Multi‐Dimensional Quality Attributes of Cooked Rice as Influenced by Nitrogen Supply and Milling Intensity

ABSTRACT

Nitrogen fertilization and milling degree are important determinants of the eating quality of rice, yet their combined effects on the multi‐dimensional quality of cooked rice remain insufficiently understood. In this study, rice from the japonica cultivar Yanjing 939 was produced under five nitrogen application rates (0, 150, 210, 270, and 330 kg/hm²) and processed to three milling degrees (0%, 5%, and 10%). Nitrogen application increased grain protein content and affected the pasting properties, texture, cooking behavior, sensory quality, and volatile characteristics of cooked rice. Moderate nitrogen application generally increased hardness, palatability, and overall sensory performance, whereas excessive nitrogen was associated with lower water absorption, swelling, and several sensory scores. Milling removed outer grain layers rich in protein, lipids, and crude fiber, thereby increasing water absorption and swelling and improving sensory quality, while reducing the number and abundance of volatile compounds. Significant interactions between nitrogen application and milling degree were observed for protein content and some pasting, textural, and sensory characteristics. Overall, moderate nitrogen application (210–270 kg/hm²) was generally favorable for Yanjing 939 in this study. DOM5 provided a more balanced combination of nutrient retention, sensory quality, and volatile composition, whereas DOM10 produced the highest overall sensory score at 270 kg/hm² but was accompanied by a broader reduction in volatile compounds.

1. Introduction

Rice remains one of the world's most important staple foods and agricultural commodities. According to the June 2026 World Agricultural Supply and Demand Estimates issued by the U.S. Department of Agriculture (USDA ²⁰²⁶), global production, total use, and trade of milled rice in 2026/27 are projected at 537.82, 541.18, and 63.01 million metric tons, respectively. The substantial scale of global rice production and trade highlights the importance of improving eating quality and cooking performance to enhance product value and meet consumer demand. Among the major grain components, protein plays an important role in cooked rice quality by affecting starch gelatinization, texture formation, and the release of volatile compounds.

The protein content and composition of rice are significantly influenced by cultivation practices, with nitrogen fertilizer application being one of the key regulatory factors (Wang et al. 2024). Appropriately increasing nitrogen application can promote nitrogen accumulation in the grains, enhance amino acid synthesis, and thereby increase protein content and its accumulation in the grains. However, the effects of nitrogen application on the sensory quality of rice are clearly complex. Ma et al. (2024) reported that, in some rice cultivars with high albumin content, high nitrogen application was associated with higher trough viscosity and final viscosity, as well as better appearance and taste value. However, some studies have indicated that excessive nitrogen application can cause excessive hardening of cooked rice and reduce the breakdown value, thereby diminishing palatability (Wang et al. 2024). These contrasting results are likely related to differences in protein distribution within the grain and in the way accumulated proteins interact with starch. Therefore, nitrogen application may affect cooked rice quality not only by changing total protein content, but also by modifying the structural environment in which starch gelatinization and texture formation occur.

Milling is a critical step in rice processing that directly affects both the physical appearance and edible quality of rice. The milling process removes the outer layers, including the husk, seed coat, and aleurone layer, thereby improving the appearance and reducing cooking time. Yi et al. (2024) reported that moderate milling improved the gelatinization characteristics, cooking performance, and sensory quality of rice but reduced the contents of volatile compounds. As the degree of milling (DOM) increases, the outer layers rich in proteins, lipids, and phenolic compounds are progressively removed. This leads to a more uniform internal composition and modifies the interactions between starch and protein. At the same time, excessive milling can cause substantial losses of nutrients, including proteins, lipids, vitamins, and phenolic compounds, and induce structural changes in the rice that ultimately influence the texture and flavor of cooked rice (Sandhu et al. 2018). Because protein accumulation resulting from nitrogen application occurs largely in the outer layers of rice grains, the effects of nitrogen on cooked rice quality are expected to vary with milling degree.

Although nitrogen application and DOM have been widely studied separately, it remains unclear how milling alters compositional differences resulting from nitrogen application and their expression in cooked rice quality. Nitrogen management establishes the compositional background of the rice, whereas milling regulates the retention or removal of outer layer components. These two stages are, therefore, closely connected in translating grain composition into the cooking and sensory properties of cooked rice. Clarifying this relationship may help rice processors adjust milling intensity according to raw material composition and meet specific product requirements for cooking performance, nutrient retention, texture, and aroma quality. These quality attributes are also closely associated with consumer eating experience and acceptance. Accordingly, this study evaluated the main and interactive effects of nitrogen application rate and DOM on the grain composition of Yanjing 939 rice and the texture and sensory quality of cooked rice, and further assessed their effects on the aroma profile of cooked rice. The findings may provide a basis for coordinating agronomic management and postharvest processing according to the intended quality of rice products.

2. Materials and Methods

2.1. Materials

Yanjing 939 (Japonica rice) was planted in separate plots at the experimental station in Dawa District, Panjin City, Liaoning Province, China (40°59′ N, 122°4′ E). Each plot covered an area of 50 m², with three replicates per treatment. Nitrogen fertilizer (urea, 46% nitrogen) was applied at five different rates: 0, 150, 210, 270, and 330 kg/hm², designated as N0, N150, N210, N270, and N330. The specific operation is the same as the method in our previous published paper (Ye et al. 2025).

2.2. Preparation of Rice With Different Milling Degrees

The rice grains were dehusked using a dehulling machine (THU35C, SATAKE, Suzhou, China) to produce brown rice. After manual sorting, 400 g of brown rice was weighed and milled using a rice polisher (TM05C, SATAKE). Unmilled brown rice was designated as DOM0, whereas rice samples with DOM values of 5% and 10% were prepared based on the mass loss during milling. Named, respectively, as DOM0‐0, DOM0‐150, DOM0‐210, DOM0‐270, DOM0‐330, DOM5%‐0, DOM5%‐150, DOM5%‐210, DOM5%‐270, DOM5%‐330, DOM10%‐0, DOM10%‐150, DOM10%‐210, DOM10%‐270, DOM10%‐330. Take 100 g of milled rice, grind it in a mill, and pass it through a 100‐mesh sieve. Seal the resulting rice flour and the milled rice separately in polyethylene bags and store them in a refrigerator at 4°C for later use. The DOM is calculated using the following formula (Ren et al. 2021):

$$DOM(\%)=(1-\frac{m}{M0})\times 100\%$$ (1)

DOM is the milling reduction rate, M ₀ is the mass of brown rice before milling, m is the mass of rice after milling.

