A novel third-generation xanthine biosensor with enzyme modified glassy carbon electrode using electrodeposited MWCNT and nanogold polymer composite film

Abstract

A novel nanobiocomposite for immobilization of xanthine oxidase (XO) was developed by incorporating functionalized MWCNT in nanogold doped poly(o-phenylenediamine) (PPD) (Au–PPD) film on glassy carbon electrode (GCE) for selective and sensitive detection of xanthine in real samples e.g. blood, urine, fish. Stable colloid of o-phenylenediamine (OPD) and HAuCl₄ in acidic environment was electropolymerized on working electrode (GCE) to form an ultrathin film of AuNP–PPD which possessed permselectivity and no interference against electroactive species such as ascorbic acid and uric acid. Spectrophotometric and microscopic analysis confirmed the doping behaviour of AuNP. Electrodeposition of carboxylated MWCNT onto the Au–PPD film increased conductivity, sensitivity and also facilitated a microenvironment to entrap XO enzyme by covalent bonding, enhancing storage stability. The conductive nature of the electrode after every step of modification was investigated by electrochemical impedance spectroscopy. High I_max/K_map value was achieved by the XO/fMWCNT/Au–PPD modified electrode. Oxidation of xanthine on this modified electrode was diffusion-controlled involving two electrons in the rate-determining step with a transfer coefficient (α) of about 0.596. Differential pulse voltammetric study of XO/fMWCNT/Au–PPD/GCE exhibited good analytical characteristics e.g. low detection limit (12 nM) (S/N = 3), a wide linear range of 0.01–300 μM (R² = 0.994), good sensitivity (14.03 μA μM⁻¹ cm⁻²), fast response (6 s) at anodic potential of +0.625 V vs. Ag/AgCl (pH 7.0). It retained 91% of its initial activity even after 210 times of use over a period of 4 months when stored at 4 °C. The applicability of the xanthine biosensor was tested by performing reproducibility, repeatability and interference study on real samples.

---

1. Introduction

Freshness of fish, meat and other derived products is a prime requirement for human health. There is worldwide demand for reliable, handheld analytical tools to monitor freshness of fish and meat. When animal tissue dies, ATP degrades to xanthine and the pathway is: ATP → ADP → AMP → IMP → HxP → Hx → X.¹ Thus the levels of xanthine in fish product can be used as an index for evaluating meat or fish freshness.^1,2 Determination of xanthine level in blood, urine tissue is also essential for medical diagnosis and management of various diseases such as hyperuricemia, gout, xanthinuria and renal failure.³ Development of a xanthine sensor is thus of immense importance in food, medical and biological research.⁴ Xanthine oxidase (EC:1.17.3.2, XO), the metalloflavoprotein is a key enzyme at the end stage of protein degradation during hydroxylation of purines in the 2, 6 and 8 ring positions to form purine followed by hypoxanthine/xanthine and finally uric acid. Biosensors, in general, show greater sensitivities compared to traditional physicochemical, biological and serological tests based on UV/vis spectrophotometry, high performance liquid chromatography and gas chromatography.^5,6 Electrochemical biosensors combine the advantages of the specificity of the enzyme for recognizing particular target molecules with direct transduction of the rate of reaction into a current. Since biomolecules used in biosensing are not conducting in nature, conducting microenvironments are required to transfer the electrons from active site of the enzyme to the electrode surface.^7–9 The sensitivity of a biosensor mainly depends on the conductivity and nature of entrapment matrix for enzyme immobilization on the electrode.

A support of polymeric matrix enhances speed, sensitivity and versatility in diagnostics of target analytes. Conducting polymers such as polyaniline (PANI), polypyrrole (PPy) and poly[3,4-ethylenedioxythiophene] (PEDOT) have been used for biosensing of xanthine, glucose and uric acids.^4,7,10–13 Although, o-phenylenediamine (OPD) is an aniline derived polymer with an extra –NH₂ group, the oxygen reduction ability at PPD film (in reduced state) is unique compared to PANI, PPy [Gajendran 2007].¹⁴ The excellent permselectivity properties of electropolymerized poly(o-phenylenediamines) (PPD) have wide application for designing oxidase-based biosensors due to high permeability to H₂O₂ and efficient blockers of interference compounds.^15–17 Nanoparticles such as Au, Ag, Pt, ZnO, Fe₃O₄ have attracted enormous interest in past years due to their superior role in acceleration of the electron transfer rate from the enzyme to the working electrode.^4,15,18–20 The main advantages of gold nanoparticles (AuNP) used in biosensing applications are nontoxicity, good biocompatibility, and high electron communication rate. Thus the application of AuNP could be useful in biosensing replacing external electron-transfer mediators.^18,20 It was demonstrated that polyaniline (PANI), polypyrrole (PPy), PANI–PPy copolymers, poly(3,4-ethylene dioxythiophene) (PEDOT), poly(phenylene vinylene) (PPV) and several other conducting polymers (CPs) could spontaneously reduce noble metal ions (e.g. Ag⁺, Au³⁺, Pd²⁺ and Pt²⁺) to the zero-valent metals.^7,19–21 Conducting polymers have been used as supporting matrices for intercalation of important nanoparticles to retain the catalytic activity of enzymes in the composite.^7,19,22 However, ultrathin Au-doped PPD film (Au@PPD) through a single step electrodeposition of Au–PPD nanocomposite has not been reported in literature.

In recent years, carbon derivatives such as carbon nanotubes (CNT), graphite and graphene have been used owing to their unique electrical, mechanical, structural and chemical properties. The potential use of functionalized MWCNT for biosensing applications have shown great promise for diagnosis of trace molecules due to their excellent electrical conductivity, ultra-high mechanical strength, good chemical stability, high specific area and high dimensional ratios.^8,9,14,23,24 In the last few years, a few reports have been published on electrochemical xanthine biosensors. The comparison of performances of some well characterized xanthine biosensors is given in Table 1.^22–31 Most of these suffered from low storage stability, non-reusability, high time of response, low electron transfer rate and complexity of fabrication process.^{4,10,20,23–31} The complexity of the enzyme entrapment would hinder acceptability for commercial use. The lack of storage stability due to leaching of enzymes could be overcome by covalent linkage of enzyme to the electrode surface and the main aim of our study concentrated on fabrication of a highly stable and sensitive xanthine biosensor with low detection limit.

Table 1 Analytical characteristics of some recently developed xanthine oxidase based biosensors for detection of xanthine^a

Reference	Electrode modification	Method of immobilization	Eapp vs. Ag/AgCl/technique	Optimum pH	Linear range (μM)	Detection limit (μM)	Sensitivity	Response time (s)	Storage stability (%)	Analyte/applications
a GCPE: glassy carbon paste electrode; CD = cyclodextrin; CHIT = chitosan; SWCNT = single walled carbon nanotube; PtE = platinum electrode; PGE = pencil graphite electrode; CPE = carbon paste electrode; X = xanthine, Hx = hypoxanthine, Naf = Nafion; GMA-co-VFc = poly(glycidyl methacrylate-co-vinylferrocene); amp = amperometry, DPV = differential pulse voltammetry.
26	XO–Au-NP–GCPE		0.70 V/amp	7.5	0.5–10	—	0.24 μA μM^−1	—	7 days (72% retain)	X/Hx
27	XO/LAPONITE®	Physico-adsorption	0.39 V/amp	7.5	0.039–21	0.01	6.54 mA M^−1
23	XO/ZnO-NP/CHIT/c-MWCNT/PANI/PtE	Covalent	0.050 V/amp	7.0	0.1–100	0.1	—	4	1 month (70% retain after 80 uses)	X/fish
10	GMC/GCE	—	0.65 V/DPV	7.0	20–320	0.388	0.062 μA μM^−1	—	—	UA, X, Hx/fish, blood, urine
28	Naf/XO–CD/pAuNP/SWNT/GCE	Physico-adsorption	0.65 V/amp	7.0	0.05–9.5	0.04	0.152 μA μM^−1	—	—	—
29	XO/GNPs–SWCNT/PtE	Physico-adsorption	0.4 V/amp	7.4	2000–37 300	0.61	0.141 μA μM^−1	—	2 week (88% retain)	—
30	XO/CHIT/Fe-NPs@Au/PGE	Covalent glutaraldehyde	0.5 V/amp	7.4	0.1–300	0.1	1.169 μA μM^−1	—	100 days (25% loss)	X/fish
31	XOD/CHT/Pt NPs/PANI/Fe3O4/CPE	Covalent glutaraldehyde	−0.35 V/amp		0.2–36	0.1	13.58 μA μM^−1 cm^−2	—	3 month (85% retain after 100 uses)	—
24	P(GMA-co-VFc)/MWCNT/XO	Physico-adsorption	0.35 V/amp	7.0	2–48	0.12	16 mA M^−1	4	25 days (70% retain)	X/fish
This work	XO/fMWCNT/Au–PPD/GCE	Covalent	0.625 V/DPV	7.0	0.01–300	0.012	14.03 μA μM^−1 cm^−2	5	4 months (91% retain after 210 uses)	X/fish, blood, urine

