Preparation of phycocyanin-rich powder from Spirulina platensis algae using freeze/thaw/osmotic treatment coupled with Refractance Window drying and its application as a colorant in set-type yoghurt

Abstract

This study developed an integrated process combining green NaCl-assisted freeze/thaw extraction with Refractance Window (RW) drying technology to produce phycocyanin-rich powder from Spirulina platensis for application as a natural colorant in set-type yoghurt. NaCl-assisted extraction gave higher phycocyanin yield and purity than water, with 12 cycles identified as optimal. The extract was dried using RW drying at 70, 80, and 90 °C to produce phycocyanin-rich powders. Powders dried at 70 °C retained the highest phycocyanin content and most intense blue color, while higher temperatures degraded phycocyanin more than allophycocyanin and phycoerythrin, though all powders showed acceptable solubility and flowability. When incorporated into set-type yoghurt, the powders imparted distinct blue-green coloration, with the 70 °C powder producing the strongest color but also highest syneresis. These findings demonstrate that NaCl-assisted freeze/thaw extraction combined with RW drying effectively produces phycocyanin-rich powder, with drying temperature critically governing pigment retention as a natural food colorant.

Highlights

• •NaCl-assisted freeze/thaw treatment enhanced phycocyanin (PC) yield and purity.
• •Twelve freeze/thaw cycles with NaCl maximized phycocyanin recovery.
• •Refractance Window drying at 70 °C best preserved PC content.
• •Allophycocyanin and phycoerythrin were more stable than PC at high temperature.
• •Yoghurt with 70 °C powder showed strongest color but highest syneresis.

1. Introduction

Phycocyanin is a blue phycobiliprotein predominantly found in cyanobacteria, especially Arthrospira platensis (commonly known as Spirulina), where it functions as a light-harvesting pigment within the phycobilisome. It is a water-soluble, non-toxic protein composed of α- and β-polypeptide subunits that assemble into monomers, trimers (αβ)₃, and hexamers [(αβ)₃]₂ (Fernández-Rojas et al., 2014; Ferraro et al., 2020). Each monomer carries one or two phycocyanobilin chromophores, an open-chain tetrapyrrole covalently linked via a thioether bond, that confer the pigment's intense blue color and its absorbance maximum near 620 nm, with the A₆₂₀/A₂₈₀ ratio commonly used to assess extract purity (values above 0.7 being considered food grade) (Zhang et al., 2021). Spirulina biomass typically contains 14–25% phycocyanin on a dry-weight basis, making it the principal commercial source of this pigment. Beyond its role as a natural colorant, phycocyanin exhibits potent antioxidant, anti-inflammatory, neuroprotective, and anticancer activities, the latter demonstrated by its ability to inhibit proliferation and induce apoptosis in pancreatic, breast, and non-small-cell lung cancer cell lines (Agrawal et al., 2021; Spínola et al., 2024; Yu et al., 2025). These combined properties, together with its regulatory status as safe for food use, have driven rapidly growing interest in phycocyanin for food, cosmetic, and pharmaceutical applications, with the global market projected to expand substantially over the coming decade.

Phycocyanin can be recovered from S. platensis biomass using chromatographic separation (Shi et al., 2024), ultrasound-assisted extraction (Khazi et al., 2024), or freeze/thaw cell-disruption (Bougatef et al., 2025). Among these, the freeze–thaw method is the most widely adopted approach owing to its simplicity, scalability, and avoidance of toxic solvents. Wet biomass is suspended in an aqueous medium and subjected to repeated cycles of freezing (typically −20 to −80 °C) and thawing (4–25 °C); the resulting ice-crystal formation ruptures the cell wall and membrane, releasing phycobiliproteins into solution. Extraction efficiency depends on the number of cycles, buffer or solvent type, and species-specific cell-wall characteristics, with three to five cycles generally reported as optimal for recovery and purity (Chittapun et al., 2020). Buffer selection strongly influences performance; sodium phosphate buffer improves yield and purity (Chittapun et al., 2020), while Tris–HCl buffer tends to give higher phycocyanin concentrations in other species (Hu et al., 2024; Pispas et al., 2024a). Salt addition, particularly NaCl, plays a critical role by maintaining ionic strength, promoting osmotic shock across the cell membrane, and minimizing co-extraction of chlorophyll and other impurities, thereby improving extract purity (Li et al., 2020). Because phycocyanin is highly sensitive to heat, light, and prolonged processing, and dry biomass resists disruption more than fresh material, careful optimization of cycle number, temperature, and salt concentration is essential to maximize yield while preserving pigment integrity (Li et al., 2020; Yuan et al., 2022).

Once extracted, phycocyanin solutions must be converted into a stable powder to enable handling, storage, and incorporation into food matrices, generally through drying or microencapsulation processes that protect the pigment from oxidative, thermal, and moisture-related degradation (Huang et al., 2023; Jiménez-González & Guerrero-Beltrán, 2021; Sun et al., 2020). Conventional methods such as oven and convective drying are simple but energy- and time-intensive, and comparative studies have shown convective drying to require substantially longer processing times than microwave-assisted alternatives (Belwal et al., 2022; Zia & Alibas, 2021). Freeze-drying, while effective, is costly and energy intensive. Non-traditional techniques, including microwave, convective infrared, and Refractance Window (RW) drying, have therefore attracted growing interest for their efficiency and superior retention of bioactive compounds and color (Belwal et al., 2022). In RW drying, a thin layer of product is spread on a transparent plastic film (typically Mylar®) floating on circulating hot water; heat is transferred by conduction through the film, convection from the water, and radiation once the film's refractive index approaches that of water, enabling rapid drying while evaporative cooling keeps the product 20–25 °C below the water temperature (Jalgaonkar et al., 2020; Kumar et al., 2024). Operating at atmospheric pressure and below 80 °C, RW drying offers energy efficiency, cost-effectiveness, and pigment retention comparable to freeze-drying, making it an industrially feasible option for processing Spirulina biomass (Silva et al., 2023). For Spirulina specifically, moisture content below 10% with good retention of phenolics, flavonoids, and phycocyanin has been achieved around 70 °C for 4.5 h, whereas higher temperatures or extended drying times accelerate pigment degradation due to its thermal sensitivity (Silva et al., 2023). Precise control of drying temperature therefore remains essential (Silva et al., 2023; Yuan et al., 2022).

The resulting phycocyanin powder holds considerable promise as a natural colorant and functional additive in dairy products, particularly fermented products such as yoghurt, which provides an ideal matrix given its wide consumption and active lactic cultures (Bai et al., 2024). Fortifying low-fat yoghurt with phycocyanin has been shown to enhance antioxidant activity and confer antimicrobial effects that reduce spoilage bacteria without impairing lactic culture growth (Ahmadi et al., 2024). Performance of the powder in this context depends strongly on particle size, moisture content, and pigment concentration which are governed by upstream extraction and drying conditions, as well as on phycocyanin's sensitivity to the acidic, fermentative environment of yoghurt, which raises questions about its color stability and interaction with the milk protein matrix during processing and storage (Yuan et al., 2022).

Although freeze–thaw extraction, RW drying, and dairy fortification with natural pigments have each been studied individually, their integration into a single process for producing phycocyanin-rich powder and its direct application as a natural dairy col.'orant has rarely been reported. Specifically, limited work has combined multiple freeze/thaw cycles with NaCl-assisted osmotic treatment to jointly maximize phycocyanin yield and purity from S. platensis; drying studies of Spirulina biomass have relied mainly on freeze-drying or conventional hot-air drying, with few evaluating how RW drying temperature affects pigment retention and powder properties. While Spirulina powders have been added to yoghurt at various stages, systematic evaluation of RW-dried phycocyanin-rich powders as natural colorants and the relationship between extraction, drying conditions, powder characteristics, and yoghurt quality remains lacking. Closing this gap is necessary to establish an integrated, industrially scalable process for producing phycocyanin powder suited to dairy applications. Accordingly, the novelty of this study lies in integrating a green NaCl-assisted freeze/thaw extraction process, Refractance Window drying technology, and the application of the resulting phycocyanin-rich powder in set-type yoghurt within a single processing framework. This study was therefore conducted to develop such a process through three objectives. First, in the extraction stage, the effects of freeze/thaw treatment combined with NaCl-assisted osmotic treatment across multiple extraction cycles were investigated to determine their influence on phycocyanin yield and quality from S. platensis biomass. Second, in the powder production stage, phycocyanin extracts were dried by the Refractance Window method at different temperatures to evaluate effects on the physicochemical properties and pigment stability of the resulting powders. Finally, in the application stage, the three phycocyanin powders obtained under different RW drying temperatures were incorporated into set-type yoghurt formulations to assess their effects on the physicochemical, structural, and quality attributes of the product during fermentation and storage.