2.3. Determination of Basic Nutritional Components in Rice

Moisture, fat, and protein contents were determined according to the Chinese National Standards using the direct drying method (GB 5009.3‐2016), Soxhlet extraction method (GB 5009.6‐2016), and Kjeldahl method (GB 5009.5‐2016), respectively. Amylose content was measured using a test kit (G0508F, Suzhou Grease Biotechnology Co., Ltd., Suzhou, China).

2.4. Determination of Rice Color and Taste Value

After calibrating the handheld colorimeter (CR‐400, Konica Minolta, Tokyo, Japan) using a white calibration plate, fill a Petri dish with the rice sample, compact it, and level the surface. Then, the colorimeter probe was positioned vertically at the center of the Petri dish, and the lightness (L*), redness (a*), yellowness (b*), and total color difference (ΔE) values were recorded. The taste value was an index on a 0–100 scale, directly generated by the algorithm built into the rice taste analyzer (JSWL, Beijing Dongfang Jiuheng Instrument Technology Co., Ltd., Beijing, China), and no additional manual calculation was performed.

2.5. Rapid Visco Analyzer Analysis

The pasting properties of the samples were determined using a Rapid Visco Analyzer (RVA 4500, Porton Instruments Ltd., Stockholm, Sweden). A 3 g sample of rice flour was weighed into an aluminum cup, and 25 mL of distilled water (based on 14% moisture content) was added. The sample was mixed using a stirring paddle, and the paddle was gently moved up and down to prevent the sample from sticking to the sides of the cup. Test procedure: Heat for 1 min at 50°C, then heat at a rate of 12°C/min to 95°C and hold at 95°C for 2.5 min. Subsequently, cool the temperature back to 50°C at a rate of 12°C/min. The stirrer speed was set to 960 rpm for the first 10 s, then maintained at 160 rpm for the remainder of the test (Li, Zhang et al. ²⁰²⁵).

2.6. Preparation of Cooked Rice

Weigh 10 g of the sample into an aluminum flask, rinse with distilled water, and drain thoroughly. Add distilled water equal to 1.3 times the mass of the sample and soak for 30 min at approximately 25°C. Bring water in a steamer to a boil, place the covered aluminum flask in the steamer, and steam for 30 min, followed by 20 min of steeping.

2.7. Cooking Characteristics

Water absorption index (WA): Weigh 10 g (M ₀) of a rice sample and place it in an aluminum dish of known mass for cooking. After cooking, determine the mass of the cooked rice (M ₁) and calculate the WA using the following formula (Fu et al. 2025):

$$WA(\%)=\frac{M1-M0}{M0}\times 100\%$$ (2)

Swelling power index (SP): Weighed 10 g of raw rice, cooked it, and recorded the mass of the cooked rice as W. Weigh out 5 g of the cooked rice and place it in a 10 mL graduated cylinder. Add 5 mL of water and calculate the volume V ₁ of the 5 g of cooked rice based on the change in water level. The total volume V ₂ of the cooked rice can be calculated from V ₁. The volume V ₀ of the raw rice is determined using the same method. The SP is calculated using the following formula (Fu et al. 2025):

$$SP(\%)=\frac{V2-V0}{V0}\times 100\%$$ (3)

2.8. Determination of Texture Properties

The textural properties of cooked rice were determined using a texture analyzer (TMS‐Pilot, FTC Company, Washington, USA), with the P/36R probe selected. The parameters were set as follows: pre‐compression and post‐compression speeds were 1 mm/s, and the compression ratio was set to 75%. Three intact grains of rice were selected and placed side by side on the sample stage for measurement.

2.9. Electronic Nose Analysis

Following the method described by Ye et al. (2025) with minor modifications, an electronic nose (PEN3 Air Sense, Schwerin, Germany) was used to analyze the flavor profiles of different rice samples. Two grams of cooked rice were placed in a headspace vial, sealed with plastic wrap. The sampling interval, flushing time, zero adjustment time, and presampling time were set to 1, 100, 10, and 5 s, respectively. The electronic nose system was operated for 60 s at a constant flow rate of 300 mL/min. Sensor responses were expressed as the normalized conductance ratio G/G ₀, where G and G ₀ represent the sensor conductance in the sample gas and clean reference air, respectively. As the sensor signals had stabilized by 60 s, the response value of each sensor at 60 s was selected for subsequent analysis. Each sample was measured in triplicate, and the mean response values were used to construct the radar plots. The electronic nose sensor array is shown in Table S1.

2.10. Gas Chromatograph‐Mass Spectrometer

Based on the methods described by Guan et al. (2025), volatile compounds in cooked rice were analyzed using gas chromatography‐mass spectrometry (7890A GC, 5975C MS, Agilent Technologies, Santa Clara, CA, USA). Transfer 5 g of cooked rice sample to a 20 mL headspace vial and add 1 µL of 2‐methyl‐3‐heptanone (0.816 µg/µL) as an internal standard. Prior to extraction, the SPME (Solid‐Phase Microextraction) fiber was conditioned at 250°C for 1 h. The sample was then equilibrated in a self‐heating incubator at 100°C for 25 min. After equilibration, the fiber was exposed to the headspace above the cooked rice for 40 min to adsorb volatile compounds. The fiber was subsequently introduced into the GC injection port at 250°C and thermally desorbed for 5 min. Volatile compounds were analyzed using a GC‐MS system consisting of a gas chromatograph (7890A, Agilent Technologies) coupled with a mass spectrometer (5975C, Agilent Technologies). Separation was performed on a DB‐WAX capillary column (30 m × 0.25 mm × 0.25 µm; Agilent J&W). Helium was used as the carrier gas at a flow rate of 1.2 mL/min in splitless mode. The oven temperature was initially held at 40°C for 3 min, then increased to 200°C at 5°C/min, followed by a further increase to 230°C at 10°C/min and held for 3 min. Mass spectrometric detection was carried out under electron impact ionization at 70 eV. The transfer line, quadrupole, and ion source temperatures were set at 280°C, 150°C, and 230°C, respectively. Mass spectra were recorded in full‐scan mode over an m/z range of 55–500. The concentrations of volatile compounds were calculated using the following formula:

$$C(\mu g/kg)=\frac{Pc}{Pis}\times Cis\times 1000/m0$$ (4)

In the equation, C represents the concentration of volatile compounds, C_is is the internal standard concentration (2‐methyl‐3‐heptanone, 0.816 µg/µL), P_c represents the peak area of the volatile compounds, P_is represents the peak area of 2‐methyl‐3‐heptanone, and M ₀ represents the mass of cooked rice.

