Ohmic Fermentation Differentiates Volatile Fingerprints of Roasted Arabica and Robusta Coffees from Four Indonesian Origins

This study evaluated whether ohmic fermentation differentiates the volatile fingerprints of roasted Arabica and Robusta coffees from four Indonesian origins. Arabica beans from Enrekang and Gowa/Malakaji and Robusta beans from Bulukumba and Bantaeng were fermented under origin-specific time–temperature conditions, roasted, and analyzed by headspace solid-phase microextraction coupled with gas chromatography–mass spectrometry. Detected compounds were organized into aroma-relevant classes and interpreted using principal component analysis. The analysis detected 477 compounds in Enrekang Arabica, 469 in Gowa/Malakaji Arabica, 316 in Bulukumba Robusta, and 425 in Bantaeng Robusta. Furans, pyrazines, pyridines, ketones, pyrroles, phenols, aldehydes, acids, thiophenes, thiazoles, alcohols, and lactones constituted the dominant interpretable chemical space. Treatment-level PCA explained 71.84% of variance in Enrekang, 73.16% in Gowa/Malakaji, 68.33% in Bulukumba, and 70.53% in Bantaeng, indicating structured treatment responses within each origin. Class-level PCA further separated Arabica and Robusta origins, with Enrekang and Gowa/Malakaji characterized mainly by furan–ketone profiles and Bantaeng Robusta by the strongest pyrazine representation. The results support ohmic fermentation as a promising pre-roasting process for origin-specific volatile modulation. Because the current dataset is based on detected-compound classes and PCA associations, future work should include peak-area normalization, retention-index confirmation, odor-activity analysis, and sensory validation.