A Green and Selective Electrochemical Depolymerization of Chitosan to Glucosamine and N-Acetylglucosamine via GONOGO-Controlled Pulsed Electrolysis
Hossein Mojtabazadeh a*, Javad Safaei-Ghomi a*,
a Department of Organic Chemistry, Faculty of Chemistry, University of Kashan, Kashan 87317-51167, Iran
*Corresponding authors: Hossein Mojtabazadeh (M.M.Ebadi@grad.kashanu.ac.ir) ; Javad Safaei-Ghomi (safaei@kashanu.ac.ir)
ORCID iDs:
Hossein Mojtabazadeh: 0009-0005-8014-442X
Javad Safaei-Ghomi: 0000-0002-9837-4478
Abstract
KEYWORDS:
Chitosan; Electrochemical depolymerization; N-Acetylglucosamine.
1. Introduction
The selective depolymerization of chitosan into structurally defined monomeric units remains a central challenge in sustainable biopolymer valorization. Chitosan is a partially N-acetylated β(1→4)-linked polysaccharide composed of glucosamine (GlcN) and N-acetylglucosamine (GlcNAc) units and is valued for its abundance, biocompatibility and wide industrial relevance [1–4]. Although numerous chemical and electrochemical methods can reduce its molecular weight, they generally generate broad oligomeric mixtures and provide limited control over the preservation of labile N-acetyl groups. As a result, existing approaches do not achieve monomer-level selectivity.
Recent progress in electro-organic synthesis has shown that temporal modulation of electron flow can unlock reactivity patterns that are not accessible under direct current or constant-potential electrolysis. Waveform-controlled strategies such as pulsed direct current (pDC) [5], pulsed step constant potential (pSCP) [6], rapid alternating polarity (rAP) [7] and pulsed alternating current (pAC) [5,8] have demonstrated improved mass transport, intermediate stabilization and double-layer organization relative to static electrolysis [9–11]. However, all these waveforms maintain continuous electrode bias or rapid polarity switching, which means the electrode never enters a true open-circuit state. This limitation prevents full relaxation of the electrical double layer and restricts diffusion-driven clearance of reactive intermediates, both of which are critical for preserving sensitive motifs such as N-acetyl groups during glycosidic cleavage.
Figure 1 summarizes the operational characteristics of these classical waveforms. In pDC and pSCP (Figure 1a–b), the electrode remains continuously biased between predefined current or potential levels. In rAP and pAC (Figure 1c–d), anodic and cathodic states alternate rapidly but no zero-current interval is introduced. These designs promote partial interfacial relaxation at best, and prolonged bias accelerates overoxidation pathways that compromise the integrity of N-acetylated residues.
Figure 1. Comparison of classical electrochemical waveforms with the GONOGO protocol. (a) pDC and (b) pSCP maintain continuous electrode bias without permitting collapse of the electrical double layer. (c) rAP and (d) pAC alternate between anodic and cathodic polarization but still lack a true open-circuit interval. (e) GONOGO introduces a unipolar Go phase followed by a genuine open-circuit No-Go interval, enabling full relaxation of the interfacial double layer and diffusion-driven clearance of reactive species. Interfacial parameters including double-layer capacitance (Cdl) and charge-transfer resistance (Rct) recover only during the No-Go period, illustrating the logic-controlled nature of the GONOGO waveform.
To address this limitation, we introduce GONOGO, a logic-gated electrolysis protocol that alternates between asymmetric Go (electrolysis) and No-Go (true open-circuit) phases. Unlike traditional pulsed methods defined solely by pre-set on/off durations, GONOGO incorporates genuine zero-current intervals whose effective timing matches the intrinsic relaxation kinetics of the system. The No-Go interval allows complete collapse of the electrical double layer, dissipation of reactive species and unforced macromolecular diffusion. These processes are not achievable under conventional pulsed or alternating waveforms (Figure 1e). This design provides precise spatiotemporal separation of radical generation from interfacial oxidation and therefore favors preservation of N-acetylated units during β(1→4) cleavage. As a result, GONOGO represents a distinct class of adaptive, logic-controlled electrolysis.
The feasibility of electrochemical depolymerization of chitosan has been demonstrated using direct current approaches. Early anodic studies achieved glycosidic scission but suffered from continuous overoxidation and formation of low-molecular-weight fragments [12,13]. Our earlier NHPI-mediated constant-current platform showed that PINO radicals can selectively cleave β(1→4) linkages to produce crystalline GlcN·HCl under mild conditions [4]. This established the viability of radical-mediated depolymerization but also highlighted the need for improved selectivity toward N-acetylated residues due to the persistent electrode bias.
Beyond chitosan, electrochemical depolymerization has also been reported for cellulose, alginate and pectin, primarily through anodic oxidation or mediator-assisted pathways in which electrogenerated reactive oxygen species promote backbone scission [14–17]. These reports confirm that polysaccharides are susceptible to controlled electrochemical cleavage, although all existing methods rely on continuous-bias electrolysis. To the best of our knowledge, no pulsed or time-segmented electrolysis strategy has been applied to any polysaccharide system, and no method has demonstrated selective preservation of labile functional motifs such as N-acetyl groups. This gap underscores the need for a waveform design capable of decoupling radical generation from overoxidation while enabling full interfacial relaxation.
Taken together, these observations show that temporal segmentation alone is insufficient and mechanistic decoupling is essential for selective depolymerization of partially acetylated polysaccharides. GONOGO fulfills this requirement through unipolar electrolysis, genuine open-circuit relaxation and decision-gated synchronization. In this work, we demonstrate that GONOGO enables selective and tunable co-production of GlcN and GlcNAc from chitosan, an outcome not achieved with any previous electrochemical method. A detailed mechanistic comparison and statement of novelty are provided in Supporting Information Section S1.
2. Materials and methods
2.1. Materials
All reagents were of analytical grade and were used as received unless otherwise noted. N-Hydroxyphthalimide (NHPI, ≥98%), glacial acetic acid, sodium acetate, hydrochloric acid (37%), sodium hydroxide, and butylated hydroxytoluene (BHT, ≥99%) were obtained from Merck (Germany) or Sigma-Aldrich (USA). Ethanol and deionized water were used as solvents where indicated. Electrochemical reactions were carried out in an undivided cell with titanium electrodes (Grade 2, 99.5%, 2 × 2 cm) as both anode and cathode. Before use, electrodes were polished, ultrasonically cleaned in ethanol and water (10 min each), and dried in air.
For the semi-industrial batch (1.0 kg chitosan), food/pharma-grade acetic acid (Dr. Mojalli Co., Iran, ≥98%) was used to reduce costs while ensuring compatibility. The reactor employed larger titanium electrodes (40 × 28 cm, same grade and pretreatment), and the acetate buffer was prepared using the same composition, 0.2 M acetic acid and 0.3 M sodium acetate (pH 4.5 ± 0.2). All materials were stored and handled under ambient conditions unless specified.
2.2. Instrumentation
Electrochemical experiments were performed using a potentiostat/galvanostat (Autolab PGSTAT302N, Metrohm) equipped with FRA32M for impedance analysis. Electrochemical experiments were performed in an undivided cell with grade 2 titanium plates serving as anode and cathode. For mechanistic and waveform analysis, a saturated Ag/AgCl reference electrode was included solely to monitor the interfacial potential, while current control was always applied in a two-electrode configuration. All runs were conducted in a two-electrode undivided cell under either constant-current operation or under the GONOGO pulse profile (Section 2.4), with real-time monitoring and pulse programming via NOVA 2.1. For scale-up, a custom-built galvanostatic controller (0–120 A, 0–60 V) with programmable ON/OFF logic was used to reproduce the optimized Go/No-Go pulse profile validated in laboratory experiments. Solution pH was measured using a calibrated benchtop meter (827 pH Lab, Metrohm), while temperature was maintained at 25 ± 2 °C, unless otherwise specified. Analytical characterization involved HPLC (Shimadzu LC-20AD, RP-C18, UV–Vis detection), FTIR (Bruker Tensor II, ATR, 400–4000 cm-1), NMR (1H, 13C, COSY, HSQC, Bruker AVANCE III 400 MHz, D2O), moisture determination (Karl Fischer titration, Mettler Toledo V20), trace metal quantification (ICP-OES, Agilent 5110), mass spectrometry (GC-MS, Agilent 7890A/5975C, EI), elemental analysis (CHNS, Thermo Flash 2000), and TLC (silica gel 60 F254, UV 254 nm) where applicable. Unless otherwise stated, key electrolysis experiments and analytical measurements were performed in triplicate, and results are reported as mean ± standard deviation.
2.3. Chitosan Source
Partially deacetylated chitosan (DD = 86.4%) was prepared from shrimp shells through sequential demineralization, deproteinization, and alkaline deacetylation. The full extraction methodology, analytical characterization (FTIR, NMR, CHNS), and DD determination are detailed in our earlier work [4].
2.4. Electrochemical Depolymerization Under GONOGO Conditions
Electrochemical depolymerization was carried out in an undivided two-electrode cell under galvanostatic control. Titanium electrodes (2 cm × 2 cm), selected as the optimal configuration (see Section S11 for electrode screening), served as both the working and counter electrodes. Prior to each experiment, electrodes were mechanically polished and rinsed with deionized water to remove surface oxides and ensure reproducibility.