The formula for calculating the relative odor activity value (ROAV) is as follows:

$$ROAV=\frac{Cx\times T\max }{C\max \times Tx}\times 100\%,$$ (5)

where C_x represents the relative percentage content of each volatile compound (%), T_x represents the corresponding odor threshold of the compound (mg/kg), C_max represents the relative percentage content of the compound with the greatest contribution to the overall aroma of the sample (%), and T_max represents the corresponding odor threshold of that compound (mg/kg).

Odor threshold values were obtained from Odor Thresholds: Compilations of Odor Threshold Values in Air, Water and Other Media (second enlarged and revised edition). When multiple values were available, those determined in aqueous systems and reported in more recent publications were preferentially adopted.

2.11. Sensory Evaluation of Cooked Rice

According to the method described by Li, Zhang et al. (2025), sensory evaluation was conducted by 12 trained panelists from the School of Food Science. The panel consisted of eight women aged 22–26 years and four men aged 21–30 years at the time of evaluation. All panelists assessed the rice samples for odor, appearance, palatability, taste, and cold rice texture. The texture of cooled rice was evaluated after the cooked rice had been kept at room temperature for 1 h. The detailed training procedure has been added to Text S1. The sensory evaluation standards are shown in Table S2. The overall sensory score was calculated as the sum of the five sensory attributes listed in Table S2, including odor (0–20 points), appearance (0–20 points), palatability (0–30 points), taste (0–25 points), and cold‐rice texture (0–5 points), with a maximum total score of 100 points. The final overall score was expressed as the mean score of the trained panelists.

2.12. Ethical Statement and Informed Consent

According to the institutional policies of Beijing Technology and Business University and relevant national guidelines, this category of low‐risk sensory analysis involving the tasting of common food products does not require formal approval from an Institutional Review Board (IRB). This study was conducted in accordance with the ethical principles for sensory science. All panelists were fully informed about the study procedures and provided their written informed consent prior to their participation. They were also advised of their right to withdraw from the session at any time without penalty, with an institutional statement confirming this ethical compliance.

2.13. Statistical Analysis

All experiments were performed at least in triplicate, and the results are expressed as mean ± standard deviation. Data analysis and figure preparation were conducted using SPSS Statistics 25.0 and Origin 2022, respectively. The significance of differences was evaluated using two‐way analysis of variance (two‐way ANOVA) followed by multiple‐comparison tests. * and ** indicate significance at p < 0.05 and p < 0.01, respectively, whereas ns indicates no significant difference.

3. Results and Discussion

3.1. Analysis of the Basic Nutritional Components of Rice

Table 1 shows that the effects of nitrogen application on the basic nutritional composition of rice were expressed differently at different milling degrees. Two‐way ANOVA indicated that nitrogen application significantly affected protein content but had no significant main effects on moisture, fat, or amylose contents, whereas milling degree significantly affected all four components. A significant interaction between nitrogen application and milling degree was observed only for protein content. Increasing nitrogen application consistently increased protein content across the three milling degrees, indicating that nitrogen application primarily promoted protein accumulation. However, this increase in protein became less pronounced as milling degree increased, suggesting that part of the protein accumulated in response to nitrogen application was removed during milling. Increasing milling degree generally decreased moisture, fat, and protein contents while increasing amylose content. This is mainly attributed to the uneven distribution of nutrients within the grain, as proteins and lipids are concentrated in the outer layers, particularly the bran and aleurone layers (Li, Zhang et al. ²⁰²⁵). Progressive removal of these tissues reduced fat and protein contents and increased the relative proportion of starch in the remaining endosperm fraction. At a milling degree of 10%, fat loss became particularly evident, confirming the extensive removal of outer tissues rich in lipids during milling. The decline in moisture content with increasing milling degree may also be related to the loss of outer tissues and heat generation during milling. Thus, nitrogen application primarily increased grain protein accumulation, whereas increasing milling degree reduced the nitrogen‐related differences in protein content through the removal of protein‐rich outer layers.

TABLE 1.

Effects of nitrogen application rate and milling degree on the nutritional components of rice.

DOM (%)	Treatment (kg/hm2)	Moisture (%)	Fat (%)	Protein (%)	Amylose (%)
0	0	9.88 ± 0.40a	4.66 ± 0.12a	6.71 ± 0.08d	19.21 ± 0.58a
	150	9.74 ± 0.43a	4.60 ± 0.07a	7.33 ± 0.07c	18.63 ± 0.58a
	210	9.87 ± 0.39a	4.76 ± 0.23a	7.57 ± 0.09b	18.51 ± 0.31a
	270	10.41 ± 0.46a	4.69 ± 0.29a	7.88 ± 0.03a	18.73 ± 0.09a
	330	10.40 ± 0.39a	4.73 ± 0.17a	7.93 ± 0.02a	18.66 ± 1.10a
5	0	9.60 ± 0.51a	3.25 ± 0.13a	6.39 ± 0.07d	21.20 ± 1.86a
	150	9.50 ± 0.29a	3.22 ± 0.09a	6.85 ± 0.14c	19.56 ± 0.43a
	210	9.88 ± 0.25a	3.16 ± 0.32a	7.15 ± 0.10b	19.65 ± 0.58a
	270	9.48 ± 0.38a	3.30 ± 0.09a	7.22 ± 0.26ab	19.76 ± 0.16a
	330	9.37 ± 0.35a	3.24 ± 0.06a	7.51 ± 0.10a	20.70 ± 0.87a
10	0	8.66 ± 0.44a	1.29 ± 0.06ab	6.18 ± 0.03c	20.97 ± 1.37a
	150	8.63 ± 0.39a	1.26 ± 0.03ab	6.37 ± 0.05c	22.79 ± 1.92a
	210	8.68 ± 0.78a	1.33 ± 0.04a	6.67 ± 0.05b	21.58 ± 1.16a
	270	8.94 ± 0.42a	1.23 ± 0.04b	6.82 ± 0.19b	21.36 ± 2.26a
	330	8.85 ± 0.19a	1.24 ± 0.04ab	7.30 ± 0.19a	20.87 ± 1.31a
ANOVA	DOM	**	**	**	**
	N	ns	ns	**	ns
	N × DOM	ns	ns	**	ns

Note: Data are presented as mean ± SD (n = 3). Within the same milling degree, different lowercase letters in the same column indicate significant differences among nitrogen application treatments (p < 0.05). ns, p ≥ 0.05; *, p < 0.05; **, p < 0.01.