2. Materials and methods

2.1. Materials and chemicals

Freeze-dried Spirulina platensis biomass was purchased from Vinh Hao Algae Joint Stock Company (Vietnam) and stored at ambient temperature until use. Sea salt (Co.op Select brand, Saigon Co.op, Vietnam) was purchased from a local supermarket, and maltodextrin (DE 10–12) was supplied by Lihua Starch (China).

For yoghurt preparation, whole milk powder (Nutifood Joint Stock Company, Vietnam; containing approximately 19% protein, 22.6% fat, and 48.2% carbohydrate) and refined white sucrose (Bien Hoa Sugar Joint Stock Company, Vietnam) were purchased from a local supermarket. The lactic acid bacteria strains Lactobacillus delbrueckii subsp. bulgaricus and Streptococcus thermophilus were isolated from commercial Vinamilk yoghurt (Vietnam Dairy Products Joint Stock Company, Vietnam).

Analytical-grade chemicals, including iron(III) chloride hexahydrate (FeCl₃·6H₂O, ≥99.0%), sodium acetate (CH₃COONa, ≥99.5%), sodium chloride (NaCl, ≥99.5%), sodium hydroxide (NaOH, ≥96.0%), acetic acid (CH₃COOH, ≥99.5%), hydrochloric acid (HCl, 36–38%), potassium persulfate (K₂S₂O₈, ≥99.0%), sodium carbonate (Na₂CO₃, ≥99.5%), gallic acid (≥98.0%), and methanol (CH₃OH, 99.9%), were purchased from Hoa Nam Scientific Chemical Co. (Ho Chi Minh City, Vietnam). Folin–Ciocalteu reagent, TPTZ (C₁₈H₁₂N₆, ≥99.0%), ABTS (C₁₈H₂₄N₆O₇S₄·2NH₃, ≥98.0%), and Trolox (6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid, ≥97.0%) were obtained from Sigma-Aldrich (St. Louis, MO, USA).

For SDS-PAGE analysis, acrylamide, Tris(hydroxymethyl)aminomethane (Tris), sodium dodecyl sulfate (SDS), ammonium persulfate (APS), N,N,N′,N′-tetramethylethylenediamine (TEMED), and Coomassie Brilliant Blue R-250 were purchased from Merck (Darmstadt, Germany), and a protein molecular weight marker (10–250 kDa) was obtained from Thermo Fisher Scientific (Waltham, MA, USA). All chemicals used were of analytical grade to ensure purity and stability for laboratory experiments.

2.2. Preparation of phycocyanin extract

Phycocyanin was extracted from S. platensis biomass using a freeze–thaw cell-disruption method adapted from previous studies (Bougatef et al., 2025). A NaCl concentration of 100 mM, adopted from the literature, was used to facilitate osmotic-assisted extraction due to its reported effectiveness in enhancing phycocyanin recovery while preserving pigment stability (Y. Li et al., 2020). Dried algal biomass was suspended in the extraction solvent (distilled water or 100 mM NaCl solution) at a biomass-to-solvent ratio of 1:80 (g/mL) and vortex-mixed (01–110 Vortex Mixer, Biologix, USA) at 2500 rpm for 1 min. The suspension was frozen at −18 °C for 24 h and then thawed at ambient temperature; this freeze–thaw cycle was repeated from 1 up to 14 times. Previous studies have generally investigated a limited number of freeze–thaw cycles (typically 3–6) (Fernandes et al., 2023), although a recent study extended the investigation to 10 cycles to characterize phycocyanin release kinetics rather than determine the optimal extraction condition (Pispas et al., 2024b). Therefore, a broader range of freeze–thaw cycles (0–14) was evaluated in this study to comprehensively assess the extraction profile of NaCl-assisted freeze/thaw treatment and identify the point beyond which further cycles provide limited improvement in phycocyanin recovery. After each cycle, the suspension was re-vortexed and centrifuged at 7000 rpm for 10 min (TG-16 high-speed centrifuge, Yingtai, China), and the resulting supernatant was collected as the crude phycocyanin extract.

2.3. Characterization of phycocyanin extracts

2.3.1. UV–Vis absorption spectrum

UV–Vis absorption spectra of the phycocyanin extract were recorded over the range 200–800 nm using a UV-1280 spectrophotometer (Shimadzu, Kyoto, Japan).

2.3.2. Color attributes

Color parameters of the extract were determined using a Color Reader CR5 handheld colorimeter (3NH, China) and expressed as Hunter coordinates L* (lightness), a* (redness (+) to greenness (−)), and b* (yellowness (+) to blueness (−)). From these values, chroma (C*) and hue angle (h°) were calculated to characterize color saturation and color tone, respectively, using the following equations:

$$C^{\ast }=\sqrt{\left(a^{\ast 2}+b^{\ast 2}\right)}$$ (1)

$$h^o=\mathit{arctan}\left(\frac{b^{\ast }}{a^{\ast }}\right)$$ (2)

The total color difference (ΔE) between samples and a reference (control) was calculated as:

$$∆E=\sqrt{{\left(L^{\ast }-L_o^{\ast }\right)}^2+{\left(a^{\ast }-a_o^{\ast }\right)}^2+{\left(b^{\ast }-b_o^{\ast }\right)}^2}$$ (3)

where L₀*, a₀*, and b₀* represent the color coordinates of the reference (control) sample, and L*, a*, and b* represent the color coordinates of the treatment sample. Each sample was analyzed in triplicate, and mean values were used for statistical analysis.

2.3.3. Phycocyanin content and purity

Total phycobiliprotein content was determined following the method described by (Bennett & Bogorad, 1973), with slight modifications. The absorbance of the phycocyanin extract was measured using a UV-1280 spectrophotometer (Shimadzu, Kyoto, Japan) at the corresponding wavelengths, and phycobiliprotein concentrations were calculated using the following equations:

$$\mathrm{PC}\left(\mathrm{mg}/\mathrm{L}\right)=\frac{\mathrm{A}620-0.474\times \mathrm{A}625}{5.34}$$ (4)

$$\mathrm{APC}\left(\mathrm{mg}/\mathrm{L}\right)=\frac{\mathrm{A}652-0.20\times \mathrm{A}620}{5.09}$$ (5)

$$\mathrm{PE}\left(\mathrm{mg}/\mathrm{L}\right)=\frac{\mathrm{A}562-2.41\times \left(\mathrm{PC}\right)-0.849\times \left(\mathrm{APC}\right)}{9.62}$$ (6)

$$\text{Phycobiliprotein}\left(\mathrm{mg}/L\right)=\mathrm{PC}+\mathit{APC}+\mathit{PE}$$ (7)

where PC, APC, and PE represent the concentrations of phycocyanin, allophycocyanin, and phycoerythrin (mg/mL), respectively; A denotes absorbance at the corresponding wavelengths.

The purity of phycocyanin in the extract was evaluated based on the absorbance ratio:

$$\text{Phycocyanin purity}=\frac{{\mathrm{A}}_{280}}{A_{620}}$$ (8)

where A₆₂₀ and A₂₈₀ represent the absorbance at 620 nm (specific to phycocyanin) and 280 nm (total soluble protein), respectively.

2.3.4. Total phenolic content

Total phenolic content was determined using the Folin–Ciocalteu method with slight modifications (T. Sun et al., 2007). Diluted extract (0.6 mL) was mixed with 10-fold diluted Folin–Ciocalteu reagent (1.5 mL) and incubated for 5 min, followed by addition of 7.5% Na₂CO₃ solution (1.2 mL) and incubation in the dark for 30 min. Absorbance was measured at 765 nm, and results were calculated from a gallic acid calibration curve and expressed as mg GAE/L.

