Comparative electrochemistry and electrocatalytic activities of cobalt, iron and manganese phthalocyanine complexes axially co-ordinated to mercaptopyridine self-assembled monolayer at gold electrodes

Electrochimica Acta 51 (2006) 2669–2677

Comparative electrochemistry and electrocatalytic activities of cobalt, iron and manganese phthalocyanine complexes axially co-ordinated to mercaptopyridine self-assembled monolayer at gold electrodes Kenneth I. Ozoemena ∗,1 , Tebello Nyokong 1 Chemistry Department, Rhodes University, Grahamstown 6140, South Africa Received 22 March 2005; received in revised form 2 June 2005; accepted 7 August 2005 Available online 8 September 2005

Abstract Comparative surface electrochemistry and electrocatalytic properties of solid ultrathin monolayer films of metallophthalocyanine (MPc) complexes of cobalt (CoPc), iron (FePc) and manganese (MnPc) self-immobilised, via axial ligation reaction, onto preformed 4mercaptopyridine self-assembled monolayers (SAMs) on gold electrodes have been described. Surface electrochemical parameters of the modified electrodes showed that these MPc-SAMs are densely packed with flat orientations. The electrochemical, electrocatalytic and stability properties of these MPc complexes follow this order: FePc > MnPc > CoPc. This finding is remarkable as it suggests that the success of using this method of self-assembling of MPc onto gold electrode is largely dependent on the bond distance between the pyridine linker and the central metal of the MPc; the shorter the distance, the faster the co-ordination and the better the electrocatalytic properties towards l-cysteine and thiocyanate. Thus, the superiority of FePc-based SAM over those of the MnPc and CoPc, has been proposed to be the result of the more favorable axial co-ordination properties of FePc with pyridine (i.e. shorter Fe–N(pyridine) bond length. © 2005 Elsevier Ltd. All rights reserved. Keywords: Metallophthalocyanine-mercaptopyridine self-assembled monolayers; Electrochemistry; Electrocatalysis; l-Cysteine; Thiocyanate

1. Introduction Metallophthalocyanine (MPc) complexes have attracted a lot of research interest because of the their physico-chemical properties that could be harnessed for the fabrication of technologically important devices, such as electrochemical sensors, molecular electronics and photovoltaic devices [1,2]. These possible applications require the use of MPc complexes in their ultrathin solid films. The established strategies for obtaining ultrathin monolayer films of MPc complexes and related macrocycles on electrode surfaces include the Langmuir-Blodgett [3–7] spin-coating [7] and self-assembly [7–11]. The advantageous properties of the SAM technique, notably high order and stability [12–14], have led to the ∗ 1

Corresponding author. Tel.: +27 46 6038801; fax: +27 46 6225109. E-mail address: [email protected] (K.I. Ozoemena). ISE member.

0013-4686/$ – see front matter © 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.electacta.2005.08.007

growing interest in the fabrication of self-assembled MPc monolayers (MPc-SAMs) on gold. Transition metal (notably Fe, Co and Mn) phthalocyanine complexes are excellent electrocatalyts that are useful for the fabrication of MPc-SAM based electrochemical sensors. Two major strategies for the fabrication MPc-SAMs are known. The first strategy involves the use of metallophthalocyanine complexes that are peripherally or non-peripherally substituted with sulfur-containing molecules. These thiolderivatised MPc complexes can bond to the gold substrates via their thiol arm(s) [7–11]. The second and more recent strategy is by axial ligation of the MPc to a selfassembled monolayer of an N-donor molecule (such as 4mercaptopyridine) onto the gold surface [15,16]. A significant advantage of this second strategy is that it avoids the synthesis of thiol-derivatised MPc complexes. Mercatopyridine molecules adsorbed onto gold surfaces as SAM exhibit suitable molecular orientation for axial ligation with MPc

2670

K.I. Ozoemena, T. Nyokong / Electrochimica Acta 51 (2006) 2669–2677

2. Experimental 2.1. Materials and reagents Manganese phthalocyanine (MnPc) was obtained from Eastman. Iron (FePc) and cobalt (CoPc) phthalocyanine complexes and 4-mercaptopyridine (MPyr, 95%) were obtained from Aldrich. The purity of the three MPc complexes was ascertained using UV–vis spectroscopy and infrared spectroscopy. UV–vis spectra were recorded on a Varian 500 UV–vis/NIR spectrophotometer. IR (KBr pellets) was recorded on a Perkin-Elmer Spectrum 2000 FTIR spectrometer. Ultra pure water of resistivity 18.2 M
was obtained from a Milli-Q Water System (Millipore Corp., Bedford, MA, USA). Potassium thiocyanate and l-cysteine were purchased from Riedel-de-Ha¨en AG and Sigma–Aldrich, respectively. Potassium hydrogen phthalate buffer tablets (pH 4.0) were purchased from Merck Pty (SAARCHEM, South Africa). All other reagents were of analytical grade and were used as received from the suppliers without further purification. Fig. 1. (a) Molecular structure of metallophthalocyanine (MPc, where M = Fe, Co and Mn) complexes and (b) schematic representation of an axially ligated MPc complex onto 4-mercaptopyridine self-assembled monolayer.

