Optimization and modeling for heat reflux extraction of total yield, stevioside and rebaudioside-A from Stevia rebaudiana (Bertoni) leaves

ABSTRACT

Stevia rebaudiana (Bertoni) leaves consist of stevioside and rebaudioside-A (Reb-A). This research sought to improve extraction of target steviol glycosides from stevia leaf powder using response surface methodology (RSM) and artificial neural networking (ANN) under these independent variables: ethanol concentration, X₁ (0–100%), extraction temperature, X₂ (55–75°C), and extraction time, X₃ (45–75 min). ANN outperformed as potential alternative to RSM in predicting optimum conditions. Maximum responses were obtained at 100% X₁, 55°C X₂, and 60 min X₃. Heat reflux extraction proved superior to maceration extraction in terms of higher extraction yields with reduced energy consumption and CO₂ emission.

KEYWORDS:

• ANN
• optimization
• RSM
• Stevia rebaudiana
• steviol glycosides

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Supplemental data

Supplemental data for this article can be accessed on the publisher’s website.

Additional information

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.