Effects of the Chemical Composition on the Antioxidant and Sensory Characteristics and Oxidative Stability of Cold-Pressed Black Cumin Oils

The antioxidant capacity (AC); amounts of tocopherols, sterols, and polycyclic aromatic hydrocarbons; oxidative parameters; fatty acid composition (FAC); and sensory quality of cold-pressed black cumin oils (CPBCOs) available on the Polish market were analyzed and compared. The AC levels of the CPBCO samples were determined using four assays, namely 2,2-diphenyl-1-picrylhydrazyl (DPPH = 226.8–790.1 μmol TE/100 g), 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS = 385.9–1465.0 μmol TE/100 g), cupric-reducing antioxidant capacity (CUPRAC = 975.3–19,823.3 μmol TE/100 g), and Folin–Ciocalteu assays (FC = 168.1–643.7 μmol TE/100 g). The FAC scores were typical for black cumin oil, except for the sample CPBCO4, which had a higher content of α-linolenic acid (C18:3 = 23.33%), pointing to possible oil adulteration. Additionally, the concentrations of total sterols (TSC = 372 mg/100 g) and tocopherols (TTC = 42.3 mg/100 g) in this sample were higher than those for other investigated oils (TSC = 159–222 mg/100 g, TTC = 1.9–10.4 mg/100 g respectively). The oxidative stability levels (IP = 8.21–37.34 h), peroxide values (PV = 21.36–123.77 meq O2/kg), acid values (AV = 6.40–22.02 mg KOH/kg), and the sums of four specific polycyclic aromatic hydrocarbons (∑4PAHs = 4.48–46.68 μg/kg) in the studied samples differed significantly (p < 0.05). A sensory lexicon including 12 attributes was developed and applied for the sensory evaluation of oils using a quantitative descriptive analysis (QDA).


Introduction
Cold-pressed vegetable oils are gaining increasing interest among conscious consumers. Nowadays, with health and nutrition becoming so important, virgin oils are being chosen instead of solvent-extracted and refined oils [1]. Recently, extra virgin olive oil, linseed oil, rapeseed oil, and other oils have been widely used in food, cosmetic, and pharmaceutical applications. It is well known that cold pressing is a mechanical process in which the oil is extracted and separated from the seeds without chemical intervention or the addition of any other substances at a temperature that does not exceed 35 • C. Therefore, cold-pressed oils are recognized as natural, healthy, and abundant in nutritious compounds such as tocopherols, polyphenols, and essential fatty acids [2]. Nigella sativa L. seeds and cold-pressed black cumin oils (CPBCOs) are part of the diet in Asia and Africa (Egypt, Tunisia). Lately, CPBCOs have gained more extensive popularity in Europe due to their beneficial fatty acid composition (FAC) and bioactive compounds, including tocopherols, sterols, phenolic compounds, and essential oils. These phytochemicals have anti-inflammatory activity and play key roles in supporting the immune system and in

Chemicals
All reagents of analytical or HPLC grade were purchased from Merck Sp. z o. o. (Warszawa, Poland). Redistilled water was used for the preparation of the solutions.

Materials
The experimental materials consisted of seven commercial cold-pressed black cumin oils originally packed into amber glass bottles (250-500 mL) purchased in shops on the Polish market. Samples were collected randomly from seven well-known Polish manufacturers who pressed black cumin oils from seeds cultivated in different localities of Poland, Egypt, and India. To maintain the discretion of the manufacturers, the oil samples were designated as CPBCO1-CPBCO7. All oil samples were within their stated shelf lives and stored in a refrigerator until analysis.
The sterols were measured using gas chromatography according to ISO 12228-1:2014 [17]. In brief, the oil samples were saponified using a 1 mol/L methanolic potassium hydroxide solution. The sterol fraction was extracted three times with a hexane/methyl tert-butyl ether (1:1) mixture. The separation and quantification of the silylated sterol fraction were performed using capillary gas chromatography (Agilent Technologies 6890, Wilmington, DE, USA) using a DB-35MS (J&W Scientific, Folsom, CA, USA) capillary column (25 m, 0.20 mm i.d. and 0.33 µm film thickness) and a flame ionization detector (FID). The detector and injector temperature was 300 • C, while hydrogen was used as a carrier gas with a 1.5 mL/min flow rate. Here, 5α-cholestane was used as an internal standard and individual sterols in oils were determined depending on their relative retention times according to the standards.
The total sum of the tocopherols and sterols and their percentage distribution were calculated.