[15,16] and its related complexes, such as the metalloporphyrins [15,17,18]. Although the facile axial ligation between the transition MPc and pyridine is now being exploited in the fabrication of MPc-SAMs [15,16], the surface electrochemistry and electrocatalytic activities of such MPc-SAMs are still grossly under-explored. As part of our work devoted to the search for the transition MPc-SAMs for the fabrication of electrochemical sensors, we explore here the electrochemistry and electrocatalytic activities of three commercially available transition metal MPc complexes (i.e. FePc, MnPc and CoPc) (Fig. 1a) when immobilised onto gold electrodes pre-modified with 4-mercaptopyridine SAMs (Fig. 1b). This study attempts to establish: (i) if there is any detectable difference in surface electrochemical properties between these three redox-active MPc catalysts when axially ligated with pyridine SAMs on gold surface and (ii) the possible roles such difference(s) can play in the successful application of this type of MPc-SAM as potential electrochemical sensors. The study employs cyclic voltammetry (CV) because it has extensively been used to characterise gold electrodes modified with SAMs of alkanethiols and their derivatives [13,14] including those of MPc [9–11,13,14] and metalloporphyrins complexes [17,18]. Electrochemical techniques are simple, low-cost and provide huge information regarding the integrity of electroactive SAMs adsorbed onto electrode surfaces. For example, it is interesting to mention that a recent study by Zhang et al. [17] showed that CV measurements provided a clearer understanding on the orientation of cobalt porphyrin attached to mercaptopyridine SAM onto gold electrode than surface enhanced Raman spectroscopy (SERS) measurements.

2.2. Apparatus and procedure Electrochemical experiments, cyclic voltammetry and square wave voltammetry (SWV), were performed using Bio-Analytical Systems, 100 B/W Electrochemical Workstation, using a conventional three-electrode system. The working electrode was either bare gold (r = 0.8 mm, BAS) or gold electrode modified with 4-mercaptopyridine SAM or MPc-linked-4-mercaptopyridine-SAM. A Ag|AgCl wire and platinum wire were used as pseudo-reference and counter electrodes, respectively. Electrocatalysis towards the determination of l-cysteine and thiocyanate were performed in freshly prepared pH 4.0 buffer solutions. SWV technique was used for the analysis. Each analysis was performed using freshly prepared pH 4.0 stock solutions (1.0 × 10−3 mol L−1 ) of each analyte, prepared daily or whenever needed. The parameters for the SWV were: step potential 4 mV; square wave amplitude 25 mV at a frequency of 15 Hz. Solutions were deaerated by bubbling nitrogen prior to each electrochemical experiment. All experiments were performed at 25 ± 1 ◦ C. 2.3. Electrode fabrication The fabrications of the MPc-linked-mercaptopyridineSAM on gold electrodes (herein referred to as MPc-MPyrSAM, where M for the MPc = Co, Mn and Fe) were carried out using established procedures [15,16]. Briefly, gold electrode was first polished using aqueous slurries of alumina (<10 ␮m) on a SiC-emery paper (type 2400 grit), and then to a mirror finish on a Buehler felt pad. The electrode was then placed in absolute ethanol and subjected to ultrasonic vibration to remove residual alumina particles that might be trapped at the surface. Finally, the electrode was etched for about 2 min in a hot ‘Piranha’ solution {1:3 (v/v) 30%

K.I. Ozoemena, T. Nyokong / Electrochimica Acta 51 (2006) 2669–2677

2671

H2 O2 and concentrated H2 SO4 } and then rinsed with copious amounts of ultrapure Millipore water. Following this pre-treatment, the electrode was rinsed with absolute ethanol and immediately placed into a nitrogen-saturated absolute ethanol solution of 4-mercaptopyridine (1 × 10−3 mol L−1 ) for 1 h at ambient temperature. The modified electrode was thoroughly rinsed in absolute ethanol solution before reacting with MPc. Ligation of MPc with the pyridine SAMs was performed by immersing the 4-mercaptopyridine modified gold into a tetrahydrofuran (THF) solution of MPc (1 × 10−3 mol L−1 ) for 6 h at ambient temperature. Upon removal from the deposition solution, the electrode was thoroughly rinsed with THF and dried in a nitrogen atmosphere prior to all electrochemical experiments.

3. Results and discussion Before immobilisation of the three commercial MPc (CoPc, FePc and MnPc) complexes used in this work, their purity was quickly confirmed by their UV–vis spectral characteristics. The UV–vis spectra of CoPc (Fig. 2a), FePc (Fig. 2b) and MnPc (Fig. 2c) in dimethylformamide are characterised by well-defined low energy or Q bands (657–705 nm region) and high energy or B bands (300–350 nm region) in agreement with literature [19]. Also, the relatively strong band in the 450 nm region found in the spectra of FePc (425 nm) and MnPc (488 nm) are typical of these complexes, which are associated with the so-called metal-to-ligand charge transfer [19]. IR was also recorded to check the purity, and all the MPc complexes exhibited the characteristic bands, such as the C–C benzene ring skeletal (1600–1670 cm−1 ) and the C–N (1638–1644 cm−1 ) stretching vibrations. 3.1. Surface electrochemistry Fig. 3a compares typical cyclic voltammograms of (i) bare gold electrode, (ii) MPyr-SAM on gold electrode and (iii) CoPc-linked-MPyr-SAM on gold electrode in pH 4.0 buffer solution. It is easily observed from Fig. 3a that the formation

Fig. 3. Cyclic voltammograms of (a): (i) bare gold electrode, (ii) MPyrSAM on gold electrode and (iii) CoPc-linked-MPyr-SAM on gold electrode. Cyclic voltammograms of gold electrodes modified with MPyr-SAM linked with (b) FePc and (c) MnPc. Electrolyte = pH 4.0 buffer. Scan rate = 25 mV s−1 .

