Ultrasound-assisted solvent-free extraction of rosemary absolute oil: A green paradigm for high-efficacy natural antioxidant in frying oil stabilization
Qiuping Chen, Huihui Zhang, Yee Ying Lee, Yong Wang, Ying Li
TL;DR
A new green method extracts rosemary oil with high antioxidant power, which can stabilize frying oil better than synthetic additives.
Contribution
Ultrasound-assisted solvent-free extraction produces rosemary oil with superior antioxidant efficacy and frying stability.
Findings
UOE-RAO had the highest yield and total phenolic content compared to conventional methods.
0.1% UOE-RAO extended frying oil stability from 31 min to 63 h.
UOE-RAO showed DPPH scavenging capacity comparable to synthetic BHT.
Abstract
•Ultrasound-assisted oleo-extraction (UOE) yielded the best rosemary absolute oil (RAO).•UOE-RAO outperformed conventional methods with the highest yield and TPC.•UOE-RAO could enhance rosemary antioxidants via carnosic acid conversion.•UOE-RAO showed comparable DPPH scavenging capacity to BHT in vitro.•0.1% (w/w) UOE-RAO extended the frying stability from 31 min to 63 h. Ultrasound-assisted oleo-extraction (UOE) yielded the best rosemary absolute oil (RAO). UOE-RAO outperformed conventional methods with the highest yield and TPC. UOE-RAO could enhance rosemary antioxidants via carnosic acid conversion. UOE-RAO showed comparable DPPH scavenging capacity to BHT in vitro. 0.1% (w/w) UOE-RAO extended the frying stability from 31 min to 63 h. This study developed a green extraction paradigm for rosemary absolute oil (RAO) as a sustainable natural antioxidant to inhibit lipid oxidation…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsEdible Oils Quality and Analysis · Essential Oils and Antimicrobial Activity · Biodiesel Production and Applications
Introduction
1
Absolute oils, defined as highly concentrated aromatic essences, are traditionally derived via ethanol extraction from concretes, which are semi-solid intermediates initially produced from plant biomass through hydrocarbon solvent extraction [1], [2]. For decades, these plant-derived absolutes have served as pivotal fragrance ingredients in perfumery, cosmetics, and flavored food products. However, conventional production methodologies predominantly rely on magnetic stirring with petroleum-based volatile organic solvents (e.g., n-hexane or petroleum ether). Notably, high solvent consumption and residual solvent traces in the final product pose significant safety concerns for the environment and food-related applications [3], [4]. In the context of food-grade absolutes, including osmanthus, jasmine, and rose, current production largely depends on animal fat enfleurage, where absolutes are extracted using animal-derived lipids and subsequently incorporated as flavor enhancers in traditional confectioneries. To address these limitations, refined soybean oil was strategically employed in this study as a green, food-grade alternative extraction solvent for rosemary absolute oil (RAO). This selection is justified by three pivotal considerations [5], [6], [7]: (1) As one of the most widely utilized frying media globally, soybean oil ensures the practical applicability of the findings; (2) Its high content of polyunsaturated fatty acids (PUFAs) renders it highly susceptible to oxidation, providing a sensitive and rigorous model system for evaluating extracted antioxidant efficacy; (3) It serves as a well-validated model in frying stability studies, offering a benchmark for comparing antioxidant performance. Employing soybean oil dual-functionally as both the extraction medium and the stabilization target establishes a direct nexus between the green extraction process and the final antioxidant application [8]. Nevertheless, this approach is constrained by low extraction efficiency and contamination with residual unsaturated fatty acids from animal fats, which elevate the risk of oxidative rancidity in end products containing high levels of unsaturated oils [[9], [10]]. These inherent limitations restrict the broader utilization of absolutes in food systems, particularly in applications where oxidative stability is critically required.
Lipid oxidation poses a critical threat to the edible oil industry, driving detrimental outcomes including rancidity, nutritional depletion, and the generation of potentially toxic secondary oxidation products. To mitigate this, synthetic antioxidants such as tert-butylhydroquinone (TBHQ) have been extensively utilized in food systems [[9], [11], [12]]. Despite their proven efficacy, accumulating toxicological evidence links chronic exposure to these compounds with adverse health effects, prompting global regulatory restrictions or bans on select synthetic antioxidants [[10], [13]]. This regulatory shift has intensified interest in natural antioxidant alternatives. While plant-derived compounds (e.g., tea polyphenols, resveratrol) have shown antioxidant promise [9], their utility is constrained by limitations in stability, solubility, and performance under high-temperature processing conditions, which are key requirements for lipid-rich food applications. Against this backdrop, rosemary extract has emerged as a particularly potent natural antioxidant, offering significant commercial potential for extending the shelf life and enhancing the stability of lipid-containing foods.
Rosemary (Rosmarinus officinalis L.), a well-characterized medicinal and aromatic plant, is enriched with terpenoids, phenolic diterpenes, and other bioactive metabolites, including carnosic acid (CA), carnosol, and rosmarinic acid (RA). These compounds act as potent free radical scavengers and singlet oxygen quenchers, thereby effectively mitigating lipid oxidation [[14], [15], [16]]. Notably, diterpenoid antioxidants such as CA exhibit exceptional thermal stability, retaining full bioactivity even under extremely high-temperature frying conditions. Since commercial rosemary extracts have historically relied on organic solvent extraction, growing regulatory pressure and consumer demand for clean-label ingredients have accelerated the development of green extraction technologies [17]. Among these emerging strategies, ultrasound-assisted extraction (UAE) has emerged as a compelling approach, harnessing acousticcavitation, characterized by localized high-pressure/temperature microenvironments, to disrupt plant cell walls, generate micro-jets, and enhance mass transfer solvent-free and efficiently. The sonochemical mechanisms underlying UAE, including cavitation-induced micro-jets, shear forces, and radical generation, enable rapid extraction while preserving thermolabile bioactives [18]. Such technological advancements now enable the production of solvent-free absolutes while preserving high levels of bioactive metabolites like CA, carnosol, and RA [[19], [20]].
