A bioinspired and degradable riboflavin-containing polypeptide as a sustainable material for energy storage

Significance

Sustainable energy storage materials can address environmental, safety, security, and ethical issues associated with conventional battery platforms. Herein, we advance the design of redox-active polypeptide materials by incorporating riboflavin, a biological redox cofactor, as redox-active pendant groups on a hydrolytically degradable polyglutamate backbone. This design retains the polypeptide backbone architecture while introducing renewable side-chain functionality, enabling the construction of a biologically sourced organic battery material. The resulting design improved sustainability, minimized cytotoxicity, and demonstrated effective degradability under hydrolytic conditions. By decoupling redox activity from fossil-based feedstocks or metals, this bioinspired strategy introduces a class of environmentally friendly materials for next-generation energy storage systems (ESSs).

Abstract

Inspired by Nature, we present a polypeptide-based organic redox-active material constructed from renewable feedstocks, L-glutamic acid (an amino acid) and riboflavin (vitamin B₂), to address challenges with start-to-end-of-life management in energy storage systems (ESSs). The amino acid was utilized to establish a degradable polymer backbone, along which many copies of riboflavin were incorporated to serve as the redox-active pendant groups that enabled energy storage. The overall synthesis involved the ring-opening polymerization (ROP) of an l-glutamic acid-derived N-carboxyanhydride (NCA) monomer, followed by side chain activation with azides and, finally, click coupling to achieve installation of alkyne-functionalized riboflavin moieties. The steric bulkiness and rich chemical functionality of riboflavin resulted in synthetic complexities that required reaction optimization to achieve the desired polymer structure. Electrochemical characterization of the resultant riboflavin polypeptide, in organic electrolyte, showed quasireversible redox activity with a half-wave potential (E_1/2) of ca. −1.10 V vs. ferrocene/ferrocenium (Fc/Fc⁺). Cell viability assays revealed biocompatibility, as indicated by negligible cytotoxicity for fibroblast cells. The polypeptide design, consisting of labile amide backbone linkages and side-chain ester functionalities that tethered the riboflavin units to the backbone, enabled hydrolytic degradation to recover building blocks for future upcycling or recycling. This bioinspired strategy advances the development of degradable redox-active polymers and promotes sustainable materials design for circular energy storage technologies.

Lithium-ion batteries (LIBs) have become the dominant energy storage systems (ESSs) for portable electronics due to their high energy density and long lifespan. Despite their ubiquity, there are considerable safety, environmental, and sociopolitical issues associated with the production and use of LIBs (1–3). Thus, there is increasing interest in the development of alternative ESSs.

Batteries constructed from organic polymer materials, having conjugated backbones or nonconjugated backbones with redox-active side chains, have each emerged as promising next-generation electrode materials in ESSs (4). Both types of systems have attributes of safety, transition metal-free compositions, flexibility, and rapid charging capability. Nonconjugated polymers that present electroactive pendant groups, however, possess additional advantages of processability and tunable structural features that define the electrochemical properties, such as capacity and redox potential (5, 6). While numerous nonconjugated polymers functionalized with electroactive pendant groups, such as anthraquinone (7–9), benzoquinone (10), catechol (11, 12), carbazole (13, 14), viologen (15–17), and persistent radical moieties (15, 16, 18–24), have been investigated and reported for battery applications, few were intrinsically degradable (25–29). Degradability of nonconjugated polymers is an essential aspect of end-of-life management and must be considered in the design of next-generation polymers to be used as one of many components required for the construction of sustainable battery devices.

To design functional polymers for a sustainable future, it is important to consider their renewable sourcing, recyclability, and/or end-of-life implications toward the environment (30). Bioinspired materials prepared from natural feedstocks offer promising solutions to ethical and environmental challenges. Due to their synthetic flexibility, bioinspired polymers can address the degradability and environmental persistence of conventional battery components, while achieving a circular material economy (31–36). Recently, we constructed polypeptides carrying redox-active viologen or TEMPO pendant groups as organic anode and cathode materials, respectively, which were studied as electrode components of both solid-state (31) and redox flow batteries (32) (Fig. 1A). These polypeptides were shown to serve as active materials during battery operation and undergo degradation by acidic hydrolysis, showcasing their adaptability and inherent degradability. Despite the polypeptides having naturally sourced, degradable backbones and reasonable energy storage performance, the redox-active moieties were petrochemically sourced and exhibited cytotoxicity, particularly the viologen polypeptide (37, 38). Therefore, we sought to enhance the renewable sourcing and safety for the anodic polypeptide materials, by shifting from viologens to nontoxic, natural product-based redox-active building blocks.

