Acetylene-sourced CVD-synthesised catalytically active graphene for electrochemical biosensing

Acetylene-sourced CVD-synthesised catalytically active graphene for electrochemical biosensing

Author’s Accepted Manuscript Acetylene-sourced CVD-synthesised catalytically active graphene for electrochemical biosensing Adeniyi Olugbenga Osikoya,...

2MB Sizes 1 Downloads 67 Views

Author’s Accepted Manuscript Acetylene-sourced CVD-synthesised catalytically active graphene for electrochemical biosensing Adeniyi Olugbenga Osikoya, Onur Parlak, N. Arul Murugan, Ezekiel Dixon Dikio, Harry Moloto, Lokman Uzun, Anthony PF Turner, Ashutosh Tiwari www.elsevier.com/locate/bios

PII: DOI: Reference:

S0956-5663(16)30250-0 http://dx.doi.org/10.1016/j.bios.2016.03.063 BIOS8574

To appear in: Biosensors and Bioelectronic Received date: 22 January 2016 Revised date: 22 March 2016 Accepted date: 23 March 2016 Cite this article as: Adeniyi Olugbenga Osikoya, Onur Parlak, N. Arul Murugan, Ezekiel Dixon Dikio, Harry Moloto, Lokman Uzun, Anthony PF Turner and Ashutosh Tiwari, Acetylene-sourced CVD-synthesised catalytically active graphene for electrochemical biosensing, Biosensors and Bioelectronic, http://dx.doi.org/10.1016/j.bios.2016.03.063 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Acetylene-sourced CVD-synthesised catalytically graphene for electrochemical biosensing Adeniyi Olugbenga Osikoya1,

2†

active

, Onur Parlak1†, N. Arul Murugan3, Ezekiel Dixon Dikio2,

Harry Moloto2, Lokman Uzun1, 4, Anthony PF Turner1, Ashutosh Tiwari1, 5,6*

1

Biosensors and Bioelectronics Centre, IFM, Linköping University, 58183, Linköping,

Sweden 2

Applied Chemistry and Nanoscience Laboratory, Department of Chemistry, Vaal University

of Technology, Private Bag X021, Vanderbijlpark, South Africa 3

Virtual Laboratory for Molecular Probes, Division of Theoretical Chemistry and Biology,

School of Biotechnology, Royal Institute of Technology, S-106 91 Stockholm, Sweden 4

Department of Chemistry, Hacettepe University, Ankara, Turkey

5

Tekidag AB, UCS, Mjärdevi Science Park, Teknikringen 4A, SE 583 30 Linköping, Sweden

6

Vinoba Bhave Research Institute, Sirsa Road, Saidabad, Allahabad 221508, India



Authors contributed equally.

*

Corresponding author. Tel.: +46 13 2823 95; Fax: +46 13 1375 68; E-mail: [email protected]

ABSTRACT In this study, we have demonstrated the use of chemical vapour deposition (CVD) growngraphene to develop a highly-ordered graphene-enzyme electrode for electrochemical biosensing. The graphene sheets were deposited on 1.00 mm thick copper sheet at 850 oC using acetylene (C2H2) as carbon source in an argon (Ar) and nitrogen (N2) atmosphere. An anionic surfactant was used to increase wettability and hydrophilicity of graphene; thereby facilitating the assembly of biomolecules on the electrode surface. Meanwhile, the theoretical

1

calculations confirmed the successful modification of hydrophobic nature of graphene through the anionic surface assembly, which allowed high-ordered immobilisation of glucose oxidase (GOx) on the graphene. The electrochemical sensing activities of the grapheneelectrode was explored as a model for bioelectrocatalysis. The bioelectrode exhibited a linear response to glucose concentration ranging from 0.2 to 9.8 mM, with sensitivity of 0.087 µA/µM/cm2 and a detection limit of 0.12 µM (S/N=3). This work sets the stage for the use of acetylene-sourced CVD-grown graphene as a fundamental building block in the fabrication of electrochemical biosensors and other bioelectronic devices. Keywords: CVD-grown graphene, bioelectronics, theoretical calculation, surfactant modification, 2D-materials.

Introduction Super-thin, flexible and smart bioelectronic devices for applications in biosensors, biofuel cells and bioreactors have become a core aim in bioelectronic design and fabrication in recent years (Nam et al., 2013; Parlak and Turner, 2016). Due to the outstanding electromechanical properties of graphene, which is capable of more than 20% elastic deformation without the perturbation of its electrical properties, graphene is rapidly becoming a viable alternative for low-cost bioelectronic device production (Chun et al., 2014; Parlak et al., 2014). Recent studies on the electrocatalytic activity of graphene have focused on the use of graphene oxide (GO) or reduced graphene oxide (rGO) (Ashaduzzaman et al., 2015; Kang et al., 2010; Shao et al., 2010); however the insertion of oxygenated functional groups to improve the hydrophilicity of the material causes a considerable reduction in the electronic transport properties of the graphene (Hu and Su, 2013). On the other hand, Puri et al. (2014) reported that electrochemically reduced graphene oxide (rGO) facilitates higher biomolecule loading capacity together with better electrical conductivity, which is one of the most desirable

