Chemical additive and physical template-free electrochemical methods to prepare carbon-supported nanostructures of catalyst metals represent an emerging technology. Formation of the metal nano/microstructures depends not only on the electrochemical method/parameters but also on the nature of the underlying carbon material. Here, we present a comparative evolution of unevenly distributed coral-like aggregates of nanocuboid-shaped gold nanostructures (AuNCBs) on the oxidised form of boron, nitrogen-doped carbon nanoonions (oxi-B,N-CNO) compared to evenly distributed bud-like aggregates of cubic shaped gold nanostructures on bare glassy carbon electrode under a similar electrochemical approach. The synthesis method provided the best availability of the surface active sites, whereas the shape of the structures showed a direct influence of both outer-sphere and inner-sphere electron transfer reactions. The higher sensitivity of AuNCBs@oxi-B,N-CNO compared to individual components and bare carbon/gold electrodes toward the inner-sphere oxidative reaction of N-acetyl-L-cysteine (NAC) was exploited in order to develop an electrochemical assay method with sensitivity and linear dynamic range of (4.70 ± 0.25) × 10⁻⁴ C∙cm⁻²∙mM⁻¹ and 0.2–2.5 mM, respectively in acetate buffer (pH 4.45). Furthermore, the sensor design was deployed in the quantitation of NAC in pharmaceutical preparations, resulting in 89%–106% recovery.

Abbreviations

AuE: gold electrode
AuNCB: gold nanocuboid
AuNP: gold nanoparticle
Au-Seed: (pre nucleated gold seeds)
CNO: carbon nano onion
DND: detonation nanodiamond
GCE: glassy carbon electrode
GSH: Glutathione
NAC: N-acetyl cysteine
oxi-B,N-CNO: oxidised form of boron and nitrogen co-doped carbon nano-onion

1 INTRODUCTION

N-acetyl-L-cysteine (NAC) is an inexpensive therapeutic agent^[¹^] used to treat various diseases including cancer, cardiovascular diseases, HIV infections, acetaminophen-induced liver toxicity and metal toxicity.^[²^] It is a well-tolerated mucolytic drug that moderates clinging mucous secretions and enhances glutathione (GSH) transferase activity. It is a useful treatment for diseases that generate oxygen free radicals,^[³^] being a GSH precursor (a naturally occurring antioxidant).^[^{4, 5}^] Oxidative stress in biological systems is caused by free radicals, such as the superoxide anion radical (O₂^−•), hydroxyl radical (OH^•) and reactive chlorine/bromine species, or non-radical species, such as hydrogen peroxide (H₂O₂), hypochlorite (OCl⁻) and nitrites (NO₂⁻). These can cause DNA and RNA mutations, alterations to protein processing and lipid oxidation.^[⁶^] GSH helps to combat oxidative stress through scavenging for these reactive oxygen species.^[⁷^] Deacetylation of NAC occurs in the liver and small intestine during oral administration, resulting in the production of cysteine (L-Cys),^[⁸^] which can then be used to promote GSH synthesis. Free L-Cys readily oxidises to the corresponding disulfide, cystine, thereby forming the cysteine/cystine redox couple.^[⁹^] GSH is the principal intracellular non-protein thiol that plays an important role in preserving the intracellular redox state, through its actions as a non-enzymatic reducing agent,^[¹⁰^] thus all of NAC's intracellular effects are mediated by GSH replenishment. Repurposed usage of NAC has been shown to be beneficial in the treatment of patients suffering from coronavirus disease 2019, where severe inflammation results in a cytokine storm and significant elevation in GSH reductase leads to rapid depletion of GSH. Oral administration of NAC is an effective way of increasing GSH levels as it has better oral and topical bioavailability than GSH. Its relatively low toxicity and wide safety margin has led to this usage, where patients administered showed a significant decrease in the severity of symptoms.^[^{11, 12}^] Therefore, the assay of NAC in developmental drugs, commercial formulations and biological fluids of patients has both scientific and technological interests.

The state of the art for NAC assay relies on high-performance liquid chromatography^[^{13, 14}^] based methods, creating a tremendous analytical load on quality control teams to meet quality by design. An on-the-spot hand-held electrochemical assay could be a viable alternative to increase the throughput of the final product without compromising quality. Hence, there is a global effort to find an efficient and effective electrochemical transducer for quantitative assay of NAC including bulk gold electrode (AuE),^[¹⁵^] catechol-containing carbon paste electrode (CPE),^[¹⁶^] copper nitroprusside adsorbed on 3-aminopyrosisica modified CPE,^[¹⁷^] copper oxide nanostructures on an ITO substrate,^[¹⁸^] cyclodextrin-carbon nanotube modified glassy carbon electrode (GCE),^[¹⁹^] multi-walled carbon nanotube and Nafion modified GCE,^[²⁰^] gold film modified carbon microelectrode^[²¹^] and an acetaminophen ruthenium oxide nanoparticle modified GCE.^[²²^] In each case, electrochemical studies of NAC were carried out at a wide range of pH (3–8), where the highest NAC anodic response was observed in acidic media at pH ∼ 4.45. Among these, the use of a modified or an unmodified AuE was shown to have the clearest response due to the interactions between the thiol group on NAC and the gold surface, whereupon electrooxidation, the NAC becomes adsorbed to the gold surface through the sulphur atom, which then combines forming the dimer N,N’-diacetylcysteine (Scheme 1).^[¹⁵^]

Details are in the caption following the image — **SCHEME 1**
Open in figure viewer PowerPoint

Electrooxidation of N-acetyl-L-cysteine to disulfide *N,N*’-diacetylcystine.

Nanostructures provide unique properties in terms of sensitivity compared to bulk materials including gold. Gold nanoparticles can be prepared through the chemical reduction of gold hydrochlorate or tetrachloroaurate by using citrate, trisodium citrate or sodium borohydride.^[^23-25^] There is a continuous effort to avoid chemical additives and use electrochemical reduction methods to synthesise the gold nanoparticles directly onto the electrode surface via cyclic voltammetry,^[²⁶^] potentiostatic,^[²⁷^] multi-potentiostatic pulse^[²⁸^] or multi-galvanostatic pulse^[²⁹^] methods. Here, we build upon our prior art with respect to additive-free electrosynthesis of gold stars on commercial GCE^[³⁰^] to explore new NAC sensor designs where the nature of the underlying carbon materials significantly influences the electrodeposition process and hence the final shape of the micro-/nano-structures. Furthermore, that synthesis method showed the advantage of producing a highly effective gold surface compared to the commercially available gold nanoparticles. This work extends knowledge in this direction, exploiting different structures of gold for NAC electroanalysis (both in standard solution and pharmaceutical formulation) by using an oxidised form of boron and nitrogen co-doped carbon nano-onions (oxi-B,N-CNOs) as an underlying carbon nanomaterial for gold electrodeposition.