In this study, three distinct extraction methodologies for RAO were systematically evaluated, including conventional heat reflux extraction (Ab-cv) with n-hexane, oleo-extraction (Ab-o) without and with ultrasound (Ab-us) using refined soybean oil as an alternative solvent. The primary objectives were to evaluate the effect of these methods on extraction efficiency, antioxidant composition, and the oxidative stability of soybean oil subjected to high-temperature frying. By integrating comprehensive chemical profiling with functional efficacy assessments, this work seeks to establish a robust scientific rationale for leveraging RAO as an effective, clean-label alternative to synthetic antioxidants in lipid-rich food systems.
Materials and methods
2
Plant material and chemicals
2.1
Dried rosemary (R. officinalis L*.*) was obtained from Huaishuntang Pharmaceutical Co. Ltd., Anhui, China. Refined soybean oil was provided by Yihai Kerry (Guangzhou, China). N-hexane and acetonitrile (>99.5%, HPLC) were purchased from OCEANPAK Scientific Co., Ltd. (Beijing, China). RA and CA (≥97%) were acquired from Aladdin Biochemical Technology Co., Ltd. (Shanghai, China). All other reagents used in this study were of analytical grade.
Heat reflux extraction (HRE)
2.2
Rosemary powder (20 g) was extracted via HRE using n-hexane (200 mL) at its boiling temperature (69℃) for 2 h. The mixture was subsequently vacuum-filtered to collect the filtrate, which was then concentrated under reduced pressure at 50°C to remove n-hexane. The resultant residue was redissolved in absolute ethanol (40 mL), followed by centrifugation at 8000 rpm for 15 min. The supernatant was collected and concentrated under reduced pressure via a rotary evaporator (EYELA N-1300, Tokyo Rika Co., Ltd., Japan). The evaporator was operated at a rotation speed of 100 rpm, with a water bath maintained at 50°C. A cooling circulation system was integrated with the condenser, maintained at a working temperature 40°C lower than that of the water bath to ensure efficient condensation and maximize solvent recovery [[21], [22]]. This process yielded the conventional RAO (Ab-cv).
Oleo-extraction (OE)
2.3
Rosemary powder (20 g) was mixed with 200 mL of refined soybean oil and stirred magnetically at room temperature for 2 h. The mixture was then vacuum-filtered, and the oil filtrate underwent liquid–liquid extraction with 500 mL of ethanol for 2 h. The separated ethanol layer was concentrated under reduced pressure at 4℃, yielding a partially concentrated extract that was stored at −15°C overnight. After freezing, phase separation was induced, and the ethanol layer was centrifuged at 8000 rpm for 20 min. The supernatant was harvested and further evaporated under vacuum at 45°C to remove residual ethanol. The final product, designated as oil-extracted RAO (Ab-o), was isolated after complete solvent elimination.
Ultrasound-assisted oleo-extraction (UOE)
2.4
Rosemary powder (20 g) was mixed with 200 mL of refined soybean oil and subjected to ultrasonic-assisted extraction (UAE) at ultrasound powers of 320–640 W for 3–23 min, with magnetic stirring at room temperature. The range of ultrasonic power and extraction time was established based on preliminary experiments and literature precedents to comprehensively encompass the potential optimal conditions for extracting phenolic compounds from plant matrices [21]. The aim was to systematically evaluate the individual and synergistic effects of these parameters on the extraction yield and total phenolic content (TPC) of RAO. UAE was conducted using a probe-type ultrasonic processor (sonicator; model 900–92; Biosafer, Nanjing, China) operating at a frequency of 20 kHz and delivering a maximum rated output power of 900 W. Following UAE, the mixture was vacuum-filtrated to remove insoluble solid residues. The filtrate was subjected to liquid–liquid extraction with 500 mL of ethanol for 2 h. Subsequently, the ethanol layer was separated and concentrated under reduced pressure at 4℃. The partially concentrated extract was then stored at −20℃ overnight to induce phase separation. After freezing, the ethanol layer was centrifuged at 8000 rpm for 20 min. The resulting supernatant was collected and further evaporated under vacuum at 45℃ to remove residual ethanol. The final absolute oil, designated as ultrasound-assisted RAO (Ab-us), was isolated after complete removal of residual solvents.
The extraction yield of HRE, OE, and UOE was calculated as follows:
Preparation of stripped soybean oils
2.5
According to the protocol of Liu et al. [23], refined soybean oil was purified by column chromatography to eliminate minor constituents. The chromatographic column was sequentially packed with diatomaceous earth, silicic acid, activated carbon, silica gel, and alumina at a mass ratio of 1:2:2:2:5:5. This purification protocol effectively eliminated pigments, phenolic compounds, tocopherols, and other minor constituents from the oil.
Determination of total phenolic content (TPC)
2.6
TPC was quantified via the Folin-Ciocalteu colorimetric method, with minor modifications based on Batool [24]. Briefly, 0.02 g of absolute oil was dissolved in 10 mL of ethanol to prepare a stock solution (2 mg/mL). The stock solution was then appropriately diluted to generate a working solution suitable for spectrophotometric analysis. For each measurement, 100 μL of the diluted working solution was transferred to a test tube, followed by the addition of 1250 μL of Folin-Ciocalteu reagent. The mixture was allowed to react in the dark for 3 min at ambient temperature, after which 1000 μL of 10% (w/v) sodium carbonate (Na_2_CO_3_) solution was added. The reaction mixture was incubated in the dark at ambient temperature for 1 h, and the absorbance was measured at 765 nm using a UV–Vis spectrophotometer. Gallic acid was employed as the calibration standard to construct a standard curve (concentration range: 10–200 μg/mL). TPC values were expressed as milligrams of gallic acid equivalents per gram of absolute oil (mg GAE/g). The calculation was performed as follows:
where CGAE (mg/mL) is the gallic acid equivalent concentration derived from the standard curve, Vtotal(mL) is the total volume of the reaction mixture, and msample (g) is the mass of absolute oil used for analysis.
