This research aimed to examine the effect of countercurrent ultrasound on extraction efficiency of cold brew coffee (CBC). The quality and flavor of CBC obtained by ultrasonic-assisted extraction (UE) were evaluated and compared with traditional extraction (TE). Meanwhile, the UE process was monitored in situ real-time using a near-infrared (NIR) spectrometer. The optimal UE parameters obtained by experiments were dual-frequency of 20/28 kHz, countercurrent ultrasonic treatment time of 50 min, ultrasonic power density of 320 W/L, and material-liquid ratio of 1:15. Under the optimal parameters, the total dissolved solids content (TDS) and extraction yield (EY) were 1.63 % and 20.52 %, respectively. This resulted in an 83.33 % improvement in efficiency matching the TDS and EY of a 5-h TE CBC. In comparison with the control, the contents of total polyphenols, total sugars, and total lipids after ultrasonic treatment increased by 13.79 %, 28.49 %, and 333.33 %, respectively, while the total acids content decreased by 27.81 %. HPLC and HS-SPME-GC/MS analysis showed similar non-volatile and volatile flavor components between UE and TE CBC. A quantitative partial least squares (PLS) prediction model was developed to monitor TDS and CGA levels during the UE CBC. This model demonstrated excellent calibration and prediction performance. (TDS: Rc = 0.9281; RMSECV = 0.0694; Rp = 0.9514; RMSEP = 0.0631. CGA: Rc = 0.9086; RMSECV = 0.0316; Rp = 0.9467; RMSEP = 0.0212). This study explored an efficient approach for the industrialization of CBC production and provided theoretical support for developing an intelligent control system for its process.

中文翻译：

---

双频逆流超声波辅助萃取冷萃咖啡，并现场实时监测萃取过程


本研究旨在检验逆流超声对冷萃咖啡 （CBC） 提取效率的影响。评价超声辅助萃取 （UE） 获得的 CBC 的品质和风味，并与传统萃取 （TE） 进行比较。同时，使用近红外 （NIR） 光谱仪对 UE 过程进行原位实时监测。实验得到的最佳UE参数为双频20/28 kHz，逆流超声处理时间50 min，超声功率密度320 W/L，料液比1：15。在最佳参数下，总溶解固体含量 （TDS） 和提取率 （EY） 分别为 1.63 % 和 20.52 %。这导致与 5 小时 TE CBC 的 TDS 和 EY 相匹配的效率提高了 83.33%。与对照相比，超声处理后总多酚、总糖和总脂质含量分别增加了 13.79 %、28.49 % 和 333.33 %，而总酸含量降低了 27.81 %。HPLC 和 HS-SPME-GC/MS 分析显示 UE 和 TE CBC 之间的非挥发性和挥发性风味成分相似。开发了一个定量偏最小二乘法 （PLS） 预测模型来监测 UE CBC 期间的 TDS 和 CGA 水平。该模型表现出出色的校准和预测性能。（TDS：Rc = 0.9281;RMSECV = 0.0694;Rp = 0.9514;RMSEP = 0.0631。CGA：Rc = 0.9086;RMSECV = 0.0316;Rp = 0.9467;RMSEP = 0.0212）。本研究探索了一种高效的 CBC 生产工业化方法，为开发 CBC 过程智能控制系统提供了理论支持。