This research presents the application of predictive regression model such as quadratic regression for estimation of CO₂ solubility in deep eutectic solvents namely allyltriphenylphosphonium bromide-triethylene glycol (ATPPB-TEG) into different molar ratios and their aqueous solutions. In doing so, a design of experiment (DOE) was applied based on Taguchi L₁₈ orthogonal array method. Four factors, namely pressure, temperature, molar ratio and water/DES concentration in mixture were selected as input parameters of model. The output parameter of model was the CO₂ solubility in terms of mole fraction of CO₂ (X_CO2). A quadratic regression model was developed after validation and confirmation through several strong approaches. The results disclose that the prediction of developed quadratic regression model is in acceptable agreement with experimental solubility data. The overall R-squared (R²) and absolute relative error (ARE) values of proposed quadratic regression model were 0.9966 and 0.0725, respectively. Moreover, analysis of variance (ANOVA) indicates that pressure is the most significant factor influencing the X_CO2. Finally, the signal to noise (S/N) ratio shows that the highest levels for pressure, concentration of DES in mixture, and molar ratio, and lowest level for temperature are the optimal levels of input parameters to obtain the highest CO₂ solubility in this system. The developed quadratic regression model and correlation are effective and provide quick, reliable and accurate predictions of CO₂ solubility in DESs without carrying out any time consuming, difficult and expensive experimental measurements. To the best of our knowledge, this is the first time a regression model was developed for prediction of CO₂ solubility in DESs and their aqueous solutions.

这项研究提出了预测回归模型（例如二次回归）在不同摩尔比及其水溶液中对CO ₂在深共熔溶剂（烯丙基三苯基溴化-三甘醇（ATPPB-TEG））中的CO ₂溶解度的估算中的应用。为此，基于Taguchi L ₁₈正交阵列方法应用了实验设计（DOE）。选择压力，温度，摩尔比和混合物中水/ DES的浓度这四个因素作为模型的输入参数。模型的输出参数是CO ₂在CO的摩尔分数计的溶解度₂（X _CO2）。经过验证和确认后，通过几种强有力的方法开发了二次回归模型。结果表明，开发的二次回归模型的预测与实验溶解度数据在可接受的范围内。提出的二次回归模型的总体R平方（R ²）和绝对相对误差（ARE）值分别为0.9966和0.0725。此外，方差分析（ANOVA）表明压力是影响X _CO2的最重要因素。最后，信噪比（S / N）表明，压力，混合物中DES的浓度和摩尔比的最高水平以及温度的最低水平是获得最大CO ₂溶解度的输入参数的最佳水平。这个系统。所开发的二次回归模型和相关性是有效的，并且可以快速，可靠和准确地预测DES中的CO ₂溶解度，而无需进行任何耗时，困难和昂贵的实验测量。据我们所知，这是第一次建立回归模型来预测DES及其水溶液中的CO ₂溶解度。