2.3.5. Antioxidant activity

Antioxidant activity was evaluated using ABTS radical-scavenging and ferric-reducing antioxidant power (FRAP) assays (Nguyen, Dang, et al., 2022). For the ABTS assay, radicals were generated by reacting ABTS (0.0115 g) with K₂S₂O₈ (0.0176 g, 1:1 v/v) and storing the mixture at 4 °C for 12–16 h; the ABTS•⁺ solution was diluted to an absorbance of 1.10 ± 0.02 at 734 nm, after which 0.2 mL of sample extract was added to 2.8 mL of ABTS•⁺ solution, incubated in the dark for 30 min, and absorbance measured at 734 nm. For the FRAP assay, the reagent was freshly prepared from sodium acetate buffer (0.3958 g CH₃COONa and 5.7 mL CH₃COOH made up to 100 mL), TPTZ solution (0.0775 g in 25 mL of 0.01 M HCl), and FeCl₃·6H₂O solution (0.0541 g in 10 mL) at a ratio of 10:1:1; 0.2 mL of diluted sample was added to 2.8 mL of FRAP reagent, vortexed, incubated in the dark for 30 min, and absorbance recorded at 593 nm. For both assays, results were expressed as mg Trolox equivalents per liter (mg TE/L) using a Trolox calibration curve. All absorbance readings were obtained using a UV-1280 spectrophotometer (Shimadzu, Kyoto, Japan).

2.3.6. SDS-PAGE electrophoresis

Protein composition was analyzed by SDS-PAGE using a 12.5% resolving gel (distilled water, 30% acrylamide, 1.5 M Tris-HCl pH 8.8, 10% SDS, 10% ammonium persulfate (APS), and TEMED) and a stacking gel (distilled water, 30% acrylamide, 0.5 M Tris-HCl pH 6.8, 10% SDS, 10% APS, and TEMED). Samples were mixed with loading buffer (2:1, v/v), heated at 100 °C for 3 min, and cooled for 2 min, with this cycle repeated three times; 10 μL of the denatured sample was loaded per well. Electrophoresis was run at 90 V for 30–60 min. Gels were stained with 0.23% Coomassie Blue R-250 for 30 min and destained with 5% methanol and 7% acetic acid until the background cleared. Protein bands were visualized with a BioDoc-It UVP imaging system (Analytik Jena AG, Germany) and compared against molecular-weight standards (10–250 kDa).

2.4. Preparation of phycocyanin-rich powder by Refractance Window drying

Phycocyanin-rich powder was produced from the S. platensis extract using a customized Refractance Window (RW) drying system, following the principles reported by (Nguyen, Nguyen, & Nguyen, 2022) with modifications. The batch laboratory-scale apparatus consisted of a hot-water reservoir (0.42 m × 0.28 m × 0.10 m) connected to a thermostatic water bath (DH.WB000106, Daihan-Scientific, Wonju-si, Korea), with a 0.25-mm-thick polyester film (Mylar™, Jiangsu, China) serving as the heat-transfer interface between the hot water and the extract layer. The phycocyanin extract was blended with maltodextrin (10%, w/w) as a carrier agent, homogenized, and spread evenly to a thickness of approximately 2 mm on the RW drying surface. Drying was carried out at three water-bath temperatures (70, 80, and 90 °C) until the samples reached a constant weight, requiring approximately 30, 20, and 15 min, respectively. After dehydration, the dried sheet was peeled from the film, milled, and sieved through a 0.5-mm mesh to obtain the phycocyanin-rich powder.

2.5. Characterization of phycocyanin-rich powder

2.5.1. Moisture content and color attribute

Moisture content of the phycocyanin-rich powder was determined according to the AOAC official method (AOAC 925.09) using the oven-drying technique. Briefly, approximately 2 g of powder sample was accurately weighed into a pre-dried and pre-weighed weighing dish and dried in a convection oven at 105 °C until a constant weight was obtained. The sample was then cooled in a desiccator for 30 min and reweighed. Moisture content was calculated as the percentage weight loss relative to the initial sample mass.

Color attributes of the powder were determined as described in Section 2.3.2.

2.5.2. Solubility and wettability

Solubility and wettability of the phycocyanin-rich powder were determined following the procedures described by (Nguyen et al., 2025), with minor modifications.

Solubility was measured by accurately weighing 0.2 g of sample and dissolving it in 10 mL of distilled water. The mixture was stirred using a ZX3 vortex mixer (Velp, Italy) at 200 rpm for 15 s to ensure uniform dispersion, then centrifuged at 5000 rpm for 10 min, and the supernatant was transferred to a weighing dish and dried in a convection oven at 105 °C until a constant weight was obtained. After drying, the sample was cooled in a desiccator for 30 min and weighed to determine solubility.

Wettability was determined by slowly adding 0.2 g of powder sample to 200 mL of distilled water in a 250-mL beaker, allowing the powder to absorb without stirring. The time was recorded until the powder particles sank and dissolved completely.

2.5.3. Flowability

Bulk and tapped densities were measured using a 50 mL graduated cylinder (Areekal et al., 2024). Bulk density was calculated as the ratio of sample mass to the initial (untapped) volume occupied, reflecting inter-particle void space, while tapped density was determined after mechanically tapping the cylinder until no further volume reduction occurred. The Carr index (CI) and Hausner ratio (HR) were calculated as:

$$\mathit{CI}=\frac{\mathit{tapped\; density}-\mathit{bulk\; density}}{\mathit{tapped\; density}}\times 100$$ (9)

$$\mathit{HR}=\frac{\mathit{tapped\; density}}{\mathit{bulk\; density}}$$ (10)

2.5.4. FE-SEM and FTIR

Surface morphology of the powders was examined by field-emission scanning electron microscopy (FE-SEM S4800, HITACHI, Japan) at an accelerating voltage of 10 kV. Functional groups were characterized by Fourier-transform infrared (FTIR) spectroscopy using a NIR-MIR Frontier system (PerkinElmer, USA).

2.5.5. Phycobiliprotein content

Phycobiliprotein content of the phycocyanin-rich powder was determined following the same procedure described in Section 2.3.3. Briefly, 0.5 g of powder sample was dissolved in 10 mL of distilled water, and the absorbance of the resulting solution was measured using a UV-1280 spectrophotometer (Shimadzu, Kyoto, Japan) at the corresponding wavelengths, with a 10% (w/v) maltodextrin solution used as the blank.

2.6. Application of phycocyanin-rich powder as colorant in set-type yoghurt formulation

Whole milk powder (15%, w/v) was reconstituted in warm water at 70 °C, followed by the addition of sucrose and stirring until a homogeneous mixture was obtained. The mixture was then cooled to 40–42 °C and inoculated with Lactobacillus delbrueckii subsp. bulgaricus and Streptococcus thermophilus at a 1:1 ratio, with an initial density of 10⁷ CFU/mL for each strain. For the control sample, maltodextrin (5%, w/v) was added instead of phycocyanin powder, while the treatment samples were prepared by adding phycocyanin-rich powder (5%, w/v) processed at 70, 80, and 90 °C during the reconstitution step. The inoculated mixtures were poured into glass jars and incubated at 40 °C for 8 h using an RGX-80B climate chamber (Xingchen SHKT, Shanghai, China). After fermentation, the samples were stored under refrigeration (4 °C) for 1 day prior to physicochemical and structural analyses.

2.7. Characterization of set-type yoghurt

2.7.1. Color attributes

Color properties of the yoghurt samples were determined as described in Section 2.3.2, using a Color Reader CR5 handheld colorimeter (3NH, China) to evaluate color variation and stability.

2.7.2. Water-holding capacity and syneresis

Water-holding capacity (WHC) was determined according to (Nguyen et al., 2024). A known mass of yoghurt (a₁) was mixed with distilled water, held at room temperature for 24 h, and centrifuged at 4800 rpm for 10 min to obtain the precipitate (a₂). WHC was calculated as:

$$\mathit{WHC}\left(\%\right)=\frac{\left(a_2-a_1\right)}{a_1}\times 100$$ (11)

Syneresis was evaluated by centrifuging a known mass of yoghurt at a fixed speed and duration, collecting and weighing the separated whey layer, and expressing the result as the percentage of whey relative to the initial sample mass:

$$\text{Syneresis}\left(\%\right)=\frac{m_{\mathit{Whey}}}{m_{\mathit{Initial}}}\times 100$$ (12)

2.7.3. Texture profile analysis

Texture parameters, including hardness (g) and springiness (mm), were measured using a CT3 texture analyzer (Brookfield Ametek, USA) fitted with a cylindrical probe (TA 11/1000; 25.4 mm diameter, 35 mm length), under a penetration distance of 10 mm, trigger force of 2 g, and pre-test speed of 2 mm/s.

