Bovine Serum Albumin-Based Sponges as Biocompatible Adsorbents: Development, Characterization, and Perfluorooctane Sulfonate Removal Efficiency

Abstract

Polyfluoroalkyl substances, particularly perfluorooctane sulfonate (PFOS), are persistent environmental pollutants with severe health risks due to their bioaccumulation and resistance to degradation. Current PFOS removal technologies are limited by either efficiency, cost, or environmental concerns. Here, we introduce a biocompatible, protein-based sponge approach using bovine serum albumin (BSA) as a scaffold for efficient PFOS removal. We developed highly porous, mechanically robust sponges by optimizing foaming parameters such as mixing speed, duration, and surfactant concentration. Advanced characterization techniques, including microcomputed tomography and cryo-scanning electron microscopy, confirmed the sponges' structural integrity. Leveraging natural BSA-PFOS interactions, the sponges demonstrate effective PFOS removal, achieving up to ≈80% efficiency at a pH of ≈7.4, similar to natural water systems. Adsorption behavior is described using Langmuir and Freundlich isotherms, showing high adsorption capacity and surface interaction. Mechanical testing confirmed durability, making the sponges suitable for real-world applications. This eco-friendly method surpasses conventional PFOS removal techniques, offering a cost-effective solution with potential applications in drug delivery, tissue engineering, and catalysis. This work paves the way for developing multifunctional, porous protein-based materials that address urgent environmental challenges while offering versatile applications in biotechnology.

Keywords: Langmuir and Freundlich isotherms; biocompatible materials; bovine serum albumin‐based sponges; microstructure optimizations; perfluorooctane sulfonate adsorptions; polyfluoroalkyl substances; protein‐based materials.

Figure 1

Synthesis, structural stabilization, and water absorption of BSA‐based sponges. a) Schematic of the BSA‐based sponge synthesis process: The process begins with a hydrogel precursor composed of BSA, TW‐20, APS, and Ru (II)bpy₃ ⁺². Homogenization incorporates air, inducing foam formation. The foam is then exposed to white light for 30 min at RT, triggering covalent crosslinking via Ru (II)bpy₃ ⁺² photolysis in the presence of APS. This reaction forms covalent bonds between adjacent tyrosine residues on BSA, yielding a porous sponge network. Inset: Zoom‐in of the BSA‐based sponge structure. b) Effect of foaming parameters on sponge structure: Foaming time, speed, and TW‐20 concentration significantly influence the microstructure and porosity of BSA‐based sponges. (i) Short time, low speed, and  TW‐20 concentration result in a dense hydrogel. (ii) Moderate speed and TW‐20 concentration produce a more open, interconnected porous hydrogel network. (iii) High‐speed and TW‐20 concentration leads to a highly porous sponge. c) Close‐up image of a porous, lightweight BSA‐based sponge after freeze‐drying, placed on a flower petal. d) Density and water absorption measurements for BSA‐based sponges synthesized with varying TW‐20 concentrations (0, 5, and 20 mM) and mixing speeds (5, 15, and 30 K rpm) over different foaming durations (1, 2.5, and 5 min). Sponges prepared with 20 mM TW‐20 at 15 K rpm exhibited the highest water absorption and lowest density (highlighted by the pink dashed circle), indicating optimal sponge formation conditions.

Figure 2

Multiscale morphological and mechanical analysis of BSA‐based sponges under varied foaming conditions. a) The 3D reconstructed microCT scans reveal the internal porosity and structural variations of BSA‐based sponges prepared under different foaming conditions, highlighting the complexity and heterogeneity in pore distribution across samples. b) Cross‐sectional CT images provide a detailed view of BSA distribution, with distinctive blue outlines marking areas enriched in BSA within the sponge matrix, showcasing structural differences between samples. c) Cryo‐SEM images illustrate the microstructure of the hydrogel struts: (i) Sponges without TW‐20, mixed at 15 K rpm for 2.5 min, exhibit densely packed structures with minimal porosity. (ii) Increasing TW‐20 concentration to 5 mM at the same mixing speed and duration enhances porosity, though the BSA network shows nonuniform integration, resulting in uneven pore distribution. (iii) Optimal pore formation and uniform hydrogel strut distribution are achieved with 20 mM TW‐20 and 15 K rpm mixing, showing clearly defined pores and well‐dispersed BSA. (iv) At 30 K rpm and 20 mM TW‐20, the structural integrity of the sponge is compromised, with disrupted hydrogel networks and decreased uniformity. d) Compressive stress–strain analysis: (i) Stress–strain curves of sponges prepared with 5 mM and 20 mM TW‐20 at 15 K and 30 K rpm, showing the effect of foaming conditions on mechanical behavior. (ii) Successive compressive tests for sponges with 5 mM TW‐20 at 15 K rpm reveal stable Young's modulus values (≈5 kPa) across five test cycles. (iii) Successive compressive tests for sponges with 20 mM TW‐20 at 15 K rpm show increasing stiffness over five cycles, with Young's modulus rising from ≈1 to ≈4 kPa, reflecting structural compaction and loss of porosity. The insets show Young's modulus progression across the test cycles.

