Nickel-DNA Complexes: Bioelectrocatalysis or Not?

Journal of The Electrochemical Society, 160 (8) H463-H468 (2013) 0013-4651/2013/160(8)/H463/6/$31.00 © The Electrochemical Society

H463

Nickel-DNA Complexes: Bioelectrocatalysis or Not? Garett G. W. Lee∗ and Shelley D. Minteer∗∗,z Departments of Chemistry and Materials Science and Engineering, University of Utah, Salt Lake City, Utah 84112, USA Alkaline fuel cells (AFC) are low temperature, quick-to-start devices that can achieve 50% operating efficiency. Low cost alternatives to platinum group electrocatalysts, which allow for direct reformation are desired. Nickel electrocatalysts are highly active in alkaline for the oxidation of fuels. However, these are ‘low-density’ catalysts, and in order to improve loading, complexing agents (e.g., citrate) and are utilized. Recently, much attention has been given to electrodeposited Ni-DNA complexes that give increased catalytic activity compared to their metal film counterparts, however, the mechanism is poorly understood. That is, are these complexes achieving this activity through physical or chemical mechanisms imparted by the DNA (bioelectrocatalysis) or merely the chemical constituents of DNA? Here, we set out to analyze these systems, focusing primarily on the effect of phosphate on nickel species, given the high concentration of phosphate groups within DNA. © 2013 The Electrochemical Society. [DOI: 10.1149/2.064308jes] All rights reserved. Manuscript submitted March 27, 2013; revised manuscript received May 7, 2013. Published May 21, 2013.

Alkaline fuel cells are among the most efficient low-temperature operating fuel cells.1 The catalysts have traditionally been platinumgroup metals or their alloys, but these quick-to-start devices are, however, highly sensitive to fouling (e.g., CO2 ), and require costprohibitive, high purity H2 for long term stability. Group 4 metal compounds, sought as a low cost alternatives, show significant activity toward the oxidation of many aliphatic species in alkaline media and are not prone to poisoning.2–4 Early work on nickel-phosphides, specifically for corrosion resistance, is a well-studied field of materials.5 Preparation of these alloys relies on elevated temperature and low pH for the deposition of a variety of stoichiometries, from hypophosphite and phosphine gas precursors. More recent analyzes of nickel phosphorous materials emphasizes the catalytic capabilities of these materials, among these are phosphides for hydroprocessing6 and phosphates for electrocatalysis.7 Metals such as nickel, manganese, and cobalt show electrochemical activity in alkaline environments toward both oxidation and reduction.8 The local environment of the electrodeposited metal, whether in an complex (e.g., citrate)9 or a co-deposited species (e.g., with Co, Fe or Cu),2,10 has been shown to greatly affect catalytic properties and longevity of the catalyst system. Recently, work on cobalt complexes has shown that inclusion of phosphate increases longevity of the system at high potentials, resulting in a ‘self-healing’ catalyst.11,12 However, no conclusive evidence indicating the role of the phosphorous in the catalyst is given. That is, the use of phosphate in preparing nickel complexes for catalytic materials has been studied, but the inclusion of phosphate as either an active component of the catalyst or active in depositing the metal during deposition, has not. Nickel-DNA complexes have more recently been explored as high-activity nickel complexes for fuel oxidation in alkaline systems, specifically for selective methanol oxidation.13 In a recently published study, we analyzed such Ni-DNA species. These ‘bioelectrocatalysts’ display increased catalytic activity to various aliphatic alcohols and sugars for applications in alkaline fuel cells.14 The catalytic mechanism and the role of DNA remain unclear. It has been hypothesized that these effects are attributed to the phosphate backbone of DNA, in that DNA provides both a physical scaffold and a conductive pathway (i.e., the phosphate backbone) for the nickel electrocatalyst. However, given the relatively high electric field experienced during electrodeposition protocols (105 V cm−1 ), it is likely that DNA is denatured at the heterogeneous interface and that merely the chemical constituents of DNA are complexing with nickel. It is hypothesized here that the DNA is merely acting as a source of phosphate for the complexation of nickel-phosphate aggregates analogous to the aforementioned cobalt phosphate species.15

