Effects of drying techniques on cowpea albumin as an egg replacement for meringue cookie applications

Abstract

In this study, we evaluated the effects of freeze-drying, spray-drying, and vacuum-drying on upcycled cowpea albumin powder and its functional performance in meringue cookie applications. Freeze-drying most effectively preserved protein structure and functionality, resulting in cookies with well-defined air cells, superior texture, and minimal color change. Conversely, spray-drying led to extensive protein denaturation and aggregation due to rapid thermal exposure, which compromised air incorporation and foam stability. Vacuum-drying exhibited intermediate performance, partially preserving protein integrity and exhibiting moderate foaming capacity. Structural and functional analyses revealed that freeze-dried albumin exhibited the highest foaming capacity (133.67 %) and foam stability (53.37 %), along with a significantly lower zeta potential (−6.79 mV) and higher surface hydrophobicity (H₀ = 363.38) compared with the other treatments. These findings highlight the significant role of the drying method in preserving the physicochemical properties of plant-based proteins.

Highlights

• •Freeze-drying minimized albumin denaturation.
• •Spray-drying resulted in albumin aggregation.
• •Freeze-dried albumin formed a strong protein network in meringue cookies.
• •Meringue cookies made from freeze-dried albumin were of excellent quality.
• •Drying techniques are essential factors in developing plant-based foods.

1. Introduction

The rising demand for sustainable and functional plant-based proteins has led to extensive research on alternative protein sources (Tufaro & Cappa, 2023). While soy, pea, and mung bean proteins have been extensively studied for their functional properties, matching the performance of animal proteins—especially in applications requiring foaming, emulsifying, and gelling—remains a significant challenge, particularly for products like whipped toppings (Choi et al., 2023; Othman et al., 2024).

Among the underutilized legume proteins, cowpea (Vigna unguiculata) albumin remains largely overlooked, despite its high solubility and foaming capacity, which render it highly promising as an egg alternative  (Mune & Sogi, 2016; Zain et al., 2023). Industrial processing of cowpeas primarily focuses on extracting globulin, producing an albumin-rich fraction as a byproduct that is typically discarded. This not only represents a significant loss of functional proteins, as albumins constitute up to 25 % of the total seed proteins, but also limits the potential of cowpeas as a valuable protein source (Lu et al., 2000). Unlike globulins, albumins remain soluble across a wide pH range, making them particularly promising for foamed food applications such as meringue-based bakery products. However, processing challenges hinder their effective use, as co-extracted compounds, such as tannins and phytic acids, bind to proteins and form insoluble complexes, thereby reducing the yield of extracted albumin (Chua & Liu, 2019).

Stabilization through drying represents a promising strategy for improving the functional usability and shelf stability of cowpea albumin. While drying effectively extends the shelf life, it also affects critical protein properties such as structure, solubility, and foaming stability, all of which are essential for food applications. Importantly, the specific drying method employed can significantly influence these functionalities (Abdul-Fattah et al., 2007; Mutukuri et al., 2021; Shen et al., 2021). Conventional techniques, including conduction and convective drying, often result in rapid surface moisture loss, leading to structural collapse and heterogeneous shrinkage in protein-based food systems. While the effects of drying on soy and pea proteins have been widely studied, there remains a lack of systematic research on the ways the different drying processes impact the physicochemical properties of cowpea albumin (Geera et al., 2011; Loveday, 2020).

Previous studies demonstrated that the different drying methods influence protein functionality. Research on soy protein isolate–maltodextrin conjugates revealed that spray-drying enhances scalability owing to its continuous operation, high productivity, and cost efficiency, but risks protein denaturation, whereas freeze-drying preserves the protein structure but is costly and less scalable (Choi et al., 2025). Similarly, cowpea protein isolates have been explored for heat-induced gel applications; however, their foaming properties, particularly in baked goods, remain understudied (Choi & Hahn, 2024). Although protein conjugation via Maillard reactions improves foaming and solubility in plant proteins, these strategies have not yet been explored for upcycled cowpea albumin (Xue et al., 2013; Zhang et al., 2019). This leaves a critical knowledge gap in determining the drying method that best preserves the functional properties of cowpea albumin for egg replacement applications.

To address this gap, we systematically investigated the ways different drying methods—vacuum drying, freeze-drying, and spray-drying—affect the physicochemical and functional properties of cowpea albumin. The impact of these drying processes on key protein attributes, such as structural properties, solubility, and interfacial characteristics, that directly influence foaming performance is thoroughly evaluated. Additionally, the practical potential of dried cowpea albumin as an egg substitute is demonstrated through the preparation and evaluation of plant-based meringue cookies. Consequently, by linking drying-induced structural changes to functional performance, this study provides a practical approach for transforming cowpea albumin from an underutilized by-product into a valuable ingredient for sustainable food applications.

2. Materials and methods

2.1. Materials

The cowpea and egg white solutions were provided by Intake Co. Ltd. (Seoul, Korea). Dehulled whole cowpeas were used to prepare the protein extract. Sugar was purchased from CJ Co. (Seoul, Korea). All the other chemicals were of analytical grade.