The electrolyte comprised 100 mL of acetate buffer prepared from 0.2 M glacial acetic acid and 0.3 M sodium acetate (pH 4.5 ± 0.2). Chitosan (300 mg), pre-dissolved in 2% (v/v) aqueous acetic acid, was introduced into the electrolysis medium, followed by 10 mol% NHPI as a redox mediator.
The GONOGO protocol was implemented as an asymmetric ON/OFF regime in which each Go phase was followed by a true open circuit No Go interval. Under the mediated NHPI and Ti conditions used in this study, the system exhibited a stable relaxation pattern that consistently converged to an effective 2 s Go and 6 s No Go timing. The pulse controller was therefore set to match this experimentally determined relaxation behaviour instead of imposing arbitrary external timing.
For semi-industrial scale-up (1.0 kg chitosan), the same GONOGO profile (2 s / 6 s) was employed using a programmable galvanostatic power controller and titanium electrodes (40 cm × 28 cm). The total reaction volume was 15 L, prepared using the same acetate buffer system (0.2 M glacial acetic acid and 0.3 M sodium acetate, pH 4.5 ± 0.2), with proportional scaling of NHPI loading and substrate concentration. Electrolysis was performed for 8.5 h under constant stirring, while maintaining all other parameters constant.
Upon completion, the reaction mixture was filtered to remove unreacted solids. The resulting filtrate was directly subjected to downstream processing as described in Section 2.5.
2.5. Isolation and Characterization of Products
Upon completion of the electrolysis, the reaction mixture was gradually neutralized to pH ~7 using solid sodium bicarbonate under continuous stirring. The suspension was filtered to remove any insoluble residues. The clear filtrate was then acidified to pH 4 with concentrated hydrochloric acid, followed by the slow addition of 200 mL absolute ethanol under stirring. The mixture was left undisturbed at room temperature for 3 h to allow co-precipitation of GlcN·HCl and GlcNAc. The resulting solid was collected by vacuum filtration and dried at 55 °C.
The dried crude solid, containing a mixture of GlcN·HCl and GlcNAc, was dissolved in a minimal volume of water/methanol (7:3, v/v) and subjected to silica gel column chromatography for separation. The chromatography was performed using a glass column (60 cm length × 3.5 cm inner diameter) packed with silica gel 60 (70–230 mesh). Elution was carried out with a stepwise gradient of water/methanol (from 9:1 to 7:3, v/v). Fractions were monitored by TLC using a water/methanol (8:2, v/v) system, with ninhydrin visualization. Fractions containing pure GlcNAc and GlcN·HCl were pooled separately, evaporated under reduced pressure, and collected as purified products.
For the semi-industrial batch (1.0 kg dry chitosan in 15 L acetate buffer), the same isolation and purification procedure was employed using 30 L of absolute ethanol for the initial precipitation step. Final gravimetric yields and purities of GlcNAc and GlcN·HCl were determined using Karl Fischer titration and the USP-based HPLC method described in Section S7 and S6, the corrected dry yields were 127.20 g (12.72%, 99.1% purity) for GlcNAc and 708.35 g (70.84%, 99.2% purity) for GlcN·HCl, with no side-products detected and excellent reproducibility (RSD < 1.5%).
The identity and purity of GlcN·HCl and GlcNAc were confirmed by comprehensive spectroscopic and analytical characterization, including 1H and 13C NMR, FTIR, MS, and CHNS elemental analysis, all of which matched the expected values. Full spectral data are provided in the Supporting Information (Section S2).
3. Results and discussion
3.1. Electrochemical Design and Implementation of the GONOGO Protocol
Building upon our previous work, which demonstrated the first direct electrochemical depolymerization of chitosan into crystalline GlcN·HCl using NHPI as a recyclable redox mediator under mild galvanostatic conditions, we herein introduce a waveform-controlled electrolysis strategy to address several key limitations observed in the constant-current system. In our earlier study, although NHPI-mediated oxidation enabled efficient β(1→4) glycosidic bond cleavage via PINO radicals, the process led to complete degradation of the N-acetylated domains, limiting the formation of GlcNAc. Moreover, continuous polarization imposed radical stress on labile functional groups and accelerated electrode fouling. These drawbacks underscored the need for a dynamically regulated electrochemical approach that temporally separates radical initiation from interfacial relaxation. To that end, we designed the GONOGO protocol, a pulsed waveform capable of balancing reactivity and selectivity by integrating electrochemical “Go” and rest-phase “No-Go” intervals.
This waveform-engineered strategy implements asymmetric pulsing, wherein a 2-second anodic “Go” phase is followed by a 6 second “No-Go” interval under open-circuit conditions (Figure 1e). During the Go phase, NHPI is oxidized to PINO radicals, which initiate site-selective hydrogen abstraction and depolymerization. The No-Go phase permits relaxation of the electrical double layer (EDL), radical diffusion, and suppression of overoxidation, thus maintaining the integrity of sensitive functional groups such as N-acetyl moieties.
The underlying hypothesis is that these transient rest phases enable three synergistic processes: reorganization of the EDL to restore charge homogeneity, diffusive clearance of localized hydroxide and PINO radicals, and protection of labile functional groups from oxidative degradation. This temporal decoupling of redox activation from interfacial dynamics provides mechanistic advantages over both conventional galvanostatic electrolysis and symmetric pulsed techniques such as rAP, which do not incorporate true open-circuit intervals.
Electrochemical depolymerization under the GONOGO protocol was carried out in an undivided cell using grade 2 titanium plates as both the working and counter electrodes. A saturated Ag/AgCl reference electrode was connected only for potential monitoring during current-controlled pulsing. Titanium was chosen for its inertness, oxidative stability, and prior validation in NHPI/PINO-mediated systems [4].Electrolyses were carried out in 0.2 M acetic acid + 0.3 M sodium acetate buffer (pH ≈ 4.5), at different current densities (60–90 mA cm-2) for 2–12 hours, using a custom-built chronopotentiometric pulse controller. Parallel Galvanostatic (Galv) experiments were performed under identical conditions but without waveform control.
The chitosan substrate exhibited a DD of 86.4%, implying that 13.6% of its monomeric units retained N-acetyl groups. Under GONOGO conditions, especially with the 2 s/6 s pulse profile, up to 13.2% GlcNAc was isolated from 300 mg of chitosan, while 13.0% was obtained from 20 g (Figure 2d), and 12.7% from 1 kg of chitosan (Table S2), indicating near-complete retention of N-acetylated domains. In contrast, galvanostatic resulted in complete acetyl loss, yielding only GlcN, consistent with non-selective radical overexposure under continuous polarization. This demonstrates the efficacy of the GONOGO protocol in modulating radical reactivity temporally to achieve chemo-selectivity.
Electrode fouling also differed markedly between the two regimes. Under galvanostatic electrolysis, ΔRct increased by ≈ 120%, whereas under GONOGO, ΔRct increased by only ≈ 35% at laboratory scale and ≈ 47% at 1-kg scale (Table S2). This reduced fouling aligns with controlled radical flux and a more stable interfacial environment.
In addition to ΔRct, the stability of the waveform provided further mechanistic evidence. Time-resolved potential traces revealed that dV/dt during Go phases remained constant throughout electrolysis, and no drift in pulse amplitude or waveform shape was observed. Because dV/dt and pulse shape reflect double-layer charging, this invariance indicates that Cdl remained effectively constant, confirming that interfacial relaxation was fully reproducible during each No-Go interval. These signatures collectively support a stable, non-accumulating electrochemical interface under GONOGO operation. A detailed mechanistic interpretation of these parameters is provided in Section 3.5.
The electrolysis performance at different current densities (60–89 mA cm-2) showed a uniform increase in GlcN·HCl production, reaching 14.44 g at 89 mA cm-2 from 20 g of chitosan. In parallel, GlcNAc formation peaked at 2.6 g under the same conditions.
Importantly, under the optimized GONOGO conditions (2 s/6 s, 89 mA cm-2), the process was successfully scaled to a 1-kg batch, yielding 708.35 g of GlcN·HCl and 127.2 g of GlcNAc with excellent selectivity and stable electrochemical performance (Table S2, Entry 94).
Figure 2. Electrochemical depolymerization of chitosan using the GONOGO protocol. (a) Proposed mechanistic pathway: (1) anodic oxidation of NHPI to the PINO radical and initiation of hydrogen atom transfer, (2) selective β(1→4) glycosidic bond cleavage leading to β GlcN formation, and (3) preservation and direct release of β GlcNAc through controlled radical propagation, followed by mutarotation to α and β anomeric forms. (b) Asymmetric GONOGO waveform consisting of 2 s Go and 6 s No Go intervals. This 2 s and 6 s pattern reflects the stable relaxation behavior established under NHPI mediated conditions rather than an arbitrarily selected timing. (c) Production rates of GlcN·HCl and GlcNAc (mg cm-2 h-1) at varying current densities, calculated for a 20 g chitosan batch. (d) Selective product yields obtained under different electrolysis modes. All entries except the GONOGO 2 s and 6 s data correspond to 300 mg scale experiments. The GONOGO 2 s and 6 s data represent the 1 kg semi industrial batch.