∗

Electrochemical Society Student Member. Electrochemical Society Active Member. z E-mail: [email protected]

∗∗

To explore this theory, we have focused on nickel-phosphate aggregates, using nickel-phosphate based electrocatalysts for analyte oxidation of methanol, formaldehyde, and formate in alkaline solutions. We demonstrate increased sensitivity of a nickel species co-deposited from phosphate solutions versus chloride analogs, and have compared them to similarly prepared Ni-DNA complexes. The electrochemical response of nickel-phosphate is different from co-deposited chloride controls at the 99.99% confidence level. The role of the phosphate is considered, but empirical evidence is not forthcoming. Characterization of the catalytic surface is performed with both electrochemical analysis, including voltammetry and amperometry, and spectroscopic analysis via X-ray photoelectron spectroscopy (XPS). To provide additional information to elucidate structure-function / compositionfunction relationships of the catalytic surface, characterization of the samples were performed via atomic force microscopy (AFM). Experimental Nickel electrocatalyst coated electrodes are prepared via electrodeposition from electrolyte solutions of either 0.1 M NaCl (control), 0.1 M K2 HPO4 solutions, or 0.1 mg/mL DNA(Calf thalamus DNA, Sigma). Electrodes are prepared on wet-proofed Toray paper, TGPH-060 (Fuel Cell Earth, Stoneham, MA). The Toray electrodes are a standard 1 cm2 geometric working area, where a wax coating is applied to define the surface area. Solutions containing either 1 M methanol (Fisher Sci, HPLC grade), 0.1 M formaldehyde (Aldrich), or 0.1 M sodium formate (Fisher Sci) are used to collect voltammetric data. Electrodes are immersed in solution to equilibrate for approximately five minutes before evaluation. (The sodium formate solution is degassed with N2 to prevent spontaneous formate oxidation.) Electrodeposition occurs in a 0.5 mg/mL solution of NiCl2 (Aldrich, anhydrous, 99.99%). Control electrodes were deposited with a 0.1 M NaCl in 18 M water (Millipore, MilliQ). Ni-DNA electrodes are prepared from 0.5 mg/mL NiCl2 and 0.1 mg/mL DNA solutions. K2 HPO4 (Fisher) solutions of 10, 50, 100, 250 and 500 mM were used for phosphate optimization. An electrode array is used for simultaneous deposition as described in Reference 3. All depositions and statistical evaluations utilize n = 3 (different) electrodes. Electrodeposition of the Ni2+ occurs for 1800 seconds at 1.8 V versus a Ag|AgCl reference electrode, consistent with M-DNA electrodeposition protocols.13,14 The counter electrode used is a large Pt mesh. Electrodeposition and electrochemical analysis are done using a Digi-Ivy 2300 bipotentiostat with the same three-electrode setup. All potentials are referenced versus Ag|AgCl. XPS and AFM samples are deposited in the same manner as Toray based electrodes, except glassy carbon plates (2 cm × 0.3 cm × 0.1 cm) are substituted instead of Toray paper. The glassy carbon is polished with 1 and 0.05 μm alumina and rinsed with copious amounts of DI before deposition. AFM, run on a Bruker Dimension

Downloaded on 2013-05-29 to IP 155.101.69.71 address. Redistribution subject to ECS license or copyright; see ecsdl.org/site/terms_use

H464

Journal of The Electrochemical Society, 160 (8) H463-H468 (2013)

Icon, analyzed surface areas of 1.19 × 106 nm2 for Ni-Pi and 1.03 × 106 nm2 for Ni-Cl. XPS and elemental analysis were performed using a Kratos Axis Ultra DLD, utilizing a corrected binding energy of 0.27 / eV for Ni-Cl sample and 0.22 / eV for Ni-Pi sample. A sixty second argon sputtering was performed to remove surface adsorbates. Results and Discussion Demonstrating the greatest activity in alkaline environments, all electrochemical analysis of nickel deposited electrodes use solutions of 0.1 M NaOH and 18 M water in addition to appropriate organic additives. The voltammetric characterization requires an initial cycling to achieve a steady-state response. Here, this response is achieved after 10 cycles. The generally accepted reaction scheme for nickel electrocatalysts in alkaline solutions is as follows:4,16 − OH− + Ni(OH)2   NiOOH + H2 O + e