2.2. Preparation of albumin powder

The albumin solution was prepared by modifying existing protein extraction methods (Wani et al., 2015). Cowpeas were peeled to form a paste, which was then combined with water at a ratio of 1:5 (w/v). The pH of the mixture was adjusted to 8.5 using 1 N NaOH upon stirring at 900 rpm for 1 h. The extraction was performed at pH 8.5 to optimize albumin solubility while minimising globulin co-extraction. After centrifugation (9000 rpm, 25 min, 4 °C), the dissolved protein was separated and the pH was further adjusted to 5.0 with 1 N HCl. The resulting albumin solution (supernatant) was obtained through centrifugation (9000 rpm, 15 min, 4 °C). Subsequently, the prepared albumin solution was spray-, freeze-, or vacuum-dried. To eliminate pH-induced variability, all albumin solutions were adjusted to pH 5.0 prior to drying. The final protein yield, determined as the percentage of recovered nitrogen (N × 6.25) relative to the initial cowpea flour, was approximately 12–13 %. The protein content was determined using the Kjeldahl method (AOAC, 2005) by applying a nitrogen-to-protein conversion factor of 6.25.

2.3. Drying conditions

Spray-drying was conducted using a spray dryer (SD-1010, EYELA, Tokyo, Japan) with the following parameters: inlet temperature, 170 °C; outlet temperature, 80 °C; atomisation pressure, 70 kPa; liquid feed pump rate, 10 mL/min; primary drying air-flow rate, 39 m³/h. Spray-drying was performed at 170 °C/80 °C to achieve rapid dehydration while minimising thermal degradation. Freeze-drying involved two steps: the albumin solution was first frozen at −78 °C for 24 h using a deep freezer and then dried in a freeze dryer (FD5508, IlShin BioBase, Yangju, Korea) at −71 °C for 72 h under a vacuum of 0.01 mbar. Vacuum-drying was performed in a vacuum oven (C-DVD3, Changshin Science Co., Seoul, Korea) at 60 °C and 0.1 kPa for 48 h. Subsequently, the dried powders were finely ground and sifted through a 40-mesh sieve (Chung-gye, Inc., Seoul, Korea). The resulting powders were categorized as SAL (spray-drying), FAL (freeze-drying), or VAL (vacuum-drying) based on the drying method employed, and they were stored at 4 °C for subsequent analysis. The moisture contents of the powders were 4.3 % for SAL, 2.8 % for FAL, and 5.2 % for VAL.

2.4. Characterization of the albumin powder

2.4.1. Chemical composition

The moisture, protein (calculated as %N × 6.25), lipid, and ash contents of the albumin powder were assessed using the AACC method (Committee, 2000). The carbohydrate content was calculated by deducting the combined values of protein, lipids, and ash from 100.

2.4.2. Zeta potential

The zeta potential of the albumin powder was determined at 25 °C using a Nano ZS90 Zetasizer (Malvern Instruments Ltd., Malvern, Worcestershire, UK). Prior to assessment, the albumin powder was diluted with distilled water to a concentration of 0.1 % (w/v). The resulting diluted solution was transferred to a cuvette to evaluate the zeta potential.

2.4.3. Thermal properties

The thermal properties of the albumin powder were analysed using differential scanning calorimetry (DSC 4000, PerkinElmer Inc., Waltham, MA, USA). The albumin powder was prepared as a 15 % (w/w) suspension in distilled water and stirred for 24 h. A portion of the prepared suspension was pipetted into a stainless-steel pan and hermetically sealed. The sample pan was heated at a rate of 5 °C/min over a temperature range of 20–120 °C, while an empty pan was used as the reference.

2.4.4. Surface hydrophobicity

The surface hydrophobicity of the albumin powder was assessed using the 1-anilinonaphthalene-8-sulfonic acid (ANS) method. A solution containing 1 mg/mL albumin powder in PBS (0.01 M pH 7.4) was prepared. The albumin solution was dissolved at 20 °C for 2 h, followed by centrifugation of the mixture at 10,000 ×g for 20 min, thus affording the supernatant. The protein concentration in the supernatant was quantified using the Bradford protein quantification method. The standard curve for Bradford quantification was prepared using bovine serum albumin (BSA) in the range of 0–1.0 mg/mL. The obtained supernatant was then diluted to concentrations of 0.15, 0.075, 0.038, and 0.019 mg/mL, and 20 μL of the ANS reagent was added to 4 mL of the protein solution. The fluorescence intensity was measured using a fluorescence spectrometer (FluoroMate FS-2, Scinco Co., Seoul, Korea) at excitation and emission wavelengths of 390 and 490 nm, respectively. ANS was used as the fluorescent probe, and the fluorescence intensity was measured at varying protein concentrations (0.019–0.15 mg/mL). Surface hydrophobicity (H₀) was calculated as the slope of the linear regression between the fluorescence intensity and protein concentration.

