Gold nanoparticles modified GC electrodes: electrochemical behaviour dependence of different neurotransmitters and molecules of biological interest on the particles size and shape
- First Online:
- Cite this article as:
- Caschera, D., Federici, F., Zane, D. et al. J Nanopart Res (2009) 11: 1925. doi:10.1007/s11051-008-9547-0
- 270 Views
Gold colloidal nanoparticles (AuNps), synthesized by gold chloride hydrate (HAuCl4) chemical reduction were used to realize a modified glassy carbon electrode (GCE). Different shapes and sizes were observed, varying the molar ratio of HAuCl4 and polyvinylpyrrolidone (PVP). The electrochemical behaviour of different neurotransmitters and molecules of biological interest (dopamine, caffeic acid, catechol, uric acid, epinephrine and serotonin) were investigated by cyclic voltammetry (CV) at the AuNps modified GCE and a dependence of the electrochemical response on the size and the shape of the particles was observed. The electrochemical responses were stable during time with a generic decreasing of the peak current after 10 days ranging from 5–10% for catechol, uric acid and serotonine to 10–15% for the other analytes. A study on the electrochemical interface of modified electrodes was also carried out by means of electrochemical impedance spectroscopy (EIS).
KeywordsGold nanoparticlesAuNp modified electrodeNeurotransmittersColloidsNanobiotechnology
The realization and the study of modified electrodes are very important to develop a new generation of sensors. For this purpose, the unique optical and conductivity properties of gold nanoparticles make them very interesting for the construction of nanoparticles-based devices. Gold is a material perfectly biocompatible, and this property makes it suitable to be used in many biological application fields.
The properties of metal nanoparticles strongly depend on their size and shape (Burda et al. 2005; Liz-Marzan 2004; Handley 1989; Puntes et al. 2002; Huang and Yang 2005), so a synthetic procedure, controlling the growth and the morphology of the nanoparticles, seems to be critical and appealing.
Usually, nanoparticles are prepared by reduction of metal salts with a chemical reducing agent (such as citric acid, borohydride or other organic compounds) (Tan et al. 2003; Bonet et al. 1999; Zhou et al. 2002; Sun and Xia 2002), by photo-reduction (Zhou et al. 1999; Esumi and Pal 2005) or sono-chemical reduction (Fujimoto et al. 2001), but always in the presence of a protecting agent which would prevent them from agglomeration. The surfactant role is to control the crystal growth and shape, by attaching itself selectively to different crystal surfaces. In addition, the surfactant avoids particles agglomeration, makes the particles stable in solution and preserves them against a further oxidation.
A possible application of AuNps for the assembling of more sensitive electrochemical probes, to detect important biological molecules, sounds interesting. For example, the monitoring of cathecol and uric acid could be important in the aetiology of brain diseases, like Parkinson’s disease and Schizophrenia (Ahlskog et al. 1996).
Evidence for the importance of the serotonin levels in anxiety disorders has increased substantially in recent years. From a clinical perspective, these disorders are very different such as panic, obsessive-compulsive disorder, post-traumatic stress disorder, and social anxiety disorder (social phobia) (Robinson and Hood 2007).
Finally, catecholamines, such as dopamine (DA) and epinephrine (adrenaline), are an important class of neurotransmitters and are involved in a variety of central nervous system functions (Kopin 1985). High catecholamine levels are known to be cardiotoxic, leading to rapid heart rate, high blood pressure, and possible death of the heart muscles (Sotonyi et al. 2004). On the contrary, a loss of DA-containing neurons may result in some serious diseases such as Parkinson’s disease (Olanow 1990).
In our work, we present an AuNps synthetic procedure, using non-toxic and biocompatible polyvinylpirrolidone, as reducing and stabilizing agent. The reaction is carried out in an aqueous solution with a procedure friendly to the environment.
Further, these colloidal AuNps are used to modify a conventional glassy carbon electrode, by dip coating technique. A study on the electrochemical interface of modified electrodes was also carried out by means of impedance measurements. An investigation on the voltammetric behaviour of dopamine, caffeic acid, epinephrine, norepinephrine and catechol and uric acid in function of gold nanoparticles size was also performed.