---

A new design of sensing matrix using the favourable effect of conducting nature of OPD, AuNP and fMWCNT matrix was used for entrapment of XO for detection of xanthine in real samples. Incorporation of AuNP in PPD film formed an ultrathin, adherent layer on which a further electrodeposition of carboxylated MWCNT was made. XO could successfully bind with activated –COOH groups of fMWCNT without altering the enzyme activity. In many publications amperometric xanthine biosensors have been reported and these were based on the cyclic voltammetry (CV) i.e. current measurements at fixed applied potential^4,23–31 whereas differential pulse voltammetry (DPV) could be a better measurement technique due to sharper response peaks i.e., higher sensitivity. In the present research DPV based analysis showed higher sensitivity and selectivity for xanthine in real samples such as human blood serum, urine, as well as fish.

2. Experimental methods

2.1 Chemicals and apparatus

Multiwall carbon nanotubes (MWCNT), o-phenylenediamine (OPD), xanthine oxidase (XO, E.C.1.1.3.22 from microorganism) and gold chloride salt (HAuCl₄·3H₂O) were purchased from Sigma Aldrich, USA. 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimidehydrochloride (EDC), N-hydroxysuccinamide (NHS), xanthine (Xn), hypoxanthine (HyX) and uric acid (UA) were obtained from Himedia, India. Potassium dihydrogen phosphate (KH₂PO₄) and dipotassium hydrogen phosphate (K₂HPO₄), sodium sulfate (Na₂SO₄), L-ascorbic acid (AA), hydrogen chloride acid, nitric acid, sulfuric acid and perchloric acid were procured from E-merck (Mumbai, India). A three-electrode system having glassy carbon working electrode (WE), Ag/AgCl reference electrode (RE) and platinum (Pt) counter electrode (CE) was used for all electrochemical analysis. The electrochemical measurements were conducted using an IVIUMStat electrochemical analyzer (Model: A09050, Iviumstat Technologies, USA) with IviumSoft software. Electrochemical impedance spectroscopy (EIS) was performed at each step of electrode modification using a frequency response analyser (Eco Chemie B.V, Utrecht, Netherlands) attached with Autolab, AUT72660 and controlled by FRA 4.9.006 software. EIS study of the modified electrodes was carried out in 5 mM of [Fe(CN)₆]₃⁻ and [Fe(CN)₆]₄⁻ with 0.1 M KCl at frequency range 1 MHz to 0.01 Hz, amplitude 10 mV and fixed potential of 0.28 V. The detailed measurement procedure for SEM, FESEM, TEM, FTIR, XRD and EDAX is given in ESI.†

2.2 Carboxylation of MWCNT

Functionalization of commercial MWCNT (average diameter ∼50 nm) was the most important step in sensor fabrication. 20 mg of commercially available MWCNT was dispersed in 4 M HCl for 2 h with the aid of ultrasonic agitation to eliminate metal oxides present within the MWCNT.^32,33 The separated MWCNT was rinsed with RO water until the pH became neutral. The dried MWCNT then was dispersed in 24 ml of a mixture of concentrated HNO₃ and H₂SO₄ (1 : 3 v/v) with constant stirring of 1 h at 50 °C followed by sonication for 2 h in an ultrasonic bath with occasional stirring to get functionalized and shortened fMWCNT. The fMWCNT was separated by centrifugation at 10 000 rpm for 10 min. The pellet of fMWCNT was washed with deionized water several times and neutralized by 1 M NaOH. Finally it was centrifuged at 10 000 rpm for 10 min to collect the fMWCNT. The pellet was dried overnight in a hot air oven.

2.3 Fabrication of modified electrode

A three-electrode assembly was used for all the electrochemical experiments. The glassy carbon working electrode and platinum counter electrode were first cleaned with polishing kit. The electrodes were washed with RO water thoroughly after sonication in ethanol–water mixture for 5 min, and then cleaned electrochemically by cyclic voltammograms of 5 cycles (−0.5 to 1.5 V) in 0.5 M sulfuric acid. Stable Au/o-PD colloidal solution was prepared by dropwise addition of 1 mM HAuCl₄ into the 0.08 M OPD monomer solution in 0.1 M HCl and 0.1 M Na₂SO₄ under continuous stirring for 10 min. The Au-doped thin conducting polymeric film of OPD was electrodeposited on GCE by cyclic voltammetry (CV) in the potential span of −0.4 to 1.4 V at a scan rate 0.05 V s⁻¹ for 20 cycles in Au/o-PD colloidal solution. A bluish green ultra-thin layer of Au-doped PPD film was formed onto the shiny surface of GCE. The unbound monomer was removed further by chronoamperometry at −0.2 V for 10 min in deionized water. 0.05% (w/v) fMWCNT was sonicated in deionized water for 10 min before electrodeposition onto the Au–PPD modified GCE surface. The chronoamperometric electrodeposition was performed at 1.7 V for 30 min to construct fMWCNT/Au–PPD/GCE. The surface coverage by the fMWCNT was calculated to be 2.78 × 10⁻⁶ mol cm⁻². The fMWCNT on the electrode surface was further activated by EDC–NHS. The electrode was first stirred for 30 min in 10 mM EDC in phosphate buffer (50 mM, pH 7.0) and 10 mM NHS was then added and stirred for a further 1 h to form a stable NHS–carbodiimide–ester onto the surface of fMWCNT. The modified electrode was washed three times with buffer to remove nonspecific chemicals. Finally 0.1 U of XO enzyme was adsorbed covalently with activated fMWCNT onto the electrode surface to construct a unique mediator free design of XO/fMWCNT/Au–PPD/GCE. The –COOH groups of fMWCNT were linked covalently to –NH₂ groups of XO which provided a more stable complex than physical aggregation. A schematic diagram of stepwise electrode fabrication of the proposed XO/fMWCNT/Au–PPD/GCE sensor is shown in Scheme 1.

Scheme 1 Fabrication steps for a unique design of the enzyme electrode XO/fMWCNT/Au–PPD/GCE.

Variation in fabrication process was considered to evaluate the best design for xanthine biosensor which could provide higher stability and sensitivity. To look into the effect of Au-doped PPD, PPD film with no added Au was electropolymerized onto the bare GCE in the absence of HAuCl₄ with the same electropolymerization conditions to construct PPD/GCE. A layer of AuNP was electrodeposited by chronoamperometry at −0.273 V in 1 mM chloroauric solution onto another PPD/GCE modified electrode to form Au/PPD/GCE. To evaluate the change in response with Nafion coverage, an electrode was modified to construct Naf/XO/fMWCNT/Au–PPD/GCE. All enzyme modified electrodes were stored in buffer (50 mM, pH 7) at 4 °C until further use. Table 2 shows the response due to variation in fabrication.