Gas chromatography-mass spectrometry (GC–MS) analysis
2.7
The chemical composition of the absolute oils was characterized via gas chromatography-mass spectrometry (GC–MS). Before analysis, each absolute oil sample (25 μL) was diluted with absolute ethanol to a final volume of 1 mL to achieve an appropriate concentration for GC–MS detection. Analyses were performed using an Agilent 6890 gas chromatograph coupled to an Agilent 5975N mass selective detector (MSD) (Agilent Technologies, Palo Alto, CA, USA). Separation was achieved on an HP-5MS capillary column (30 m × 0.25 mm, 0.25 μm film thickness). Samples (2 μL) were injected in splitless mode, with the injector temperature maintained at 280°C. High-purity helium (>99.999%) was used as the carrier gas at a constant flow rate of 1.0 mL/min. The temperature program was as follows: initial hold at 50°C for 1 min, ramped at 6°C/min to 280°C, and final hold at 280°C for 10 min. Mass spectrometric detection was operated in electron ionization (EI) mode (70 eV ionization energy), with a mass scan range of m/z 41–510. The temperature for injector, ion source, and detector was maintained at 280℃, 230°C, and 150°C, respectively.
High-performance liquid chromatography (HPLC) analysis
2.8
The absolute oil samples were prepared for HPLC quantification of RA, ursolic acid (UA), CA, and carnosol [26, 27]. For RA and UA analysis, 0.5 g of absolute oil was dissolved in absolute ethanol and diluted to 5 mL. The solution was filtered through a 0.45 μm membrane filter before injection. For CA and carnosol analysis, 0.5 mL of the aforementioned solution was further diluted with absolute ethanol to 25 mL and filtered through a 0.45 μm membrane filter. Analyses were performed using a Waters e2695 system (Milford, MA, USA) equipped with a diode array detector (DAD). Separation was performed on a Waters Symmetry C18 column (4.6 mm × 250 mm × 5 μm, For the quantification of RA and UA, the mobile phase comprised (A) acetonitrile and (B) 0.1% (v/v) trifluoroacetic acid aqueous solution, delivered via isocratic elution at a ratio of A:B = 32:68 (v/v). For CA and carnosol, the mobile phase consisted of acetonitrile and 0.5% (v/v) phosphoric acid aqueous solution at a ratio of A:B = 68:32 (v/v) [25]. Chromatographic parameters were optimized as follows: flow rate, 1.0 mL/min; column temperature, 35°C; injection volume, 5 μL; total runtime, 20 min. Typical retention times under these conditions were approximately: RA (10.2 min), UA (12.5 min), CA (8.5 min), and carnosol (9.2 min). Detection wavelengths were set individually: 328 nm for RA, and 210 nm for CA, UA, and carnosol.
The purity and content of studied compounds in rosemary absolute oils from different extraction methods were calculated as follows:
DPPH radical scavenging assay
2.9
The 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical scavenging activity of the absolute oils was evaluated based on Baliyan with minor modifications [26]. DPPH (≥97.0% purity) was supplied by Aladdin (Shanghai, China). Each absolute oil sample (2 mL) was dissolved in ethanol to prepare a series of working solutions (0–120 μg/mL). For each assay, the sample solution was mixed with 2 mL of methanolic DPPH solution (0.01 mmol/L) in a quartz cuvette. The mixture was vortexed and incubated in the dark at room temperature for 30 min. Absorbance was measured at 517 nm using a UV–Vis spectrophotometer. A blank control was prepared by mixing ethanol with the DPPH solution with the same volume. Butylated hydroxytoluene (BHT), a synthetic antioxidant, was used as a positive control under identical conditions (0–120 μg/mL). Radical scavenging activity was calculated using the following equation:
(5)where Ai is the absorbance of the reaction mixture (50 μL sample solution + 150 μL DPPH solution), Aj is the absorbance of the sample blank (50 μL sample solution + 150 μL ethanol), and A0 is the absorbance of the negative control (50 μL ethanol + 150 μL DPPH solution).
DSC analysis
2.10
Oxidative stability was evaluated using a Mettler Toledo DSC 1 differential scanning calorimeter (Schwerzenbach, Switzerland). Samples of 9–12 mg were weighed in an open aluminum pan to ensure direct exposure to the airflow during heating. An identical empty open aluminum pan was used as the reference. The sample and reference pans were loaded into the DSC furnace and heated isothermally at 140°C for 180 min under a constant airflow rate of 100 mL/min. Oxidative induction times (OITs) corresponding to the onset of primary and secondary oxidation were determined from the DSC thermograms.
Frying procedure
2.11
The oxidative stability of soybean oil enriched with different rosemary-derived absolute oils (RAOs; 0.1%, w/w) was assessed under simulated domestic frying conditions [27]. A total of 400 g of each RAO-fortified soybean oil sample was continuously heated in a commercial domestic deep-fat electric fryer (CZ2001, Guangzhou, China) set at 180°C. French fries (50 g) were fried in batches every 15 min, with each batch fried for 5 min. The oil temperature was maintained at 180 ± 5°C, which was monitored via an ATC-300 thermocouple inserted into the oil bulk for real-time tracking. Oil samples were collected at 4-h intervals, immediately cooled to ambient temperature, and stored in amber glass vials at 4°C in the dark until analysis.
Total polar materials (TPM) analysis
2.12
TPM content was quantified with minor modifications. The analysis was performed using a commercial deep-frying oil tester (Testo270, Testo SE & Co. KGaA, Titisee-Neustadt, Germany), which measures TPM via dielectric spectroscopy. The frying experiment was terminated when the TPM content reached the regulatory discard threshold of 27%.
Statistic analysis
2.13
All experiments were conducted in triplicate (n = 3), and results are presented as mean ± standard deviation (SD). Statistical analysis was performed using IBM SPSS Statistics 26 (IBM Corp., Armonk, NY, USA). Differences among groups were evaluated by one-way analysis of variance (ANOVA), followed by Duncan’s multiple range test for post hoc pairwise comparisons. Statistical significance was defined as p ≤ 0.05.