Fig. 1.

(A) Structural design of polypeptides carrying redox-active pendant groups. (B) Riboflavin was selected as a potential electroactive alternative to viologen to enhance the sustainability and biocompatibility of the resulting redox-active polypeptide.

Inspired by electron transport reactions in biological systems (39, 40), it was hypothesized that naturally occurring redox-active species could be used as pendant groups in polypeptide-based electrode materials. Prior work has demonstrated that naturally occurring motifs with inherent redox activity, such as flavins (39, 41–43), catechols (11, 44, 45), niacin (46), ascorbic acid (47, 48) and quinones (7, 10, 39, 49, 50), could be directly incorporation into polymers, eliminating the need for complicated multistep syntheses of redox-active moieties. Among these, riboflavin stands out as a naturally occurring alternative to viologen, exhibiting a relatively low redox potential and moderate theoretical capacity, while also offering the benefits of biocompatibility and low toxicity (Fig. 1B). Additionally, the redox behavior of riboflavin is highly sensitive to the electrochemical environment and careful selection of operating conditions can be used to tune the electrochemical performance (51–54).

Guided by these concepts, a bioinspired riboflavin-containing polyglutamate was designed and synthesized for potential use as an anode material, with investigation into its electrochemical properties in the solution state, evaluation of cytotoxicity, examination of degradability under hydrolytic acidic or enzymatic environments, and determination of degradation products. While this study does not include the construction of a complete energy storage device, the development of such an electroactive material provides advances toward meeting the increasing demand for renewable energy and mitigating environmental impacts.

Results and Discussion

Design of a Riboflavin-Derived Redox-Active Polypeptide.

In pursuit of an anode material derived from bioresources, including riboflavin as pendant groups along the backbone of a polypeptide, riboflavin was readied for copper(I)-catalyzed alkyne-azide cycloaddition (CuAAC) click reaction by protection and installation of an alkynyl group, and a polypeptide was constructed having an azido group at each repeat unit; the two components were then coupled together (Fig. 2). Riboflavin first underwent acetylation to protect the hydroxyl groups, followed by nucleophilic substitution to install the propargyl moiety. For the synthesis of the polypeptide, L-glutamic acid was utilized and a 9-carbon (C9) aliphatic spacer length was employed to allow for sufficient distance between the resulting polyglutamate backbone and the sterically bulky riboflavin redox-active side chain groups. L-Glutamic acid underwent esterification to incorporate a chloro-terminated C9 aliphatic side chain. Subsequently, a ring-closure reaction was utilized to establish the NCA cyclic monomer, followed by ROP to obtain the poly(α-L-glutamate) scaffold. To ensure efficient conjugation of redox-active motifs, the chloro side chains were converted into clickable azido groups. Multiple propargyl-bearing riboflavin derivatives were then clicked onto the azido-functionalized polyglutamate by CuAAC. As shown in Fig. 2, the multistep synthesis led to the development of an L-glutamate-based polypeptide carrying redox-active riboflavin units. Through careful manipulation of chemical reactions and functional group transformations, this synthetic strategy enabled access to a riboflavin-grafted polypeptide with defined functionality.

Fig. 2.

Synthetic schemes for the stagewise and overall route for the production of a riboflavin-functionalized redox-active polypeptide as an electroactive material.

Synthesis and Physicochemical Characterization of a Riboflavin-Derived Redox-Active Polypeptide.

Synthesis of the alkynyl-functionalized riboflavin involved two steps: protection followed by propargylation. As riboflavin is poorly soluble in typical organic solvents, such as dichloromethane (DCM), chloroform, tetrahydrofuran (THF), acetonitrile (ACN), and N,N-dimethylformamide (DMF), the hydroxyl groups on the ribityl tail were protected with acetyl groups using acetic anhydride and pyridine, leading to the formation of riboflavin tetraacetate (RBTA, 1) (55, 56). This acetylation step not only improved solubility and photostability of riboflavin but also circumvented undesired side reactions during the propargylation (57). The protected flavin was then derivatized at the flavin imide moiety with a propargyl substituent, affording 3-propargyl RBTA, 2.

The N-carboxyanhydride (NCA) cyclic monomer, γ-(9-chlorononyl)-L-glutamate NCA (CNLG NCA, 4), was synthesized by Fisher esterification of L-glutamic acid with 9-chloro-1-nonanol to afford γ-(9-chlorononyl)-L-glutamate (CNLG, 3), followed by cyclocarbonylation with phosgene (generated in situ from triphosgene) in the presence of propylene oxide as an HCl scavenger. This NCA synthesis method, based on Lu’s protocol (58), enabled the formation of 4 without extensive purification and with the avoidance of the need for strictly anhydrous and air-free conditions. The chemical structure of the resulting monomer was characterized by NMR and attenuated total reflectance-Fourier transform infrared (ATR-FTIR) spectroscopies (SI Appendix, Figs. S4 and S12).