2.4.5. SDS-PAGE

SDS-PAGE analysis of the albumin powders was conducted using Mini-PROTEAN® TGX™ Precast Gel (Bio-Rad Laboratories, Hercules, CA, USA). The albumin powder was dissolved in distilled water to obtain a solution with a concentration of 1 mg/mL. Thereafter, a mixture of 2× Laemmli sample buffer (10 μL) and protein supernatant (10 μL) was heated at 100 °C for 5 min and then centrifuged at 6000 rpm for 1 min. The running buffer consisted of 10× Tris/Glycine/SDS buffer diluted 1× using distilled water. Subsequently, 10 μL of the prepared supernatant sample and 3 μL of the protein molecular markers were loaded onto the precast gel. Electrophoresis was performed at 200 V for 30 min. The gel was then stained with the Sun-Gel staining solution (LPS solution; Daejeon, Korea) upon continuous agitation for 24 h during the staining process. Following staining, the gels were rinsed with 200 mL deionized water and desalinated via continuous agitation for 48 h, with the solution renewed periodically.

2.4.6. Foaming properties

The foaming capacity (FC) and foaming stability (FS) of the albumin powders were evaluated using established methodologies (Liang & Kristinsson, 2007). The albumin powder was combined with distilled water at a concentration of 1 % (w/v) and allowed to hydrate overnight. Subsequently, the prepared albumin solution was mixed at 6000 rpm for 1 min using a homogenizer (Ultra-Turrax T25D; IKA Werke GmbH & Co., Staufen, Germany), followed by an additional homogenization step at 9000 rpm for 3 min. FC and FS were calculated as follows:

$$\mathrm{FC}\left(\%\right)=\frac{V_0}{V}\times 100$$

$$\mathrm{FS}\left(\%\right)=\frac{V_{20}}{V}\times 100$$

where V, V₀, and V₂₀ represent the total volume before homogenization, foam volume at 0 min after homogenization, and foam volume at 20 min after homogenization, respectively.

2.5. Characterization of meringue cookies

2.5.1. Preparation of meringue cookies

Meringue cookies were prepared using either the albumin powder or the egg white solution. Preparation of the albumin powder cookies involved the following steps: In a KitchenAid mixer (KitchenAid Artisan®, Whirlpool Corporation, MI, USA), 10 g of albumin powder, 90 g of distilled water, and 10 g of sugar were combined and mixed for 2 min at speed setting 6. Subsequently, 20 g of sugar was gradually added and the mixture was stirred at speed setting 8 for 2 min, followed by an additional 2 min at speed setting 10. The resulting meringue batter was divided into portions of approximately 45 mm in diameter and 30 mm in height, arranged on a baking tray and then placed in an oven (MA324BGS, LG Electronics, Seoul, Korea), and baked at 100 °C for 90 min. A similar procedure was followed for the control meringue made from an egg white solution, involving mixing 71.4 g of egg white solution and 28.6 g of distilled water to maintain a solid content equivalent to that of the albumin powder.

2.5.2. Batter specific density (BSD)

The batter specific density (BSD) of the meringue batter prepared using albumin powder or the egg white solution was determined by adapting a previously reported method with modifications (Yüceer & Caner, 2021). The weight of the meringue batter and the volume of distilled water were determined using the same calibrated container. BSD was calculated as follows:

$$\mathrm{BSD}\left(\frac{\mathrm{g}}{{\mathit{cm}}^3}\right)=\text{Weight of meringue batter in calibrated container}/\text{Volume of water in calibrated container}.$$

2.5.3. Rheological properties of meringue batters

The rheological characteristics of the meringue batter prepared from the albumin powder or the egg white solution were examined at 25 °C, using a rotary rheometer (RheoStress RS 1, HAAKE Instruments, Karlsruhe, Germany) with a plate–plate geometry system of 20 mm diameter. The distance between the plates and the geometry was adjusted to 1 mm. To investigate the flow behavior, steady shear tests were performed at shear rates ranging from 0.1 to 100 s⁻¹. Furthermore, frequency sweep tests were performed at 1 % strain by examining frequencies ranging from 0.1 to 10 Hz.

2.5.4. Appearance of meringue batters and cookies

The visual transformations of the meringue batters and cookies made from albumin powder or the egg white solution were captured using a stationary camera (450D; Canon, Tokyo, Japan) positioned in a black box.

2.5.5. Color of meringue cookies

The color of the meringue cookies made from albumin powder or the egg white solution was assessed using a colorimeter (CR-400; Konica Minolta Sensing Inc., Tokyo, Japan). Color parameters including L* (brightness), a* (redness), and b* (yellowness) were measured against a standard reference tile (L* = 94.51, a* = 0.09, b* = 2.69) before the measurements. The total color difference (∆E) was calculated as follows:

$$\mathrm{\Delta }E=\sqrt{{\left(∆L^{\ast }\right)}^2+{\left(∆a^{\ast }\right)}^2+{\left(∆b^{\ast }\right)}^2}$$

2.5.6. Microstructure of meringue cookies

Each cookie made from albumin powder or the egg white solution was examined as a whole and cut in half. Its internal structure was examined using a stereoscopic microscope (SMZ745T; Nikon, Tokyo, Japan). The surface of each cookie was also examined (at 50× magnification) to determine the arrangement of the bubbles developed during baking.