2

properties to promote easy and efficient electron transfer to achieve high sensitivity of the bioelectrodes (Puri et al., 2014). The outstanding electromechanical properties of graphene are mainly influenced by the sp2 hybridisation of the carbon atoms in the hexagonal lattice of the graphene sheet, which involves the mixing together of one S orbital with two P orbitals (2S, 2Px, and 2Py). This hybridisation leads to the formation of an σ bond and a trigonal planer structure with the C-CC bond angle at 120o (Batzill, 2012; Geim and Novoselov, 2007; Ivanovskii, 2012; Singh et al., 2009). Overlapping of the lone-pair electrons present in the un-hybridised 2Pz orbitals within three neighbouring carbon atoms generates two different bands, valence band with filled π orbitals and conduction band with empty π* orbitals (Geim and Novoselov, 2007), thus endowing graphene with its extremely high electron transport mobility (Shahil and Balandin, 2012; Wang et al., 2011). Although monolayer graphene has been reported to possess zero band gaps, it is unsuitable for the fabrication of electronic devices, especially when digital on/off control is desired. Graphene with a few layers, however, has a quadratic low energy band structure and thus possess different scattering properties from those of monolayer graphene. This feature allows the band gap to be easily tuned for switchable bioelectronics through the synthesis of few-layer and multi-layer graphene (Avouris, 2010). Current trends in graphene research involves the use of graphene-based hybrid materials in the design of bioreactors, biosensors and bioelectrodes (Parlak et al., 2013). Parlak et al. (2013) used a graphene-enzyme-nanoparticle based hybrid system in the design of bioelectrode for electrochemical biosensing of hydrogen peroxide and cholesterol, and reported that the bioelectrodes exhibited a high sensitivity of 3.14 µA/µM/cm2. Zhang et al. (2011) also reported the fabrication of immobilised horseradish peroxidase on graphene oxide-multi-walled carbon nanotube (MWCNT) bioelectrodes. It was found that the HRP/GO-MWCNT/GC electrode exhibited excellent electrocatalytic behaviour for the

3

reduction of H2O2 and NaNO2, respectively (Zhang et al., 2011). In most cases, the graphene synthesis involved the use of strong oxidisers and reducers, which causes the formation of undesirable functional groups such as epoxides, alcohols, carbonyl and carboxylic acid groups in the graphene oxide (GO), which are not completely eliminated even in rGO forms. The presence of these functional groups and the consequent numerous defects, cause a considerable disruption in the sp2 hybridisation of carbon atoms in the graphene network, which results in poor conductivity in comparison to other graphene forms (Huang et al., 2011). Although various different synthesis methods of graphene from various sources have been reported, chemical vapour deposition (CVD) is one of the most promising approaches in the design and fabrication of bioelectronic devices (Kibena et al., 2014). In this study, we developed a CVD graphene sheet grown from acetylene as a carbon source, without any hash chemical treatment and incorporated other materials (usually metals) to improve conductivity. In this way, we grew CVD graphene sheets with a larger available surface area, due to the absence of the strong oxidisers and reducers used in the preparation of graphene by the chemical method. In the literature, the charge carrier mobility in CVD grown graphene has been reported to be in excess of 4000 cm2/Vs, while the sheet resistance was reported to be as low as about 125 Ω/sq (Castro Neto et al., 2009; Cermak et al., 2014; Kumari et al., 2014), which makes this form of graphene an ideal material for electrochemical reactions. In addition, the larger strong and flexible surface area is an excellent ultimate platform for enzymes during the fabrication of very stable, durable and ultra-sensitive bioelectrodes for electrochemical (bio)sensing (Huang et al., 2011). There are relatively few reported studies on the electrochemical behaviour of CVD graphene (Valota et al., 2011); Brownson et al., 2014; Kibena et al., 2014). Kibena et al. (2014) developed a CVD-grown graphene electrode, using methane as a carbon source, and reported the electrochemical behaviour of highly ordered pyrolytic graphite (HOPG) for oxygen reduction with good electrocatalytic activity

4

(Kibena et al., 2014). Browson et al. (2014) reported that experimental observation on the electrochemical properties of CVD-grown graphene shows that the heterogeneous electron transfer (HET) rates in multi-layer graphene are about 2-8 times faster than in single-layer graphene and, of course, much more fast than that in HOPG (Brownson et al., 2014). Valota et al. (2011) reported that the HET is quite dependent on the density of the edge plane and defect sites on the graphene, therefore the HET is higher in multi-layer graphene than that in single-layer graphene (Valota et al., 2011). They also compared the electrochemical behaviour of mono-layer and multi-layer graphene and confirmed that both mono-layer and multi-layer graphene microelectrodes showed quasi-reversible behaviour during voltammetry measurements using potassium ferricyanide (Valota et al., 2011). Another problem with pristine graphene is it poor wettability due to its extremely hydrophobic nature. However, the use of anionic surfactants to overcome this challenge is one of easy physical surface modification. Surfactants are used to lower surface tension in aqueous media, increasing wettability and forming emulsions, while locating between aqueous and hydrophobic solid phases by virtue of their lipophilic and hydrophilic ends (Kim et al., 2010; Li et al., 2010; Liu et al., 2012). Similarly, we also introduced an anionic surfactant, sodium dodecyl benzene sulfonate (C18H29NaO3S, SDBS), for the surface modification of CVD-grown graphene to increase its wettability and enhance immobilisation of the hydrophilic glucose oxidase (GOx) enzyme on the surface. Based on our previous experience with graphene-based hybrid bioelectronics, we here focused our attentions on smart combination of CVD graphene, surfactant and GOx, while monitoring/assessing surface modification through both theoretical and practical approaches and evaluating electrocatalytic performance of the hybrid system developed. In this study, we developed GOx-based bioelectrodes using the CVD-grown graphene as a carbon based 2Dmaterial to enhance the electrochemical and electrocatalytic features of the bioelectrodes.