Abbreviations: DOM, milling degree; N, nitrogen application rate; N × DOM, interaction between nitrogen application rate and milling degree.

3.2. Analysis of Rice Color and Taste Value

Table 2 shows that both milling degree and nitrogen application rate significantly affected all color parameters. Significant interactions between nitrogen application and milling degree were observed for a* and b*, indicating that their responses to nitrogen depended on milling degree. In brown rice, increasing nitrogen application decreased the L* value and increased the a* and b* values, indicating a darker appearance with enhanced redness and yellowness. Ren et al. (2021) reported that components such as proteins and minerals contribute to the red and yellow color attributes of cereal flour. However, this color response became progressively weaker as milling degree increased. At DOM5, nitrogen application mainly affected L* and ΔE. At DOM10, L* and b* varied only slightly among nitrogen treatments, whereas a* and ΔE still showed significant differences. These results suggest that the color response to nitrogen was closely associated with components in the outer layers, because proteins, minerals, and pigments concentrated in bran tissues were gradually removed during milling. As a result, brown rice retained greater nitrogen‐related differences in outer layer composition, whereas extensive milling reduced these differences by removing tissues rich in proteins and pigments. Li et al. (2025) similarly reported that increasing milling degree reduced rice color values.

TABLE 2.

Effects of nitrogen application rate and milling degree on rice color.

DOM (%)	Treatment (kg/hm2)	L*	a*	b*	ΔE
0	0	62.42 ± 0.78a	3.31 ± 0.11b	18.91 ± 0.31b	33.12 ± 0.67c
	150	62.58 ± 0.38a	3.38 ± 0.21b	19.41 ± 0.43b	33.19 ± 0.52c
	210	61.94 ± 0.22ab	3.94 ± 0.51a	20.69 ± 0.77a	34.38 ± 0.50b
	270	61.27 ± 0.12bc	4.00 ± 0.24a	20.62 ± 0.26a	34.95 ± 0.24ab
	330	60.61 ± 0.34c	4.22 ± 0.19a	20.35 ± 0.43a	35.45 ± 0.46a
5	0	70.35 ± 0.15a	0.96 ± 0.13b	13.04 ± 0.36bc	23.41 ± 0.28c
	150	69.37 ± 0.07b	1.07 ± 0.03b	13.04 ± 0.30bc	24.36 ± 0.09b
	210	68.46 ± 0.44c	0.96 ± 0.11b	12.68 ± 0.39c	25.12 ± 0.52b
	270	67.71 ± 0.55cd	1.40 ± 0.22a	14.04 ± 0.81a	26.27 ± 0.62a
	330	67.48 ± 0.49d	1.43 ± 0.16a	13.86 ± 0.32ab	26.43 ± 0.54a
10	0	73.39 ± 0.52a	0.09 ± 0.06ab	10.16 ± 0.16a	19.67 ± 0.48b
	150	73.01 ± 1.00a	−0.15 ± 0.09bc	9.77 ± 0.54a	19.96 ± 0.89ab
	210	72.36 ± 0.66a	−0.15 ± 0.19bc	10.11 ± 0.45a	20.66 ± 0.75ab
	270	72.24 ± 1.29a	−0.34 ± 0.06c	9.59 ± 0.31a	20.68 ± 1.21ab
	330	71.56 ± 0.70a	0.13 ± 0.18a	10.30 ± 0.55a	21.49 ± 0.71a
ANOVA	DOM	**	**	**	**
	N	**	**	**	**
	N × DOM	ns	**	**	ns

Note: Data are presented as mean ± SD (n = 3). Within the same milling degree, different lowercase letters in the same column indicate significant differences among nitrogen application treatments (p < 0.05). ns, p ≥ 0.05; *, p < 0.05; **, p < 0.01. L* represents lightness, a* indicates redness or greenness, b* indicates yellowness or blueness, and ΔE represents the overall color difference between samples.

Abbreviations: DOM, milling degree; N, nitrogen application rate; N × DOM, interaction between nitrogen application rate and milling degree.

As shown in Figure 1, two‐way ANOVA indicated that both nitrogen application and milling degree significantly affected taste value, whereas their interaction was not significant. Taste value decreased with increasing nitrogen application, possibly because nitrogen‐induced changes in grain composition, particularly the increase in protein content, altered the transparency and uniformity of the rice grains. By contrast, milling removed outer layer components rich in protein, lipids, and crude fiber, thereby improving the taste value of rice.

FIGURE 1.

Effects of nitrogen application and milling degree on the taste value of rice. ns, p ≥ 0.05; *, p < 0.05; **, p < 0.01. Abbreviations: DOM, milling degree; N, nitrogen application rate; N × DOM, interaction between nitrogen application rate and milling degree.