An initial formation of nickel oxyhydroxide is followed by substrate adsorption: RCH2 OHsol   RCH2 OHadsorbed An intermediate complex that is associated with a proton shift to the catalytic surface then occurs: [RCH2 OHadsorbed + NiOOH]   [RCHOHadsorbed + Ni(OH)2 ] Finally, product desorption occurs, which is accompanied by the regeneration of the heterogeneous catalytic surface: − [RCHOHadsorbed + Ni(OH)2 ]   Ni(OH)2 + RCHOO or RCHOOH

Figure 1 shows the cyclic voltammetric response (swept initially in the oxidative direction here, and throughout the study) of the nickel electrodes in a 0.1 M NaOH solution, deposited from 0.1 M sodium chloride as control (Ni-Cl), 0.1 M monobasic phosphate (Ni-Pi), and 0.1 mg/mL Ni-DNA. Chloride was chosen here as control as the nickel counter ion source, NiCl2 , contributes to an approximate chloride concentration of 0.1 mM in all deposition solutions. Chloride ion was speculated by Horkans as a possible competitive deposition species with nickel; however, this is not a well understood process.17 The phosphate concentration was analyzed between 10 and 500 mM. The response of these films is given in Figure 2; the inlay of 250 and 500 mM phosphate are separated for clarity. Concentrations above 100 mM phosphate in the deposition baths leads to the loss of nickel activity (i.e., Ep ox ≈ 0.560 V vs. Ag|AgCl). When the phosphate concentration in the deposition bath is increased to 500 mM, the current response at the characteristic nickel oxidation

Figure 2. Effects of phosphate concentration on electrocatalytic response; representative cyclic voltammograms for electrochemical response at 50 mV/sec in 0.1 M NaOH, concentrations labeled in the figure.

peak potential (i.e., Ep ox ≈ 0.560 V vs. Ag|AgCl), again increases, but as a purely capacitive current. This current response does not correlate to nickel deposition or to the electrocatalytic capabilities of the deposited films. In effect, a traditional faradaic analysis does not accurately correlate deposition current to the resulting electrocatalysis of substrate. Differences in current densities between Ni-Cl and Ni-Pi deposited films are found in Table I. Where again, averages account for n = 3 (different) electrodes. Statistical evaluations include traditional Spooled and Tcalc analyzes, with values compared to two-tailed Student T values. Table II gives the effective oxidative current response for Ni2+/3+ by varying the concentration of phosphate between 10 and 500 mM. The current response of the Ni-DNA complex is also given. The NiDNA films give highly variable current response. However, if the mass concentration of phosphate within DNA is approximately 10%, the current response of the Ni-DNA system agrees with the trend seen for Ni-Pi complexes at the low concentration end. The reason for the current response at the high end for Ni-DNA complexes is less forthcoming. It could be as simple as increased capacitance from electrodeposited DNA not participating in any electrocatalytic activity. Table I. Oxidative response of fuels in alkaline solution at Ni-Cl and Ni-Pi electrodes, peak ratio of Ni3+ /fuel oxidation. Solution

J / mA cm−2

J-ratio

Ni-Cl Ni-Pi

0.1 M NaOH

−1.1 (± 0.1) −3.5 (± 0.6)

4.8 (± 0.3) 2.1 (± 0.1)

Ni-Cl Ni-Pi

1.0 M methanol

−2.9 (± 0.2) −8.6 (± 0.5)

12 (± 2) 5 (± 1)

Ni-Cl Ni-Pi

0.1 M formaldehyde

−3.2 (± 0.3) −5.2 (± 0.4)