2.5.7. Textural properties of meringue cookies

The hardness and fracturability of the meringue cookies prepared with albumin powder or the egg white solution were assessed using a texture analyser (TA.XT Plus, Stable Micro Systems, Godalming, Surrey, UK). The cookies were formed into cylindrical shapes of 45 mm in diameter and 20 mm in height and were stabilized at 25 °C. Hardness was measured as the maximum force required to fully slice the sample using a Warner–Bratzler blade operated at a speed of 2 mm/s.

2.6. Statistical analysis

The meringue batters and cookies produced using albumin powder or the egg white solution were prepared at least thrice, and all measurements were conducted at least thrice. ANOVA was employed to analyse the experimental data, and the outcomes are presented as mean values along with their respective standard deviations. The significance of each characteristic was determined using the Tukey test, considering p < 0.05 as statistically significant. All statistical computations and analyses were performed using SPSS software (version 25.0; SPSS Inc., Chicago, IL, USA).

3. Results and discussion

3.1. Chemical composition of albumin powders

A chemical composition analysis of cowpea-derived albumin powders revealed a protein content between 60 and 61 % across the different drying methods employed (Table 1). These values are marginally higher than previously reported protein contents (47.6–57.7 %) for legume albumins (Yang et al., 2022), potentially due to differences in extraction parameters such as pH, extraction time, and drying conditions. The albumin powder also contained more than 18 % non-protein components, primarily fats and carbohydrates, indicating incomplete removal via centrifugation alone. The relatively high fat content can be attributed to the absence of a degreasing step, which can influence interfacial activity and subsequent foaming or emulsification properties. Furthermore, the presence of soluble oligosaccharides such as raffinose, stachyose, and verbascose, common in legumes, may have contributed to the functional characteristics of albumin powders (Han et al., 2024). While the total protein content remained stable across drying methods, macromolecular retention varied with drying conditions. These compositional differences highlight the less-refined nature of the albumin powders, aligning with the aim of utilizing underexploited protein fractions without extensive purification.

Table 1.

Mean values of the proximate composition of albumin powder according to the drying method.

	Proximate compositions (%, d.b)
	Ash	Crude fat	Crude protein	Carbohydrate
SAL	1.65 ± 0.02a	18.22 ± 0.54a	61.12 ± 1.13a	19.01 ± 1.01a
VAL	1.54 ± 0.03b	18.01 ± 0.32a	60.23 ± 1.32a	20.22 ± 0.88a
FAL	1.58 ± 0.01b	18.04 ± 0.68a	60.59 ± 1.45a	19.79 ± 0.87a

All data represent the mean of triplicates. Different letters in the same column represent significant differences (p < 0.05).

3.2. Zeta potential of albumin powders

The zeta potential provides important insights into the surface charge characteristics of proteins, significantly affecting their dispersion stability and interactions within food matrices. As shown in Table 2, the albumin powders prepared via different drying methods exhibited distinct surface charge behaviors. Given that the pH of all albumin solutions was standardized to 5.0, the observed variations in the zeta potential can be primarily attributed to structural changes induced by processing rather than pH differences.

Table 2.

Changes in thermal properties, zeta potential, and surface hydrophobicity (H₀) of albumin powder depending on the drying method.

Samples	Thermal properties				Zeta potential (mV)	H0
	To (°C)	Tp (°C)	Tc (°C)	ΔH (J/g)
SAL	N.D.1	N.D.	N.D.	N.D.	0.74 ± 0.14a	188.48 ± 8.08c
VAL	N.D.	N.D.	N.D.	N.D.	−3.49 ± 0.08b	324.05 ± 3.11b
FAL	68.55 ± 0.34	74.40 ± 0.13	79.73 ± 0.09	0.52 ± 0.02	−6.79 ± 0.17c	363.38 ± 6.10a

All data represent the mean of triplicates. Different letters in the same column represent significant differences (p < 0.05). T_o represents the onset temperature, T_p indicates the peak temperature, T_c refers to the completion temperature, and ΔH denotes the enthalpy.

¹Not detected.

SAL exhibited a near-neutral zeta potential (0.74 ± 0.14 mV), indicating minimal electrostatic repulsion between protein molecules. Such reduced surface charges may promote protein aggregation and coaggregation with carbohydrates, thereby reducing the availability of charged functional groups and limiting dispersion stability and solubility. In contrast, the freeze-dried sample (FAL) exhibited a significantly negative zeta potential (−6.79 ± 0.17 mV), indicative of stronger electrostatic repulsion and enhanced protein dispersibility. This characteristic is typically beneficial for maintaining protein solubility and dispersion in aqueous food systems. The vacuum-dried sample (VAL) showed an intermediate zeta potential (−3.49 ± 0.08 mV), suggesting moderate electrostatic interactions and partial structural modifications.

These differences highlight the impact of the drying methods on the surface properties of the albumin powders, potentially affecting their functionality in food applications such as whipped toppings, meringues, and foamed desserts. Subsequent structural analyses, including SDS-PAGE, can further clarify the relationship between these surface charge differences and protein aggregation.

3.3. Thermal properties of albumin powders

Thermal analyses provide valuable insights into the stability and structural characteristics of proteins by assessing their thermal transition behaviors. Table 2 summarizes the thermal properties of cowpea albumin powders prepared using various drying techniques. Among the samples, only FAL exhibited clear thermal transitions, with a peak temperature (T_p) of approximately 74 °C and an enthalpy change (ΔH) of 0.52 J/g. The presence of this thermal peak suggests relatively well-preserved native-like protein structures and stable intermolecular interactions in the freeze-dried albumin powders.