The chloro-containing polypeptide, poly(α-(γ-(9-chlorononyl))-L-glutamate)₅₀ (PCNLG₅₀, 5), was constructed by N₂-flow accelerated ROP (59) of CNLG NCA, 4 in anhydrous DMF, with neopentylamine as the initiator (monomer/initiator = 50:1). The reaction was monitored using ATR-FTIR spectroscopy until the anhydride absorbance bands at 1,852 cm⁻¹ and 1,767 cm⁻¹ disappeared (60) to ensure complete monomer conversion. The polymerization reaction was then quenched by addition of acetic anhydride to acetylate the N terminus of PCNLG₅₀, 5 and prevent potential reactions of the terminal amine, such as intramolecular cyclization with side chain ester groups. MALDI-TOF analysis confirmed the end-group identity (SI Appendix, Fig. S17) (61). According to size exclusion chromatography (SEC) analysis, this well-defined polypeptide exhibited a monomodal molar mass distribution with a number-average molar mass (M_n) of 15.5 kDa and a dispersity (Đ) of 1.13 (SI Appendix, Fig. S13). A number-average degree of polymerization (DP_n) of ca. 50 was determined from ¹H NMR spectroscopy, by comparing the integration values of the tert-butyl protons (ca. 0.89 ppm) of the neopentylamido α-chain terminus and the methine protons (ca. 4.61 ppm) of the repeat units (SI Appendix, Fig. S5).

To install redox-active riboflavin motifs through CuAAC chemistry, the chloro side chains of 5 underwent S_N2 reaction with NaN₃ to obtain an azide-functionalized polypeptide, poly(α-(γ-(9-azidononyl))-L-glutamate)₅₀ (PANLG₅₀, 6) (62). Quantitative conversion was confirmed by an upfield shift of the methylene protons adjacent to the side-chain chloro vs. azido groups in the ¹H NMR spectra (SI Appendix, Figs. S5 and S6), in agreement with elemental analysis results (SI Appendix, Table S1). Additionally, FTIR spectroscopy (SI Appendix, Fig. S12) confirmed the presence of a characteristic azide absorbance at ~2,100 cm⁻¹. Maintenance of α-helical conformations for the polypeptides in the solid state before and after transformation of the chloroalkyl side chains to azidoalkyl groups was indicated by the presence of amide absorbances at 1,651 cm⁻¹ (amide I region) and 1,543 cm⁻¹ (amide II region) in the FTIR spectra for both PCNLG₅₀, 5, and PANLG₅₀, 6 (SI Appendix, Fig. S12) (63, 64). Circular dichroism (CD) spectra also confirmed both 5 and 6 adopted α-helical conformations in THF, with characteristic negative peaks at 208 nm (π–π*) and 222 nm (n–π*) (SI Appendix, Fig. S15). This azido-functionalized polyglutamate PANLG₅₀, 6 (M_n,SEC = 17.4 kDa, Đ = 1.11) was then clicked together with propargyl-bearing flavins 2 by CuAAC.

Given the structural complexity of the riboflavin moiety, the conditions for the CuAAC reaction were not straightforward. The click reaction of 6 with 2 was initially attempted in anhydrous DMF with CuSO₄⋅5H₂O, which was reduced in situ by sodium ascorbate (NaAsc) to form the catalytically active Cu(I) species (65). To drive the coupling reaction to completion, a slight excess of 2 was used (1.2 equiv.), and when the reaction was conducted on scales larger than ~100 milligrams, elevated temperatures were required. Despite achieving efficient conjugation using CuSO₄ with NaAsc, the molar mass distribution was multimodal and broadened significantly (Đ ~ 3) (entry 1 in Table 1 and Fig. 3 and SI Appendix, Fig. S14), even at room temperature. These observations suggested the occurrence of side reactions, possibly arising from copper-mediated oxidative coupling (66). We also postulated that solubility issues could lead to partial aggregation in DMF (67); which would present as higher hydrodynamic volume species at shorter retention times in the SEC traces. To determine whether side reactions and/or aggregation was involved, CuAAC coupling of 9, a small molecule analog of 6, with 2 under the CuSO₄/NaAsc conditions was studied, which indicated that in addition to the expected product 10, trace amounts of other products were observed by thin-layer chromatography. Those side products were not isolated, however, simply treating 2 with the CuSO₄/NaAsc conditions led to no observable reaction, indicating that the propargyl-functionalized riboflavin moieties remain stable under these conditions but in the presence of azide, small amounts of apparent side reactions may occur.