Although GONOGO is conceptually defined as a logic-gated protocol in which each Go phase is triggered by kinetic decision points rather than preset ton/toff cycles, the mediated conditions used here (NHPI, acetate buffer and Ti anode) establish a quasi-steady-state interfacial environment. As a result, the experimentally observed Go/No-Go durations remain essentially constant. While these durations were implemented through a programmed 2 s/6 s waveform, this apparent constancy does not arise from arbitrary pre-programming; instead, it reflects mediator-driven self-regulation of radical flux, suppression of E–t drift, and reliable confinement of each Go excursion within the PINO activation window.
Notably, both GlcN and GlcNAc were detected in their α- and β-anomeric forms in aqueous solution. Following PINO-mediated cleavage, the β-anomers initially generated interconverted via classical mutarotation, consistent with formation of an open-chain aldehyde intermediate (Figure 2a, step 3). This behavior, previously known for GlcN, is here extended and verified for GlcNAc by NMR analysis. The coexistence of both anomers in solution underscores the structural fidelity of the cleavage products and further confirms the mildness and selectivity of the electrochemical environment.
In conclusion, the GONOGO protocol provides a distinct waveform-controlled strategy that temporally separates radical generation from interfacial relaxation, enabling highly selective depolymerization of partially acetylated chitosan into GlcN and GlcNAc. This approach overcomes the limitations of our earlier galvanostatic system and establishes a more generalizable framework for electro-organic synthesis and biopolymer valorization.
3.2. Comparative Evaluation of Electrolysis Modes and Mediator Loading Effects
The chitosan used in this study was extracted from shrimp shell waste and was extensively characterized in our previous work. It had a DD of 86.4%, corresponding to 13.6% residual N acetyl content. Its structural integrity and electrochemical suitability were confirmed through FTIR, 1H NMR, CHNS elemental analysis and DSC, and therefore full recharacterization was not repeated in the present study.
To evaluate the effect of waveform modulation on product selectivity and depolymerization performance, electrolysis experiments were carried out under five distinct operational modes: CPE, galvanostatic electrolysis (as applied in our previous work), pDC, rAP, and the newly developed GONOGO protocol. All reactions were performed in acetate buffer (pH ≈ 4.5), using a titanium working electrode and 10 mol% NHPI under constant current control, with waveform-dependent cycling conditions applied as specified in each experiment.
Although pDC is architecturally closer to GONOGO in terms of having explicit ON and OFF intervals, rAP provided substantially more mechanistic information and therefore received broader experimental coverage. rAP exaggerates redox-cycling intensity, interfacial polarization, fouling progression, and mass-transport stress, which allowed us to map the operational boundaries that GONOGO must avoid. In contrast, pDC was used primarily as a benchmark to illustrate the limitations of fixed ton and toff pulsing, and once its performance plateau had been identified, additional pDC experiments did not yield further mechanistic insight.
Electrolysis durations varied across the different experiments depending on the pulse regime and applied current density. In each case, the reaction was monitored by cyclic voltammetry (CV), TLC, and HPLC, and was terminated when no further increase in GlcN or GlcNAc concentration was observed.
The choice of 10 mol% NHPI was based on our systematic screening (Table S4), in which this loading offered an optimal balance between redox efficiency, mediator stability, and cost-effectiveness. Lower NHPI concentrations (e.g., 2–8 mol%) resulted in incomplete depolymerization and slower reaction kinetics, likely due to insufficient PINO radical generation. Although 9 mol% provided a moderate yield, full efficiency and selectivity were only achieved at 10 mol%. Conversely, higher loadings (≥11 mol%) did not lead to any significant improvement in product yield and instead promoted side reactions, including nonselective oxidation and partial NHPI decomposition during extended electrolysis. Therefore, 10 mol% NHPI was selected as the mechanistically and operationally optimal dosage under the applied GONOGO-controlled electrochemical conditions.
The electrochemical window for selective NHPI oxidation to the PINO radical was determined by CV (Figure S1). Under galvanostatic conditions, complete degradation of N-acetyl groups occurred, yielding only GlcN·HCl, in line with our previous findings (Figure 2d). In contrast, application of the GONOGO waveform (e.g., 2 s/4 s or 2 s/6 s) resulted in partial preservation of N-acetyl groups, affording up to 13.0% isolated GlcNAc, which closely matches the theoretical maximum based on the starting DD of the biopolymer. The corresponding 1H NMR spectra showed clear methyl signals (δ ~2.0 ppm) attributed to GlcNAc, confirmed by HPLC and 2D NMR techniques.
Importantly, the GlcN/GlcNAc ratio could be selectively tuned by adjusting the Go/No-Go cycle duration. For example, extending the No-Go interval from 4 to 6 seconds (while keeping the Go time at 2 s) slightly improved GlcNAc yield, suggesting better protection of labile acetyl groups due to extended interfacial relaxation. Conversely, shortening the No-Go phase (e.g., 2 s/2 s) reduced selectivity and increased side reactions. These intentional variations were performed to challenge the intrinsic 2 second and 6 second relaxation pattern observed under NHPI mediated conditions and to confirm that deviations from this natural timing disrupt interfacial recovery, weaken PINO mediated selectivity and promote overoxidation pathways.
Comparison with rAP revealed key mechanistic differences. Although symmetric current reversal (e.g., 2 s anodic / 2 s cathodic) partially mitigated electrode fouling and allowed limited GlcNAc preservation (~5%), the continuous bidirectional cycling and absence of true rest intervals produced higher radical stress and significantly reduced selectivity. For completeness, a pDC (2 s ON / 6 s OFF) was also evaluated to clarify whether the enhanced selectivity observed under GONOGO stems merely from the duty cycle or from the mechanistic constraints imposed by the logic-gated design. Although pDC employs the same nominal ON/OFF timing as GONOGO, its OFF period does not enforce full relaxation to open circuit, resulting in only partial discharge of the double layer and incomplete dissipation of interfacial hydroxide or oxidizing species.
This distinction is critical because the selective NHPI → PINO activation window lies between approximately +0.68 and +0.80 V, whereas potentials above +0.85 V promote nonselective oxidation pathways that destroy N-acetyl groups. Under pDC, the electrode potential progressively drifts into this overoxidation region, leading to higher ΔRct (+44%), increased interfacial polarization and reduced GlcNAc selectivity. In contrast, the enforced open-circuit reset in GONOGO ensures that each Go cycle begins from a fully discharged interface, maintains controlled PINO-mediated radical flux, suppresses drift and confines every Go excursion to the selective window. The pDC dataset in Table S2 confirms that waveform-enforced interfacial relaxation, not merely ON/OFF timing, is essential for high selectivity and stable reactor operation.
To further clarify the advantages of GONOGO, we benchmarked it against representative pulsed and alternating polarity modes using the dataset in Table S2. Under rAP (5 s anodic / 5 s cathodic, 85 mA cm-2), GlcNAc was obtained in only 4.6% yield and ΔRct increased by ≈ 70%, indicating substantial electrode passivation. Under traditional pDC conditions (2 s ON / 6 s OFF, 89 mA cm-2), GlcNAc formation improved to 7.8% and ΔRct decreased to ≈ 44%, but acetyl groups were still only partially preserved. In contrast, GONOGO at 2 s / 6 s and 89 mA cm-2 delivered 13.2% GlcNAc, very close to the DD-defined ceiling, while limiting ΔRct to ≈ 35%. Since these experiments were performed at comparable current densities and cell voltages, the shorter electrolysis time under GONOGO (8.0 h, versus 10.6–10.8 h for rAP and pDC) translated into approximately 25–30% lower specific electrical energy consumption.
A comprehensive dataset summarizing over 130 electrolysis conditions (138 in total) is provided in Table S2. This includes all tested operational modes, CPE, conventional galvanostatic electrolysis, rAP (across multiple symmetric pulse intervals), pDC, and GONOGO (across fifteen asymmetric waveform designs), as well as NHPI-free controls. The matrix spans a wide range of current densities, pulse profiles, and electrolysis durations, and includes scale-up experiments up to 1 kg of chitosan. While performance metrics for optimized GONOGO conditions at various scales were detailed in Section 3.1, Table S2 enables direct comparison of product yields, GlcN/GlcNAc selectivity, and electrode fouling (ΔRct %) across all conditions. Notably, the data confirm that GONOGO not only minimizes radical overexposure but also significantly reduces electrode passivation compared to CPE and galvanostatic modes, further validating its utility as a scalable, selective, and electrochemically stable depolymerization strategy.
Taken together, these comparisons demonstrate that true open-circuit relaxation, unique to GONOGO, is the key mechanistic factor enabling selective preservation of N-acetyl groups—an outcome unattainable under CPE, galvanostatic or rAP waveforms.
