Superstable Advanced Hydrogen Peroxide Transducer Based on

Feb 24, 2011 - The only disadvantage of the Prussian Blue based sensing layers in respect to long-term continuous monitoring is their inherent instabi...
0 downloads 0 Views 889KB Size
ARTICLE pubs.acs.org/ac

Superstable Advanced Hydrogen Peroxide Transducer Based on Transition Metal Hexacyanoferrates Natalya A. Sitnikova, Anastasiya V. Borisova, Maria A. Komkova, and Arkady A. Karyakin* Chemistry and Material Science faculties of M.V. Lomonosov Moscow State University, 119991, Moscow, Russia ABSTRACT: We report on a superstable hydrogen peroxide (H2O2) transducer made by sequential deposition of the iron- and nickel-hexacyanoferrate (NiHCF) layers. Both chemical and mechanical stability of the latter, as well as similarity of its structure to Prussian Blue (PB) provide a substantial stabilization of the most advantageous H2O2 transducer. The electrochemically deposited five bilayers of PBNiHCF exhibit a complete stability under the continuous wall-jet flow of 1 mM of H2O2 during more than 2 h, maintaining current at a level of 0.2 mA cm-2, whereas common Prussian Blue loses half of its response within the first 20-25 min. Even being deposited in the open circuit regime on screen-printed electrodes, PB-NiHCF bilayers dramatically improve tolerance of the resulting transducer to alkaline solutions and iron ligands. Despite their 2-2.5 times decreased sensitivity (compared to common Prussian Blue), the sequentially deposited bilayers of PB-NiHCF provide a similar dynamic range of the transducer due to the decreased noise level.

H

ydrogen peroxide (H2O2) is becoming one of the most important analytes nowadays. We always referred to it as to the chemical threat agent, whose excessive concentration as a product of industry and atomic power stations dramatically affects the environment,1,2 as well as to disinfecting agent for water pools, food, and beverage packages,3,4 which makes it important to measure its residual concentration. However, the most important role as an analyte H2O2 plays in the greatly expanding area: clinical diagnostic. First, it is the most valuable marker for oxidative stress, recognized as one of the major risk factors in progression of disease-related pathophysiological complications in diabetes, atherosclerosis, renal disease, cancer, aging, and other conditions.5-9 Second, hydrogen peroxide is also a side product of oxidases, the enzymes used as terminal ones in the majority of analytical kits. As shown already almost 40 years ago, the detection of H2O2 provides the highest sensitivity of the corresponding biosensors.10,11 We already reported on Prussian Blue (PB) as the most advantageous hydrogen peroxide transducer.12-14 Comparing with the most widely used platinum, Prussian Blue modified electrodes are (i) 3 orders of magnitude more active in H2O2 reduction and oxidation in neutral media and (ii) 3 orders of magnitude more selective for hydrogen peroxide reduction in the presence of oxygen.15 The attractive performance characteristics of the electrochemically deposited Prussian Blue allowed to denote it as artificial enzyme peroxidase.13,14 Moreover, nanostructuring of Prussian Blue was resulted in elaboration of electrochemical sensor with record performance characteristics, namely linear calibration range prolonged over seven orders of magnitude of analyte (H2O2) concentration.16,17 The only disadvantage of the Prussian Blue based sensing layers in respect to long-term continuous monitoring is their inherent instability, particularly in neutral solutions. First, the Prussian Blue layer is mechanically unstable. Second, the longr 2011 American Chemical Society

