Black bean protein isolate-gum Arabic coacervated citronellal microcapsules: Preparation, characterization, antibacterial activities, and application in strawberry preservation Authors: Haifang Mao, Congle Wang, Zijia Zhang, Yanzhen Long, Linlin Zhao, Jie Chen, Liwei Fu, Liqing Qiu Published: 2026 ABSTRACT AbstractBlack bean protein isolate-gum Arabic (BBPI-GA) coacervates were employed for citronellal encapsulation, and the feasibility of citronellal-loaded microcapsules for strawberry preservation was investigated. The optimal BBPI-to-GA ratio and pH value for fabricating BBPI-GA coacervates were 4:1 and 4.1, with hydrogen bonding and electrostatic interactions serving as two primary drivers for complex coacervation. The characteristic core-shell morphology of the prepared wet microcapsules and the intact surface of the freeze-dried microcapsules indicated the successful encapsulation of citronellal in coacervates, with a high encapsulation efficiency (EE) of 82.95 ± 0.37%. The prepared citronellal microcapsules displayed a uniform particle size distribution, with an average size of 25.63 ± 2.69 μm. The encapsulated citronellal displayed improved thermal and storage stability, sustained release property, and stronger antimicrobial activity than free citronellal. Furthermore, compared with free citronellal, microencapsulated citronellal more effectively inhibited spoilage microorganisms, mitigated weight loss, and retarded firmness degradation in stored strawberries, verifying its great potential for fruit preservation.Keywords: Essential oil; Microcapsules; Fruits, complex coacervation; Preservation Introduction Strawberries (Fragaria×ananassa ‘Benihoppe’) enjoy widespread popularity on a global scale owing to their vibrant red appearance, pleasant taste, soft and juicy texture, abundant bioactive and nutritional substances (e.g., polyphenols, vitamin C, tannins, and folate), as well as prominent antioxidant characteristics (Huang et al., 2025). Nevertheless, significant quantities of strawberries are discarded each year as a result of spoilage, mechanical damage, and fungal infections, which consequently give rise to considerable food waste and economic losses in the food field (Chen et al., 2025). For the purpose of addressing these issues, certain traditional preservatives (e.g., sorbic acid, phenolic compounds, and benzoic acid) have presented efficacy in maintaining food quality during storage periods. Nevertheless, the potential toxicity and the adverse effects of these preservatives are not negligible, and the development of efficient, safer alternatives are therefore urgently required (Chen et al., 2025). Given this context, the exploration and development of preservatives featuring natural sources, high safety, and excellent efficacy have emerged as an imperative and pivotal research priority in food-related fields (Chen et al., 2026; Qiao et al., 2026). Owing to their inherent antifungal and antitoxigenic activities, various plant products have long been employed for microbial inactivation in food preservation (Varma & Dubey, 2001). Notably, essential oils represent one of the most valuable natural resources for the development of safe antifungal agents (Tian et al., 2012). According to the literature, the low-molecular-weight and highly lipophilic compounds in essential oils can penetrate cell membranes to induce biological responses, thereby disrupting bacterial membranes and prolonging the shelf life of food products (Tian et al., 2012). Citronellal not only exhibits multiple desirable attributes, including natural sources, high safety, and non-toxicity, but also demonstrate potent antibacterial activity, thereby effectively suppressing the growth of spoilage microorganisms in food (Charret et al., 2024). On the contrary, the hydrophobicity, volatility, and instability of citronellal limited its application in the field of food preservation (Charret et al., 2024). To solve these problems, various encapsulation techniques (e.g., emulsions, microcapsules, liposomes) have been developed (Venancio et al., 2025). Microencapsulation is a widely adopted encapsulation technique that uses polymer-based materials (defined as wall materials) to entrap gaseous, liquid or solid components (referred to as core materials) (Yao et al., 2020). It achieves dual functions of protecting active ingredients and regulating controlled release, thereby maintaining the biological activity and extending the functional performance of core materials (Yao et al., 2020). For example, pummelo essential oil was encapsulated with glutenin as the wall material to improve its stability and broaden its application in white mushroom preservation (Zhang, Jiang, Tang, & Li, 2024). Experimental results indicated that the prepared pummelo essential oil microcapsules possessed excellent potential for inhibiting cap opening, viscosity increase and off-flavor generation of white mushrooms during storage (Zhang et al., 2024). Similarly, compared to the active core compounds unencapsulated, linalool microcapsules (Xu et al., 2026) and cinnamon essential oil microcapsules (Hu et al., 2024) exhibited superior stability and sustained release properties, thus effectively extending the shelf life of sweet cherries and strawberries during storage. Therefore, we speculate that the encapsulation of citronellal can not only mitigate its volatilization, but also enhance its stability and hydrophilicity, thus expanding its application prospects in the food field (Gan et al., 2022). Among all encapsulation techniques, complex coacervation serves as a highly effective approach for encapsulating hydrophobic active ingredients. This phenomenon refers to electrostatic interaction-dominated liquid-liquid phase separation driven by the interaction between two biopolymers with opposing charges, and proteins and polysaccharides are the two most frequently used substrates in relevant applications (Dakhili et al., 2026). Compared to other encapsulation methods, complex coacervation boasts distinct merits, including easy operation, superior encapsulation efficiency (EE), strengthened stability of core materials, and desirable controlled-release characteristics (Dakhili et al., 2026). Therefore, complex coacervation has been widely utilized for encapsulating a variety of hydrophobic active ingredients. For example, Jamshidian and Rafe (2024) successfully adopted wheat germ protein-high methoxy pectin complex coacervates as shell materials to encapsulate d-limonene, thereby enhancing its stability and controlled release properties. In addition, Hasanvand and Rafe (2019) also realized the encapsulation of vanillin via β-cyclodextrin inclusion complex coacervates. The resulting microcapsules exhibited excellent thermostability and sustained release profiles, showing great potential for application in dairy and pastry products (Jamshidian & Rafe, 2024). As reported, amphoteric proteins (e.g., gelatin, casein and, soy protein) and anionic polysaccharides (gum Arabic (GA), sodium alginate, and