Determination of Antioxidant Capacity
Before the antioxidant capacity (AC) determination, methanolic extracts of the studied oils were prepared according to the extraction procedure used for phenolic antioxidants from cold-pressed black cumin seed oil described by Lutterodt et al. [9] with minor modifications. Briefly, 2 g of oil was weighed into test tubes and extracted with 5 mL of methanol for 30 min using an orbital shaker (SHKA25081 CE, Labo Plus, Warszawa, Poland). Then, the test tubes were placed in a refrigerator for 30 min, and the methanolic extracts were separated from the oil samples. The extracted procedure was repeated three times. The methanolic extracts were collected and stored in glass bottles. The extracts were kept in a cold dark place until analysis.
DPPH method: The 0.5 mL of DPPH solution in methanol (304.0 µmol/L) was added to 0.1-0.2 mL of methanolic oil extracts diluted by methanol to 1.4-1.3 mL, respectively. The obtained mixtures were shaken vigorously and allowed to stand at room temperature in darkness for 15 min. Then, the absorbance of each solution was measured at 517 nm against a reagent blank (0.5 mL of DPPH methanolic solution mixed with 2 mL of methanol).
ABTS method: To generate the ABTS •+ solution, potassium persulfate (2.45 mmol/L) and ABTS (7 mmol/L) were mixed (1:0.5) and kept for 16 h in dark conditions. After this, the initial absorbance ABTS •+ solution was adjusted to 0.70 at 734 nm through dilution with ethanol. Next, 0.05-0.10 mL of each methanolic oil extract was allowed to react with the ABTS •+ solution (2.45-2.40 mL). After 5 min incubation at 30 • C, the absorbance of the reaction mixture was measured at 734 nm against a reagent blank (2.5 mL of ABTS •+ solution).
CUPRAC method: Firstly, 2 mL of CuCl 2 (0.01 mol/L), 2 mL of neocuproine solution (0.0075 mol/L), and 2 mL of ammonium acetate buffer (1 mol/L, pH = 7.0) were added into a 10 mL volumetric flask. Then, 0.1-0.2 mL of oil methanolic extracts were mixed, and the total volume was brought up to 10 mL with redistilled water. The mixture absorbance was recorded against a blank (all reagents without oil extracts) at 450 nm after 30 min incubation at room temperature.
FC method: In a 10 mL volumetric flask, 1.0 mL of methanolic oil extract and 0.5 mL of FC reagent were combined and mixed until homogeneous at room temperature for 3 min. Next, the reaction was neutralized with 1 mL of Na 2 CO 3 (22.0%) made up to the mark with redistilled water and then allowed to react in darkness for 1 h. The absorbance measurements of the blue-colored solutions were carried out at 765 nm against a reagent blank.
The absorbance of each reaction mixture in a 1 cm glass cell was read with a Hitachi U-2900 spectrophotometer (Hitachi, Tokyo, Japan). The AC levels of methanolic extracts are expressed as micromoles of Trolox equivalents (TE) per 100 g sample.

Determination of Oxidative Stability
The oxidative stability index values of the analyzed oils were determined using a 743 Rancimat instrument made by Metrohm, applying the AOCS Official Method Cd 12b-92 [19]. Here, 3 g of each CPBCO was measured in the Rancimat test instrument at a temperature of 100 • C ± 0.3 • C with a gas flow rate of 20 L/h and 60 mL of water in the measuring vessel.
The peroxide value (PV) of each oil was measured according to ISO 27107 (2010) [20] via potentiometric titration with the use of a 905 Titrando high-end titrator from Metrohm (Warszawa, Poland) after dissolving the oil in the chloroform-glacial acetic acid mixture in the presence of a saturated solution of potassium iodide. The iodide was then oxidized using oil peroxides, and the formed iodine was titrated with sodium thiosulfate to an equivalence point. The result was reported in milliequivalents of oxygen per kilogram of oil (meq O 2 /kg). However, the acid value (AV) was analyzed according to ISO 660 (1996) [21] via the directly acid-base titration of oil in an alcoholic medium against standard potassium hydroxide. The AV is defined as the number of milligrams of potassium hydroxide required to neutralize the free fatty acids (FFA) present in one gram of oil.

Determination of Water and Volatile Matter Content
The water and volatile matter contents in the studied cold-pressed black cumin oils were measured gravimetrically according to ISO 662 (2016) [22]. The amounts of both the moisture and volatile matter were determined in triplicate by drying 10 g of each oil in an oven at the temperature of 103 • C. The drying process was repeated until reaching a constant weight.

Determination of Polycyclic Aromatic Hydrocarbons
Four polycyclic aromatic hydrocarbons (PAHs) limited by EU regulations (benzo(a)pyrene (B(a)P), chrysene (Chry), benzo(a)anthracene (B(a)A), and benzo(b)fluoranthene (B(b)F)) in the analyzed CPBCO samples were determined according to the internal method using high-performance liquid chromatography with a fluorescence detector (HPLC-FLD, Shi-madzu, Kyoto, Japan), Zorbax Eclipse PAH column (particle size 3.5 µm, length 150 mm, diameter 4.60 mm, Agilent Technologies, Santa Clara, CA, USA), and Eclipse XDB-C18 precolumn (3.5 µm, 4.6 × 150 mm, Agilent Technologies) at an oven temperature of 30 • C. The oil samples were dissolved in cyclohexane and extracted to dimethyl formaldehyde. Benzo(b)chrysene was used as an internal standard. The calibration standards were prepared via the dilution of the PAH standards into acetonitrile. These four compounds were quantified using calibration curves plotted for each PAH in the concentration range of 0.25-8.50 µg/kg. The limit of detection (LOD) values (0.08, 0.06, 0.04, and 0.07 µg/kg) and limit of quantification (LOQ) values (0.26, 0.20, 0.14, and 0.23 µg/kg) were calculated for B(a)P, Chry, B(a)A, and B(b)F, respectively. The LOD is the lowest concentration of each PAH in the studied oils that can be detected by the proposed method, while the LOQ is the lowest concentration at which an analyte can be quantitated with a linear response and acceptable precision and accuracy.

Determination of Fatty Acid Composition
The fatty acid profile of each CPBCO was determined according to the official method ISO 5508 [23]. Fatty acid methyl esters prepared using the ISO 5509 standard [24] method were separated on a gas chromatograph (HP 5890 GC) equipped with a flame ionization detector (FID) (Hewlett-Packard, Avondale, PA, USA) and a high polar capillary column BPX 70 (60 m × 0.25 mm, 0.25 µm). The temperatures on the injector and detector were adjusted to 250 • C, while the oven temperature was as follows: heating from 150 to 210 • C at 1.3 • C/min, holding at 210 • C for 5 min. The carrier gas was helium at a flow rate of 0.6 mL/min. The identification of fatty acids was accomplished using external fatty acid methyl ester (FAME) standards, and the results are presented as weight percentages of the total fatty acids.