In contrast, SAL and VAL showed no detectable thermal transitions, possibly due to the structural modifications induced by the drying processes. The absence of such transitions suggests possible disruption of the native protein structure or interactions, caused by the thermal stress or dehydration conditions during drying. This preliminary interpretation aligns with previous studies on the impact of drying methods on protein stability (Nozawa et al., 2016). Further structural analyses, particularly SDS-PAGE and functional assessments, are essential to elucidate the extent and nature of the structural changes in these powders.

3.4. Surface hydrophobicity of albumin powders

Table 2 presents the surface hydrophobicity of cowpea-derived albumin powders processed using different drying methods, providing insights into how these methods influence protein-protein interactions and interfacial behavior, which are critical in food formulations (McClements & Keogh, 1995). The drying techniques led to distinct variations in surface hydrophobicity among the tested albumin powders, as summarized in Table 2.

SAL exhibited the lowest surface hydrophobicity (188.48 ± 8.08), possibly due to protein aggregation during spray-drying, which traps hydrophobic domains within the aggregated protein structures and reduces their surface accessibility. While moderate heating typically increases surface hydrophobicity by exposing hydrophobic residues, the intense thermal stress from spray-drying may have triggered structural changes such as protein unfolding and subsequent aggregation, ultimately limiting the availability of surface-exposed hydrophobic residues (Ryan et al., 2012). VAL exhibited intermediate surface hydrophobicity (324.05 ± 3.11), suggesting partial retention of accessible hydrophobic domains. The moderate drying conditions of vacuum drying likely induced limited protein unfolding or structural rearrangements, offering a balance between structural preservation and modification. FAL showed the highest surface hydrophobicity (363.38 ± 6.10), indicating superior preservation of the native protein structure. Freeze-drying possibly minimized aggregation and preserved the well-organized hydrophobic domains accessible at the protein surface. Such structural preservation is generally beneficial for interfacial activities critical to protein functionality.

Overall, these findings suggest that drying-induced structural variations significantly affect the surface hydrophobicity. Further investigation into the relationship between these structural changes and functional properties, such as foaming stability, is presented in the subsequent section.

3.5. SDS-PAGE analysis of albumin powders

SDS-PAGE analysis of the albumin powders was conducted to assess the structural integrity of the protein fractions under reducing conditions, which facilitate dissociation into subunits and monomers, providing valuable insights into the protein structure and stability. The electrophoresis results presented in Fig. 1A reveal prominent albumin (23–26 kDa) and globulin (75–80 kDa) bands, consistent with previous reports on legume-derived protein extracts (Yang et al., 2022).

Fig. 1.

Structural and functional properties of albumin powder according to the drying method. (A) SDS-PAGE: lane M: marker; lane 1: SAL; lane 3: VAL; lane 4: FAL. (B) Foaming properties. Different lowercase letters within the same foaming property (i.e., capacity or stability) indicate significant differences among treatments (p < 0.05).

SAL exhibited pronounced fragmentation of globulin bands, characterized by the appearance of multiple lower molecular weight subunits. This fragmentation indicates structural degradation, likely caused by the intense thermal stress of spray-drying. VAL exhibited intermediate structural changes, with less pronounced fragmentation compared to SAL, suggesting partial preservation of the protein structures due to the relatively moderate drying conditions employed. In contrast, FAL exhibited minimal structural disruption, maintaining intact albumin and globulin bands. This minimal fragmentation reflects superior structural preservation, attributable to the gentle drying conditions of freeze-drying. Collectively, these SDS-PAGE results underscore the substantial structural differences induced by the drying methods.

3.6. Foaming properties of albumin powders

Fig. 1B compares the foaming capacity and foam stability of albumin powders subjected to different drying treatments, providing visual confirmation of the observed trends. Distinct differences were evident among the drying methods, closely correlating with structural properties identified in previous analyses (Table 2 and Fig. 1A).

Proteins in FAL exhibited superior interfacial properties, owing to the minimal structural damage and retained native conformations caused by freeze-drying. SDS-PAGE revealed minimal fragmentation, high surface hydrophobicity, and significantly negative zeta potential. Such characteristics enabled FAL proteins to effectively spread at the air-liquid interface and form stable, cohesive films. The enhanced interfacial elasticity promoted strong stabilization of air bubbles, resulting in optimal foam capacity and stability. In contrast, SAL, subjected to high-temperature spray drying, showed the lowest foaming capacity and stability. SDS-PAGE indicated extensive protein fragmentation and reduced globulin intensity, indicating significant protein unfolding and hydrolysis. Combined with its near-neutral zeta potential and reduced surface hydrophobicity, these structural degradations promoted protein aggregation, impairing effective adsorption and cohesive film formation at the interface, thus severely compromising foam stability. VAL, produced via moderate-temperature vacuum-drying, demonstrated intermediate foaming properties. Although some aggregation likely occurred under moderate drying conditions, VAL retained intermediate structural stability, evidenced by its negative zeta potential and relatively high surface hydrophobicity. These properties allowed the partial retention of protein dispersibility and interfacial adsorption capability, enabling VAL to support foam stability more effectively than SAL, yet less effectively than FAL. In conclusion, VAL represents a balanced option, suitable for applications requiring moderate foam performance.