Figure 3

BSA secondary structure preservation and scheme of BSA‐PFOS interactions in BSA‐based sponges. a) ATR‐FTIR spectra of 2 mM BSA in TRIS , 2 mM BSA with 20 mM TW‐20 mixture, and a 2 mM BSA‐based sponge prepared with 20 mM TW‐20. The spectra show effective removal of TW‐20 after washing the sponge with TRIS, indicated by the green‐dashed circle. b) Fourier deconvolution of the amide I band in the ATR‐FTIR spectra is used to analyze the secondary structures of BSA. The deconvoluted spectra reveal the preservation of BSA's key secondary structures—intramolecular β‐sheets, α‐helices, and β‐turns—throughout the sponge formation process. c) The analysis of the amide I band shows that BSA's native secondary structure remains intact across all conditions, demonstrating that neither TW‐20 nor the sponge preparation process significantly alters the protein's conformation before and after gelation. d) XRD patterns of native BSA powder and BSA‐based sponges’ powder. Native BSA exhibited characteristic peaks at 8.94° (θ ₁) and 19.30° (θ ₂), while the sponges showed corresponding peaks at 9.14° ± 0.03° (θ ₁) and 19.20° ± 0.24° (θ ₂). The first peak corresponds to ordered regions, potentially α‐helices and β‐sheets, while the second peak reflects amorphous regions. The similarity between the patterns confirms the preservation of BSA's semicrystalline structure during sponge fabrication. e) Schematic representation of serum albumin acting as a PFOS carrier in blood plasma. Inset: The formation of a BSA‐PFOS complex via hydrophobic interactions, electrostatic forces, and hydrogen bonding between BSA amino acids and PFOS. This natural affinity was leveraged in designing BSA‐based sponges to effectively adsorb PFOS for environmental remediation.

Figure 4

BSA‐based sponges for PFOS Removal: Adsorption mechanism, Efficiency, and Reusability. a) A scheme of (i) BSA‐based sponges immersed in PFOS solutions prepared at varying concentrations. (ii) A passthrough experiment where PFOS solution diluted in TRIS is passed through a sponge‐filled column to assess sponge PFOS removal efficiency. The PFOS removal was assessed using LC‐MS/MS. b) Adsorption isotherm study showing PFOS adsorption on BSA‐based sponges at three soaking times: 20 min, 6 h, and 12 h. The curves show that extended soaking times lead to significantly higher equilibrium adsorption capacities. This suggests that longer exposure allows more PFOS molecules to interact with available binding sites on the sponge surface, contributing to a higher adsorption capacity. c) Adsorption Isotherm Modeling and Data Fitting: (i) Langmuir Isotherm Model – Equilibrium adsorption data fitted to the Langmuir model, assuming monolayer adsorption on a homogeneous surface. The linear fit confirms the model's suitability, indicating that PFOS adsorption onto BSA‐based sponges predominantly follows a monolayer mechanism. Higher equilibrium adsorption capacities at extended soaking times suggest increased PFOS interaction with available binding sites. (ii) Freundlich Isotherm Model – Equilibrium adsorption data fitted to the Freundlich model, which describes multilayer adsorption on heterogeneous surfaces. The linear relationships observed indicate that PFOS initially adsorbs onto binding sites of varying affinities before progressing toward a more uniform distribution over time. The strong fits to both models highlight the complementary nature of the adsorption process, with early‐stage heterogeneity transitioning into monolayer adsorption at longer exposure durations.   d) Time‐dependent PFOS removal by BSA‐based sponges (20 mM TW‐20, 15 K rpm), evaluated after 20 min, 6 h, and 12 h of immersion. At lower concentrations (60 ppb), the sponges exhibited rapid and high adsorption, achieving ≈80% PFOS removal within 20 min. Despite a slight decline in removal efficiency as the PFOS concentration increased, the sponges continued to show robust adsorption, maintaining ≈80% removal at 60 ppb and ≈60% at 1000 ppb, even after 12 h. e) PFOS removal efficiency (%) of BSA‐based sponges over six reuse cycles. The sponges exhibited high initial removal efficiency, with values of ≈74% in the first cycle and ≈68% by the third cycle. This gradually declined to ≈24% in the fifth cycle and ≈10% in the sixth. The decline in PFOS removal efficiency is attributed to the progressive exposure to alkaline pH during washing, which disrupt the protein's native conformation and prevent the complete restoration of its original secondary structure. f) Comparative PFOS removal efficiency of sponges prepared with 5 and 20 mM TW‐20, tested with a 1000 ppb PFOS solution over multiple pass‐through cycles (one, three, and six). The sponges prepared with 20 mM TW‐20 exhibited consistently higher removal rates compared to those with 5 mM TW‐20, with the difference becoming more pronounced as the number of cycles increased. After six cycles, sponges with 20 mM TW‐20 reached a removal efficiency of ≈45%, while sponges with 5 mM TW‐20 achieved ≈36%. A column packed with three 20 mM TW‐20 sponges achieved ≈89% PFOS removal, attributed to increased surface area and interaction time. g) Comparative radar chart analysis of BSA‐based sponges versus other PFOS removal methods, including magnetic activated carbon (MAC), powder activated carbon (PAC), anion‐exchange resins, and electrocoagulation. The chart highlights the superior biocompatibility, microstructure controllability, and competitive PFOS removal efficiency of BSA‐based sponges, along with advantages in cost and synthesis complexity over traditional methods.