Sensory Evaluation and Quality Assessment
Two sensory leaders with experience in the sensory field of cold-pressed oils sensory evaluations screened differences in the studied samples and reviewed the previous literature in order to identify a sensory flavor lexicon for the CPBCOs [25,26]. The initial list of attributes was presented during the sensory training session (6 sessions of 45 min) with reference samples and definitions. The selection of attributes was conducted according to ISO 11035 (1994) [27]. As can be seen in Table 1, the final lexicon for CPBCOs consisted of 12 attributes, including 2 basic tastes (sweet, bitter), 6 flavors (overall flavor intensity (OFI), herb-like, flower-like, medical-like, fuel-like, spicy-like), 3 mouth feel sensory terms (astringency, pungency, painty), and the color intensity.
The expert panel consisted of ten (three males and seven females, ages 22-52 years), well-trained individuals (100 h of training) experienced in the sensory evaluation of coldpressed oils. The selection, training, and monitoring of the assessors were conducted according to ISO 8586:1 [28] and ISO 8586:2 [29].
The CPBCO samples were evaluated in triplicate by each panelist in a randomized, balanced black design. The panelists were instructed to assess the intensity of each of the attributes using a 10-cm intensity scale with an anchoring point from 0 (not perceived) to 10 (very high intensity). Additionally, the overall sensory quality (OSQ) was determined using a 5-point scoring scale ranging from 1 (very poor quality) to 5 (very good quality) based on the DGF Standard [26]. The quality level and characteristics used to determine the OSQ values of the oils are presented in Table 2.  Table 2. Overall sensory quality. General characteristics of cold-pressed black cumin oils.

Very good
Extra fresh taste characteristic for cold-pressed black cumin oil, high intensity taste with positive attributes such as herb-like, flower-like, spicy-like, and sweet taste.

Good
Positive tastes such as herb-like, flower-like, and sweet-like flavor at medium intensity. Slightly bitter and astringency is acceptable.

Fair
Negative taste attributes such as fuel-like, medical-like, and bitter taste and mouth feeling attributes such as pungency, astringency, and painty at low intensity were identified.

Bad
Negative taste attributes such as fuel-like, medical-like, and bitter taste and mouth feeling attributes such as pungency, astringency, and painty at moderate intensity were identified. 1 Very bad Negative taste attributes such as fuel-like, medical-like, and bitter taste and mouth feeling attributes such as pungency, astringency, and painty at high/very high intensity were identified.
The industrial laboratories used the 5-point quality scale to determine the OSQ values to check if oils meet the product specification requirements. In this study, we based our assessment on internal industry requirements, where the minimum acceptable scores for fresh cold-pressed and stored oils were 4.0 and 3.5, respectively.
The oil samples were coded with three-digit code numbers and served at room temperature in the amount of 80 mL in a blue jar. The panelists were instructed to clean their palates by rinsing their mouths with a weak warm black tea or eating a piece of apple. All sensory sessions were conducted in sensory laboratory designs according to ISO 8586:1 [28] and ISO 8586:2 [29].

Statistical Analysis
All measurements were conducted in triplicate, while the AC results were taken in five repetitions. All measurements were reported as means ± standard deviation (SD) and were processed using an analysis of variance (ANOVA) and post hoc Duncan test to determine the differences among means. A probability of p > 0.05 was deemed significant. The Pearson correlation and principal component analyses (PCA) were performed using Statistica (Windows software package, version 8.0; StatSoft Inc., Tulsa, OK, USA), while Fizz software (Biosystems, Courtenon, France) was applied for the collection of all sensory data.

Tocopherol and Phytosterol Contents in Cold-Pressed Black Cumin Oils
It is well known that tocopherols exhibit potent antioxidant properties due to their ability to donate phenolic hydrogen. Therefore, these lipophilic antioxidants are crucial for the protection of unsaturated fatty acids against peroxidation. In addition, the tocopherol profile is an important indicator of the authenticity of vegetable oils.
The results of the qualitative and quantitative analyses of tocopherols in CPBCOs are presented in Table 3. Table 3. The compositions and contents of tocopherols and sterols in cold-pressed black cumin oils.