pectin) are the most commonly utilized encapsulating materials in complex coacervation process (Delaporte, Paraskevopoulou, Grisel, & Gore, 2025; Du et al., 2025). Studies have confirmed that wall material types, properties, and compositions markedly affect the encapsulation performance, release profile, and stability characteristics of the fabricated microcapsule samples (Chen, Feng, et al., 2025). The exploitation of novel wall materials featuring high EE and low economic cost remains a research hotspot in the field of encapsulation. GA, a naturally derived anionic polysaccharide with inherent amphiphilic properties, has been extensively utilized in the food sector as a wall material and emulsifier due to its advantages such as low viscosity, high solubility, and high EE (Vinterbladh, Soussi, Forsman, Bouhallab, & Lund, 2025). Black bean protein isolate (BBPI), an amphoteric protein sourced from black beans, is capable of interacting electrostatically with polysaccharides (e.g., GA) to form protein-polysaccharide coacervates (Zanghelini, Benvenutti, Laurindo, Zielinski, & Carciofi, 2025). BBPI is also characterized by various advantages, including wide availability, low cost, non-toxicity, biocompatibility, excellent emulsifying properties, a balanced essential amino acid profile, and favorable bioactivities (Song et al., 2024). Furthermore, as a plant-derived protein, BBPI has a favorable content and well-balanced profile of essential amino acids, which endows it with relatively high digestibility and utilization efficiency, with a digestible indispensable amino acid score (DIAAS) of approximately 80 in humans (Mali & Kumar, 2025; Sousa et al., 2023; Yang, Dias, Pham, Barile, & de Moura Bell, 2024). These features make BBPI-GA coacervates a highly ideal wall material for encapsulating active compounds. Nevertheless, no reports have illustrated the complexation between BBPI and GA thus far. The interaction mechanism underlying BBPI-GA complexation, together with the stability, release dynamics, and antibacterial activity of the corresponding microcapsule samples, and their practical application in fruit preservation remains to be thoroughly explored. Considering the protective effects and controlled release properties of bioactive compounds afforded by protein-polysaccharide coacervates as illustrated above, microcapsules for strawberry preservation were fabricated using BBPI-GA coacervates as the wall material and citronellal as the core material. This study aimed to improve the stability and prolong the functional performance of bioactive compounds through encapsulation by complex coacervation, thereby achieving excellent preservation effects on strawberries. Accordingly, the effects of protein-to-polysaccharide ratio and pH on the formation of BBPI-GA coacervates were investigated, and the formation mechanism of the complex was also elucidated. Furthermore, the physicochemical and functional properties of citronellal-loaded microcapsules were characterized. Finally, the prepared microcapsules were applied to strawberry preservation, which provides a novel technical strategy for improving the postharvest storage quality of fruits. Materials and methods Raw materials Strawberries (Fragaria×ananassa ‘Benihoppe’) were sourced from a local ALDI supermarket in Shanghai, China. Peeled black beans were purchased from Gannan County Jilang Rice Industry Co., Ltd. (Heilongjiang, China). GA and citronellal (99% purity) were obtained from Titan Technology Co., Ltd. (Shanghai, China). The remaining chemicals and reagents utilized in the article were of analytical grade and obtained from Shanghai Macklin Biochemical Technology Co., Ltd. (Shanghai, China). Preparation of black bean protein isolate BBPI was obtained using the methodology developed by Zanghelini et al. (2025) with appropriate alterations. Firstly, the peeled black beans were ground for 10 min, followed by passing through a sieve of 60-mesh size. Subsequently, the prepared black bean powder was defatted repetitively for three cycles with a mixed solvent system (ethanol:n-hexane = 1:9, v/v) at a 3:1 (w/v) solvent-to-solid ratio. Afterwards, the powder after defatting was blended with ultrapure water to achieve a 10% (w/v) suspension, whose pH was increased to 8.0 using NaOH (0.1–1 M) before being stirred continuously at 25 °C for 120 min. The resultant mixture was subsequently filtered and centrifuged for 20 min at 10000 rpm, with the corresponding supernatant harvested and reserved for subsequent experimental procedures. The supernatant was then treated with an acetic acid solution to bring its pH value to 4.5, and the obtained mixture was incubated statically for a duration of 60 min. The obtained precipitate was harvested via centrifugation at 10000 rpm for 15 min, followed by resuspension in deionized water at pH 7.0. The supernatant collected following centrifugation at 5000 rpm for 10 min was subjected to lyophilization and then preserved at −20 °C for subsequent experiments. The protein purity of BBPI was measured as 87.34% by the Kjeldahl method (Yanu & Jakmunee, 2017). Preparation of BBPI and GA stock solutions Separate solutions of BBPI and GA were prepared by dispersing each component in deionized water at 1% (w/w). The solution of BBPI was increased to pH 8.0 and subjected to stirring at 200 rpm for half an hour. Afterwards, the mixed solution was acidified to 7.0 and then magnetically agitated at 200 rpm for 12 h below 10 °C to achieve complete hydration. Determination of turbidity Turbidity measurements were conducted on BBPI-GA mixed solutions with varying mass ratios (1:2, 1:1, 2:1, 4:1, 6:1, 8:1, w/w) at a wavelength of 600 nm, with solutions' pH adjusted across a range of 2.5–5.5. To ensure that the absorbance of the dispersion at 600 nm is within an appropriate range, the BBPI-GA mixed solutions were subjected to dilution to a final concentration of 0.05% (w/w). Determination of zeta potential The zeta potential of BBPI solutions, GA solutions, and their mixtures (0.05%, w/w) at different BBPI-to-GA ratios (1:2, 1:1, 2:1, 4:1, 6:1, and 8:1, w/w) was tested using a Zetasizer Nano ZS instrument (Malvern Panalytical, UK) as a function of pH (2.5–5.5). All experimental procedures were implemented in three parallel replicates, and the ambient temperature was kept constant at 25 °C throughout the trials. Determination of coacervate yield The yield BBPI-GA coacervates across varying BBPI-to-GA ratios (1:2, 1:1, 2:1, 4:1, 6:1, and 8:1, w/w) was determined as described in earlier work (Qiu, Zhang, Adhikari, & Chang, 2022). In short, for 1% (w/w) BBPI-GA solutions with distinct protein-to-gum ratios, the pH was adjusted to their corresponding isoelectric points, as optimized in Section 2.5. After stirring at 4 °C for half an hour, the precipitate was first obtained via centrifugation (9000 ×g, 15 min), then lyophilized, and finally accurately weighed. The coacervate yield (YC) was determined via the equation listed below: where, WC, WBBPI, and WGA represent the weight of BBPI-GA