pubs.acs.org/Langmuir © 2010 American Chemical Society
The Influence of Solution-Phase HNO2 Decomposition on the Electrocatalytic Nitrite Reduction at a Hemin-Pyrolitic Graphite Electrode Matteo Duca, Sima Khamseh, Stanley C. S. Lai, and Marc T. M. Koper* Leiden Institute of Chemistry, Leiden University, PO Box 9502, 2300 RA Leiden, The Netherlands Received March 24, 2010 The mechanism of nitrite electroreduction by hemin adsorbed at pyrolitic graphite is investigated. Two main issues are addressed: the effect of the medium pH and the selectivity of the reaction, which was determined by the combined use of the rotating ring disk electrode (RRDE) and online electrochemical mass spectroscopy (OLEMS). In acidic media, the behavior observed is indicative of the presence of NO, as the main reactant, generated from the solution-phase decomposition of HNO2. Reduction of the NO-heme complex shows a Tafel slope of 59 mV/dec-1 and a pH dependence of 42 mV/pH, indicative of a so-called EC mechanism. In acidic media, HNO2 and NO are reduced to hydroxylamine (NH2OH) with almost 100% selectivity at low potentials, nitrous oxide (N2O) being only a minor side product. In neutral media, the hemin is largely unresponsive to the presence of nitrite, giving only a very small reduction current. The comparison of our simple heme catalyst to the behavior of the naturally occurring heme-containing nitrite reductases, which operate under biological conditions, suggests that these enzymes dissociate nitrite at neutral pH either via a complexation step favored by a specific ligating environment or by locally regulating the pH to induce HNO2 dissociation.
1. Introduction Nitrite is a highly reactive ion1 which can pose serious threats to human health, being the main cause of the so-called blue-baby syndrome.2 For this reason, it is important to avoid accumulation of this ion (and its precursor, nitrate) in wastewater and regulate its use as a food additive in cured meat. Among the various species taking part in the nitrogen biogeochemical cycle, nitrite occupies a central position in two reaction pathways, known as nitrification (from NH3 to NO3-) and, most important, denitrification (from NO3- to N2),2 which allows microorganisms to use nitrate as an electron acceptor instead of oxygen (anaerobic respiration). The nitrite-reducing enzymes can selectively form a single product, whereas;as a comparison;various products are to be expected for most metal electrode systems.1 Many electrochemical studies1 have addressed the properties of the heme-containing globins (myoglobin, hemoglobin), which are known to form iron-nitrosyl complexes upon exposure to NO2under physiologic conditions.3 These globins share a similar behavior, as shown in a comparative work by Mimica et al.,4 who immobilized these proteins in a surfactant (didecyldimethylammonium bromide, DDAB) film cast on glassy carbon and detected a certain activity toward NO2- reduction. Evidence was found that an Fe-nitrosyl adduct is found upon exposure of these proteins to a neutral nitrite solution and that the first electron transfer to the adduct is rate-determining. Literature on these catalysts shows that two main topics have been investigated so far: the influence of the pH of the medium on the reduction activity and on the product selectivity, as will be briefly discussed below. As general background, the pH is a key parameter determining the reactivity of N(III) species as it influences the HNO2/NO2*Corresponding author. E-mail:
[email protected]. (1) Rosca, V.; Duca, M.; de Groot, M. T.; Koper, M. T. M. Chem. Rev. 2009, 109, 2209–2244. (2) Averill, B. A. Chem. Rev. 1996, 96, 2951–2964. (3) Gladwin, M. T.; Kim-Shapiro, D. B. Blood 2008, 112, 2636–2647. (4) Mimica, D.; Zagal, J. H.; Bedioui, F. J. Electroanal. Chem. 2001, 497, 106– 113.