Compound
Content * ± SD (mg/100 g) Note: * n = 3; different letters ( a-f ) within the same row indicate significant differences between amounts of tocopherols and sterols in the studied oils (one-way ANOVA and Duncan test, p < 0.05). Abbreviations: SDstandard deviation; LOD-detection limit; CPBCO-cold-pressed black cumin oil.
In the studied CPBCO samples, the total tocopherol content (TTC) values ranged from 1.9 to 42.3 mg/100 g ( Table 3). The highest TTC (42.3 mg/100 g) for CPBCO4 indicates pressing this oil from an entirely different seed variety or the adulteration of this oil with another vegetable oil.
For comparison, the total concentrations of tocopherols in black cumin oils pressed from the seeds grown in Macedonia, Turkey, and Egypt were 4.5 mg/100 g, 6.5 mg/100 g and 34.0 mg/100 g, respectively [6,30,31].
Concerning the tocopherol composition of the studied oils, six samples of CPBCO2-CPBCO7 (except CPBCO1) contained high percentages of the two homologous α-+ γtocopherols varying from 94% up to 100%. A somewhat lower sum of α-+ γ-tocopherol levels (57-85%) in cold-pressed and extracted black cumin oils from different countries was observed by other authors [6,8,30,31]. However, a similar percentage of α-+ γ-tocopherols (98.7%) for black cumin oil extracted with chloroform-methanol from a local Nigella sativa L. seed variety was reported by Hassanein et al. [32].
It is noteworthy that among all analyzed samples, CPBCO4 had the lowest amount of α-tocopherol (15% of the total tocopherol content) and the highest level of γ-tocopherol (83% of the total tocopherol content). Additionally, the CPBCO1 sample had the lowest level of sum α-+ γ-tocopherols (66% of total tocopherol content), whereas this oil contained the highest amount of β-tocopherol (2.9 mg/100 g). For comparison, CPBCO6 and CPBCO7 had significantly (p < 0.05) lower concentrations of β-tocopherol (0.4 and 0.2 mg/100 g, respectively). Interestingly, similar levels of δ-tocopherol (0.6-0.7 mg/100 g) were found only in two samples, CPBCO1 and CPBCO4 (Duncan test, p > 0.05, Table 3). The significant differences indicated by the Duncan test (p < 0.05) for the tocopherol composition and concentration results for seven commercial CPBCOs can be explained by the influences of genetic, agronomic, environmental (seed varieties, seed growing location, etc.), and technological factors [6,8,[30][31][32].
Moreover, sterols are bioactive compounds that naturally occur in cold-pressed oils. They have a tendency to lower LDL cholesterol levels in the blood, reducing the risk of cardiovascular diseases. Apart from this hypocholesterolemic function, these phytochemicals and their derivatives can also be potent antioxidants [34]. Their antioxidant activities can be attributed to the formation of an allylic free radical and its isomerization to other relatively stable free radicals. In addition, the sterol composition can be considered as an important parameter for determining the adulteration or authenticity of cold-pressed oils due to each seed variety having a specific sterol profile.
It is noteworthy that the sterol compositions of the analyzed CPBCOs from the seven different manufacturers differed significantly (Duncan test, p < 0.05, Table 3). The total concentration of phytosterols (TSC) in CPBCO samples varied from 159 to 372 mg/100 g. As seen in Table 3, β-sitosterol was the most predominant sterol in all oil samples, followed by ∆5-avenasterol, campesterol, and brassicasterol. CPBCO4 and CPBCO2 revealed the highest and lowest β-sitosterol amounts, which were 182 and 81 mg/100 g, respectively. However, insignificant differences (p > 0.05) in this sterol concentrations were observed between the CPBCO3, CPBCO6, and CPBCO7 samples. The determined high β-sitosterol (48.9-56.1%) was in close agreement with those for black cumin oils from different origins (32.3-59.1%), as reported by other authors [8,[30][31][32].
However, the ∆5-avenasterol and campesterol concentrations were similar for all studied oils (except CPBCO4). In addition, brassicasterol was found only in three oil samples (CPBCO1, CPBCO4, and CPBCO6), while a significantly higher brassicasterol content was present in CPBCO4 (Duncan test, p < 0.05, Table 3).
It can be noted that the sterol composition of CPBCO4 was not typical for black cumin oils. The high amounts of β-sitosterol, campesterol, and brassicasterol found in CPBCO4 suggest that it was adulterated with cold-pressed linseed oil.

Antioxidant Capacity of Cold-Pressed Black Cumin Oils
The AC results for the seven commercial CPBCOs are summarized in Table 4. The CPBCO samples had their antioxidative properties confirmed by four in vitro antioxidant assays, namely radical scavenging (ABTS, DPPH) and reducing power (CUPRAC, FC) methods. It is noteworthy that the AC results measured using different analytical methods differed significantly (p < 0.05). The different mechanisms of the used analytical methods-namely single electron transfer (SET), hydrogen atom transfer (HAT), or a combination of both-probably caused these discrepancies between the AC results. It can be noted that the CUPRAC and ABTS values of the CPBCOs were higher than the DPPH and FC results. This can be explained by the fact that CUPRAC and ABTS are reactive with both hydrophilic and hydrophobic antioxidants. However, the DPPH test only allows the determination of antioxidants that can quench the purple DPPH free radicals in the alcoholic solution by providing hydrogen atoms or via electron donation and conversion to the yellow-colored non-radical form (DPPH-H). In contrast, the FC method is performed for the analysis of phenolic antioxidants capable of reducing a mixture of phosphomolybdic/phosphotungstic acid complexes in an alkaline medium, yielding a blue-colored product. Unfortunately, some non-phenolic compounds (reducing sugars and amino acids) with an FC-reductive ability can overestimate the AC values determined by FC assays [35]. The AC variability was visible between the different CPBCO samples (Duncan test, p < 0.05, Table 4). Definitely, the CPBCO2 had the highest AC values determined by all analytical methods (DPPH = 790.1 µmol TE/100 g, ABTS = 1465.0 µmol TE/100 g, CUPRAC = 19,823.3 µmol TE/100 g, FC = 643.7 µmol TE/100 g). The AC (except by CUPRAC value) for CPBCO2 was around four times higher than the AC of the oil with the lowest antioxidant potential (Table 4). Moreover, CPBCO5 had high radical scavenging activity against ABTS and DPPH radicals, whereas the reducing ability of CPBCO1 was found to be high in both CUPRAC and FC methods.
The Duncan test indicated that insignificant differences (p > 0.05) in DPPH and FC results were observed between the CPBCO6 and CPBCO7, while the CPBCO1 and CPBCO7 had similar ABTS values. As can be seen, the mean CUPRAC results significantly differed (p < 0.05) from each other, but the CPBCO4 and CPBCO5 samples revealed the same power to reduce Mo(VI) to Mo(V), with the subsequent formation of a green phosphate/Mo(V) complex.
The variability in AC between different the oils can be explained by genetic (seed varieties), agronomic (plant growing conditions), and technological factors, as well as the storage conditions and time. The agronomic influence on the antioxidant properties of crude oils cold-pressed from Nigella sativa L. seed varieties from Tunisia and Iran was confirmed by Cheikh-Rouhou and co-workers [5]. The polyphenol content in the oil from Tunisian variety was lower (245 mg gallic acid (GA)/kg) than the total phenolic content (TPC) in the oil from Iran (309 mg GA/kg).
The extraction technique affected the amounts of antioxidants in oils from Nigella sativa L. seeds. The TPC was the highest in cold-pressed black cumin oil (36.05 mg GA/kg), followed by Soxhlet-extracted oil (21.44 mg GA/kg), and was the lowest in the oil obtained from black cumin seeds using microwave-assisted extraction (15.19 mg GA/kg) [31]. However, the cold-pressed oil from Nigella seeds was a poorer source of phenolic compounds (TPC = 94.40 mg GA/100 mL), and had lower antiradical activity as determined by the DPPH method (IC 50 = 2.30 mg/mL) and lower reducing ability as analyzed via the FRAP assay (329 mmol/100 mL) than the oil extracted using supercritical fluid extraction (TPC = 160.51 mg GA/100 mL, IC 50 = 1.58 mg/mL, 538.67 mmol/100 mL) [36].
These differences between the AC and TPC results for black cumin oils observed by other authors can be explained by the use of various types of raw material, the country of origin, the cultivation conditions, the oil extraction techniques, and the applied analytical methods based on different reaction mechanisms.

