Modeling the Physicochemical Properties and Antioxidant Activity of Lemon Pomace-Based Water Kefir Using Response Surface Methodology

Published: 29 September 2026| Version 1 | DOI: 10.17632/k378nz9w7s.1
Contributor:
sefanadia putri

Description

Dataset Overview: This dataset contains empirical measurements from a 15-run split-plot Central Composite Design evaluating lemon pomace-based water kefir. Independent Variables: The experimental matrix includes lemon pomace concentration at 5%, 10%, and 15% (w/v), sucrose at 6%, 8%, and 10% (w/v), and pineapple juice at 0%, 5%, and 10% (v/v). Dependent Variables: Measured responses include pH, supernatant fraction, total soluble solids, absolute viscosity, total titratable acidity, vitamin C content, total phenols, IC50 values, and DPPH radical scavenging inhibition percentages. Software and Analysis: Statistical models and optimization data were processed using Design-Expert software version 13.0 via Type III partial ANOVA.

Files

Steps to reproduce

Substrate Preparation: Prepare Eureka lemon pomace at concentrations of 5%, 10%, and 15% (w/v), food-grade sucrose at 6%, 8%, and 10% (w/v), and commercial pineapple juice at 0%, 5%, and 10% (v/v). Inoculation and Split-Plot Design: Activate water kefir grains in a 5% sucrose solution for 24 hours. Assign lemon pomace to the whole-plot stratum and liquid sub-substrates to the sub-plot stratum under a 15-run Central Composite Design. Two-Stage Fermentation: Conduct primary fermentation for 48 hours at 25 °C with 4% (w/v) water kefir grains. Separate grains via a 3 mm sieve, add pineapple juice for secondary fermentation, and continue for 24 hours. Analytical Procedures: Measure pH, total soluble solids, sedimentation stability, absolute viscosity, total titratable acidity, vitamin C via iodometric titration, total phenols via Folin-Ciocalteu assay, and antioxidant activity via DPPH free radical scavenging assay. Statistical Modeling: Evaluate experimental data using Design-Expert software with Type III partial ANOVA based on the split-plot error structure.

Institutions

Categories

Characterization of Food

Licence