Results and discussion
3
Optimization of UOE conditions
3.1
The effects of extraction time and ultrasonic power on the yield of RAO were investigated (Fig. 1). All optimization experiments for these parameters were conducted using a constant solid-to-liquid ratio of 1:10 (w/v; 20 g rosemary powder per 200 mL refined soybean oil) at room temperature (25°C), with ultrasonic power ranging from 320-640 W and extraction time spanning 3–23 min. Regarding extraction time, prolonged durations generally improved extraction efficiency, but extending beyond 13 min did not yield a statistically significant increase in TPC. For ultrasonic power, its efficacy stems from acoustic cavitation, which induces micro-jets, shock waves, and shear forces to disrupt plant cell walls and promote the release of intracellular components [[27], [28], [29], [30]]. However, excessively high power (> 400 W) risks thermal degradation of thermosensitive bioactives via localized heat accumulation from cavitation collapse. Based on a systematic evaluation of power-time interactions within the tested ranges (power: 320–640 W; time: 3–23 min), the optimal UOE conditions were determined as follows: solid-to-liquid ratio 1:10, ultrasonic power 400 W, temperature 25°C, and extraction time 13 min. This combination (400 W, 13 min) was selected because it consistently maximized both RAO yield and TPC across all parameter sets (Fig. 1), indicating a clear optimum within the experimental domain. Under these optimized conditions, the extracted absolute oil (Ab-us) demonstrated a superior polyphenol content compared to other samples, underscoring the efficacy of the UOE parameter optimization.Fig. 1. Effects of ultrasound-assisted oleo-extraction (UOE) time and power on the yield of rosemary absolute oil at 25°C with a solid-to-liquid ratio of 1:10. Extraction time varied from 3 to 23 min at 400 W (Left); Ultrasonic power varied from 320 to 640 W for 13 min (Right). Different letters indicate significant differences between groups (p < 0.05), with lowercase and uppercase letters indicating significant differences for extraction time and ultrasonic power, respectively.
The dynamic extraction profiles of the absolute oils, Ab-cv (from HRE), Ab-o (from OE), and Ab-us (from UOE), are shown in Fig. 2. For HRE and OE, polyphenol content in the resulting absolute oils increased with extraction time from 5 to 90 min. In HRE, the majority of extraction occurred within the first 30 min, with no significant increase observed thereafter (1.15 ± 0.01 mg/mL at 90 min vs. 1.07 ± 0.01 mg/mL at 30 min). At room temperature, OE for 30 min yielded a lower polyphenol content (0.88 ± 0.01 mg/mL) than HRE, but extending the extraction to 90 min increased it to 0.98 ± 0.01 mg/mL, suggesting that target polyphenols were predominantly released within the initial 30 min. The difference between HRE and OE may be attributed to variations in extraction temperature and solvent properties (e.g., polarity and volatility). For UOE at room temperature (400 W), a polyphenol content of 1.08 ± 0.01 mg/mL was achieved within just 3 min, surpassing the levels of HRE and OE at 30 min. The maximum polyphenol content with UOE was observed at 13 min (1.15 ± 0.01 mg/mL), exceeding that of both HRE and OE at their respective optimal times. These results collectively demonstrate that ultrasonication markedly enhances the extraction efficiency of polyphenols from rosemary by accelerating mass transfer and disrupting cellular structures.Fig. 2. Comparison of extraction yield and total phenolic content among rosemary absolute oils from different extraction methods. HRE (heat reflux extraction): n-hexane, 69°C; OE (oleo-extraction) and UOE (ultrasound-assisted oleo-extraction): refined soybean oil, 25°C; UOE: 400 W, 13 min. Different letters indicate significant differences between groups (p < 0.05), with lowercase and uppercase letters indicating significant differences for yield and total phenols, respectively.
UOE is fundamentally driven by acoustic cavitation, which is a sonochemical phenomenon arising from high-power, low-frequency ultrasound wave propagation in a liquid medium. Here, alternating compression-rarefaction cycles induce microbubble formation, growth, and violent implosion. The implosive collapse of these cavitation bubbles generates localized high-pressure/temperature microenvironments, producing micro-jets, shock waves, and intense shear forces that disrupt plant cell walls, increase membrane permeability, and enhance solvent (oil)-mediated intracellular penetration. These effects collectively accelerate the mass transfer of bioactive compounds into the oil-based extraction medium [19]. In contrast to conventional thermal extraction, UOE achieves efficient extraction under milder bulk temperature conditions, which is a critical advantage for preserving thermolabile phenolic antioxidants [20]. This mild-temperature operation, coupled with cavitation-driven mass transfer enhancement, underscores the unique synergy of ultrasound in green, solvent-free extraction processes.
Comparison of extraction yield and TPC among different methods
3.2
Fig. 2 presents a comparative analysis of the extraction yield and TPC among three RAO samples. For the extraction yield, Ab-us exhibited a significantly higher yield (20%) compared to Ab-cv (13%), while Ab-o showed an intermediate yield (16%). Regarding TPC, Ab-us demonstrated the highest value (347.16 ± 2.81 GAE mg/g), significantly exceeding that of Ab-o (213.68 ± 3.78 GAE mg/g) and Ab-cv (330.52 ± 13.71 GAE mg/g). It is noted that the TPC of Ab-o was the lowest among the three methods, though it achieved a considerable extraction yield, highlighting a trade-off between yield and bioactive retention. The superior performance of ultrasound-assisted solvent-free oil extraction (UOE) is mechanistically rooted in ultrasound-induced acoustic cavitation: acoustic waves propagate through the liquid medium, generating localized high-pressure/temperature microenvironments that drive microbubble formation, growth, and violent collapse [31]. This implosion produces micro-jets, shock waves, and intense shear forces to disrupt plant cell walls, while mechanical vibrations augment mass transfer efficiency between the oil (solvent) and matrix [19]. Previous studies confirm that cavitation-induced physical effects, including cell fragmentation, erosion, and sonoporation (enhanced membrane permeability), accelerate extraction kinetics while minimizing thermal/oxidative degradation of sensitive bioactives (e.g., RA and CA) [18]. Recent kinetic and thermodynamic insights further reveal that optimized sonication parameters (frequency, power density, extraction time) enhance mass transfer coefficients and antioxidant yield without compromising compound stability [18]. Collectively, these findings demonstrate that UOE not only elevates extraction efficiency but also preserves bioactive integrity, highlighting its superiority over conventional extraction approaches (Ab-cv, Ab-o).