Table 2 Analytical parameters obtained from xanthine calibration curve for different modified electrodes

Modified electrode	Imax (μA)	Kmap (μM)	Imax/Kmap (nA μM^−1)	LRS (nA μM^−1)	LOD (μM μA^−1)	Range (μM)
XO/PPD/GCE	1.4645	218.136	6.714	1.56	1.2	100–400
XO/Au/PPD/GCE	7.8549	611.995	12.835	6.8	330.84	10–500
XO/Au–PPD/GCE	9.738	74.265	131.125	86.6	1.03	5–50
XO/fMWCNT/Au–PPD/GCE	43.464	122.879	353.715	991.8	0.012	0.01–35
Naf/XO/fMWCNT/Au–PPD/GCE	37.174	142.658	225.817	187.8	5.667	0.5–50

---

2.4 Sample preparation

Three different samples were tested. The fish samples e.g. Labeo rohita (F1) and Lates calcarifer (F2) were purchased from a local market in Kolkata. The fish extract was prepared by following published report.^20,23,30 1 g of fish flesh was converted into a fine paste using 5 ml of 0.5 M perchloric acid (HClO₄) by a motor pestle. The extract was stirred mechanically for 10 min to make a homogenized mixture. Centrifugation was then performed at 6000 rpm for 15 min. The supernatant was collected and neutralized by dropwise addition of 0.6 M NaOH. Blood serum and urine samples were collected from a medical diagnostic centre in Kolkata, India (not collected directly from human volunteers). These were filtered through 0.45 micron membrane (Millipore, India) before analysis. 50 μl of each sample was added to the electrochemical cell to monitor the xanthine level in the real samples. The xanthine levels in real samples were also quantified using a C-18 HPLC column (Nova-Pak C18, 3.9 × 150 mm) with 5 μm pore size and a binary pump system of WATERS 2487 (Massachusetts, USA) by using a mobile phase of methanol–water–acetic acid (7.5 : 92 : 0.5 v/v/v) with 0.5 ml min⁻¹ of flow rate at 272 nm. All experiments were performed in compliance with the relevant laws and institutional guidelines. The institutional committee has also approved the experiments.

3. Results and discussion

3.1 Electrochemical deposition of Au-doped poly(o-phenylenediamine) [Au–PPD] on GCE

Fig. 1 depicts the cyclic voltammetric growth profile of PPD and Au–PPD onto bare GCE along 20 scans in the range of −0.4 to 1.4 V. In the first cycle, a broad oxidation peak with 1.556 times higher anodic current appears at 0.74 V for Au–PPD than only monomer oxidation (at 0.88 V), indicating the faster formation of OPD radical cation through one-electron oxidation of amino groups.¹⁴ For both the electrodes, the oxidation peak current decreased gradually. In the case of Au–PPD (Fig. 1b), the second scan produced two oxidation peaks at 0.67 and 1.01 V, both of which diminished with number of scans. On the reverse scan, there were no corresponding cathodic peaks, indicating that active cation radicals underwent polymerization reaction immediately.¹⁴ For the second scan another oxidation peak was observed at −0.08 V with larger peak current than only PPD, that decreased gradually until scan 5, after that two peaks appeared at −0.11 and 0.01 V with increasing current in the following scans, that was not observed for polymerization of PPD. At the lower potential, two cathodic peaks at −0.14 and 0.04 V merged into one peak at −0.02 V with continuous scan for monomer polymerization (Fig. 1a), whereas two distinct peaks at −0.15 and −0.04 V appeared with higher I_p value during formation of Au–PPD film (Fig. 1b). The increase in peak current of the redox pair at 0.11/−0.15 V in the presence of HAuCl₄ indicated successful incorporation of gold nanoclusters into the Au–PPD film. The thickness of the Au–PPD film was calculated to be 178.2 ± 12 nm using the equation d = mQ/FAρ,¹¹ where, m denotes the molecular weight of the monomer; Q the electric charge during the electropolymerization; F the Faraday constant (F = 96 485 C mol⁻¹); A the surface area of the working electrode (i.e. 0.0707 cm²) and ρ the density of PPD. The ultrathin film of polymeric matrix could be synthesized only by electro-polymerization ensuring many advantages compared to other coating methods for enzyme based electrode fabrication: (i) high uniform polymeric matrix in a controllable manner for enzyme immobilization, (ii) fast electron transfer and mass transport, (iii) removal of intralayer diffusion mass transport limitations. Thus supportive matrix for enzyme immobilization could remain constant in repeated synthesis during sensor fabrication. The surface coverage Γ of the electrode was evaluated by integrating the anodic peak current, determining the average charge Q using Faraday’s law [eqn (1)]¹¹ and was found to be 1.648 × 10⁻⁹ ± 0.26 mol cm⁻² for Au–PPD film, whereas surface coverage by only PPD film was calculated as 1.254 × 10⁻⁹ ± 0.31 mol cm⁻².

Γ = Q/nFA (1)

where n is the number of electrons transferred in redox reaction, the other symbols have their usual meaning.

Fig. 1 Cyclic voltammogram during electropolymerization of poly(o-phenylenediamine) in absence (a) and presence of HAuCl₄ (b). Inset: change of CV patterns from scan 1 to 20. Conditions: scan rate: 0.05 V s⁻¹, potential range: −0.4 to 1.4, electrodes: WE: bare GCE, RE: Ag/AgCl, CE: platinum wire.

3.2 Spectrophotometric and microscopic characterization of Au-doped PPD film

Scheme 2 represents the proposed chemical structure of the Au-doped PPD film that may be formed by electrochemical growth through head-to-head and/or head-to-tail coupling of OPD cation radicals. Delocalization of electrons occurred between chains and neighbouring redox sites of polymer during oxidative or reductive electro-polymerization. Previous work,^34,35 explained that the electron rich N atom of amines had strong affinity towards electron deficient AuNP, which could be comparable to the Au–S bond energy. OPD has two sp² nitrogen atoms in the aromatic amine ring, thus a strong N–Au bond may form easily. As a result, OPD can act as an oxidant in presence of AuCl₄⁻. Fig. 2a displays a rapid color change from yellow to brown with addition of increasing concentration (0.01, 0.05, 0.1, 0.25, 0.5, 0.75, 1 mM) of AuCl₄⁻ suggesting the formation of a stable Au–PPD colloidal solution by reduction of Au³⁺ to Au⁰ and simultaneously chloride ions were incorporated during polymerization of OPD in an acidic environment.³⁴ This phenomenon could be characterized by spectrophotometric analysis of the colloidal mixture before and after addition of AuCl₄⁻. The UV/vis spectra are shown in Fig. 2a. The peak at 281 nm could be assigned to the π–π* transition of the benzenoid ring.³⁶ Other peaks at 415, 451 and 474 nm were observed after electro-oxidation of OPD and these were absent in the monomer. These peaks were gradually more pronounced with increase in concentration of Au doping due to the charge transfer excitation e.g. transition related to the benzoid unit in the reduced state of the polymer.^14,37 The peak at 451 appeared for phenazine-like dimers/oligomers formation by cyclization (internal coupling); the peak at 415 nm could be attributed from intermediates of the dimeric or oligomeric species containing phenazine structure. The absorption bands at 474 nm were assigned to the low intensity of cation radicals of aniline-type dimer that further was coupled to the PANI-like radical or its dicationic form.³⁸ The variation of fluorescence spectra also confirmed the interaction of gold nanoclusters to polymer, as is depicted in Fig. 2c. Only PPD exhibited an emission peak at 466 nm when excited at 395 nm, whereas Au-doped PPD showed two distinct maxima at 467 and 567 nm in acidic environment. With the addition of HAuCl₄ from 0.01 to 1 mM, the emission peak intensity became enhanced at 566 nm but subsequently decreased at 466 nm gradually. Thus the new composite shows unique physical characteristics owing to Au doping.

Scheme 2 The proposed mechanism of electrochemical growth of Au-doped PPD film.