CNOs are concentric multi-layered fullerenes consisting of concentric graphitic shells^[³¹^] and morphology, size and physico-chemical properties are dependent on the synthesis process.^[³²^] Thermal annealing of detonation nanodiamonds is a common low-cost and high-yield methodology to produce quasi-spherical shaped CNOs composed of 4–6 concentric graphitic shells measuring between 4–6 nm in diameter.^[^33-38^] Thermal annealing has been shown to be an effective way to introduce heteroatoms such as B, N and S^[^{31, 39-41}^] onto the CNO surface, leading to a significant impact on their electrochemical properties.^[^{42, 43}^] Little has been reported about the electrochemical behaviour of B-doped-CNOs, however, a high capacitance was observed upon B doping^[⁴⁰^] attributed to the increased porosity of the CNO surface and the surface defects in the carbon network induced upon boron and/or nitrogen doping.^[^{44, 45}^] We have recently reported a co-doping process to yield BN-CNOs bearing nitrogen and boron, showing remarkable oxygen reduction reaction electrocatalytic activity.^[⁴¹^] There is a scarcity of reports regarding the combination of CNOs with metallic nanomaterials^[^35-37^] Such works investigated the deposition of nanoparticles onto pristine CNOs for durable oxygen reduction, ultra-high energy supercapacitors and enhanced field emission behaviour.^[^{35, 36}^] The CNOs used in previous studies were synthesised via different methodologies, including thermal annealing of nanodiamonds, and combustion of white and thin cotton as well as a methane cracking method.^[^35-38^] The CNOs were first activated followed by the treatment of the product with ethyleneglycol, deionised water and H₂PtCl₆ at 140°C, giving rise to Pt-CNOs.^[³⁶^] A composite was also prepared by dropwise addition of H₂PtCl₆ into a CNO suspension followed by ultrasonication and ageing.^[³⁷^]

The hydrophilicity of the B,N-CNO increases significantly upon introducing carboxylic acid groups by acid treatment, resulting in their oxidised forms (oxi-B,N-CNOs).^[⁴⁶^] The acid treatment also leads to the creation of some B and N deficient sites on the first layer of the onion structures.^[⁴⁷^] In this context, it is interesting to investigate the influence of the layer defect structures and surface functionality of oxi-B,N-CNOs on the gold electrodeposition and nano-cuboid structure evolution of gold particles (AuNCBs). This surface architecture proved advantageous for thiol electrooxidation and was exploited here for the first time determination of NAC over the relevant concentration range, with an extension to real sample analysis.

2 EXPERIMENTAL

2.1 Materials and reagents

N-acetyl-L-cysteine (C₅H₉NO₃S, 99%), sodium acetate (CH₃COONa, 99%), glacial acetic acid (CH₃COOH), sulfuric acid solution in water (H₂SO₄, 96%), 5 M hydrochloric acid solution in water (HCl), chloroauric acid (HAuCl₄, 99.4%), absolute ethanol (C₂H₅OH, 99.8%), potassium chloride (KCl), potassium ferricyanide (K₃[Fe(CN)₆], 99%), potassium hexacyanoferrate trihydrate (K₄[Fe(CN)₆].3H₂O, 99.5%), N-acetyl-L-cysteine commercial sample (Holland and Barratt) (0.6 g per capsule). B,N-CNOs were synthesised via thermal annealing of DNDs in the presence of H₃BO₃ and further functionalised with oxygen-containing groups (oxi-B,N-CNO) following a previously reported oxidative procedure.^[^{32, 41, 47, 48}^]

2.2 Instrumentation

Electrochemical investigations were carried out using a three-electrode cell, with a GCE (0.07 cm²) or an AuE (0.032 cm²) as the working electrodes, a platinum wire as the auxiliary and an Ag/AgCl (3 M KCl) as the reference electrode. Measurements were made at room temperature (25 ± 2°C) using a Solartron SI2187 Electrochemical Interface and a CHI Instruments 600E Potentiostat/Galvanostat. Materials and reagents were weighed out using a Sartorius LA230S analytical balance, and a VWR ultrasonic cleaner model USC100T was used for nanoparticle suspension and electrolyte/analyte dissolution. A calibrated Jenway 550 pH meter was employed to measure the solution pH. The high-resolution scanning electron microscopy (HRSEM) was performed on an FEI (Thermo) Helios G4 CX Dual Beam SEM-FIB with Oxford Instruments EDS at the Bernal Institute, University of Limerick, Ireland.

2.3 Procedures

2.3.1 Preparation of carbon nano-anion-modified electrodes

A 1 mg/mL suspension of oxi-B,N-CNOs was prepared in absolute ethanol with ultrasonication for 45 min to ensure well-dispersed suspensions. The oxi-B,N-CNO suspensions were sonicated for a further 5 min immediately prior to use. A GCE was polished in a figure-of-eight pattern on a polishing pad with a 1 μm monocrystalline diamond suspension for 1 min and rinsed with a jet of deionised water. The electrode was then sonicated in deionised water for 30 s, rinsed in ethanol and allowed to dry. 10 μL of the oxi-B,N-CNOs suspension was drop-cast onto the GCE surface and the layer was dried using a heat lamp. The oxi-B,N-CNOs loading on GCEs was optimised by monitoring the anodic peak current of cyclic voltammogram (CV) for 1 mM NAC in 0.1 M sodium acetate buffer (pH 4.45) on the modified GCE prepared by drop-casting 10 μL of 1.00, 0.10, 0.05 and 0.02 mg/mL suspension of oxi-B,N- CNOs in absolute ethanol (Figure S1).

