Effect of milling degrees on volatile profiles of raw and cooked black rice (Oryza sativa L. cv. Sintoheugmi)

Abstract

Volatile compounds in raw and cooked black rice (cv. Sintoheugmi) samples with different degrees of milling (step 0, 0%; step 1, 4.2%; and step 2, 10.5%, w/w) were investigated by headspace solid-phase microextraction and gas chromatography–mass spectrometry. A total of 101 volatile compounds were found. Among them, 44 compounds found in raw black rice were absent in cooked black rice and 20 compounds were newly formed in cooked black rice. The 8 identified major odor-active volatile compounds in raw and cooked black rice included 3 phenols (guaiacol, 4-vinylphenol, and 2-methoxy-4-vinylphenol), 2 benzenes (benzaldehyde and p-xylene), 2 furans (2-butylfuran and 2-pentylfuran), and 1 terpene (calamenene). Additionally, fatty acid oxidation products such as hexanal, 2-nonenal, octanal, and 2-pentylfuran were found in raw and cooked black rice samples. The relative concentrations of these volatile compounds were significantly higher in step 0 than in step 2 of raw and cooked black rice (p < 0.05). Partially milled cooked black rice (i.e., step 1) contained ~ 80% guaiacol (a favorable unique black rice flavor) of unpolished rice (step 0), with similar levels of several lipid oxidation indicator volatile products (e.g., 2-nonenal and 2-pentyl furan) of fully milled rice (step 2). Thus, partially milled black rice should be consumed rather than fully milled black rice.

Introduction

Black rice has received attention from the food industry and researchers because of its high levels of anthocyanins such as cyanidin-3-O-glucoside and peonidin-3-O-glucoside [1]. Moreover, black rice has a relatively intense flavor that is distinctly different from that of white rice. 2-Acetyl-1-pyrroline, guaiacol, indole, and p-xylene are unique volatile compounds found in cooked black rice but not in cooked white rice [2]. Particularly, guaiacol is responsible for a smoky flavor, making cooked black rice acceptable to customers. Guaiacol and its derivatives have been used as flavor agents in the food and perfume industries. The smoky flavor imparts a roasted flavor to processed foods including bacon. Guaiacol can be also used as a starter compound for vanillin synthesis [3].

Black rice has been used as rice flour and kernel after milling. Rice bran is a by-product of rice milling and contains large amounts of nutritional components such as anthocyanins and phytic acids. The degree of milling is the weight ratio of removed rice bran to unpolished rice. The degree of milling alters rice quality characteristics such as nutritional composition and cooking quality. Polished rice, for which the bran has been completely removed from unpolished rice, shows not only improved taste but also low degeneration caused by lipid-associated rancidity during distribution [4]. However, when rice bran and germ are completely removed, favorable flavor components such as guaiacol are also removed.

The analysis of volatile compounds using headspace solid-phase microextraction fibers in conjunction with GC–MS has proven to be an effective method since its development in the 1990s [5]. Several different phases on fibers are available to choose [6]. The DVB/CAR/PDMS combination fiber has been used to analyze volatiles in rice [7, 8]. This combination has been found to trap a greater range of volatile compounds with different polarities, such as aldehydes, ketones, alcohols, esters, and terpenic hydrocarbons, than other fibers, which is important when analyzing complex mixtures of volatile components, such as rice volatiles [6, 9].

The volatile profiles of raw and cooked black rice milled to different degrees have not been characterized. Thus, the objectives of this study were to (1) identify and quantify the volatile compounds in raw black rice varying in degree of milling and (2) observe changes in volatile composition according to milling degrees by the cooking process by using headspace solid-phase microextraction (HS-SPME) and gas chromatography–mass spectrometry (GC–MS).

Materials and methods

Chemicals and reagents

Authentic standards for GC–MS analysis, C7–C40 saturated hydrocarbon standard, methyl alcohol, hexanal, octanal, nonanal (E)-2-octenal, benzaldehyde, 2-nonenal, acetone, 2-heptanone, 2-octanone, ethanol, 1-pentanol, 1-hexanol, 1-octanol, 1-nonanol, acetic acid, 3-methylbutanoic acid, hexanoic acid, methyl myristate, octanoic acid, nonanoic acid, methyl palmitate, methyl (9Z)-9-octadecenoate, benzoic acid, decane, undecane, dodecane, tridecane, tetradecane, hexadecane, heneicosane, chloroform, and acetonitrile were purchased from Sigma-Aldrich (St. Louis, MO, USA). Internal standards (octanal-d₁₆, 2-methylpyrazine-d₆, and n-hexyl-d₁₃ alcohol) were purchased from C/D/N Isotope, Inc. (Quebec, Canada).