12418 DOI: 10.1021/la101172f
ratio and the occurrence of homogeneous-phase decomposition reactions1,5;active for pH < 5 - an important one being: 2HNO2 f NO þ NO2 þ H2 O
ð1Þ
The influence of such homogeneous-phase chemistry on HNO2/ NO2- electroreduction was recently elucidated on Pt electrodes.6 NO is a highly reactive molecule1 which readily adsorbs at Pt;6,7 its presence in the aqueous phase during nitrite reduction has been positively related to the formation of N2O at a Pt electrode. On the other hand, an increase in pH suppresses the decomposition (1) and the nitrite ion will predominate in solution. This correlates with a remarkable change in the voltammetry and lower reduction currents.6 With respect to heme-modified electrodes, the role of pH was mentioned in an early study by Younathan et al. concerning nitrite reduction at electropolymerized films of iron protoporphyrin(IX) dimethyl ester on glassy carbon (GC) in mildly acidic media (pH = 3).8 The importance of the decomposition of HNO21 in acidic media as the NO-generating reaction was already pointed out in this paper, the dissolved amount of NO being small but sufficient to saturate the heme centers. Among other studies involving heme-containing enzymes, a marked pH effect was reported for nitrite electroreduction with DDAB-embedded myoglobin (DDAB-Mb on PG) by Lin et al.:9 in the pH range 5-9, the reduction current measured was maximum at the acidic end of the range and close to zero at pH 9. A comparable “catalytic” effect of a pH decrease was also reported for nitrite reduction at a different heme-containing protein immobilized in DDAB (thermophilic (5) Park, J. Y.; Lee, Y. N. J. Phys. Chem. 1988, 92, 6294–6302. (6) Duca, M.; Kavvadia, V.; Rodriguez, P.; Lai, S. C. S.; Hoogenboom, T.; Koper, M. T. M. J. Electroanal. Chem., in press. (7) de Vooys, A. C. A.; Koper, M. T. M.; van Santen, R. A.; van Veen, J. A. R. Electrochim. Acta 2001, 46, 923–930. (8) Younathan, J. N.; Wood, K. S.; Meyer, T. J. Inorg. Chem. 1992, 31, 3280– 3285. (9) Lin, R.; Bayachou, M.; Greaves, J.; Farmer, P. J. J. Am. Chem. Soc. 1997, 119, 12689–12690.
Published on Web 04/23/2010
Langmuir 2010, 26(14), 12418–12424
Duca et al.
cytochrome P450 CYP119).10 It thence appears clear that pH exerts a significant influence regardless of the type of enzyme or (bio)catalyst used. The selectivity of the heme-catalyzed nitrite reduction is the second major research topic that remains to be elucidated. In this respect, it must be borne in mind that two different types of analytical techniques for product detection can be used, which we will define as “online and offline”, typically performed during voltammetry or during long-term bulk electrolysis, respectively. Younathan et al.8 carried out long-term electrolysis experiments with electropolymerized films of iron protoporphyrin(IX) dimethyl ester on GC and showed that the reaction mixtures after several hours contain different proportions of most of the possible products, depending on the electrode potential. The ratio N2O/NH3 was almost 1 for higher potentials, but NH3 became the dominant product at lower potentials, accounting for the 63% of the total products, with a high current efficiency. NH2OH and N2 were also consistently detected, N2 being a relevant product (ca. 30% independently of E applied). The formation of N2 was ascribed to the coupling of nitrido [FeV(N)(PP)]0 intermediates. On the other hand, Lin et al.9 determined with offline techniques that the reaction solution after long-term nitrite reduction at DDAB-Mb on PG at pH 5.5 contains total amounts of NH4þ and NH3OHþ never exceeding 60%. The authors also detected (by means of online gas chromatography) N2O and N2 formation with qualitatively different rates; gas evolution is retarded (the onset was recorded 5 min after the beginning of the potentiostatic experiment) and N2O formation also takes place at more positive potentials. DDAB-Mb evolves N2O during NO reduction11 as well, and the mechanism proposed involved the coupling of a Fe-bound nitrosyl with an NO molecule, leading to a N2O2-- heme complex and ultimately to N2O. As a final remark, it must be stressed that the pH effects discussed above are expected to have a remarkable impact on selectivity, since the pH affects the presence of NO in solution, a favored substrate of heme proteins. The aim of this paper is to investigate the reactivity and selectivity of iron protoporphyirin IX (hemin) directly adsorbed at a pyrolytic