All chemicals (HAuCl4, Poly Vinyl-Pyrrolidone (PVP, MW: 10000), potassium chloride (KCl), potassium ferricyanide (K3[Fe(CN)6]), hexaammine ruthenium(III) chloride ([Ru(NH3)6]Cl3), dopamine, caffeic acid, serotonin, uric acid, epinephrine and catechol) were purchased from Sigma Aldrich and used without further purification.
Glassy carbon electrode (CGE/492/2) of apparent area 0.031 ± 0.003 cm2 was purchased from Amel Milan, Italy.
All aqueous solutions were prepared using deionized-distilled water.
All glassware used were cleaned in a bath of freshly prepared aqua regia solution, then it was rinsed thoroughly with distilled water prior to use.
The obtained gold nanoparticles were characterized by UV–Vis spectroscopy, by transmission electron microscopy (TEM) and by XRD diffraction.
UV–Vis absorbance spectra were collected using a double beam UV–Vis Jasco 7850; spectral range 190–900 nm, with a slid width of 2 nm.
TEM was carried out with a Philips CM120 operating at 120 kV, with a single-tilt holder. Specimen for inspection by TEM was prepared by depositing a drop of the gold colloidal solution on a 200 mesh FORMVAR grid, coated by an amorphous carbon film and evaporating the solvent in air at room temperature.
XRD diffraction spectra were collected with an SEIFERT ID 3000 operated at 30 KV voltage and 30 mA, with CuKα1 radiation (1.54059 Å), using a scanning rate of 0.05 s−1 in 2θ ranging from 5° to 65°.
For the powder analysis, 3 mL of gold nanoparticles solution was evaporated on a Petri dish, dried at 70 °C for 3 h and then grinded to obtain a pink powder. The powder was preserved from humidity in a sealed container.
The electrochemical experiments were performed with an Autolab 12PG potentiostat/galvanostat with an FRA2 module, Autolab GPES and FRA software system was used for cyclic voltammetry (CV) and for Electrochemical Impedance Spectroscopy (EIS).
Synthesis and characterization of Au colloidal nanoparticles
Au nanoparticles were obtained mixing aliquots of HAuCl4 3.8 × 10−3 M aqueous solution, and PVP 0.01 M solution in order to investigate the effect of different Au/PVP molar ratios on the procedure. The synthesis was carried out at room temperature.
In few hours (approximately 2 h), the transparent initial solution turned into red-violet colour, depending to the Au/PVP molar ratio used, and it was stable for more than 6 months.
Different molar ratios of Au/PVP were investigated: the final products were controlled by UV–Vis analysis and TEM measurements.
Assembling of AuNps modified glassy carbon electrode
Glassy carbon surfaces were polished with alumina powder (Al2O3, Buehler, Evanston, IL) having particle sizes of 1, 0.3, and 0.05 μm, washed with double distilled water, and finally sonicated for 15 min before use.
Afterwards, the glassy carbon (GCE) electrode was modified by dip coating, using different immersion times (15 min, 2 h and 2 days) in the colloidal nanoparticles solutions. The resulting modified electrode was dried under nitrogen flow and used for the electrochemical characterization. All experiments were carried out at room temperature. The electrochemical cell was assembled as a conventional three-electrode system: the modified electrode as working electrode, an Ag/AgCl electrode as reference and a Pt electrode as counter.
The impedance of the samples was measured at room temperature by means of a frequency response analyser in the 104–0.1 Hz frequency range in open circuit voltage conditions (OCV).
The modified electrode was electrochemically characterized by cyclic voltammetry (CV) in KCl 0.2 M, using K3[Fe(CN)6] 0.01 M and [Ru(NH3)6]Cl3 0.01 M as electrochemical probes.
Finally, this AuNPs modified gold electrode was used in phosphate buffer for the electrochemical determination of different biological molecules and neurotransmitters at a concentration of 0.005 M, such as dopamine, caffeic acid serotonin, uric acid, epinephrine and catechol at pH 7.0. Phosphate buffer solutions (PBS; pH 7.0 0.1 M) were prepared using sodium dihydrogen phosphate dihydrate (Fluka, Microselect for molecular Biology), adjusting pH with NaOH pellets.
The buffer solutions were prepared using distilled-deionized water and were kept refrigerated to minimize bacterial growth.
Further, the CV experiments were carried out at a scan rate ranging from 20 to 100 mV/s.