Table 1.

Data table details for the CuAAC reaction conditions, yields, and dispersity values for coupling of the alkynyl-functionalized riboflavin, 2, onto the azido-functionalized polypeptide, 6

Entry*	Copper source (5 mol%)	Reducing agent† (5 mol%)	Ligand‡ (10 mol%)	Condition	Yield (%)	Dispersity§(Đ)
1	CuSO4	NaAsc	–	r.t., 24 h	78	2.94
2	CuI	–	–	r.t., 20 h	<5	–
3	CuI	–	–	40 °C, 3 d	98	1.45
4	CuI	–	THPTA	r.t., 2 h	NR	–
5	CuI	–	THPTA	r.t., 20 h	77	1.29
6	CuI	–	THPTA	r.t., 3 d	76	1.47
7	CuI	–	THPTA	40 °C, 3 d	88	1.45
8	CuI	–	PMDETA	r.t., 20 h	70	1.47

^*Reaction mixtures were degassed by performing four freeze–pump–thaw cycles before being carried out in anhydrous DMF under nitrogen atmosphere.

^†NaAsc = sodium ascorbate.

^‡THPTA = tris(3-hydroxypropyltriazolylmethyl)amine, PMDETA = N,N,N’,N’’,N’’-pentamethyldiethylenetriamine.

^§Dispersities of polymers were determined by size exclusion chromatography using DMF with 0.01 M LiBr as the eluent, calibrated with poly(methyl methacrylate) (PMMA) standards.

Fig. 3.

SEC traces of the resulting riboflavin-functionalized polypeptides, 7 from different CuAAC conditions listed in Table 1.

To address this synthetic hurdle and avoid these potential complications, several alternative CuAAC catalyst systems were examined. By screening various conditions, a change to copper iodide (CuI) with the addition of ligands, including tris(3-hydroxypropyltriazolylmethyl)amine (THPTA) or N,N,N′,N′′,N′′-pentamethyldiethylenetriamine (PMDETA), was found to accelerate the reaction, enhance the selectivity, and minimize side reactions. Consequently, CuAAC could be carried out at room temperature with high conversions and minimized observance of apparent higher molar mass products (entries 5 to 8 in Table 1 and Fig. 3). In the best case of the tested conditions, using 5 mol% CuI and 10 mol% THPTA at room temperature was found to be an effective combination of catalyst system (Fig. 4A, entry 5 in Table 1 and Fig. 3), which enabled the synthesis of riboflavin polypeptide (7) as a yellow solid having the lowest dispersity (M_n,SEC = 30.8 kDa, Đ = 1.29) and least pronounced high molar mass shoulder, as observed by SEC at shorter retention time. This set of experimental conditions and 20 h reaction time provided an isolated yield of 77% after purification by filtering through celite to remove CuI, precipitating in water to remove DMF and THPTA, dissolving in DCM and precipitating in diethyl ether to remove residual 2. In contrast, using 5 mol% CuI in the absence of ligand at r.t. gave trace yield after 20 h (entry 2 in Table 1). Elevation of the temperature to 40 °C for 3 d resulted in significantly greater high molar mass product without or with ligand (entry 3 and entry 7 in Table 1, respectively). A shorter reaction time (2 h) with 5 mol% CuI and 10 mol% THPTA led to negligible conjugation (entry 4 in Table 1). Use of PMDETA rather than THPTA as the ligand gave higher Đ (entry 8 in Table 1 and Fig. 3; the apparent lower yield for this experiment may have been partially due to precipitation into diethyl ether twice in an attempt to remove the high molar mass species).

Fig. 4.

(A) CuAAC of 6 and 2 conducted at room temperature under N₂ atmosphere using 5 mol% CuI and 10 mol% THPTA, and images of the resulting product under visible and UV light. (B) ¹H NMR spectra (500 MHz, DMSO-d₆) of PANLG₅₀ (6, black trace), 3-propargyl RBTA (2, red trace), and riboflavin polypeptide (7, blue trace). (C) ATR-FTIR spectra of PANLG₅₀, 3-propargyl RBTA, and riboflavin polypeptide. The highlighted region showed the disappearance of the azide peak from 6 and the terminal alkyne peak from 2 after CuAAC.