Fig. 2 (a) Change of color profile due to addition of 0 to 1 mM AuCl₄⁻; (b) change of UV/vis spectra pattern for Au-doped PPD, inset: increasing peak intensity pattern for addition of 0, 0.01, 0.05, 0.1, 0.25, 0.5, 0.75, 1 mM AuCl₄⁻; (c) fluorescence spectra for increasing concentration (0–1 mM) of Au doping in PPD film; (d) FTIR spectra of OPD monomer (i), PPD (ii), Au-doped PPD (iii); (e, f) SEM and FESEM image of Au–PPD respectively; (g) TEM image of Au–PPD, inset showed the HRTEM image of the selected area of (g); (h) diffraction pattern of Au–PPD showing crystallinity of AuNP; (i) EDAX of Au–PPD film.

The FTIR spectra of OPD, PPD were compared with the Au-doped PPD film that was synthesized on the GCE surface in Fig. 2d. A typical IR spectrum profile of OPD with two peaks at 3384 and 3362 cm⁻¹ could be characterized for asymmetric and symmetrical N–H stretching vibrations, while the bands at 1271.44 and 1154.74, 1057.77 cm⁻¹ could be ascribed for C–N stretching vibrations. The two bands at 1497.7 and 1456.96 cm⁻¹ corresponded to C=C stretching vibrations of benzenoid rings, and shifted after electropolymerization. An IR band of OPD at 3384 cm⁻¹ was due to the presence of primary amine that changed into N–H bending of secondary amine at 1504 cm⁻¹ in the PPD film during electropolymerization. The IR spectrum of PPD and Au-doped PPD was nearly the same indicating no structural changes during polymerization. A ladder polymeric chain of PPD was characterized by the following peaks at 3405, 1628, 1537, 1315 which might be ascribed to N–H, C=N, C=C, C–N stretching vibration of phenazine structures in PPD, respectively.^14,36–39 Two other peaks at 977 and 761 cm⁻¹ could be attributed to the out of plane C–H bending vibrations of aromatic benzene with the phenazine skeleton.^36–39 The peak intensity of the IR band for sulfate dopant at 1093 cm⁻¹ increased in the presence of AuNP of PPD film. Since, incorporation of AuNP did not change the IR band position of Au–PPD film, successful electrodeposition of Au-doped PPD film could be predicted. A shift of 2 cm⁻¹ for the sulfate group stretching at 1092 cm⁻¹ in the presence of AuNP in PPD film was likely due to a change in their dipole moment when AuNP becomes bound to the surface of high electron density.

A spherical disposition of nanoclusters of AuNP was observed clearly in SEM and FESEM image of Au-doped PPD film (Au–PPD) in Fig. 2e and f respectively, that was not observed for only PPD film (image not shown). The TEM image in Fig. 2e, also clearly showed that spherical AuNPs (4–16 nm) (11.2 ± 4.5) were monodispersed as well as nanoclustered in the Au–PPD film, while the EDAX spectra (Fig. 2g) confirmed the presence of AuNP in the Au–PPD film. The electron diffraction (SAED) pattern of thin film displayed polycrystalline diffraction rings with bright spots that particularly reveal good orientations of microcrystals (Fig. 2b) in a number of different directions. The crystalline property of Au–PPD thin film was characterized by XRD, as shown in Fig. S1.† A number of Bragg reflections with 2θ values of 22.91, 28.38, 32.91, 38.68, 45.84, 55.27 and 61.78° were observed. Several main peaks centred at 2θ = 22.91, 28.38, 32.91° were characterized for doped PPD, corresponding to the periodically parallel and perpendicular chains of the polymer matrix.³⁹ The familiar peaks appearing at 2θ values of 38.68, 45.84, 55.27, 61.78° were for the (111), (200), (220) and (311) sets of lattice planes, respectively, depicting the face centred cubic (fcc) structure of gold nanocrystals.

3.3 Functionalization of MWCNT

Functionalization of as-received MWCNT was a crucial step for formation of a uniform suspension before electrochemical deposition (Fig. 3a). After acid treatment, abundantly negative charged MWCNT provided a microenvironment to form a uniform stable aqueous suspension through electrostatic interaction (Fig. 3b). In the FTIR spectra (Fig. 3c), several drastically enhanced absorption peaks in the range of 3400–3800 cm⁻¹ were observed in the fMWCNT, compared to commercial MWCNT, confirming the presence of –OH groups on the backbone of fMWCNT. Two peaks at 3485 and 1660 cm⁻¹ were due to the O–H stretching vibration and C=O vibrations of carboxylic group, respectively, indicating a successful incorporation of –COOH to the end or sidewalls of the MWCNT. Fig. 3d shows the FESEM image of fMWCNT. From the TEM images, it could be clearly observed that the fMWCNT became shortened and thinner (Fig. 3f and g) during acid treatment, compared to the as-received MWCNT (Fig. 3e). The average diameter of the fMWCNT was reduced to 18.36 ± 7.35 nm. Fig. 3g displays the defects of the side walls of the nanotubes due to chemical oxidation by strong acids. This defect increased the reactivity due to presence of functional groups on its surface. Two XRD peaks at 2θ of 25, 42° were changed after functionalization. The crystallization pattern was altered to 2θ values of 28, 33, 48, 54, 59, 73° confirming the carboxylation of MWCNT (Fig. S2a†). EDAX spectra of fMWCNT (Fig. S2b†) also confirmed the presence of oxygen on the surface of fMWCNT.

Fig. 3 (a) A schematic diagram on functionalization of MWCNT; (b) the Eppendorf tubes show higher solubility of FMWCNT after acid treatment, (c) change of FTIR spectra for functionalized MWCNT, (d) FESEM and (e) TEM images of MWCNT; (f and g) TEM image of fMWCNT.

3.4 Electrochemical kinetic studies for xanthine determination and data analysis

Different immobilization designs (already mentioned in Section 2.3) were followed to achieve optimum catalytic activity of the enzyme. To evaluate the catalytic properties of electrodes to the oxidation of xanthine, characteristic CVs were recorded (Fig. 4) in the potential range of 0.3–1.0 V with a scan rate of 50 mV s⁻¹. Fig. 4 shows the CVs of XO/PPD/GCE (a), XO/Au–PPD/GCE (b), XO/fMWCNT/Au–PPD/GCE (c), and Naf/XO/fMWCNT/Au–PPD/GCE (d) in 0.05 M potassium phosphate buffer containing 50 μM xanthine, respectively. As can be seen in Fig. 4, the responses of XO/PPD/GCE and XO/Au–PPD/GCE towards xanthine are very weak, which might be due to the very slow electrode kinetics. It was observed that the anodic peak current at 0.65 V was enhanced 42-fold after incorporation of fMWCNT on Au–PPD film, whereas, it could decrease 1.5 times due to an outer layer coating of Nafion. Thus, the new design of biopolymeric matrix consisting of XO/fMWCNT/Au–PPD provided the best microenvironment for direct electron transfer from catalytic sites of the enzyme to the electrode surface by means of conducting tunnel formation through the novel nanobiocomposite layer.

Fig. 4 Cyclic voltammetric response of 50 μM xanthine in 50 mM phosphate buffer for XO/PPD/GCE (a), XO/Au–PPD/GCE (b), XO/fMWCNT/Au–PPD/GCE (c) and Naf/XO/fMWCNT/Au–PPD/GCE (d). Conditions: potential range: 0.3–1.0 V, scan rate: 0.1 V s⁻¹. Electrodes: RE: Ag/AgCl, CE: platinum wire.