Oxidative Stability
The oxidative stability of CPBCO samples was analyzed using the Rancimat method, and the obtained results were expressed via the induction period (IP). It is known that the IP is associated with the oil shelf life. The more extended the IP, the higher the stability against oil oxidation over time. The IP values for the studied CPBCO samples ranged between 8.31 and 37.34 h (Table 5). Note: * n = 3; different letters ( a-g ) within the same row indicate significant differences between oxidative stability and quality parameters of cold-pressed black cumin oils (one-way ANOVA and Duncan test, p < 0.05). As can be seen, the IP values of the examined oils differed significantly (Duncan test, p < 0.05, Table 5). The differences were probably related to the various cultivars used and quality levels of seeds, the different technological processes, and the country of Nigella sativa L. seed origin, as well as the amounts of natural antioxidants and profiles of fatty acids [5]. The highest significant oxidative stability was found for CPBCO2, and then decreased in the following order: CPBCO2 > CPBCO5 > CPBCO1 > CPBCO4 > CPBCO3 > CPBCO7 > CPBCO6 ( Table 5). As expected, in the same order, the examined oils lost their radical scavenging activity, as measured using the DPPH and ABTS methods (Table 4). It is noteworthy that the low IP values for CPBCO6 (8.31 h), CPBCO7 (9.60 h), and CPBCO3 (10.14 h) indicated the low antioxidant potential of these oils, as analyzed using different analytical methods (Tables 4 and 5).
Similar IP values that ranged between 12.00-50.33 h at 100 • C and 76. .58 h at 80 • C for black cumin oils were found by other authors [5,9,11]. The higher oxidative stability of black cumin oils (13.45-38.22 h) compared to other cold-pressed oils (IP = 3.67-13.98 h for linseed, camelina, evening primrose, hempseed, poppy, and rapeseed oils) can be explained by the different fatty acid compositions and strong antioxidants such as thymoquinone [9,11,38].

Amounts of Primary Oxidation Products and Free Fatty Acids
The CPBCO samples were evaluated for severity in oxidation and hydrolysis using respectively the PV and AV, as shown in Table 5. The PV and AV are regular indices of quality control in oils. The Duncan test indicated that the overall intensity levels of the primary oxidation products (PV = 21.36-123.77 meq O 2 /kg) were significantly different (p < 0.05) between the investigated oils. The amount of hydroperoxides in each CPBCO was higher than permitted by legal requirements (PV = 15 meq O 2 /kg for cold-pressed oils) [39]. Unexpectedly, the formation of hydroperoxides in the tested CPBCOs increased almost linearly (correlation coefficient, r = 0.9064) with the increase in IP values. This can be explained by the fact that lipid oxidation is a multifactorial phenomenon that depends on a number of variables.
For comparison, two cold-pressed black cumin oils and oils extracted from two Nigella seeds having an Iranian and Tunisian origin with similar PV (1.03-1.56 meq O 2 /kg and 4.35-5.65 meq O 2 /kg, respectively) showed different IP (13.45-38.22 h and 12.00-50.33 h, respectively) [5,11]. In contrast, black cumin oils obtained using microwave-assisted extraction and Soxhlet extraction with comparable PV (21. 45 [39]. In our investigation, the CPBCO samples showed significantly higher acidity compared to the AV results (0.23 and 0.35 mg KOH/g) for cold-pressed black cumin oils obtained by Symoniuk et al. [11].
The obtained PV and AV results suggest that none of the assessed CPBCO samples were of acceptable oxidative or hydrolytic status, meaning they could potentially pose a health risk to consumers. In this regard, Gotoh and Wada [40] reported that oxidized fats and oils with PVs at a level of about 100 meq O 2 /kg could be neurotoxic. Furthermore, the presence of these oxidized compounds affected the sensory quality of the analyzed oils.
The estimated oxidation and hydrolytic status of the investigated oils available for retail indicated the use of low-quality raw seeds and/or inappropriate oil extraction and preservation.