5

Herein, we designed three main experimental steps: (i) synthesis and characterisation of graphene via a CVD method; (ii) theoretical calculations for the interaction between surfactant, enzyme and CVD graphene; and (iii) determination of the actual electrocatalytic behaviour of the bioelectrodes developed. Graphene, synthesised via the CVD method from acetylene as a carbon source, was characterised by X-ray diffraction, Raman spectroscopy, scanning electron microscopy (SEM), zeta potential and contact angle measurements. Theoretical calculations were performed according to density functional theory/molecular dynamics using Gaussian09 software. Electrochemical characterisation and electrocatalytic activity of the bioelectrodes developed were investigated through cyclic voltammetry and amperometry.

Experimental Chemicals Acetylene (C2H2, 99.0 %, Afrox, Gauteng, South Africa); argon (Ar, 99.999 %, Afrox, Gauteng, South Africa); nitrogen (N2, 99.99 %, Afrox, Gauteng, South Africa); ethanol (C2H5OH, 99 %, Labchem, South Africa) and copper sheet 1 mm x 15.0 cm x 15.0 cm (thickness x width x length, Labchem, South Africa) 100 % were used. Glucose oxidase (GOx) from Aspergillus niger (100 U mg-1), potassium dihydrogen phosphate monobasic (KH2PO4) (99.99%), dipotassium hydrogen phosphate (K2HPO4) (≥ 99%), potassium chloride (KCl) (≥ 99%), ferrocene carboxylic acid (97%) and sodium dodecyl benzene sulfonate (SDBS) (≥ 99%) were purchased from Sigma-Aldrich (St. Louis, MO, USA) and used without further purification. PBS (1X, pH 7.4) and ferrocene carboxylic acid aqueous solutions were used as supporting electrolyte for all amperometric measurements. Aqueous solutions were prepared with double-distilled water from Millipore system (18.2 MΩ cm).

6

Characterisation Characterisation of the as-synthesised graphene samples was performed through X-ray diffraction, Raman spectroscopy, scanning electron microscopy (SEM), zeta potential and contact angle measurements. X-ray diffraction spectra were obtained using a Shimadzu X-ray diffractometer (Tokyo, Japan) X-ray 7000 operated at 50 kV and 50 mA, using a copper source with generated x-ray wavelength (λ) of 1.540498. Raman spectra were obtained using HR800 UV-VIS-NIR Raman spectrometer (Horiba Jobin-Yvon, Germany) equipped with an Olympus BX-40 attachment. The excitation wavelength and energy setting from a coherent Innova model argon-ion laser were applied as 514.5 nm and 1.2 mV, respectively. The Raman spectra were collected by means of back scattering geometry with an acquisition time of 50 seconds. In order to characterise the external morphology and elemental analysis, we also used an SEM instrument (Jeol XL30, MA, USA) with image magnification range of x5 – x1,000,000; a resolution of 2 nm and an accelerating voltage of 1-30 keV using an aluminium stub. The zeta potential of graphene dispersions before and after surface modification was determined using a Nano ZS dynamic light-scattering (DLS) Zeta potential instrument (Malvern Instruments, Worcestershire, UK). Water contact-angle measurement was performed using a CAM 200 optical contact-angle meter (KSV Instrument, Helsinki, Finland). All voltammetric measurements were carried out with an Ivium Stat.XR electrochemical analyser (Eindoven, Netherlands). A three-electrode cell with glassy carbon working electrode, having 0.07 cm2 surface area, platinum wire auxiliary and Ag/AgCl (3 M KCl) reference electrode was used for the voltammetric measurements. Preparation of CVD-synthesised graphene Copper sheet (1 mm thickness and 2.0 x 6.0 cm) was used as collector after extensively cleaning with absolute ethanol. The experimental procedure involved placing a metal strip right into the middle of the horizontal quartz tube, which was then inserted into a Nabertherm 7

P330 furnace with digital and programmable controller. A Teledyne Power Pod 400 massflow controller was connected to the system via quartz glass connectors and rubber tubes. The deposition of the graphene on the copper sheet was achieved using a mixture of argon (Ar) and nitrogen (N2) as carrier gases at flow rates of 350 cm3/min and 250 cm3/min, respectively, while acetylene (C2H2) was used as the carbon source. The process involved purging the quartz tube with argon gas flowing at the rate of 50 cm3/min for 30 min, after which the programmed temperature controller started heating the CVD system. The carrier gases at the flow rates stated earlier were introduced into the system after the attainment of the respective synthesis temperatures and allowed to stabilise for 10 min before the commencement of the synthesis reaction. The synthesis reaction was started by the introduction of acetylene gas (at the flow rate of 10 cm3/min) for 2 min and was ended by stopping the acetylene gas flow. The reaction products were allowed to cool to 500oC under the inert gas mixture of Ar and N2 and they were then cooled at room temperature under argon gas flow (alone) at a flow rate of 50 cm3/min. The experimental details and characterisation of the products are reported in the Supporting Information (Supporting information scheme S1). Fabrication of graphene-enzyme self-assembled hybrid structures The graphene-enzyme assembly was prepared by dispersing 1.0 mg of as-synthesised graphene powders in 5 mL of aqueous SDBS solution (0.1 M). The dispersion was sonicated for 3h, and incubated at room temperature (RT) under mild stirring conditions for further 3h. After that, the resulting suspension was subjected to three centrifugation/resuspension (in deionised water) cycles to isolate surfactant-capped graphene nanosheets and remove unbounded surfactant molecules. The graphene nanosheets were finally washed with distilled water and dried at room temperature prior to conjugation with glucose oxidase (GOx).