Result and discussion
Synthesis and morphological characterization of AuNps
Au nanoparticles were synthesized by HAuCl4 chemical reduction, using PVP as reducing-dispersing agent. The PVP coordinating mechanism on metal particles is well known (Fievet et al. 1989; Ducamp-Sanguesa et al. 1992). PVP, as surfactant, has a great effect on preventing the nanoparticles from aggregating and, such an effect could largely depend on the steric effect, arising from the long polyvinyl chain of PVP on the surface of the growing particles (Wang et al. 2005). Since the steric effect was largely determined by the covered fraction of PVP on the surface of the particles, the amount of PVP should be important to control both the shape and the size of the final product.
The related mechanism is radicalic, involving a concerted coordination and reduction of Au+3 and according to a well-known literature scheme (Staszewska 1983; Kaczmarek et al. 1998; Sionkowska et al. 2005).
Summary of experimental parameters and some chemical-physical characteristics of gold nanoparticles solutions obtained at different Au/PVP molar ratios
Au/PVP molar ratio
λmax absorption (nm)
Medium diameter (nm ± 0.2)
Shape of particles
7.60 × 10−3
2.00 × 10−2
5.00 × 10−2
9.50 × 10−2
It is well known that, when metallic clusters are not growing as spherical particles, but as oblate, truncated particles, or even as wires, the surface plasmon resonance (SPR) will deviate from the single peak feature and split into transverse resonance (high energy) and longitudinal resonance (low energy) (Chang et al. 1999). In sample (c) (molar ratio ≥5 × 10−2), the spectrum showed the presence of another broad peak in the range of 600 to 700 nm, probably originating from longitudinal Au nanorods SPR, according to TEM observations (see further discussion). Also the broadening peak for sample (d) was due to the presence of another signal in 600–700 nm range. For samples (a) and (b), obtained with lower molar ratio, all spectra showed only one peak for Au nanoparticles.
Moreover, according to the literature (Mie 1908; Haiss et al. 2007), the average gold particle diameter could be estimated from the maximum absorption band; the diameter of gold nanoparticles increases when the maximum absorption band shifts to longer wavelength. So we expected smaller particles could be formed in solutions with lower Au/PVP molar ratio, and these particles should have a more spherical shape, as confirmed by TEM analysis.
Using a relative high concentration of PVP, TEM images (Fig. 2a and its corresponding histogram of distribution 2e) showed that only small spherical gold nanoparticles (average diameter 3.5 nm) could be obtained and the particle’s distribution was quite homogeneous. It should be due to the protective action of PVP molecules; in fact, PVP prevented the growth of gold particles by absorbing on the surface of the Au0 nanoparticles.
Lowering the PVP content, larger and polygonal gold nanoparticles started to appear (Fig. 2c, d). This clearly indicated that there were few polymeric molecules surrounding the surface of just-formed Au0 particles. Moreover, the growth of particles was not isotropic and preferred direction of growth could give rise to polygonal gold nanoparticles (Mandal et al. 2002; Sau and Murphy 2005; Brown et al. 2000; Jana et al. 2001). We assumed that a higher concentration of PVP could act as an effective capping agent for the preparation of gold nanoparticles. Gold nanoparticles prepared in this way showed better stability under the same preserving conditions.
When a poor content of PVP was used, the solution turned bluish-violet due to the formation of larger and irregular particles, which finally collapsed in few hours. Increasing the molar ratio, the particle size also increased and the medium diameter grew until it reached the maximum of 21.5 nm for the highest molar ratio. Moreover, the distribution of the nanoparticles into the matrix also changed; in the first sample analysed, the particles were homogeneously distributed in the observed area, while, on increasing the Au/PVP ratio, the particles could form large clusters, often larger than 100 nm in extension, dispersed in PVP. Nevertheless, no agglomeration was visible in those clusters and the particles were always well separated, even if sometimes polygonal particles resulted placed on top of the smallest ones. Only in few cases we observed a partial coalescence between different particles, which was always comparable in dimension.
These nanoparticles were analysed by transmission electron diffraction. The electron diffraction patterns showed concentric circles resulting from the random orientation of crystal planes.