2.8. Statistical analysis

Normality (Shapiro–Wilk test) and homogeneity of variance (Levene's test) were assessed prior to hypothesis testing. Student's t-test was used to compare the two extraction solvents (distilled water vs. NaCl solution), while one-way ANOVA followed by Tukey's post-hoc test was used to compare groups across freeze/thaw cycles, drying temperatures, and yoghurt formulations. All statistical analyses were performed at a 5% significance level using R version 4.5.2 with the agricolae package. Graphs were plotted using OriginPro 25 (OriginLab Corporation, Northampton, MA, USA).

3. Results and discussion

3.1. Effects of freeze/thaw/osmotic treatment on phycocyanin extraction

3.1.1. Phycocyanin yield and purity

The changes in phycocyanin (PC), allophycocyanin (APC), phycoerythrin (PE), total phycobiliprotein, and phycocyanin purity of S. platensis extracts obtained using distilled water and 100 mM NaCl solution across different freeze/thaw cycles are presented in Fig. 1. For the NaCl-treated extracts, phycocyanin content increased significantly (p < 0.05) with increasing cycle number, rising from 65.05 mg/L at cycle 1 to a maximum of 367.42 mg/L at cycle 12, before declining to 298.43 mg/L at cycle 14 (Fig. 1A). Allophycocyanin, phycoerythrin, and total phycobiliprotein content followed a similar trend, all peaking at cycle 12 (45.38, 33.37, and 446.17 mg/L, respectively) and decreasing thereafter at cycle 14. Phycocyanin purity also increased progressively with cycle number, from 0.10 at cycle 1 to a maximum of 0.29 at cycle 12, remaining statistically comparable through cycle 14 (0.27–0.29) (Fig. 1E). In contrast, extraction with distilled water produced a different profile. Phycocyanin content was highest at cycle 1 (74.91 mg/L) but decreased steadily with increasing cycle number, reaching a minimum of 51.32 mg/L at cycle 12, before increasing again to 72.11 mg/L at cycle 14. Total phycobiliprotein content remained relatively stable (82–92 mg/L) through cycle 12 before rising sharply to 144.48 mg/L at cycle 14, driven mainly by increases in APC and PE. Phycocyanin purity of the water extracts remained consistently low (0.07–0.11) across all cycles and showed a slight downward trend with increasing cycle number.

Fig. 1.

Changes in (A) phycocyanin, (B) allophycocyanin, (C) phycoerythrin, (D) total phycobiliprotein contents, and (E) phycocyanin purity of Spirulina platensis algae extracts over different freeze/thaw cycles using water and 100 mM NaCl as solvents.

This contrasting behavior between solvents can be explained by the combined mechanical, osmotic, and ionic effects exerted on the structural organization of the S. platensis biomass. Spirulina cells are enclosed by a peptidoglycan-based cell wall and contain an internal phycobilisome, a supramolecular light-harvesting complex anchored to the thylakoid membrane (Li et al., 2023). Within this complex, phycocyanin occupies the peripheral rod segments, held in place by relatively weak, non-covalent linker interactions, whereas allophycocyanin forms the core, which is more intimately associated with the membrane and requires greater structural disruption to be released (Kawakami et al., 2022). During each freezing step, the formation and growth of ice crystals within the cytoplasm and extracellular medium exert mechanical pressure on the cell wall and thylakoid membrane, while the subsequent thawing step causes rapid water re-entry that further intensifies this stress (Rodrigues et al., 2023). Because this disruptive action accumulates incrementally, repeated freeze/thaw cycling progressively weakens and ruptures an increasing proportion of Spirulina cells. As cellular disruption becomes more extensive, phycobilisome components are sequentially released, beginning with the more loosely bound phycocyanin in the rod region before extending toward the core allophycocyanin (Sato et al., 2024). This sequential release explains the parallel increase in PC, APC, PE, and total phycobiliprotein observed with increasing freeze/thaw cycles under NaCl treatment.

The addition of NaCl amplified the freeze/thaw disruption through two complementary mechanisms. First, the osmotic gradient generated by the salt solution draws water out of the cells before freezing, inducing plasmolysis and increasing membrane tension. Consequently, the cells become more susceptible to ice-crystal-induced rupture during subsequent freeze/thaw cycles, allowing greater structural disruption to accumulate than in a low-ionic-strength medium such as water (Kato et al., 2023). Second, the elevated ionic strength of NaCl weakens the electrostatic interactions between linker polypeptides that stabilize the connection between the peripheral phycocyanin rods and the allophycocyanin core. This promotes more rapid and selective dissociation of phycocyanin into the surrounding solvent (Marx & Adir, 2014), analogous to the disruption of electrostatic protein aggregates reported in other biological systems. In addition to enhancing cell disruption, NaCl also improves extraction selectivity by influencing the partitioning behavior of cellular components during phase separation (Sakuta et al., 2026). Hydrophobic membrane-associated constituents, particularly chlorophyll and its pigment–protein complexes, are preferentially retained within membrane fragments and cellular debris under high-ionic-strength conditions (Saga et al., 2025). In contrast, the intrinsically hydrophilic phycobiliproteins partition into the aqueous phase (Chen et al., 2022). This differential solubility explains why NaCl-assisted extraction increased both phycocyanin yield and purity by limiting the co-extraction of competing pigments and other cellular impurities. In the absence of NaCl, the weaker osmotic driving force and lower ionic strength reduced both plasmolysis-assisted membrane disruption and the ionic-mediated dissociation of phycobilisome components, resulting in less efficient and less selective phycobiliprotein extraction (Sato et al., 2024). Accordingly, the disproportionate increase in APC and PE at cycle 14, despite the decline in phycocyanin, likely reflects non-specific co-release of degraded proteins and cellular debris rather than continued selective extraction of intact phycobiliproteins.

3.1.2. UV–Vis absorption spectra

The UV–Vis absorption spectra of the S. platensis extracts obtained using distilled water and 100 mM NaCl solution are presented in Fig. 2A. Both extracts exhibited a similar overall spectral profile, with a prominent absorption peak at approximately 260 nm and a broad absorption band centred around 620 nm. The absorption peak at 260 nm is commonly associated with nucleic acids and other UV-absorbing cellular constituents released during cell disruption, whereas the band at 620 nm corresponds to the characteristic absorption of phycocyanin. Although both extracts displayed these characteristic absorption features at similar wavelength positions, the NaCl-treated extract exhibited substantially greater absorbance at 620 nm than the water-treated extract, consistent with its markedly higher phycocyanin concentration.

Fig. 2.

(A) UV–Vis absorption spectrum and (B) SDS-PAGE gel electrophoresis of Spirulina platensis algae extract after 12 freeze/thaw cycles using water and 100 mM NaCl as solvents. Notes: M is the standard protein marker with molecular weights ranging from 10 to 250 kDa.

The strong absorption at 260 nm observed in both extracts indicates that, besides phycobiliproteins, other intracellular components were simultaneously released during extraction. Consequently, the relatively low A₆₂₀/A₂₆₀ ratio, particularly for the water-treated extract, reflects the crude nature of the extracts rather than a shift in phycocyanin absorption characteristics, in agreement with the purity values. Importantly, no additional absorption peaks or noticeable spectral shifts were observed in either extract, suggesting that the freeze/thaw process did not introduce detectable structural modification of the phycocyanin chromophore, regardless of solvent. Instead, the principal difference between treatments was the relative intensity of the characteristic phycocyanin absorption band, indicating that NaCl-assisted extraction primarily improved pigment recovery rather than altering its optical properties. Similar spectral characteristics have been reported for crude phycocyanin extracts, in which phycocyanin coexists with nucleic acids, soluble proteins, and residual photosynthetic pigments (Ismail et al., 2023).