13 (± 3) 3.1 (± 0.6)

Ni-Cl Ni-Pi

0.1 M formate

−4.3 (± 0.2) −6.6 (± 0.7)

18 (± 3) 4 (± 1)

Electrode

Table II. Voltammetric current response for deposition of nickel in varying [phosphate] deposition baths; currents taken from Ni ip ox for catalysts prepared between 0 and 100 mM phosphate; for 250 and 500 mM deposition baths, currents measured at 0.56 V vs. Ag|AgCl when no nickel signal is observed. [Phosphate] (mM) Figure 1. CVs of nickel electrodes in a 0.1 M NaOH solution. The gray curve corresponds to nickel electrodes formed by electrodeposition in a 0.1 M NaCl solution, the black curve corresponds to nickel electrodes formed by electrodeposition in a 0.1 M phosphate solution, and the dashed gray corresponds to Ni-DNA; scan rate of 50 mV/sec. The peaks correspond to the oxidation and reduction of Ni2+/3+ , while the current at 0.8 V corresponds to water oxidation.

0 10 50 100 250 500 DNA (0.1 mg/mL)

Ep

ox

(V)

0.522 (± 0.008) 0.492 (± 0.003) 0.518 (± 0.006) 0.540 (± 0.007) N/A N/A 0.499 (± 0.031)

Jp ox (mA cm−2 ) −0.24 (± 0.04) −0.09 (± 0.03) −0.47 (± 0.17) −1.72 (± 0.39) −0.19 (± 0.06) −0.57 (± 0.05) −0.47 (± 0.33)

Downloaded on 2013-05-29 to IP 155.101.69.71 address. Redistribution subject to ECS license or copyright; see ecsdl.org/site/terms_use

Journal of The Electrochemical Society, 160 (8) H463-H468 (2013)

Figure 3. Electro-oxidation of methanol: a) Representative cyclic voltammograms at 50 mV/sec in a 0.1 M NaOH and 1 M methanol solution: Ni-Cl (gray), Ni-Pi (black); inlay: representative chronoamperometry, Ni-Cl (gray), Ni-Pi (black); b) calibration curves of Ni-Cl (gray) and Ni-Pi (black); linear fit of y = 3.63 × 10−5 x –2.67 with an R2 = 0.98 for chloride (gray) and −2.18 × 10−4 x + 8.89 with an R2 = 0.98 for phosphate (black).

Given these observed trends, the Ni-Pi complexes were further explored for their oxidative catalytic ability toward a variety of hydrocarbons and are compared to chloride controls. Voltammetric characterization shows the enhanced activity of phosphate deposited electrodes for both nickel oxidation and reduction, as well as the onset for water oxidation (oxygen evolution), seen in Figure 1, versus chloride controls. This is rationalized as the formation of catalytic hydroxyl groups on the surface of the Ni2+ catalyst, as given in the mechanism above. After stabilizing, the enhanced Ni2+ oxidation signal at 0.52 V is increased by 85% versus Ni-Cl controls. This increase is statistically significant at the 99% confidence level from Ni-Cl. A 68% increase in water oxidation current, at 0.8 V vs. Ag|AgCl, is also statistically significant at the 99% confidence level, while Ni3+ reduction is increased by 84% at 0.44 V vs. Ag|AgCl, which is statistically significant at the 99% confidence level. Activity of nickel complexes in alkaline solutions for oxidation of various aliphatic alcohols has led to the following accepted oxidation mechanism.4 For methanol oxidation, a six electron process occurs as follows with carbonate as the final oxidation product. 2− − − CH3 OH → H2 CO + 2e− → HCO− 2 + 2e → CO3 + 2e

Increased affinity for hydrocarbons helps to explain the oxidative character of the current response for the system in 1 M methanol and 0.1 M NaOH. The voltammograms seen in Figure 3a and tabulated in Table I show the oxidation of methanol at nickel electrodes electrodeposited in phosphate electrolyte and chloride co-deposited controls. The current response at 0.8 V vs. Ag|AgCl shows a statistically significant increase of 68% at the 99% confidence level.