To precisely quantify the contribution of the extraction medium to the final product’s composition, a blank control experiment was performed using stripped soybean oil subjected to identical UOE conditions. The resulting extract exhibited a TPC of 204.21 ± 1.50 mg GAE/g (Fig. 2), indicating that ultrasonic treatment indeed helps to enhance the enrichment of both rosemary-derived bioactives and beneficial minor constituents inherent to soybean oil in UOE-RAO, as only a minor fraction of TPC (≈143 mg GAE/g) originated from the soybean oil matrix itself. This clarifies that UOE-RAO is a composite product enriched with beneficial compounds from both rosemary and the oleaginous solvent, which may contribute synergistically to its overall antioxidant capacity.
Comparison of chemical composition in RAOs
3.3
The volatile profiles of RAOs obtained by HRE, OE, and UOE were characterized by GC–MS. A total of 26, 28, and 29 compounds were identified in HRE-, OE-, and UOE-derived absolute oils, respectively. Total ion chromatograms (TIC) revealed similar peak distributions in the 8–20 min retention window across all samples (Fig. 3), whereas the blank control (soybean oil) exhibited no peaks in this region, confirming the rosemary-derived nature of the identified volatiles (Fig. 3a). The major volatile aroma components, including dolcymene, eucalyptol, limonene, linalool, terpinen-4-ol, camphor, borneol, α-terpineol, and caryophyllene, dominated the profiles (Table 1). All absolute oils were enriched in eucalyptol, camphor, borneol, and α-terpineol, aligning with previous reports on rosemary essential oils [32]. Notably, HRE yielded the highest proportion of eucalyptol (23.32%) and dolcymene (3.05%) but the lowest contents of α-terpineol (3.01%), linalool (0.61%), and terpinen-4-ol (0.48%), indicating that the extraction method significantly modulates volatile composition. Moreover, stereoisomeric differences were pronounced: (+)-limonene (dextrorotatory isomer) predominated in Ab-cv and Ab-us, whereas rac-limonene (racemate) was detected in Ab-o. For caryophyllene, Ab-cv contained both α- and β-isomers, Ab-o exclusively β-caryophyllene, and Ab-us solely α-caryophyllene, suggesting that extraction conditions induce stereoselective enrichment of specific isomers.Fig. 3GC–MS total ion chromatograms of rosemary absolute oils and blank control. (A) Ab-cv from HRE, (B) Ab-o from OE, (C) Ab-us from UOE, (D) Blank control (refined soybean oil, no extraction).Table 1. Volatile and non-volatile chemical compositions of rosemary absolute oils obtained by different extraction methods.CompoundsAb-cvAb-oAb-usContent (%)RT (min)Content (%)RT (min)Content (%)RT (min)Bicyclo[2.2.1]hept-2-ene−−0.068.1550.138.32Dolcymene3.058.340.128.3410.198.52Limonene1.338.450.118.440.178.62Eucalyptol23.328.5212.098.51019.338.69γ-Terpinene0.149.15−−0.079.33Linalool0.6110.100.8910.100.8510.29Camphor6.6411.244.8811.238.8511.44Borneol2.9111.734.5211.732.9311.94Terpinen-4-ol0.4812.000.6212.001.0012.20α-Terpineol3.0112.304.1712.315.5012.52Linalyl Anthranilate0.1813.780.0913.78−−Carvacrol0.0414.440.0614.44−−Caryophyllene4.1017.530.9017.530.1618.44δ-Cadinene0.9619.630.1319.63−19.49Methyl Jasmonate−−0.3322.000.4722.20Ferruginol1.5332.981.1332.98−−Squalene0.7839.21−−0.6339.38Vitamin E0.6843.694.6643.706.4544.26
For non-volatile constituents, distinct peaks in the 30–40 min retention window (absent in the blank control) were observed in Ab-o and Ab-us (Fig. 3b-d), attributed to esters, ketones, and heterocyclic compounds. These findings imply that soybean oil acts not only as a solvent for diverse rosemary constituents but also as a source of intrinsic bioactives. For instance, Ab-us contained linoleic acid ethyl ester and oleic acid ethyl ester at retention times of 30.68 and 30.76 min, respectively, which are likely formed via ethanolysis of soybean oil triglycerides under ultrasonic cavitation (free radical generation) and mild thermal effects.
Vitamin E levels varied significantly: 0.68% (Ab-cv), 4.66% (Ab-o), 6.45% (Ab-us), and 12.07% (blank control), confirming its primary origin from soybean oil. Sterol profiling revealed that Ab-o retained campesterol (1.42%), stigmasterol (2.37%), and β-sitosterol (13.05%), which were lower than the total sterol content (47.07%) in the blank control, indicating incomplete transfer of soybean oil sterols. In contrast, no sterols were detected in Ab-us, likely due to ultrasound-promoted degradation into ketones (evidenced by peaks at 32–35 min; Fig. 3c), driven by mechanical shear and localized heating during sonication.
Squalene content further differentiated the extracts: 0.78% (Ab-cv), 0% (Ab-o), and 0.36% (Ab-us) (Table 1). This variation may be ascribed to solvent polarity and temperature: n-hexane (used in HRE) exhibits high affinity for nonpolar squalene, while UOE enhanced squalene release via cavitation-induced cell wall disruption despite lower operating temperatures. These results collectively demonstrate that UOE not only enhances extraction efficiency but also broadens the spectrum of recoverable compounds (volatile and non-volatile), highlighting its superiority as a safe and efficient method for producing high-quality RAOs.