term operational stability is affected by the product of hydrogen peroxide reduction at Prussian Blue modified electrodes, known to be hydroxyl ion (OH-).18 Moreover, different iron complexing agents (for instance, EDTA used as blood anticoagulant) are known to solubilize ferric hexacyanoferrate. Strategies used to stabilize the advanced hydrogen peroxide transducer were (i) covering with organic polymers, including dip-coating with Nafion19 and electropolymerization of nonconductive polymer on the top surface of Prussian Blue,20 and (ii) involvement in sol-gel21-23 or conductive polymer matrixes.24,25 However, neither absolute long-term stability, nor prolonged shelf life can be achieved due to possibility to degradation of organic polymers and their structural changes upon dryingswelling. In contrast to ferric hexacyanoferrate, other transition metal ferrocyanides possess relatively low catalytic activity in H2O2 reduction-oxidation.15 However, even iron triad-mates (see periodic table)—cobalt and nickel—form both chemically and mechanically stable hexacyanoferrates. It was, hence, highly attractive to use these closely related compounds to stabilize Prussian Blue. We report here on a superstable hydrogen peroxide transducer based on mixed iron-nickel hesacyanoferrates. The bilayers of PB-NiHCF exhibit a complete stability under the continuous wall-jet flow of 1 mM of H2O2. This fact, taking into account the dramatically improved tolerance to alkaline solutions and iron ligands, allows to denote the resulting transducer as a superstable one.

Received: December 24, 2010 Accepted: February 4, 2011 Published: February 24, 2011 2359

dx.doi.org/10.1021/ac1033352 | Anal. Chem. 2011, 83, 2359–2363

Analytical Chemistry

ARTICLE

’ EXPERIMENTAL SECTION Materials. Experiments were carried out with Milli-Q water from a Millipore Milli-Q system. All inorganic salts and hydrogen peroxide (30% solution) were obtained at the highest purity from Reachim (Moscow, Russia) and used as received. Planar electrodes were made by screen printing (Screen Printer SCF-550, Technical Industrial Co. Ltd., Hong Kong) on polyester films. For both working and auxiliary electrodes, the carbon inks C10903P14 and C2030519D4 (Gwent, UK) providing reversible voltammograms of hexacyanoferrate in neutral media were chosen. The working electrode diameter was 2 mm. Glassy carbon disk electrodes (2 mm in diameter) used as working electrodes were made by pressing glassy carbon (Su2500) rods in Teflon. Prior to use, the glassy carbon electrodes were mechanically polished with alumina powder (Al2O3, 0.05 μm) until a mirror finish was observed. Instrumentation. Electrochemical experiments were made in a three-compartment electrochemical cell containing a platinum net auxiliary electrode and Ag|AgCl reference electrode in 1 M KCl. Cell construction allowed deaeration of the working electrode space. The flow injection (FIA) system consisted of a Cole Parmer (Vernon Hills, IL) peristaltic pump (7519-10), homemade flowthrough wall-jet cell with 0.5-mm nozzle positioned at 1-2 mm distance from the surface of disk electrode, (Ag|AgCl|1 M KCl) reference, homemade injector, and PalmSens electrochemical interface interfaced to an IBM PC. Flow rates used were in the range 0.5-1 mL min-1. In FIA experiments, the peak current density values were taken for data treatment, sample volume was 50 μL, and working electrode potential was 0.00-0.05 V, allowing hydrogen peroxide reduction on Prussian Blue-modified electrodes. The carrier solution in FIA experiments was 0.05 M phosphate buffer pH 6.0 containing 0.1 M KCl as the supporting electrolyte. The concentration of hydrogen peroxide in stock solutions was controlled by optical density at 230 nm with an LKBUltraspec UII spectrophotometer (Broma, Sweden). Electrodeposition of Prussian Blue. This process was made in cyclic voltammetric conditions with switching potentials of 0.4 V (cathodic) and of 0.7-0.8 V (anodic) at a sweep rate of 40 mV s-1 or by applying a constant potential of 0.4 V as described elsewhere.16 The growing solution contained 0.54 mM K3[Fe(CN)6] and 0.5-4 mM FeCl3. A solution of 0.1 M HCl and 0.1 M KCl was used as supporting electrolyte. After deposition, Prussian Blue films were electrochemically activated in the same supporting electrolyte by cycling in the range from -0.05 to 0.35 V at a rate of 40 mV s-1 until a stable voltammogram was obtained. Then, the electrodes were heated at 100 °C for 1 h. Interfacial Synthesis of Prussian Blue. This was carried out by dipping a droplet of the growing solution containing 2-4 mM K3[Fe(CN)6] and 2-4 mM FeCl3 in a mixture of 0.1 M HCl and 0.1 M KCl onto the target support. Deposition was initiated adding a reductant (hydrogen peroxide) to the droplet to a final concentration of 50-200 mM. After 2-40 min of deposition, the electrode was washed with water and heated for 1 h at 100 °C. Deposition of Nickel Hexacyanoferrate. This process was made using 0.5 M KCl and 0.1 M HCl as a background electrolyte. Concentration of precursors (Ni2þ, [Fe(CN)6]3-) was varied in the range 0.5-2 mM. The highest electrocatalytic activity in H2O2 reduction was recorded for NiHCF deposited