coacervates, BBPI, and GA, respectively. Preparation of BBPI-GA coacervates and citronellal microcapsules The preparation of the BBPI-GA coacervates was carried out in line with the approach proposed by Chen, Zhang, Mujumdar, and Wang (2023), with minor alterations. Specifically, the two solutions (1%, w/w BBPI and 1%, w/w GA) were mixed thoroughly at a BBPI-to-GA weight ratio of 4:1. Afterwards, the pH of the mixed solutions was adjusted downward from 7.0 to 4.1 under constant stirring (200 rpm). Afterwards, the mixtures were agitated (100 rpm) at 4 ± 1 °C for half an hour to promote the generation of BBPI-GA coacervates. The citronellal microcapsules were fabricated by adopting the methodology proposed by Silva, Sato, and Perrechil (2025), with minor adjustments. Initially, citronellal was dispersed in the BBPI-GA solution at a mass ratio of 1:2 (citronellal:BBPI-GA, w/w), which was optimized based on our pre-experiments. Then, the mixed solution was homogenized (15,000 rpm, 5 min). The identical protocol utilized for the preparation of BBPI-GA coacervates earlier was adopted for the production of citronellal microcapsules. BBPI-GA coacervates and citronellal microcapsules obtained via the aforementioned procedure were subjected to freeze-drying and then stored in an aluminum foil bag at 4 ± 1 °C for subsequent experiments. Intrinsic fluorescence analysis A fluorescence spectrometer (F-7000, Hitachi High-Tech Corporation, Japan) was utilized to investigate the intrinsic fluorescence intensity of BBPI, GA, and their BBPI-GA coacervates. Emission spectra within the wavelength range of 310–420 nm were acquired at 25 °C using 295 nm excitation and a 5 nm excitation slit width (Muhoza, Harindintwali, Qi, Zhang, & Li, 2023). Fourier transform infrared spectroscopy Tablets containing testing samples and potassium bromide (KBr) at 1:100 (w/w) mixing ratios were fabricated by a tablet press (HYP-15, Tianjin Hench Technology Co., Ltd., China) prior to analysis. The FTIR spectrum of BBPI, GA, and BBPI-GA coacervates were acquired via an FTIR spectrometer (IS10, Nicolet Company, USA). All tests were conducted over a wavelength range of 400–4000 cm−1 and a scan number of 64 (Chen, Zhang, Wang, Mujumdar, & Deng, 2024). Particle size and morphology analysis The distribution of particle size of freshly prepared complex coacervates and citronellal microcapsule samples (section 2.7) were measured by a laser particle size analyzer (Mastersizer3000, Microtrac Inc., USA) using deionized water (pH 7.0, 25 °C) as the dispersion medium (Yu et al., 2025). The microstructure of the BBPI-GA complex coacervates and citronellal microcapsules before freeze drying was observed via an optical microscope (OM) (primo star, ZEISS, Germany) at 100× magnification (Wang et al., 2025). The microstructure of lyophilized citronellal microcapsule samples was captured via a scanning electron microscope (SEM) (MAIA3, TESCAN, Czech Republic) at 600× and 10 k × magnifications (Cui, Chen, Chen, Li, & Jiang, 2025). Encapsulation efficiency and payload For the determination of surface citronellal content in the microcapsules, 0.05 g of freeze-dried microcapsules was initially distributed in 30 mL of absolute ethanol, followed by repeated gentle shaking. The mixture was then subjected to filtration to prepare the ethanol solution containing surface citronellal. Total citronellal content in the microcapsules was quantified by subjecting the ethanol suspension (0.05 g sample in 30 mL) to three 20-min cycles of ultrasonic treatment at 25 °C. Following filtration, the supernatants were tested for absorbance at 202 nm. These absorbance data were further applied to calculate the EE and payload (PL) of citronellal microcapsule samples using the following formulas: where, WT and WS represent the mass of the total and surface citronellal, respectively; WM and WW represent the mass of the microcapsules and shell materials, respectively. Thermal stability Citronellal, BBPI-GA coacervates, and citronellal microcapsule samples were subjected to thermal stability analysis via a thermogravimetric analyzer (TGA2, Mettler Toledo Instrument Co., Ltd., USA). Specifically, approximately 3 mg of each freeze-dried sample was accurately weighed into an alumina crucible. Afterwards, the alumina crucible was subjected to programmed heating, with the temperature increasing from 25 to 500 °C at a constant rate of 10 °C/min under a nitrogen flow of 50 mL/min as part of the programmed thermal treatment (Qiu, Zhang, Adhikari, & Chang, 2022). Release profiles in food simulators Deionized water, 10% and 50% ethanol, as well as 3% acetic acid, representing four distinct food simulants, were utilized to characterize the release profiles of citronellal corresponding to aqueous, alcoholic/alkaline, oily, and acidic environments. Specifically, 20 mg of citronellal microcapsules was mixed with 50 mL of each type of food simulant, followed by stirring at 100 rpm throughout the entire testing period. At predetermined time intervals, the supernatant was assayed at 202 nm. Storage stability analysis To measure the stability of citronellal and its microcapsule formulations during storage, the measurements described by Chen et al. (2023) were employed with suitable adjustments tailored to the specific experimental needs of the samples. Briefly, free citronellal and the microcapsules were evenly placed in glass dishes and stored at 25 °C and 35 °C, respectively. The citronellal retention rate for all samples was tested every 5 days throughout the 30-day storage period. The citronellal content was determined following the procedure outlined in Section 2.11, and the citronellal retention rate was determined using the formula listed below: where, WO and WS are the citronellal contents at the initial storage period (day 0) and different storage intervals, respectively. Determination of antioxidant capacity Pre-weighed samples of citronellal and citronellal microcapsules were dispersed in identical volumes of anhydrous ethanol to maintain consistent citronellal concentrations in all samples. At pre-determined time periods, 3.0 mL of each sample was withdrawn. To evaluate the scavenging efficiencies against DPPH and ABTS radicals, the supernatants were collected via shaking and centrifugation for subsequent analysis (Wang et al., 2025). DPPH free radical scavenging rates The DPPH free radical scavenging behavior of different samples was assessed refereed to the protocol described by Zhang et al. (2026), with results expressed as DPPH radical inhibition percentage. ABTS free radical scavenging rates The ABTS free radical scavenging capacity of different samples was investigated following the protocol described by Qiu, Zhang, and Chang (2023), with results expressed as ABTS radical inhibition percentage. Determination of antibacterial activity Inhibition rate Three typical bacterial strains: Escherichia coli (E. coli), Staphylococcus aureus (S. aureus), and Bacillus subtilis (‌B. subtilis) were selected to determine the antibacterial