8

For the conjugation of enzyme with SDBS-capped graphene, SDBS-capped graphene (1.0 mg) was added to 1 mL of GOx solution (1 mg/mL) in PBS (pH 7.4), and the mixture was incubated for 3h at room temperature. The mixture was then centrifuged at 5000 rpm for 0.5h, then the precipitate was resuspended in PBS to remove excessive surfactant/enzymes. This cycle was repeated three times to successively remove loosely attached enzyme from the modified graphene surface. Assembly of bio-electrodes Prior to immobilisation of enzyme on the surface of modified graphene, glassy carbon electrodes (GCE) were carefully polished using 1.0, 0.3 and 0.05 micron Buehler alumina slurry on Buehler polishing micro-cloths (Buehler, Ltd. USA), respectively. The bioconjugate solution, which contained surface modified graphene-GOx enzyme, was sonicated for 5 min to achieve a homogenous dispersion. This suspension (15 µL) was drop-cast onto the GCE surface and dried at 4 °C for 8 h. The other electrodes, such as with/without enzyme, were prepared using the same procedure described above.

Computational studies Molecular dynamic simulations were carried out to study the stability of graphene and SDBS interface. In addition, the interaction of this hybrid material with glucose oxidase enzyme was also studied. The graphene coordinates with 5x4 dimension were obtained using VMD software and the edge atoms were suitably capped with hydrogens. Calculations at a PM3 semi-empirical level were carried out to obtain the Mullikan charges which were used in all subsequent simulations. Similarly for the single molecule of SDBS surfactant, a model was prepared by keeping all the CH2 groups in trans conformation. The geometry was optimised for the SDBS anion at density functional level of theory and 6-31g* basis set as implemented in Gaussian09 software. The ESP charges as implemented in a Merz-Singh-Kollman scheme, 9

were derived for the surfactant which was used in the molecular dynamics simulations. Finally, the coordinates for GOx were obtained from the protein database (reference number 1GPE). The general amber force-field (GAFF) was used for the description of both graphene and SDBS surfactants arranged on the surface of graphene in such a way that the alkane axis was perpendicular to the plane of the graphene. The entire system was neutralised by adding enough sodium ions and then the whole system was solvated with TIP3P water models. Another system was also prepared which was like the aforementioned system, but also included a monomer of GOx which was placed above the surfactant layer. The second system also included around 60000 water molecules. Molecular dynamic simulations using amber/14 software were carried out for these two systems to study the structure and stability. These two systems were studied at two different temperatures (200 and 300 K) to assess temperature dependency of these hybrid structures. The simulations were carried out in an isothermalisobaric ensemble so that the density of the system can be predicted. The temperature and pressure controls were performed by using a Nose thermostat and Barendsen's barostat, respectively. The time step for the integration of equation of motion was 1 fs and both the systems at two temperatures were studied for a total time step of 10 ns. The simulations were carried out until various structural and energetics parameters reached a constant value (or were fluctuating around a constant value). Results and discussion Preparation and characterisation of acetylene-sourced CVD graphene bioelectrodes Scheme 1 is a schematic representation of the steps involved in the fabrication of a grapheneGOx/GCE bioelectrode, starting from CVD graphene and culminating in the analyte (glucose) sensing process. As the first step, we computationally calculated both the systems namely (i) graphene + dodecyl benzene sulfonate ions solvated in water and (ii) graphene + dodecyl benzene sulfonate ions + glucose oxidase solvated in water at two different 10

temperatures, 200 and 300 K. The simulations were carried out until the density reached a constant value. The structures after the equilibration were analysed and the snapshots of the simulations were combined in Scheme 1, while being given in Supporting Information file in more detail. In the Figures, the counter ions (Na+) and water solvents are not shown to improve the clarity. As can be seen from Scheme 1 (also Figure S1 and S2), the graphene and dodecyl benzene sulphonate surfactant form a stable hybrid structure. All the SDBS molecules assembled on the graphene surface and in such a way that the molecular axis of SDBS was mostly perpendicular to plane of the graphene. It was worth noting that most of the SDBS molecules were inclined towards surface, which was attributed to the bulky benzene sulfonate group connected at the head of the SDBS molecule. At high temperature, the inclination angle increased and most of the SDBS molecules made non-90 degree angles with graphene. It is again worth noting that none of the molecules escaped from the graphene surface and it is clearly evident that the hydrophobic interaction between the SDBS alkane moiety and graphene was the most dominant and stabilising interaction in such structures. With these structures, the hydrophilic interactions between the polar sulfonate groups and water molecules could also be maximised. The simulation also showed that stable SDBSgraphene/GOx structures were obtained due to possible hydrophilic interactions between the polar sulphonate groups of SDBS and GOx. The immobilisation of GOx on the graphene surface was not usually possible due to the hydrophobic nature of the latter structure. However, the hydrophobic nature of graphene was successively modified through the SDBS assembly on the surface, which now made possible easy immobilisation of GOx on the surface of the graphene (since the modification with SDBS surfactant had provided a much more hydrophilic-like surface). The schematic of the CVD set-up is presented in figure S1 (Supporting information), and the overall equation of the reaction representing the CVD synthesis of the graphene