To further prove the electrocatalytic activity of XO/fMWCNT/Au–PPD/GCE, the analytical parameters obtained from the calibration plot using DPV has been noted in Table 2. The Michaelis–Menten equation applied to electrochemistry is given by:

1/I_s = 1/I_max + [(K_map/I_max)/S] (2)

where, I_max is maximum current value at enzyme–substrate saturation and K_map is the substrate concentration at which the current response is I_max/2; that actually represents the enzyme affinity to the substrate. The values of I_max and K_map were evaluated by non-linear regression analysis using MATLAB 7.1. Highest I_max value related to optimum enzyme activity was achieved by the sensor design of XO/fMWCNT/Au–PPD/GCE (shown in Table 2). The conservation of the native structure of enzyme increased the electroactive surface, improving the performance of the transducer. Since, enzymatic turnover number increased proportionately with the ratio of I_max to K_map, the optimum design was considered on the basis of higher value of I_max/K_map.^7,11,16 Nanostructure variation of immobilization matrix could be responsible for increase of active surface area for enzyme binding that could further enhance the I_max/K_map value. The advantages of gold doping in PPD film was clearly noted in Table 2. Enzymatic efficiency (I_max/K_map) was enhanced to 19-fold for XO/Au–PPD/GCE as compared to XO directly absorbed on PPD film (XO/PPD/GCE), whereas it was increased only two times when AuNP was electrodeposited separately onto the PPD film (XO/Au/PPD/GCE). The linear range sensitivity (LRS) of the XO/Au–PPD increased to 55.5-fold due to the unique microspheric nanostructure of Au–PPD film that could enhance the active surface area for enzyme binding. The defects of functionalized MWCNT promoted fast electron transfer through the matrix. The electrodeposition of highly conducting fMWCNT on the Au–PPD film enhanced the sensitivity even to the nanomolar range, leading to lower LOD (limit of detection) for xanthine detection i.e. 12 nM. The covalent interaction of XO–NH₂ with the –COO⁻ group of fMWCNT could also stabilize the enzyme loading and hence effectively achieved a turnover number of 2.7 compared to that obtained by XO/Au–PPD/GCE. Though Nafion is commonly used to reduce the interference of ascorbic acid and uric acid in real samples a coating of Nafion reduced the enzyme affinity, denatured the enzyme and thus reduced current response.^15,28 The sensitivity of the Nafion coated sensor design (Naf/XO/fMWCNT/Au–PPD/GCE) was observed to be lower than that achieved without Nafion coating.

3.5 Characterization of modified electrode by FESEM, AFM, FTIR and XRD

Fig. 5a and b displays the FESEM image of successful electrodeposition of fMWCNT onto the Au–PPD surface and XO/fMWCNT/Au–PPD/GCE, respectively. The XO molecules were aggregated on the surface of fMWCNT through peptide bonds (Fig. 5b). Successful incorporation of XO enzyme on the fMWCNT/Au–PPD changed FTIR spectra (Fig. 5c) significantly showing characteristic bands of polypeptide. The IR spectra of XO/fMWCNT/Au–PPD showed two distinguishable absorption peaks at 1685 and 1570 cm⁻¹ for amide I and amide II vibrations of peptide bonds, whereas a broad absorption peak at 3480 cm⁻¹ peak was attributed for N–H stretching of amide or free amino groups present in the protein backbone. Fig. 5d displays the 3D AFM images of XO/fMWCNT/Au–PPD obtained in tapping phase in 1 μm scale. A mesh like porous structure was observed after immobilization of XO on the electro-deposited surface of fMWCNT (Fig. 5d). The “cavities” or “holes” on the fMWCNT-modified film might also be helpful for substrates or small inorganic ions in buffers to move into or out of the films, thus improving the electrocatalytic performances. The surface area of the PPD and Au–PPD image was calculated by NanoScope analysis software version 1.4. For Au-doped PPD film it was 76.4 μm², whereas the corresponding value for unmodified PPD film was 4.39 μm² (images shown in ESI, Fig. S3†). The huge change could be possible due to gold nanocluster deposition in the Au–PPD film (Fig. S3a and b†). The roughness and depth of XO/fMWCNT/Au–PPD/GCE was calculated to be 32.9 ± 1.3 nm and 21.6 ± 0.8 nm, respectively (Fig. S3c†).

Fig. 5 FESEM images of (a) fMWCNT/Au–PPD and (b) XO/fMWCNT/Au–PPD; (c) FTIR spectra of modified electrode fMWCNT/Au–PPD/GCE (i), XO/fMWCNT/Au–PPD/GCE (ii); (d) tapping mode AFM images of XO/fMWCNT/Au–PPD.

3.6 Electrochemical characterization of the modified electrode

Electrochemical impedance spectra (EIS) were obtained at each stage of modification of the electrode to analyze the change of electrical properties at the interface due to characteristic changes in the interface at frequencies between 0.01 Hz and 1 MHz at an applied potential of 0.28 V (anodic peak potential for K₃Fe(CN)₆/K₄Fe(CN)₆ with modified electrode). Nyquist plots for different layers of modification in Fig. 6A shows changes of charge transfer resistance (R_ct) value. The R_ct and parallel double layer capacitance (C_dl) varied at higher frequency range that represented a semicircle structure of the spectra while at lower applied frequencies, the straight line represented the Warburg-diffusion impedance (Z_w), indicating a diffusion control charge transfer process. The parameters for surface variation were obtained by fitting the Randles equivalent circuit at the higher frequency as well as lower frequency range. R_s was related to the uncompensated solution resistance which remained almost constant. The electron transfer resistance (R_ct) value for the bare GCE, PPD/GCE, Au–PPD/GCE, fMWCNT/Au–PPD/GCE and XO/fMWCNT/Au–PPD/GCE, electrodes were 3.481, 2.298, 0.0851, 64.798 and 4.645 kΩ, respectively. The electron transfer resistance decreased due to polymerization of OPD that forms a thin conducting polymer matrix (PPD/GCE). The novel Au–PPD film provided 27 times lower R_ct value than PPD/GCE, suggesting that the thin film of Au–PPD had higher conductivity than a simple polymeric film of PPD due to the presence of Au nanoclusters in Au–PPD film. The R_ct value increased after electrodeposition of fMWCNT owing to the electrostatic repulsion of negative charges of electrode surface and anions in the solution [Fe(CN)₆³⁻/Fe(CN)₆⁴]. Further reduction of R_ct value could explain the successful immobilization of enzyme onto the fMWCNT/Au–PPD/GCE surface.

Fig. 6 (A) Nyquist, (B) Bode phase and (C) Bode amplitude plots of electrochemical impedance spectra of each step of modifications of XO/fMWCNT/Au–PPD/GCE; (A, inset) shows the equivalent Randles circuit for fitting the circuit of the modified enzyme electrode. The curves represent the results obtained for bare GCE (curve a), PPD/GCE (curve b), Au–PPD/GCE (curve c), fMWCNT/Au–PPD/GCE (curve d) and XO/fMWCNT/Au–PPD/GCE (curve e), in 5 mM of [K₃Fe(CN)₆/K₄Fe(CN)₆] within the frequency range of 1 MHz to 0.01 Hz at a constant potential of 0.28 V with modulation amplitude 0.01 mV using Ag/AgCl as standard RE and Pt wire as CE.

Fig. 6B depicts Bode phase angles and Fig. 6C represents Bode amplitude plots for bare GCE, PPD/GCE, Au–PPD/GCE, fMWCNT/Au–PPD/GCE, XO/fMWCNT/Au–PPD/GCE. In Fig. 5b, the PPD/GCE and Au–PPD/GCE showed lower phase angles such as 43.17 and 31.2° compared to bare GCE (56.13°) indicating increased charge transfer rate at the electrode surface due to the presence of the gold doped polymer composite film. The fMWCNT/Au–PPD/GCE showed a phase angle of 75.68° at low frequency (1.3 Hz), indicating ideal capacitive behaviour due to electrodeposition of negatively charged fMWCNT; which further moved to 56.7° at 56.7 Hz after enzyme immobilization. The change of phase angle pattern suggests facilitated charge transfer reaction on XO/fMWCNT/Au–PPD/GCE. In the frequency range from 10⁵ to 10⁴ Hz, the phase angles approached zero and the |Z| value was almost constant, indicating the same solution resistance (R_s) for all the modification stages of GCE; the R_ct value also decreased with modification of electrodes, thus facilitating electron transfer.¹¹ In order to calculate the electron transfer rate constant k₀ (data shown in Table 3), eqn (3) was applied at different steps of electrode modification.