effect of different samples (Noptana, McClements, McLandsborough, Onsaard, & Onsaard, 2026). Before testing, three strains were inoculated in an oscillating incubator (at 37 ± 1 °C) for 12 h. Subsequently, the concentration of bacterial strains was diluted to the proper range (<300 cells/mL) using sterile Luria-Bertani (LB) broth. Bacterial strain suspension of l mL was incorporated into the nutrient agar, followed by uniform smearing with a sterile coating rod. Different testing samples (citronellal concentration of 50 mg) were evenly placed on the covers of Petri dishes. Each Petri dish was sealed using sterile tape, placed in an inverted position, and incubated at 37 ± 1 °C for 12 h. Subsequent colony enumeration was performed, with the resulting data expressed in terms of inhibition rate. Microscopic observation of bacteria The WH-RE sample was inoculated with bacterial suspensions of E. coli, S. aureus, and B. subtilis, and cultured at for 24 h 37 °C. Following incubation, the bacterial precipitates derived from different strains were individually collected via centrifugation (8000 rpm, 10 min) and rinsed with PBS (pH 7.2) to get rid of residual culture medium. Afterwards, the bacterial pellet was subjected to fixation in glutaraldehyde solution (2.5%, v/v) at 4 °C for a 12-h duration to preserve cellular morphology. After washing using PBS, the precipitate collected was dried using gradient ethanol dehydration (30%, 50%, 70%, 80%, 90%, and 100%). Finally, the microscopic morphology of the strains after dehydration was determined utilizing a SEM (ZEISS Sigma 360, ZEISS, Germany). Application to strawberry preservation Strawberries with uniform color, ripeness, and size, free from mechanical damage and microbial contamination, were randomly assigned to the control, citronellal treatment, and citronellal microcapsule treatment groups. The strawberry samples (30 ± 0.5 g) were stored in sterilized square plastic airtight containers (18.7 cm × 12.5 cm × 5 cm). Then, equal amounts of citronellal and microcapsules (3% (w/w) of the total strawberry mass) were uniformly dispersed inside the glass dishes (30 mm in diameter) and left uncovered. Afterwards, the glass dishes were sealed with breathable non-woven fabric and placed into each container to evaluate the preservation effect of different treatments. All samples were subjected to an 8-day storage period under ambient temperature conditions (25 °C). The quality profiles, including firmness (Chen, Feng, et al., 2025), weight loss (Li et al., 2025), and the total number of microbial colonies (Nong et al., 2025), were measured for strawberries at a two-day interval. Statistical analysis All tests were carried out three times, except specified otherwise. Data are reported as mean ± SD and subjected to one-way analysis of variance (ANOVA). Significance difference between any two mean values was calculated by Duncan's multiple range tests at a 95% confidence level (p < 0.05). SPSS Statistics 19.0 software package (IBM SPSS, Chicago, USA) was utilized to perform these analyses. Results and discussion Optimal complex coacervation conditions for BBPI and GA Zeta potential It can be observed from Fig. 1A that the zeta potentials of BBPI and GA drastically declined as the pH increased within the range of 2.5 to 5.5. BBPI attained its corresponding isoelectric point (pI) at approximately pH 4.8, with the molecule carrying a negative charge at pH values above this threshold. In addition, GA remained negatively charged throughout the entire testing pH range applied in the experiments (2.5–5.5). As evidenced by these findings, the pH values below the pI of BBPI (pH<4.8) facilitated the formation of BBPI-GA complex coacervates, wherein the two polymers possessed opposite charges. Zeta potential changes of individual BBPI and GA (A); Zeta potential changes of BBPI-GA mixed solutions with varying BBPI-to-GA ratios (B); Turbidity of BBPI-GA systems with varying BBPI-to-GA ratios at different pH values (C); Complex coacervate yield of BBPI-GA systems with varying BBPI-to-GA ratios at their optimal pH values (D). Note: Different letters in the columns indicate a significant difference (p<0.05). Zeta potentials of BBPI-GA samples across varying BBPI-to-GA ratios are presented in Fig. 1B. As illustrated in Fig. 1B, these systems reached zero zeta potential (pI) over the pH range of 2.5–4.5, indicating that charge neutralization was achieved within this pH interval. Within this specified pH range, an efficient complex coacervation among BBPI and GA was predicted to be induced, given that the strongest electrostatic attractive forces typically arise when the total net charges of the two biopolymers are neutralized (Chen et al., 2023). In addition, a gradual shift in the pI of the BBPI-GA system toward higher pH values was observed with the increase in the BBPI-to-GA ratio. This might be attributed to the higher BBPI-to-GA ratio, which reduced the negative charge in the mixed system and thereby ensured sufficient positive charges even at a pH near the system's pI, owing to the elevated BBPI concentrations (Qiu, Zhang, Chitrakar, Adhikari, & Yang, 2022). Turbidity and coacervates yield From Fig. 1C, the turbidity of the BBPI-GA system at varying BBPI-to-GA ratios (1:2, 1:1, 2:1, 4:1, 6:1, and 8:1, w/w) presented a tendency of initial increase followed by a decrease when pH was reduced from 5.5 to 2.5. This phenomenon suggested that BBPI-GA complex coacervation could be categorized as a classic pH-induced process. Systems with a BBPI-to-GA ratio of 1:2, 1:1, 2:1, 4:1, 6:1 and 8:1 (w/w) reached their highest turbidity at around pH 3.0, 3.3, 3.8, 4.1, 4.2, and 4.2, respectively, which were almost the same as their pI values (Fig. 1B). In addition, the highest turbidity of the BBPI-GA complex system was observed at a BBPI-to-GA weight ratio of 4:1 (w/w) and a pH of 4.1. Nevertheless, decreased turbidity was detected in the BBPI-GA system at BBPI-to-GA ratios below or above 4:1 (w/w), demonstrating that the polymer blending ratio exerts a crucial influence on the generation of protein-polysaccharide coacervates. From one perspective, the lower turbidity of the BBPI-GA system at relatively low BBPI-to-GA ratios (1:2, 1:1, and 2:1, w/w) could be attributed to the enhanced steric repulsion triggered by redundant polysaccharide molecules. From another perspective, the lower turbidity of BBPI-GA system at relatively high BBPI-to-GA ratios (6:1 and 8:1, w/w) was probably attributed to excess BBPI molecules, which interacted with each other and thus restricted the complex coacervation between BBPI and GA (Chen et al., 2023). Fig. 1D displays the yield of BBPI-GA coacervates with different BBPI-to-GA ratios at their optimal pH. As displayed in Fig. 1D, the yield of BBPI-GA coacervates exhibited a mass ratio-dependent trend. Specifically, the maximum yield of BBPI-GA coacervates (89.83 ± 0.24%) was observed at a BBPI-to-GA mass ratio of 4:1 (w/w), while the yields of the other systems were lower, ranging from 37.00 ± 0.04% to 