These results indicated that the morphology of Au colloids was dependent on the preparation conditions. In particular, as indicated in Fig. 3c and d, the icosahedral crystals, formed at higher Au/PVP molar ratios, showed stress patterns probably imposed by the polymeric matrix, causing the symmetric growth of these ultrathin gold plates (Porel et al. 2005; Shao et al. 2004).
To confirm the crystallinity of the samples, XRD diffraction was performed. It is observed that PVP maintained its dispersing-protecting property unchanged even in the solid state, and the powder remained unaltered for several months.
The value of the lattice constant calculated from XRD pattern was 4.0770(7) Å, which was consistent with that of pure bulk gold. The significant broadness of the scattering peaks is caused by the very small size of crystallites. For this sample, the average size of Au crystals was calculated by the Scherrer’s equation and it was 21.0 ± 0.3 nm. Such a value was in good agreement with the corresponding TEM measurement (medium diameter 21.5 nm), thus proving a mono-crystalline nature of gold cluster.
XRD measurements were carried on the other samples, but in those cases, due to the smaller particle size and the major contribution of the amorphous matrix to the diffraction, most of the peaks related to gold crystalline were hidden in the noise background.
Electrochemical characterization of Au modified GCE
The AuNps modified electrodes were firstly electrochemically characterized by some “test compounds”, such as 0.01 M K3[Fe(CN)6] and 0.01 M ([Ru(NH3)6]Cl3) in KCl 0.2 M solution.
Ipc (μA) at AuNps modified GC electrode using different immersion times (15 min, 2 h and 2 days) and at GC bare electrode (2 mm diameter) in a 0.01 M K3[Fe(CN)6]− 0.2 M KCl solution scan rate 20 mV/s
Probably, the observed current is related to two opposite factors: the diffusion of the analyte towards the modified surface electrode and the PVP wrap on the gold nanoparticles. The first one became important for dipping time up to 2 h, while for longer time the latter effect became prevalent.
Differences in interfacial features of the electrodes modified by AuNps of different sizes have been measured by faradaic electrochemical impedance spectroscopy (EIS), at bare GCE and at AuNps modified GCE in a 0.01 M K3[Fe(CN)6] 0.2 M KCl solution. The equivalent circuit used to describe the electrochemical interface is a Randles circuit (Bard and Faulkner 1980), in which the electrical simple elements were: the ohmic resistance of the electrolyte solution (Re), the Warburg impedance (Zw) resulting from the diffusion of ions from the bulk electrolyte to the electrode interface or inside the material, the double layer capacitance (Cdl) and charge transfer resistance (Rct), relevant if a redox probe is present in the solution. Since our EIS experiments were recorded in OCV conditions without any overpotential acting on the electrodes, the results gave information about the surface coverage.
Values for CPE and Rct were evaluated (sample c: CPE = 0.180 μF and Rct = 199 Ω; sample b: CPE = 0.658 μF and Rct = 298 Ω). They clearly indicated that mainly, the capacity underwent changes, as we expected, because of its dependence on the surface area of material. Moreover, regarding the values obtained for Rct, we have kept in mind the different superficial areas of samples, due to the differences in AuNps size, and this resistance could be considered as an interfacial resistance. We suggest that these results could support the presence on the electrode of a larger redox surface according to CV data (see below); in fact the higher peak current could be related to a larger electrodic surface.
The effect of PVP polymer is also important; in fact the charge transfer resistance values are high for the presence of such insulating binder, but not enough to block completely the electrochemical activity. Concerning the region of low frequencies, it is clearly controlled by a Warburg impedance behaviour, indicating a diffusion limited process, caused by PVP, partially limiting the transport of ions to the surface of nanoparticles.
Summary of the cyclic voltammetric data in a 0.01 M K3[Fe(CN)6] and 0.01 M [Ru(NH3)6]Cl3, KCl 0.2 M, scan rate 20 mV/s for the AuNps-GC electrode and for the bare electrode
Similarly, diffusion-controlled electron transfer at AuNP modified electrodes was also found for ([Ru(NH3)6]Cl3 (data not shown).
Electrochemical behaviour with biological analytes
The responses of AuNp modified GC electrodes were compared with those of bare GC electrodes for a more complete investigation of the electrochemistry of several interesting biological molecules. Bare GC electrodes were chosen here as a reference because of the satisfactory electrochemistry that had been previously observed with inorganic substrates as well as with these same molecules of biological interest (Shahrokhian and Hamzehloei 2003; Shen-Ming and Kuo-Tzu 2003; Wang et al. 2001a, b; Jian-Shan et al. 2003; Zhang and Bond 2003). Moreover, we tried to dropcast AuNPs on bare gold electrode, but the adhesion on the electrodic surface was very poor, probably depending on the used dispersing agent.