11

sheets is presented in Figure S2 (Supporting information), the reaction took place at atmospheric pressure at 850oC while using argon, nitrogen and acetylene gases at the flowrates of 350 cm3/min, 250 cm3/min and 10 cm3/min, respectively. The dimensions of the copper sheets used as substrates were: 6 cm x 2 cm x 1 mm (length x width x thickness). The duration of the synthesis reaction was 2 min. After the synthesis step, the surface modification process involved the dispersion of 1.0 mg of graphene in 5 mL of aqueous SDBS solution (0.1 M), the mixture was then sonicated for 1 h. The most vital consequence of this surface modification with SDBS is the introduction of electrostatic charges on the surface (basal plane) of the graphene through the ions present in the surfactant. The next step is the immobilisation of enzymes on the surface of the modified graphene sheet, this was achieved by dissolving 10 mg of GOx in 1 mL of PBS solution, followed by the incubation of the mixture at 4oC for 3 h. 1 mg of the modified graphene was added to the mixture then centrifuged at 5000 rpm for 30 min. The supernatant was collected for the determination of enzyme loading efficiency, while the precipitate was also recovered and washed three times to remove the loosely attached enzymes. The XRD pattern of the as-synthesised graphene sheets is given in Figure S3 (Supporting information). In the diffraction pattern, we evaluated specific graphene phases by using a reference diffraction pattern contained using the PDF2 software. Four (4) peak positions were observed at 37.8, 44, 64.6 and 77.6 (2θ degrees), respectively. From the PDF2 reference software, all four peaks correspond to reference peaks with the Miller’s index values given respectively as 021, 101, 203 and 110, and stem from hexagonal carbon peak positions with bond angles of 120o. The inter-atomic distances and d-spacings were calculated as, respectively, 2.399540Å, 2.031700Å, 1.451650Å, 1.230800Å and as 11 and 87 using fixed slit intensity Cu K1 1.54056Å (Fayos, 1999; Howe et al., 2003). The consistency of these four peaks, though with different intensities, in the entire experimental XRD pattern 12

gives the appearance of a fingerprint experimental diffraction pattern for few-layers graphene. The shape of the peak also indicated the material synthesised had a highly crystalline structure. Figure 1 combines Raman spectroscopy and electron microscopy results. As shown in the Raman spectrum of the as-synthesised graphene sheets, we found two specific graphene bands at 1361 and 1589 cm-1, corresponding to the D-band and G-band respectively. Raman spectroscopy has become a sort of “finger print” technique for the characterisation of all forms of graphitic carbon materials, giving valuable information about the presence or otherwise of sp2 hybridised carbon atoms, unwanted by-products (e.g., amorphous carbon), the presence of structural defects or damage to the plane or lattice of the synthesised carbon material structure, all of which are likely to occur during synthesis or in the course of postsynthesis treatment (Fayos, 1999; Ferrari, 2007). There are few prominent peaks in the Raman spectra of graphene namely: the G band which is incidental at around 1580 cm-1 and the D band, which is located at around 1350 cm-1. Ferrari (Ferrari, 2007), stated that in the molecular approach, the G band is attributed to the stretching of all pairs of sp2 carbon atoms in both rings and chains, while the D band is due to the breathing mode of sp2 atoms in rings. Zhu et al. (2010) however, said that the G band is attributed to the in-plane optical vibrations and the second order zone boundary phonons respectively, while the D-band is due to first order zone boundary phonons and it is due to the presence of defects in the graphene plane (Zhu et al., 2010). In Figure 1a, both the G band and the D band are notably present, signifying the presence of both sp2 carbon atoms and defects in the graphene plane. Figure 1b shows the optical image (at magnification of x1000) of the regions being analysed, taken during the Raman analysis. These optical Raman images clearly show wave-like (or corrugated iron-like) graphene structures. The SEM images of the as-synthesised graphene sheets also confirmed a plate-like structure (Figure 1c-d) with a thickness of about 2 nm.