Water and Volatile Matter Contents
The water and volatile compound (WVC) levels, some of basic quality parameters for vegetable oils, were in the range of 0.05-0.26% (Table 5). As can be seen, the WVC results (0.05-0.06%) for the five CPBCO samples available for retail did not differ significantly (Duncan test, p > 0.05, Table 5). According to the Codex Alimentarius [39], the volatile matter of cold-pressed oils should not exceed 0.20%. The amounts of water and volatiles in only one sample, CPBCO4, were somewhat higher (0.26%) than the prescribed limit in the Codex Alimentarius [39]. It is well known that the water excess in the oil can influence its stability through triacylglycerol hydrolysis. The higher amount of water in the oil may be the result of excessive seed humidity or technological processes.
For comparison, a lower water content (0.03%) in commercial cold-pressed black cumin oil was found by Mikołajczak et al. [41].

Polycyclic Aromatic Hydrocarbons Content
All investigated samples were analyzed for carcinogenic compounds, namely PAHs, as it is known that black cumin seeds are often dried with a conventional method where the air used for drying is heated by charcoal or gasoline.
Unfortunately, the Σ4PAHs results for the four CPBCO samples ranged between 16.09 and 46.68 µg/kg (Table 5) and exceeded the legal limit of 10 µg/kg [42]. Moreover, the level of B(a)P in the CPBCO2 (3.76 µg/kg) was significantly higher than that recommended by the European Commission [42] for foodstuffs (2 µg/kg). Interestingly, the B(a)P values of all studied oils differed significantly (p < 0.05), while CPBCO3, CPBCO5, and CPBCO7 showed similar B(b)F contents (Duncan test, p > 0.05, Table 5). Badary et al. [43] reported that thymoquinone, which is present in all black cumin seeds and oils, can reduce B(a)Pinduced forestomach carcinogenesis.
It is evident that the content of Σ4PAHs was the highest in CPBCO2 (46.68 µg/kg) due to the presence of the highest amounts of B(a)P (3.76 µg/kg), B(a)A (36.98 µg/kg) and B(b)F (3.14 µg/kg) in this oil. It is noteworthy that the Σ4PAHs value for CPBCO2 was approximately 10 times higher compared to CPBCO3 containing the lowest amounts of these four carcinogenic compounds (Σ4PAHs = 4.48 µg/kg). However, the Duncan test indicated that Chry concentrations in CPBCO2 and CPBCO7 were similar (p > 0.05, Table 5). Insignificant differences (p > 0.05) in results of B(a)A concentration were observed between oil samples CPBCO3, CPBCO6, and CPBCO7 (3.20-4.00 µg/kg), as well as CPBCO1, CPBCO4, and CPBCO5 (14.48-16.47 µg/kg), respectively.
The high levels of PAHs in all investigated oil samples and a lack of professional references for PAH analyses in black cumin oils indicate it is not a well-researched topic and should be explored for human and livestock health.

Fatty Acid Compositions of Cold-Pressed Black Cumin Oils
The fatty acid compositions of the investigated CPBCO samples available on the Polish market are presented in Table 6. It is noteworthy that most studied oils had the typical fatty acid profile for black cumin oil reported by other authors [5,6,8,9,11,[30][31][32]37,38,41,44]. The percentages of fatty acids cannot be referenced to the Codex Alimentarius, the recognized standard for vegetable oils, because black cumin oil and seeds belong to the novel food category according to the European Commission Novel Food Catalogue; thus, these values are not present in this document. As can be seen, all CPBCO samples contained moderately low amounts of saturated fatty acids (SAFA = 12.14-16.76%) and moderate levels of monounsaturated fatty acids (MUFA = 24.49-31.71%), with the predominant component being oleic acid (C18:1, omega-9 = 23.89-31.01%). It is known that MUFA, among all unsaturated fatty acids, are the most resistant to lipid oxidation, and oleic acid plays a major role in the prevention of cardiovascular diseases. However, the polyunsaturated fatty acids (PUFA), ranging between 55.27 and 59.27%, predominantly linoleic acid (C18:2, omega-6 = 32.80-59.08%), are much more prone to lipid oxidation, but on the other hand are supportive of the human cardiovascular and immune systems [44].
All investigated samples contained summed unsaturated fatty acids in the range of 83.11-87.65%, allowing the recognition of black cumin oil as a healthy fat in general.
Surprisingly, the α-linolenic acid content (C18:3, omega-3 = 23.33%) and the C18:1 percentage (31.01%) in the CPBCO4 sample were significantly higher than in all other investigated samples and were not typical for the usual fatty acid profile of black cumin oil (Duncan test, p < 0.05, Table 6). However, the lowest concentrations of C18:2 (32.80%) and C16:0 (8.27%) were determined in this oil. This was probably due to the high-linolenic black cumin seed cultivar used for oil production or the adulteration of this oil with other omega-3 and omega-9 fatty acid-rich oils. Similar amounts of C16:0 (8.70%) and C18:2 (34.20%) were found in cumin seed oil by Ramadan et al. [44]. Note: * n = 3; different letters ( a-e ) within the same row indicate significant differences between the percentages of fatty acids of cold-pressed black cumin oils (one-way ANOVA and Duncan test.

