84
Views
0
CrossRef citations to date
0
Altmetric
RESEARCH ARTICLE

Optimization of β-cyclodextrin based ultrasonic-assisted extraction of apigenin and luteolin from celery seed using support vector regression and different nature-inspired optimization algorithms

, , , & ORCID Icon
 

Abstract

This study aimed to extract apigenin and luteolin from celery seeds using 2-hydroxypropyl-β-cyclodextrin (HP-β-CD) based ultrasonic-assisted technique. Optimal extraction conditions were determined through the integration of support vector regression combined with various optimization algorithms, including Genetic Algorithm, Particle Swarm Optimization, Gray Wolf Optimizer, Bat Algorithm, Firefly Algorithm, and Cuckoo Search Algorithm. The identified optimal conditions included an extraction time of 52 min, an HP-β-CD concentration of 24 mg/mL, and a solvent-to-solid ratio of 24 mL/g. Under these conditions, the predicted apigenin and luteolin content in the extract were 0.156 and 0.814 mg/g, respectively, closely aligning with the actual values of 0.158 ± 0.005 mg/g and 0.807 ± 0.008 mg/g. Notably, the HP-β-CD/water extract exhibited significantly higher concentrations of active ingredients and antioxidant activity compared to the water extract, with the content of apigenin and luteolin in the water extract being only 25 and 45%, respectively, compared to the HP-β-CD/water extract. Regarding the antioxidant effect, the water extract displayed a 50% scavenging capacity (SC50) of 331.29 ± 20.5 g/mL, whereas the SC50 for the HP-β-CD/water extract was 89.04 ± 4.70 g/mL. Molecular modeling and IR spectra analysis also suggested the formation of inclusion complexes between HP-β-CD and apigenin/luteolin. These findings suggested that HP-β-CD could be a preferable alternative to conventional solvents due to its advantages, such as easier control, safety, and cost-effectiveness. In addition, the use of support vector regression and nature-inspired optimization algorithms could be promising alternatives to traditional response surface methodology.

Acknowledgment

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

Disclosure statement

No potential conflict of interest was reported by the author(s).

Reprints and Corporate Permissions

Please note: Selecting permissions does not provide access to the full text of the article, please see our help page How do I view content?

To request a reprint or corporate permissions for this article, please click on the relevant link below:

Academic Permissions

Please note: Selecting permissions does not provide access to the full text of the article, please see our help page How do I view content?

Obtain permissions instantly via Rightslink by clicking on the button below:

If you are unable to obtain permissions via Rightslink, please complete and submit this Permissions form. For more information, please visit our Permissions help page.