R_ct = RT/(nF)²Ak₀C; (3)

where, C is the molar concentration [Fe(CN)₆³⁻/Fe(CN)₆⁴] in solution. The calculated k₀ value from the equivalent Randles circuit model was smaller than any previously reported sensing system.^27,30

Table 3 Change of impedance values (R_ct and k₀) for each step of modification determined from equivalent circuit models

Modified electrode	Rct (kΩ)	k0 (cm s^−1)	Phase angle (°)@frequency
Bare GCE	3.481	2.16 × 10^−7	56.133@543 Hz
PPD/GCE	2.298	3.28 × 10^−7	43.175@175.8 Hz
Au–PPD/GCE	0.0851	8.85 × 10^−6	31.2@121 Hz
fMWCNT/Au–PPD/GCE	64.798	1.16 × 10^−8	75.68@1.3 Hz
XO/fMWCNT/Au–PPD/GCE	4.645	1.62 × 10^−7	56.7@50.65 Hz

---

The active surface area of the modified fMWCNT/Au–PPD/GCE electrode was determined by performing CV in the range of −0.5 to 0.6 V for 1 mM K₃Fe(CN)₆ in 0.1 M KCl and the slope of the I_p versus ν curve was obtained by varying scan rate from 10 to 200 mV. The active surface area was calculated following Randles–Sevcik eqn (4) for a reversible process.¹¹

I_p + (2.69 × 10⁵)n^3/2A_cD_r^1/2ν^1/2C (4)

where I_p refers to the anodic peak current (A), n the number of electrons transferred, A the active surface area (cm²), D_r diffusion coefficient (cm² s⁻¹), ν the scan rate (V s⁻¹) and C denotes the concentration of the analyte (mol cm⁻³), i.e. 1 mM K₃Fe(CN)₆. For 1 mM K₃Fe(CN)₆ in 0.1 M KCl electrolyte, n = 1, C = 1 and D_r = 7.6 × 10⁻⁶ cm² s⁻¹.⁴² The linear slope of 3 × 10⁻⁵A/(V s⁻¹), the active surface area is 0.403 cm².

The average surface coverage (Γ*) of XO enzyme on the fMWCNT/Au–PPD modified glassy carbon electrode was estimated according to the following eqn (5) of the Brown–Anson model (Dhyani, Ali, Pandey, Malhotra and Sen, 2012).²⁷

I_p (μA) = n²F²Γ*νA/4RT (5)

where, n denotes number of electrons transferred; A area of electrode surface; F the Faraday constant; R universal gas constant; T temperature in Kelvin and ν scan rate. The value of Γ* was calculated from the slope of the I_p versus ν plot and the average surface coverage by XO molecules was about 4.37 × 10⁻⁵ M cm⁻² under the saturated adsorption conditions, indicating a sub-monolayer of enzyme immobilized on the modified electrode surface, that was very large compared to the previous report i.e. 7.62 × 10⁻¹² M cm⁻² of XO immobilized on LAPONITE® nanoparticles modified GEC electrode.²⁷

3.7 Electrochemical studies of XO/fMWCNT/Au–PPD/GCE

3.7.1 Influence of pH. The redox potential of an electrochemical reaction depends on the solution pH that indicates the participation of protons in the redox reaction. The influence of buffer pH was displayed in Fig. 7a in the pH range of 5.0–9.0 in 50 mM potassium phosphate buffer. CV was performed with 5.0 × 10⁻⁵ M of xanthine at a scan rate of 100 mV s⁻¹. The redox potential (E_p) shifted to negative direction i.e. less positive value with increasing the pH from 5 to 9. The linear regression equation E°′ = 1081.7 − 53.33pH was obtained with correlation coefficient of 0.9946 (inset of Fig. 6a). Since, the value of slope was −53.33 mV per pH i.e. approximately close to the theoretical value of −59 mV per pH (Nernstian case); suggested that the number of e⁻ transferred is equal to the number of protons coupled with the redox process on the modified electrode surface. The buffer pH = 7 was chosen for further electrochemical analysis owing to the highest peak current for enzymatic oxidation. Since pK_a of xanthine is 7.7 and 11.9, it is slightly positively charged at pH 7 whereas XO enzyme (pK_a = 4.2) was negatively charged at pH of electrolyte 7.0. The active site of XO having Gly–COOH, was responsible for nucleophilic attraction on C8 of xanthine molecule. A stable enzyme–substrate complex was formed by reduction of the fully oxidized form of XO–Mo(VI). A water molecule was attributed to oxidize xanthine into the enol tautomer of uric acid, that further reforms to the keto tautomer of uric acid. The reduced XO–Mo(IV) form of enzyme was further oxidized by two-electron transfer in the electrochemical reaction. As there were no alterations in the peak potentials with respect to the pK_a values it could be concluded that the protonation reaction was not a rate determining step in the electrochemical oxidation processes. Thus like other FAD-containing flavoenzymes, transfer of two protons was accompanied with two electrons during the electrochemical reaction onto the XO/fMWCNT/Au–PPD/GCE electrode surface and the electrochemical reaction could be described as below:

Xanthine + XO–FADH₂ ↔ uric acid + XO–FAD + 2e⁻ + 2H⁺ (6)

2e⁻ → working electrode (7)

Fig. 7 Cyclic voltammograms of modified electrodes in the presence of 5.0 × 10⁻⁵ M of xanthine in 50 mM phosphate buffer; (A) at different pH range of 5.0–9.0 at a scan rate of 100 mV s⁻¹ (inset: plot of E°′ vs. pH); (B) at different scan rates of 10, 25, 50, 75, 100, 125, 150, 175, 200 mV s⁻¹; inset: (i) dependence of the logarithm peak current on logarithm of scan rate (R² = 0.995); another (ii) calibration plot for I_p versus ν^1/2. Conditions: potential range: 0.3–1.0 V, scan rate: 0.1 V s⁻¹ Electrodes – WE: XO/fMWCNT/Au–PPD/GCE, RE: Ag/AgCl, CE: platinum wire.

3.7.2 Influence of scan rate. The kinetics of the electrode reactions were investigated by studying the effects of scan rate (ν) on the anodic peak currents (I_p) for xanthine at XO/fMWCNT/Au–PPD/GCE (Fig. 7b). The peak current of 5.0 × 10⁻⁵ M xanthine increased linearly between 10 and 200 mV s⁻¹ and this indicated a typical diffusion controlled process and the equation could be expressed as follows: I_p = 6.5817ν^1/2 − 0.0295; R² = 0.9842. In addition, there was a linear relationship between log I_p and log ν with slope of 0.52 that was close to the theoretical value of 0.5 for a diffusion controlled process^40,41 given by log I_p = 0.5203 log ν + 0.832; R² = 0.989.

The redox peak potential shifted linearly towards positive value with increasing scan rates, corresponding to the following relation: E_p = 0.0428 log ν + 0.7658 with R² value 0.9614. According to Laviron^40,41 for irreversible reaction process, E_p is calculated by the following equation:

where α is the transfer coefficient, k° is the standard heterogeneous rate constant of the reaction, n is the number of electrons transferred through the electrode, ν is the scan rate and E°′ is the formal redox potential. Other symbols are universal standard R = 8.314 J K⁻¹, F = 96 480 C mol⁻¹, T = 298 K. The value of αn calculated as 1.382 from the slope of E_p vs. log ν i.e., 0.0482. The α value was calculated from Bard and Faulkner (2004)⁶ eqn (9):where E_p/2 is defined as the potential (mV) where the current is the half of the peak current. For our system α is 0.596. Thus the number of electrons transferred (n) during the oxidation of xanthine was calculated as 2.32 (∼2).