Figure 1. Cyclic voltammograms of NiHCF films deposited electrochemically (solid) and in open circuit conditions (dash) and of Prussian Blue (dotted); 0.1 M KCl and 0.1 M HCl, sweep rate 20 mV s-1; hydrodynamic voltammogram of H2O2 reduction at NiHCF modified electrode, wall-jet cell, flow rate 0.8 mL min-1, 0.05 M phosphate pH 7.0 with 0.1 M KCl.

from the initial solution containing the two-fold excess of Ni2þ over [Fe(CN)6]3-.

’ RESULTS AND DISCUSSION As mentioned, except for Prussian Blue, other transition metal hexacyanoferrates displayed only a minor electrocatalytic activity in hydrogen peroxide reduction.15 However, taking into account both high stability and structural similarity to Prussian Blue, we tested cobalt and nickel hexacyanoferrates as stabilizers of the advanced hydrogen peroxide transducer. The best results were obtained for nickel hexacyanoferrate (NiHCF). Deposition of Nickel Hexacyanoferrate. In contrast to Prussian Blue, which is deposited from stable solutions of ferrous (Fe3þ) and ferricyanide (Fe(CN)63-) ions by electroreduction of Fe3þ to Fe2þ, the latter forms an insoluble precipitate with Fe(CN)63-, nickel-ion precipitates with both ferro- and ferricyanide. Electrodeposition of nickel hexacyanoferrate from noncolloid solution became possible due to stabilization of the mixture of Ni2þ and Fe(CN)63- ions in aqueous media with the excessive amount of supporting electrolyte (KCl).26 Cycling the electrode potential between 0 and 0.7-0.8 V in such solution results in deposition of nickel hexacyanoferrate (NiHCF) film. Typical cyclic voltammogram (CV) of the electrodeposited NiHCF modified electrode in supporting electrolyte solution is shown in Figure 1 (solid line). The observed two sets of peaks are attributed to different forms of nickel hexacyanoferrate: the positive one to the so-called potassium rich (K2Ni[Fe(CN)6]) and the negative one to potassium deficient (KNi1.5[Fe(CN)6]).27 In addition to electrochemical synthesis, an open-circuit deposition (highly important for mass production) of nickel hexacyanoferrate is also possible. Figure 1 also displays CV of carbon electrode kept in similar growing solution without applying of electrical force for 30 min (dash line). As seen, a quite similar electroactivity is observed. Moreover, varying content of the growing solution, it is possible to deposit as individual (nickelrich or -deficient) hexacyanoferrate, as the mixed polycrystal with the two sets of peaks of similar intensity (data not shown). Considering various reports on electrocatalytic activity of nickel hexacyanoferrate in hydrogen peroxide (H2O2) reduction, 2360

dx.doi.org/10.1021/ac1033352 |Anal. Chem. 2011, 83, 2359–2363

Analytical Chemistry

Figure 2. Optimization parameter (I0/kin) for simultaneous deposition of iron and nickel hexacyanoferrates as a function of both molar fraction of ferric ion in the growing solution (O) and the sweep rate (9).