2.3.2 Electrodeposition of AuNCBs on oxi-B,N-CNOs/GCE and bare GCE

A polished bare or oxi-B,N-CNO modified GCE was decorated with electrodeposited AuNCBs of monodispersed size using a multi-pulse potentiostatic sequence. The AuNCBs were deposited from a 1 mM HAuCl₄ solution in 0.1 M HCl. Test solutions were purged with high-purity nitrogen for 10 min prior to use. Three potentials and pulse durations were selected from the respective CVs and chronoamperograms (CAs) of 1 mM HAuCl₄ in 0.1 M HCl as discussed in the next section. The strategy to employ the pulses arises from prior know-how and the objectives of this work. The modified electrodes were carefully cleaned with deionised water prior to performing electroanalytical experiments.

2.3.3 Electrochemical characterisation of AuNCBs/GCE and AuNCBs/oxi-B,N-CNOs/GCE

The electrodes modified with AuNCBs were studied by electrochemical impedance spectroscopy (EIS) in 5 mM [Fe(CN)₆]^3-/4− in 0.1 M NaOH at 0.221 V (for AuNCBs/GCE) and 0.207 V (for AuNCBs/oxi-B,N-CNOs/GCE), respectively over a frequency range of 0.01–100,000 Hz with a potential amplitude of 5 mV. The electrodes modified with AuNCBs were also studied by cyclic voltammetry at different scan rates (10–200 mV/s) over the range of 0.4–1.5 V in 0.1 M H₂SO₄ solution.

2.3.4 Electrochemical studies of NAC

NAC standards were prepared in 0.1 M sodium acetate buffer (pH 4.45) and electrochemistry was studied at the modified electrodes via cyclic voltammetry at different scan rates. A calibration curve was plotted with charge recorded for 5 s using constant potential coulometry at an applied potential of 1.15 V versus Ag/AgCl. The measurements were recorded in triplicate in 5 mL of the background electrolyte with 5 s quiet time between each measurement over the range of 0.2–2.53 mM NAC. The charge was measured at 4.5 s from each transient and averaged over the three replicates of each experiment. These calibration studies were carried out in both 0.1 M H₂SO₄ and in 0.1 M sodium acetate buffer (pH 4.45), ultimately opting to utilise the sodium acetate buffer (pH 4.45) in the electroanalysis of NAC in a commercial formulation.

2.3.5 NAC recovery from a pharmaceutical formulation

A commercially available NAC solid dose tablet containing 80% w/w (600 mg label content) was employed. Analytical samples containing 1 and 2 mM NAC were prepared by weighing out the calculated amount of tablet powder, followed by dissolution in sodium acetate buffer (pH 4.45) with filtering to remove insoluble excipients. 5 mL of the prepared sample was then measured, and standard addition was performed via chronocoulometry over the range of 0.199–0.99 mM. The sample concentration and subsequent recovery was then estimated (n = 3).

3 RESULTS AND DISCUSSION

3.1 Electrodeposition of AuNCBs on oxi-B,N-CNO/GCE and bare GCE

A strategy for electrodepositing discrete clusters of AuNCBs was developed by decoupling the nucleation and growth mechanism of gold electrodeposition on oxi-B,N-CNO/GC. For comparative purpose, the same method was employed for the electrodeposition of gold on GC to understand the differences in the morphological and electroanalytical performances.

Figure 1 shows the CVs of 1 mM HAuCl₄ in 0.1 M HCl on (A) oxi-B,N-CNO/GCE and (B) GCE at a scan rate of 25 mV/s in the potential window −0.2 V and 1.4 V for five consecutive cycles starting from 0.7 V toward the negative scan direction. The onset and cathodic peak of gold deposition on oxi-B,N-CNO/GCE were observed at 0.40 and 0.21 V, respectively in the first cycle, whereas the onset and anodic peak of gold dissolution were observed at 0.90 and 1.175 V, respectively (Figure 1A). However, all of the deposited gold did not dissolve during the anodic process and thus the onset and cathodic peak of gold deposition progressively shifted to 0.71 and 0.55 V, respectively, in the successive potential cycles.

A similar phenomenon, but at different potentials, was observed for the electrodeposition of gold on GCE under similar conditions (Figure 1B). The onset and cathodic peak of gold deposition on GCE were observed at 0.40 and 0.56 V, respectively in the first cycle, whereas the onset and anodic peak of gold dissolution were observed at 0.90 and 1.12 V, respectively (Figure 1B). Once more, not all deposited gold dissolved during the anodic process and thus the onset and cathodic peak of gold deposition progressively shifted to 0.65 and 0.47 V, respectively, in the successive potential cycles. Hence, it can be understood from Figure 1 that the reduction of HAuCl₄ on the gold seeds deposited on oxi-B,N-CNO (0.02 mg/mL)/GCE became more energetically favourable compared to the same process on the gold seed containing GC surface, although the reverse preference was true for the initial deposition of gold on these surfaces. This might be attributed to the concentric onion-type porous structure, acid functional groups along with the dopant depleted defects on the first layer of the oxi-B,N-CNO compared to the more uniform GCE surface.

Towards the objective of the decoupled nucleation and growth phenomenon, a very high overpotential (at −0.13 V as shown in Figure 1) was selected to instantaneously nucleate the gold seeds on the active sites of the oxi-B,N-CNO/GCE or GCE surface. A potential (at 0.92 V as shown in Figure 1), very near to the onset potential of the gold dissolution on both surfaces was selected to strip some of the under-critical size gold seeds from the surfaces while simultaneously creating some active sites on the deposited gold seeds themselves. Then, 0.55 V (as shown in Figure 1) was chosen for the controlled growth of the gold structures.

Besides the energies of the electrodeposition process as discussed above, understanding the kinetics of the same is very important to achieve the final size, shape and population of the gold structures on these materials. CAs, as shown in Figure 2, represent the kinetics of the electrodeposition of gold on (A,B) oxi-B,N-CNO/GCE and (C,D) GCE at −0.13 V (A,C) and 0.55 V (B,D) from 1 mM HAuCl₄ in 0.1 M HCl.