Rice sample preparation

Sintoheugmi was harvested in October 2016 at a local farm in Hoengseong-gun, Gangwon-do, Republic of Korea. Rough rice was dehulled at laboratory in the form of unpolished rice using a hand-operated rice huller. After dehulling, an FS-2000 rice polisher (Mosul, Seoul, Korea) was used to mill unpolished rice to 3 different milling degrees (step 0, 0%; step 1, 4.2%; and step 2, 10.5%, w/w). Step 0 and step 2 were milled to the typical brown rice level and typical white rice level.

Rice was cooked as described previously [10, 11]. The black rice samples (steps 0–2) were cooked using a consistent ratio of 1:1.8 (black rice/water, w/w). Black rice samples were presoaked in water for 1 h prior to cooking in a rice cooker (DWX-200K, Daewoong Co., Ltd., Seoul, Korea). Rice was cooked for 30 min in the cooker, and all cooked rice was allowed to steam-cool for 10 min after the heating was stopped [12]. Cooked samples were cooled for 2–3 min at 25 °C and then placed in a heat-insulating container and used for GC/MS analysis.

Color measurement of raw and cooked black rice

The color of raw and cooked black rice kernels varying in three different milling degrees was measured using an UltraScan PRO colorimeter (Hunterlab, Reston, VA, USA). Color characteristics were indicated as L* (whiteness), a* (redness), and b* (yellowness) values. The colorimeter was calibrated using a standard white tile (L* = 100.03, a* = 0.09, b* = 0.15). One hundred grams of each rice sample was added to a glass dish (40 Ø × 12 mm). The color was analyzed in the middle part of the glass dish. Samples were prepared in triplicate (n = 3). The total color difference (ΔE*) between different milling degree levels of rice samples of step 0 (control) was calculated using the following equation.

$$\Delta E^{*} = \sqrt {\Delta L^{*2} + \Delta a^{*2} + \Delta b^{*2} }$$

Black rice sampling for HS-SPME–GC/MS analysis

To prepare samples for analysis, a 60-g random raw rice sample was removed from each batch and ground for 60 s at low speed using a SFM-353NK food blender (Sinil, Ansung, Korea). After passing through an 18-mesh standard sieve, 2 g (± 1%) of the raw rice powder or cooked rice kernels was transferred into a 22.5 × 75 mm (20 mL) glass headspace vial (Gerstel, Baltimore, MD, USA). Samples were prepared in duplicate (n = 2). Internal standards (octanal-d₁₆, 2-methylpyrazine-d₆, n-hexyl-d₁₃ alcohol, and 2,4,6-trimethylpyridine) were dissolved in methanol and then diluted with nano-pure water. The mixed internal standard solution was added to each headspace vial with a final internal standard concentration of 10 ng/g. The vials were immediately sealed with a magnetic crimp cap (Gerstel). An MPS 2L-XT SPME System (Gerstel) was used for all sequences for automated headspace extraction and analysis.

After a 5-min equilibration time and 18-min headspace extraction time at 80 °C, a 1-cm-long DVB/Car/PDMS fiber (Supelco, St. Louis, MO, USA) was injected into the GC and remained in the GC inlet for 25 s. This fiber is commonly used for HS-SPME analysis for complex mixtures of volatiles [13].

GC–MS analysis

Volatile analysis was conducted using GC–MS on a 7890A coupled to an Agilent 5975 mass selective detector (Agilent Technologies, Santa Clara, CA, USA). Compounds were separated on a DB-Wax fused silica capillary (30 m × 0.25 mm i.d., 0.25 μm film thickness, Agilent Technologies) at 50 °C for 1 min, and then the temperature was increased by 5 °C/min to a final temperature of 260 °C, with a final holding time of 4 min. Injection was performed in splitless mode, and the injector temperature was 260 °C. Helium (99.999%) was used as the carrier gas at a flow rate of 1.0 mL/min. The MSD was fitted with an electron impact ionization source set at 250 °C at 70 eV. Total ion chromatograms were recorded by scanning from m/z 40–350 at a rate of 3.06 scans/s.