Figure 4

BSA‐based sponges for PFOS Removal: Adsorption mechanism, Efficiency, and Reusability. a) A scheme of (i) BSA‐based sponges immersed in PFOS solutions prepared at varying concentrations. (ii) A passthrough experiment where PFOS solution diluted in TRIS is passed through a sponge‐filled column to assess sponge PFOS removal efficiency. The PFOS removal was assessed using LC‐MS/MS. b) Adsorption isotherm study showing PFOS adsorption on BSA‐based sponges at three soaking times: 20 min, 6 h, and 12 h. The curves show that extended soaking times lead to significantly higher equilibrium adsorption capacities. This suggests that longer exposure allows more PFOS molecules to interact with available binding sites on the sponge surface, contributing to a higher adsorption capacity. c) Adsorption Isotherm Modeling and Data Fitting: (i) Langmuir Isotherm Model – Equilibrium adsorption data fitted to the Langmuir model, assuming monolayer adsorption on a homogeneous surface. The linear fit confirms the model's suitability, indicating that PFOS adsorption onto BSA‐based sponges predominantly follows a monolayer mechanism. Higher equilibrium adsorption capacities at extended soaking times suggest increased PFOS interaction with available binding sites. (ii) Freundlich Isotherm Model – Equilibrium adsorption data fitted to the Freundlich model, which describes multilayer adsorption on heterogeneous surfaces. The linear relationships observed indicate that PFOS initially adsorbs onto binding sites of varying affinities before progressing toward a more uniform distribution over time. The strong fits to both models highlight the complementary nature of the adsorption process, with early‐stage heterogeneity transitioning into monolayer adsorption at longer exposure durations.   d) Time‐dependent PFOS removal by BSA‐based sponges (20 mM TW‐20, 15 K rpm), evaluated after 20 min, 6 h, and 12 h of immersion. At lower concentrations (60 ppb), the sponges exhibited rapid and high adsorption, achieving ≈80% PFOS removal within 20 min. Despite a slight decline in removal efficiency as the PFOS concentration increased, the sponges continued to show robust adsorption, maintaining ≈80% removal at 60 ppb and ≈60% at 1000 ppb, even after 12 h. e) PFOS removal efficiency (%) of BSA‐based sponges over six reuse cycles. The sponges exhibited high initial removal efficiency, with values of ≈74% in the first cycle and ≈68% by the third cycle. This gradually declined to ≈24% in the fifth cycle and ≈10% in the sixth. The decline in PFOS removal efficiency is attributed to the progressive exposure to alkaline pH during washing, which disrupt the protein's native conformation and prevent the complete restoration of its original secondary structure. f) Comparative PFOS removal efficiency of sponges prepared with 5 and 20 mM TW‐20, tested with a 1000 ppb PFOS solution over multiple pass‐through cycles (one, three, and six). The sponges prepared with 20 mM TW‐20 exhibited consistently higher removal rates compared to those with 5 mM TW‐20, with the difference becoming more pronounced as the number of cycles increased. After six cycles, sponges with 20 mM TW‐20 reached a removal efficiency of ≈45%, while sponges with 5 mM TW‐20 achieved ≈36%. A column packed with three 20 mM TW‐20 sponges achieved ≈89% PFOS removal, attributed to increased surface area and interaction time. g) Comparative radar chart analysis of BSA‐based sponges versus other PFOS removal methods, including magnetic activated carbon (MAC), powder activated carbon (PAC), anion‐exchange resins, and electrocoagulation. The chart highlights the superior biocompatibility, microstructure controllability, and competitive PFOS removal efficiency of BSA‐based sponges, along with advantages in cost and synthesis complexity over traditional methods.