Collectively, these results clearly illustrate that drying-induced structural modifications significantly influence protein interfacial behavior and subsequent foaming characteristics. Among the drying methods evaluated, freeze-drying (FAL) proved the most effective in preserving foaming performance, vacuum drying (VAL) exhibited moderate effectiveness, and spray-drying (SAL) was the least effective.

3.7. Physicochemical properties of meringue batters

3.7.1. Batter specific density (BSD)

To evaluate the degree of air incorporation into meringue batters, the batter specific density (BSD) was measured, and the results are presented in Fig. 2A. Previous studies suggested a strong correlation between the batter density and internal air content, significantly affecting the texture and stability of the final product (Yüceer & Caner, 2021). Although the BSD value of the egg white-based meringue batter was approximately 0.13, the values for SAL, VAL, and FAL were approximately 0.20, 0.12, and 0.14, respectively. Such results suggest that a lower BSD is associated with greater air incorporation into the batter matrix, which generally indicates more stable protein foams. During whipping, a lower BSD enables more efficient entrapment of air, thereby enhancing foam stability and contributing to a lighter, more voluminous meringue texture. The relatively high BSD observed in SAL suggests that the protein network was unable to incorporate sufficient air, likely because of the increased protein aggregation and reduced electrostatic repulsion. These structural changes could have limited protein unfolding at the interface, restricting air entrapment, and leading to a denser batter with reduced foam stability. The high density of the SAL meringue batter may have contributed to excessive fluidity, preventing the batter from maintaining its shape, and indicating difficulties in forming a stable protein foam structure capable of trapping air.

Fig. 2.

Batter specific density (BSD) and rheological properties of meringue batters prepared using egg white or albumin powder according to the drying method. (A) BSD. (B) Relationship between shear stress and apparent viscosity. (C) and (D) Variation of G′ and G″ with frequency.

Conversely, the FAL and VAL methods produced BSD values comparable to those of egg whites, suggesting that the protein interactions facilitated by the increased surface hydrophobicity and negative zeta potential allowed better air incorporation and contributed to the formation of a stable structure. The lower BSD values observed in the FAL and VAL samples suggest the maintenance of a finer and more evenly distributed air-cell structure, which is essential for achieving a light and stable meringue texture. These findings suggest that albumin powder produced via spray-drying (SAL) may disrupt the structure of meringue cookies, preventing them from maintaining their desired shapes and textures. Compared to the SAL method, the FAL and VAL methods, with their closer-to-ideal densities, are more likely to enhance the structural integrity and support the formation of a stable meringue.

3.7.2. Rheological properties

Figs. 2B–D present the rheological profiles of the batters under varying shear and frequency conditions, illustrating their pseudoplastic behavior and gel-like consistency. Rheological characteristics play a crucial role in determining the texture and shape of meringue cookies and eggless cakes (Ashwini et al., 2009; Vega & Sanghvi, 2012). All batters, either based on egg whites or albumin powders, exhibited a pseudoplastic behavior characterized by shear-thinning properties, in which the viscosity decreased with increasing shear rate (Fig. 2B). The pseudoplastic behavior of the FAL and VAL batters possibly arises from protein unfolding and subsequent intermolecular interactions, forming transient networks that respond to shear stress by aligning and disentangling. Despite this similarity, the egg white-based meringue batter exhibited the highest viscosity, suggesting that stronger protein interactions occurred within its network. Previous studies suggested that egg proteins undergo partial unfolding under shear stress, allowing reformation of hydrogen bonds and enhancing viscosity, which helps maintain the structural consistency of the batter (Chang et al., 1999). Conversely, the albumin powders subjected to different drying methods exhibited variable viscosities. The higher viscosities of FAL and VAL may also reflect partial protein unfolding, which increases the intermolecular entanglement and enhances the formation of a structured network capable of resisting shear deformation, thus retaining viscosity (Yang et al., 2023). Conversely, the lower viscosity observed in SAL suggests that heat-induced aggregation leads to the formation of large, non-uniform protein clusters that limit molecular mobility, thereby reducing the effective intermolecular interactions necessary for network formation (Yang et al., 2024). In dynamic rheology, G′ (storage modulus) represents the elastic, solid-like behavior of a material, while G″ (loss modulus) describes its viscous, liquid-like behavior. Frequency sweep tests (Fig. 2C and D) further demonstrated that G′ consistently surpassed G″ for batters made from egg whites using FAL and VAL across a frequency range of 0.1–10 Hz. This trend indicated the formation of a gel-like network, which is essential for maintaining the stability of the meringue batter. In contrast, SAL exhibited a crossover, where G″ exceeded G′ at higher frequencies, indicating a transition toward a dominant viscous behavior under dynamic deformation. This shift suggests that the batter exhibited reduced elasticity and impaired protein network connectivity, which limited its ability to sustain air incorporation and maintain its foam structure, as observed in the BSD and foaming property analyses (Figs. 2A and 1B).