graphite (PG) electrode toward HNO2/NO2- reduction during voltammetry with online analysis techniques, thereby bypassing possible long-term reactivity effects. This simple electrode, which avoids the complications due to the presence of a surfactant film (i.e., diffusion limitations of the reacting species), has been applied for the study of NO reduction in neutral media12 and at other pH values,13 although the hemin-PG electrode may be less active than the hemin-DDAB-PG electrode.14 The occurrence of direct NO take-up from solution8 or heme-mediated generation of this species from HNO2 will be addressed in the present paper, along with a detailed analysis of selectivity of the reaction. Since gaseous molecules play pivotal roles as intermediate (NO) or as possible products (N2O, N2), we will employ online electrochemical mass spectrometry (OLEMS), a technique closely related to the well-known differential electrochemical mass spectrometry15 and successfully applied in our previous papers on
Article
the electrocatalytic properties of hemin-PG electrodes for NO reduction.12-14,16 Rotating ring-disk electrodes will be employed to study the formation of NH2OH.12
2. Experimental Section Materials. Hemin (Fluka, 98%) was used as received and not purified further. All other chemicals were p.a. ACS grade (Merck). Phosphate buffer solutions (ionic strength 0.1 M) were used as electrolytes. Pyrolytic Graphite (various suppliers) was machined into cylindrical stationary working electrodes (typical diameters 4/5 mm). Home-made disk electrodes for rotating disk experiments were machined from the same material and shrouded in PTFE holders to obtain a total diameter of 5 mm (disk electrode þ holder). This size was suitable for use in combination with commercial Ring-Disk PTFE tips (Pine Instruments) equipped with a Pt ring (vide infra). Working solutions were prepared with Millipore Milli-Q water (resistivity 18.2 MΩ cm). Electrochemical Instrumentation and Techniques. All glassware was carefully rinsed with Milli-Q water prior to the experiments. Electrolyte solutions were deoxygenated by bubbling Ar (purity grade 6.0, Hoekloos) for at least 15 min. Argon blanketing was used to protect the solutions during the experiments. All experiments have been performed at room temperature (ca. 23 °C). An Autolab PGSTAT20 (bi)potentiostat was used throughout this work. Electrochemical experiments were carried out in homemade three-electrode cells: the working electrode was either a stationary hemin-modified PG or a hemin-modified PG-disk with Pt-ring electrode; the counter electrode was a Pt flag, which was flame-annealed and quenched in air before use. A reversible hydrogen electrode (RHE) was used as a reference electrode for all experiments, to which all potentials in this paper are referred. Rotating ring-disk (RRDE) experiments were performed with a Pine Instrument motor generator (MSR rotator). Adsorbate transfer experiments were performed with a stationary hemin-modified PG electrode using a standard procedure.7 The hemin-modified PG (heme-PG) electrode was immersed in a nitrite-containing solution (0.8 mM) buffered at a chosen pH, and under potential control (0.5 V vs RHE) for 120 s, unless otherwise specified in the Results section. The electrode was then thoroughly rinsed and transferred to another cell containing clean buffer, where the electrochemical experiments on the adsorbate were performed. Typical adsorbates (nitrosyl adducts on heme) are robust and do not decompose in aerobic buffer solutions for at least 10 min;12 in our experiments, the electrode was exposed to air for less than 1 min. Preparation of Hemin-Modified PG Electrodes. Details of the immobilization method employed have been published previously.12 As the only difference, in the present work a RRDE tip with a removable disk was used; therefore, the PG disk and the Pt ring were pretreated separately. The former was abraded with SiC sandpaper (500- and 1000-grit) and sonicated for 5 min in Milli-Q water prior to use, the latter was polished with 0.05 μm Al2O3, rinsed and sonicated for 5 min twice in Milli-Q water. The ringdisk tip was then reassembled and immersed in the hemin-containing solution. Finally, the Pt ring was cleaned by repeatedly cycling between the oxygen and the hydrogen evolution region in the working electrolyte (in a separate cell). This procedure does not harm the hemin adsorbed at PG.12 Online Electrochemical Mass Spectrometry (OLEMS).