3.8 Evaluation of performance parameters of the proposed xanthine biosensor

Fig. 8 displays the differential pulse voltammetric (DPV) response for oxidation of xanthine on the modified electrode for a wide range of xanthine concentrations of 0.01–300 μM. DPV was conducted in 50 mM potassium phosphate buffer, pH 7 with 50 mV pulse amplitude and step potential of 10 mV in the range of 0.4–1.0 V. The oxidation potential of xanthine was shifted towards higher potential with addition of increasing concentration of xanthine.¹⁷ The xanthine standard curve was linear for 0.01–0.250 μM given by the equation: I_p (μA) = 0.9979[xan] + 0.0114 (R² = 0.99); 0.1 to 35 μM with I_p (μA) = 0.2559[xan] + 0.0857 (r² = 0.998) and for 35 to 200 μM with I_p (μA) = 0.0999[xan] + 6.392 (R² = 0.982). The limit of detection (LOD = [(3 × SD of blank)/slope]) and limit of quantification (LOQ = [(10 × SD of blank)/slope]) were calculated as 12 nM and 40 nM, respectively (for lowest range of calibration curve) based on S/N = 3. The lowest detection limit of the new sensing system compared very well with reported xanthine sensors i.e. modified glassy carbon electrode coated with graphitized mesoporous carbon (0.388 μM),¹⁰ XO immobilized on LAPONITE® NP modified electrode (0.01 μM),²⁷ P(GMA-co-VFc)/MWCNT/XO (0.12 μM),²⁴ Naf/XO–CD/pAuNP/SWNT/GCE (0.04 μM),²⁸ DNA–polyaniline (PAn) complex Langmuir–Blodgett film (30 nM)¹⁷ (see note at bottom of Table 1 for abbreviations).

Fig. 8 (A) Differential pulse voltammetric response of different concentrations of xanthine such as 0.01, 0.1, 0.5, 1, 2.5, 5, 10, 25, 50, 75, 100, 150, 200, 300 in 50 mM phosphate buffer, pH 7; (B) the current response of xanthine in the range of 1–300 μM, inset: calibration curve of the linear range of 0.01–35 μM and 0.5–35 μM; conditions: potential range: 0.3–1.0, scan rate: 0.1 V s⁻¹ in 50 mM phosphate buffer using a three-electrode system of WE: XO/fMWCNT/Au–PPD/GCE, RE: Ag/AgCl, CE: platinum wire.

The analytical performance of the proposed sensor was better than any of the previously reported xanthine sensors with respect to the stability. The long-term storage stabilities of the proposed modified XO electrode were tested every week for six months. After each experiment, the electrode was washed with reaction buffer and stored in phosphate buffer at 4 °C. The stability was calculated using the formula:

% stability = 100I_n/I_o (10)

where I_o is the obtained current in first day and I_n is the obtained current on the n^th day. It was observed that the modified electrode could be used comfortably up to 180 times. The electrode retained more than 91% activity after storing it for 130 days (Fig. S4†). The higher stability of XO/fMWCNT/Au–PPD/GCE than previous reports^{4,10,17,23–31} might be due to the strong covalent bonding of carboxylated MWCNT and free –NH₂ of the enzyme that prevented leakage of enzyme from the electrode surface.

The biosensor showed highest sensitivity of 14.03 μA μM⁻¹ cm⁻² towards xanthine detection. The reproducibility of the sensory system was investigated by modification of four glassy carbon electrodes and observing their anodic response towards 50 μM xanthine by five repeat measurements. The peak current was determined five times with the same electrode (within batch) and four different electrodes (between batches). The response for same fish extract showed almost consistent results of xanthine content within and between batches i.e. coefficient of variation (CV) were <1.93 and <3.08% (Table S1†). The repeatability of the biosensor was analyzed at 50 μM xanthine and the relative standard deviations for six determinations were 2.1% and 2.8%, respectively.

3.9 Selectivity of xanthine biosensor

The selectivity of the immobilized enzyme electrode was evaluated by its ability to protect the sensor from other electroactive interfering compounds that might be present in the real samples. The specificity of the biosensor was monitored in presence of some interfering chemicals (1 mM) such as hypoxanthine, inosine, inosine monophosphate, uric acid (UA), ascorbic acid (AA), oxalic acid, L-cysteine, glucose, sucrose, theobromine and theophylline. Since, co-electro-oxidation of potential interferences such as AA and UA is one of the major problems in the amperometric detection of analytes, DPV was performed owing to its ability to give distinct oxidation potentials for each substance. The interference of the above reagents (1 mM) was investigated by monitoring the DPV response in the range of 0.3–1.0 V. It could be seen from Fig. 9 that there was no interference for the common interfering compounds near the oxidation potential of xanthine, i.e. 0.625–0.725 V. Uric acid depicted an oxidation potential at 0.29 V and ascorbic acid oxidized at 0.05 V; both these oxidation potentials occur far away from the target potential of interest, i.e. nearly 0.65 V. Most importantly, the enzyme electrode did not show any response towards the common electroactive biologically important molecules such as ascorbic acid (AA) (1 mM) and uric acid (UA) (0.5 mM) in the potential scan range of 0.4–1.0 V. Since the active form of anionic XO (pK_a = 4.2 ± 0.1) was immobilized on carboxylated MWCNT at the electrode surface, the modified XO/fMWCNT/Au–PPD electrode was negatively charged and could repel negatively charged interfering substance (such as AA, UA etc.) present in the real sample. Thus the response for electroactive anions present in biological media declined to a great extent.

Fig. 9 DPV response for interfering substances (1 mM) present in biological media. DPV response of some interfering substances. Conditions: potential range: 0.3–1.0 V, scan rate: 0.1 V s⁻¹ in 50 mM phosphate buffer using WE: XO/fMWCNT/Au–PPD/GCE, RE: Ag/AgCl, CE: platinum wire.

It is also known that PPD depicts good permeability and permselectivity towards AA. The apparent permeability of AA for XO/fMWCNT/Au–PPD was calculated as 2.34% using the following eqn (11).¹⁶

where, I_AA (1 mM) at PEC/GCE was determined as the nanoampere value at the fixed potential of 0.65 V for modified polymer enzyme composite (PEC).

3.10 Real sample analysis

Freshness of fish could be monitored by determination of xanthine during its storage. The level of xanthine in blood serum and urine are of importance in clinical diagnosis. Hence the proposed biosensor was tested with human serum and urine samples as well as fish extract. The accuracy of the proposed sensor was investigated by standard addition method to determine xanthine content in real samples. In this method, additions of standard xanthine solution were made several times to the each sample. The results were compared in Table 4 with standard HPLC based method. The xanthine values obtained by the present biosensor (y) matched with standard HPLC data with good correlation (y = 0.9984x + 0.0124, r = 0.998, significant at 1% level), showing high accuracy of the proposed method. Table 4 represents a good agreement between biosensor and HPLC methods, indicating that the present method could be used for real sample analysis. Spoilage of fish could be monitored by registering the concentration of xanthine with storage time. The spoilage rate of fish was determined by storing it in 0 °C, −20 °C temperature and monitoring the xanthine content in 30 μl of fish extract. The response increased abruptly with increasing storage temperature (Table 5).

Table 4 Determination of xanthine in real samples using XO/fMWCNT/Au–PPD/GCE^a

Sample	Added (μM)	Found (μM) by biosensor	Recovery (%)	Original (μM) by HPLC	Relative error (%)	RSD (%)
a All the experimental data show P value (n = 5) < 0.05.
Fish sample 1 (Leabeo rohita)	0 (30 μl)	4.8	101.05	4.86	1.235	1.597
	5	9.75	99.08	9.84	0.915	0.785
	10	14.82	98.66	15.021	1.338	1.617
Fish sample 2 (Lates calcarifer)	0 (30 μl)	3.609	98.79	3.653	1.204	1.269
	5	8.458	98.44	8.592	1.560	0.993
	10	14.019	101.40	13.825	1.403	1.561
Blood serum	0 (50 μl)	2.514	98.67	2.514	1.334	3.813
	5	7.805	102.68	7.960	1.947	1.667
	10	12.537	99.82	12.56	0.183	1.518
Urine	0 (50 μl)	0.301	103.44	0.291	3.436	3.32
	5	5.230	97.74	5.351	2.261	1.68
	10	10.463	101.85	10.273	1.849	0.893

---

Table 5 Determination of xanthine concentration in fish extracts (Leabeo rohita) solution stored at different temperature