3.3. Selective Formation of GlcN and GlcNAc
To quantitatively assess the product distribution achieved under GONOGO electrolysis, reaction mixtures were analyzed by both HPLC and 1H NMR spectroscopy. The baseline experiment employing a 2 s “Go” and 4 s “No-Go” pulsing scheme yielded a GlcN·HCl to GlcNAc ratio of approximately 68.5:11.5, as determined by integration of the acetyl methyl proton signal at δ ~1.97 ppm in the 1H NMR spectrum, in close agreement with the HPLC results (Figure 3(1)). Extending the “No-Go” interval to 6 s reproducibly enhanced GlcNAc production, reaching 12.72% in a 1 kg scale reaction, approaching the theoretical maximum (13.6%) based on the residual acetylation of the starting chitosan (Figure 3(2)).
To confirm the identity and purity of the GlcNAc, the crude reaction mixtures were compared to an authentic sample obtained via silica gel column purification followed by recrystallization. The isolated GlcNAc matched standard references in both retention time and NMR profile (Figure 3(4)), supporting its structural assignment. Likewise, pure GlcN·HCl (>99%) was isolated and characterized as a reference (Figure 3(3)).
Overlaid HPLC chromatograms and 1H NMR spectra of crude and purified samples are presented in Figure 3 for direct comparison. These analyses confirm the chemoselective generation of both GlcN and GlcNAc under the GONOGO protocol and validate the temporal control over radical exposure as a decisive factor in preserving labile acetyl groups. Notably, this selectivity is not attainable under constant-potential or traditional galvanostatic electrolysis conditions, where overoxidation typically leads to complete deacetylation.
Figure 3. (a) HPLC chromatograms and (b) 1H NMR spectra of the purified products obtained via the GONOGO method. Sample 1 corresponds to the product isolated under the GONOGO (2 s/4 s) profile, consisting of 68.5% GlcN·HCl and 11.5% GlcNAc as determined by HPLC. Sample 2 represents the product obtained under the GONOGO (2 s/6 s) profile, containing 70.84% GlcN·HCl and 12.72% GlcNAc for 1 kg scale. Sample 3 is purified GlcN·HCl with >99% purity, and Sample 4 is purified GlcNAc with >99% purity. The combined HPLC and 1H NMR analyses confirm the successful separation and precise compositional determination of the products.
Further pulse tuning revealed that shortening the No-Go phase (e.g., 2 s/2 s) decreased GlcNAc yield and increased side-product formation, supporting the hypothesis that sufficient interfacial relaxation time is critical for protecting sensitive functional groups during oxidative depolymerization.
3.4. Spectroscopic Characterization and Anomeric Assignment of GlcNAc
The structural elucidation of GlcNAc, isolated from the electrochemical depolymerization of partially deacetylated chitosan in a semi-industrial batch mediated by the GONOGO protocol, was performed using high-resolution 1H NMR, 2D COSY spectroscopy, and USP-compliant HPLC analysis. The data clearly confirmed both the chemical integrity and anomeric composition of the product, in full agreement with established reference standards [18-20].
The 1H NMR spectrum (Figure 4a), acquired in D2O (400 MHz), displayed a doublet at δ = 5.12 ppm (J = 2.3 Hz) corresponding to the anomeric proton of α-GlcNAc, and a doublet of doublets at δ = 4.63 ppm (J = 8.4 Hz) assigned to the anomeric proton of β-GlcNAc. These coupling constants are consistent with axial–equatorial (α, small J1,2) and axial–axial (β, large J1,2) geometries at the C1–C2 bond, validating the presence of both anomers in equilibrium, typically favoring the β-form (ca. 65:35), as also observed in solution-state GlcNAc studies [18,20].
A prominent singlet at δ = 1.97 ppm (6H) was attributed to the methyl protons of the N-acetyl moiety in both α- and β-anomers. This chemically isolated signal serves as a reliable quantitative handle for estimating GlcNAc content in crude reaction mixtures, owing to its stoichiometric correspondence and lack of overlap with sugar-ring protons.
The spectrum further exhibited a series of well-resolved multiplets in the range δ = 3.37–3.84 ppm, consistent with the overlapping envelope of ring protons (H2–H6) from both anomers. A triplet at δ = 3.60 ppm (J = 9.7 Hz) is likely attributable to H5, indicative of preserved chair conformation and regular scalar coupling along the pyranose ring.
Structural confirmation was reinforced by the 2D COSY spectrum (Figure 4b), which revealed all expected 3JHH correlations. Notably, strong cross-peaks between H1β–H2β and H1α–H2α corroborated the anomer-specific assignments. Sequential couplings (H2–H3, H3–H4, H4–H5, and H5–H6) were also identified, supporting complete structural integrity of the pyranose core. The absence of extraneous or unresolved signals suggests high chemical purity and minimal oxidative degradation.
Figure 4. (a) 1H NMR spectrum of electrochemically derived GlcNAc with assignments for α\β-anomeric protons. (b) 2D COSY spectrum confirming proton–proton correlations and identification of α\β-GlcNAc.
Figure 5. (a) 13C NMR spectrum of isolated GlcNAc showing distinct signals for α- and β-anomers. (b) 2D HSQC spectrum confirming C–H correlations for both anomers, validating the structural integrity of GlcNAc after electrochemical depolymerization.
The 13C NMR spectrum (Figure 5a) of the purified GlcNAc showed a strong resonance at δ ~ 174 ppm, corresponding to the amide carbonyl (C=O), and a distinct signal at δ ~ 22 ppm for the acetyl methyl carbon (CH3). These chemical shifts are fully consistent with the literature values reported for GlcNAc and its analogs [18,20].
FTIR analysis (Figure S2), further confirmed the presence of characteristic functional groups. The spectrum revealed prominent absorption bands at 1654 cm-1 (amide I, C=O stretching), 1560 cm-1 (N–H bending), and a broad stretch between 3450–3200 cm-1 attributed to overlapping O–H and N–H stretching vibrations.
To assess purity and verify identity based on United States Pharmacopeia (USP) standards, HPLC analysis was carried out without derivatization, following a validated Sigma-Aldrich protocol for GlcNAc quantification. A single sharp peak was observed at Rt ≈ 4.5 min, matching the retention time of the certified reference standard (Sigma-Aldrich, ≥99%). No additional peaks were detected, confirming a chemical purity >99%. The standard calibration curve exhibited excellent linearity in the range of 25–1000 μg/mL (R2 = 0.998), as shown in Figure 3(4).
Details of the full HPLC method, column specifications, mobile phase, and calibration parameters are provided in the Supporting Information (Section S6 and S7).
GlcN·HCl analyses are provided in the Supporting Information (Section S8). GlcNAc structural profile complies with pharmacopeial standards (Table S3), and the synthesis, being acid-free and metal-free, ensures excellent biocompatibility for biomedical applications.
3.5. Mechanistic Insights and DFT-Supported Selectivity Control
CV (Figure S1) defines the selective NHPI→PINO oxidation window (+0.68–0.80 V vs Ag/AgCl), enabling productive radical initiation while avoiding direct oxidation of chitosan or the acetate buffer. Although the electrolysis is performed under galvanostatic control, the instantaneous anodic potentials reached during each GONOGO Go interval consistently remain within this PINO-activation window. This congruence between CV and chronopotentiometric behavior confirms that NHPI turnover is both selective and stable throughout GONOGO operation.
Once generated, the PINO radical performs hydrogen abstraction at the C2 or C6 positions of the GlcN repeating units, initiating β(1→4) homolytic cleavage (TS1). As shown in Figure 6, DFT calculations confirm that this productive depolymerization pathway proceeds through TS1 with a modest barrier (ΔG‡ ≈ 17.8 kcal mol-1), fully compatible with operation inside the PINO-mediated potential window. Under conventional galvanostatic electrolysis, however, continuous PINO generation combined with local alkalization drives the anode potential upward over time, eventually surpassing +0.85 V. In this high-potential regime, the DFT-predicted N-acetyl degradation steps TS2 (ΔG‡ ≈ 27.9 kcal mol-1) and TS3 (ΔG‡ ≈ 40.0 kcal mol-1), also illustrated in Figure 6, become thermodynamically accessible, explaining the complete loss of GlcNAc and exclusive formation of GlcN·HCl under galvanostatic conditions.
Figure 6. DFT-derived free energy profile for the NHPI/PINO-mediated depolymerization and N-acetyl degradation pathways. The initial β(1→4) homolytic scission step (TS1, ΔG‡ ≈ 17.8 kcal mol-1) corresponds to productive depolymerization and operates within the selective NHPI→PINO oxidation window (+0.68–0.80 V vs Ag/AgCl). In contrast, the high-energy degradation steps TS2 (ΔG‡ ≈ 27.9 kcal mol-1), involving PINO-mediated hydrogen abstraction from the N-acetyl methyl group, and TS3 (ΔG‡ ≈ 40.0 kcal mol-1), corresponding to C–N bond homolysis, require access to the high-potential overoxidation region (>0.85 V). These steps account for the complete loss of GlcNAc under galvanostatic conditions and sharply contrast with the selective N-acetyl preservation achieved under GONOGO operation. All geometries were optimized at the B3LYP/6-31+G(d,p) level with implicit water solvation (PCM model).