Summary of the peak potential data for several biological redox systems (0.005 M) at AuNps modified GC electrodes and at bare GC electrode (dipping time 2 h) in a pH 7.0 phosphate buffer solution (0.1 M) at scan rate 100 mV/s
Electrochemical parameters (mV)
Summary of the peak potential data for several biological redox systems (0.005 M) at AuNps modified GC electrodes and at bare GC electrode (dipping time 2 h) in a pH 7.0 phosphate buffer solution (0.1 M) at scan rate 100 mV/s
Electrochemical parameters (μA)
Considering Tables 4–5, the corresponding Epa and Epc (mV) values detected at AuNP modified electrodes for the biological compounds were similar to those detected at GC bare electrode (see previous paragraph, concerning the test compounds) and the evident electroanalytical improvement concerned the higher peak current values obtained.
The electrochemical behaviour of dopamine at AuNPs GC modified electrode is quite similar to that at other electrodes modified with other nanomaterials (see Valentini et al. 2005), probably it is related to a particular interaction with a nanostructured material, which seems to be easier for this neurotransmitter.
The regenerability of the modified GCEs was also tested. Evidence for the non-adsorption behaviours of biological molecules was demonstrated by the following experiment. When AuNP modified GC electrodes were switched to 0.1 M phosphate buffer (pH 7.0) after being used in biological molecules solutions, there were no peak signals at all.
The electrochemical response is stable during the time with a generic decreasing of the peak current after 10 days ranging from 5–10% for catechol, uric acid and serotonin to 10–15% for the other analytes.
We observed a generic peak current increase with respect to the bare GC electrode, and, further, a different behaviour depending not only on the particle size, but also on the nature of the biological molecules.
For catechol and its derivatives such as caffeic acid and dopamine, we observed a higher response using nanoparticles with a diameter ranging from 3.5 to 9.8 nm. On the other hand, for heterocyclic molecules such as uric acid and serotonin, the highest peak current resulted, using AuNps with the largest diameter 21.8 nm.
The epinephrine behaviour is unusual because even if it could be considered a structural derivative of catechol, its electrochemical response is almost similar to uric acid and serotonine. This behaviour could be explained considering other characteristics of epinephrine than its structure, such as polarity, size, and its possible interactions with NP. Other elements that could be evaluated are the role of the PVP on the electrode surface and the different geometry of the nanoparticles, which could affect the electrochemical response and the interaction with the biological analytes.
In our opinion, this electrochemical behaviour depending on the size and shape of AuNps could be conveniently exploited in order to assemble sensors for detecting selectively biological molecules.
We obtained stable aqueous solutions of gold nanoparticles of different shapes and sizes, using PVP as protecting as well reducing agent. The diameter of nanoparticles varies from 3.5 to 21.5 nm, and their shape from spherical to polygonal, depending on Au/PVP molar ratios. The colloidal solution is very stable for about 6 months.
This synthetic green method is particularly useful when the presence of any other interfering agent is not desirable, such as in sensing applications. In fact, PVP is completely biocompatible, friendly for the environment and completely non-invasive material. Then gold nanoparticles modified GCE was prepared to study the amplification of electrochemical signals of different neurotransmitters with respect to a conventional GCE.
The electrochemical response was studied in function of the nanoparticles dimension and of deposition time. However, the responses are stable during the time and no degradations are present even after several days.
Modified GCE with different gold nanoparticles showed a stable and amplified signal respect to the conventional and depending on the nanoparticles size and shape and on the neurotransmitters nature.
Our results seem interesting in order to assemble selective sensors for a target neurotransmitter, changing only the dimensions and the shape of the nanoparticles used and so it could be possible to detect a biological molecule in a very specific way.
Thanks are due to the group of Prof. Bemporad for the helpful assistance in TEM analysis in the LIME Laboratory of University of RomaTre and to Mr C. Veroli (X-Ray Laboratory ISMN-CNR) for his assistance in XRD measurements.