These viscoelastic characteristics indicate that the FAL and VAL samples possess stronger internal protein networks, which enhance air retention and structural stability in the final product. Additionally, maintaining bubble integrity during baking possibly contributes to the crisp and light texture favored by consumers. These findings underscore the importance of selecting appropriate drying methods to optimize the rheological functionality of albumin powders in aerated food systems.

3.8. Physicochemical properties of meringue cookies

3.8.1. Appearance and color

The visual appearance and internal structure of the baked meringue cookies are shown in Fig. 3. Meringue cookies made from egg whites exhibited a well-distributed internal air structure, which indicates stable protein foams capable of retaining air bubbles during baking. In contrast, SAL-based meringue cookies exhibited a deformed exterior and collapsed internal structure, possibly due to heat-induced protein aggregation and Maillard reaction-driven modifications, which weakened the protein-carbohydrate interactions and impaired foam stability. The VAL meringue cookies maintained a well-defined outer shape, however, they exhibited limited internal air retention, suggesting that, while the protein network provided sufficient rigidity, partial protein unfolding during vacuum drying may have reduced their elasticity, thereby limiting effective aeration and foam expansion. The intermediate air retention observed in VAL cookies likely stems from the partial preservation of the structural integrity of proteins during vacuum-drying. While the moderate drying temperature (60 °C) under vacuum conditions minimized extensive denaturation, it still induced some degree of protein unfolding. This led to a moderately functional protein network with sufficient interfacial activity to incorporate air but with insufficient elasticity and strength to stabilize it effectively. To address the structural instability observed in spray-dried albumin powders, future studies should incorporate carrier agents such as maltodextrin or gum arabic during the spray-drying process. These carriers may provide a protective matrix that preserves protein conformation and improves interfacial functionality. In contrast, the FAL-based meringue cookies exhibited a well-defined exterior and a dense, evenly distributed internal air structure, likely due to reduced protein damage and preserved intermolecular interactions, which promoted stable foam formation.

Fig. 3.

Visual images of meringue batters and cookies made using egg white or albumin powder according to the drying method. (A1–A4) Meringue batter. (B1–B4) Whole meringue cookie. (C1–C4) Cross-section of meringue cookie. Numbers 1 to 4 indicate meringues made from (1) egg whites, (2) SAL, (3) VAL, or (4) FAL.

As summarized in Table 3, the color variations among the meringue cookies further supported the observed structural differences. SAL exhibited the highest ΔE value (13.87 ± 0.35), indicating the most significant color change. This pronounced difference may be attributed to heat-induced Maillard browning and protein modifications, which are more pronounced during spray drying due to prolonged exposure to high temperatures (Mustafa et al., 2018). The darker coloration and higher ΔE in SAL align with its observed structural collapse, suggesting that Maillard reaction-induced crosslinking may have contributed to its weaker foam matrix. In contrast, FAL exhibited the lowest ΔE value (4.05 ± 0.12), indicating that freeze-drying better preserved the original color of the meringue cookies. Such a minimal color shift suggests that a lower processing temperature limits the Maillard reaction activity, reducing browning and maintaining protein solubility. The VAL sample exhibited an intermediate ΔE value (8.80 ± 0.16), suggesting that while some thermal modifications occurred, they were not as extensive as those observed in SAL, possibly due to partial protein denaturation and controlled oxidation under vacuum conditions.

Table 3.

Changes in color and textural properties of meringue cookies using egg white or albumin powder depending on the drying method.

Samples	L*	a*	b*	ΔE	Fracturability (gf)	Hardness (gf)
Egg white	71.70 ± 0.87c	0.67 ± 0.33b	21.31 ± 0.37b	–	101.60 ± 3.18c	142.00 ± 9.44c
SAL	58.80 ± 0.06d	3.73 ± 0.78a	25.26 ± 0.90a	13.87 ± 0.35a	128.77 ± 2.77a	196.27 ± 8.48a
VAL	78.68 ± 0.58a	−2.27 ± 0.08d	16.92 ± 0.92c	8.80 ± 0.16b	92.83 ± 2.00d	144.43 ± 8.01c
FAL	74.78 ± 0.23b	−1.76 ± 0.14c	21.74 ± 1.09b	4.05 ± 0.12c	119.53 ± 3.52b	174.63 ± 5.91b

All data represent the mean of triplicates. Different letters in the same column represent significant differences (p < 0.05).

Overall, the color differences among the meringue cookies reflect the degree of the Maillard reaction and protein structural changes induced by thermal processing during drying. To minimize Maillard-induced browning in SAL, strategies such as reducing inlet temperature, shortening residence time, or incorporating antioxidant compounds may be considered. In contrast, the low ΔE in the FAL sample suggests that the minimal thermal processing helped preserve the protein structure, which may have contributed to its superior foam stability and protein integrity. The VAL sample exhibited an intermediate ΔE value, indicating partial protein unfolding and oxidation, resulting in moderate structural and color changes. These findings highlight the importance of controlling the thermal exposure during drying to minimize structural and color degradation and ensure the optimal functional properties of albumin-based meringue products.