GONOGO fundamentally prevents access to this overoxidation regime. The short 2-s Go interval produces PINO and drives TS1 selectively, while the subsequent 6-s No-Go interval returns the electrode fully to open-circuit potential (OCP), allowing EDL relaxation, dissipation of hydroxide, and restoration of interfacial neutrality. This temporal decoupling suppresses overoxidation pathways and preserves N-acetyl groups.
These mechanistic distinctions are directly observable in the extended chronopotentiometric analysis shown in Figure 7a–c. The potential trace demonstrates that every Go excursion remains confined to +0.68–0.80 V, and every No-Go interval returns completely to OCP (~0.65 V), with no Go pulse exceeding +0.85 V. The instantaneous voltage-slope map (dV/dt) exhibits perfectly periodic charging and relaxation spikes, with invariant amplitude across all examined cycles, confirming the absence of drift or mediator degradation. The logic-state overlay shows that the reconstructed dV/dt-derived state perfectly matches the imposed Go/No-Go sequence, demonstrating full reproducibility and stability of the waveform over prolonged operation.
Figure 7. Electrochemical behavior of the GONOGO waveform (2 s Go / 6 s No-Go). (a) Potential trace showing that all Go intervals remain confined to the NHPI→PINO activation window (+0.68–0.80 V), whereas all No-Go intervals return fully to OCP (~0.65 V). No pulse enters the high-potential overoxidation region (>0.85 V). (b) Instantaneous voltage slope (dV/dt), displaying periodic EDL charging and relaxation spikes with invariant amplitude, indicating complete waveform stability and absence of drift. (c) Reconstructed logic-state derived from dV/dt overlaid with the programmed Go/No-Go sequence. The perfect overlap demonstrates state-dependent relaxation, reproducible duty cycles, and full interfacial compliance with the imposed logic rule.
To formalize the logic element of GONOGO, the instantaneous voltage slope (dV/dt) at the onset of each Go pulse was treated as a quantitative descriptor of interfacial polarization. Go intervals persisted while dV/dt remained within its characteristic PINO-activation baseline, and No-Go intervals continued until dV/dt had fully relaxed back to this baseline. Under the NHPI/Ti/acetate conditions, this state-dependent relaxation reproducibly converged to an effective 2 s Go / 6 s No-Go duty cycle, precisely matching the invariant waveform morphology observed experimentally. This demonstrates that the imposed waveform is reinforced by interfacial kinetics governing mediator turnover and EDL relaxation, rather than merely programmed externally.
Further mechanistic validation is provided by radical-scavenging controls (Section S11). BHT significantly reduces total monomer formation and nearly eliminates GlcNAc, consistent with interception of the initial PINO-mediated hydrogen-abstraction step. TEMPO almost fully suppresses GlcNAc and substantially decreases GlcN yield, confirming the participation of carbon-centered radicals downstream of TS1. Importantly, no aldehydic or overoxidized byproducts are detected under GONOGO conditions, demonstrating that depolymerization proceeds through a controlled mediated-radical pathway rather than direct anodic oxidation.
Electrochemical impedance spectroscopy (Table S2) further supports the distinct interfacial behavior of GONOGO. Galvanostatic electrolysis induces severe fouling and strong polarization (ΔRct ≈ +120%), conditions that activate TS2/TS3 and destroy N-acetyl groups. In contrast, GONOGO produces only a moderate increase in Rct (≈ +35%), reflecting reduced interfacial stress and controlled radical flux consistent with selective GlcNAc preservation.
Finally, the no-NHPI control (Entry 138, Table S2) clearly demonstrates that mediated radical chemistry is essential for waveform stability and selectivity. Without NHPI, Go-phase potentials rapidly exceed the PINO window, waveform morphology collapses, and <5 mg of monomers are formed with no detectable GlcNAc. This confirms that NHPI not only initiates productive depolymerization but also stabilizes the interfacial environment required for the logic-gated waveform to function.
Taken together, CV analysis, dV/dt-resolved logic-state mapping, DFT calculations, EIS measurements, and radical-trapping studies converge to a unified mechanistic conclusion: GONOGO achieves chemoselectivity by temporally structuring radical initiation and interfacial relaxation. This logic-gated control confines the system to the TS1 depolymerization pathway while preventing entry into the high-potential TS2/TS3 regime, thereby enabling simultaneous and selective formation of GlcN and GlcNAc.
3.6. Kinetic Modeling and Activation Energy Determination of Selective GlcNAc Formation under GONOGO Pulsed Electrolysis
To elucidate the kinetic and thermodynamic parameters governing the selective formation of GlcNAc during GONOGO-assisted electrochemical depolymerization of chitosan, a temperature-dependent kinetic study was performed under pulsed electrolysis conditions. The GONOGO (2 s/6 s) protocol was specifically engineered to suppress overoxidation, limit radical-induced deacetylation, and maintain the integrity of labile N-acetyl groups during depolymerization.
Experiments were conducted using 1000 g of partially deacetylated chitosan (degree of deacetylation, DD = 86.4%) in 0.2 M acetic acid and 0.3 M sodium acetate (pH 4.5 ± 0.2), with 10 mol% NHPI as the redox mediator. A titanium anode (40 cm × 28 cm) was used, and the effective electrolysis current during the Go phases was maintained at 89 mA cm-2. The system was stirred continuously and maintained at isothermal conditions (25, 30, 35, 40, 45 and 50 °C). Aliquots were withdrawn at predetermined intervals, quenched, and analyzed by HPLC using pre-calibrated GlcNAc standards. Quantitative HPLC data showed co-generation of GlcNAc and GlcN, with GlcNAc contributing ~13% (127.2 g, 0.575 mol) of the total monomer yield under optimal conditions.
Although the depolymerization involves a complex interplay of radical generation and polymer fragmentation, the experimental GlcNAc concentration–time profiles were most accurately described by a pseudo-first-order model. This approximation was therefore adopted for kinetic analysis, expressed as:
where 𝑘app is the apparent rate constant. The extracted 𝑘app values at different temperatures were used to construct an Arrhenius plot (Figure 8a). Linear regression of ln(𝑘app) versus 1/T showed excellent linearity (R2 = 0.9897), confirming an activation-controlled process.
From the Arrhenius plot, the activation energy (Ea) and pre-exponential factor (A) were determined to be 75.50 kJ mol-1 and 1.85 × 108 s-1, respectively. Based on the regression analysis, the standard error of the intercept (ln A) was estimated as ± 0.18, yielding a 95% confidence interval for the pre-exponential factor (A) in the range of (1.26 × 108 to 2.71 × 108 s-1). This statistical analysis confirms the robustness of the fitted kinetic model and supports its validity in describing the experimental data.
Figure 8. (a) Arrhenius plot based on pseudo-first-order rate constants (kapp) extracted from GlcNAc formation kinetics under GONOGO electrolysis. The linear fit yields an Ea of 75.5 kJ mol-1 and A of 1.85 × 108 s-1 (R2 = 0.9897). (b) Time-dependent GlcNAc concentration profile at 40 °C, fitted to a pseudo-first-order kinetic model with kapp = 5.70 × 10-5 s-1. Experimental data are shown as red dots; fits are shown as dashed lines.
Details of the calculation of Qtotal, Qtheoretical, and ΦF, along with model validation and comparison with higher-order kinetic fits, are provided in the Supporting Information (Section S10).
Table 1 summarizes the key kinetic parameters determined at 40 °C under GONOGO conditions.
| Parameter | Value (Mean ± SE)* |
| Total charge passed (Qtotal) | 765,000 C |
| Theoretical useful charge (Qtheoretical) | 732841.22 C |
| Experimental useful charge (Quseful) | 483872.27 C |
| Theoretical Faradaic efficiency (ΦF) | 95.79% ± 1.0% |
| Experimental Faradaic efficiency (ΦF) | 63.25% ± 1.20% |
| GlcNAc yield (mol) | 0.575 ± 0.001 mol |
| Final GlcNAc mass (g) | 127.196 g |
| Pseudo-first-order rate constant (kapp) | 5.70 × 10-5 s-1 |
* All values (except fixed/calc. quantities) are reported as mean ± standard error (SE) obtained from triplicate experiments under the same electrochemical conditions.
The selectivity of GlcNAc production under GONOGO conditions is attributed to the rest phases, which reduce the steady-state radical concentration and mitigate hydroxide-induced N-acetyl hydrolysis. This was confirmed by the absence of GlcNAc in control experiments under galvanostatic. In addition, the combined HPLC and NMR data confirmed partial preservation of acetyl functionality, validating the kinetic model.
Complete mathematical modeling, kinetic simulations, and details of GlcNAc quantification and calibration are provided in the Supporting Information (Sections S6 and S10).