88.33 ± 0.01%. These observations aligned well with the results of the turbidity tests (Fig. 1C). Specifically, higher turbidity values in the system usually correspond to a higher production of complex coacervates during the coacervation reaction. Based on the turbidity data presented above, pH 4.1 and a BBPI-GA weight ratio of 4:1 were identified as the optimal parameters for fabricating BBPI-GA coacervate samples. Characterization of BBPI-GA coacervates Intrinsic fluorescence analysis Changes in the three-dimensional spatial structure of proteins can be identified through the detection of their intrinsic fluorescence intensity, which is predominantly caused by the fluorescence emissions of protein residues, including tryptophan (Trp) and tyrosine (Tyr) (Luan et al., 2025). In the present article, fluorescence spectroscopy was utilized to characterize the binding interactions of BBPI and GA in the complex coacervation process. The emission profiles of BBPI, GA, and their complex coacervates at 295 nm are displayed in Fig. 2A. GA did not exhibit obvious fluorescence absorption, which could be attributed to the lack of amino acid residues in its structure. In contrast, strong fluorescence absorption was observed in BBPI and BBPI-GA coacervates, which was ascribed to their elevated protein concentration. Moreover, a reduced fluorescence intensity was detected in BBPI-GA coacervates compared with BBPI, and a concomitant blue shift of λmax was also noted. Aggregation of BBPI-GA coacervates during complex coacervation may have triggered the relocation of amino acid residues from the protein's polar surface to its hydrophobic interior, resulting in this observed phenomenon. Moreover, the shielding effect of GA also contributed to the weaker fluorescence intensity of BBPI. Fluorescence emission (A) and FTIR spectra of BBPI, GA, and BBPI-GA coacervates. Fourier transform infrared spectroscopy analysis To verify the generation of BBPI-GA coacervates through molecular interactions, the FTIR spectra of BBPI, GA, and their coacervates were measured (Fig. 2B). From the spectrum, BBPI has multiple characteristic absorption peaks corresponding to its protein structural features. Stretching vibrations of the O—H and N—H groups within amino acid residues are responsible for the broad absorption peak at around 3305.00 cm−1 (Li et al., 2025). The peaks that appeared at 2961.58 cm−1 were identified as the signal generated by the stretching vibrations of aliphatic C—H groups (Chen, Zhang, Adhikari, & Wang, 2022). Moreover, the three characteristic peaks: amide I (1651.78 cm−1), amide II (1539.58 cm−1), and amide III (1237.86 cm−1) could be ascribed to the characteristic amide vibrations (Zhou et al., 2025). For GA, the broad absorption peak detected around 3423.62 cm−1 was determined to originate from the stretching vibrations of O—H groups (Qiu, Zhang, Adhikari, et al., 2022). The peak observed at 2926.18 cm−1 corresponded to the aliphatic C—H stretching vibration (Zhou et al., 2025). Furthermore, the peaks associated with the symmetric and asymmetric stretching vibrations of the carboxyl groups were found at 1384.40 and 1617.16 cm−1, respectively (Chen et al., 2023). The intense absorption signal situated at 1073.01 cm−1 verified the existence of GA's inherent polysaccharide backbone structure (Chen et al., 2023). The BBPI-GA coacervate spectrum preserved the majority of characteristic peaks of the two biopolymers but exhibited slight shifts in peak positions, suggesting intermolecular interactions between BBPI and GA (Tabak & Karakullukçu, 2025). As shown in Fig. 2B, the spectrum of BBPI-GA displayed combined characteristic features of BBPI and GA, with distinct peak shifts relative to pure BBPI and GA. Specifically, the carboxyl group-related characteristic peaks of GA at 1617.16 and 1384.40 cm−1 disappeared in the BBPI-GA coacervate spectrum. In addition, the peaks of BBPI at 1651.78 (amide I) and 1539.58 cm−1 (amide II) shifted after complexing with GA. These findings revealed the occurrence of electrostatic interaction between BBPI (amino groups) and GA (carboxylic groups) (Li, Wang, & Mei, 2021). Furthermore, the -OH stretching vibration peak of BBPI-GA coacervates (3416.56 cm−1) shifted to lower wavelength compared to that of GA (3423.62 cm−1), while the peak of amide II (1540.24 cm−1) in coacervates shifted to higher wavelength compared with that of BBPI (1539.58 cm−1), indicating the existence of hydrogen bonding interactions (Guo et al., 2022). Based on the above results, the complex coacervation between BBPI and GA involves not only electrostatic interaction, but also hydrogen bonding interaction. Micromorphology, particle size distribution, and encapsulation efficiency of citronellal microcapsules As illustrated in Fig. 3, the micromorphology of the BBPI-GA coacervates and citronellal microcapsules was obtained via OM and SEM. The wet BBPI-GA coacervates presented an irregular and compact microstructure owing to the electrostatic interaction and the subsequent agglomeration of BBPI and GA during the process of complex coacervation (Cui et al., 2025). As for the citronellal microcapsules (Fig. 3B), citronellal droplets were uniformly encapsulated in wall materials composed of BBPI and GA. In addition, negligible amounts of free citronellal were found in the system, demonstrating that the microcapsules possessed a satisfactory encapsulation capacity, with an EE of 82.95% ± 0.37% and a PL of 32.00 ± 2.83%. BBPI-GA coacervate particles had a unimodal normal distribution (average size: 24.31 ± 1.41 μm), and the microcapsules showed a similar average particle size of 25.63 ± 2.69 μm. This indicated that the loading of citronellal into BBPI-GA coacervates at a shell-to-citronellal ratio of 2:1, exerting little effect on the micromorphology of BBPI-GA coacervates. Additionally, the wet citronellal microcapsules were featured by a “multi-core” microstructure, confirming their better mechanical strength, storage stability, and release properties than “single-core” microcapsules (Timilsena, Akanbi, Khalid, Adhikari, & Barrow, 2019). The freeze-dried citronellal microcapsules showing a typical morphology consistent with that of complex coacervated microcapsules. The surface of the prepared microcapsule samples exhibited a continuous structure without obvious holes or fractures (Fig. 3D), indicating that citronellal was effectively encapsulated within the wall materials after freeze-drying, and the wall materials were capable of shielding the core essential oil from environmental perturbations. Optical microscopy images of BBPI-GA coacervates (A) and citronellal microcapsules (B) before freeze-drying, and scanning electron microscopy (SEM) images of freeze-dried citronellal microcapsules at different magnifications (C, D). Release profiles Fig. 4 displayed the release behaviors of citronellal from microcapsules in four distinct food simulates. It can be observed from Fig. 4A that the citronellal release rate was relatively fast within the initial 15 