Characterization of the alginate oxidation reaction on catalytic materials
Based on our previous studies on the catalytic oxidation of cellulose, we first investigated alginate oxidation on a conventional Au electrode as alginate and cellulose share the common polysaccharide structure with polyol groups.11 The alginate oxidation reaction was examined by using an Au-coated glassy carbon (GC) electrode in 0.2 % sodium alginate dissolved in different media, which ranged from acidic to basic. Different from basic and neutral solvents, the use of an acidic solution resulted in the precipitation of sodium alginate, thus the measurement could not be performed. Cyclic voltammograms of Au-coated GC in 0.5 M NaOH featured the oxidation peak at 0.43 V in the forward scan and the reduced reduction peak of gold oxide at 0.06 V in the backward scan. This reduction indicates a strong adsorption of alginate and/or intermediates on the Au surface under alkaline conditions (Figure S1 A in the Supporting Information). In contrast, the solution of 0.2 % alginate in neutral solvent (phosphate-buffered saline (PBS), pH 7) did not induce significant changes in the forward and backward peaks, which indicates that alkaline conditions are preferred over neutral or acidic conditions (Figure S1 B). This is in agreement with previous studies on direct biomass-derived fuel oxidation on Au surfaces.11–13 Compared to pretreated cellulose, alginate oxidation is advantageous in that it does not require any energy-consuming pretreatment.11 Nevertheless, the fact that the alginate oxidation peak occurred at a positive potential indicated a relatively low power output of this direct alginate-based fuel cell.
To overcome this hurdle, we utilized Au nanoparticles (AuNPs) conjugated to functionalized multiwalled carbon nanotubes (AuNPs/MWCNTs) to enhance both the catalytic activity and total active surface area.14 For a better correlation between the respective geometric surface areas and for comparison to conventional electrodes, we compared AuNPs/MWCNT-modified carbon sheets (AuNPs/MWCNTs/C) with Au sheets (instead of Au on glassy carbon). We considered the structure–activity relationships of AuNP-catalyzed monosaccharide oxidation12 and tuned the size of the AuNPs to find the catalytic system with the highest homogeneity and activity for the alginate oxidation reaction by utilizing carbon nanotubes (Figure 2) of different sizes and loading amounts (Figure S2).
Figure 2. The effect of the carbon nanotube diameter on the sintered-AuNPs size (upon thermal treatment at 400 °C) and the corresponding alginate oxidation current. A) SEM images of AuNPs on i, i′) small MWCNTs (S_MWCNTs, diameter of 3–20 nm), ii, ii′) medium MWCNTs (M_MWCNTs, diameter of 40–70 nm), and iii, iii′) large MWCNTs (L_MWCNTs, diameter of 110–170 nm). B) Summary of the sintered-AuNP diameter data shown in part A.
Download figure to PowerPoint
Medium MWCNTs induced the smallest and most homogeneous AuNPs upon thermal treatment (Figure 2, Scheme S1). Both small and large MWCNTs resulted in a bigger AuNP size and less homogeneous distribution than that from medium MWCNTs, which led to a lower total surface area of AuNPs. The optimal system consisted of 6.9 nm AuNPs deposited on medium-sized functionalized carbon nanotubes (40–70 nm in diameter) at loadings of 0.45 mg cm−2 Au and 0.5 mg cm−2 CNTs.
To study alginate oxidation by the catalytic materials, we first performed cyclic voltammetry (CV) within a potential window of −0.4 to 0.8 V versus Ag/AgCl in the absence and presence of alginate in 0.5 M NaOH. The CV scans of the AuNPs/MWCNTs/C electrode resulted in a clear alginate oxidation peak at 0.108 V followed by a gold oxide formation peak at 0.264 V (Figure 3 A).15 In contrast, the Au sheet electrode only demonstrated a small shoulder at around 0.3 V, which indicated the enhanced reaction kinetics on AuNPs owing to the 3D carbon nanotubes.12, 14 Importantly, the current density of the AuNPs/MWCNTs/C electrode was ∼53 times higher than that of the Au sheets, even though the active surface area of Au (calculated based on the Au oxide reduction peak on the backward scans) of AuNPs/MWCNTs/C was ∼24 times larger than that of the Au sheet. In both cases, the alginate oxidation starts at E>0.0 V versus Ag/AgCl, which suggests that the primary reaction is the oxidation of secondary alcohols on a surface with a high OH coverage rather than the oxidation of aldehyde (or hemiacetal) groups at the head of the polymer chain. To further elucidate the peak assignment, we performed the same CV experiment on AuNPs/MWCNTs/C by using 0.2–5 % glucuronic acid in 0.5 M NaOH, the monosaccharide that has a similar structure to the alginate monomer (Figure S3). At the same concentration (0.2 %), the CV of glucuronic acid featured the oxidation peak of secondary alcohol groups at 0.171 V near the alginate oxidation peak, albeit at a higher current density and earlier onset potential. If we increased the concentration of glucuronic acid to 5 %, a clear oxidation peak of hemiacetal groups at negative potential (−0.256 V) was observed, which is not the case of alginate.