H465

Figure 4. Electro-oxidation of formaldehyde: a) Representative cyclic voltammograms at 50 mV/sec in a 0.1 M NaOH and 0.1 M formaldehyde solution: Ni-Cl (gray), Ni-Pi (black); inlay: representative chronoamperometry, Ni-Cl (gray), Ni-Pi (black); b) calibration curves of Ni-Cl (gray) and Ni-Pi (black); The linear fit for Ni-Cl (gray) is y = −8.88 × 10−5 x –3.39 × 10−4 , with R2 = 0.99, while for Ni-Pi (black) the fit is −3.21 × 10−4 x + 6.65 × 10−4 with R2 = 0.99.

Chronoamperometry demonstrates catalytic activity toward solution-based analytes over a range of concentrations. A calibration curve was generated by standard addition of each analyte in an electrolyte solution. The corresponding curves show strong linearity (R2 = 0.99) for most systems. All chronoamperometric measurements use a potential bias of 0.65 V vs. Ag|AgCl, with run times of 300 seconds. Samples are introduced to the stirring system every 30 seconds; the current measurements are taken after 10 seconds of mixing. Representative amperometric data is given in Figures 3–5 for methanol, formaldehyde, and formate analytes, along with the corresponding calibration curves (and relative error). The amperometric calibration curve for methanol, with additions of 25 μL of 1 M methanol (in 0.1 M NaOH) every 30 seconds, gives a linear plot. Sensitivity increases by nearly an order of magnitude for Ni-Pi. The responses are different at the 95% confidence level. Relative standard deviations for chloride and phosphate are 17 and 22%, respectively, with no baseline correction. Formaldehyde analyzed at 0.1 M in 0.1 M NaOH solutions, gives similar current responses to analyte oxidation, yet surprisingly lacks the nickel character of alkaline and methanol solutions. (The reason for this is not known, but is reproducible.) The oxidative wave of formaldehyde at Ni-Pi (black) shows a statistical increase of 32% over Ni-Cl at the 99% confidence level. The amperometric response to standard additions in formaldehyde solutions is given in the inlay of Figure 4a and tabulated in Table I, where 250 μL of 0.1 M formaldehyde in 0.1 M NaOH was added to solution every 30 seconds. The corresponding calibration curve is seen in Figure 4b. Again, the difference in response is significant at the 99% confidence level and more

Downloaded on 2013-05-29 to IP 155.101.69.71 address. Redistribution subject to ECS license or copyright; see ecsdl.org/site/terms_use

H466

Journal of The Electrochemical Society, 160 (8) H463-H468 (2013)

Figure 5. Electro-oxidation of formate: a) Representative cyclic voltammograms at 50 mV/sec in a 0.1 M NaOH and 0.1 M formate solution: Ni-Cl (gray), Ni-Pi (black); inlay: representative chronoamperometry, Ni-Cl (gray), Ni-Pi (black); b) calibration curves of Ni-Cl (gray) and Ni-Pi (black); with fits of y = −8.8 × 10−6 x–6.29 × 10−5 for Ni-Cl (gray) at R2 = 0.97, and y = −9.62 × 10−6 –8.83 × 10−4 with R2 = 0.99 for Ni-Pi (black).