(10) Immoos, C. E.; Chou, J.; Bayachou, M.; Blair, E.; Greaves, J.; Farmer, P. J. J. Am. Chem. Soc. 2004, 126, 4934–4942. (11) Bayachou, M.; Lin, R.; Cho, W.; Farmer, P. J. J. Am. Chem. Soc. 1998, 120, 9888–9893. (12) de Groot, M. T.; Merkx, M.; Wonders, A. H.; Koper, M. T. M. J. Am. Chem. Soc. 2005, 127, 7579–7586. (13) de Groot, M. T.; Merkx, M.; Koper, M. T. M. C. R. Chim. 2007, 10, 414–420. (14) de Groot, M. T.; Merkx, M.; Koper, M. T. M. J. Am. Chem. Soc. 2005, 127, 16224–16232. (15) Willsau, J.; Heitbaum, J. J. Electroanal. Chem. 1985, 194, 27–35.
Langmuir 2010, 26(14), 12418–12424
Detailed description of the setup can be found in previous publications.6,16 Briefly, the hemin-saturated PG surface was exposed to the electrolyte in a hanging-meniscus configuration and a PTFE tip, connected to the mass spectrometer, was positioned at approximately 10 μm from the PG electrode. The solution was not stirred during the experiments, while a flow of blanketing Ar was maintained to protect the solution from oxygen. A scan rate of (16) Wonders, A. H.; Housmans, T. H. M.; Rosca, V.; Koper, M. T. M. J. Appl. Electrochem. 2006, 36, 1215–1221.
DOI: 10.1021/la101172f
12419
Article
Figure 1. Adduct stripping profile for a hemin-PG electrode in a
clean pH 3 buffer, v=500 mV 3 s-1. The adduct was generated in a pH 3 buffer þ0.8 mM NaNO2 by keeping E = 0.5 V for 120 s.
1 mV/s was chosen for all OLEMS experiments. Currently, our OLEMS setup does not allow a quantitative analysis (and therefore, a direct comparison of results obtained in different experiments) of the signals. The absolute magnitude of the signals depends on various factors, such as the porosity of the PTFE sheet, the tipelectrode distance and the pressure in the mass spectrometer, which might vary (slightly) from experiment to experiment. However, the relative magnitude of ion currents measured in different experiments is highly reproducible. The data obtained were analyzed with commercial scientific software: further details will be reported below when the results are presented.
3. Results Adsorbate Formation at þ0.5 V. The state of a heminmodified PG electrode under stationary conditions in 0.8 mM NaNO2 solutions was first investigated in the pH range 3-7, with so-called transfer experiments, in order to ascertain or to exclude the formation of an adduct/adsorbate at a certain potential, and to obtain information on the nature of this adsorbed species. Figure 1 shows the first two cycles of an adsorbate stripping experiment at pH 3: in the first sweep, there is no trace of the well-known signal of the reversible couple heme-Fe(II)/hemeFe(III), and only a sharp peak, centered at ca. -0.21 V, appears during the scan. In the return sweep, however, the oxidation peak at 0.18 V corresponding to Fe(II) f Fe(III) reappears, and in the second scan, the voltammogram superimposes with the profile of a bare hemin-modified PG electrode. The same behavior has been reported in an experiment involving a heme-modified PG-electrode pretreated in a saturated NO solution at pH 7.12 In that case the peak was positively identified with the stripping of NO, which formed a heme-NO adduct. Also in that experiment, a single negative-going sweep was enough in order to remove the adsorbate. From Figure 1, the charge exchanged during adsorbate stripping and the charge exchanged in the process Fe(II) f Fe(III) can be calculated, giving a ratio of ca. 3: 1, in agreement with the previous study,12 and supporting the identification of the peak as the NO-adduct being irreversibly reduced to NH2OH. The formation of the adduct was found to be dependent on the pH in both a qualitative and quantitative way. As illustrated in Figure 2, the peak potential for the adsorbate stripping shifts to more negative potentials (on the SHE scale) with increasing pH with a slope of 42 mV/pH. This value, which remarkably differs from more common pH-slope values (59 mV/pH and multiples), was obtained reproducibly. Along with this shift, the charge corresponding to the adduct stripping decreases with increasing pH. Simultaneously, the reversible peak due to the heme-Fe(II)/ heme-Fe(III) become already visible during the first negative-going 12420 DOI: 10.1021/la101172f
Duca et al.