Quantification of key bioactive antioxidants
3.4
The contents of major antioxidant compounds in RAOs extracted via HRE, OE, and UOE were quantified by HPLC, with results summarized in Fig. 4a. CA, the predominant phenolic diterpenoid, exhibited the highest concentration in Ab-cv (97.4 ± 0.7 mg/g), followed by Ab-us (89.45 ± 1.89 mg/g) and Ab-o (57.01 ± 1.56 mg/g). Carnosol content varied markedly, peaking in Ab-us (16.81 ± 0.07 mg/g) compared to Ab-o (14.86 ± 0.93 mg/g) and Ab-cv (8.53 ± 0.76 mg/g). UA was present at lower levels, with Ab-us containing 7.43 ± 0.02 mg/g, versus 6.66 ± 0.02 mg/g (Ab-o) and 1.32 ± 0.02 mg/g (Ab-cv). These data indicate that UOE significantly enhances the recovery of key diterpenoid antioxidants.Fig. 4. Content and purity of key bioactives in rosemary absolute oils from different extraction methods. (A) Contents of carnosic acid, carnosol, and ursolic acid; (B) Purity of carnosic acid and carnosol. Different lowercase letters indicate significant differences between groups (p < 0.05).
Compared with conventional HRE, the superior preservation of thermosensitive diterpenoids (e.g., CA, carnosol) in UOE-RAO mechanistically attributed to fundamental differences in heat generation regimes and mass transfer dynamics. Under the applied ultrasonic conditions (400 W, 25°C), heat generation is primarily localized and transient, originating from acoustic cavitation rather than sustained bulk heating. While cavitation bubble collapse generates microscopic hot spots (transient local temperatures > 1000 K), the overall system temperature remains mild due to rapid heat dissipation. Experimental monitoring corroborated that ultrasound-induced heating caused only a transient temperature elevation to ∼ 48–50°C, significantly lower than the prolonged boiling temperature (69°C) employed in HRE. Previous studies confirm that CA and carnosol retain structural integrity below 60°C, with significant degradation occurring under prolonged high-temperature (>60°C) or oxidative conditions [[18], [20]]. Thus, the moderate, short-term temperature excursion during UOE is insufficient to compromise the bioactive integrity of these diterpenoids. In contrast, HRE’s extended heating (prolonged exposure to 69°C) promotes oxidative degradation and isomerization of labile diterpenoids. Furthermore, ultrasound-enhanced mass transfer, driven by cavitation-induced micro-jets, shock waves, and shear forces, accelerates the release of bioactives from plant matrices, reducing cumulative extraction duration and minimizing thermal exposure. This synergy of a controlled temperature regime (mild bulk conditions, transient local heating) and rapid cavitation-driven mass transfer underpins the enhanced preservation of thermosensitive diterpenoids in UOE-RAO, highlighting its advantage over conventional thermal extraction methods.
Elevated temperatures may induce thermal degradation of CA, reducing its purity while concomitantly elevating the relative content of its degradation product, carnosol [33]. As Fig. 4b illustrated, the purity of CA and carnosol in the absolute oils revealed that carnosol purity (0.84 ± 0.01–1.7 ± 0.08%) was consistently lower than that of CA (5.70 ± 0.07–9.74 ± 0.17%), with both compounds exhibiting the highest purity in Ab-us (CA: 9.74 ± 0.17%; carnosol: 1.68 ± 0.09%), followed by Ab-o (CA: 8.95 ± 0.20%; carnosol: 1.49 ± 0.10%) and Ab-cv (CA: 5.70 ± 0.08%; carnosol: 0.84 ± 0.02%). This trend suggests that UOE under optimized conditions not only improves extraction yields but also enhances the relative purity of these actives owing to minimal thermal degradation or conversion of CA. The elevated CA content and purity in Ab-us may stem from ultrasound-induced cavitation, which disrupts plant cell walls and enhances intracellular component leaching at a low bulk temperature regime, coupled with improved mass transfer efficiency. Notably, the disproportionate increase in CA content relative to carnosol purity implies potential conversion or degradation pathways: CA may partially transform into derivatives (e.g., epirosmanol, rosmanol, rosmaridiphenol) or undergo minor degradation under UOE conditions, rather than solely degrading into carnosol.
RA, a moderately polar phenolic, was uniquely detected in Ab-us (0.18 ± 0.01 mg/g) but absent in Ab-cv and Ab-o. This observation aligns with two factors: (1) the higher polarity of RA, favoring solubility in the aqueous phase generated during ultrasonic emulsification; and (2) ultrasound-mediated cell wall disruption, which facilitates the partitioning of water-soluble phenolics into the oil phase. Collectively, these results demonstrate that UOE is a superior strategy for extracting a broad spectrum of rosemary antioxidants, including lipophilic (CA, carnosol) and moderately polar (RA) constituents, while maintaining high purity at low temperature. Its ability to enhance both yield and selectivity positions UOE as a promising method for producing nutritionally and functionally enriched RAOs.
In vitro antioxidant activity
3.5
The DPPH radical scavenging capacities of RAOs extracted via HRE (Ab-cv), OE (Ab-o), and UOE (Ab-us) are summarized in Fig. 5. All samples exhibited concentration-dependent scavenging behavior, with percentages increasing progressively from 20 to 120 μg/mL. At the highest test concentration (120 μg/mL), the scavenging rates of the synthetic antioxidant BHT (positive control), Ab-cv, Ab-o, and Ab-us were 89.16 ± 0.36%, 79.92 ± 0.50%, 75.78 ± 0.63%, and 88.81 ± 1.02%, respectively. Notably, while BHT displayed potent activity at low concentrations, the disparity in scavenging capacity between BHT and Ab-us diminished markedly at higher concentrations, with no significant difference observed at 120 μg/mL. The IC_50_ values further validated these trends: BHT (33.43 ± 0.39 μg/mL), Ab-cv (73.37 ± 0.22 μg/mL), Ab-o (81.54 ± 1.59 μg/mL), and Ab-us (71.26 ± 0.81 μg/mL). The enhanced performance of Ab-us, evidenced by both higher scavenging rates and lower IC_50_ correlates strongly with its superior TPC observed in previous analyses, aligning with established structure–activity relationships linking phenolic diterpenoids to antioxidant efficacy [[34], [35], [36]]. Mechanistically, the concentration-dependent scavenging profiles indicate that bioactive constituents in Ab-us (notably CA, carnosol, and ursolic acid) act synergistically to neutralize DPPH• radicals via single electron transfer mechanisms, leveraging their redox-active hydroxyl groups for radical stabilization. The minimal gap between Ab-us and BHT at 120 μg/mL underscores UOE’s ability to preserve or concentrate these electron-donating antioxidants more effectively than conventional methods (HRE/OE), which likely suffer from thermal degradation or incomplete component recovery. Collectively, these results position UOE-derived Ab-us as a viable natural alternative to synthetic antioxidants like BHT, combining high extraction efficiency with retained/potentiated antioxidant functionality. Its performance highlights the potential of ultrasound-assisted processing for producing nutraceutical-grade RAOs with tailored bioactivity.Fig. 5DPPH radical scavenging activities of rosemary absolute oils (Ab-cv, Ab-o, Ab-us) and butylated hydroxytoluene (BHT). Different lowercase and uppercase letters indicate significant differences between and within groups (p < 0.05), respectively.