The chemical composition and structure of the riboflavin polypeptide were characterized by several spectroscopic techniques. Although the solubility of the precursor azido-functionalized polypeptide was limited in DMSO-d₆, giving broadened signals relative to the ¹H NMR spectrum collected in CDCl₃-TFA-d (SI Appendix, Fig. S6), the overlaid ¹H NMR spectra of each of the polypeptide and riboflavin precursor along with the final riboflavin-conjugated polypeptide in a consistent DMSO-d₆ NMR solvent are included in Fig. 4B for direct comparison. The appearance of the 1,2,3-triazole proton resonating at 7.7 ppm (peak d of 7, Fig. 4 B, Bottom), together with downfield shifts for the methylene protons adjacent to the alkyne from 4.6 (peak a of 2, Fig. 4 B, Middle) to 5.1 ppm (peak e of 7, Fig. 4 B, Bottom) and the azide groups from 3.2 (peak of 6, Fig. 4 B, Top) to 4.2 ppm (peak s ▲ of 7, Fig. 4 B, Bottom) confirmed the formation of the desired riboflavin-conjugated polypeptide. By FTIR spectroscopy, disappearance of the azide peak at 2,100 cm⁻¹ (Fig. 4 C, Top vs. Bottom) further suggested the transformation of the azide moieties to the triazole groups, while the absence of the alkynyl C ≡ C stretch in the FT-IR spectrum of 7 confirmed an absence of residual 2. Grafting of the riboflavin moieties was quantitative, as evidenced by the elemental analysis results (SI Appendix, Table S1). Additionally, an obvious peak shift to shorter retention time for the SEC profiles of 6 vs. 7 was observed, indicating that coupling of the large riboflavin moieties gave 7 a greater hydrodynamic diameter than its precursor polypeptide 6, and confirming modification of the side chains was successful (SI Appendix, Fig. S13). The absorption and fluorescence behaviors of 7 were investigated in chloroform, revealing the characteristic absorption and emission patterns of flavin species (SI Appendix, Fig. S18) (68). Specifically, the absorption spectrum of 7 displayed peaks at 350 and 450 nm, and upon excitation at 450 nm, the emission spectrum exhibited a peak at 506 nm. Interestingly, unlike the precursor polypeptides, 5 and 6, 7 did not exhibit a stable helical conformation as measured by CD in organic solutions [THF, ACN, or hexafluoroisopropanol (HFIP)] (SI Appendix, Fig. S16), likely due to disruptions of intramolecular backbone hydrogen bonding and steric bulkiness from side chain riboflavin substituents (69).

In parallel, a small-molecule analog (10) mimicking the repeat unit of riboflavin polypeptide was synthesized to evaluate CuAAC reaction conditions (vide supra), guide structural determination, investigate redox potentials, and predict degradation products. Synthetic details of 10 are available in the Supporting Information (SI Appendix, Scheme S1), with structural characterizations conducted by ¹H and ¹³C NMR (SI Appendix, Figs. S8–S10), ATR-FTIR (SI Appendix, Fig. S11), UV-Vis, and photoluminescence (PL) (SI Appendix, Fig. S18) spectroscopies.

Electrochemical Characterization in the Solution State.

Electrochemical studies of the riboflavin polypeptide in solution were performed using cyclic voltammetry (CV) in a metal-free supporting electrolyte of 0.5 M tetrabutylammonium hexafluorophosphate (TBAPF₆) in propylene carbonate (PC). The three-electrode setup consisted of a glassy carbon working electrode, a coiled platinum wire counter electrode, and an Ag/Ag⁺ reference electrode (0.01 M AgNO₃ and 0.1 M TBAClO₄ in ACN). Prior to measurement, the 1 mM riboflavin polypeptide solution was preconditioned with CV cycles at 10 mV⋅s⁻¹ (3×) (SI Appendix, Fig. S21), followed by a series of sequential CV cycles (2×) at variable scan rates (10, 25, 50, and 100 mV⋅s⁻¹) (Fig. 5A), under a nitrogen atmosphere. At 10 mV⋅s⁻¹ (Fig. 5A), the CV showed a quasireversible redox couple with a half-wave potential (E_1/2) of ca. −1.10 V vs. ferrocene/ferrocenium (Fc/Fc⁺) and a peak separation (ΔE_p) of ca. 0.10 V. The peak currents are linearly proportional to the square root of the scan rate for both oxidation and reduction processes (Fig. 5B), revealing that the electrochemical reaction was diffusion-controlled (70). The observed E_1/2 of the redox couple is in agreement with literature reports for (54, 71) for the consecutive (stepwise) proton-coupled electron transfer (PCET) involving one-electron reduction/oxidation of riboflavin to the flavosemiquinone radical anion in aprotic organic environments (Fig. 5C). Potentials of less than −1.5 V vs. Fc/Fc⁺ are required to access the second redox couple, which falls outside the scope of this study. Additionally, solution-state CVs of the riboflavin precursor 2 and riboflavin analog 10 showed similar CV responses, suggesting that the nonconductive polypeptide backbone and side chain linker did not impact the redox potentials of the pendant riboflavin motifs (SI Appendix, Fig. S22 and Table S2).