3.1.3. SDS-PAGE

The protein profiles of the NaCl- and water-extracted samples, analyzed by SDS-PAGE, are shown in Fig. 2B. The NaCl-treated extract displayed a distinct, intense band at approximately 17–18 kDa, corresponding to the β-subunit of phycocyanin, together with a faint higher-molecular-weight band around 70 kDa, which likely represents residual oligomeric or aggregated protein species (Ismail et al., 2023; Mogany et al., 2020). The α-subunit of phycocyanin, typically reported at approximately 14–15 kDa, was not clearly resolved as a separate band, probably because of its close electrophoretic mobility to the more intense β-subunit under the experimental conditions employed. In contrast, the water-treated extract lacked a clearly resolved band in the 17–18 kDa region and instead exhibited only a faint diffuse band around 70 kDa, together with a poorly defined smeared region below 20 kDa. The marked difference in band definition between the two extraction media indicates that NaCl-assisted freeze/thaw treatment recovered a greater amount of intact phycocyanin protein than extraction with water alone. This observation is in good agreement with the higher phycocyanin concentration and purity determined spectrophotometrically (Section 3.1.1), demonstrating that the increased pigment content reflected recovery of structurally intact phycocyanin rather than solely an increase in soluble protein concentration.

The well-resolved β-subunit band observed for the NaCl-treated extract further suggests that the extraction conditions were sufficiently mild to preserve the primary structure of phycocyanin during release from the biomass. In contrast, the diffuse bands and protein smearing observed in the water-treated extract may indicate lower extraction selectivity and a greater contribution of non-specific soluble proteins released during repeated freeze/thaw treatment. These electrophoretic results therefore complement the UV–Vis spectra by confirming that NaCl-assisted extraction not only enhanced phycocyanin recovery but also preferentially enriched the target phycobiliprotein within the crude extract. Although SDS-PAGE provides information on molecular weight distribution rather than protein conformation, the presence of a distinct β-subunit band without obvious fragmentation suggests that repeated freeze/thaw treatment up to the optimum extraction condition largely preserved phycocyanin integrity. This finding is consistent with the absence of spectral shifts in the UV–Vis analysis (Section 3.1.2), indicating that the extraction process primarily increased pigment recovery while maintaining the characteristic structural features of phycocyanin.

3.1.4. Color attributes

The color parameters of the S. platensis extracts obtained using distilled water and 100 mM NaCl solution across different freeze/thaw cycles are presented in Table 1, and the corresponding visual appearance is shown in Fig. 3. For the NaCl-treated extracts, L* decreased progressively with increasing freeze/thaw cycles, while b* became increasingly negative and chroma (C*) increased significantly (p < 0.05). In addition, the a* value gradually shifted from negative to positive after cycle 9, and the hue angle changed from 24.19° at cycle 1 to approximately 277–282° at cycles 9–14, indicating a clear transition toward the blue region of the CIELAB color space. These instrumental measurements were consistent with the visual appearance of the extracts, which changed from a pale blue-green color at low cycle numbers to a more vivid and intense blue after prolonged NaCl-assisted extraction (Fig. 3). In contrast, the water-treated extracts exhibited only minor variations in color parameters throughout the extraction process. L* remained relatively high, whereas a*, b*, chroma, and hue angle fluctuated within narrow ranges without a clear directional trend. Consequently, the extracts retained a relatively pale blue-green appearance across all freeze/thaw cycles, indicating limited color development compared with the NaCl-treated samples.

Table 1.

Changes in color attributes of Spirulina platensis algae extracts over different freeze/thaw cycles using water and 100 mM NaCl as solvents.

Color attribute	Solvent	Freeze/thaw cycles
		1	3	5	9	12	14
L*	NaCl	59.55 ± 3.00a	38.53 ± 1.25Ab	26.03 ± 1.79Ac	28.83 ± 1.29Ac	25.78 ± 0.97Ac	29.55 ± 1.22Ac
	Water	55.58 ± 2.50Aa	46.6 ± 4.06Bbc	43.55 ± 2.45Bc	56.33 ± 3.19Ba	52.85 ± 2.75Bab	51.30 ± 2.24Bab
a*	NaCl	−11.6 ± 0.95Ac	−10.15 ± 0.6Ac	−5.8 ± 1.68Ab	6.48 ± 0.74Aa	8.53 ± 0.61Aa	6.50 ± 0.64Aa
	Water	−12.53 ± 0.61Aab	−13.55 ± 0.64Bbc	−11.18 ± 0.57Ba	−11.55 ± 0.53Ba	−11.30 ± 1.01Ba	−14.78 ± 0.52Bc
b*	NaCl	−4.78 ± 0.45Aa	−26.80 ± 0.54Ab	−25.30 ± 0.6Ab	−38.55 ± 1.32Ac	−38.85 ± 0.87Ac	−39.53 ± 1.02Ac
	Water	−9.55 ± 0.53 Bd	−10.83 ± 0.57 Bd	−5.15 ± 0.6Bab	−4.53 ± 0.61Ba	−7.28 ± 0.49Bc	−6.3 ± 0.87Bbc
C*	NaCl	12.56 ± 0.73Ad	28.66 ± 0.61Ab	25.99 ± 0.95Ac	39.09 ± 1.37Aa	39.78 ± 0.89Aa	40.06 ± 1.07Aa
	Water	15.77 ± 0.17Bb	17.36 ± 0.33Ba	12.32 ± 0.48Bc	12.42 ± 0.55Bc	13.46 ± 0.61Bc	16.07 ± 0.79Bb
h°	NaCl	24.19 ± 0.44Ad	69.26 ± 1.04Ac	77.17 ± 3.31Ab	279.53 ± 0.92Aa	282.38 ± 0.83Aa	279.33 ± 0.79Aa
	Water	37.35 ± 2.86Ba	38.64 ± 2.61Ba	23.25 ± 1.03Bc	19.21 ± 1.09Bc	30.98 ± 1.62Bb	23.02 ± 2.30Bc
Notes: Data are presented as mean ± standard deviation of three replicates. Values in the same row with different lowercase letters and in the same column with different uppercase letters indicate statistically significant differences at the 5% significance level (p < 0.05).

Fig. 3.

Changes in the visual appearance of Spirulina platensis algae extracts over different freeze/thaw cycles using water and 100 mM NaCl as solvents.

The progressive decrease in L*, together with the increase in chroma and the shift of hue angle toward the blue region, indicates that NaCl-assisted freeze/thaw treatment enhanced not only the quantity but also the visual intensity of the extracted pigment. Rather than simply reflecting higher pigment concentration, these color changes demonstrate improved color quality, characterized by a darker, more saturated, and purer blue appearance that is desirable for food colorant applications. The close agreement between the instrumental color measurements and the visual observations further supports the effectiveness of NaCl-assisted extraction in producing phycocyanin-rich extracts with superior color characteristics.

The color behavior closely paralleled the changes in phycocyanin concentration and purity observed in Section 3.1.1. As phycocyanin is the principal blue chromophore in Spirulina, increased recovery of intact phycocyanin resulted in greater absorption of visible light, producing lower lightness and higher color saturation. Conversely, the relatively stable color parameters of the water-treated extracts are consistent with their limited phycocyanin recovery and lower extract purity. These findings indicate that instrumental color analysis provides a rapid indirect indicator of phycocyanin extraction efficiency and complements the spectrophotometric and electrophoretic analyses. Importantly, no deterioration in color quality was observed up to cycle 12 despite the substantial increase in pigment concentration, suggesting that the extracted phycocyanin largely retained its chromophore integrity under the optimized extraction conditions. This observation is consistent with the UV–Vis and SDS-PAGE results, which showed preservation of the characteristic absorption behavior and protein profile of phycocyanin following NaCl-assisted freeze/thaw extraction.

3.1.5. Total phenolic content and antioxidant activities

The changes in total phenolic content and ABTS and FRAP antioxidant activities of S. platensis extracts obtained using distilled water and 100 mM NaCl solution across different freeze/thaw cycles are presented in Fig. 4. For the NaCl-treated extracts, total phenolic content increased significantly (p < 0.05) with increasing freeze/thaw cycles, rising from 19.11 mg GAE/L at cycle 1 to a maximum of 46.11 mg GAE/L at cycle 12 before declining to 40.28 mg GAE/L at cycle 14. A similar trend was observed for antioxidant activities. ABTS radical-scavenging activity increased from 45.46 to 89.40 mg TE/L at cycle 12 and remained statistically unchanged at cycle 14 (80.55 mg TE/L), while FRAP activity increased from 9.44 to 13.81 mg TE/L at cycle 12, with no significant difference between cycles 12 and 14. In contrast, the water-treated extracts exhibited only limited changes throughout the extraction process. Total phenolic content remained relatively constant between cycles 1 and 12 and increased only slightly at cycle 14. Likewise, ABTS and FRAP activities varied within a relatively narrow range and remained consistently lower than those obtained using NaCl-assisted extraction.