Fig. 2. UV–vis spectra of CoPc (a), FePc (b) and MnPc (c).

of mercaptopyridine SAM and a typical MPc-linked-MPyrSAM led to a dramatic reduction of the capacitive charging current of the bare gold electrode. The reduction of the capacitive current is a common phenomenon for SAMs of thiols [13,14,20] and thiol-derivatised metallophthalocyanine [10,11,15,17] and metalloporphyrin [17,18] complexes when

2672

K.I. Ozoemena, T. Nyokong / Electrochimica Acta 51 (2006) 2669–2677 Table 1 Surface cyclic voltammetric characteristics for self-immobilised metallophthalocyanine (MPc, M = Co, Mn and Fe) complexes axially ligated to 4mercaptopyridine self-assembled monolayers onto gold electrodes in pH 4.0 buffer solutionsa MPc-SAM

E1/2 /V (vs. Ag|AgCl)

Ep /V (vs. Ag|AgCl)

Ipa /Ipc

Γ MPc /mol cm−2

CoPc FePc MnPc

0.13 −0.04 0.07

0.20 ∼0.10 0.14

1.4 1 1

1.93 × 10−10 7.22 × 10−11 6.85 × 10−11

a

All the data were estimated under similar experimental conditions and at scan rate of 25 mV s−1 . All values are averages from five measurements performed during 2 weeks (R.S.D. < 3%).

Fig. 4. Cyclic voltammetric responses of FePc-linked-MPyr-SAM on gold electrode during 20 continuous scans in pH 4.0 buffer. Full and broken arrows indicate the directions of scans and voltammograms, respectively. Scan rate = 50 mV s−1 .

immobilised onto gold surfaces. There was no noticeable redox peak at the MPyr-SAM gold electrode within the potential range (between −0.50 and +0.50 V versus Ag|AgCl) investigated. The scan direction used in this work (from negative direction to the positive, i.e. oxidation followed by reduction of central metal of the MPc) resulted in wellbehaved cyclic voltammograms as evident in Fig. 3. Fig. 3b and c are examples of cyclic voltammograms of MPyr-SAMs linked with FePc and MnPc, respectively. Co-ordination of MPc to MPyr has been reported in solution [15]. Each SAM of the three MPc complexes on gold electrode shown in Fig. 3 was obtained after voltammetric conditioning or annealing in pH 4.0 buffer solutions. An essential factor to a successful SAM formation on gold surfaces is the monolayer annealing process, achieved by cycling SAMmodified electrode in buffer solution until there is no change in current response, described by Finklea [13,14]. Voltammetric annealing is an important conditioning procedure for SAMs, since it helps to break out the metastable disordered states of the SAMs. Fig. 4 shows typical evolution of cyclic voltammograms of FePc-based SAM when subjected to 20 continuous voltammetric cycles. A similar decrease in current density was also observed for the SAMs of CoPc, MnPc and mercaptopyridine (not shown). There was a steady decrease in current of the voltammetric wave from the first scan until about the 18th scan when it became constant and remained approximately similar with subsequent scans. The steady decrease in current could be attributed to the voltammetric ‘shock’ received by the immobilised SAM species leading to its restructuring or reorganisation and subsequent removal of the physically adsorbed, metastable disordered MPc molecules. The stable voltammogram at the 18th cycle is an indication of the complete formation of the true ordered MPc-SAM. As discussed below, the final CVs for each of the three MPc complexes showed different surface-confined parameters.