The fast decrease of current at the initial short time span represents the interfacial rearrangement upon application of the deposition potential. The discharge current was also overlapped with the cathodic current of the Au(III) reduction to Au(0). Then the cathodic current started to rise to a maximum current (i_m) at time (t_m) during the continuous increase in the effective surface area owing to the formation of discrete gold seeds and hemispherical diffusion flux of HAuCl₄ around those. At maximum stages (i.e., at t_m), the hemispherical diffusion flux around adjacent gold seeds overlapped with each other, resulting in linear diffusion flux of the same toward the electrode. Beyond this time, the cathodic current of the electrodeposition process followed normal Cottrell behaviour. The t_m and i_m for the electrodeposition of gold on oxi-B,N-CNO/GCE at −0.13 V were found as 0.17 s and −0.34 mA, respectively (Figure 2A). Hence, 0.06 s appeared to be sufficient to nucleate discrete gold seeds on the oxi-B,N-CNO/GCE at −0.13 V. On the other hand, the cathodic current slowly but steadily increased, but no peak appeared up to 100 s for the electrodeposition of gold on oxi-B,N-CNO/GCE at 0.55 V (Figure 2B). The longer t_m and lower i_m on oxi-B,N-CNO(0.02 mg/mL)/GCE compared to that on GCE (0.045 s and −0.79 mA) indicated slower kinetics of gold nucleation on oxi-B,N-CNO/GCE at −0.13 V (Figure 2C). Furthermore, the gold electrodeposition process on GCE became sluggish at 0.55 V representing t_m and i_m 4.84 s and −0.073 mA, respectively (Figure 2D).

The presence of pre-nucleated gold seeds (Au-Seed) on oxi-B,N-CNO/GCE and GCE made the growth process (at 0.55 V) faster, where the i_m of −0.039 mA appeared at t_m 13.6 s for Au-Seed-oxi-B,N-CNO/GCE (Figure 3A) and overlapped with the double layer discharge current on Au-Seed-GCE (Figure 3B). Therefore, it was decided to use 1 s for the growth of gold seeds avoiding the overlaps of hemispherical diffusion flux neighbours. Hence, the AuNCBs were electrochemically prepared on oxi-B,N-CNO/GCE from the solution of 1 mM HAuCl₄ in 0.1 M HCl by utilising the following decoupled pulse strategy (see Table S1):

\begin{eqnarray*} &&\mathrm{Nucleation}\left(1\, \mathrm{cycle}\right)\!\!: {\mathrm{E}}_{1}=0.91\,\,\mathrm{V},\, {\mathrm{t}}_{1}=10\,\,\mathrm{s},\\ &&\quad{\mathrm{E}}_{2} =-0.13\,\,\mathrm{V},\, {\mathrm{t}}_{2} = 0.06\,\,\mathrm{s}, {\mathrm{E}}_{3}=0.92\,\,\mathrm{V},\, {\mathrm{t}}_{3} = 0.005\,\,\mathrm{s}.\\ &&\mathrm{Growth}\left(749\, \mathrm{cycles}\right)\!: {\mathrm{E}}_{1}= 0.55\,\,\mathrm{V}, {\mathrm{t}}_{1} = 1\, \mathrm{s},\\ &&\,\,\,\,\,\,{\mathrm{E}}_{2}=0.92\,\,\mathrm{V},\, {\mathrm{t}}_{2}\ =\ 0.005\, \mathrm{s}. \end{eqnarray*}

The same parameters were employed for the GCE electrode in order to understand the differences in the morphological and electroanalytical performances.

3.2 Morphology of the modified electrodes after gold electrodeposition

The HRSEM images of the AuNCB@oxi-B,N-CNO/GCE showed that AuNCBs of aspect ratios 1.7 ± 0.4 with average height and width of 170 ± 30 nm and 100 ± 10 nm, respectively (calculated over 20 independent structures shown in the inset of Figure 4A), were grown over coral-like aggregates, irregularly distributed over oxi-B,N-CNO/GCE (Figure 4A). The drop-casted oxi-B,N-CNOs were irregularly distributed on GCE and thus the distribution of coral clusters of AuNCBs represents the preferential formation of clusters on those sites populated with oxi-B,N-CNOs. These uniformly sized finger-like protrusions were consistent in size, providing high-density active sites for Au-thiol interactions.

On the other hand, a high population of bud-like gold particles (inset of Figure 4B) were observed randomly, but uniformly distributed on the GCE surface (Figure 4B). Using an analysis over 435 clusters, it could be understood that clusters with a diameter of 300–319 nm accounted for 33% of the overall population followed by 320–339 nm (16.6%) and 280–299 nm (7%) indicating that the cluster sizes were uniform across the GCE surface. Clusters of average diameter < 279 nm were present and accounted for 12% of the distribution while clusters of average diameter > 339 nm accounted for 30% of the size distribution. Many of the larger clusters appeared to be formed from the overlap and aggregation of neighbouring clusters resulting in longer linear chain-like structures. From this result, it could be understood that the nature of the underlying surface plays a crucial role in the shape, size and population of clusters grown under similar conditions.

3.3 Outer sphere electron transfer properties of the modified electrodes after gold electrodeposition compared to bare GCE

The outer sphere electron transfer reaction of [Fe(CN)₆]^3/4− (5 mM) was then used to further characterise the AuNCBs in 0.1 M NaOH. This type of investigation is usually carried out in 0.1 M KCl background electrolyte, however, the presence of Cl⁻ ions in solution was found to degrade the gold surface due to the high formation constant of [AuCl₄]⁻ which will in turn diffuse from the electrode surface and strip the Au off the electrode.^[^{21, 49, 50}^] This was not observed in the alkaline electrolyte and thus, was found more suitable to further study the electrochemical behaviour of the gold-modified GCEs. Based on the voltammograms as shown in Figure S2, the electron transfer to/from [Fe(CN)₆]^3/4− was sluggish at the bare GCE resulting in peak-to-peak separation (ΔE_p) of 0.666 V with significant response improvements at the modified electrodes with ΔE_p of 0.171 and 0.157 V for the AuNP/GCE and AuNCB@oxi-B,N-CNO/GCE respectively. The decrease in ΔE_p clearly showed that the barrier to electron transfer was significantly reduced at the Au-modified surfaces in comparison to the bare GCE^[⁴⁹^] and furthermore, AuNCB@oxi-B,N-CNO/GCE is a better material compared to AuNP/GCE.