Identification and relative quantification of volatile compounds

Volatile compounds were identified by comparing their mass spectra and retention times with those of authentic standard compounds [13, 14]. Volatiles without authentic standard compounds were tentatively identified by comparison of mass spectrum with those reported in the Wiley 9 and NIST 08 with < 80% as a cutoff to match compounds and/or comparison of the Kovats’ retention index (RI). Kovats’ RIs were determined using a polar DB-Wax column and C7–C40 n-alkanes and compared with previously reported RIs at https://www.nist.gov/ or www.pherobase.com.

The entire spectrum was scanned in total ion chromatogram mode. The relative concentration of each volatile compound in black rice was determined using a unique extracted ion peak area at its respective retention time and by comparison with the extracted ion peak area of one of 3 internal standards (i.e., octanal-d₁₆, 2-methylpyrazine-d₆, and n-hexyl-d₁₃ alcohol for aldehydes, nitrogen-containing compounds, and alcohols, respectively). The remaining volatile compounds were quantified by comparison with the internal standard that eluted closest to each of these compounds [15]. Concentration was calculated as described by Baek and Cadwallader [16] and Lee et al. [15].

$${\text{Concentration}}\left( {\frac{\text{ng}}{\text{g}}} \right) = \frac{{{\text{extracted}}\;{\text{ion}}\;{\text{peak}}\;{\text{area}}}}{{{\text{extracted}}\;{\text{ion}}\;{\text{peak}}\;{\text{area}}\;{\text{of}}\;{\text{I}} . {\text{S}} .}}\left[ {{\text{I}}.{\text{S}}.\left( {\frac{{10\;{\text{ng}}}}{\text{g}}} \right)} \right]$$

Statistical analysis

To determine whether differences in volatile compound levels in black rice samples of varying milling degrees were significant, the results were tested by one-way analysis of variance followed by Duncan’s multiple range test (p = 0.05). All statistical analysis was performed using SPSS statistical software version 23 (SPSS, Inc., Chicago, IL, USA).

Results and discussion

Color of raw and cooked black rice samples

Table 1 shows the L*, a*, and b* values, which differed among the three milling degree levels. The L* values of raw black rice kernels significantly increased with increased milling degree levels (p < 0.05), as reported previously [17]. In raw rice samples, the L* values was the highest at 60.34 for step 2 and lowest at 39.68 for step 0 (i.e., unpolished rice) (p < 0.05). The a* value was the highest at 4.53 for step 2 and lowest at 0.66 for step 0 (i.e., unpolished rice) (p < 0.05). The b* value was significantly higher (1.94) in step 2 raw black rice than in other black rice samples (p < 0.05).

Table 1 Hunter color of various milling degrees of black rice (cv. Sintoheugmi)

The changes in a* and b* of cooked black rice samples with different milling degrees were similar to those of raw black rice samples. The a* and b* values of cooked black rice increased as the milling degree increased from step 0 to step 2 (p < 0.05). Compared to the a* and b* values, the L* value changed only minimally as the milling degree increased. There was a slight decrease in the L* value of cooked black rice as the milling degree increased. For example, there was no significant difference in the L* value in step 1 and step 2 of cooked black rice.

Additionally, ΔE*, color differences between unpolished (step 0) and polished rice (steps 1–2) were determined. In raw rice samples, ΔE* was 3.4 (obvious difference to an untrained eye) and 21 (obvious difference) at steps 1 and 2, respectively. In contrast, the cooked rice sample showed a value of 0.38 (normally invisible difference) and 1.07 (only obvious to a trained eye) at steps 1 and 2, respectively. Thus, the color difference was smaller in cooked black rice than in raw rice after cooking as the milling degree increased. This can be explained by the diffusion of black pigments, such as anthocyanin pigments, during cooking.