3.3. Analysis of the Pasting Properties of Rice

The pasting properties of rice under different nitrogen application rates and milling degrees were determined using an RVA, which measures the gelatinization, swelling, breakdown, and setback of starch (Hu et al. 2024). The effects of nitrogen application and milling degree on the pasting properties of rice are shown in Figure 2 and Table S3. Two‐way ANOVA showed that milling degree significantly affected all RVA parameters, whereas nitrogen application affected all except breakdown. Significant interactions occurred for peak, trough, and final viscosities and pasting temperature, indicating that their responses to nitrogen depended on milling degree. In brown rice, peak viscosity began to decrease numerically when the nitrogen application rate exceeded 150 kg/hm², whereas breakdown decreased after 210 kg/hm², suggesting that high nitrogen application restricted starch granule swelling and reduced starch paste breakdown during heating and shear (Shi et al. 2023). At 210 kg/hm², pasting temperature decreased from 75.75°C in DOM0 to 75.22°C in DOM5 and 74.18°C in DOM10, indicating that milling reduced the temperature required for pasting. This trend may be associated with the progressive removal of proteins, lipids, and dietary fiber from the outer grain layers, thereby weakening their constraints on starch hydration and swelling. Increasing milling degree generally increased peak, trough, and final viscosities as well as breakdown values across nitrogen treatments. The higher breakdown values at DOM10 indicate greater susceptibility of swollen starch granules to disruption and lower starch paste stability during heating. Overall, milling altered the nitrogen response of several pasting characteristics and may have facilitated starch hydration and gelatinization by removing outer layer components.

FIGURE 2.

Effects of nitrogen application and milling degree on the pasting properties of rice.

3.4. Analysis of the Cooking Properties of Cooked Rice

Cooking properties reflect water penetration and starch swelling during heating and are closely associated with eating quality. As shown in Figure 3, two‐way ANOVA indicated that both nitrogen application and milling degree significantly affected the water absorption index and swelling power index, whereas their interactions were not significant. Both indices decreased with increasing nitrogen application but increased with increasing milling degree. Higher nitrogen application increased grain protein accumulation, which likely strengthened starch−protein interactions and promoted the formation of a denser matrix around starch granules. Such structural constraints hindered water penetration and diffusion during cooking, thereby limiting starch swelling and reducing grain expansion (Shi et al. 2023). In contrast, milling removed outer‐layer tissues enriched in protein, lipids, and crude fiber, which weakened the physical barriers to water transfer and exposed the more porous endosperm (Zhu et al. 2020). As a result, water uptake and starch swelling were facilitated, leading to higher water absorption and swelling indices. Thus, nitrogen application reduced water uptake and swelling, whereas increasing milling degree promoted both processes through the removal of outer‐layer components.

FIGURE 3.

Effects of nitrogen application and milling degree on the cooking properties of rice. ns, p ≥ 0.05; *, p < 0.05; **, p < 0.01. Abbreviations: DOM, milling degree; N, nitrogen application rate; N × DOM, interaction between nitrogen application rate and milling degree.

3.5. Analysis of the Texture Properties of Cooked Rice

The effects of nitrogen application and milling degree on the textural properties of cooked rice are shown in Table 3. Two‐way ANOVA indicated that milling degree significantly affected all textural parameters, whereas nitrogen application significantly affected hardness, cohesiveness, springiness, and chewiness but not adhesiveness. Significant interactions between nitrogen application and milling degree were observed for hardness, adhesiveness, and chewiness, indicating that the responses of these parameters to nitrogen depended on milling degree. At lower milling degrees, hardness and chewiness generally increased with nitrogen application and then declined under excessive nitrogen. This suggests that moderate nitrogen favored the formation of a more stable rice matrix. This trend may be associated with protein accumulation resulting from nitrogen application. Higher protein content may have strengthened starch−protein interactions and promoted a denser internal structure, thereby restricting starch swelling and gelatinization during cooking and resulting in a firmer texture (Li et al. 2020; Xiong et al. ²⁰²²). Under excessive nitrogen, however, further protein accumulation may disrupt the continuity of the starch matrix, thereby reducing hardness and chewiness (Lin et al. 2022). The marked decreases in cohesiveness, springiness, and chewiness at DOM10 indicate that extensive milling weakened the structural integrity and elastic recovery of cooked rice. These findings indicate that cooked rice texture was jointly determined by nitrogen‐induced compositional changes and milling‐related alterations in grain structure.

TABLE 3.

Effects of nitrogen application rate and milling degree on the texture characteristics of rice.

DOM (%)	Treatment (kg/hm2)	Hardness (N)	Cohesiveness	Springiness (mm)	Adhesiveness (mj)	Chewiness (N)
0	0	9.93 ± 0.41d	0.27 ± 0.04a	0.84 ± 0.07b	2.62 ± 0.37c	2.22 ± 0.41c
	150	12.58 ± 1.03c	0.25 ± 0.00a	0.74 ± 0.06c	3.10 ± 0.23bc	2.29 ± 0.10bc
	210	14.39 ± 0.68b	0.23 ± 0.02a	0.69 ± 0.04c	3.31 ± 0.20b	2.29 ± 0.20bc
	270	16.35 ± 1.31a	0.27 ± 0.01a	1.00 ± 0.04a	4.39 ± 0.40a	4.40 ± 0.40a
	330	13.93 ± 0.32bc	0.26 ± 0.02a	0.78 ± 0.05bc	3.64 ± 0.36b	2.82 ± 0.24b
5	0	11.22 ± 0.85d	0.27 ± 0.04a	0.84 ± 0.16ab	3.03 ± 0.68b	2.53 ± 0.71a
	150	12.21 ± 0.65cd	0.28 ± 0.04a	0.81 ± 0.08ab	3.39 ± 0.50ab	2.77 ± 0.60a
	210	12.79 ± 0.85bc	0.29 ± 0.04a	0.89 ± 0.09a	3.65 ± 0.70ab	3.29 ± 0.86a
	270	15.57 ± 0.98a	0.25 ± 0.03a	0.82 ± 0.09ab	3.84 ± 0.33a	3.14 ± 0.46a
	330	13.63 ± 1.24b	0.24 ± 0.04a	0.75 ± 0.06b	3.31 ± 0.69ab	2.49 ± 0.64a
10	0	9.33 ± 1.08c	0.10 ± 0.02b	0.45 ± 0.09a	0.95 ± 0.08b	0.44 ± 0.12b
	150	10.75 ± 1.09bc	0.09 ± 0.01b	0.43 ± 0.04a	0.97 ± 0.10b	0.42 ± 0.08b
	210	13.56 ± 1.80a	0.14 ± 0.01a	0.51 ± 0.10a	1.78 ± 0.11a	0.91 ± 0.21a
	270	12.06 ± 1.52ab	0.10 ± 0.02b	0.43 ± 0.08a	1.16 ± 0.22b	0.51 ± 0.15b
	330	11.72 ± 0.90abc	0.09 ± 0.02b	0.38 ± 0.03a	1.12 ± 0.26b	0.43 ± 0.12b
ANOVA	DOM	**	**	**	**	**
	N	**	*	**	ns	*
	N × DOM	**	ns	ns	**	**

Note: Data are presented as mean ± SD (n = 3). Within the same milling degree, different lowercase letters in the same column indicate significant differences among nitrogen application treatments (p < 0.05). ns, p ≥ 0.05; *, p < 0.05; **, p < 0.01.