EIS was also carried out in 5 mM [Fe(CN)₆]^3/4− over a frequency range of 0.01–100,000 Hz with an amplitude of 5 mV. From these data, information regarding the charge transfer resistance (R_ct) was obtained using Nyquist plots for bare and modified electrodes (Figure 5A1,2,B1,2). The bare GCE had a significantly larger R_ct value as expected (13,560.7 Ω relative to 357.5 Ω for AuNP/GCE and 305.3 Ω for AuNCB@oxi-B,N-CNO/GCE). Therefore, the AuNCB surface accelerated the [Fe(CN)₆]^3/4− electron transfer process (see Table 1 for relevant data). The Bode phase angle plot is shown in Figure 5A3,B3 and the maximum phase was determined as −72.7°, −38.0° and 35.9° for the bare GCE, AuNP/GCE and AuNCB@oxi-B,N-CNO/GCE, respectively, being reached at higher frequencies for the modified electrodes. This provides information on the behaviour of each electrode, where angles measured at 45° indicated high ionic permeability.^[⁵¹^] The slope of the Bode magnitude plot relates to the resistance and capacitance of the system^[⁵¹^] in question, with capacitive properties evident here at low and high frequencies in the case of the AuNP/GCE and AuNCB@oxi-B,N-CNO/GCE and decreased resistance (indicating good electron transfer) evident in the mid-frequency range. Partial surface coverage ϑ may be estimated using

\begin{equation}\vartheta \ = \ 1 - \frac{{{R}_{CT}}}{{R_{CT}^o}}\end{equation}

(1)

where

{R}_{CT}

and

R_{CT}^o

are the charge transfer resistances of the bare and modified GCE respectively, resulting in values of 36.1% and 43.4% for AuNPs and AuNCB@oxi-B,N-CNO.

TABLE 1. Comparison of the R_ct and capacitance obtained from the electrochemical impedance spectra for both simulated and experimental data obtained at the bare GCE, AuNP@GCE and AuNCB@oxi-B,N-CNO/GCE.

		R_ct (Ω)	Capacitance (F)
Bare GCE	Experimental	13,560	7.56 × 10⁻⁷
	Fitted	13,660	6.33 × 10⁻⁷
	% error	0.73
AuNP@GCE	Experimental	357.5	1.55 × 10⁻⁷
	Fitted	357.5	2.21 × 10⁻⁷
	% error	0.00
AuNCB@oxi-B,N-CNO/GCE	Experimental	305.3	1.31 × 10⁻⁶
	Fitted	304.1	1.74 × 10⁻⁶
	% error	0.39

The AuNCB@oxi-B,N-CNO/GCE was studied in 0.1 M H₂SO₄ by cyclic voltammetry over 0.4–1.5 V at different scan rates in the range of 10–200 mV/s. Figure 6 data indicates the adsorption-controlled behaviour of both anodic and cathodic processes, having a linear relationship with respect to scan rate. The AuNCB@oxi-B,N-CNO/GCE electroactive surface area was calculated by estimating the charge passed during the reduction process of gold oxide and comparing it to the gold charge density factor (340 μC/cm²).^[⁵²^] The resulting electroactive surface area of the AuNP@oxi-B,N-CNO/GCE was calculated as 0.111 cm².

3.4 Electrochemical determination of N-acetyl cysteine at the AuNCB@oxi-B,N-CNO/GCE

The availability of surface active Au sites was examined for the inner sphere electron transfer reaction of NAC at the AuNCB@oxi-B,N-CNO/GCE surface (Figure 7A) and a clear NAC oxidative response was observed at 1.05 V. It was also noteworthy that the Au reduction peak was suppressed in the presence of NAC. This was then studied over 10–200 mV/s and from plotting the scan rate against the peak height, it was deduced that the process was adsorption controlled with clear dependence on the scan rate and a 144 mV anodic shift of the NAC oxidation process over the scan rate range examined (Figure 7B).

NAC was then calibrated using two supporting electrolytes, 0.1 M H₂SO₄ and 0.1 M sodium acetate buffer (pH 4.45). These calibration studies employed constant potential coulometry at 1.15 V for 5 s (Figure 8 with analytical performance data summarised in Table 2), a facile method which gives scope for use of simple allied instrumentation suitable for portable/onsite measurements.^[⁵³^] First, NAC calibration studies were performed in H₂SO₄ at three electrodes in triplicate: at the bare AuE, the oxi-B,N-CNO/GCE and at the AuNCB@oxi-B,N-CNO/GCE with charge taken at t = 4.5 s. The oxi-B,N-CNO/GCE displayed the lowest sensitivity at 1.39 × 10⁻⁴ C·cm⁻²·mM⁻¹, followed by the bare AuE (6.25 × 10⁻⁴ C·cm⁻²·mM⁻¹) and the AuNCB@oxi-B,N-CNO/GCE (1.11 × 10⁻³ C·cm⁻²·mM⁻¹). NAC calibration in 0.1 M sodium acetate buffer (pH 4.45) followed at oxi-B,N-CNO/GCE and AuNCB@oxi-B,N-CNO/GCE. The oxi-B,N-CNO/GCE yielded a sensitivity of 8.18 × 10⁻⁵ C·cm⁻²·mM⁻¹ while the AuNCB@oxi-B,N-CNO/GCE resulted in sensitivity of 4.76 × 10⁻⁴ C·cm⁻²·mM⁻¹, representing an overall 6-fold increase in sensitivity at the AuNCB@oxi-B,N-CNO/GCE relative to the oxi-B,N-CNO/GCE.

TABLE 2. Comparison of the sensitivities, limits of detection (LODs) and limits of quantification (LOQs) of the oxi-B,N-CNO/GCE and the AuNCB@oxi-B,N-CNO/GCE in 0.1 M sodium acetate buffer (pH 4.45).

Electrode	Sensitivity (μC / cm².mM¹)	95% confidence limit (μC / cm².mM¹)	LOD (μM)	LOQ (μM)
oxi-B,N-CNO/GCE	75.5	17.5	770	2600
AuNCB@oxi-B,N-CNO/GCE	470.0	25.0	51	170

In both cases the AuNCB@oxi-B,N-CNO/GCE yielded the highest sensitivity with a limit of detection (LOD) of 51 μM while 770 μM was estimated in the case of the oxi-B,N-CNO/GCE, indicating that the AuNCB@oxi-B,N-CNO/GCE could detect NAC at concentrations 15 times lower than the oxi-B,N-CNO/GCE with the active AnNCBs facilitating this process, aided by the underlying carbon nanomaterial. Although the sensitivity was higher in the 0.1 M H₂SO₄, sodium acetate buffer solution was selected as the most suitable background electrolyte for follow-up sample recovery purposes.