As is clearly evident in Fig. 3, unlike the MPyr-SAM, each of the MPc-linked-MPyr-SAMs showed a pair of remarkably well-defined, reversible redox peaks. From the well established redox chemistry of CoPc, MnPc and FePc complexes [21] the redox waves are associated with the metal-centred, one-electron ([M(III) Pc(−2)]+ /[M(II) Pc(−2)] with M = Co, Mn and Fe) processes. Table 1 lists the estimated surfacedconfined, electrochemical parameters, which included the half-wave peak potentials (E1/2 ), the ratio of anodic to cathodic peak currents (Ipa /Ipc ), potential peak separations (Ep ) and surface coverage (Γ MPc ). The redox (E1/2 ) values of adsorbed MPc complexes have been reported [22], and were observed at ∼0 V for MnPc, −0.1 V for FePc and 0.3 V for CoPc. Thus, the trend of M(III)/(II) of CoPc > MnPc > FePc observed in this work is the same as reported. The differences in the actual values of potential observed in this work with those reported is certainly due to the presence of the axial mercapto ligand in this work. The Ep values for the MPc-SAMs (Table 1) were found to be similar at the investigated scans rates (10–300 mV s−1 ), however, the peaks become broader at the >300 mV s−1 . Ideal, surface-confined, diffusionless species is characterised by a zero Ep value [13,14]. In the present study, however, peak separations of 100, 140 and 200 mV (versus Ag|AgCl) for FePc, MnPc and CoPc, respectively, suggest kinetic limitations or electrostatic interactions of the molecules in the films. The Ep trend FePc < MnPc < CoPc indicates that the rates of the heterogeneous electron transfer in these MPc molecules are fastest for the FePc molecules and slowest for the CoPc molecules in the films. This disparity may be postulated to be due to electron tunnelling distances between each MPc and the gold electrode. Due to the dπ(metal) → dπ(phthalocyanine) back donation, the transition metal ions (Fe, Co and Mn) of the MPc complexes exhibit very strong M–N(pyridine) bonds [23–27]. Interestingly, the differences in the bond distances can exert some strong influence on the chemistries of these three MPc complexes. According to recent studies by Janczak et al. [26,27], the bond distance between the central metal of the MPc and that of the nitrogen atom of the pyridine ˚ < MnPc (M–N(Pyr)) follow this order: FePc ([2.039(2) A]) ˚ ˚ ([2.114(2) A]) < CoPc ([2.340(2) A]), hence making the FePc closest to the gold electrode while the CoPc farthest from it. The same authors [26,27] also proved that the sta-

K.I. Ozoemena, T. Nyokong / Electrochimica Acta 51 (2006) 2669–2677

bility of the MPc-pyridine co-ordination follow the order: FePc > MnPc > CoPc. In a previous report by Nyokong [25], it was shown that the rate of pyridine co-ordination of FePc is much faster than for CoPc. Long electron-tunnelling distance, coupled to slow and sluggish co-ordination ability of the CoPc with pyridine could invariably lead to its slow electron transport and, hence the relatively large Ep value. Another likely explanation for our observed trend (FePc > MnPc > CoPc) may be found on the reorganisation energies associated with the M(III)/(II) oxidation state changes. It has long been known [28,29] that metal-centered oxidation of cobalt (Co(III)/(II) ) in porphyrins exhibits sluggish heterogeneous and homogeneous electron transfer (ET) rates in comparison to the analogous Fe(III)/(II) oxidation in iron porphyrins. Indeed, cobalt complexes frequently exhibit significant changes in bond length and spin changes with changes in oxidation state [28,29]. Both events lead to large reorganisation energies, and thus slow standard rate constants. For example, the self-exchange ET rate constants (kex ) for tetraarylmetalloporphyrin complexes fall in the range of 10−2 to 104 M−1 s−1 for the Co(III)/(II) process [30] and 107 to 108 M−1 s−1 for the Fe(III)/(II) process [31–34]. The value of the Ipa /Ipc was unity for the FePc and MnPc but slightly higher than unity for the CoPc (Table 1). The larger anodic current for the CoPc is not understood at this time, however, it should be mentioned that sometimes this is observed for the Co(III) Pc/Co(II) Pc [35], where one redox wave exhibits higher peak current than its reverse wave. The surface concentrations of these redox-active MPc films (Γ MPc /mol cm−2 ) (Table 1) were estimated from the background-corrected electric charge, Q, under the anodic

2673

peaks in accordance with the theoretical relationship [13,14] (Eq. (1)): ΓMPc =

Q nFA

(1)

where n represents number of electrons transferred (assume ≈1), F the Faraday constant (96,485 C mol−1 ) and A is the geometric surface area of the electrode (0.02011 cm2 ). The estimated surface concentrations confirm monolayer coverage (∼10−10 mol cm−2 ) of flat orientation for MPc complexes [9–11,36,37]. The coverage estimated for the CoPcSAM is almost three times higher than those of the FePc and MnPc, which may suggest a perpendicular orientation. Next, we investigated the voltammetric response resulting from the reductive desorption of the modified electrodes. It is well known [13,14,38–40] that thiol complexes (RSH) are desorbed from the gold surface by the following reduction process in an alkaline aqueous solution (Eq. (2)); Au-SR + e− + M+ → Auo + RS− M+

(2)

where M+ is the cation from the electrolyte. Fig. 5 shows typical cyclic voltammetric responses of reductive desorption of gold electrodes modified with (a) MPyr-SAM, and MPyr-SAMs axially ligated with (b) CoPc (c) MnPc and (d) FePc in 0.2 M KOH solution obtained by scanning from −0.2 to −1.2 V (versus Ag|AgCl). Two well-defined mercaptopyridine desorption peaks were observed for the MPyrSAM at −0.66 and −1.03 V versus Ag|AgCl (Fig. 5a). The appearance of two peaks in this work is consistent with literature precedent [38], however, it should be pointed out that some laboratories have also observed just one peak at

Fig. 5. Cyclic voltammograms for the reductive desorption of (a) MPyr-SAM and MPyr-SAM axially ligated with (b) CoPc, (c) MnPc and (d) FePc. Electrolyte = 0.2 M KOH solution. Scan rate = 50 mV s−1 .