Fig. 4.

Changes in (A) phenolic content, (B) ABTS, and (C) FRAP antioxidant activities of Spirulina platensis algae extracts over different freeze/thaw cycles using water and 100 mM NaCl as solvents.

The close parallel between the extraction profiles of total phenolic content, antioxidant activities, and phycobiliprotein content indicates that the same osmotic plasmolysis and ionic-mediated cell disruption mechanism enhanced the recovery of phenolic compounds and other water-soluble antioxidant constituents from the Spirulina biomass, alongside phycocyanin. Since phenolic compounds are distributed throughout the cytoplasm and cell wall matrix rather than confined to the phycobilisome (Del Mondo et al., 2021), their release is expected to track the overall extent of cellular disruption rather than the specific rod-to-core dissociation sequence governing phycobiliprotein release, which is consistent with the closely matched, cycle-dependent trends observed for both classes of compounds. Although phycocyanin is recognized as one of the major antioxidant components of Spirulina, the observed antioxidant activity is therefore likely the result of multiple bioactive constituents acting collectively, and the assays should be interpreted as reflecting the overall functional quality of the crude extract rather than the activity of phycocyanin alone.

The decline in total phenolic content and antioxidant activities beyond cycle 12 was consistent with the reduction in phycocyanin recovery, suggesting that prolonged freeze/thaw treatment may gradually reduce the stability of antioxidant constituents after their release into solution, offsetting the benefit of additional cellular disruption. Nevertheless, the relatively small decrease in ABTS and FRAP activities compared with the reduction in phycocyanin concentration indicates that the antioxidant functionality of the crude extracts remained largely preserved, likely because multiple antioxidant compounds contributed to the overall activity. Collectively, these results demonstrate that NaCl-assisted freeze/thaw treatment not only improved pigment recovery but also enhanced the functional value of the resulting extracts, supporting their potential application as natural food ingredients with both colouring and antioxidant properties.

Overall, NaCl-assisted freeze/thaw treatment resulted in substantially higher phycocyanin yield and purity than water alone at all cycle numbers tested, with 12 cycles identified as the optimal condition, giving the highest phycocyanin content (367.42 mg/L) and purity (0.29). This combination was therefore selected for subsequent characterization and drying experiments. Although additional freeze/thaw cycles beyond cycle 12 continued to expose the biomass to mechanical disruption, phycocyanin recovery declined at cycle 14. This decline indicates that once cell disruption is largely complete, continued freeze/thaw exposure no longer contributes additional pigment release but instead imposes cumulative physical and thermal stress on the phycobiliproteins already solubilized. The chromophore of phycocyanin, phycocyanobilin, is covalently attached to the protein backbone via a thioether bond, while the overall quaternary structure is maintained through non-covalent interactions between monomeric subunits (Z. Zhang et al., 2021); repeated freezing and thawing can progressively destabilize these interactions, promoting partial unfolding, aggregation, or degradation of the pigment–protein complex. Therefore, 12 freeze/thaw cycles represented the optimum balance between efficient cellular disruption and preservation of phycocyanin integrity under the extraction conditions employed in this study.

3.2. Effects of Refractance Window drying temperature on phycocyanin-rich powder properties

3.2.1. Moisture content and color attributes

Moisture content of the phycocyanin-rich powder differed significantly (p < 0.05) among drying temperatures, ranging from 3.43% to 5.21%, with all three powders showing relatively low moisture content, indicating that the RW drying process was effective in achieving adequate dehydration across the tested temperature range (Table 2).

Table 2.

Selected physicochemical properties of phycocyanin-rich powder prepared by Refractance Window drying at 70, 80, and 90 °C using maltodextrin as the carrier.

Attribute	Drying temperature
	70 °C	80 °C	90 °C
Moisture (%)	5.21 ± 0.26a	3.43 ± 0.06c	4.56 ± 0.17b
Solubility (%)	99.67 ± 17.04a	96.67 ± 11.55a	96.67 ± 5.77a
Wettability (s)	451.2 ± 29.4c	771.7 ± 21.1a	678.6 ± 22.2b
Bulk density (g/mL)	0.62 ± 0.01a	0.59 ± 0.01b	0.62 ± 0.01a
Tapped density (g/mL)	0.72 ± 0.01a	0.64 ± 0.02b	0.68 ± 0.03ab
HR	1.15 ± 0.01a	1.10 ± 0.03b	1.10 ± 0.04b
CI	12.90 ± 0.58a	10.28 ± 1.68ab	6.97 ± 1.02b
L*	12.67 ± 0.25a	18.32 ± 0.42b	12.51 ± 0.41a
a*	−4.33 ± 0.18a	−4.83 ± 0.08b	−2.36 ± 0.05c
b*	−10.58 ± 0.30a	−9.40 ± 0.05b	−6.15 ± 0.14c
C*	11.86 ± 0.58a	11.63 ± 0.33a	5.86 ± 0.19b
h°	247.80 ± 3.65a	249.13 ± 5.51b	247.31 ± 2.67a

Notes: Data are presented as mean ± standard deviation of triplicatesa and different letters in the same row indicate that the mean values were significantly different at 95% confidence level (p < 0.05).

Color attributes also showed clear and significant differences among drying temperatures (Table 2). L* was significantly higher at 80 °C, indicating that the powder dried at this temperature was visibly lighter than those dried at 70 °C and 90 °C, which were statistically similar to each other. The a* values remained within a relatively narrow negative range across all drying temperatures, reflecting a consistently weak green tone in all powders. In contrast, b* became significantly less negative with increasing drying temperature, indicating a progressive loss of blue tone at higher drying temperatures, consistent with the thermal degradation of phycocyanin (Ruprecht & Kohlus, 2023). This was reflected in chroma (C*), which was significantly lower at the highest drying temperature, confirming a marked reduction in color intensity and saturation. Hue angle remained relatively stable across all three temperatures, consistently within the blue region, indicating that the underlying hue characteristic of phycocyanin was preserved despite the decline in color intensity. Taken together, these results indicate that increasing drying temperature, particularly to 90 °C, resulted in a paler, less saturated blue powder, consistent with the known thermal sensitivity of phycocyanin.

3.2.2. Solubility

Solubility of the phycocyanin-rich powder did not differ significantly among drying temperatures, remaining consistently high across all samples (96.67–99.67%) (Table 2). This high solubility can be attributed to the presence of maltodextrin as the carrier agent used during Refractance Window drying, which is well known for its high water solubility and hygroscopic nature; its incorporation into the powder matrix likely facilitated rapid dissolution of the phycocyanin-rich particles upon contact with water, regardless of the drying temperature applied (Gharehbeglou et al., 2023). The lack of a significant temperature effect on solubility further suggests that the drying conditions used in this study did not induce substantial structural changes, such as excessive particle fusion or surface crust formation, that would otherwise hinder the dissolution behavior of the powder.

3.2.3. Wettability

Wettability was significantly affected by drying temperature, with the shortest wetting time observed at 70 °C and the longest at 80 °C, while the powder dried at 90 °C showed an intermediate value (Table 2). This difference in wetting behavior, despite comparable solubility across all samples, may be explained by temperature-induced changes in particle surface characteristics and porosity during drying. At higher drying temperatures, more rapid moisture evaporation can promote the formation of a denser, less porous particle surface or induce greater particle agglomeration, both of which can hinder the initial penetration of water into the powder matrix and thereby prolong the time required for the powder to fully wet and sink, even though the particles remain readily soluble once water uptake begins (Wang et al., 2024). The comparatively longer wetting time observed at 80 °C rather than 90 °C suggests that this effect on surface structure may not increase in a straightforward manner with drying temperature and could instead be influenced by additional factors such as localized particle size distribution or moisture content differences among the powders.