Table 3 below sets out the scope of literature in relation to NAC-modified electrodes with a range of electrochemical analytical approaches (cyclic voltammetry, chronoamperometry, pulse voltammetry and impedance). The methods vary in complexity with respect to synthesised catalysts, and comparative studies specifically with respect to Au surfaces are limited, reflecting the novelty of this work with no such report of the use of carbon nano onions in this manner. Table 4 lays out the comparative analytical parameters for the optimal modified electrode in both sodium acetate and sulphuric acid-supporting electrolytes.

TABLE 3. Comparison of analytical performance data with respect to N-acetyl cysteine sensor literature.

Electrode design	Sensitivity	LOD	Reference
Catechol @carbon paste electrode	38.25 μA/mM	60 μM	Raoof et al.^[¹⁶^]
Copper nitroprusside @carbon paste electrode	30.2 μA/mM	41.8 μM	Acelino Cardoso de Sá^[¹⁷^]
CuO nanostructures on ITO	6862.2 μA/μM	1.2 × 10⁻³ μM	Mohamed Tunesi^[¹⁸^]
Multiwalled carbon nanotubes	0.00892 μA/μM	0.43 μM	Kuyumcu Savan^[²⁰^]
Acetaminophen on RuO₂ nanoparticles	Not mentioned	2.84 μM	Zare^[²²^]
rGO/Imidazole derivative	Not mentioned	61 nM	Benvidi^[⁵⁴^]
AuNCB@oxi-B,N-CNO/GCE	476 μC/ cm² mM¹	51 μM	This work

TABLE 4. Validation parameters for NAC determination at AuNCB@oxi-B,N-CNO/GCE in electrolytes examined in this work.

Parameter	Acetate buffer pH 4.45 (0.1 M)	H₂SO₄ (0.1 M)
Operating potential (V)	1.15	1.15
Linear concentration range (mM)	0.2–2.53	0.2–2.53
Slope (C/cm².mM¹)	4.70 × 10⁻⁴ ± 2.5 × 10⁻⁵	0.0012 ± 1.34 × 10⁻⁴
Intercept (C/cm²)	3.12 × 10⁻⁵ ± 4.1 × 10⁻⁵	-3.98 × 10⁻⁵ ± 7.99 × 10⁻⁵
Correlation coefficient	0.998	0.992
Limit of detection (μM)	51	21
Limit of quantitation (μM)	170	69
Time to result (s)	4.5 s	4.5 s
95% confidence limits for NAC in solid dose formulation (n = 6)	1.93 × 10⁻³ ± 1.51 × 10⁻⁴ C/cm²
%RSD sample analysis (n = 9)	5.4%

3.5 Analysis of NAC in a commercial formulation

Tablets containing 600 mg NAC were analysed according to the method described in procedure section 2.3.5 such that the resulting solutions contained 1 and 2 mM NAC. The samples were measured using coulometry via standard addition of 0.1 M NAC (raw data shown in Figure S3), with RSD 5.4% (n = 9) (Table 5). The resulting sample recoveries were between 89% and 106% with a percentage variability between 2.7% and 10.1%, verifying the analytical test capability of the sensor design.

TABLE 5. Analysis of NAC in samples prepared from the commercial formulations.

Sample number	Expected NAC concentration (mM)	NAC recovered (mM)	% Recovery of NAC
1	1.97	1.98	101 ± 9.7
2	1.98	1.76	89.1 ± 10
3	2.03	2.00	98.9 ± 2.7
4	0.99	0.97	98.1 ± 3.9
5	1.01	1.07	106 ± 7.5

4 CONCLUSION

Electrosynthesis of metal nanoparticles on GCE and nanocarbon-modified GCEs was achieved using a simple chemical additive-free multistep potentiostatic deposition approach with the creation of exposed nanostructures with active metal surfaces. Current-time transients were examined to achieve suitable heterogeneous phase formation of Au seeds on GCEs and oxi-B,N-CNO/GCE with nanocuboid clusters realised upon repeated growth cycles having high particle population, with interesting form and morphology as evident by HRSEM imaging. Some overlapping occurred with rings/chain formation making for a wider average size distribution (approx. 500–800 nm) relative to the AuNCBs formed in the absence of oxi-B,N-CNOs (300–330 nm). Key deposition parameters included the applied potential, its duration and sequence and optimal conditions enabled the proton-dependent irreversible electro-oxidation process of NAC at AuNCB@oxi-B,N-CNO, being superior to that of both GCE and bulk Au electrodes. Electroanalytical studies involved NAC quantitation in 0.1 M sodium acetate buffer (pH 4.45) resulting in sensitivity 4.76 × 10⁻⁴ C·cm⁻²·mM⁻¹, LOD 51 μM with a 6-fold sensitivity increase relative to oxi-B,N-CNO modified GCE surface alone, verifying the important role that AuNCB plays in this process. NAC quantitation from commercial formulations resulted in recovery (89%–106%) ± 6.8%, signifying the potential for use of this low cost, simple and fast electroanalytical technique complimentary to complex and expensive chromatographic methods being routinely used in the quality control of commercial drugs.

AUTHOR CONTRIBUTIONS

Eoghain Murphy: Methodology, investigation, data curation and visualisation and writing—original draft; Saurav K. Guin: Conceptualisation, data curation and visualisation, formal analysis, writing and review and editing; Alexandra Lapiy: Investigation; Adalberto Camisasca: Methodology and resources; Silvia Giordani: Methodology, resources and supervision; Eithne Dempsey: Conceptualisation, supervision, funding acquisition, project administration, writing and review and editing.

ACKNOWLEDGEMENTS

We acknowledge the assistance of Dr. Vasily Lebedev for HRSEM analysis at the Bernal Institute, University of Limerick, Ireland. Drs. Guin and Dempsey acknowledge Enterprise Ireland and the European Union's Horizon 2020 Research and Innovation Programme under the Marie Skłodowska-Curie Career FIT PLUS grant agreement no. 847402. Alexandra Lapiy acknowledges the Irish Research Council Postgraduate Scholarship Grant No. GOIPG/2023/3838.