3.8.2. Microstructure

The stereoscopic microscopic images in Fig. 4 reveal the internal microstructures of the meringue cookies prepared using albumin powders processed using different drying techniques. Meringue cookies made with egg whites exhibited a well-distributed and finely structured air cell network, which is consistent with previous findings that emphasize the role of albumin (α-livetin) in stabilizing air bubbles and forming a robust foam structure (Lomakina & Mikova, 2006). In contrast, the SAL- and VAL-based meringue cookies exhibited larger and more irregular air cells, suggesting less effective air encapsulation and weaker structural integrity (Fig. 4). This structural irregularity observed in SAL and VAL align with the decreased protein solubility and increased protein aggregation discussed in 3.2, 3.5, respectively. The structural weakness of SAL is likely associated with heat-induced protein aggregation, which impairs the formation of a continuous protein matrix, resulting in a more fragile foam structure. Conversely, the FAL-based meringue cookies exhibited a dense and evenly distributed internal air-cell arrangement, closely resembling that of egg white-based meringues. This well-structured microarchitecture, depicted clearly in Fig. 4, suggests that FAL maintained a more intact protein structure, allowing for enhanced protein-protein interactions and stable foam formation (Zhu et al., 2024). The preservation of the protein solubility and hydrophobic interactions in FAL likely facilitated stronger intermolecular bonding, thereby improving foam stability and gas retention. Similar microstructural characteristics have been reported in foams stabilized by aquafaba, where fine, uniform air cells enhanced foam stability and texture (Tufaro & Cappa, 2023). Additionally, soy protein isolate–maltodextrin conjugates have demonstrated the ability to enhance foam stability by forming cohesive interfacial films and well-defined internal structures (Choi et al., 2025).

Fig. 4.

Stereoscopic microscopic image (50× magnification) of a cross-section of meringue cookies made using egg white or albumin powder according to the drying method. (A) Egg white. (B) SAL. (C) VAL. (D) FAL.

As evidenced by the batter density (Fig. 2A) and foaming property (Fig. 1B) data, which reflect air incorporation and foam stability, respectively, these microstructural findings aligned with the foaming capacity and batter density results, where FAL exhibited the highest foam stability and lowest BSD, suggesting a more efficient protein network for air retention and structure formation. These results underscore the importance of the drying method in determining the final textural properties of meringue products, particularly for applications where foam stability and uniform air distribution are critical for product quality.

3.8.3. Textural properties

The textural properties of the meringue cookies, specifically their fracturability and hardness, were significantly influenced by the drying method used on the albumin powder (Table 3). SAL-based meringue cookies exhibited the highest hardness and fracturability, primarily due to severe heat-induced protein aggregation. Such structural changes likely restricted protein flexibility and interfacial activity, hindering adequate air incorporation and resulting in a denser, excessively firm texture. FAL-based cookies also showed relatively high hardness and fracturability compared to egg white-based samples; however, their increased hardness was attributed to a dense, compact, and uniform internal air-cell structure. This structural arrangement provided a desirable crispness, distinctively superior to the excessively dense texture observed in the SAL-based cookies. In contrast, VAL-based cookies demonstrated comparatively lower hardness and fracturability, closely resembling the texture of egg white-based cookies. Their intermediate texture was consistent with their moderately porous internal structures, reflecting balanced protein structural preservation and moderate foam stabilization. These findings underscore the critical influence of drying-induced structural modifications on cookie texture.

4. Conclusion

In this study, we demonstrated that drying methods significantly influence the structural and functional properties of upcycled cowpea albumin powders, directly impacting their suitability for aerated food products such as meringues. Among the drying methods evaluated, freeze-drying (FAL) best preserved protein structure and functionality, yielding meringue cookies with superior foaming capacity, uniform and finely structured air cells, desirable crispness, and textures comparable to those produced using traditional egg whites. Vacuum-drying (VAL) provided moderate functionality, resulting in meringue cookies with acceptable air retention and intermediate textural and mechanical properties. Spray-drying (SAL), despite its industrial scalability, significantly impaired protein integrity and produced meringue cookies with dense, irregular air cells and compromised structural stability, substantially reducing the overall product quality. Our findings provide clear guidance for selecting drying methods in food applications: freeze-drying is recommended for premium aerated products requiring high stability, vacuum-drying for moderately functional products, and spray-drying, while scalable, necessitates further optimization. Suggestions for future research include refining vacuum and spray-drying conditions through protein–polysaccharide conjugation or enzymatic structuring, as well as exploring alternative drying techniques (e.g., drum-drying or refractance window drying). Furthermore, evaluating how variations in macromolecular composition specifically influence functional properties such as foaming and emulsification could significantly enhance the practical applicability of cowpea albumin powders. Overall, our study highlights the critical role of drying methods in optimizing the functional quality of plant-based proteins, particularly emphasizing their potential for enhancing the quality and consumer acceptability of plant-based meringue cookies and related aerated products.

CRediT authorship contribution statement

Hyun Woo Choi: Writing – review & editing, Writing – original draft, Validation, Methodology, Investigation, Formal analysis, Conceptualization. Youngsang You: Writing – review & editing, Visualization, Validation, Methodology. Jungwoo Hahn: Writing – review & editing, Resources, Methodology, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

The authors do not have permission to share data.