Sensory Analysis of Cold-Pressed Black Cumin Oils
To the best of our knowledge, the sensory characteristics and sensory quality of oils cold-pressed from black cumin seeds have not been reported. In the literature, we have only found overviews on the chemical composition, nutritional properties, or influence of black cumin on the sensory characteristics of mayonnaise [45]. Bendini et al. [13] and Wroniak et al. [14] employed the QDA methodology to describe only the sensory profiles of cold-pressed sunflower oils and cold-pressed rapeseed oils, respectively.
The sensory quality and attribute intensity data of the studied CPBCO samples are listed in Tables 1 and 2.
From the sensory quality scores, the analyzed samples can be divided into the acceptance sensory quality group (OSQ > 4.0 for CPCBO3, CPBCO6, and CPBCO7) and the unacceptance group (OSQ < 3.5 for CPBCO1, CPBCO2, CPBCO4, and CPBCO5). The PCA applied to all samples involved the mean attributes scores presented in Table 7, which were used to interpret and correlate data from 12 attributes and OSQ measured on the seven oils ( Figure 1).
The first two principal components accounted for 88.69% (PC1 = 78.54% and PC2 = 10.15%, respectively) of the variability in the data set. The PCA sensory map ( Figure 1) depicted that three samples (CPBCO3, CPBCO6, and CPBCO7) located on the right side of the PCA map were characterized by good sensory quality scores, as presented by the use of typical flavor descriptors (herb-like, flower-like, spicy-like and sweet-taste) for oils cold-pressed from black cumin seeds. However, some oil samples (CPBCO1, CPBCO2, CPBCO4, and CPBCO5) had poor sensory quality scores (score below 3.5 on a 5-pointing scale) with negative flavor descriptors (fuel-like, medical-like) and were situated on the left side of the PCA map. As shown in Figure 1, the CPBCO2 sample characterized by very high mouth feeling scores, mainly for astringency and pungency (about 9 points on the 10 cm scale), created an evidently distinct cluster. Furthermore, the CPBCO4 sample having a brighter (light-yellow) color and the highest bitterness perception indicated that it might be adulterated with cold-pressed linseed oil. According to Brühl et al. [46], freshly pressed linseed oil provides a delicate nutty flavor, and a lingering bitter off-taste develops upon storage at room temperature. Table 7. Means for the scoring of the overall sensory quality and sensory attributes.

Sensory Attribute
Mean Value * ± SD CPBCO1 CPBCO2 CPBCO3 CPBCO4 CPBCO5 CPBCO6 CPBCO7 Figure 1. Biplot for PC1 and PC2 scores generated by the PCA of 12 attributes and OSQ for coldpressed black cumin oils. Table 7. Means for the scoring of the overall sensory quality and sensory attributes. Nevertheless, high positive correlations were found between the OSQ and flavor attributes such as herb-like (r = 0.9462), flower-like (r = 0.9668), spicy-like (r = 0.9440), and sweet taste (r = 0.9273). However, negative correlations were observed between the OSQ and attributes such as medical-like (r = −0.9464) and fuel-like (r = −0.8642), and mouth feeling attributes such as astringency (r = −0.8724) and painty (r = −0.8880). As can be seen in Table 7, all oil samples presented high scores in terms of the OFI and were positively correlated with the color intensity (r = 0.9587). Generally, in all studied samples, the pungency, astringency, and bitter taste intensities were perceived in medium-high scores. An intense taste with the presence of bitter notes associated with pungent and astringent mouth feeling might be due to bioactive compounds such as polyphenols [47]. Different factors might contribute to off-flavors in cold-pressed oils, such as incorrect storage conditions causing the oxidative degradation of the oil, the presence of impurities in the raw material, or even the use of improper processing technology.