we recorded the corresponding hydrodynamic voltammogram. As expected, the current of H2O2 reduction was approximately 1000 times lower compared to Prussian Blue. Taking into account, that equilibrium potential of H2O2/H2O couple in aqueous solutions is above 1.5 V (NHE),28 in the range of NiHCF redox activity hydrogen peroxide reduction has to be at the current plateau region. As seen in Figure 1, the current of H2O2 reduction at nickel hexacyanoferrate modified electrodes around redox potentials of the latter is negligible. A significant current of H2O2 reduction is observed at more negative potentials: the half-wave potential of the hydrodynamic votammogram is close to the redox potential of the Prussian Blue-modified electrode (Figure 1, dotted curve). Hence, the electrocatalytic activity of NiHCF films in H2O2 reduction is due to the presence of Prussian Blue structures as defects in nickel hexacyanoferrate polycrystal. Figure 1 also indicates that the sets of peaks of the ferric and nickel hexacyanoferrates are perfectly separated, which makes it possible to control the structural composition of the resulting mixed films. Electrosynthesis of Mixed Nickel-Iron Hexacyanoferrate. Considering Prussian Blue-related electrocatalysis by nickel hexacyanoferrates in H2O2 reduction (above), it was highly attractive to synthesize the mixed nickel-iron hexacyanoferrate film for a highly active and stable hydrogen peroxide transducer. Obviously, the simultaneous deposition of both hexacyanoferrates seemed to be preferable. Simultaneous deposition of iron and nickel hexacyanoferrates is possible from the growing solution containing as nickel (Ni2þ) and ferric (Fe3þ) ions, as ferricyanide (Fe(CN)63-). Potential cycling in the range between 0.7 and 0.8 V and þ0.4 V is favorable for synthesis of both hexacyanoferrates. A typical cyclic voltammogram of the simultaneously deposited mixed nickel-iron hexacyanoferrate displays redox activities peculiar to both hexacyanoferrates. We’ve found that the ratio of the iron-to-nickel hexacyanoferrate (estimated from cyclic voltammograms) is dependent not only on the ion concentration ratio ([Fe3þ] to [Ni2þ]) but also on the sweep rate. Higher sweep rates are preferable for synthesis of nickel hexacyanoferrate (data not shown). We’ve chosen operational stability in wall-jet cell under continuous flow of 1 mM of hydrogen peroxide as a stability

ARTICLE

criterion of hydrogen peroxide transducer. The corresponding inactivation constant (kin) was calculated as a slope in semilogarithmic plots. Considering the prolonged calibration range of the Prussian Blue modified electrodes,13,14,17 the initial response (I0) to continuous flow of 1 mM of H2O2 can be taken as a criterion of the transducer sensitivity. Since for an optimal transducer both operational stability and sensitivity are important, it is convenient to choose the ratio of the initial response to inactivation constant (I0/kin) as the parameter for optimization. Figure 2 presents the dependence of the optimization parameter (I0/kin) as a function of both molar fraction of ferric ion in the growing solution and the sweep rate. As seen, in both cases the sharp extremes occur allowing one to determine optimal conditions for electrosynthesis. The resulting mixed nickel-iron hexacyanoferrate was characterized by the 10 times decreased inactivation constant and 2.5-3 times decreased sensitivity as compared with Prussian Blue. The linear calibration range was, however, similar because of the decreased noise in the case of mixed nickel-iron hexacyanoferrate. Despite the significantly improved stability, the elaborated hydrogen peroxide transducer based on mixed nickel-iron hexacyanoferrate deposited simultaneously did not give satisfactory results in continuous monitoring. This provided an interest to other approaches for synthesis of mixed transition metal hexacyanoferrates, in particular to their layer-by-layer deposition. Layer-by-Layer Deposition of Transition Metal Hexacyanoferrates. An apparent lack of success with simultaneous synthesis of nickel-iron hexacyanoferrate (above) caused an interest to its layer-by-layer deposition. Obviously, a layer of Prussian Blue was covered by a layer of Ni-hexacyanoferrate for stabilization. Electrosynthesis of nickel hexacyanoferrate film over Prussian Blue modified electrodes has led to a significant stabilization of the hydrogen peroxide transducer. Depositing 0.8 nmol cm-2 of NiHCF as a stabilizing layer, we decreased the corresponding inactivation constant (under a continuous flow of 1 mM H2O2) three times. The thicker layer of NiHCF seemed to improve further the transducer stability. However, the increased thickness of the stabilization Ni-hexacyanoferrate layer causes both the decreased sensitivity and higher response time (data not shown). Hence, the increase of NiHCF thickness over 1.5-2 nmol cm-2 is not plausible, and further improvement of the transducer stability can be made only by sequential deposition of several PBNiHCF layers. Indeed, whereas for one bilayer of PB-NiHCF the remaining response after 30 min under continuous wall-jet flow of 1 mM H2O2 was at the level of 50% (compared to 44% for uncovered Prussian Blue), in the case of the two PB-NiHCF bilayers, this value was already at the level of 70%. Three and five PB-NiHCF bilayers were characterized by the remaining response (after 30 min under continuous wall-jet flow of 1 mM H2O2) of 82% and 100%, respectively. Further increase of the number of PB-NiHCF bilayers seems to be not plausible due to both decreased reproducibility of the resulting transducer and the apparent irregularity of the transition metal hexacyanoferrates noticed from their cyclic voltammograms (data not shown). Accordingly, for electrochemical deposition the five bilayers of PB-NiHCF were found to be optimal in terms of both reproducibility and operational stability. In the case of an open circuit deposition, no valuable improvement of the transducer stability in case of the three bilayers compared with the two bilayers was found. On the contrary, the substantial background increase even 2361