Overall, the kinetic analysis under the GONOGO protocol highlights a moderately high activation energy (Ea = 75.50 kJ mol-1) and a pre-exponential factor (A = 1.85 × 108 s-1), which are consistent with radical-mediated surface reactions in heterogeneous electrochemical systems. The real Faradaic efficiency (63.25%) is fully compatible with the theoretical value (95.79%) when intrinsic radical-consumption pathways are taken into account, indicating that the measured efficiency reflects the true chemical fate of PINO-derived radicals under GONOGO conditions. Although the theoretical Faradaic efficiency for complete NHPI-mediated depolymerization is calculated as 95.79%, the experimentally measured value at the 1 kg scale (63.25%) is consistent with the behavior of NHPI and PINO-based radical systems, where self-recombination, non-productive quenching and mediator decay frequently limit charge utilization [21]. Our process analysis (Sections S13–S14) indicates that most of this gap does not originate from classical electrochemical or energetic losses, since ohmic dissipation, electrode polarization and thermal contributions are minimized under the GONOGO-controlled pulsed regime. This is supported by the moderate increase in charge-transfer resistance (ΔRct ≈ +47%), the small voltage fluctuation (<3%), and the low specific energy demand of approximately 3.2 kWh per kg of chitosan. Faradaic efficiencies and loss pathways of comparable magnitude have been documented in radical-mediated electro-oxidative depolymerization of lignocellulosic substrates, where complex radical networks and mass-transport constraints inherently limit electron utilization [22,23]. Instead, the discrepancy mainly arises from mediator-related radical pathways, including PINO self-recombination, non-productive radical quenching and partial mediator decay during extended operation, which consume charge without contributing to glycosidic bond scission. In this context, the Faradaic efficiency of 63.25% is comparable to or higher than values commonly observed in radical-mediated electro-oxidative depolymerization of lignocellulosic substrates, and further improvements, if needed at industrial scale, would stem from engineering-level optimizations such as improved electrode spacing and enhanced mixing, as widely emphasized in electrochemical engineering studies of organic electrosynthesis and biomass electroconversion [24]. These insights validate the proposed kinetic model and underscore the critical role of the pulsed electrolysis waveform in modulating selectivity and energy efficiency during electrochemical chitosan conversion.
3.7. Comparative Evaluation of Methods for GlcNAc Recovery from Chitosan
While the production of GlcN from chitosan has been extensively explored using a variety of chemical, enzymatic, and electrochemical approaches, the selective recovery of GlcNAc and concurrent formation of GlcN remains synthetically underexplored. This is primarily due to the inherent lability of the N-acetyl group, which is prone to hydrolysis and oxidation under harsh or uncontrolled reaction conditions. Most conventional methods for chitosan depolymerization, including acid hydrolysis and galvanostatic electrolysis, result in complete deacetylation, yielding GlcN as the sole product, with no recovery of intact N-acetylated monomers.
In our previous study, we evaluated and compared various strategies for GlcN production from chitosan, including acid-mediated, enzymatic, and electrochemical methods. However, none of these approaches were able to preserve the N-acetyl groups to yield GlcNAc. The present study addresses this critical gap by demonstrating, for the first time, that time-modulated electrochemical depolymerization via the GONOGO protocol enables the selective recovery of GlcNAc directly from partially deacetylated chitosan, alongside GlcN.
To contextualize the performance of the GONOGO method, Table 2 summarizes the reported yields of GlcNAc obtained from various representative chemical, enzymatic, and electrochemical strategies. As shown, conventional acid hydrolysis and galvanostatic electrolysis lead to complete deacetylation, while enzymatic methods suffer from poor conversion and low scalability. Notably, previously reported waveform-based electrochemical approaches (e.g., rAP or AC methods) have not been applied to GlcNAc production and lack selectivity for N-acetylated units due to their uncontrolled radical flux and polarity switching. In contrast, the GONOGO method achieves up to 13.0% GlcNAc yield under optimized pulsing conditions.
Beyond these classical strategies, several state-of-the-art depolymerization methods, including solution plasma discharge, Fenton-type oxidative systems, γ- or UV-assisted radical degradation, ultrasonic scission, and TEMPO-mediated selective oxidations, were also examined (Table 2). Although these approaches effectively reduce the molecular weight of chitosan, none of them preserve the N-acetyl functionality, and no structurally confirmed GlcNAc has been reported. The underlying limitation across these methods is the uncontrolled radical flux or harsh oxidative conditions, which lead to rapid cleavage or transformation of the N-acetyl group. This highlights the lack of any existing depolymerization protocol capable of selectively retaining N-acetylated units and underscores the unique advantage of the time-modulated GONOGO approach.
As summarized in Table 2, and to the best of our knowledge, no chemical, oxidative, physical, or electrochemical depolymerization strategy reported to date has demonstrated selective GlcNAc formation. Rapid deacetylation and overoxidation following glycosidic bond cleavage fundamentally limit N-acetyl retention in conventional methods. This long-standing gap underscores the need for a time-regulated electrochemical protocol capable of preserving N-acetyl groups, a requirement that is addressed here through the GONOGO pulsed method.
Table 2 Comparative performance of representative methods for GlcNAc production from chitosan or related precursors.
| No. | Method | Conditions | GlcNAc Yield | Notes | Ref. | |
| 1 | Acid hydrolysis | 6 N HCl, 100 °C, 6 h | 0 | Complete deacetylation; GlcN only | [25] | |
| 2 | Enzymatic digestion | Chitosanase + NAGase, 37 °C, 48 h | <5 | Requires multiple enzymes; low conversion and low throughput | [26] | |
| 3 | Galvanostatic electrolysis (Ti/NHPI) | 60 mA, pH 4.5, 40 °C, 2 h | 0 | Overoxidation of N-acetyl groups; only GlcN detected | [4] | |
| 4 | Solution Plasma Depolymerization | High-voltage discharge in aqueous chitosan | 0 | Produces low-MW fragments, N-acetyl groups not preserved | [27] | |
| 5 | Oxidative – Fenton/Peroxide | H2O2 (1–5%) + metal catalyst (Fe2+, Cu2+); often with UV or heat. | 0 | Strong OH• radicals cause full deacetylation; no GlcNAc preserved | [28] | |
| 6 | Oxidative – Radiation (γ, UV) | γ-rays (10–100 kGy) or UV with H2O2; often in solid or acidic solution. | 0 | Lower-MW chitosan; oligomers (DP 2–10), monomers heavily oxidized | [29] | |
| 7 | Physical – Ultrasonic | Ultrasound (20–40 kHz, high power) in dilute acid or water; 0–60 °C | 0 | Reduced-viscosity chitosan; MW down to ~10 kDa; no monomers | [30] | |
| 8 | Selective Oxidation (TEMPO) | TEMPO/NaOCl/NaBr at pH ~10, 0–5 °C. Often applied to chitosan dissolved in dilute acid, then pH adjusted. | 0 | TEMPO-oxidized chitin/chitosan (polyglucuronic acid derivatives); some depolymerization if GlcN present. | [31] | |
| 9 | pDC (2 s/6 s) | 88 mA cm-2, NHPI, pH 4.5, 40 °C, 10.8 h | 7.8 | Partial N-acetyl preservation; pulsing reduces overoxidation compared to DC | This work | |
| 10 | rAP (5 s/5 s) | 85 mA cm-2, NHPI, pH 4.5, 40 °C, 10.6 h | 4.6 | Not suitable for GlcNAc; uncontrolled radical flux | This work | |
| 11 | GONOGO (2 s/6 s) | 89 mA cm-2, NHPI, pH 4.5, 40 °C, 8.5 h | 12.72 | Near-theoretical limit (DD = 13.6%) | This work | |
In the context of practical application, both GlcN and GlcNAc are industrially valuable products. However, the emphasis of the present work is on the selective preservation and recovery of GlcNAc, since, to the best of our knowledge, no electrochemical depolymerization method has previously succeeded in producing intact GlcNAc. While GlcN is readily accessible through conventional acid, enzymatic or galvanostatic routes, GlcNAc remains synthetically challenging due to the extreme lability of the N acetyl group under oxidative or acidic conditions. Therefore, the ability of the GONOGO protocol to recover GlcNAc at yields approaching the DD defined theoretical limit represents a meaningful improvement over existing methods and demonstrates clear advancement in selective polysaccharide depolymerization.
3.8. Analytical Benchmarking, Industrial Relevance, and Applications of Recovered GlcN/GlcNAc
Electrochemical valorization of chitosan has historically produced a narrow range of products, most notably acetate, hydrogen, low-molecular-weight chitosan (LMWC), and fully deacetylated GlcN, reflecting fundamental limitations in selectivity under static or non-modulated conditions. For example, systems employing nickel-based electrodes, as in entries 1 [32] and 2 in Figure 9, predominantly yielded small molecules such as acetate and hydrogen, without any formation of higher-order products. In our own nickel-based experiments (entry 2), acetate was also detected together with hydrogen, corroborating the intrinsic tendency of Ni systems toward complete backbone cleavage (Section S11). The remaining oxidation current in Ni-based systems was diverted toward acetate and hydrogen formation (Section S12).
These outcomes underscore the intrinsic non-selectivity of such setups, where oxidative cleavage of the polysaccharide backbone leads to total molecular breakdown. While acetate and hydrogen have industrial uses (e.g., as buffering agents, preservatives, or clean fuels [33]), these outputs lack structural complexity and provide limited value in the context of targeted biopolymer depolymerization. Similarly, random-chain scission methods yielding LMWC (entries 3 [12] and 4[13]) are not suitable for selective depolymerization, as they fail to generate well-defined monomers such as GlcN or GlcNAc and cannot preserve labile functional groups like the N-acetyl moiety.