2674

K.I. Ozoemena, T. Nyokong / Electrochimica Acta 51 (2006) 2669–2677

about −0.55 V (versus Ag|AgCl) [20,39] and −0.90 V (versus SCE) [40]. The differences in the various reports are definitely due to variations in experimental conditions (such as the nature of electrodes, electrolytes, deposition times, etc.) employed. From the reductive desorption peak of the MPyr (and assuming the one-electron process from Eq. (2)) the surface coverage of the MPyr (Γ MPyr ) was estimated as 4.73 × 10−10 mol cm−2 . There have been several reports on the surface coverage of MPyr monolayer on gold; and they are in the range of 4.4–5.4 × 10−10 mol cm−2 [20,38,41] much less than the 7–8 × 10−10 mol cm−2 [42–44] of those of the nalkanethiol monolayers on gold. Also, Jin et al. [45] reported that the average molecular area in the MPyr monolayer on ˚ 2 , which is greater than 21.6 A ˚ 2 the average gold to be 26.4 A molecular area in the n-alkanethiol monolayers. Thus, the Γ MPyr value obtained in this work is comparable to the literature reports for the SAM MPyr on gold electrode. The lower Γ MPc value compared to the Γ MPyr is attributed to the small molecular sized MPyr compared to the large-sized flat-lying ˚ 2 [36,37]) MPc complexes. Presumably, there is a (∼200 A type of stereochemical frustration where MPc rings cover or block the perpendicular-lying pre-formed MPyr SAMs from binding onto the gold electrode. As evident from Fig. 5, on ligation with an MPc molecule, the major peak (at −0.66 V versus Ag|AgCl) shifted to a more negative values of −0.69 V (CoPc, Fig. 5b), −0.73 V (MnPc, Fig. 5c) and −0.78 V (FePc, Fig. 5d), while the minor peak (at −1.03 V versus Ag|AgCl) dramatically decreased in current density (>95%) and shifted to a more negative potential with a abroad peak around −1.10 V (versus Ag|AgCl) for CoPc and MnPc and not observed for FePc. A shift to a more negative value for the reductive potential desorption is indicative of a monolayer more effectively chemisorbed [39], therefore, this result points to FePc-SAM as the best formed and most stable SAM while CoPc-SAM is the least stable. This result corroborates the superior chemisorption of the FePc-SAM over MnPc and CoPc as already described above. Also, unlike the CV of the FePc-SAM, the CVs due to the CoPc and MnPc displayed weak peaks around the −0.3 V. This peak may be assigned to the M(II) Pc/M(I) Pc redox processes since such is a usual region for the occurrence of the Co(II) Pc/Co(I) Pc redox processes in alkaline conditions [37]. The absence of this peak for the FePc is not fully understood but may partly be due to its relatively smaller reductive desorption current density compared to those of the CoPc and MnPc complexes. In general, the different shapes of the reductive desorption waves, exhibiting different peak potentials, prove the formation of different redox-active MPc species on the gold surface. 3.2. Electrocatalytic activity Electrocatalytic activities of the MPc-based SAM gold electrodes were investigated using two biologically relevant molecules, l-cysteine and thiocyanate. l-Cysteine was chosen as a model for sulfhydryl molecules known for their physiological and clinical importance [46,47].

Fig. 6. Comparative cyclic voltammetric responses of (i) bare gold electrode and (ii) MnPc-linked-MPyr-SAM-modified gold electrodes in pH 4.0 buffer solution containing 8.5 × 10−5 mol L−1 thiocyanate.

Thiocyanate (SCN− ) is known [48,49] to be harmful to aquatic life. Clinical studies indicate that certain levels (ca. 10−5 –10−3 mol L−1 ) of thiocyanate in biological samples, such as saliva and urine, correlate to excessive cigarette smoking [48,49]. All studies were carried out in pH 4.0 buffer conditions. Fig. 6 shows typical cyclic voltammograms obtained on (i) bare gold electrode and (ii) MnPc-linked-MPyr-SAM on gold electrode in pH 4.0 buffer solution containing 8.5 × 10−5 mol L−1 thiocyanate. It can be seen that while bare gold showed a broad anodic peak at ca. 0.65 V, the same gold electrode modified with SAM of MnPc-linkedMPyr exhibited well-defined oxidation peaks at 0.58 V (versus Ag|AgCl). The anodic wave due to bare gold for SCN− is characteristic of a slow electron-transfer process while the waves exhibited by the electrode modified with the MnPcMPyr-SAM indicates that this SAM can be used to improve the electronic communication between the gold electrode and thiocyanate. Also, the MnPc-linked-MPyr-SAM clearly showed a negative potential shift and an enhanced current density (when compared to the bare gold or MPyr-SAMmodified gold electrode) suggesting its electrocatalytic activity towards SCN− . Fig. 7 is a typical example of the cyclic voltammetric responses of the electrodes based on the SAMs of CoPc (Fig. 7a), MnPc (Fig. 7b) and FePc (Fig. 7c) towards the anodic oxidation/detection of l-cysteine in pH 4.0 buffer solutions. We found that the three electrodes exhibited similar voltammetric shapes for the oxidation of the two analytes, except for the slight difference in the shapes of the background/buffer voltammograms, and also that the FePc-based electrode exhibited better response characteristics than those of the CoPc and MnPc (see Table 2). As summarised in Table 2, the FePc-modified electrodes showed higher catalytic peak current (Ip ) density for the two analytes than with the CoPc-SAM or MnPc-SAM based electrodes. However, the oxidative peak potentials (Ep ) values for the analytes are approximately similar (∼0.20 and ∼0.6 V (versus Ag|AgCl) for cysteine and thiocyanate, respectively) for the three different MPc-modified electrodes, suggesting similar rate of electron transfer. Although the Ip values obtained for the CoPcSAM for the two analytes are higher than those of the MnPc-