min, after which a stable release phase continued until a plateau was reached, indicating that citronellal release was effectively regulated in all food simulants. The rapid release observed in the initial minutes may be attributed to the citronellal present on the microcapsule surface and the free citronellal within the fabricated microcapsules (Chen et al., 2025). The subsequent slow-release stage may be ascribed to the gradual diffusion of citronellal through the pore structure of the microcapsule wall and from the microcapsule interior, resulting from microcapsule swelling and rupture (Chen et al., 2025). As shown in Fig. 4, the maximum citronellal release was detected in 50% ethanol, followed by that in 10% ethanol and distilled water, while the lowest release was observed in 3% acetic acid. These results demonstrated that citronellal released more rapidly under oily (50% ethanol), alcoholic/alkaline (10% ethanol), and aqueous conditions (distilled water) compared to an acidic environment (3% acetic acid). The cumulative release profiles of citronellal from microcapsules in different food simulants (A), and the fitting curves of encapsulated citronellal based on the Zero-order (B), First-order (C), Weibull (D), Ritger-Peppas (E), and Higuchi (F) models. For the purpose of gaining deeper insights into the release characteristics and inherent release mechanism of citronellal, the release curves of citronellal were fitted with the five typical kinetic models (Zero-order, First-order, Weibull, Ritger-Peppas, and Higuchi models). As presented in Fig. 4, the release curves of citronellal in four food simulants fitted well with the Weibull model, with a high correlation coefficient (R2) of 0.99. As reported in previous studies, the β parameter of the Weibull model serves as the characteristic indicator for defining the release mechanism of the core compounds. Typically, in the Weibull model, a β value≤0.75 indicates Fickian diffusion-dominated release mechanism, while a β value of 0.75 < β ≤ 1 denotes non-Fickian diffusion release mechanism, which involves both diffusion-driven and swelling-controlled release behaviors (Li et al., 2025; Papadopoulou, Kosmidis, Vlachou, & Macheras, 2006). As illustrated in Fig. 4B, the β values ranged from 0.33 and 0.61 indicated that the release of citronellal from microcapsules in all simulated media (50% ethanol, 10% ethanol, distilled water, and 3% acetic acid) followed Fickian diffusion. Therefore, the swelling of the microcapsules can be neglected at this process. This result may be attributed to the limited swelling of the wall materials in these media and the slow diffusion of citronellal. In summary, the BBPI-GA coacervates slowed down the release rate of citronellal in different food simulants. Similarly, the release mechanism of rose essential oil from perilla protein isolate‑sodium alginate matrices in the four food-simulated media was also identified as Fickian diffusion by Weibull model in our previous study (Qiu, Zhang, Adhikari, & Chang, 2023). Likewise, the release mechanism of Sichuan pepper essential oil from soybean protein isolate-Sichuan pepper seed soluble dietary fiber coacervates in distilled water, 3% acetic acid solution and 10% ethanol solution was identified as Fickian diffusion, while that in 50% ethanol solution corresponded to non-Fickian diffusion (Chen et al., 2022). The distinct release mechanisms observed can be attributed to differences in the properties of the wall materials in these solutions. Thermal stability The stability of free citronellal, BBPI-GA coacervates, and citronellal microcapsules at high temperatures was determined, with the corresponding results summarized in Fig. 5. According to the thermogravimetric (TGA) and differential thermogravimetric (DTG) curves displayed in Fig. 5A and Fig. 5B, free citronellal showed a mass loss of 14.34% at temperatures below 100 °C, which could be caused by the volatilization of citronellal and the evaporation of residual water. A marked weight loss (75.16%) was observed for free citronellal when the heating temperature rose from 100 to 300 °C, which was attributed to the substantial volatilization and degradation of citronellal (boiling point: 205–208 °C) (Kanza, Silvianti, Ramadhan, Cahayo, & Sukardi., 2021). As for BBPI-GA coacervates, a small weight loss (10.64%) was observed before 100 °C, likely caused by bound water evaporation. A mass loss of 20.95% was observed in BBPI-GA complex coacervates between 100 and 300 °C, which might be caused by the decomposition of BBPI, GA, or both components. A dramatic weight loss of around 53.77% was observed with the temperature rising from 300 to 500 °C indicated the severe degradation of the shell materials. In the case of citronellal microcapsules, the moisture evaporation and oil volatilization (15.00%) occurred at a temperature below 100 °C. Afterwards, a weight loss of 25.44% was determined with a temperature ranging from 100 to 300 °C, suggesting the decomposition of BBPI-GA walls and the oil volatilization/degradation. With the temperature increasing from 300 to 500 °C, the complete degradation of citronellal microcapsule samples occurred, and a weight loss of 33.82% was noticed. This part of the mass loss might be ascribed to the decomposition of shell materials, which left citronellal unprotected and thus prone to volatilization and degradation. The obtained results indicated that the stability of citronellal at high temperatures was significantly improved when BBPI-GA coacervates were employed as the shell matrix for microencapsulation. Stability of citronellal microcapsules. TGA (A) and DTG (B) curves of citronellal, BBPI-GA coacervates, and citronellal microcapsules; Citronellal retention rate during storage at 25 °C (A) and 35 °C (B). Storage stability The stability of free citronellal and citronellal-loaded microcapsules during storage at different temperatures (25 and 35 °C) is presented in Fig. 5. As displayed in this figure, the citronellal retention rate continued to decline as the storage duration increased. Meanwhile, the retention rate of free citronellal at the end of the storage period was 9.16% at 25 °C and 8.57% at 35 °C, respectively. These findings illustrated that storage led to a significant loss of citronellal, with higher temperatures resulting in greater losses. The occurrence of this phenomenon was likely ascribed to the combined effects of citronellal volatilization, oxidation, and degradation (Chen et al., 2023). Compared with free citronellal, the citronellal microcapsules exhibited considerably higher retention rates (25.73% at 25 °C and 24.51% at 35 °C), which suggests that BBPI-GA coacervates could significantly retard the volatilization and degradation of citronellal, thus protecting it from environmental stresses at various temperatures (25 °C and 35 °C). Antioxidant capacity In this article, the antioxidant performance of both free and encapsulated citronellal was tested by DPPH and ABTS radical scavenging experiments. As displayed in Fig. 