sensitive for Ni-Pi by 260%. Relative error for Ni-Cl is 5.6% and 9.9% for Ni-Pi. When considering the complete oxidation of aliphatic alcohols (e.g., methanol) to CO2 , the next logical analyte is formate. Formate, however, does not exhibit the enhanced sensitivity at Ni-Pi as in the cases of methanol and formaldehyde. Although the current response at 0.8 V vs. Ag|AgACl is enhanced again by 24% at the 99% confidence level, at the potential chosen here for amperometric analysis (0.65 V), the increased sensitivity (i.e., slope) is not present (Figure 5). The calibration curve, generated in the same manner as formaldehyde, shows an increase in sensitivity of only 9%, different at the 90% confidence level. However, if baseline is normalized for increased current response at 0 mM formate, this difference is negated. This result implies that selectivity is present in the case of Ni-Pi catalyst for the first two oxidative steps in methanol oxidation. Nickel electrodeposits into nanoparticle aggregates on electrode surfaces. Surface analysis of nickel deposited substrates by atomic force microscopy (AFM) in Figure 6 show a relative increase of 40% (4.3 to 7.3 nm) in the half-height of deposited aggregates when nickel is co-deposited with phosphate. Elemental analysis via X-ray photoelectron spectroscopy (XPS), given in Table III, reveals increased deposited nickel content (3x increase); this may account for the difference. However, changes to inter-aggregate structure are not revealed. (Ni-DNA aggregates were analyzed in Reference 14.) As was true with the cobalt system studied by Nocera et. al.,11,12 increased Ni content may be attributed to a deposition mechanism that relies on phosphate present in the system as seen in Table I. In the cobalt-phosphate system, phosphate is speculated as acting as a proton accepting electrolyte; in that during water oxidation, the abundance of

Figure 6. AFM image of nickel deposited surfaces: a) Ni-Pi with average film half-height of 7.3 nm and b) Ni-Cl with average film half height of 4.3 nm. Larger particulate remnants present in a) due to adventitious material. (RMS roughness of polished glassy carbon ≈ 0.6 nm).

phosphate in the system prevents electrodeposited catalyst leaching from the surface. Other common electrolytes, including sulfate, nitrate and hypochlorite did not demonstrate the same properties. DNA is a phosphate rich system, comprised of approximately 10% (by wt) phosphate. For Ni-DNA complexes, it is not without reason to hypothesize that DNA acts as a phosphate source, encouraging metal deposition. Gileadi offers significant insight on the theory of metal deposition.18 Analogous to metal deposition, the evolution of gasses at metal electrodes first requires the formation of modified layers, e.g., the formation of PtO before oxygen evolution; a process that requires a larger overpotential than necessary. Gileadi also offers a theory on charge transfer, in short, that adsorbed species (such as anions)

Table III. XPS data for Ni-Cl and Ni-Pi electrode compositions. Peak

Electrode

Atomic%

Mass%

O 1s

Ni-Cl Ni-Pi Ni-Cl Ni-Pi Ni-Cl Ni-Pi Ni-Cl Ni-Pi Ni-Cl Ni-Pi Ni-Pi

1.81 8.38 1.40 1.08 95.55 86.06 0.96 0.57 0.29 0.96 0.13

2.32 9.75 1.57 1.10 92.03 75.24 2.73 1.48 1.36 4.10 0.29

N 1s C 1s Cl 2p Ni 2p P 2p

Downloaded on 2013-05-29 to IP 155.101.69.71 address. Redistribution subject to ECS license or copyright; see ecsdl.org/site/terms_use

Journal of The Electrochemical Society, 160 (8) H463-H468 (2013)

H467

Figure 7. CV of Ni-Cl (solid gray), Ni-Pi (dash black), and Ni from a two-step deposition (solid black); 0.1 M NaOH, v = 50 mV/sec.