Abbreviations: DOM, milling degree; N, nitrogen application rate; N × DOM, interaction between nitrogen application rate and milling degree.

3.6. Analysis of Electronic Nose Results

The overall volatile characteristics of cooked rice were further evaluated using an electronic nose. As shown in Figure 4, the radar profiles of cooked rice differed distinctly across treatments. Across the three milling groups, W1W, W2W, W1S, and W2S showed relatively high responses. By contrast, W1C, W3C, and W5C showed low responses and limited variation. According to the sensor characteristics, the observed differences were mainly associated with sulfur‐containing compounds, methyl compounds, alcohols, and aldehyde‐related signals. Signals corresponding to aromatic hydrocarbons and alkanes varied little among treatments. With increasing nitrogen application, the W1W response generally decreased. In contrast, W1S and W2W tended to be relatively higher at moderate nitrogen levels, especially under DOM5 and DOM10. This suggests that nitrogen application altered the relative contribution of different volatile groups rather than simply changing overall aroma intensity. Increasing milling degree generally reduced the sensor responses, especially those of W1W and W2W. This indicates that milling reduced the sensor responses to certain volatile signals originating from the outer layers. These results suggest that milling attenuated the aroma differences established under different nitrogen treatments by removing outer‐layer components. As electronic nose analysis mainly reflects overall odor fingerprints, these changes should be interpreted together with the GC‐MS results.

FIGURE 4.

Effects of nitrogen application and milling degree on the electronic nose results of cooked rice.

3.7. Analysis of Volatile Compounds

GC‐MS analysis further showed that nitrogen application and milling degree affected the number and composition of volatile compounds in cooked rice, and the detailed profiles are presented in Tables S4–S6. As shown in Figure 5, the detected volatile compounds mainly comprised aldehydes, alcohols, ketones, hydrocarbons, and heterocyclic compounds, with aldehydes being the predominant class across all treatments. Within each milling group, the number of detected volatile compounds varied only moderately with nitrogen application but declined clearly with increasing milling degree. Specifically, 32, 33, 32, 31, and 30 volatile compounds were detected under DOM0 at 0, 150, 210, 270, and 330 kg/hm², respectively. The corresponding numbers were 31, 29, 27, 27, and 26 under DOM5 and 24, 24, 25, 26, and 25 under DOM10. These results indicate that milling had a stronger effect than nitrogen application on the number of detected volatile compounds, whereas aldehydes remained predominant regardless of treatment. The predominance of aldehydes suggests that lipid oxidation, particularly the oxidative degradation of unsaturated fatty acids, was an important pathway of volatile formation in cooked rice, consistent with previous reports (Bryant and McClung ²⁰¹¹). Representative aldehydes, including hexanal, octanal, nonanal, and decanal, were detected in all samples, with nonanal remaining one of the most abundant compounds across the three milling groups. Several heterocyclic compounds, including 2‐pentylfuran, 2,3‐dihydrobenzofuran, indole, and 2‐acetyl‐1‐pyrroline (2‐AP), were also detected.

FIGURE 5.

Effects of nitrogen application and milling degree on the number of volatile compounds in cooked rice.

The heatmap of volatile compounds is shown in Figure 6. The relative abundances of individual volatiles also responded nonlinearly to nitrogen application. Under DOM5, several representative volatiles and some compounds with ROAV values greater than 1 showed relatively high abundances at 210 kg/hm². Under DOM0 and DOM10, several compounds reached relatively high levels at 270 kg/hm². These results suggest that the effect of nitrogen application on volatile composition may be related to differences in precursor retention caused by milling. However, a higher volatile abundance does not necessarily indicate better aroma quality. Nitrogen application not only affects protein accumulation but may also alter the balance among amino acids, lipids, and carbonyl precursors, thereby influencing the formation of aldehydes derived from lipid oxidation and compounds associated with Maillard reactions during cooking (Ye et al. 2025). This milling‐dependent response was particularly evident for 2‐AP, whose measured content reached relatively high levels at 210 and 270 kg/hm² in the DOM5 and DOM10 samples, respectively. Furans and other heterocyclic compounds also exhibited clear treatment‐dependent responses. For example, 2‐pentylfuran reached its highest content at 270 kg/hm² under DOM0 and at 210 kg/hm² under DOM5, whereas it showed a different response pattern under DOM10. Therefore, the effect of nitrogen application on cooked rice aroma was mainly reflected in changes in the formation and relative contributions of volatiles originating from different pathways. However, volatile abundance alone does not directly indicate the sensory importance of individual compounds.

FIGURE 6.

Heatmap of volatile compounds in cooked rice under different nitrogen application rates and milling degrees.