Volatile compound identification in raw and cooked black rice

A total of 101 volatile compounds were identified using Wiley 9, NIST 08 libraries, and Kovats’ retention index in black rice (Table 2). The identities of 32 volatile compounds were confirmed using authentic standards and denoted as * after peak numbers in Table 2. Representative chromatograms of raw and cooked black rice are shown in Fig. 1. Among the 101 volatile compounds, 80 and 56 compounds were identified in raw and cooked rice, respectively. These include 9 aldehydes, 10 ketones, 8 alcohols, 16 acids and esters, 13 alkanes, 5 olefins, and 19 additional compounds in raw samples (Fig. 1A), and 13 aldehydes, 13 ketones, 2 alcohols, 6 acids and esters, 5 alkanes, 2 olefins, and 15 additional compounds in cooked sample (Fig. 1B). As reported previously in a study using SPME fiber containing PDMS [18], siloxane derivatives affected the chromatogram peaks. These compounds were identified by comparing blank tests with the black rice samples. Other contaminants were also detected and are shown in Table 2. Some contaminants may be absorbed from the storage container (e.g., plastic or cloth bags) and/or pesticides used during cultivation [6]. Additionally, when other black rice samples purchased from local markets grown in different regions were analyzed, contaminants were detected in the samples. Five volatile compounds, not detected in previous studies, were identified as 4,4,7a-trimethyl-5,6,7,7a-tetrahydro-1-benzofuran-2(4H)-one (ripe, apricot, woody odor), 1-methoxy-2-propanol (bitter taste), 1,4,7,10,13,16-hexaoxacyclooctadecane, methyl acetate (ethereal, sweet odor), and calamenene (herb odor), in raw black rice samples.

Table 2 Identified volatiles in raw and cooked black rice varying with milling degrees

Fig. 1

Typical chromatogram of (A) raw and (B) cooked unpolished black rice using SPME/GC–MS showing the peak of volatile compounds and internal standards

Of the volatile compounds identified in the black rice cultivar, 52 and 42 volatile compounds, excluding contaminants, were reported as odor-active compounds in raw and cooked black rice, respectively. The odor description and threshold of all aldehydes in black rice identified in this study were determined previously by GC–olfactometry (Table 2) [19, 20]. Some volatile compounds were unique to the raw and/or cooked black rice. Twenty volatile compounds were newly formed during cooking. Newly formed volatile compounds included aldehydes (2-furaldehyde, 2,6,6-trimethyl-1-cyclohexene-1-carbaldehyde, phenylacetaldehyde, 2-butyl-2-octenal, 4-hydroxy-3-methoxybenzaldehyde), ketones (2,5-octanedione, 2-nonanone, 3-octen-2-one, 6-methyl-3,5-heptadien-2-one, (5E)-6,10-dimethyl-5,9-undecadien-2-one, 2-pentadecanone, 5-pentyl-2(5H)-furanone, and 6,10,14-trimethyl-2-pentadecanone), acids and esters (dibutyl (2Z)-2-butenedioate), alkane (tricyclo[5.2.1.0^2,6]decane), olefins (3-ethyl-2-methyl-1,3-hexadiene), and additional compounds (propylbenzene, isopropenylbenzene, 2-sec-butylphenol, and 2-methoxy-4-vinylphenol). Additionally, 44 volatile compounds in raw black rice samples were absent in cooked black rice samples. This may be because of the low abundance of these volatile compounds or because they were removed during cooking.

Changes in volatile profiles of raw black rice samples by milling

Changes in volatile compounds in raw and cooked black rice (cv. Sintoheugmi) of three milling degree levels are shown in Table 3. Milling commonly refers to the process of removing rice bran and germ from unpolished rice. Thus, polished rice contains high endosperm contents. The milling degree levels of rice are commonly related to lower concentrations of surface lipids of rice [21]. Therefore, odor-active compounds in rice bran and germ, such as lipid-derived compounds, alcohols, and aldehydes, were effectively removed as the milling degree level increased.

Table 3 Concentration of volatile compounds identified in raw and cooked black rice varying with milling degrees

In raw black rice, identified aldehydes included hexanal, octanal, (E)-2-heptenal, nonanal, (E)-2-octenal, decanal, benzaldehyde, 2-nonenal, and pentadecanal. The relative concentrations of these aldehydes generally decreased as the milling degrees increased (p < 0.01). For example, the relative concentration of hexanal (green tomato odor) showed a decrease of 52% after milling from step 0 to step 2. Odor-active aldehyde compounds have a low odor threshold (0.1–13 ppb except benzaldehydes) and octanal, (E)-2-heptenal, nonanal, (E)-2-octenal, decanal, 2-nonenal, and pentadecanal showed higher levels in the by-products than in the rice kernel [2].