at the inner Helmholtz plane facilitate electron transfer between the electrode and metal in solution which acts to lower or eliminate the activation energy.19 However, as Gileadi states, this theory has not been treated in literature and no experimental evidence exists. Likewise, Conway and Bockris extensively analyzed mechanisms for the electrolytic deposition of metals.20 The authors propose that deposition occurs through a series of steps, including ion transfer (i.e., metal ion) from solution to the electrode surface, surface diffusion to more favorable sites, loss of the solvation shell (dehydration), and ultimately the redox process from metal ion (Mn+ ) to deposited metal (M0 ). Direct deposition and redox is unlikely, as prohibitively high H values exist. The analysis specifically speculates on the possibility of intermediate ion states assumed during deposition, and their role on facilitating a decrease in the energy of activation. However, energy calculations for these processes by Conway and Bockris are based upon initial and final states of metal species, based upon enthalpies or chemical potential energy. Neither Gileadi nor Conway and Bockris offer empirical evidence for this deposition mechanism. Conway and Bockris, do in a later work,21 indicate that current exchange densities for Ni2+ deposition are low (relative to Ag+ and Cu2+ ) due to the instability of Ni+ ions in aqueous solutions. We hypothesized that phosphate acts to stabilize this intermediate species, providing an environment for Ni+ . In a recently published study on the electrolyte effects of manganese oxygen reduction electrocatalysts, it was observed that the electrolyte deposition bath has a substantial effect on the electrocatalytic properties of the manganese.22 The trends in the manganese study are also unclear, as it does not obey the trends of phosphate, nor metal-oxygen coordination. To test our hypothesis and determine if phosphate facilitates the deposition of nickel through the formation of a surface modified layer, glassy carbon electrodes were surface modified through three difference mechanisms. Glassy carbon was used here to lend higher confidence to the known surface area. Before deposition (as described above) glassy carbon electrodes were polished with alumina powder, rinsed, and sonicated in DI to remove any adsorbates. Two depositions proceeded as above, from 0.5 mg/mL NiCl2 solutions containing either 0.1 M NaCl or K2 HPO4 . The third variation required two depositions: first, a 3 minute, preferential phosphate deposition from 0.1 M K2 HPO4 solution. And second, 30 minute deposition for a 0.5 mg/mL NiCl2 in 0.1 M NaCl solution. As can be seen in the CV in Figure 7, a drastic change in both the nickel character and nickel content of the Ni2+/3+ peak in 0.1 M NaOH is observed. Further analysis by XPS into the oxidation state of the nickel reveals no difference in nickel character between Ni-Cl and Ni-Pi deposited systems. Analyzed as per Smart and Biesinger,23,24 fitting the nickel peak shows that the deposited nickel species are primarily metallic in nature. Spectra of the Ni 2p signal, shown in Figure 8, reveals increased nickel in the Ni-Pi system; these values are given in Table III. Typically, electrodeposited nickel is considered to be diva-

Figure 8. XPS spectra for Ni 2p signal: a) Ni-Pi, b) Ni-Cl; for both spectra: Ni 2p spectra (black), overall fit (red), Ni0 (metallic, gray), satellite 1 (green), satellite 2 (orange), background (blue).

lent (e.g., NiO) in alkaline solutions.4 Analyzes on planar or ‘massive’ nickel metal surfaces do not generate the same catalytic character as the deposited species.25 The phosphorus, it appears, enables both greater nickel content to be deposited, and deposited in a manner (structural) that enables greater catalytic activity. Conclusions We have analyzed electrodeposited nickel complexes that function as electrocatalysts for water and aliphatic alcohol oxidation. Complexing agents have been shown to increase the activity of these nickel species, among them DNA.13 However, the role of the DNA in these complexes was unclear. Therefore, it was hypothesized here that the DNA component of these ‘biocatalysts’ serves merely as a source of phosphate, and are not an active component of the complex. To test this theory, nickel-phosphate complexes were formed through the deposition of nickel in K2 HPO4 electrolyte solutions and were compared to previously studied to Ni-DNA complexes.14 These Ni-Pi complexes demonstrated increased sensitivity to methanol and formaldehyde oxidation versus chloride controls. For the Ni-DNA species, at approximately 10% phosphate (m/m), the current response correlates well to the concentration of phosphate in solution. However, the Ni-DNA complexes are highly variable in their electrochemical behavior. The incorporation of phosphate through HPO4 2− produces both increased activity and decreased variance versus the DNA complex system. Additionally, these results show that DNA is not needed for improved electrocatalysis when phosphate acts as a more consistent and cheaper alternative. The relationship between increased catalytic activity and phosphate incorporation could be one of many: increased surface area (as shown by AFM), increased Ni content (as shown by XPS), or a change in the carbon surface layer that facilities nickel deposition possibly

Downloaded on 2013-05-29 to IP 155.101.69.71 address. Redistribution subject to ECS license or copyright; see ecsdl.org/site/terms_use