Figure 2. Influence of the pH on parameters of NO bonding to hemin: relative NO-on-heme occupancy (Θest) and stripping peak potential (Epeak).
scan. Figure 2 shows this effect expressed in terms of normalized charge (Θest),17 which is supposed to be a measure for the relative number of heme centers that bind NO. Additionally, it was found that the adduct stripping charge could be increased by a longer contact time with the nitrite-containing electrolyte, before the subsequent transfer and stripping. For a pH 5 nitrite solution, the fraction of heme centers binding NO on the heme-PG electrode increased to 0.18 after a 13-min immersion, as compared to 0.02 obtained during a standard 2-min contact with the nitrite-containing electrolyte. These experiments clearly show that NO generation, and its subsequent binding to hemin at a relatively high potential, takes place in a submillimolar nitrite solution, the pH being a key parameter in determining the amount of adsorbed NO (for a constant adsorption time). The almost zero NO coverage at pH 7 suggests that, at neutral pH, heme does not significantly catalyze the nitrite decomposition to NO. Continuous Reduction at Stationary and Rotating Electrodes. The plot of the current measured during the continuous reduction of nitrite at a stationary hemin-modified PG electrode (in a pH 3 buffer) is reported in Figure 3 along with the NO stripping current in the same electrolyte for comparison. It can be observed that the onset of the continuous reduction occurs at a potential essentially identical to that of NO stripping, evidencing that the NO adduct is the key intermediate of the continuous reduction of nitrite. The voltammetric features obtained at a rotating hemin-PG electrode are shown in Figure 4 (the negative-going scan only). A main wave can be recognized, centered at -0.2 V, followed by a less steep region and a second wave starting at ca. -0.35 V. If plotted on a logarithmic scale (see inset in Figure 4), the main wave displays a slope of 59 mV 3 dec-1, suggesting that an EC mechanism (rate-determining chemical step following the first electron transfer) is operative. This is another feature in common with the continuous reduction of NO.12 The effect of rotation rate is also shown in Figure 4: below -0.25 V the current responds weakly to the rotation rate, while the main wave is unaffected by rotation. The influence of an increase in nitrite concentration is reported in the Supporting Information (Figure S1), suggesting a reaction order slightly larger than 0.5. Two pH values, pH 1.7 and pH 7, have been studied for comparison to pH 3. The reduction of 0.8 mM NaNO2 at these pH values is shown in Figure 5. The main observation is that the reduction current in a neutral medium is almost zero, except for a slight increase at the most negative potentials. The inset of Figure 5, (17) The normalized charge (i.e., the coverage) is calculated from the stripping peak charge (Qstrip) and the blank FeII/FeIII charge (Qblank) with the formula (Qstrip/3)/Qblank, which assumes that the adsorbate forms only hydroxylamine upon stripping.
Langmuir 2010, 26(14), 12418–12424
Duca et al.
Article
Figure 3. Comparison of voltammetric profiles of heme-bound NO (solid line) and nitrite continuous reduction (0.8 mM NaNO2, dashed line) at a stationary hemin-PG electrode in a pH 3 buffer, v = 500 mV 3 s-1.
Figure 4. Linear sweep voltammetric (LSV) profiles for the reduction of 0.8 mM NaNO2 at pH 3 at a rotating hemin-PG electrode, v = 10 mV 3 s-1, ω = 900-2500 rpm. Inset: logarithmic plot of the scan at 900 rpm.
Figure 6. NH2OH detection during reduction of 0.8 mM NaNO2 at a RRDE (hemin-PG disk, Pt ring). Electrolyte: pH 3 buffer. Main figure: profile of the ring current with the disk at open circuit potential (o.c.p., bold line) and with Edisk =-0.3 V (thin line). Inset: highlight on the potential range 0.8 < Ering < 1.5 V showing the “loop” signal typical of NH3OHþ formation in a nitrite solution.
especially it gives rise to a lower current at E