Open access funding provided by IReL.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest.

Open Research

DATA AVAILABILITY STATEMENT

The raw data of the reported results will be made available on request.

Supporting Information

REFERENCES

1G. Youssef, M. A. Ali, N. Alaa, B. Makin, M. Waly, A. AbouSetta, Middle East Fertil. Soc. J. 2006, 11, 109.
Google Scholar
2M. Zafarullaha, M. Q. Li, J. Sylvester, M. Ahmadb, Cell. Mol. Life Sci. 2003, 60, 6.
10.1007/s000180300001
PubMedWeb of Science®Google Scholar
3V. Mokhtari, Afsharian, P., Shahhoseini, M., Kalantar, S. M., Moini, A., Cell J. 2016, 19, 11.
PubMedWeb of Science®Google Scholar
4E. A. Elgindy, A. M. El-Huseiny, M. I. Mostafa, A. M. Gaballah, T. A. Ahmed, Reprod. Biomed. Online 2010, 20, 789.
10.1016/j.rbmo.2010.03.001
CASPubMedWeb of Science®Google Scholar
5C. Gaucher, A. Boudier, J. Bonetti, I. Clarot, P. Leroy, M. Parent, Antioxidants 2018, 7, 62.
10.3390/antiox7050062
Google Scholar
6H. Sies, C. Berndt, D. P. Jones, Annu. Rev. Biochem. 2017, 86, 715.
10.1146/annurev-biochem-061516-045037
CASPubMedWeb of Science®Google Scholar
7S. R. Thom, J Appl Physiol (1985) 2009, 106, 988.
10.1152/japplphysiol.91004.2008
CASPubMedWeb of Science®Google Scholar
8R. N. Dilger, D. H. Baker, J. Anim. Sci. 2007, 85, 1712.
10.2527/jas.2006-835
CASPubMedWeb of Science®Google Scholar
9G. J. McBean, Antioxidants 2017, 6, 62.
10.3390/antiox6030062
PubMedWeb of Science®Google Scholar
10L. Arranz, C. Fernandez, A. Rodriguez, J. M. Ribera, M. De la Fuente, Free Radic Biol Med 2008, 45, 1252.
10.1016/j.freeradbiomed.2008.07.014
CASPubMedWeb of Science®Google Scholar
11S. De Flora, R. Balansky, S. La Maestra, FASEB J. 2020, 34, 13185.
10.1096/fj.202001807
CASPubMedWeb of Science®Google Scholar
12R. Borgohain, J. Yang, J. P. Selegue, D. Y. Kim, Carbon 2014, 66, 272.
10.1016/j.carbon.2013.09.001
CASWeb of Science®Google Scholar
13J. H. T. M. Ploemen, A. Van Schanke, B. Van Ommen, P. J. Van Bladeren, Cancer Res. 1994, 54, 915.
CASPubMedWeb of Science®Google Scholar
14W. Wu, G. Goldstein, C. Adams, R. H. Matthews, N. Ercal, Biomed. Chromatogr. 2006, 20, 415.
10.1002/bmc.583
CASPubMedWeb of Science®Google Scholar
15C. Barus, P. Gros, M. Comtat, S. Daunes-Marion, R. Tarroux, Electrochim. Acta 2007, 52, 7978.
10.1016/j.electacta.2007.06.065
CASWeb of Science®Google Scholar
16J. B. Raoof, R. Ojani, M. Amiri-Aref, F. Chekin, J. Appl. Electrochem. 2010, 40, 1357.
10.1007/s10800-010-0093-y
CASWeb of Science®Google Scholar
17L. L. P. Acelino Cardoso de Sá, U. de Oliveira Bicalho, D. Ribeiro do Carmo, Int. J. Electrochem. Sci. 2011, 6, 3754.
10.1016/S1452-3981(23)18286-1
Web of Science®Google Scholar
18M. Mohamed Tunesi, R. A. Soomro, R. Ozturk, J. Mater. Chem. C 2017, 5, 2708.
10.1039/C6TC04611H
CASWeb of Science®Google Scholar
19J. Zhang, Y. Chang, C. Dong, Surf. Eng. Appl. Electrochem. 2015, 51, 111.
10.3103/S1068375515020155
Google Scholar
20E. Kuyumcu Savan, Sens. Actuators B 2019, 282, 500.
10.1016/j.snb.2018.11.092
Web of Science®Google Scholar
21L. H. Wang, W. S. Huang, Sensors 2012, 12, 3562.
10.3390/s120303562
CASWeb of Science®Google Scholar
22H. R. Zare, F. Chatraei, Sens. Actuators B 2011, 160, 1450.
10.1016/j.snb.2011.10.012
CASWeb of Science®Google Scholar
23J. Kimling, Maier, M., Okenve, B., Kotaidis, V., Ballot, H., Plech, A., J. Phys. Chem. B., 2006, 110, 15700.
10.1021/jp061667w
CASPubMedWeb of Science®Google Scholar
24H. N. Verma, P. Singh, R. M. Chavan, Vet. World 2014, 7, 72.
10.14202/vetworld.2014.72-77
CASGoogle Scholar
25C. Deraedt, L. Salmon, S. Gatard, R. Ciganda, R. Hernandez, J. Ruiz, D. Astruc, Chem. Commun. 2014, 50, 14194.
10.1039/C4CC05946H
CASPubMedWeb of Science®Google Scholar
26N. E. Tarolla, S. Voci, J. Reyes-Morales, A. D. Pendergast, J. E. Dick, J. Mater. Chem. A 2021, 9, 20048.
10.1039/D1TA02369A
CASWeb of Science®Google Scholar
27N. D. Nikolić, K. I. Popov, L. J. Pavlović, M. G. Pavlović, Surf. Coat. Technol. 2006, 201, 560.
10.1016/j.surfcoat.2005.12.004
CASWeb of Science®Google Scholar
28S. K. Guin, J. S. Pillai, A. S. Ambolikar, A. Saha, S. K. Aggarwal, RSC Adv. 2013, 3, 17977.
10.1039/c3ra42198h
CASWeb of Science®Google Scholar
29S. K. Guin, K. Parvathi, A. S. Ambolikar, J. S. Pillai, D. K. Maity, S. Kannan, S. K. Aggarwal, Electrochim. Acta 2015, 154, 413.