Storage time (days)	Storage temperature
	−20 °C	0 °C
0	3.034 ± 0.23	3.109 ± 0.34
1	3.612 ± 0.27	5.012 ± 0.43
3	4.032 ± 0.31	9.245 ± 0.61
5	6.258 ± 0.53	15.179 ± 0.76
7	7.676 ± 0.68	30.810 ± 1.03

---

4. Conclusions

A novel interference free xanthine biosensor was successfully fabricated. The most important achievement of this study was rapid formation of an ultrathin Au decorated PPD film (thickness-178.2 ± 12 nm) having good permselectivity and permeability by one-step electrodeposition. EIS study depicted that the nanospheric structure of gold doped PPD film enhanced the conductivity when compared with only PPD film. The large number of hydroxyl and carboxyl groups in fMWCNT provided increased the active surface area for loading higher amount of xanthine oxidase enzyme and higher surface energy that led to the favourable conformational change of protein for direct electron transfer between active site of enzyme to the underlying electrode. The covalent interaction of the enzyme enhanced enzyme stability on the modified electrode and retained its catalytic activity up to 6 months. The sensor depicted good catalytic activity (I_m/k_m) and higher substrate affinity (k_m = 0.123 mM). The oxidation of xanthine onto the modified electrode surface was found to be an irreversible and diffusion controlled process. The newly designed XO/fMWCNT/Au–PPD/GCE sensor exhibited good electrode characteristics including high sensitivity (14.03 μA μM⁻¹ cm⁻²) with low LOD of 12 nM, rapid response (5 s). A wide range of linear calibration curve was obtained for 0.01–200 μM of xanthine (R² = 0.99). The coefficients of variation for reproducibility were found to be only 1.9 and 3.1% within and between the assays. It also exhibited good repeatability of 2.1%. Furthermore, the XO/fMWCNT/Au–PPD/GCE showed excellent selectivity towards xanthine in the presence of interfering agents such as ascorbic acid, glucose and uric acid. The proposed sensor also showed good correlation (r = 0.998) with standard HPLC data having high recovery rate 97–101%. Hence, the sensor could be useful for rapid analysis of xanthine in real samples such as blood, urine and tissue. In addition, this novel immobilization matrix could be extended to other enzymes for fabrication of efficient biosensors.

Acknowledgements

The authors are very much thankful to CRNN, University of Calcutta to provide the facility of SEM, FESEM, TEM and DBT-IPLS for AFM facility.

References

1. G. Volpe and M. Mascini, Talanta, 1996, 43, 283–289.
2. V. Venugopal, Biosens. Bioelectron., 2002, 17, 147–157.
3. N. Cooper, R. Khosravan, C. Erdmann, J. Fiene and J. W. Lee, J. Chromatogr. B: Anal. Technol. Biomed. Life Sci., 2006, 837, 1–10.
4. R. Devi, M. Thakur and C. S. Pundir, Biosens. Bioelectron., 2011, 26, 3420–3426.
5. L. D. Mello and L. T. Kubota, Food Chem., 2002, 77, 237–256.
6. A. J. Bard and L. R. Faulkner, Electrochemical Methods Fundamentals and Applications, Wiley, New York, 2nd edn, 2004.
7. P. Q. Pandey, S. K. Arya, Z. Matharu, S. P. Singh, M. Datta and B. D. Malhotra, J. Appl. Polym. Sci., 2008, 110, 988–994.
8. Rajesha, T. Ahuja and D. Kumar, Sens. Actuators, B, 2009, 136, 275–286.
9. W. Fenga and P. Ji, Biotechnol. Adv., 2011, 29, 889–895.
10. R. Thangaraj and A. S. Kumar, Anal. Methods, 2012, 4, 2162–2171.
11. T. Ghosh, P. Sarkar and A. P. F. Turner, Bioelectrochemistry, 2015, 102, 1–9.
12. M. Ates, Mater. Sci. Eng., C, 2013, 33, 1853–1859.
13. L. Xia, Z. Wei and M. Wan, J. Colloid Interface Sci., 2010, 341, 1–11.
14. P. Gajendran and R. Saraswathi, J. Phys. Chem. C, 2007, 111, 11320–11328.
15. E. Turkmen, S. Z. Bas, H. Gulce and S. Yildiz, Electrochim. Acta, 2014, 123, 93–102.
16. A. Khan and S. A. Ghani, Biosens. Bioelectron., 2012, 31, 433–438.
17. L. Zou, Y. Li, S. Cao and B. Ye, Talanta, 2014, 129, 346–351.
18. S. A. Ansari and Q. Husain, Biotechnol. Adv., 2012, 30, 512–523.
19. K. Ding, H. Jia, S. Wei and Z. Guo, Ind. Eng. Chem. Res., 2011, 50, 7077–7082.
20. R. Devi, S. Yadav and C. S. Pundir, Colloids Surf., A, 2012, 394, 38–45.
21. P. Xu, X. Han, B. Zhang, Y. Du and H.-L. Wan, Chem. Soc. Rev., 2014, 43, 1349.
22. A. B. Moghaddam, T. Nazari, J. Badraghi and M. Kazemzad, Int. J. Electrochem. Sci., 2009, 4, 247–257.
23. R. Devi, S. Yadav and C. S. Pundir, Analyst, 2012, 137, 754–759.
24. M. Dervisevic, E. Custiuc, E. Çevik and M. Şenel, Food Chem., 2015, 181, 277–283.
25. Ü. A. Kirgöz, S. Timur, J. Wang and A. Telefoncu, Electrochem. Commun., 2004, 6, 913–916.
26. M. Çubukçu, S. Timur and Ü. Anik, Talanta, 2007, 74, 434–439.
27. D. Shan, Y.-N. Wang, H.-G. Xue, S. Cosnier and S.-N. Ding, Biosens. Bioelectron., 2009, 24, 3556–3561.
28. R. Villalonga, P. Díez, M. Eguílaz, P. Martínez and J. M. Pingarrón, ACS Appl. Mater. Interfaces, 2012, 4, 4312–4319.
29. L. Zhang, J. Lei, J. Zhang, L. Ding and H. Ju, Analyst, 2012, 137, 3126–3131.
30. R. Devi, S. Yadav, S. Nehra and C. S. Pundir, J. Food Eng., 2013, 115, 207–214.
31. S. Sadeghi, E. Fooladi and M. Malekaneh, Anal. Biochem., 2014, 464, 51–59.
32. S. Goyanesa, G. R. Rubiolo, A. Salazard, A. Jimenod, M. A. Corcuerad and I. Mondragon, Diamond Relat. Mater., 2007, 16, 412–417.
33. V. Datsyuk, M. Kalyva, K. Papagelis, J. Parthenios, D. Tasis, A. Siokou, I. Kallitsisand and C. Galiotis, Carbon, 2008, 46, 833–840.
34. J. Zhang, C. Yang, X. Wang and X. Yang, Analyst, 2012, 137, 3286–3292.
35. H. Chi, B. Liu, G. Guan, Z. Zhang and M. Y. Han, Analyst, 2010, 135, 1070–1075.
36. S. Bilal and R. Holze, Electrochim. Acta, 2007, 52, 5346–5356.
37. U. Olgun and M. Gülfen, React. Funct. Polym., 2014, 77, 23–29.
38. L. Zhang, W. Yuan and Y. Yan, Electrochim. Acta, 2013, 113, 218–228.
39. T. Li, C. Yuan, Y. Zhao, Q. Chen, M. Wei and Y. Wang, J. Macromol. Sci., Part A: Pure Appl. Chem., 2013, 50, 330–333.
40. C. Lanzellotto, G. Favero, M. L. Antonelli, C. Tortolini, S. Cannistraro, E. Coppari and F. Mazzei, Biosens. Bioelectron., 2014, 55, 430–437.
41. E. Laviron, J. Electroanal. Chem., 1979, 101, 19.
42. A. L. Eckermann, D. J. Feld, J. A. Shaw and T. J. Meade, Coord. Chem. Rev., 2010, 254, 1769–1802.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra18889j

---