13

Energy dispersive spectral (EDS) results of the as synthesised graphene sheets are presented in Figure S4 (Supporting information). EDS is an important tool for the determination of the elemental composition of the sample under analysis (Osikoya et al., 2015). However, it does not give any information about bonding types between the constituent elements. The EDS spectra in Figure S4 show the presence of a single peak due to carbon. This result obviously confirmed the synthesised material as pure graphene. The absence of oxygen peaks in the spectra indicates that the synthesised material was not graphene oxide, which is usually synthesised via the Hummer’s method (Ago, 2013; Hu and Su, 2013; Huang et al., 2011; Ivanovskii, 2012). The quantitative analysis of the graphene is presented in Table S1 (Supporting Information). It shows that the synthesised material consisted of highly pure carbon with other elements including silicon (Si), aluminium (Al) and potassium (K) present at trace levels. Si could be from quartz materials used for the CVD synthesis, Al could be from the aluminium stub used for the generation of electron in the Jeol XL30 SEM machine, which was also used for the elemental analysis, and K could be from handling during the characterisation. Figure 2a-b shows representative tapping mode AFM phase images of the different surface-modified sheets at different magnifications (a-b). The samples were prepared by using dilute water solution of SDBS-modified graphene dropped onto a surface of mica. We can clearly see that they all display a similar appearance, i.e., irregular shapes with lateral sizes typically between a few and several hundred nanometers, and that modified graphene nanosheets disperse in water very well without any strong aggregation; the corresponding height profile showed ∼ 2-10 nm, which indicates that the majority of the sheets are multilayered. The images also show that at this level of surface modification, the nanosheets were not seen to undergo any significant morphological change.

14

The surface charge of the colloidal system is an important parameter that not only defines the stability of the dispersion, but also helps to understand and characterise the electrochemical equilibrium at the interface. In the present study, we modified the graphene nanosheets with SDBS via hydrophobic-hydrophobic interaction between the alkyl chain of the surfactant and the basal plane on the graphene. In order to prepare the graphene-enzyme assembly, the graphene surface was first functionalised with anionic surfactant to produce a negatively charged graphene surface in well-dispersed solution. The surface modification and the surface charge formed after modification generate electrical potentials due to the electrostatic repulsion force between the modified graphene and the medium. Therefore, the formation of this interaction determines the colloidal stability of the dispersion, the zeta potential value and also provides qualitative (sign) and quantitative (magnitude) information about the effective surface charge related to the electrical double layer around the surface. The isoelectronic point (IEP), which is described as the pH at which the surface is neutrally charged, can also be derived from the zeta potential measurement. The zeta potential of graphene nanosheets before and after surface modification by SDBS was measured to evaluate the dispersion quality and charge characteristics at different pH values (Figure 3a). Zeta potential values of unmodified graphene decreased from +14.3 mV to -17.4 over a wide pH range. The presence sulphonated aliphatic hydrocarbon ion (from SDBS) functionalities on the surface can dissociate and form negatively charged sulphonated alkyl-benzene groups (-RC6H4SO3-), which may make the surface of graphene more negative, even though there are no extra charged ions on the surface. However, SDBS-modified graphene shows more stability and homogeneity over a wide pH range in 0.1 M SDBS solution. It was relatively stable over a range of pH values (pH 6-10) with low zeta potential values (-27.8 to -67.0 mV). The movement of the valence electron through the π* conduction band may cause a slight repulsion of negatively charged (-SO3-) anionic surfactant molecules via Coulombic

15

interactions between the two negative sides. These measurements clearly confirmed that SDBS is a good dispersant in terms of delivering a zeta potential even at relatively low concentrations. The wettability of modified electrode was also analysed using contact angle measurements (Figure 3b). The electrodes were modified with unmodified, SDBS-modified and enzyme-immobilised graphene, and each one showed different wettability characteristics under the same conditions. The electrode modified with bare unmodified graphene showed an intrinsically quite hydrophobic character (i.e., contact angle, 90° ± 8°) due to the production method. However, when the surface was modified with SDBS and the modified graphene was drop-cast on the electrode surface, we observed that wettability increased significantly resulting in a contact angle of 45° ± 6°, due to the strong hydrogen bonding between the graphene surface and the surrounding aqueous environment. Moreover, when GOx enzyme was immobilised on the same electrode, the surface changed to a more hydrophilic state (i.e., contact angle, 40° ± 6°). The change in hydrophobicity/hydrophilicity as measured by watercontact angle is one of the driving forces behind the diffusion mechanism of electroactive species or substrate to the electrode surface. Electrochemical properties of bio-electrodes The electrochemical properties of all modified electrodes were characterised by measuring voltammetry responses in 1.0 M PBS solutions containing 1 mM ferrocene carboxylic acid and 0.1 M KCl. The results are shown in Figure 4a-b, Figure 5a-b and Figure S5 (Supporting Information). The CV analysis of the redox reaction of ferrocene carboxylic acid is a wellknown method to characterise the electrochemical behaviour of modified and bare electrodes and is the simplest way to determine the electron transfer rate. The cyclic voltammetry responses of all modified electrodes displayed a classical sigmoidal shape with different