Principal Component Analysis on the Entire Set of Chemical and Sensory Data
The PCA was applied to compare the multidimensional chemical and sensory quality levels of the seven CPBCO samples. The PCA model was applied to all data to determine the most important variables that explain the relationships between the investigated oils and to identify any group patterns. The PCA model retained two principal components (PC1 and PC2), which gave eigenvalues greater than 1.00 (8.63 and 4.50, respectively) and explained 87.52% of the total variability. Therefore, only the first two PCs were used to understand the similarities or dissimilarities of the CPBCO samples available for retail, as illustrated in Figure 2 It is noteworthy that CPBCO3, CPBCO4, CPBCO6, and CPBCO7, with high OSQ and AV scores, as well as low antioxidant properties were located to the right in the score biplot and had positive values for PC1, while the three oils CPBCO1, CPBCO2, and CPBCO5, with the highest oxidative stability (IP), hydroperoxide content (PV), and antioxidant potential (DPPH, ABTS, CUPRAC, and FC values), were situated to the left in the diagram and had negative values for PC1. However, the samples CPBCO3, CPBCO5, CPBCO6, and CPBCO7, with moderate concentrations of total phytosterols (TSC = 201-222 mg/100 g) and tocopherols (TTC = 3.9-7.2 mg/100 g), were located under the A1 axis.
It should be noted that the studied oils fell into three distinct groups ( Figure 2). The CPBCO4 with the longest distance from the other oils revealed the highest contents of tocopherols (TTC), phytosterols (TSC), MUFAs, and water and volatiles (WVC). Three oils (CPBCO3, CPBCO6 and CPBCO 7) with the highest sensory quality (OSQ) and free fatty acid (AV) contents but the lowest amounts of peroxides (PV), ∑4PAHs, and antioxidant potential as determined by different analytical assays (DPPH, ABTS, CUPRAC, and FC) created an evidently distinct cluster. Moreover, three oil samples (CPBCO1, CPBCO2, and CPBCO5) with the highest oxidative stability (IP), antioxidant properties, PV results, and ∑4PAHs, as well as the lowest AV and sensory quality, were separated from the other investigated oils. Additionally, the oils with the highest concentrations of the primary oxidation products were characterized by poor sensory quality.
Regarding the influence of impurities on the sensory profile, Wroniak et al. [14] reported a negative correlation between the level of contaminants in cold-pressed rapeseed oil and the sensory quality with the presence of off-flavors.
The positive and negative correlations between the chemical and sensory characteristics of the seven commercial CPBCO samples are presented as a correlation matrix in Figure 3. It should be noted that the studied oils fell into three distinct groups ( Figure 2). The CPBCO4 with the longest distance from the other oils revealed the highest contents of tocopherols (TTC), phytosterols (TSC), MUFAs, and water and volatiles (WVC). Three oils (CPBCO3, CPBCO6 and CPBCO 7) with the highest sensory quality (OSQ) and free fatty acid (AV) contents but the lowest amounts of peroxides (PV), ∑4PAHs, and antioxidant potential as determined by different analytical assays (DPPH, ABTS, CUPRAC, and FC) created an evidently distinct cluster. Moreover, three oil samples (CPBCO1, CPBCO2, and CPBCO5) with the highest oxidative stability (IP), antioxidant properties, PV results, and ∑4PAHs, as well as the lowest AV and sensory quality, were separated from the other investigated oils. Additionally, the oils with the highest concentrations of the primary oxidation products were characterized by poor sensory quality.
Regarding the influence of impurities on the sensory profile, Wroniak et al. [14] reported a negative correlation between the level of contaminants in cold-pressed rapeseed oil and the sensory quality with the presence of off-flavors.
The positive and negative correlations between the chemical and sensory characteristics of the seven commercial CPBCO samples are presented as a correlation matrix in Figure 3. Unexpectedly, TTC and TSC did not correlate with AC, as determined by four analytical methods (r = −0.4238-0.0060, p = 0.3433-0.9897 for TTC and AC; r = −0.2239-−0.5376, p = 0.2133-0.6293 for TSC and AC). These low negative correlation coefficients suggest that the antioxidant properties of the oils may be due to more hydrophilic compounds such as polyphenols in the prepared extracts or to the synergistic interactions of tocopherols and sterols with other antioxidants present in the CPBCO samples. Therefore, the TTC and TSC in the studied oils are not reliable indicators of their overall AC. The applied DPPH, ABTS, CUPRAC, and FC methods for the determination of the AC were not equivalent by virtue of their varying sensitivities to different antioxidants found in CPBCOs. The effects of the different antioxidants did not appear to be additive but were most likely governed by synergistic and antagonistic interactions among them. Unexpectedly, TTC and TSC did not correlate with AC, as determined by four analytical methods (r = −0.4238-0.0060, p = 0.3433-0.9897 for TTC and AC; r = −0.2239-−0.5376, p = 0.2133-0.6293 for TSC and AC). These low negative correlation coefficients suggest that the antioxidant properties of the oils may be due to more hydrophilic compounds such as polyphenols in the prepared extracts or to the synergistic interactions of tocopherols and sterols with other antioxidants present in the CPBCO samples. Therefore, the TTC and TSC in the studied oils are not reliable indicators of their overall AC. The applied DPPH, ABTS, CUPRAC, and FC methods for the determination of the AC were not equivalent by virtue of their varying sensitivities to different antioxidants found in CPBCOs. The effects of the different antioxidants did not appear to be additive but were most likely governed by synergistic and antagonistic interactions among them.

Conclusions
This study was one of the first attempts to thoroughly investigate the similarities between the sensory attributes, the composition of the CPBCO samples, and their oxidative stability.
The CPBCOs (except CPBCO4) available on the Polish market had similar chemical and nutritional values. The amounts of antioxidants and fatty acid compositions of the studied samples can be considered typical for black cumin oil. The studied oils were rich sources of unsaturated fatty acids and consisted of large amounts of linoleic acid (18:2, omega-6, >50%) and oleic acid (18:1, omega-9, >20%), while a low level of SAFA (about 16%) was found. The CPBCO4 sample was probably adulterated or unintentionally contaminated with cold-pressed linseed oil. Additionally, the sensory attributes of typical black cumin oil (including bitter taste) can be similar to linseed oil's characteristics.
The antioxidant properties determined as the AC, TTC, and TSC and the oxidative stability of the investigated oils differed significantly (p < 0.05) due to differences in the raw material varieties, country of cultivation, or oil age. However, the high antioxidant potential enhanced the oxidative stability of the CPBCOs and can provide potential health benefits to consumers. On the other hand, some analyzed chemical parameters indicating the oil oxidative and hydrolytic status (PV, AV) and cancerogenic contaminant content (∑4PAHs) were above the legal limits recommended by the EU food regulations and the Codex Alimentarius standards.
Moreover, the sensory quality levels of three oil samples with the perception of offflavor attributes were unacceptable for fresh cold-pressed oils (OSQ = 2.0). All of these defects can occur due to poor raw material quality, improper storage of the seeds, or the industrial process. For the first time, a sensory evaluation of CPBCOs was conducted by applying the QDA methodology. The sensory glossary developed in this study can be helpful for the assessment of CPBCO quality.

Conflicts of Interest:
The authors declare no conflict of interest. Two authors are employees of Bunge Polska Sp. z o.o. These studies were carried out as part of the "Implementation Doctorate" Programme for favourable cooperation between university and business.