3.2.4. Flowability

Bulk density of the phycocyanin-rich powder ranged from 0.59 to 0.62 mg/L, and tapped density ranged from 0.64 to 0.72 g/mL, with both parameters showing significant differences among drying temperatures. These relatively narrow ranges indicate that the overall packing behavior of the powder particles under both static (bulk) and mechanically consolidated (tapped) conditions was broadly similar across the three drying temperatures, suggesting that particle size and morphology were not drastically altered by the RW drying process within this temperature range. The Hausner ratio (HR), derived from the ratio of tapped to bulk density, ranged from 1.10 to 1.15 and also differed significantly among samples, reflecting subtle differences in the degree of volume reduction achieved through tapping, which relates to the extent of inter-particle cohesion and rearrangement capacity within each powder.

The Carr index (CI) decreased significantly with increasing drying temperature, from 12.90% at 70 °C to 10.28% at 80 °C and 6.97% at 90 °C, indicating a progressive improvement in powder flowability as drying temperature increased. This trend can be attributed to the combined effects of moisture content and particle surface characteristics on inter-particle cohesion. Powders dried at higher temperatures generally undergo more extensive moisture removal and may develop a smoother or less sticky particle surface, both of which reduce inter-particle friction and cohesive forces (Jiang et al., 2025). Lower cohesion, in turn, allows particles to settle more readily into a compact arrangement under gravity alone, resulting in a smaller difference between bulk and tapped density and, consequently, a lower CI value. According to standard flowability classifications, HR values between 1.00 and 1.11 and CI values between 5% and 15% generally correspond to good flow properties, while HR values between 1.12 and 1.18 and CI values between 12% and 16% are typically classified as fair flow (Areekal et al., 2024). Based on these references, the powder dried at 90 °C (HR ≈ 1.10, CI ≈ 6.97%) exhibited good flow characteristics, whereas the powders dried at 70 °C and 80 °C fell within the fair-to-good flow range, indicating that all three powders possessed acceptable flowability for handling and further processing, with flow properties improving progressively at higher drying temperatures.

3.2.5. FTIR

The FTIR spectra of maltodextrin (MD) and the phycocyanin-rich powders dried at 70, 80, and 90 °C are shown in Fig. 5A. All four spectra displayed a broad absorption band in the region of 3600–3200 cm⁻¹ (Karrar et al., 2021), corresponding to O—H stretching vibrations of hydroxyl groups (Kang et al., 2019), together with a weaker band around 2900 cm⁻¹, attributed to C—H stretching vibrations, and a series of strong, sharp absorption bands in the fingerprint region between 1150 and 1000 cm⁻¹ (C. Zhang et al., 2020), characteristic of C—O—C and C—O stretching vibrations associated with the glycosidic linkages of the maltodextrin polysaccharide backbone. The close similarity of these dominant spectral features across all three powder samples and pure MD confirms that maltodextrin, used as the carrier material during Refractance Window drying, constituted the major structural component of the resulting powders. The preservation of these characteristic carbohydrate bands also suggests that the structural integrity of the maltodextrin matrix was maintained during drying, which is consistent with the favorable solubility and flowability observed for the resulting powders.

Fig. 5.

(A) FTIR spectra and (B) phycobiliprotein contents of phycocyanin-rich powder prepared by Refractance Window drying at 70, 80, and 90 °C using maltodextrin (MD) as the carrier. Notes: The FTIR spectrum of MD alone is included for comparison.

Despite this overall similarity, the phycocyanin-rich powders exhibited additional absorption bands in the 1650–1550 cm⁻¹ region that were less pronounced or absent in the pure MD spectrum. These bands are typically attributed to amide I (C O stretching) and amide II (N—H bending) vibrations, which are characteristic of peptide bonds in proteins (Ma et al., 2024), and their presence in the powder spectra provides evidence for the successful incorporation of phycocyanin protein within the maltodextrin matrix. The relative intensity of these amide-related bands appeared to decrease slightly with increasing drying temperature, particularly at 90 °C, which may reflect partial protein degradation or conformational changes in phycocyanin induced by greater thermal exposure during drying, consistent with the reduced color intensity and chroma observed for this sample, thereby reducing the integrity of the chromophore-protein complex and contributing to the lower color intensity and chroma observed for this sample.

Overall, the FTIR results indicate that the functional group profile of the phycocyanin-rich powders was dominated by the maltodextrin carrier, while additional protein-associated amide bands confirmed the co-existence of phycocyanin within the powder matrix. The observed spectral changes further support the functional measurements, indicating that higher drying temperatures induced mild structural modifications in phycocyanin while largely preserving the carrier matrix, thereby influencing pigment quality without substantially altering the overall physicochemical characteristics of the powders.

3.2.6. Phycobiliprotein content

The phycocyanin (PC), allophycocyanin (APC), and phycoerythrin (PE) content of the phycocyanin-rich powders dried at 70, 80, and 90 °C are presented in Fig. 5B. Phycocyanin content was significantly higher (p < 0.05) in the powder dried at 70 °C (40.18 mg/g) compared with those dried at 80 °C (23.94 mg/g) and 90 °C (21.26 mg/g), which did not differ significantly from each other. This marked decline in phycocyanin content at higher drying temperatures is consistent with the known thermal sensitivity of the pigment and corroborates the reduced color intensity and chroma observed for the powders dried at 80 °C and 90 °C, indicating substantial thermal degradation of phycocyanin as drying temperature increased above 70 °C (Ruprecht & Kohlus, 2023).

Allophycocyanin content showed a different pattern, being significantly lower at 80 °C (19.68 mg/g) compared with 70 °C and 90 °C (21.59 and 23.11 mg/g, respectively, not significantly different from each other). Similarly, phycoerythrin content was significantly lower at 80 °C (8.73 mg/g) than at 70 °C and 90 °C (11.22 and 12.02 mg/g, respectively, also not significantly different from each other). The relatively stable or even slightly increased APC and PE content at 90 °C, despite the continued decline in PC, suggests that these two phycobiliproteins may possess comparatively greater thermal stability than phycocyanin under drying conditions (Pittera et al., 2017), or that the proportional contribution of APC and PE to the total phycobiliprotein pool increased simply as a consequence of the more pronounced loss of phycocyanin at the higher drying temperatures.

Overall, these results indicate that phycocyanin was the most thermally labile of the three phycobiliproteins during Refractance Window drying, with its content decreasing sharply once the drying temperature exceeded 70 °C, while APC and PE remained comparatively more stable, further supporting 70 °C as the most favorable drying temperature for preserving the pigment quality of the phycocyanin-rich powder.

3.2.7. FE-SEM

The surface morphology of the phycocyanin-rich powders dried at 70, 80, and 90 °C, examined by field-emission scanning electron microscopy, is shown in Fig. 6. The powder dried at 70 °C exhibited an irregular, angular morphology composed of small, glassy, sharp-edged fragments with a rough, fractured surface texture. In contrast, the powders dried at 80 °C and 90 °C displayed a markedly different morphology, characterized by larger, smoother, sheet-like or flake-like structures with more continuous surfaces and fewer sharp fracture edges compared with the 70 °C sample.

Fig. 6.

SEM images of phycocyanin-rich powder prepared by Refractance Window drying at (A) 70, (B) 80, and (C) 90 °C using (D) maltodextrin as the carrier.

This shift from small, angular fragments at 70 °C to larger, smoother flakes at 80–90 °C likely reflects the influence of drying temperature on the physical state of the maltodextrin-based matrix during dehydration and subsequent milling. At higher drying temperatures, more rapid moisture evaporation may promote the formation of a more continuous, glassy sheet upon the RW drying film, which, upon milling, tends to fracture into larger, flatter flakes rather than the finer, more irregular fragments obtained from the comparatively less rigid or more brittle sheet formed at 70 °C. These morphological differences are consistent with the flowability results reported in Section 3.2.3, where the larger, smoother particles obtained at higher drying temperatures were associated with lower Carr index values and improved flow behavior, likely due to reduced surface irregularity and inter-particle friction compared with the smaller, more angular particles observed at 70 °C.