K.I. Ozoemena, T. Nyokong / Electrochimica Acta 51 (2006) 2669–2677

2675

Fig. 8. Typical square wave voltammetric responses of gold electrode modified with FePc-linked 4-MPyr SAM following addition of increasing concentrations of cysteine ((i) 0, (ii) 10, (iii) 20, (iv) 40, (v) 80 and (vi) 160 ␮mol L−1 ).

Fig. 7. Cyclic voltammetric responses of gold electrodes modified with MPyr-SAM axially ligated with (a) CoPc, (b) MnPc and (c) FePc in pH 4.0 buffer solution not containing (i) and containing (ii) 2 × 10−5 mol L−1 l-cysteine. Scan rate = 25 mV s−1 .

SAM, the Ep values for the MnPc are same as that of the CoPc-SAM. The Ip for the analytes were found to be directly proportional to square root of the scan rate (ν1/2 ), characteristic of a diffusion-controlled reaction. Fig. 8 shows a typical SWV obtained with the FePc-based SAM showing the current response with increasing l-cysteine concentration. A linear dependence of current versus concentration of l-cysteine was obtained over the range 4.0 × 10−6 –1.0 × 10−4 mol L−1 with a detection limit of 2.0 × 10−6 mol L−1 . A closer look at the voltammogram shows a slight shift of anodic peak potential Ep (from 0.16 V (curve ii) to a more positive peak potential (0.20 V (curve vii)) as the concentration of the l-cysteine was increased from 10 to 160 ␮mol L−1 . This behaviour was observed with the three MPc-SAM electrodes studied but was more pronounced with the CoPc-based elelctrode (∼30% shift) compared to FePc or MnPc. The behaviour is characteristic of slow electron transfer processes resulting, probably, from the competition of the analyte molecules for the electrode surface. It is noteworthy that this behaviour was not

Table 2 Cyclic voltammetric parameters of self-immobilised metallophthalocyanine (MPc, M = Co, Mn and Fe) complexes axially ligated to 4-mercaptopyridine self-assembled monolayers onto gold electrodes towards electrocatalysis of l-cysteine and thiocyanate in pH 4.0 buffer solutionsa MPc-SAM

CoPc FePc MnPc

Cysteine (2 × 10−5 mol L−1 )

Thiocyanate (8 × 10−5 mol L−1 )

Ep /V (vs. Ag|AgCl)

Ip /␮A cm−2

Ep /V (vs. Ag|AgCl)

Ip /␮A cm−2

0.20 0.18 0.20

13.92 16.41 9.95

0.58 0.55 0.58

87.52 114.37 76.58

a All the data were estimated under similar experimental conditions and at scan rate of 25 mV s−1 . All values are averages from five measurements performed during 2 weeks (R.S.D. < 3%).