Figure 3. Electrochemical characterization of alginate oxidation on AuNPs/MWCNTs/C and Au sheets. A) For each sample, data was acquired in the first cycle at a scan rate of 10 mV s−1. The electrolyte was 0.5 M NaOH with (—/—) and without (••••/••••) 0.2 % sodium alginate. B) Chronoamperometry was performed at a potential of 0.1 V. The electrolyte was 0.2 % sodium alginate in 0.5 M NaOH. C) Effect of oxidation current density on sodium alginate concentration. D) Nyquist plots of Au sheet and AuNPs/MWCNTs/C (inset) at different applied potentials in 0.2 % sodium alginate in 0.5 M NaOH.
Download figure to PowerPoint
To investigate the long-term stability of the material towards oxidation, we conducted chronoamperometry in sodium alginate at 0.1 V (Figure 3 B), which is close to the oxidation potential observed in the voltammogram shown in Figure 3 A. In the first 10 s, the Au sheet exhibited a drastic decrease in current density owing to poisoning from the alginate oxidation. In contrast, AuNPs/MWCNTs/C showed a monotonous decay and reached a steady state after approximately 2000 s, which indicated that poisoning was significantly reduced. Accordingly, the current density of the Au sheet reached its maximum saturated value at 0.1 % alginate concentration (data not shown), whereas that of the AuNPs/MWCNTs/C exhibited a linear increase up to 0.5 % alginate at a much higher current density (Figure 3 C). Furthermore, under the same measurement conditions, the Au sheet demonstrated an impedance that was more than 10 times higher than that of AuNPs/MWCNTs/C, which indicates that charge transfer occurred at a higher rate for AuNPs/MWCNTs/C (Figure 3 D). As the Au loading of the AuNPs/MWCNTs/C (0.45 mg cm−2 Au) was much lower than that of the Au sheet, the increased current density and stability is credited to the enhanced reaction kinetics, which in turn result from the unique architecture of the AuNPs/MWCNTs.
Product analysis of the alginate oxidation reaction on catalytic materials
To explore possible reaction pathways, the products that resulted from the alginate oxidation were analyzed. Based on previously reported studies on alginate oxidation by enzymes, irradiation,10, 16, 17 and strong chemical oxidants,18–21 we suggest two possible pathways (Scheme 1). The first pathway is similar to irradiation-induced and enzymatic oxidation processes and involves glycosidic bond cleavage and β-elimination, followed by oxidation at the OC site. The second pathway resembles the chemical oxidation mechanism and involves cleavage of the C2C3 bond in the monosaccharide unit, followed by electrochemical oxidation to form a binding platform for cell and tissue engineering.21
To differentiate between these mechanisms, we subjected 0.2 % sodium alginate to the electrolysis process, which was performed at the AuNPs/MWCNTs/C electrode in comparison with the Au sheet electrode in 0.5 M NaOH. The potential program consisted of the oxidation plateau, which was set at 0.108 V for 60 s, following by a cleaning pulse at −1.2 V for 1 s. This two-potential-step-process was repeated for 12 h then the final solution was collected and analyzed by HPLC, UV/Vis spectrophotometry, and attenuated total reflectance (ATR) FTIR spectroscopy.
Figure 4. Product analysis of the electrochemical oxidation of alginate. A) HPLC analysis of alginate before and after electrolysis. Calibration standard: pullulan (Shodex), column: Shodex OHpak SB-802.5 HQ, detector: refractive index (RID) and UV (inset), flow rate: 0.5 mL min−1, eluent: 0.2 M phosphate buffer (pH 7), concentration: 0.2 %. B) UV/Vis spectra. C) ATR-FTIR analysis of dried sodium alginate in 0.5 M NaOH before and after electrolysis compared to commercial sodium alginate (without NaOH). The molecular weight of P5-800 pullulans are 0.59×104, 2.11×104, 4.71×104, 20.0×104, and 70.8×104, respectively.