Frying stability
3.6
TPM, the primary markers of thermal-oxidative degradation in frying oils, are associated with impaired lipid metabolism and chronic health risks (e.g., atherosclerosis, carcinogenesis). Regulatory agencies stipulate a maximum permissible TPM threshold of 27% for edible oils, beyond which disposal is mandated [[37], [38], [39]]. As depicted in Fig. 6a, all oil systems, including control soybean oil (unfortified) and soybean oil supplemented with RAOs (Ab-cv, Ab-o, and Ab-us), exhibited comparable initial TPM levels. Upon progressive heating (180°C, continuous frying simulation), TPM accumulation escalated in all groups, yet the time to reach the 27% discard threshold differed significantly. Specifically, control oil reached the threshold at 28 ± 0h, whereas supplementation with Ab-cv, Ab-o, and Ab-us extended this duration to 53 ± 0h, 50 ± 0h, and 60 ± 0.01 h, respectively. Notably, Ab-us demonstrated the most pronounced protective efficacy, delaying TPM formation by 114% compared to the control, which is also an outcome directly correlated with its superior TPC, underscoring the role of phenolic diterpenoids (e.g., CA, RA) as chain-breaking antioxidants that suppress hydroperoxide decomposition into polar polymers [[40], [41]].Fig. 6. Impact of rosemary absolute oils on thermal oxidation stability of refined soybean oil. (A) Formation of total polar materials (TPMs) in refined soybean oil fortified with Ab-cv, Ab-o, or Ab-us during 180°C frying. (B) Dose-dependent inhibition of TPM accumulation by Ab-us (UOE-derived) at 0.2–2.0% (w/w) addition.
As illustrated in Fig. 6b, a concentration-dependent protective effect was further revealed when Ab-us was added to soybean oil at varying doses (0.2–2.0%, w/w). The 1.0% Ab-us supplementation conferred the longest induction period (63 ± 18 h) to reach the 27% TPM threshold, representing a 125% extension versus the control. lower doses (0.2%: 42 ± 0h) were insufficient to fully inhibit oxidation, while higher doses (2.0%: 58 ± 0.35 h) showed diminishing returns, likely due to solubility limits or pro-oxidant effects at supraoptimal concentrations. Thus, 1.0% Ab-us emerges as the optimal dosage for maximizing frying stability, balancing antioxidant potency with practical application feasibility. These findings highlight the superiority of UAE again in producing Ab-us enriched in thermally stable phenolics, which outperforms conventional extraction methods in mitigating TPM formation. This positions Ab-us as a viable natural alternative antioxidant for enhancing the oxidative shelf life of frying oils.
Oxidative stability evaluated by differential scanning calorimetry (DSC)
3.7
Lipid oxidation proceeds via sequential stages: primary oxidation (initiated by free radical attack, generating hydroperoxides, conjugated dienes, and free fatty acids) and secondary oxidation (involving hydroperoxide decomposition, producing aldehydes, ketones, and other carbonyl compounds). To assess the oxidative stability of soybean oil supplemented with RAOs, DSC was employed to measure oxidation induction times (OITs) as key indicators of thermal-oxidative resistance. As summarized in Table 2, supplementation with RAOs significantly prolonged both primary (OIT_1_) and secondary (OIT_2_) oxidation induction times compared to the control (unfortified soybean oil). Specifically, OIT_1_ increased from 12.75 ± 3.51 min (control) to 61.85 ± 1.36 min (Ab-cv), 15.22 ± 2.27 min (Ab-o), and 14.14 ± 0.20 min (Ab-us). For OIT_2_, extensions were more pronounced: 40.47 ± 6.02 min (control) to 99.13 ± 2.71 min (Ab-cv), 50.34 ± 0.89 min (Ab-o), and 61.41 ± 3.10 min (Ab-us), respectively. These results demonstrate that RAOs, irrespective of extraction method, effectively delay both stages of lipid oxidation. This aligns with previous studies demonstrating that rosemary-derived phenolics (e.g., CA, carnosol) act as chain-breaking antioxidants, scavenging peroxyl radicals and inhibiting hydroperoxide propagation [42].Table 2. Primary and secondary oxidation induction times of soybean oil with and without the supplementation of rosemary absolute oils. Different lowercase letters within a column indicate significant differences between groups.SampleTime of oxidation (min)First Induction timeSecond induction timeControl (Soybean oil)12.75 ± 3.51^a^40.47 ± 6.02^a^Soybean oil + Ab-cv61.85 ± 1.36^a^99.13 ± 2.71^b^Soybean oil + Ab-o15.22 ± 2.27^a^50.34 ± 0.89^c^Soybean oil + Ab-us14.14 ± 0.20^b^61.41 ± 3.10^d^
Extraction method significantly modulated antioxidant performance. Notably, all RAO supplements induced a markedly greater prolongation of the secondary oxidation induction time (OIT_2_) than the primary oxidation induction time (OIT_1_). The extension of induction times relative to the control (ΔOIT = OIT_RAO_ – OIT control) were significantly greater for secondary oxidation (e.g., ΔOIT_2_: Ab-cv = 58.66 min; Ab-us = 20.94 min) than for primary oxidation (e.g., ΔOIT_1_: Ab-cv = 49.10 min; Ab-us = 1.39 min). This observation demonstrates that RAOs exert a disproportionately potent inhibitory effect on hydroperoxide decomposition (secondary oxidation), a critical step driving to rancidity, consistent with their established role as chain-breaking antioxidants (via peroxyl radical scavenging and hydroperoxide suppression). Ab-cv exhibited the longest OITs, attributed to hexane’s high affinity for non-polar antioxidants and efficient disruption of plant matrices. Ab-us showed intermediate OITs but outperformed Ab-o in delaying secondary oxidation. The superior secondary oxidation inhibition by Ab-us may reflect ultrasound-induced preservation of polar antioxidants, which preferentially suppress hydroperoxide decomposition. The slightly lower performance of Ab-o and Ab-us compared to Ab-cv may stem from residual soybean oil triglycerides, which act as pro-oxidants at elevated temperatures (180°C).