2676

K.I. Ozoemena, T. Nyokong / Electrochimica Acta 51 (2006) 2669–2677

observed with our previous studies on the “octopus-oriented” MPc-SAMs [10,11], thus it can be speculated to be due to orientation of the MPc-MPyr-SAMs on the gold electrodes. It could be easier for an analyte to cause some changes (e.g. by tilting) on the umbrella (Fig. 1b) than on the octopus orientation. The electrocatalytic activities of these MPc complexes towards the detection of thiocyanate and l-cysteine occur near the formal potentials of the M(III) Pc/M(II) Pc redox processes of the MPc-linked-MPyr-SAMs investigated in this work. For example, a closer look at the Fig. 8 (curve i), one can easily observe the Fe(III) Pc/Fe(II) Pc redox peak. Similar observations have been made in previous reports on MPcbased SAMs [10,11], thus the basic mechanism through which electrocatalytic reactions operate at the MPc-MPyrlinked-SAM-modified gold electrodes for thiocyanate and l-cysteine involve the M(III) Pc/M(II) Pc redox processes. That is, initial oxidation of the M(II) Pc to M(III) Pc followed by the oxidation of the analyte to its products via M(III) Pc and subsequent regeneration of the M(II) species. 3.3. Electrode stability The cyclic voltammogram of each of the sensors after annealing (shown in Fig. 4) did not show any detectable change in shape after 100 scans in acidic and neutral aqueous solutions (pH 4.0 or 7.0 solutions) at potentials between −0.5 and +0.9 V (versus Ag|AgCl), which indicates electrochemical stability. Variation of catalytic peak currents with scan number (25 scans) was investigated at a fixed concentration of each of the analytes. A decrease in peak currents was observed after the first scan (∼10 and 42% for SCN− and cysteine, respectively), which is a typical behaviour for electrode fouling. However, by simply rinsing the electrode in a fresh pH 4.0 buffer solutions the electrode surface was renewed and the initial catalytic current was obtained, indicating that the oxidation products of these analytes are physically adsorbed onto the electrode surface and are incapable of making these modified electrodes unusable. The oxidation products of these analytes are responsible for the electrode fouling. The oxidation products of cysteine and thiocyanate when catalysed by MPc or MPc-like catalysts in pH 4.0 conditions have been reported to be cystine (for cysteine) [10,11] and thiocyanate radicals (for thiocyanate) [50,51]. The thiocyanate radicals easily dimerize to form the species (SCN)2 . This dimer (SCN)2 is unstable in the pH 4.0 region due to hydrolysis [50,51]. The long-term stability was checked by repeated analysis of a fixed concentration (8.5 × 10−5 mol L−1 ) of SCN− or cysteine on a twice-a-week basis. The catalytic currents (stability) decreased in the following the order; FePc < MnPc  CoPc. Unlike the other electrodes, FePcbased electrodes could reliably be used for the analysis of SCN− (up to 4 weeks) and cysteine (up to a week) without significant deterioration in catalytic current. The eightring-legged (octopus-oriented) MPc-SAMs [9–11] showed

much better long-term stability that these one-axial-legged (umbrella-oriented) MPc-SAMs. Presumably, the one-legged anchorage makes the umbrella orientation vulnerable to defects (e.g. tilting) and so exhibits shorter long-term stability than the octopus.

4. Conclusion A comparative surface electrochemistry and electrocatalytic investigations on the self-immobilised solid ultrathin films of metallophthalocyanine (FePc, MnPc and CoPc) complexes onto a preformed 4-mercaptopyridine self-assembled monolayer on gold electrode has been described. The study provides interesting electrochemical and electrocatalytic features that follow a similar trend found for the bond distance between the pyridine linker and the central metal of the MPc. A remarkable finding of this work is the superiority of FePc-based SAM over those of the MnPc and CoPc. The superior behaviour of FePc-SAM is proposed to be due to the favorable axial ligation properties of FePc with pyridine (i.e. smaller Fe–N(Pyr) bond length, faster co-ordination times and superior stability) over the MnPc and CoPc complexes. The advantageous properties of this type of MPc-SAM electrodes as potential electrochemical sensors could be found in their ease of fabrication and good catalytic activity. These findings are also likely to apply to the MPc related complexes, such as the Fe, Co and Mn porphyrin complexes.

Acknowledgements This work was supported by Rhodes University, the National Research Foundation (NRF, FA2004042800005 and GUN 2053657) and the Andrew Mellon Foundation for Accelerated Development Programme, South Africa.

References [1] C.C. Leznoff, A.B.P. Lever (Eds.), Phthalocyanines: Properties and Applications, vol. 1–4, VCH Publishers, New York, 1989–1996. [2] N.B. McKeown, Phthalocyanine Materials: Synthesis, Structure and Function, Cambridge University Press, Cambridge, 1998. [3] S. Baker, M.C. Pretty, G.G. Roberts, M.V. Twigg, Thin Solid Films 99 (1983) 53. [4] A. Arrieta, M.L. Rodriguez-Mendez, J.A. de Saja, Sens. Actuators B 95 (2003) 357. [5] Q. Zhang, D. Huang, Y. Liu, Synth. Met. 137 (2003) 989. [6] S. Palacin, Adv. Colloid Interface Sci. 87 (2000) 165. [7] M.J. Cook, Pure Appl. Chem. 71 (1999) 2145. [8] Z. Li, M. Lieberman, in: J.P. Blitz, C.B. Little (Eds.), Fundamental and Applied Aspects of Chemically Modified Surfaces, RSC, Lettchworth, 1999, p. 24. [9] Z. Li, M. Lieberman, W. Hill, Langmuir 17 (2001) 4887. [10] K. Ozoemena, T. Nyokong, Electrochim. Acta 47 (2002) 4035. [11] K.I. Ozoemena, T. Nyokong, P. Westbroek, Electroanalysis 15 (2003) 1762.