The relative concentrations of alcohols (e.g., ethanol, 1-methoxy-2-propanol, 1-pentanol, 1-hexanol, 1-octen-3-ol, 1-octanol, 1-nonanol, 2,2′-[1,2-ethanediylbis(oxy)]diethanol) also significantly decreased as the milling degree increased in raw black rice samples (p < 0.05). Rice is typically contaminated by compounds present in water and/or pesticides used in the cultivation process [22]. The relative concentrations of the contaminants decreased as milling degree increased (p < 0.05). For example, 2,2′-[1,2-ethanediylbis(oxy)]diethanol was not detected in step 2, but was detected in step 0.

The 7 identified major odor-active compounds, including 2 phenols (guaiacol, 4-vinylphenol), 2 benzenes (benzaldehyde and p-xylene), 2 furans (2-butylfuran and 2-pentyl furan), 1 terpene (calamenene), and 1 nitrogen-containing compound (1H-indole), were detected in raw and/or cooked black rice (Fig. 2A, B) [2]. The relative concentration of the major odor-active compounds significantly decreased as the milling degree increased (p < 0.05). Guaiacol, a characteristic odor-active compound in black rice, is a phenolic compound with a low odor threshold (3 ppb) [19, 20]. According to previous studies, guaiacol is reported to impart a smoked odor such as those of roasted coffee, lightly roasted sesame seed, and smoked salmon [23]. As the milling degree increased from step 0 to step 2, the relative concentrations of guaiacol (smoky, black rice-like odor) decreased by 65% (Table 2). However, partially milling the rice such as step 1 retained 92% of the guaiacol in unpolished rice (step 0). As an off-flavor producing volatile, 1H-Indole is a nitrogen-containing compound conferring mothball and fecal odors. The relative concentration of 1H-indole in raw black rice decreased by 74% when milled from step 0 to step 2 (Table 2).

Fig. 2

Change in relative concentrations of volatile compounds in varying milling degrees of black rice: (A, B) aromatic compounds in raw black rice, (C) lipid oxidation products in raw black rice, (D, E) aromatic compounds in cooked black rice, and (F) lipid oxidation products in cooked black rice. Error bars are standard deviations (n = 2)

Additionally, 5 lipid oxidation products (i.e., hexanal, octanal, 2-nonenal, 1-pentanol, and 2-pentylfuran), representative lipid oxidation indicators of stored rice [24], were detected in the black rice samples. Among the lipid oxidation products, hexanal, 2-nonenal, octanal, 1-pentanol, and 2-pentylfuran were chosen based on their high relative concentrations and low odor threshold values. Lipid oxidation products may have been formed by the oxidation of unsaturated fatty acids in black rice. More than 60% of fatty acids are composed of unsaturated fatty acids in rice and unsaturated fatty acids act as precursors of lipid oxidation products [24]. Major unsaturated fatty acids in rice include oleic, linoleic, and linolenic acids [24]. The relative concentrations of hexanal, 1-pentanol, and 2-nonenal, which are indicators of secondary oxidation of linoleic acid [25], decreased significantly with increasing milling degrees of raw black rice samples (p < 0.05) (Fig. 2C). Step 1 black rice contained hexanal concentrations by 60% of unpolished rice (step 0). The relative concentration of octanal, a representative secondary oxidation product of oleic and linoleic acids [24], decreased significantly with increasing milling degree (p < 0.01). The concentration of 2-pentylfuran, a secondary oxidation product of linolenic and linoleic acid, also decreased significantly with increasing milling degrees (p < 0.01). Additionally, the 1-pentanol concentration decreased significantly with increasing milling degree levels (p < 0.001). Thus, milling can also remove lipid oxidation products.

Changes in volatile profiles of cooked black rice samples by milling

In cooked black rice samples, (E)-2-heptanal (pungent odor), an odor-active compound, was not detected, regardless of the milling degree. In contrast, favorable volatile compounds (e.g., 2-furaldehyde (sweet, woody, almond, bread baked odor), 2,6,6-trimethyl-1-cyclohexene-1-carbaldehyde (saffron, herbal odor), phenylacetaldehyde (sweet, floral, nutty odor), 2-butyl-2-octenal (fruity, pineapple), and 4-hydroxy-3-methoxybenzaldehyde (vanilla odor)) were newly detected in cooked black rice.