Fig. 5.

(A) Cyclic voltammograms of riboflavin polypeptide (1 mM in 0.5 M TBAPF₆/PC) collected at different scan rates (2×, with the second of each scan rate cycle being shown), after a 10 mv/s (3×) preconditioning. (B) Peak current vs. square-root of scan rate from the cyclic voltammograms. (C) Redox state of riboflavin (neutral form) and flavosemiquinone (radical anion). (D) Cottrell plots from chronoamperometry of riboflavin polypeptide. The Inset shows the potential steps and the current responses.

To further understand the electron transfer process of the riboflavin polypeptide in aprotic organic solution, the electrochemical kinetics during oxidation and reduction were experimentally examined by chronoamperometry (CA) and further analysis of the CV data (SI Appendix, Table S3). CA was recorded for both reductive and oxidative potential steps (step 1 = −0.3 to −1.5 V for 2 s, step 2 = −1.5 to −0.3 V for 2 s, SI Appendix, Fig. S23). The resulting current densities were plotted against t^−1/2 to obtain the Cottrell plots (Fig. 5D), from which the values of the apparent diffusion coefficient (D_app) were determined. Under the assumption of semi-infinite diffusion, linear fits were applied to Cottrell plots at t ≤ 0.1 s, and the resulting slopes were used to calculate D_app following the Cottrell equation (72). The D_app values are (3.49 ± 0.12) × 10⁻⁷ for oxidation and (6.07 ± 0.76) × 10⁻⁷ for reduction, which are comparable to literature-reported values of other nonconjugated redox-active polymers (73–77).

To further consider the electron transfer mechanism for the riboflavin polypeptide, the modes of diffusion of the riboflavin polypeptide must be identified. Since the electron transfer was evaluated in solution state, where the riboflavin polypeptide was able to freely diffuse, the physical diffusion coefficient (D_phy) of the polymer should be determined. If D_phy is greater than the measured D_app, D_phy must be decoupled from D_app to solely evaluate the electron hopping, since D_app reflects the combined effects of physical diffusion and electron hopping (self-exchange) (4). Therefore, the value of D_phy was estimated by measuring the hydrodynamic radius (r_h) of the riboflavin polypeptide using dynamic light scattering (DLS) and applying the Stokes–Einstein equation. Although DLS does not directly measure the mobility of the redox-active pendant groups, we approximated their diffusion coefficients by assuming they diffused with the polymer backbone, and thus shared the same diffusion behavior measured by DLS (73). From DLS, the D_phy of the riboflavin polypeptide in solution (1 mM riboflavin polypeptide in 0.5 M TBAPF₆/PC) was on the order of 10⁻⁷ cm²⋅s⁻¹, which is comparable to D_app determined using CA. The similarity between D_phy and D_app suggests that the diffusion of the riboflavin polypeptide had a significant impact on the measured D_app values, so D_phy could not be ignored when determining the rate constants for electron transfer.

Knowing both D_app and D_phy must be accounted for, the rate constants for homogeneous hopping (k_ex) and heterogeneous (k⁰) charge transfer were determined. To understand charge hopping between neighboring redox species, k_ex was determined using the Dahms–Ruff equation, with the average hopping distance approximated by (C_EN_A)^−1/3 (78, 79). k_ex was determined to be (8.07 ± 0.62) × 10¹⁰ M⁻¹⋅s⁻¹ for oxidation and (2.17 ± 0.33) × 10¹¹ M⁻¹⋅s⁻¹ for reduction. To evaluate charge transfer between the active material and the current collector, k⁰ was calculated using the Nicholson method with the peak separation (ΔE_p) from CV at 10 mV⋅s⁻¹ (80). The k⁰ values were found to be within the scale of 10⁻⁴ cm⋅s⁻¹, with the values of (3.85 ± 0.02) × 10⁻⁴ cm⋅s⁻¹ for oxidation and (5.08 ± 0.43) × 10⁻⁴ cm⋅s⁻¹ for reduction. The detailed calculations of kinetic parameters (D_app, D_phy, k_ex, and k⁰) are provided in the SI Appendix.

Degradation Studies.

To address environmental concerns associated with the persistence and accumulation of plastics or microplastics, the riboflavin polypeptide 7 was designed to be degradable. As a result, identifying its degradation pathways under hydrolytic environments is important to ensure effective breakdown. In this study, acidic and enzymatic degradation methods were explored to investigate the degradation behaviors of riboflavin polypeptide in different media.