16

peak-to-peak potential separations. The larger peak separations indicated slow electron transfer kinetics. The electrodes containing graphene-based assemblies demonstrated narrow peak-topeak separations, which indicate the achievement of fast electron transfer for each system. The bare GCE (without any graphene-based assembly) also showed a very narrow ΔEp value around 100 mV, which indicated fast electron transfer. After modification of the electrode surface with SDBS modified graphene, the peak current of ferrocene carboxylic acid increased relative to electrodes modified with the as-synthesised graphene or bare GCE, indicating that the surface modification played an important role in creating a higher electroactive surface area and providing the conductive interlayer for the electron transfer of ferrocene carboxylic acid. Moreover, when the peak currents and ΔEp of modified electrodes are compared, it is clear that surface modification in the assembled system enhanced the electron transfer. The effects of scan rates on the cathodic (Ipc) and anodic peaks potentials (Ipa) of modified electrode surface are shown in Figure 4b and Figure S6a-d (Supporting Information). Both anodic and cathodic peak currents increased linearly with increasing scan rates (10 to 150 mV/s), which suggest a typical diffusion-controlled electrode process. Moreover, the electrocatalytic oxidation of glucose was investigated, using grapheneGOx modified bioelectrodes, by amperometric measurement (Figure 5a). Successive additions of glucose were made with an applied voltage of +0.35 V (vs. SCE) in the presence of 0.1 M ferrocene carboxylic acid and aqueous KCl solution (0.1 M) as a supporting electrolyte. SDBS-modified graphene-GOx conjugate modified electrode showed a rapid and sensitive response to the addition of glucose, and a steady-state current was obtained within 5 seconds. The relatively fast response may be attributed to the fast diffusion in the open threedimensional structure of the assembly and the synergistic catalytic effect of the immobilised 17

systems toward glucose. Furthermore, well-defined current responses proportional to glucose concentrations were observed. The calibration curve of the electrocatalytic currents of the assembly to the concentrations of glucose is shown in Figure 5b. The catalytic current displays a linear relationship to glucose concentration in the range from 0.2 to 9.8 mM. The sensitivity and detection limit of modified electrode was 0.087 µA/µM/cm2 and 0.12 µM (S/N=3), respectively. The results show that SDBS-modified graphene-GOx bioelectrode provided good catalytic performance for the detection of glucose over a reasonably wide linear range, high sensitivity and a low detection limit. Table S2 shows comparative recent studies on graphene bioelectrodes. Such good catalytic properties of the enzyme electrode are ascribed to the open structures of the SDBS-modified graphene nanosheets for enzyme immobilisation, fast mass transport due to the large interlayer distance of ordered structure and the biocompatibility of SDBS-graphene.

Conclusions Hybrid systems such as carbon based 2D-materials have been attracting the attention researchers due to continuing demands for more efficient sensing platforms. Based on our previous experience of graphene-based smart hybrid interfaces, we focused here on combining CVD-grown, pure, few-layers graphene with anionic surfactants and GOx based electrocatalysis, while exploring both theoretical and practical design routes simultaneously. The results showed that the CVD method is a promising way to produce few-layers graphene at extremely high purity. SDBS is excellent surfactant for easy and efficient surface modification of the graphene sheets in order to increase surface wettability and enzyme loading capacity. Theoretical calculations also confirmed that the SDBS modification enhanced the surface chemistry for enzyme loading and electrocatalytic performance. Finally, the electrochemical characterisation of the modified CVD graphene-based bioelectrodes,

18

revealed that this approach is applicable for developing smart bioelectronic interfaces with excellent electrocatalytic performance.

Acknowledgement The authors wish to acknowledge the Swedish Research Council (VR-2011-6058357), Sweden and the Research Directorate of the Vaal University of Technology, South Africa for the generous financial support to carry out this research. References Ago, H., 2013. Epitaxial CVD Growth of Graphene: Growth Mechanism, Nanofabrication, and Properties. Proceedings of 2013 Twentieth International Workshop on ActiveMatrix Flatpanel Displays and Devices (Am-Fpd 13): Tft Technologies and Fpd Materials, 55-58. Ashaduzzaman, M., Anto Antony, A., Arul Murugan, N., Deshpande, S.R., Turner, A.P.F., Tiwari, A., 2015. Biosens. Bioelectron. 73, 100-107. Avouris, P., 2010. Nano Lett. 10(11), 4285-4294. Batzill, M., 2012. Surface Sci. Reports 67(3-4), 83-115. Brownson, D.A.C., Varey, S.A., Hussain, F., Haigh, S.J., Banks, C.E., 2014. Nanoscale 6(3), 1607-1621. Castro Neto, A.H., Guinea, F., Peres, N.M.R., Novoselov, K.S., Geim, A.K., 2009. Rev. Modern Phys. 81(1), 109-162. Cermak, J., Yamada, T., Ledinsky, M., Hasegawa, M., Rezek, B., 2014. J. Mater. Chem. C 2(42), 8939-8948. Chun, S., Choi, J., Ashraf, A., Nam, S., 2014. Three-dimensional, flexible graphene bioelectronics. 2014 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014, pp. 5268-5271. Fayos, J., 1999. J. Solid State Chem. 148(2), 278-285. Ferrari, A.C., 2007. Solid State Comm. 143(1-2), 47-57. Geim, A.K., Novoselov, K.S., 2007. Nature Mater. 6(3), 183-191. Howe, J.Y., Rawn, C.J., Jones, L., Ow, H., 2003. Powder Diffr. 18(02), 150-154.