To further evaluate the sensory relevance of individual volatiles, ROAVs were calculated, and the detailed results are presented in Tables S7–S9. A total of 13, 12, and 13 compounds with ROAV values greater than 1 were identified in the DOM0, DOM5, and DOM10 groups, respectively. Most compounds with relatively high ROAV values were aldehydes, indicating their important contributions to cooked rice aroma. However, 2‐acetyl‐1‐pyrroline (2‐AP) and 1‐octen‐3‐ol were also of considerable sensory importance because their low odor thresholds enabled them to contribute markedly even at relatively low contents (Dias et al. 2021; Wei et al. ²⁰¹⁷). Hexanal, octanal, nonanal, and (E)‐2‐nonenal were associated with grassy, metallic, woody, or fatty notes, respectively, and their excessive accumulation may alter the overall aroma quality of cooked rice. In contrast, 2‐AP, characterized by popcorn‐like and cooked cereal‐like notes, is an important characteristic aroma compound of cooked rice (Deng et al. 2013). The ROAVs of individual compounds responded nonlinearly to nitrogen application and were modulated by milling degree. In the DOM0 group, the ROAVs of aldehydes showed different responses to nitrogen application, with those of hexanal and octanal generally peaking at 270 kg/hm², nonanal remaining at a high level from 150 to 270 kg/hm², and (E)‐2‐nonenal peaking at 210 kg/hm², whereas the ROAV of 2‐AP fluctuated across nitrogen treatments. Under DOM5, 2‐AP and nonanal both showed relatively high contributions at 210 kg/hm², with ROAV values of 68.87 and 100.00, respectively. Under DOM10, the ROAV of 2‐AP reached 100.00 at both 210 and 330 kg/hm², while that of nonanal reached 66.79 and 70.38 under the same treatments, respectively. The ROAV of indole, which was associated with floral and animal‐like notes, also increased to 25.20 at 330 kg/hm². These changes do not necessarily indicate that nitrogen application intensified off‐odors. Rather, nitrogen application simultaneously altered the relative contributions of characteristic aroma compounds, oxidation‐associated aldehydes, and other volatiles with low odor thresholds, and the resulting sensory effect depended on the overall balance among multiple odor‐active compounds.

Milling further modified the differences in volatile composition caused by nitrogen application. As milling degree increased, both the number and overall abundance of detected volatiles decreased, with particularly pronounced reductions in several aldehydes derived from lipid oxidation. Under the treatment without nitrogen application, the nonanal content decreased from 45.25 µg/kg in DOM0 to 18.56 µg/kg in DOM5 and 8.48 µg/kg in DOM10. Similar decreases were observed for hexanal, 1‐octen‐3‐ol, 2‐pentylfuran, 2,3‐dihydrobenzofuran, and indole. These results are consistent with the relative enrichment of lipids and other volatile precursors in the bran‐rich outer layers, whose progressive removal may reduce the precursor supply for aldehydes derived from lipid oxidation and some heterocyclic compounds (Ma et al. 2020; Mi et al. ²⁰²³). However, milling did not selectively remove compounds potentially associated with undesirable odors. (E)‐2‐Nonenal exhibited an ROAV value greater than 1 only in the DOM0 group, suggesting that removal of the outer layers may reduce its potential odor contribution. At 210 kg/hm², however, the measured content of 2‐AP also decreased with increasing milling degree, from 2.19 µg/kg in DOM0 to 1.18 µg/kg in DOM5 and 0.72 µg/kg in DOM10, whereas its ROAV increased from 60.71 to 68.87 and 100.00, respectively. This result indicates that extensive milling increased the relative importance of 2‐AP within the volatile profile rather than its measured content. The number and overall abundance of volatiles in DOM5 were generally intermediate between those in DOM0 and DOM10. While reducing some oxidation products associated with the outer layers, DOM5 maintained higher contents of 2‐AP and several compounds with low odor thresholds than DOM10 under most nitrogen treatments. In contrast, DOM10 showed a broader reduction in volatiles, involving both compounds with potentially undesirable odor attributes and characteristic aroma compounds. It should also be noted that the sensory roles of aldehydes were not uniformly negative. (E,E)‐2,4‐Decadienal was associated with fresh and peanut‐like notes and maintained relatively high ROAV values in most treatments. In addition, 1‐octen‐3‐ol, another compound with a low odor threshold, contributed mushroom‐like and peanut milk‐like notes. Therefore, the influence of milling on cooked rice aroma mainly arose from changes in volatile precursor distribution and the redistribution of the relative contributions of different odor‐active compounds, rather than merely from a reduction in total volatile abundance.

3.8. Sensory Analysis

The sensory quality of cooked rice under different nitrogen application rates and milling degrees is shown in Table 4. Two‐way ANOVA indicated that milling degree significantly affected all sensory attributes, whereas nitrogen application significantly affected odor, palatability, taste, and overall sensory score but not appearance or cold rice texture. Significant interactions between nitrogen application and milling degree were observed for odor, palatability, taste, and overall sensory score, indicating that the responses of these attributes to nitrogen depended on milling degree. Across the three milling groups, nitrogen application at 210 or 270 kg/hm² resulted in higher scores for odor, palatability, and taste. The overall sensory score was numerically highest at 210 kg/hm² in DOM0, reaching 70.64, whereas the highest scores in DOM5 and DOM10 occurred at 270 kg/hm², reaching 84.64 and 85.09, respectively. These results indicate that nitrogen application within the range of 210–270 kg/hm² was generally favorable for sensory quality, whereas excessive nitrogen application adversely affected sensory quality.

TABLE 4.

Effects of nitrogen application rate and milling degree on sensory quality of rice.