Acidic degradation was accomplished by suspending 7 as a powder at 1 mg/mL in 1 M HCl and magnetically stirring the solution at elevated temperature. Precautions included the use of a portable blast shield in addition to the chemical fume hood sash and performing the degradation within a vial of 40× volume vs. the solution, which was securely clamped within an oil bath heated at 110 °C. Over a period of 24 h, the initial suspension gradually transformed into a homogeneous yellow solution. The solution was then neutralized to pH 7 using 1 M NaOH and the degradation products were analyzed by liquid chromatography–mass spectrometry (LC–MS) (Fig. 6A).

Fig. 6.

(A) A schematic diagram of the acidic degradation study of riboflavin polypeptide (7). Figures were made at BioRender.com. (B) Degradation species detected by LC–MS.

The degradation products, arising from the hydrolytic cleavage of amide linkages in the backbone and ester bonds in the side chains, were separated and detected (Fig. 6B and SI Appendix, Figs. S24 and S25). The side chain-derived products were separated using a C18 reversed-phase column, while the backbone degradation product, L-glutamic acid, was selectively isolated using a hydrophilic interaction liquid chromatography (HILIC) column for polar amino acid separation. The polypeptide backbone was hydrolyzed into L-glutamic acid P1, whereas the ester bonds at the side chains were cleaved and formed the riboflavin derivative P2, which further transformed into the chlorinated derivative P3 in the presence of HCl_(aq). In addition, these flavin derivatives resulting from side chain degradation were light-sensitive under acidic media. These side chain degradation products underwent photodegradation processes (81), leading to the formation of various photoproducts, such as lumichrome compounds (P4 and P5). Due to the complexity of degradation pathways, a full characterization of all degradation products was not feasible. A parallel study of the riboflavin analog 10 was conducted (SI Appendix, Fig. S26), which revealed similar degradation patterns.

While acidic degradation effectively facilitates polymer hydrolysis, it typically requires harsh reaction conditions, such as elevated temperatures at extreme pH levels. These conditions could introduce complications in degradation and generate unwanted degradation species, posing challenges for raw material recovery. One of the alternative approaches for conquering these limitations is to incorporate enzymatically degradable motifs within the polymer backbones or side chains. Enzymatic degradation of polymers can be operated under milder conditions, such as neutral pH and ambient temperatures, to achieve a more selective and controlled polymer breakdown. As such, assessing degradability of the riboflavin polymer 7 was of interest.

First, a preliminary study was performed using the riboflavin analog 10 and esterase, with an enzymatic activity of ca.16 U/mL, suspended in a phosphate-buffered saline solution (PBS, pH = 7.4). Porcine liver esterase was selected for this small-molecule study, to mimic the enzymatic cleavage of ester bonds in the side chains of 7. As the mixture was incubated for 24 h in a shaker (37 °C, 60 rpm), the original turbid suspension gradually turned into a homogeneous yellow solution, consistent with previous observations from acidic degradation studies. Subsequently, the mixture was purified to remove the enzyme, and the supernatant was analyzed by LC–MS (SI Appendix, Fig. S27). As expected, there were two major degradation products detected (SI Appendix, Figs. S28-S29), revealing that esterase efficiently cleaved the ester bonds with minimal formation of secondary degradation products.

Following this encouraging result, the same enzymatic degradation method was applied to the riboflavin polypeptide 7, with expectations of a similar controlled degradation outcome. However, the degradation proceeded significantly slower and identification of the degradation products was inconclusive by LC–MS, likely indicating more complex enzyme–substrate interactions with the polypeptide than with the small-molecule analog. This complexity is attributed to the hydrophobic nature of 7 and the introduction of bulky pendant groups, all of which can contribute to slower and more limited degradation of the side chains (82, 83). With this in mind, future studies should focus on optimizing enzymatic degradation conditions for this polypeptide system (e.g., pH, buffer, choice of enzyme), or better design of materials, including the incorporation of hydrophilicity or electrostatic charges (84), to achieve efficient interactions and enzymatic degradation.

Cell Viability Studies.

