Abstract It is desirable to predict the surface tension of liquid mixtures for a wide range of compositions, temperatures, and pressures, but current state-of-the-art calculations (e.g., density gradient theory) are computationally expensive. We propose a computationally simple—but accurate—semi-empirical mathematical model of surface tension for a wide variety of multicomponent mixtures, including those with a supercritical compound when coupled with an equation of state (by introducing a reduced mole fraction scaled by a critical composition). Our predictions for binary systems with one supercritical component are an average of 0.22 mN/m away from literature experimental data (466 data points), and those for systems with two subcritical components (293–333 K) are within 0.09 mN/m (236 data points). We make predictions for methanol + ethanol + water using binary coefficients within an average of 0.71 mN/m (196 data points). Given its computational simplicity and wide applicability, the proposed model will be useful for many applications.

中文翻译：

---

表面张力随温度和成分变化而变化的范围广泛的混合物

摘要 预测液体混合物在各种成分、温度和压力下的表面张力是可取的，但当前最先进的计算（例如，密度梯度理论）在计算上是昂贵的。我们提出了一种计算简单但准确的表面张力半经验数学模型，适用于各种多组分混合物，包括那些与状态方程结合时具有超临界化合物的混合物（通过引入按临界组成缩放的降低摩尔分数） ）。我们对具有一个超临界组分的二元系统的预测与文献实验数据（466 个数据点）的平均偏差为 0.22 mN/m，而对具有两个亚临界组分（293–333 K）的系统的预测在 0.09 mN/m（236数据点）。我们使用平均为 0.71 mN/m（196 个数据点）的二元系数对甲醇 + 乙醇 + 水进行预测。鉴于其计算简单性和广泛的适用性，所提出的模型将适用于许多应用。