6, free citronellal and encapsulated citronellal demonstrated obvious antioxidant capacity when the citronellal concentration ranged from 1.5 to 9 mg/mL. In addition, the free radical (DPPH and ABTS) scavenging rates of free and encapsulated citronellal increased with citronellal concentration (Fig. 6), proving their concentration-dependent antioxidant capacity. Particularly, the scavenging rate of DPPH and ABTS free radicals were 81.78% and 75.79% at 9 mg/mL citronellal, respectively. The free radical scavenging ability of encapsulated citronellal was comparable to that observed for free citronellal. Similarly, Chen et al. (2022) found that the coacervate shell which is composed of soluble dietary fiber from Sichuan pepper seed and soybean protein isolate had almost no effect on the antioxidant ability of the encapsulated oil. On the contrary, Huang et al. (2025) demonstrated that encapsulating resveratrol with whey protein and flaxseed gum could effectively enhance its antioxidant capacity, which was caused by the protective behavior of the composite. These results might be caused by the form of active substances, as microcapsules improved their hydrophilicity and dispersion, thus leading to the higher antioxidant activity of the encapsulated citronellal (Timilsena et al., 2019). On the contrary, the antioxidant capacity in this study was determined using extracted citronellal in its free form rather than in microcapsule form, which accounted for the observed differences in the microcapsule-related results. DPPH (A) and ABTS (B) radical scavenging activity of free and encapsulated citronellal at different concentrations. Note: Different letters in the columns with the same pattern indicate significant differences (p<0.05). Antibacterial activity The inhibitory effects of both free citronellal and citronellal microcapsules on three bacterial strains, namely, E. coli, S. aureus, and B. subtilis were investigated and the outcomes are illustrated in Fig. 7. As displayed in Fig. 7, citronellal and microcapsules exhibited effective antibacterial capacity against these three strains. In comparison to free citronellal, the encapsulated citronellal exhibited higher antibacterial activity against these three bacterial strains. This might be caused by the sustained release profile of citronellal after encapsulation. In addition, all samples (free citronellal and citronellal microcapsules) showed a higher inhibition rate against S. aureus and B. subtilis than that against E. coli. These results might be resulted from the structural differences in the cellular components between these strains (Qiu, Zhang, Chitrakar, et al., 2022). Gram-positive bacteria are characterized by a single outer peptidoglycan layer with only small amounts of protein, while Gram-negative bacteria exhibit a more complex cellular architecture, encompassing lipid-containing peptidoglycan, supplementary proteins, and a phospholipid membrane (Qiu, Zhang, Chitrakar, et al., 2022). Thus, it is evident that the loose external peptidoglycan layer of Gram-positive bacteria cannot function as an effective protective barrier against hydrophobic and macromolecular substances, resulting in decreased tolerance to essential oils (Wu et al., 2025). Inhibition rates of free citronellal and citronellal microcapsules against E. coli, S. aureus, and B. subtilis (A, B); and the microscopic morphology of untreated strains and citronellal microcapsule-treated strains (C). Note: Different letters in the columns with the same pattern indicate significant differences (p<0.05). To further explore the damaging effect of citronellal microcapsules on E. coli, S. aureus, and B. subtilis, the morphological characteristics of these three bacterial strains were tested via SEM. As displayed in Fig. 7, the untreated strains are full, intact, and without deformation damage. After adding citronellal microcapsules, significant damage to the cell membrane and obvious changes in the bacteria's shape were observed in these three bacterial strains. The incorporation of citronellal microcapsules induced the dissolution of bacterial cell membranes, thereby causing severe damage and the shrinkage of the tested bacterial strains. These observations could be explained by the fact that polycyclic structures in essential oils impair cell membranes, as these structures show good affinity for cell membrane macromolecules such as lipids, proteins, and nucleic acids (Qiu, Zhang, Chitrakar, et al., 2022). Application in strawberry preservation The effects of free citronellal and citronellal microcapsules on the freshness retention capacity of strawberry samples during storage were evaluated by determining the changes in appearance, firmness, weight loss, and total microbial colony counts. The appearance variations of the strawberry samples over an 8-day storage duration were presented in Fig. 8A. It can be seen from the figure that the mold appeared on control-group strawberries after 4 days of storage, whereas no visible spoilage microorganism growth was detected in the experimental group. Obvious decay signs were observed on strawberries treated with free citronellal on the 8th day of storage, while strawberries treated with citronellal microcapsules maintained an intact and fresh state, with no obvious spoilage or microbial proliferation detected throughout the storage duration. Changes of appearance (A), firmness (B), weight loss (C), and microbial colony count (D) of strawberries during storage at 25 °C. In terms of firmness, all strawberries presented a decreasing trend (Fig. 8B). Additionally, the firmness of the control group and strawberries treated with free citronellal and citronellal microcapsules decreased by approximately 56.70%, 49.39%, and 38.41%, respectively. As reported in the literature, the softening of the strawberry fruits is mainly caused by the breakdown of cell wall components and the action of spoilage microorganisms (Li et al., 2025). As a result, strawberries were susceptible to degradation by spoilage microorganisms, which disrupted cell walls and cellular matrices and ultimately induced cell collapse. In contrast, the growth of spoilage microorganisms was effectively inhibited by citronellal microcapsules, which in turn delayed the softening of strawberry fruits (Chen et al., 2025). The rate of weight loss is recognized as a key metric for estimating changes in fruits' external appearance and internal quality during storage. As presented in Fig. 8C, the control one displayed a high rate of weight loss, reaching approximately 29.24% after 8 days of storage. In contrast, the two experimental groups showed a relatively lower weight loss rate, which might be ascribed to the respiratory inhibitory effect of essential oil (Qiu, Zhang, Chitrakar, et al., 2022). Additionally, the strawberries treated with citronellal microcapsules exhibited the lowest weight loss rate among all groups, implying that the sustained release of citronellal from BBPI-GA coacervates during storage prolonged its action time. Thus, it