K.I. Ozoemena, T. Nyokong / Electrochimica Acta 51 (2006) 2669–2677 [12] A. Ulman, An Introduction to Ultrathin Organic Films: From Langmuir-Blodgett to Self-assembly, Academic Press, San Diego, 1991. [13] H.O. Finklea, in: A.J. Bard, I. Rubinstein (Eds.), Electroanalytical Chemistry, vol. 19, Marcel Dekker, New York, 1996, p. 109. [14] H.O. Finklea, in: R.A. Meyers (Ed.), Encyclopedia of Analytical Chemistry: Applications, Theory and Instrumentations, vol. 11, Wiley, Chichester, 2000, p. 10090. [15] G. Kalyuzhny, A. Vaskevich, G. Ashkenasy, A. Schanzer, I. Rubinstein, J. Phys. Chem. B. 104 (2000) 8238. [16] K.I. Ozoemena, T. Nyokong, J. Electroanal. Chem. 579 (2005) 283. [17] Z. Zhang, S. Hou, Z. Zhu, Z. Liu, Langmuir 16 (2000) 537. [18] Z. Zhang, T. Imae, Langmuir 17 (2001) 4564. [19] M.J. Stillman, T. Nyokong, in: A.P.B. Lever, C.C. Leznoff (Eds.), Phthalocyanines: Properties and Applications, vol. 1, VCH Publishers, New York, 1989, p. 133. [20] T. Sawaguchi, F. Mizutani, S. Yoshimoto, I. Taniguchi, Electrochim. Acta 45 (2000) 2861. [21] A.B.P. Lever, E.R. Milaeva, G. Speier, in: A.P.B. Lever, C.C. Leznoff (Eds.), Phthalocyanines: Properties and Applications, vol. 3, VCH Publishers, New York, 1993. [22] J.H. Zagal, Coord. Chem. Rev. 119 (1992) 89. [23] A. Yamamoto, L.K. Phillips, M. Calvin, Inorg Chem. 7 (1968) 847. [24] G. Pennesi, C. Ercolani, P. Ascenzi, M. Brunori, F. Monacelli, J. Chem. Soc., Dalton Trans. 6 (1985) 1107. [25] T. Nyokong, Polyhedron 14 (1995) 2325. ´ z, H. Borrmann, Y. Grin, Polyhedron [26] J. Janczak, R. Kubiak, M. Sled´ 22 (2003) 2689. [27] J. Janczak, R. Kubiak, Inorg. Chim. Acta 342 (2003) 64. [28] S. Fukuzumi, in: R. Guilard (Ed.), The Porphyrin Handbook, vol. 8, Academic Press, San Diego, 2000, p. 115. [29] H. Sun, V.V. Smimov, S.G. DiMagno, Inorg. Chem. 42 (2003) 6032. [30] R.D. Chapman, E.B. Fleischer, J. Am. Chem. Soc. 104 (1982) 1575. [31] D.W. Dixon, M. Barbush, A. Shirazi, J. Am. Chem. Soc. 106 (1984) 4638. [32] D.W. Dixon, M. Barbush, A. Shirazi, Inorg. Chem. 24 (1985) 1081.

2677

[33] A. Shirazi, M. Barbush, S. Ghosh, D.W. Dixon, Inorg. Chem. 24 (1985) 2495. [34] D.W. Dixon, S. Woehler, X. Hong, A.M. Stolzenberg, Inorg. Chem. 27 (1988) 3682. [35] Z.X. Zhao, K.I. Ozoemena, D.M. Maree, T. Nyokong, Dalton Trans. (2005) 1241. [36] N. Kobayashi, P. Janda, A.B.P. Lever, Inorg. Chem. 31 (1992) 5172. [37] Y.-H. Tse, P. Janda, H. Lam, W.J. Pietro, A.B.P. Lever, J. Porphyrins Phthalocyanines 1 (1997) 3. [38] B.D. Lamp, D. Hobara, M.D. Porter, K. Niki, T.M. Cotton, Langmuir 13 (1997) 736. [39] J.R. de Sousa, A.A. Batista, I.C.N. Di´ogenes, G.F.S. Andrade, M.L.A. Temperini, L.G.F. Lopes, I.S. Moreira, J. Electroanal. Chem. 543 (2003) 93. [40] N. Kanayama, T. Kanbara, H. Kitano, J. Phys. Chem. B 104 (2000) 271. [41] T. Sawaguchi, F. Mizutani, I. Taniguchi, Langmuir 14 (1998) 3565. [42] D.E. Weisshaar, B.D. Lamp, M.D. Porter, J. Am. Chem. Soc. 114 (1992) 5860. [43] M.M. Walczak, C.A. Alves, B.D. Lamp, M.D. Porter, J. Electroanal. Chem. 396 (1995) 103. [44] C.J. Zhong, J. Zak, M.D. Porter, J. Electroanal. Chem. 421 (1997) 9. [45] Q. Jin, J.A. Rodfiguez, C.Z. Li, Y. Darici, N.J. Tao, Surf. Sci. 425 (1999) 101. [46] M.T. Goodman, K. McDuffie, B. Hernandez, L.R. Wilkens, J. Selhub, Cancer 89 (2000) 376. [47] J.K. Liu, H.C. Yeo, E. Overvik-Douki, T. Hagen, S.J. Doniger, D.W. Chu, G.A. Brooks, B.N. Ames, J. Appl. Physiol. 89 (2000) 21. [48] D. Gao, J.-Z. Li, R.-Q. Yu, Anal. Chem. 66 (1994) 2245. [49] M.K. Amini, S. Shahrokhian, S. Tangestaninejad, Anal. Chim. Acta 402 (1999) 137. [50] S. Modi, S.S. Deodhar, D.V. Behere, S. Mitra, Biochemistry 30 (1991) 118. [51] S. Adak, A. Mazumdar, R.K. Banerjees, J. Biol. Chem. 272 (1997) 11049.