H468

Journal of The Electrochemical Society, 160 (8) H463-H468 (2013)

through double-layer disruption or an electron tranport mechanism. As is often the case, this analysis raises more questions, specifically the mechanism by which phosphate facilitates increased nickel deposition. Further research will focus on understanding the detailed structure/function relationship of nickel electrocatalysts formed by electrodeposition of nickel in phosphate electrolytes and anion effects on metal deposition. Acknowledgment The authors thank the Utah Science and Technology Initiative (USTAR) and the National Science Foundation for financial support. References 1. A. Kirubakaran, S. Jain, and R. K. Nema, Renewable and Sustainable Energy Reviews, 13, 2430 (2009). 2. I. Danaee, M. Jafarian, F. Forouzandeh, F. Gobal, and M. G. Nahjani, International Journal of Hydrogen Energy, 33, 4367 (2008). 3. M. Jafarian, R. Moghaddam, M. Mahjani, and F. Gobal, Journal of Applied Electrochemistry, 36, 913 (2006). 4. A. Kowal, S. Port, and R. Nichols, Catalysis Today, 38, 483 (1997). 5. R. L. Zeller and U. Landau, J. Electrochem. Soc., 139, 3464 (1992). 6. S. T. Oyama, X. Wang, Y. K. Lee, K. Brando, and F. G. Requejo, Journal of Catalysis, 210, 207 (2002).

7. R. M. A. Hameed and K. M. El-Khatib, International Journal of Hydrogen Energy, 35, 2517 (2010). 8. M. Dinca, Y. Surendranath, and D. Nocera, Proceedings of the National Academy of Sciences, 107, 10337 (2009). 9. O. Berkh, Y. Shacham-Diamand, and E. Gileadi, J. Electrochem. Soc., 158, F85 (2011). 10. T. Lambert, D. Davis, W. Lu, S. Limmer, P. Kotula, A. Thuli, M. Hungate, G. Ruan, Z. Jin, and J. Tour, Chem. Comm., 48, 7931 (2012). 11. D. Lutterman, Y. Surendranath, and D. Nocera, JACS Communications, 131, 3838 (2009). 12. Y. Surendranath, M. Kanan, and D. Nocera, Journal of the American Chemical Society, 132, 16501 (2010). 13. Y. Liu, W. Wei, X. Liu, X. Zeng, Y. Li, and S. Luo, Microchim Acta, 168, 135 (2010). 14. D. Chen, G. G. W. Lee, and S. D. Minteer, ECS Electrochemistry Letters, 2, F9 (2013). 15. Y. Surendranath, M. Dinca, and D. G. Nocera, J. Am. Chem. Soc., 131, 2615 (2009). 16. N. Leventis and X. Gao, Anal. Chem., 73, 3981 (2001). 17. J. Horkans, Journal of the Electrochemistry Society, 126, 1861 (1979). 18. E. Gileadi, Journal of Solid State Electrochemistry, 15, 1359 (2011). 19. E. Gileadi, Journal of Electroanalytical Chemistry, 660, 247 (2011). 20. B. E. Conway and J. O. M. Bockris, Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences, 248, 394 (1958). 21. B. E. Conway and J. O. M. Bockris, Electrochimica Acta, 3, 340 (1961). 22. G. G. W. Lee and S. D. Minteer, ACS Sustainable Chemistry & Engineering, 1(3), 359 (2013). 23. M. C. Biesinger, B. P. Payne, L. M. W. Lau, A. Gerson, and R. S. C. Smart, Surf. Interface Anal., 41, 324 (2009). 24. A. P. Grosvenor, M. C. Biesinger, R. S. C. Smart, and N. S. McIntyre, Surface Science, 600, 1771 (2006). 25. S. Maximovitch and S. Bronel, Electrochemcial Acta, 26, 1331 (1981).

Downloaded on 2013-05-29 to IP 155.101.69.71 address. Redistribution subject to ECS license or copyright; see ecsdl.org/site/terms_use