can be concluded from these results that the weight loss of strawberries could be retarded by citronellal microcapsules through the inhibition of respiratory activity. Fig. 8D exhibits the change in the total number of microbial colonies during 8 days of storage. It is apparent from Fig. 8D that the total number of microbial colonies of strawberry samples increased as the storage duration increased. Citronellal effectively inhibited the increase in total microbial counts in strawberries during storage. Similarly, Wang et al. (2015) found that nerol could inhibit the mycelial growth and spore germination of A. niger, and alleviate black mold disease in postharvest grapes, thereby effectively suppressing grape spoilage during storage. As reported, the antibacterial effects of some lipophilic substances (e.g., nerol) are attributed to the disruption of cell membrane integrity, thereby altering membrane permeability and inducing cell death (Wang et al., 2015). This antibacterial mechanism is consistent with our findings in Section 3.8. The total microbial colony count of the control group strawberries exhibited the highest growth rate, rising from 2.02 to 7.15 Log CFU/g. The treated strawberries exhibited a significantly lower total microbial colony count, and the group treated with citronellal microcapsules presented the lowest total microbial colony count among all samples. This result implied the strongest antimicrobial activity of citronellal microcapsules, which might be ascribed to the sustained release profiles of citronellal from coacervates, thus extending the antimicrobial duration of citronellal. Mirmoeini, Moradi, Tajik, and Almasi (2025) also found that active packaging incorporated with Zataria multiflora essential oil could effectively reduce the counts of total mesophilic bacteria (TMC), total psychrotrophic bacteria (TPC), and total coliform bacteria (TCC) in minced meat over 9 days of refrigerated storage due to the gradual release of active compounds. These phenomena indicated that compared with free citronellal, citronellal microcapsules possessed superior potential in maintaining strawberry freshness. This is mainly attributed to the sustained-release property of the core active components, allowing the core materials to sustain stable functional performance (antibacterial and antioxidant activities) throughout the whole storage period. As reported, Algarni (2025) fabricated an edible coating with oregano essential oil and chitosan, and applied the coating solution for strawberry preservation. Compared with single chitosan coating and pure oregano essential oil treatment, the composite coating of chitosan incorporated with oregano essential oil showed better performance in maintaining the storage quality of strawberries and inhibiting microbial growth (Algarni, 2025). Similarly, Han et al. (2025) encapsulated tea tree oil within a composite of β-cyclodextrin and nano-montmorillonite, and the obtained microcapsules were applied for strawberry preservation. Experimental results showed that the prepared microcapsules maintained the appearance and quality of strawberries, inhibited spoilage during refrigeration, reduced weight loss, retained firmness, delayed the declines in soluble solid content and titratable acidity at the later storage stages, and enhanced the activities of superoxide dismutase, catalase, and ascorbate peroxidase (Han et al., 2025). The microcapsules also restrained the accumulation of anthocyanins and inhibited the further ripening of stored strawberries (Han et al., 2025). In addition, essential oil microcapsules have also been successfully applied to improve the storage quality of various fruits and vegetables, such as mango fruits (Liu et al., 2024), cherry tomatoes (Lou, Tao, Yu, Cui, & Chan, 2025), and white mushrooms (Zhang et al., 2024), thereby enhancing the functional performance of active core materials. According to the literature, the sustained release of core materials might account for their enhanced preservative effects on fruits and vegetables (Liu et al., 2024). In addition, the higher stability of core materials after encapsulation may also contribute to their superior preservation performance (Liu et al., 2024). Conclusion This research was designed to fabricate microcapsules containing citronellal through complex coacervation using BBPI and GA as the wall material matrix, and to assess their applicability in the field of strawberry fresh-keeping. The results revealed that BBPI and GA interacted electrostatically to generate coacervates, and the critical BBPI-to-GA ratio and pH for efficient complex coacervation were optimized to 4:1 and 4.1, respectively. It was found that the liquid citronellal microcapsules exhibited a multi-core structure, and their mean particle size was determined to be 25.63 ± 2.69 μm. After freeze-drying, the resultant microcapsules displayed a smooth and intact surface free of obvious cracks and holes, which verified the successful encapsulation of citronellal within the BBPI-GA coacervate matrix. The BBPI-GA coacervates exhibited excellent performance in citronellal encapsulation, achieving an encapsulation efficiency (EE) of 82.95 ± 0.37%. In comparison with free citronellal, the resultant microcapsules showed enhanced thermal stability, superior sustained-release property and more potent antibacterial activity. For strawberry shelf-life extension, citronellal microcapsules showed greater potential than free citronellal in inhibiting weight loss, spoilage microbial growth, and firmness reduction of strawberries over 8 days of storage. Our study revealed that BBPI-GA coacervates not only exhibited suitability for citronellal encapsulation but also effectively exploited the potential of BBPI in the encapsulation of bioactive agents. Moreover, this study furnishes critical information to support the future exploitation of encapsulated natural antioxidants and antimicrobials, as well as their prospect in practical application within the field of food preservation. CRediT authorship contribution statement Haifang Mao: Writing – review & editing, Methodology, Conceptualization. Congle Wang: Writing – original draft, Data curation. Zijia Zhang: Investigation, Data curation. Yanzhen Long: Investigation, Data curation. Linlin Zhao: Investigation, Data curation. Jie Chen: Writing – review & editing, Investigation. Liwei Fu: Supervision, Project administration. Liqing Qiu: Writing – review & editing, Software. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements This research was financially supported by Shanghai Sailing Program (24YF2743800), National Natural Science Foundation of China (NSFC, Project No. 32502318), and Science and Technology Commission of Shanghai Municipality (No. 25ZR1402473). Contributor Information Liwei Fu, Email: fuliwei1993@sit.edu.cn. Liqing Qiu, Email: qiuliqing1993@sit.edu.cn. Data availability Data will be made available on request. References Associated Data Data Availability Statement Data will be made available on request.