DOM (%)	Treatment (kg/hm2)	Odor (score)	Appearance (score)	Palatability (score)	Taste (score)	Cold rice texture (score)	Overall score (score)
0	0	12.27 ± 1.90bc	11.09 ± 2.02a	19.73 ± 2.10b	17.00 ± 1.27b	2.82 ± 0.98a	60.91 ± 3.02c
	150	14.00 ± 2.00ab	11.45 ± 2.88a	21.09 ± 2.12ab	18.09 ± 2.30ab	2.82 ± 1.17a	67.45 ± 4.63ab
	210	14.91 ± 1.51a	11.73 ± 3.74a	23.00 ± 2.24a	18.18 ± 2.44ab	2.82 ± 0.87a	70.64 ± 3.04a
	270	15.09 ± 1.45a	12.18 ± 2.04a	21.36 ± 2.20ab	19.18 ± 1.60a	2.64 ± 1.12a	70.45 ± 5.77a
	330	11.55 ± 2.25c	10.82 ± 1.83a	19.82 ± 1.99b	18.36 ± 1.91ab	2.82 ± 0.75a	63.36 ± 3.47bc
5	0	15.91 ± 0.83b	14.00 ± 1.95a	24.82 ± 1.83c	18.45 ± 1.64b	3.45 ± 0.69a	76.64 ± 3.72c
	150	16.73 ± 1.35ab	14.91 ± 1.58a	25.00 ± 1.00c	19.73 ± 1.85ab	3.55 ± 1.04a	79.91 ± 4.57bc
	210	17.55 ± 1.29a	14.73 ± 2.10a	26.73 ± 1.62ab	21.36 ± 1.80a	3.73 ± 1.10a	84.09 ± 4.18a
	270	17.55 ± 0.52a	15.36 ± 2.16a	27.55 ± 1.81a	20.09 ± 1.76ab	4.09 ± 0.70a	84.64 ± 4.20a
	330	16.27 ± 0.65b	15.73 ± 2.53a	25.91 ± 1.04bc	20.09 ± 1.76ab	4.18 ± 0.98a	82.18 ± 3.43ab
10	0	13.00 ± 2.15d	16.73 ± 1.68a	24.36 ± 1.21b	19.36 ± 1.86ab	3.73 ± 0.91a	77.18 ± 4.60c
	150	14.45 ± 1.86c	17.64 ± 1.36a	25.18 ± 0.87ab	19.91 ± 1.51a	4.18 ± 0.75a	81.36 ± 3.67b
	210	16.00 ± 1.55ab	17.64 ± 1.75a	26.00 ± 2.05a	20.64 ± 2.16a	4.18 ± 0.87a	84.45 ± 2.51ab
	270	16.64 ± 1.36a	18.09 ± 2.07a	25.64 ± 2.11ab	20.64 ± 2.20a	4.09 ± 0.83a	85.09 ± 4.46a
	330	14.82 ± 1.08bc	17.36 ± 1.63a	21.73 ± 0.79c	17.82 ± 1.47b	4.00 ± 1.10a	75.73 ± 2.15c
ANOVA	DOM	**	**	**	**	**	**
	N	**	ns	**	**	ns	**
	N × DOM	*	ns	**	*	ns	*

Note: Within the same milling degree, different lowercase letters in the same column indicate significant differences among nitrogen application treatments (p < 0.05). ns, p ≥ 0.05; *, p < 0.05; **, p < 0.01.

Abbreviations: DOM, milling degree; N, nitrogen application rate; N × DOM, interaction between nitrogen application rate and milling degree.

Milling degree significantly altered the sensory characteristics of cooked rice. Compared with DOM0, both DOM5 and DOM10 showed markedly higher overall sensory scores. This improvement may be associated with the removal of bran‐rich outer tissues containing crude fiber, lipids, and other components that restrict water penetration and starch gelatinization during cooking. However, sensory quality did not improve consistently as milling degree increased. Both DOM5 and DOM10 maintained relatively high overall sensory scores at 210–270 kg/hm². At 330 kg/hm², DOM5 retained relatively high odor, palatability, and overall sensory scores, whereas DOM10 showed marked decreases in palatability, taste, and overall score. These results suggest that both milling degrees performed well under moderate nitrogen application, while DOM5 showed greater sensory stability under high nitrogen application.

PCA (Principal Component Analysis) was conducted separately for DOM0, DOM5, and DOM10 samples and for all samples combined (Figures S1−S4). Considerable overlap among nitrogen treatments was observed in the score plots, indicating limited discrimination based on sensory profiles. The first two principal components explained 60.8%, 67.2%, 67.8%, and 73.0% of the total variance for DOM0, DOM5, DOM10, and the combined dataset, respectively. Loading plots showed that odor was most closely associated with the overall sensory score in DOM0 and DOM10, whereas taste showed the closest association in DOM5. When all samples were analyzed together, palatability was most closely related to the overall sensory score. Collectively, odor, taste, and palatability were identified as potential sensory drivers associated with the overall sensory score, with their relative importance varying with milling degree.

3.9. Potential Economic and Environmental Implications

The present findings provide preliminary insight into the potential economic and environmental implications of nitrogen and milling management. Under the present conditions, nitrogen application beyond 270 kg/hm² did not further improve cooked rice quality and, in some cases, adversely affected sensory attributes. Maintaining nitrogen input within 210–270 kg/hm² may, therefore, help avoid unnecessary fertilizer expenditure and potentially reduce nitrogen losses associated with excessive fertilization. Regarding milling, DOM10 produced the highest sensory score at 270 kg/hm², whereas DOM5 retained more nutrients and aroma‐active compounds with less grain material removal. Milling degree should, therefore, be selected according to product targets to balance sensory quality, nutrient and aroma retention, and material recovery. Because production costs, milling energy use, nitrogen losses, and emissions were not directly measured, these implications should be regarded as preliminary. Further assessments of costs, benefits, and environmental impacts are needed to evaluate the sustainability of combined nitrogen and milling strategies.

4. Conclusion

Nitrogen application and milling degree both influenced cooked rice quality in Yanjing 939 by affecting grain composition, pasting and cooking behavior, texture, sensory properties, and volatile characteristics. Nitrogen application within the range of 210–270 kg/hm² generally resulted in favorable textural properties, odor, palatability, and overall sensory performance, whereas taste value, water absorption, and swelling decreased with increasing nitrogen application. Milling improved cooking uniformity and sensory quality but decreased the number and abundance of volatile compounds. DOM5 provided a more balanced combination of nutrient retention, sensory quality, and volatile composition, whereas DOM10 produced the highest overall sensory score at 270 kg/hm² but was accompanied by a broader reduction in volatile compounds. These findings demonstrate that the effects of nitrogen application on cooked rice quality were modulated by milling degree. Several limitations should be acknowledged. This study used a single japonica cultivar from one growing season, and the proposed mechanisms were inferred mainly from physicochemical analysis and compositional data rather than from direct structural observations. Accordingly, the findings are specific to Yanjing 939 under the tested conditions and require further validation across a broader range of cultivars, growing environments, and processing scales. Future studies should also incorporate direct microstructural observations and characterization of starch−protein interactions at the molecular level.

Author Contributions

Zhuojia Lv: Conceptualization, data curation, methodology, formal analysis, software, Writing – original draft. Guodong Ye: Investigation, methodology, formal analysis, writing – review and editing. Wenhui Zhang: Investigation, formal analysis. Sixuan Li: Investigation, formal analysis. Feng Liang: Investigation, formal analysis. Longfei Zhang: Investigation, formal analysis. Min Zhang: Conceptualization, supervision, resources, funding acquisition, validation, writing – review and editing.

Conflicts of 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.

Supporting information

Supplementary Material: jfds71411‐sup‐0001‐SuppMat.docx