10.1016/j.electacta.2014.11.126
CASWeb of Science®Google Scholar
30S. K. Guin, A. Lapiy, E. Dempsey, Electrochem. Sci. Adv. 2023, https://doi.org/10.1002/elsa.202300009
10.1002/elsa.202300009
Google Scholar
31Q. Abbas, R. Raza, I. Shabbir, A. G. Olabi, J. Sci. Adv. Mater. Devices 2019, 4, 341.
10.1016/j.jsamd.2019.07.007
Web of Science®Google Scholar
32A. Camisasca, S. Giordani, Inorg. Chim. Acta 2017, 468, 67.
10.1016/j.ica.2017.06.009
CASWeb of Science®Google Scholar
33O. Mykhailiv, H. Zubyk, M. E. Plonska-Brzezinska, Inorg. Chim. Acta 2017, 468, 49.
10.1016/j.ica.2017.07.021
CASWeb of Science®Google Scholar
34S. Lettieri, A. Camisasca, M. d'Amora, A. Diaspro, T. Uchida, Y. Nakajima, K. Yanagisawa, T. Maekawa, S. Giordani, RSC Adv. 2017, 7, 45676.
10.1039/C7RA09442F
CASWeb of Science®Google Scholar
35J. Yang, S. H. Kim, S. K. Kwak, H. K. Song, ACS Appl. Mater. Interfaces 2017, 9, 23302.
10.1021/acsami.7b04410
CASPubMedWeb of Science®Google Scholar
36S. R. Suryawanshi, V. Kaware, D. Chakravarty, P. S. Walke, M. A. More, K. Joshi, C. S. Rout, D. J. Late, RSC Adv. 2015, 5, 80990.
10.1039/C5RA12364J
CASWeb of Science®Google Scholar
37H. Lu, P. K. Eggers, C. T. Gibson, X. Duan, R. N. Lamb, C. L. Raston, H. T. Chua, New J. Chem. 2015, 39, 915.
10.1039/C4NJ01378F
CASWeb of Science®Google Scholar
38J. Bartelmess, S. Giordani, Beilstein J. Nanotechnol. 2014, 5, 1980.
10.3762/bjnano.5.207
PubMedWeb of Science®Google Scholar
39M. P. Thomas, N. Wanninayake, M. De Alwis Goonatilleke, D. Y. Kim, B. S. Guiton, Nanoscale 2020, 12, 6144.
10.1039/D0NR00335B
CASPubMedWeb of Science®Google Scholar
40O. Mykhailiv, K. Brzezinski, B. Sulikowski, Z. Olejniczak, M. Gras, G. Lota, A. Molina-Ontoria, M. Jakubczyk, L. Echegoyen, M. E. Plonska-Brzezinska, Chem. 2017, 23, 7132.
10.1002/chem.201700914
CASGoogle Scholar
41A. Camisasca, A. Sacco, R. Brescia, S. Giordani, ACS Appl. Nano Mater. 2018, 1, 5763.
10.1021/acsanm.8b01430
CASWeb of Science®Google Scholar
42L.-N. Han, X. Wei, Q.-C. Zhu, S.-M. Xu, K.-X. Wang, J.-S. Chen, J. Mater. Chem. A 2016, 4, 16698.
10.1039/C6TA05607E
CASWeb of Science®Google Scholar
43J. Goclon, Appl. Surf. Sci. 2020, 532, 147267.
10.1016/j.apsusc.2020.147267
CASGoogle Scholar
44B. K. Mutuma, B. J. Matsoso, D. Momodu, K. O. Oyedotun, N. J. Coville, N. Manyala, Nanomaterials 2019, 9, 446.
10.3390/nano9030446
CASPubMedWeb of Science®Google Scholar
45K. Takagi, K. Natsui, T. Watanabe, Y. Einaga, ChemElectroChem 2019, 6, 1683.
10.1002/celc.201801702
CASWeb of Science®Google Scholar
46J. Bartelmess, S. J. Quinn, S. Giordani, Chem. Soc. Rev. 2015, 44, 4672.
10.1039/C4CS00306C
CASPubMedWeb of Science®Google Scholar
47M. d'Amora, A. Camisasca, R. Arenal, S. Giordani, Nanomaterials 2021, 11, 3017.
10.3390/nano11113017
PubMedGoogle Scholar
48L. R. Cumba, A. Camisasca, S. Giordani, R. J. Forster, Molecules 2020, 25, 3884.
10.3390/molecules25173884
CASPubMedGoogle Scholar
49N. Elgrishi, K. J. Rountree, B. D. McCarthy, E. S. Rountree, T. T. Eisenhart, J. L. Dempsey, J. Chem. Educ. 2017, 95, 197.
10.1021/acs.jchemed.7b00361
Web of Science®Google Scholar
50X. Xu, A. Makaraviciute, J. Pettersson, S.-L. Zhang, L. Nyholm, Z. Zhang, Sens. Actuators B 2019, 283, 146.
10.1016/j.snb.2018.12.008
CASWeb of Science®Google Scholar
51N. O. Laschuk, E. B. Easton, O. V. Zenkina, RSC Adv. 2021, 11, 27925.
10.1039/D1RA03785D
CASPubMedWeb of Science®Google Scholar
52J. Y. Kim, M. G. Han, M. B. Lien, S. Magonov, Y. Zhu, H. George, T. B. Norris, N. A. Kotov, Sci. Adv. 2018, 4, e1700682.
10.1126/sciadv.1700682
PubMedWeb of Science®Google Scholar
53G. Halpin, S. McEntee, C. Dwyer, F. Lawless, E. Dempsey, J. Electrochem. Soc. 2022, 169, 037528.
10.1149/1945-7111/ac5e41
CASGoogle Scholar
54A. Benvidi, S. Dalirnasab, S. Jahanbani, M. D. Tezerjani, M. M. Ardakani, B.-B. F. Mirjalili, R. Zare, Electroanalysis 2016, 28, 1625.
10.1002/elan.201501076
CASWeb of Science®Google Scholar

Early View

Online Version of Record before inclusion in an issue

e2300027

Exploitation of carbon surface functionality toward additive-free formation of gold nanocuboids suitable for sensitive assay of N-acetylcysteine in pharmaceutical formulations

Graphical Abstract