dx.doi.org/10.1021/ac1033352 |Anal. Chem. 2011, 83, 2359–2363

Analytical Chemistry

Figure 3. Optimization parameter (I0/kin) for layer-by-layer deposition of iron and nickel hexacyanoferrates (O) two PB-NiHCF bilayers, open circuit regime, (9) five bilayers, electrochemical deposition.

Figure 4. Operational stability in wall-jet cell under continuous flow of 1 mM H2O2 of the electrodeposited (1) Prussian Blue and (2) five PBNiHCF bilayers: 0.05 M phosphate pH 7.0 with 0.1 M KCl.

in case of the three bilayers deteriorated analytical performances of the resulting transducer. Hence, for open circuit deposition (as a tool for mass production), further optimization was made for two bilayers of PB-NiHCF. The amount of the deposited Prussian Blue plays a crucial role in analytical properties of the hydrogen peroxide transducer. Obviously, increasing the amount of PB, one improves the transducer sensitivity. However, as was found, a higher amount of Prussian Blue reduces the operational stability of the transducer. Figure 3 presents the I0/kin ratio (as an optimization parameter, above) as a function of the Prussian Blue amount deposited via each layer. At present, for each layer, we deposited similar amounts of each transition metal hexacyanoferrate. As seen in Figure 3, for both transducers synthesized electrochemically and in open circuit regime there are the recognizable extremes of the optimization parameter (I0/kin). Both extremes are achieved at approximately at half amount of Prussian Blue used as a single layer (5 nmol cm-2 for screen printed and 7-8 nmol cm-2 for glassy carbon electrodes). Transducer Stability. Layer-by-layer deposition of Prussian Blue and nickel hexacyanoferrate resulted in a highly stable hydrogen peroxide transducer. Figure 4 presents operational stability of the electrochemically deposited five-bilayered

ARTICLE

Figure 5. Calibration graph for two PB-NiHCF bilayers synthesized in the open circuit regime on a screen printed electrode: wall-jet cell, flow rate 0.8 mL min-1, 0.05 M phosphate pH 7.0 with 0.1 M KCl.

PB-NiHCF in hard conditions of continuous flow (0.8 mL min-1) of 1 mM H2O2 in a wall-jet cell. It is seen that whereas Prussian Blue alone loses half of its response within the first 2025 min, the PB-NiHCF based transducer is completely stable within 2 h. Despite the fact that the response of the latter is less compared to Prussian Blue, the lower noise (Figure 4) provides a similar detection range of the transducer (below). An apparent complete stability for hours in such hard conditions allows the distinction of the transducer to be a superstable one. As mentioned, for mass production technology screen printed electrodes were modified with PB-NiHCF bilayers in the open circuit regime. The sensors with the two bilayers of PB-NiHCF in similar hard conditions (continuous flow of 1 mM H2O2 in a wall-jet cell) were completely stable within the first 30-55 min. After 1 h of operation more, than 90% of the initial response remained. Moreover, the response value of the stabilized transducer is only