Download figure to PowerPoint
ATR-FTIR spectroscopy was employed to (1) investigate the interaction between NaOH and the untreated sodium alginate and (2) to determine the functional groups involved and generated in the electrochemical oxidation. Upon dissolution in 0.5 M NaOH and subsequent drying, a crystalline structure resulted, which gave rise to drastic changes in the IR vibrations (Figure 3 C). The most significant changes occurred at 3200, 1590, and 1020 cm−1, which corresponds to the ν(OH), ν(COO)asym, and ν(COC)+ν(CCO)+δ(CC) bands (i.e., bands that involve the alcohol group and the glycosidic ring).22, 23 After crystallization, these peaks were strongly diminished, which indicates that the Na+-bound OH and COO− groups of alginate might induce the dissociation of H-bonded OH to form a monodentate coordination with Oδ− atoms. As a result, the initially broad OH peak was sharpened, whereas the strong ν(COO)asym vibration decreased considerably. In contrast, the associated peak at 1410 cm−1, which corresponds to ν(COO)sym, increased as a result of the change in pH.23 If we consider the IR vibrations before and after oxidation, the CH peak at 3000 cm−1 decreased after chemical oxidation, which indicates the alteration of the CHOH bond as a result of the oxidation process. The oxidized sample further showed an increase in the intensity of ν(COO)asym, which indicates that there might be an additional formation of CO groups upon the oxidation of OH at C2 and C3 followed by C2C3 scission (i.e., the aldehyde or ketone groups), which is in agreement with the UV/Vis spectroscopy, HPLC, and electrochemical results. As the products are similar to the reported oxidized alginate used for cell and tissue engineering purposes,21 our data further suggest that the electrochemical oxidation process is not only capable of inducing partial oxidation to generate bioenergy but also yields valuable chemicals (Figure 1).
The performance of direct alginate fuel cells
A fuel cell assembly with an air-cathode chamber and a 1 mL Si anode chamber was employed to evaluate the performance of the as-prepared AuNPs/MWCNTs/C electrode as the anode in comparison with an Au sheet by using 0.2 % alginate in 0.5 M NaOH at room temperature. The AuNPs/MWCNTs/C electrode was freshly synthesized and the Au sheet was cleaned thoroughly by boiling in Piranha solution followed by hot ultrapure water. A cleaned Pt mesh served as the cathode and was pressed on to the opposite side of a Nafion membrane and mounted inside the fuel cell as illustrated in Figure 5. As expected from the electrochemical measurements, the direct alginate fuel cell (DAlgFC) that comprised AuNPs/MWCNTs/C as the anode exhibited a maximum power density of 53.7 mW m−2 (5.37 W m−3), which is 2.1 times higher than that of the DAlgFC with an Au sheet as the anode (Figure 5 A). In addition, although the open-circuit voltage of the Au sheet-based fuel cell (0.62 V) was 0.1 V higher than the AuNPs/MWCNTs/C-based fuel cell (0.53 V), which suggests less activation loss in the case of the Au sheet (Figure S4), the operating voltage steeply decreased with increasing current density. This result indicates that the Au sheet electrode induces a higher internal resistance and mass transfer limitation because of the insufficient alginate active adsorption on the Au surface compared to the AuNPs/MWCNTs/C electrode. Although this is still a factor of ∼20 lower than a direct glucuronic fuel cell (Figure S5), it does fall in the acceptable power range of microbial fuel cells and enzymatic fuel cells that do not consume any external energy (e.g., thermocontrolled bioreactors, pumps for continuously supplied fuels) and function in air-breathing mode with similar reactant concentration.24–28 Notably, the testing concentration of alginate was 0.2 %, whereas in the case of glucuronic acid the concentration was 5 %. Furthermore, because the alginate molecule is a hundred times larger than other compared monosaccharides, this difference in power density is expected and acceptable. As this is the first report on direct alginate fuel cells, further improvements both in the catalysts and in the design are needed to advance this type of fuel cell to the stage of practical application.
Figure 5. Performance of the direct alginate fuel cell that used AuNPs/MWCNTs/C as the anode in comparison with that with an Au sheet anode. A) Polarization and power density curves of direct alginate FCs fuelled with 0.2 % sodium alginate in 0.5 M NaOH. B) Photograph of our fuel cell prototype.
Download figure to PowerPoint