The consistently larger ΔOIT_2_ values demonstrate that all RAOs were more effective in delaying secondary oxidation relative to primary oxidation, highlighting their specific capacity to inhibit hydroperoxide decomposition, which is a rate-limiting step in the secondary oxidation cascade. Importantly, Ab-us retained substantial antioxidant activity while avoiding organic solvents, underscoring its potential as a green extraction strategy for natural antioxidants. This positions UOE-derived Ab-us as a sustainable alternative to solvent-based extracts for enhancing the oxidative stability of edible oils, particularly in high-temperature applications such as frying.
Conclusion
4
This study comprehensively assessed the antioxidative efficacy of RAOs extracted via three methods (HRE, OE, and UOE) with a focus on their potential to enhance the frying stability of refined soybean oil. The results demonstrated that RAO incorporation significantly extended the frying lifespan of soybean oil (up to 125% extension vs. control, reaching the 27% TPM discard threshold at 63 h with 1% Ab-us), delayed TPM accumulation during 180°C prolonged heating. Notably, the antioxidant activity of UOE-derived Ab-us was comparable to that of the synthetic antioxidant BHT in the DPPH radical scavenging assay, particularly at higher concentrations. Among the extraction techniques, UOE emerged as the superior strategy, yielding Ab-us with the highest TPC and exceptional antioxidant activity. Mechanistically, UOE leverages ultrasound-induced cavitation to disrupt plant cell walls, enhance mass transfer, and preserve thermally sensitive bioactives without organic solvents, addressing the limitations of conventional methods (HRE: solvent dependency; OE: lower TPC). These findings confirm that UOE-derived Ab-us, enriched in synergistic phenolic diterpenoids (CA, carnosol) and polar antioxidants (RA), serves as a promising natural alternative to synthetic antioxidants for inhibiting lipid oxidation in frying oils. Although this straightforward operational workflow demonstrates promising scale-up feasibility, the efficacy in diverse oil matrices and the techno-economic feasibility of upscaling merit further investigation to validate industrial applicability. Furthermore, since this solvent-free UOE process has been established as a sustainable option for enhancing edible oil thermostability, it aligns with the food industry’s growing consumer- and regulatory-driven demand for clean-label ingredients. This supports the transition toward environmentally benign processing paradigms, positioning UOE as a green alternative to conventional thermal extraction methods.
CRediT authorship contribution statement
Qiuping Chen: Writing – original draft, Visualization, Validation, Software, Methodology, Investigation, Formal analysis, Data curation. Huihui Zhang: Investigation, Formal analysis, Data curation. Yee Ying Lee: Validation, Supervision, Investigation, Data curation. Yong Wang: Resources, Project administration, Funding acquisition. Ying Li: Writing – review & editing, Validation, Supervision, Resources, Project administration, Investigation, Funding acquisition, Conceptualization.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Bleus D.Blockx H.Gesquiere E.Adriaensens P.Samyn P.Marchal W.Vandamme D.High-temperature hydrothermal extraction of phenolic compounds from brewer's spent grain and malt dust biomass using natural deep eutectic solvents Molecules 292024198310.3390/molecules 2909198338731474 PMC 11085089 · doi ↗ · pubmed ↗
- 2Ayyildiz S.S.Pelvan E.Karadeniz B.Optimization of accelerated solvent extraction, ultrasound assisted and supercritical fluid extraction to obtain carnosol, carnosic acid and rosmarinic acid from rosemary Sustain. Chem. Pharm.37202410142210.1016/j.scp.2023.101422 · doi ↗
- 3Dassoff E.S.Li Y.O.Mechanisms and effects of ultrasound-assisted supercritical CO 2 extraction Trends Food Sci. Technol.86201949250110.1016/j.tifs.2019.03.001 · doi ↗
- 4Musarurwa H.Tavengwa N.T.Switchable solvent-based micro-extraction of pesticides in food and environmental samples Talanta.224202112180710.1016/j.talanta.2020.12180733303131 · doi ↗ · pubmed ↗
- 5Zhou W.Peng Y.Wu Z.Zhang W.Cong Y.Study on the frying performance evaluation of refined soybean oil after PLC enzymatic degumming Foods.13202427510.3390/foods 1302027538254576 PMC 10815329 · doi ↗ · pubmed ↗
- 6Zhang Q.Saleh A.S.M.Chen J.Shen Q.Chemical alternations taken place during deep-fat frying based on certain reaction products: a review Chem. Phys. Lipids 165201266268110.1016/j.chemphyslip.2012.07.00222800882 · doi ↗ · pubmed ↗
- 7Manzoor S.Masoodi F.A.Rashid R.Dar M.M.Improving oxidative stability of soyabean oil by apple pomace extract during deep frying of french fries Food Biosci.49202210187410.1016/j.fbio.2022.101874 · doi ↗
- 8Chemat F.Vian M.A.Fabiano-Tixier A.S.Nutrizio M.Jambrak A.R.Munekata P.E.S.Lorenzo J.M.Barbara F.J.Binello A.Cravotto G.A review of sustainable and intensified techniques for extraction of food and natural products Green Chem.2220202325235310.1039/C 9GC 03878 G · doi ↗