The relative concentrations of the newly detected volatile compounds decreased significantly as milling degree increased (p < 0.05). For example, vanillin (4-hydroxy-3-methoxybenzaldehyde), a unique odor-active compound in cooked black rice, levels in step 0 (78.4 ng/10 g), decreased significantly in step 1 (37.4 ng/10 g) and step 2 (20.9 ng/10 g) (p < 0.05).

In cooked black rice samples, 7 volatile ketone compounds (2-nonanone (fruity, cheesy odor), 3-octen-2-one (creamy, earthy, oily odor), 6-methyl-3,5-heptadien-2-one (cinnamon, coconut odor), (5E)-6,10-dimethyl-5,9-undecadien-2-one (green, fruity, waxy odor), 2-pentadecanone (jasmine, celery odor), 5-pentyl-2(5H)-furanone (minty odor), and 6,10,14-trimethyl-2-pentadecanone (fresh jasmine, celery odor)), which were absent in raw black rice samples, were newly found after cooking. The concentrations of the ketones decreased significantly in polished black rice (i.e., steps 1–2) compared to that in unpolished black rice (step 0) (p < 0.05). For example, 3-octen-2-one was not detected in polished black rice.

Among the odor-active alcohols found in raw black rice samples, only 1-hexanol and 1-octen-3-ol were detected in cooked black rice samples, while the other odor-active alcohols were not detected in cooked black rice samples. Cooking appeared to greatly influence the alcohol concentration.

The major odor-active compounds of cooked black rice included 3 phenols (guaiacol, 4-vinylphenol, and 2-methoxy-4-vinylphenol), 2 benzenes (benzaldehyde and p-xylene), 2 furans (2-butylfuran and 2-pentyl furan), 1 terpene (calamenene), and 1 nitrogen-containing compound (1H-indole) (Fig. 2D, E) [2]. The relative concentration of odor-active compounds in cooked black rice samples significantly decreased as milling degree increased (p < 0.05). By milling unpolished black rice (step 0) to polished black rice (step 2), the relative concentration of guaiacol (smoky, black rice-like odor) decreased by ~ 50% from 397.4 ng/10 g (step 0) to 185.3 ng/10 g (step 2) (Table 2). However, partially polished rice (step 1) maintained guaiacol levels (311.2 ng/10 g) of 78% in unpolished rice (step 0). As an off-flavor producing volatile, 1H-Indole is responsible for mothball and fecal odors. In step 2, the relative concentration of 1H-indole decreased by 42%, compared to that in step 0.

Among the 5 lipid oxidation products (i.e., hexanal, octanal, 2-nonenal, 1-pentanol, and 2-pentylfuran) found in raw black rice, hexanal, octanal, 2-nonenal, and 2-pentylfuran were found in cooked black rice, and 1-pentanol was not detected in cooked black rice (Fig. 2F). As the milling degree increased, the sum of the relative concentrations of hexanal, octanal, 2-nonenal, and 2-pentylfuran decreased to 65% (step 1) and 49% (step 2) of step 0. Hexanal concentrations of step 0 cooked black rice were significantly higher than those of polished rice (steps 1–2) (p < 0.05). In contrast, step 1 cooked rice showed no significant difference from step 2 (p > 0.05). The relative concentrations of octanal decreased significantly (step 0 > step 1 > step 2) as milling degrees increased (p < 0.05). The relative concentrations of 2-nonenal and 2-pentylfuran in step 0 were significantly higher than those of steps 1–2; however, those of step 1 and step 2 showed no significant difference (p > 0.05). Thus, by partially milling (step 1), 2-nonenal and 2-pentylfuran concentrations decreased significantly as did the concentrations of these compounds in fully milled rice (step 2).

In summary, various volatile compounds differed significantly among black rice samples. As milling degrees increased, unique aromatic compounds (e.g., guaiacol and vanillin) of black rice as well as off-flavor (e.g., lipid oxidation products) compound levels decreased in raw and cooked black rice. Partially milled rice (step 1) maintained guaiacol levels of unpolished rice (step 0) by up to 90 and 80% for raw and cooked black rice, respectively. The relative concentrations of several lipid oxidation products such as 2-nonenal and 2-pentylfuran were not significantly different in partially milled black rice (step 1) and fully milled black rice (step 2) (p > 0.05). Thus, partially milled rice can be used to retain favorable unique volatile compounds in black rice with low lipid oxidation levels.