we show that the mean activity coefficient, osmotic coefficient and vapor pressure of aqueous electrolyte solutions can be successfully predicted through combining the Pitzer-Debye-Huckel model for long range interactions and the COSMO-SAC model for short range interactions. The interactions between different types of species, including ions, ionic groups, hydrogen bonding, and non hydrogen bonding, are considered explicitly (with 13 universal parameters). This approach does not require any ion-solvent pair interaction parameters and does not contain any ion specific parameter other than the ion radius. We have examined this method for three types of systems, including a single salt in water (100 electrolyte systems at 298.15K and 32 electrolyte systems at different temperature), mixture salts in water (8 electrolyte systems), and a single salt in solvent mixture containing water and alcohols (2 electrolyte systems). The predicted result from this method (deviation for 100 electrolyte systems at 298.15K, %ARD = 9.76) presents a similar accuracy as a recent work based on SAFT (%ARD = 9.12) and is suitable for completely dissociated electrolyte solutions.