19

Hu, Y., Su, X., 2013. Chemically Functionalized Graphene and Their Applications in Electrochemical Energy Conversion and Storage. In: Aliofkhazraei, D.M. (Ed.), Advances in Graphene Science, pp. 161-189. InTech. Huang, X., Yin, Z.Y., Wu, S.X., Qi, X.Y., He, Q.Y., Zhang, Q.C., Yan, Q.Y., Boey, F., Zhang, H., 2011. Small 7(14), 1876-1902. Ivanovskii, A.L., 2012. Russ. Chem. Rev. 81(7), 571-605. Kang, X., Wang, J., Wu, H., Liu, J., Aksay, I.A., Lin, Y., 2010. Talanta 81(3), 754-759. Kibena, E., Marandi, M., Sammelselg, V., Tammeveski, K., Jensen, B.B.E., Mortensen, A.B., Lillethorup, M., Kongsfelt, M., Pedersen, S.U., Daasbjerg, K., 2014. Electroanalysis 26(12), 2619-2630. Kim, J., Cote, L.J., Kim, F., Yuan, W., Shull, K.R., Huang, J., 2010. J. Am. Chem. Soc. 132(23), 8180-8186. Kumari, A., Prasad, N., Bhatnagar, P.K., Mathur, P.C., Yadav, A.K., Tomy, C.V., Bhatia, C.S., 2014. Diamond Relat. Mater. 45, 28-33. Li, W., Yoon, J.A., Matyjaszewski, K., 2010. J. Am. Chem. Soc. 132(23), 7823-7825. Liu, J., Tang, J., Gooding, J.J., 2012. J. Mater. Chem. 22(25), 12435-12452. Nam, S., Chun, S., Choi, J., 2013. All-carbon graphene bioelectronics. Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS, pp. 5654-5657. Osikoya, A.O., Dikio, C.W., Ayawei, N., Wankasi, D., Afolabi, A.S., Dikio, E.D., 2015. Adv. Mater. Sci. Appl. 4(2), 53-62. Parlak, O., Tiwari, A., Turner, A.P., Tiwari, A., 2013. Biosens. Bioelectron. 49, 53-62. Parlak, O., Turner, A.P., Tiwari, A., 2014. Adv. Mater. 26(3), 482-486. Parlak, O., Turner, A.P.F., 2016. Biosens. Bioelectron. 76, 251-265. Puri, N., Niazi, A., Srivastava, A.K., Rajesh, 2014. Appl. Biochem. Biotechnol. 174(3), 911925. Shahil, K.M.F., Balandin, A.A., 2012. Solid State Comm. 152(15), 1331-1340. Shao, Y., Wang, J., Wu, H., Liu, J., Aksay, I.A., Lin, Y., 2010. Electroanalysis 22(10), 10271036. Singh, G., Rice, P., Mahajan, R.L., McIntosh, J.R., 2009. Nanotechnology 20(9), 095701. Valota, A.T., Kinloch, I.A., Novoselov, K.S., Casiraghi, C., Eckmann, A., Hill, E.W., Dryfe, R.A.W., 2011. ACS Nano 5(11), 8809-8815. Wang, X., Yang, H., Song, L., Hu, Y., Xing, W., Lu, H., 2011. Compos. Sci. Technol. 72(1), 1-6. 20

Zhang, Q., Yang, S., Zhang, J., Zhang, L., Kang, P., Li, J., Xu, J., Zhou, H., Song, X.M., 2011. Nanotechnology 22(49). Zhu, Y., Murali, S., Cai, W., Li, X., Suk, J.W., Potts, J.R., Ruoff, R.S., 2010. Adv. Mater. 22(35), 3906-3924.

21

Surface functionalisation

Graphene

SDBS

SDBS-Graphene

Enzyme Immobilization

GOx

Glucose

Mediator HO

Mediator HO 2

2

2

2

Glucose

GOx Graphene/Gox Graphene/Gox bioelectrode bioelectrode

Graphene/GOx

Scheme 1. CVD graphene-GOx/GCE bioelectrode fabrication steps and the sensing process.

22

c

d

Figure 1. (a) Raman spectra, (b) Raman optical image, (c) flat and large surface area and (d) multiple layers for the as-synthesised CVD graphene.

23

Figure 2. Tapping mode phase images by AFM for the surface modified graphene at different magnifications (a-b).

24

Figure 3. Zeta potential (a) and water contact (b) angle measurement of unmodified graphene and SDBS-modified graphene -based assemblies and representative images for SDBSgraphene (c) and SDBS-graphene-GOx interactions (d).

25

Figure 4. Cyclic voltammetry, Scan rate: 50 mV s–1 vs. Ag/AgCl reference electrode in 1.0 M PBS solutions containing 1 mM ferrocene carboxylic acid and 0.1 M KCl (a) and the effect of scan rates (b) on the cathodic (Ipc) and anodic peaks (Ipa) of bare and modified electrodes.

Figure 5. Amperometric response (a) and calibration curve (b) for the sensing of glucose with SDBS-modified graphene electrode-GOx assembled bioelectrode in 0.1 M ferrocene carboxylic acid-0.1 M KCl solution at 0.35 V applied potential vs. Ag/AgCl.

26

TOC Figure

Glucose

Mediator HO 2

Mediator H2O2

2

GOx Graphene/Gox bioelectrode

Graphene /GOx bioelectrode and glucose sensing process.

27

Highlights

1.

Synthesis and characterisation of graphene via a CVD method.

2.

Theoretical calculations for the interaction between surfactant, enzyme and CVD graphene.

3.

Investigation of electrocatalytic behaviour of the bioelectrodes.

4.

Demonstrated the use of a graphene sheet as a fundamental building block to obtain a highly ordered graphene-enzyme electrode for electrochemical biosensing.

28