Cell viability testing with mouse fibroblast cells (NIH/3T3) was performed to assess the cytotoxicity of the riboflavin polypeptide 7, its repeat-unit analog 10, and their products from acidic aqueous degradation. The riboflavin polypeptide did not exhibit cytotoxic effects toward fibroblast cells after 72 h of exposure. Even at a concentration as high as 100 µg/mL (achieved by transitioning from DMSO to cell media), cell viability was nearly 80%, exceeding the ISO 10993-5 biocompatibility threshold of 70% (SI Appendix, Fig. S30) (85). In contrast, the previously reported analogous viologen polypeptide exhibited an IC₅₀ of only 2.68 µg/mL (31). This result highlights that the incorporation of natural redox-active moieties provides a strategy to decrease the cytotoxicity in organic redox-active materials. Compared to the riboflavin polypeptide 7, the repeat-unit analog 10 showed reduced cell viability (IC₅₀ = 33.2 µg/mL, SI Appendix, Fig. S30), which is still higher than that of viologen polypeptide. To evaluate cytotoxicity of the degradation products, Live/Dead and MTS assays were performed on cells exposed to degradation products from 7 and 10 prepared by acidic hydrolysis (1 M HCl, 110 °C, 24 h). In each case, the degradation products revealed viability greater than 70% toward fibroblast cells, as verified visually in fluorescence confocal imaging (Fig. 7A) and quantitatively through MTS cell viability measurements (Fig. 7B). Overall, these findings indicated the riboflavin polypeptide is biocompatible, with no signs of cytotoxicity in its intact and degraded forms. This study highlights the importance of rational raw material selection in the development of redox-active polymers to promote the transition toward sustainable energy storage.

Fig. 7.

(A) Fluorescence confocal microscopy images of Live/Dead test for the degradation products at 100 µg/mL. Images show live cells (green, stained with calcein AM) and dead cells [red, stained with ethidium homodimer-1 (EthD-1)] of fibroblast cells after incubation with degradation products of riboflavin polypeptide, 7, riboflavin analog, 10, and control (unexposed). (Scale bar, 200 µm.) (B) Quantitative cell viability assessment by MTS assay for the degradation products at 100 µg/mL. Viability percentages were normalized to the control group, calculated from different confocal microscopy areas with error bars showing SD (n = 3).

Conclusions

This work advances the design of organic energy materials by introducing an electroactive polymer composed of naturally sourced components for both the framework of the polymer backbone and the redox-active side chain substituents. In comparison to our earlier report of redox-active polypeptide battery materials, this current system incorporates riboflavin as a naturally sourced and nontoxic alternative to the more common, petrochemically sourced viologen. While the theoretical capacity of the riboflavin polypeptide is 61.0 mAh⋅g⁻¹, which is lower compared to the previously studied viologen system at 118.3 mAh⋅g⁻¹, this shift toward bioinspired redox functionality enhances sustainability and promotes biocompatibility. In addition, the resulting polypeptide maintained the characteristic ability to undergo hydrolytic degradability under acidic environments, reinforcing its potential as a degradable component of electronic devices with minimal end-of-life concerns.

While the riboflavin polypeptide showed an appropriate redox potential and appreciable degradability, the integration of riboflavin moieties presented several new challenges. The steric bulk and multifunctionality of riboflavin complicated its conjugation using copper-catalyzed alkyne-azide click chemistry, disrupted the backbone helical conformation, and hindered enzymatic degradability. Furthermore, its light sensitivity led to photoproduct formation during hydrolytic degradation, which made recovery and recycling of natural building blocks difficult. Overcoming these hurdles will require molecular redesign, including developing alternative conjugation strategies or incorporating other bioderived natural products to balance synthetic chemistry with material stability and degradability. One promising approach may involve the intentional production of partially conjugated polypeptides, wherein a fraction of the side chains remains unmodified to promote helix formation, inhibit uncontrolled aggregation, preserve macromolecular order, and enhance enzymatic degradability. In combination, alternate linker lengths (17) may lead to an optimization of intramolecular side chain interactions to promote helicity. Restoring the secondary structure would further enable investigations into the role of the helical dipole moment in modulating electrochemical behavior (35). Despite these considerations, this current riboflavin-functionalized polypeptide represents a promising direction of electrode design for energy applications.

While the current study focuses on organic anode development and characterization, transition of this redox-active polypeptide into a functional battery or other electronic device will require significant additional work. Next steps are underway to determine how these materials perform as components within a larger and more complex full battery device, which includes a cathode, an electrolyte, a separator, and current collectors in addition to this redox-active anodic material. Expanding to other electronic device applications will constitute the foci of several future investigations. Further evaluation will be pursued with goals to realize their potential sustainability attributes and environmental impacts, while also expanding their potential applications toward wearable and implantable electronic materials (86, 87).

Materials and Methods

See SI Appendix for the Materials, Methods, Analysis, Synthetic Protocols, Electrochemical Kinetic Parameters Calculations, enzymatic degradation study of riboflavin analog 10, and cell viability study.

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

All study data are included in the article and/or SI Appendix.