3.3. Application of phycocyanin-rich powder as a natural colorant in set-type yoghurt

3.3.1. Color attributes

The color parameters of the control and phycocyanin-fortified set-type yoghurt samples are presented in Table 3 and Fig. 7. Compared with the control, all phycocyanin-fortified samples showed a marked decrease in L*, more negative a*, and a shift of b* from positive to negative values, reflecting a clear transition from the pale, faintly yellowish appearance of plain yoghurt toward a blue-green coloration upon phycocyanin incorporation. This shift was further confirmed by hue angle, which moved from the yellow region in the control into the blue-green region in all fortified samples.

Table 3.

Water-holding capacity (WHC), syneresis, textural, and color attributes of set-type yoghurt supplemented with phycocyanin-rich powder prepared by Refractance Window drying at 70, 80, and 90 °C using maltodextrin (MD) as the carrier.

Attribute	Control	70 °C	80 °C	90 °C
WHC (%)	95.78 ± 1.96a	90.46 ± 5.48b	91.13 ± 1.65b	90.35 ± 0.56b
Syneresis (%)	29.40 ± 1.21a	34.33 ± 1.06b	39.19 ± 1.96c	46.03 ± 2.95d
Hardness (g)	21.85 ± 0.13a	13.52 ± 0.46b	17.37 ± 0.55c	14.82 ± 0.20d
Springiness (mm)	2.22 ± 0.01a	2.31 ± 0.11a	2.62 ± 0.10b	1.33 ± 0.04c
L*	78.54 ± 3.90a	51.70 ± 1.77b	63.41 ± 1.85c	65.06 ± 2.54c
a*	−2.05 ± 0.08a	−8.95 ± 0.3b	−9.48 ± 0.21c	−8.91 ± 0.42b
b*	10.90 ± 0.16a	−9.41 ± 0.38b	−4.87 ± 0.21c	−0.92 ± 0.03d
C*	11.05 ± 0.18a	12.99 ± 0.59b	10.66 ± 0.26c	8.94 ± 0.44d
h°	99.52 ± 1.27a	225.91 ± 0.65b	207.17 ± 0.82c	185.05 ± 1.52d
ΔE	–	34.36 ± 3.36a	23.08 ± 2.84b	19.20 ± 3.27c

Notes: Data are presented as mean ± standard deviation of triplicates and different letters in the same row indicate that the mean values were significantly different at 95% confidence level (p < 0.05).

Fig. 7.

Visual appearance of the phycocyanin-rich powder, and visual appearance of set-type yoghurt supplemented with phycocyanin-rich powder prepared by Refractance Window drying at 70, 80, and 90 °C using maltodextrin (MD) as the carrier.

Among the phycocyanin-fortified samples, color intensity decreased progressively with increasing powder drying temperature. Chroma and the negative magnitude of b* were highest in the yoghurt fortified with the 70 °C powder and declined significantly at 80 °C and 90 °C, while hue angle shifted gradually away from the blue-green region toward a more neutral tone as drying temperature increased. Consistently, the total color difference (ΔE) relative to the control was greatest for the 70 °C treatment and decreased significantly with increasing powder drying temperature. Taken together, these results indicate that the phycocyanin powder dried at 70 °C imparted the most intense and visually distinct coloration to the yoghurt, while powders dried at higher temperatures produced progressively weaker color effects, consistent with the greater phycocyanin content and pigment stability observed for the 70 °C powder.

3.3.2. Water-holding capacity and syneresis

Water-holding capacity (WHC) of the yoghurt was significantly higher in the control sample (95.78%) compared with all phycocyanin-fortified samples (90.35–91.13%), which did not differ significantly from one another regardless of powder drying temperature (Table 3). This reduction in WHC upon phycocyanin powder incorporation suggests that the addition of the powder, and its associated maltodextrin carrier, may have interfered with the formation of a homogeneous protein gel network during fermentation, resulting in a slightly weaker water-binding capacity compared with the unfortified control (Bazrafshan et al., 2024).

Syneresis showed a clear and progressive increase with increasing phycocyanin powder drying temperature, from 34.33% at 70 °C to 39.19% at 80 °C and 46.03% at 90 °C, all of which were significantly higher than the control (29.40%) and significantly different from one another. This trend indicates that yoghurt fortified with powder dried at higher temperatures exhibited greater whey separation and reduced gel stability (Table 3). Given that WHC did not differ significantly among the three phycocyanin-fortified samples, the progressive increase in syneresis with powder drying temperature suggests that the impact of the phycocyanin powder on gel network structure may be more directly related to changes in powder properties, such as reduced pigment and protein integrity, as reflected in the lower phycocyanin content observed at higher drying temperatures, affecting the fine structure and stability of the yoghurt gel network rather than its overall water-holding capacity per se.

3.3.3. Texture profile

Hardness of the yoghurt differed significantly among all samples, being highest in the control (21.85 g) and lower in all phycocyanin-fortified samples, with the 80 °C sample showing the highest hardness among the treatments (17.37 g), followed by 90 °C (14.82 g) and 70 °C (13.52 g). This overall reduction in hardness upon phycocyanin powder addition is consistent with the lower WHC observed for the fortified samples, suggesting that the incorporation of phycocyanin powder and its associated carrier material weakened the yoghurt gel network structure relative to the control.

Springiness was similar between the control and the 70 °C sample (2.22 and 2.31 mm, respectively, not significantly different), while the 80 °C sample showed significantly higher springiness (2.62 mm) and the 90 °C sample showed a marked and significant reduction (1.33 mm), the lowest among all samples. This sharp decline in springiness at 90 °C, together with its correspondingly high syneresis, suggests that yoghurt fortified with the powder dried at the highest temperature exhibited the least resilient and most structurally compromised gel network among the samples tested, consistent with the greater loss of phycocyanin integrity and associated color intensity observed for this powder in previous sections.

4. Conclusions

This study demonstrates that combining NaCl-assisted freeze/thaw extraction with Refractance Window drying provides a simple, energy-efficient, and scalable pathway for converting Spirulina platensis biomass into a functional, phycocyanin-rich powder suitable for use as a natural food colorant. The reliance on low-cost inputs, such as common salt, water, and a thin-film hot-water drying system operating at atmospheric pressure, together with the avoidance of organic solvents or specialized encapsulation equipment, underscores the practical and economic feasibility of scaling this process beyond the laboratory for small- and medium-scale food manufacturers seeking natural pigment alternatives. The clear dependence of pigment retention and powder functionality on drying temperature further indicates that this process can be readily tuned to match specific application requirements, whether prioritizing maximal color intensity or optimized powder handling characteristics.

The successful incorporation of the resulting powders into set-type yoghurt highlights the broader potential of this phycocyanin-rich powder as a viable, plant-derived alternative to synthetic blue colorants in dairy and other food systems, aligning with growing consumer demand for clean-label and naturally sourced ingredients. Beyond yoghurt, the favorable solubility and flow properties of the powder suggest suitability for a wider range of food applications, including beverages, confectionery, and other fermented products, where both visual appeal and functional integration into aqueous or semi-solid matrices are required. Nevertheless, this study was limited to physicochemical characterization and did not include sensory evaluation, storage stability of the phycocyanin-rich powder, or color stability of the fortified yoghurt during storage. These aspects should be investigated in future studies to provide a more comprehensive assessment of product performance and consumer acceptance. Future work should focus on extending this process to pilot- or industrial-scale production and further optimizing powder-yoghurt compatibility to mitigate structural changes such as increased syneresis. More exhaustive phycocyanin extraction and purification strategies should also be investigated to improve pigment recovery while maintaining its stability and functionality. Together, these advances would facilitate the broader application of phycocyanin-rich powder as a sustainable natural colorant in the food industry.

CRediT authorship contribution statement

Duc-Tin Nguyen: Writing – original draft, Methodology, Investigation, Data curation. Thanh-Thuy Dang: Writing – original draft, Methodology, Investigation, Data curation. Quoc-Duy Nguyen: Writing – original draft, Visualization, Methodology, Data curation. Thi Tuong Vi Tran: Writing – original draft, Methodology, Data curation. Thi-Thuy-Dung Nguyen: Writing – original draft, Methodology, Data curation. Quoc-Trung Huynh: Writing – original draft, Data curation. Anh-Duy Do: Writing – original draft, Data curation. Thi-Van-Linh Nguyen: Writing – review & editing, Writing – original draft, Visualization, Methodology, Investigation, Data curation, Conceptualization.

Funding statement

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

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

Data will be made available on request.