In contrast, the GONOGO protocol enables the selective and simultaneous production of two structurally distinct monosaccharides: GlcN and GlcNAc. These compounds are not only structurally relevant but also industrially valuable. GlcN is well-established as a joint health supplement and is commercially available as its hydrochloride salt (GlcN·HCl) in pharmaceutical-grade formulations [34]. GlcNAc, on the other hand, is a rarer and more labile compound essential for the biosynthesis of hyaluronic acid and other glycoconjugates, with broad applications in dermatology, wound healing, and anti-aging formulations [35]. Its higher solubility, stability in biological media, and reduced immunogenicity make it preferable in many cosmetic and therapeutic settings [36].
The ability of the GONOGO protocol to preserve N-acetyl groups is directly linked to the temporal decoupling of radical generation and interfacial reactivity, which limits overoxidation and allows labile motifs to survive. Under optimized 2 s/6 s pulsing conditions, GlcNAc was obtained at 12.72% yield together with 70.84% GlcN·HCl, giving an overall Faradaic efficiency of 63.25% (entry 6, Figure 9; see Table 1 for details). This demonstrates that electrode material (Ti vs Ni) and waveform control (GONOGO vs DC) jointly determine product selectivity.
Figure 9. Comparative overview of electrochemical depolymerization strategies for chitosan, highlighting product selectivity, electrode configurations, and applied conditions across recent and current studies.
To validate the industrial relevance of this method, a semi-industrial batch (1 kg chitosan) was processed under GONOGO conditions. A detailed engineering-scale assessment, including real energy consumption (~3.2 kWh per kg), NHPI recovery efficiency (91.2%), voltage–current stability profiles, and comparative techno-environmental metrics, has been provided in Supporting Information Sections S13 and S14, fully substantiating the semi-industrial scalability of the GONOGO protocol. The isolated GlcN and GlcNAc were comprehensively characterized and benchmarked against USP monographs. As detailed in Supporting Information Sections S9, orthogonal techniques including HPLC (USP method), FTIR, 1H and 13C NMR, COSY, HSQC, CHNS, and ESI-MS confirmed the identity, purity, and structural integrity of both products. The final purity of GlcN·HCl and GlcNAc exceeded 99%, with compliance across all USP parameters.
Additionally, microbiological testing confirmed acceptable bioburden and endotoxin levels for dermal and cosmetic applications (Table S3). These GlcNAc samples are now being utilized for the chemoenzymatic synthesis of hyaluronic acid, demonstrating the high-value potential of the protocol for biomedical and skincare industries.
Compared to traditional acid or enzymatic hydrolysis, the GONOGO-controlled protocol offers significantly lower energy input and a markedly reduced environmental footprint, as supported by green metrics and Life-Cycle Assessment (LCA) data in Sections S13 and S14.
This multifaceted validation confirms that the products obtained via GONOGO-controlled depolymerization not only meet regulatory standards but also hold promise for large-scale pharmaceutical and cosmetic use cases.
4. Conclusion
This study set out to test the hypothesis that logic-gated, time-asymmetric GONOGO pulsed electrolysis can decouple radical generation from interfacial chemistry, thereby stabilizing N-acetyl motifs in partially deacetylated chitosan and enabling selective β(1→4) scission to afford both GlcN and GlcNAc. The experimental findings validate this hypothesis across mechanistic, analytical, and scale-up dimensions.
First, the GONOGO protocol using NHPI as a mediator and titanium electrodes under mild aqueous conditions consistently yielded both GlcN·HCl and GlcNAc, whereas constant galvanostatic operation produced only GlcN. Importantly, the GlcNAc fraction reached 12.72% in a 1 kg batch at pH 4.5, approaching the theoretical maximum (13.6%) defined by the degree of acetylation (DD = 86.4%). This confirms that GONOGO preserves acetylated units in alignment with the central hypothesis.
Second, the waveform-dependence of product distribution was diagnostic. Extending the No-Go interval to 6 s enhanced GlcNAc recovery and minimized acetyl degradation, whereas shorter relaxation periods led to increased side reactions. These results support the role of time-gated double-layer re-equilibration and hydroxide dissipation in protecting N-acetyl groups during depolymerization.
Third, mechanistic controls demonstrated the involvement of a PINO-mediated radical pathway. Radical scavengers (BHT, TEMPO) suppressed monomer formation and diminished GlcNAc selectivity. Electroanalytical data located the NHPI/PINO redox couple within the accessible potential range during Go phases, and DFT calculations indicated that N-acetyl cleavage requires prohibitively high free-energy unless driven by sustained oxidative bias. These results mechanistically explain why GONOGO’s relaxation phases safeguard labile acetylated motifs.
Fourth, interfacial stability was superior under GONOGO relative to constant current. Electrochemical impedance spectroscopy showed lower growth of charge-transfer resistance (ΔRct), consistent with reduced electrode passivation and enhanced surface recovery during relaxation. This directly validates the prediction of the hypothesis regarding electrode fouling control.
Finally, the method proved scalable without loss of selectivity. The optimized 2 s ON / 6 s OFF protocol preserved the GlcN:GlcNAc ratio from milligram to kilogram scale, while maintaining >99% purity confirmed by USP-compliant HPLC and orthogonal spectroscopy. The process delivered a Faradaic efficiency of 63.25%, with theoretical limits of 95.79%, establishing realistic benchmarks for further engineering optimization.
Taken together, the results confirm that time-resolved, logic-gated pulsing is a decisive determinant of chemoselectivity, more influential than electrode material or mediator loading within the tested ranges. By approaching the DD-defined ceiling for GlcNAc recovery and reducing interfacial resistance growth, GONOGO validates a carbohydrate-centric, hypothesis-driven strategy for selective depolymerization of chitosan. Limitations remain in closing the gap between theoretical and experimental Faradaic efficiency and in extending the method to other chitin derivatives with diverse acetylation profiles. Future work will couple pulse optimization with hydrodynamic control, mediator recovery, and life-cycle assessment to expand the scope and enhance industrial viability of this electrochemical platform.
CRediT authorship contribution statement
Hossein Mojtabazadeh: Conceptualization, Methodology, Investigation, Data curation, Formal analysis, Resources, Writing – original draft. Javad Safaei-Ghomi: Writing – review & editing, Supervision, Resources, Project administration.
Acknowledgements
The authors gratefully acknowledge the financial support of Rah Enqelab Sanat Company and the research facilities provided by Kashan University.
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.
SUPPLEMENTAL INFORMATION
Supporting Information includes conceptual innovation of the GONOGO method, CV of NHPI (Figure S1), comparison of electrolysis modes and pulse strategies (Table S2), FTIR analysis of GlcNAc (Figure S2), USP-based HPLC analysis of GlcNAc and GlcN·HCl (Figure S3), semi-industrial product assessment, spectroscopic characterization of GlcN·HCl (Figures S4–S6), comparative specification data (Table S3), kinetic modeling and faradaic efficiency analysis, reaction optimization data (Table S4), and HPLC quantification of acetate during nickel-mediated electrolysis (Figure S7) and process engineering metrics for industrial relevance (Figure S8, Table S5) and a detailed life-cycle and environmental impact assessment (Table S6).
References
[1] I. Younes, M. Rinaudo, Chitin and Chitosan Preparation from Marine Sources. Structure, Properties and Applications, Marine Drugs 13(3) (2015) 1133–1174. https://doi.org/10.3390/md13031133.
[2] Y. Wang, Z. Wang, W. Lu, Y. Hu, Review on chitosan-based antibacterial hydrogels: Preparation, mechanisms, and applications, International Journal of Biological Macromolecules 255 (2024) 128080. https://doi.org/10.1016/j.ijbiomac.2023.128080.
[3] M. Mousavi-Ebadi, J. Safaei-Ghomi, H. Mojtabazadeh, Anchoring cobalt nanoparticles on to the grafted mono(6-ethylene-diamino-6-deoxy)-β-cyclodextrin to magnetic chitosan for enhanced catalytic performance of the synthesis of pyrano[2,3-c]pyrazole-3-carboxylates, Carbohydrate Polymer Technologies and Applications 11 (2025) 100923. https://doi.org/10.1016/j.carpta.2025.100923.
[4] H. Mojtabazadeh, J. Safaei-Ghomi, Sustainable electrochemical depolymerization of chitosan into glucosamine hydrochloride using N-hydroxyphthalimide as a redox catalyst, Carbohydr. Polym (2025) 124008. https://doi.org/10.1016/j.carbpol.2025.124008.
[5] A.P. Atkins, A.K. Chaturvedi, J.A. Tate, A.J.J. Lennox, Pulsed electrolysis: Enhancing primary benzylic C(sp3)–H nucleophilic fluorination, Org. Chem. Front. 11 (2024) 802–808. https://doi.org/10.1039/D3QO01865B.
[6] A.P. Atkins, A.J.J. Lennox, Application of pulsed electrolysis in organic electrosynthesis, Curr. Opin. Electrochem 44 (2024) 101441. https://doi.org/10.1016/j.coelec.2024.101441.
[7] Y. Kawamata, K. Hayashi, E. Carlson, S. Shaji, D. Waldmann, B.J. Simmons, P.S. Baran, Chemoselective electrosynthesis using rapid alternating polarity, J. Am. Chem. Soc 143 (2021) 16580–16588. https://doi.org/10.1021/jacs.1c06572.
[8] S. Rodrigo, C. Um, J.C. Mixdorf, D. Gunasekera, H.M. Nguyen, L. Luo, Alternating current electrolysis for organic electrosynthesis: Trifluoromethylation of (hetero)arenes, Org. Lett 22 (2020) 6719–6723. https://doi.org/10.1021/acs.orglett.0c01906.
[9] Y. Hioki, M. Costantini, J. Griffin, K.C. Harper, M.P. Merini, B. Nissl, P.S. Baran, Overcoming the limitations of Kolbe coupling with waveform-controlled electrosynthesis, Science 380 (2023) 81–87. https://doi.org/10.1126/science.adf4762.
[10] E.O. Bortnikov, S.N. Semenov, Coupling of alternating current to transition-metal catalysis: Examples of nickel-catalyzed cross-coupling, J. Org. Chem 86 (2021) 782–793. https://doi.org/10.1021/acs.joc.0c02350.
[11] D.E. Blanco, B. Lee, M.A. Modestino, Optimizing organic electrosynthesis through controlled voltage dosing and artificial intelligence, Proc. Natl. Acad. Sci. U. S. A 116 (2019) 17683–17689. https://doi.org/10.1073/pnas.1909985116.
[12] Z. Gu, Q. Cai, Y. Liu, F. Li, Electrochemical degradation of chitosan using Ti/Sb–SnO2 electrode, J. Polym. Environ 21 (2013) 479–486. https://doi.org/10.1007/s10924-012-0518-5.
[13] P.C. Deng, G. Wang, J.Z. Hu, Y.W. Tian, C. Weng, X. Gao, Electrochemical depolymerization of chitosans using the IrO2 electrode with interlayers as anode, Mater. Sci. Forum 847 (2016) 281–286. https://doi.org/10.4028/www.scientific.net/MSF.847.281.
[14] S. Xiao, Y. Yang, Cellulose Valorization via Electrochemical Oxidation: Efficient Formate Generation for Green Energy Storage, Biomass 5(2) (2025) 27. https://www.mdpi.com/2673-8783/5/2/27.
[15] D. Gao, D. Ouyang, X. Zhao, Electro-oxidative depolymerization of lignin for production of value-added chemicals, Green Chem 24(22) (2022) 8585–8605. http://dx.doi.org/10.1039/D2GC02660K.
[16] Y. Fan, Y. Zou, S. Wang, Electrocatalytic oxidation of biomass derived molecules, Sci. Bull 68(22) (2023) 2695–2699. https://doi.org/10.1016/j.scib.2023.09.024.
[17] L.Q. Hoa, H. Yoshikawa, M. Saito, M. Ueda, T. Shibata, E. Tamiya, Direct Energy Extraction from Brown Macroalgae-Derived Alginate by Gold Nanoparticles on Functionalized Carbon Nanotubes, ChemCatChem 6(1) (2014) 135–141. https://doi.org/10.1002/cctc.201300531.
[18] F.-C. Liu, C.-R. Su, T.-Y. Wu, S.-G. Su, H.-L. Yang, J.H.-Y. Lin, T.-S. Wu, Efficient 1H-NMR quantitation and investigation of N-acetyl-D-glucosamine (GlcNAc) and N,N′-diacetylchitobiose (GlcNAc)2 from chitin, Int. J. Mol. Sci 12 (2011) 5828–5843. https://doi.org/10.3390/ijms12095828.
[19] United States Pharmacopeia (USP), N-Acetylglucosamine. Dietary Supplement Monographs, USP–NF Online (2022). https://doi.org/10.31003/uspnf_m5949_04_01.
[20] Y. Xue, G. Nestor, Determination of amide cis/trans isomers in N-acetyl-D-glucosamine: Tailored NMR analysis of the N-acetyl group conformation, ChemBioChem23 (2022) e202200338. https://doi.org/10.1002/cbic.202200338.
[21] J.E. Nutting, M. Rafiee, S.S. Stahl, Tetramethylpiperidine N-Oxyl (TEMPO), Phthalimide N-Oxyl (PINO), and Related N-Oxyl Species: Electrochemical Properties and Their Use in Electrocatalytic Reactions, Chem. Rev 118(9) (2018) 4834–4885. https://doi.org/10.1021/acs.chemrev.7b00763.
[22] K. He, S. Feng, W. Yang, K. Yang, Z. Yin, Y. He, J. Ma, Synthetic Electrochemical Technologies, Energy Environ. Prot 39 (2025) 34–47. https://doi.org/10.20078/j.eep.20241206.
[23] R. Hu, Y. Zhao, C. Tang, Y. Shi, G. Luo, J. Fan, J.H. Clark, S. Zhang, Electrochemical Biorefinery toward Chemicals Synthesis and Bio-Oil Upgrading from Lignin, Engineering 27 (2023) 178–198. https://doi.org/10.1016/j.eng.2022.10.013.
[24] M. Regnier, C. Vega, D.I. Ioannou, T. Noël, Enhancing electrochemical reactions in organic synthesis: the impact of flow chemistry, Chemical Society Reviews 53(21) (2024) 10741–10760. http://dx.doi.org/10.1039/D4CS00539B.
[25] Y. Niu, W. Hu, Preparation, characterization and application in environmental protection of low-molecular-weight chitosan: A review, Sustain. Environ. Res 34 (2024) 29. https://doi.org/10.1186/s42834-024-00236-8.
[26] X. Zhou, Q. Liu, X. Chen, N. Zhou, G. Wei, F. Chen, K. Chen, Enzymatic hydrolysis of chitinous wastes pretreated by deep eutectic solvents into N-acetyl glucosamine, Polym. Degrad. Stab 227 (2024) 110907. https://doi.org/10.1016/j.polymdegradstab.2024.110907.
[27] A. Pandit, A. Indurkar, C. Deshpande, R. Jain, P. Dandekar, A systematic review of physical techniques for chitosan degradation, Carbohydr. Polym. Technol. Appl 2 (2021) 100033. https://doi.org/10.1016/j.carpta.2021.100033.
[28] A.B. Muley, M.R. Ladole, P. Suprasanna, S.G. Dalvi, Intensification in biological properties of chitosan after γ-irradiation, Int. J. Biol. Macromol 131 (2019) 435–444. https://doi.org/10.1016/j.ijbiomac.2019.03.072.
[29] N. Tabassum, S. Ahmed, M.M. Ittisaf, M. Rakid-Ul-Haque, M.A. Ali, A green approach for depolymerization of chitosan: applications in hydrogels, Cellulose 30(14) (2023) 8769–8787. https://doi.org/10.1007/s10570-023-05372-9.
[30] S. Suryani, A.Y. Chaerunisaa, I.M. Joni, R. Ruslin, L.O.A.N. Ramadhan, Y.W. Wardhana, S.H. Sabarwati, Production of Low Molecular Weight Chitosan Using a Combination of Weak Acid and Ultrasonication Methods, Polymers 14(16) (2022) 3417. https://www.mdpi.com/2073-4360/14/16/3417.
[31] Y. Kato, J. Kaminaga, R. Matsuo, A. Isogai, TEMPO-mediated oxidation of chitin, regenerated chitin and N-acetylated chitosan, Carbohydr. Polym 58(4) (2004) 421–426. https://doi.org/10.1016/j.carbpol.2004.08.011.
[32] H. Zhao, D. Lu, J. Wang, Y. Wu, Y. Yang, X. Zhang, Q. Xu, X. Sun, Raw biomass electroreforming coupled to green hydrogen generation, Nat. Commun 12 (2021) 2008. https://doi.org/10.1038/s41467-021-22250-9.
[33] H. Mojtabazadeh, J. Safaei-Ghomi, High conductivity graphite paste for radio frequency identification tag with wireless hydrogen sensor based on CeO2–Fe2O3–graphene oxide, RSC Advances 15(16) (2025) 12773–12784. https://doi.org/10.1039/D5RA00587F.
[34] N.X. Vo, N.N.H. Le, T.D.P. Chu, H.L. Pham, K.X.A. Dinh, U.T.T. Che, T.T.T. Ngo, T.T. Bui, Effectiveness and Safety of Glucosamine in Osteoarthritis: A Systematic Review, Pharmacy 11(4) (2023) 117. https://www.mdpi.com/2226-4787/11/4/117.
[35] M. Romanelli, Unique combination of hyaluronic acid and amino acids in the management of patients with a range of moderate-to-severe chronic wounds: Evidence from international clinical trials, International Wound Journal 21(S1) (2024) 4–8. https://doi.org/10.1111/iwj.14617.
[36] A. Połubinska, J. Cwalinski, E. Baum, A. Bręborowicz, N-Acetylglucosamine modulates function of the skin fibroblasts, Int. J. Cosmet. Sci 35(5) (2013) 472–476. https://doi.org/10.1111/ics.12067.