Co-reactant Free Novel Self-enhanced Solid-state ECL Platform of Polyluminol-gold Nanocomposite for Highly Selective and Sensitive Signal-on Detection of Hg2+ ion

Development of self-enhanced solid-state ECL platform creates a straightforward experimental design for the fabrication of point-of-care applications. Herein, we develop a promising method for self-enhanced solid-state ECL platform of polyluminol gold nanocomposite on glassy carbon electrode ((PL-Au)nano/GCE) via simple one step electrochemical deposition process without involving any additional co-reactants. The presence of AuNPs augments the electron transfer kinetics of PL and enhances the solid-state ECL intensity and promotes label free, excellent sensitivity, and selectivity to detect Hg2+ in physiological pH through signal-on mode. Unlike pristine PL/GCE, electrochemically co-deposited AuNPs in the (PL-Au)nano/GCE composite, enable the co-reactant accelerator by improving the catalytic activity of PL towards oxygen reduction reaction (ORR) yielding in-situ ROS (co-reactant) generation. Further, the ECL intensity of (PL-Au)nano/GCE composite, gradually increases by each addition of Hg2+ ion. This is because of the formation of an amalgamation of Au-Hg on (PL-Au)nano/GCE composite surface which further accelerate the yield of in-situ ROS and enhance the intensity of ECL. Whereas no ECL signals changes were observed for PL/GCE composite. The proposed self-enhanced solid-state ECL platform is selectively sensing the Hg2+ ion in the linear range of 0.3 nM to 200 nM with a detection limit of 0.1 nM. The demonstrated (PL-Au)nano/GCE platform might pave new avenues for further studies in the solid-state ECL platform which could be more useful in on-site monitoring of clinical bioassay and immunosensors.


Introduction
Electrogenerated-chemiluminesence (ECL) is a process of light emission which is due to an energetic interaction between electrogenerated species on electrode and electrolyte interface. 1 ECL has a great advantage over chemiluminescence (CL) because of low background signals, simple optical setup, good spatial-temporal control and versatility. 2,3 Recently, ECL becomes a popular and widely used analytical technique in clinical diagnostics, material science and environmental monitoring. 4,5 Even though Ru(bpy) 3 2+ based ECL-active luminophores systems have been reported in literature such as Ru(bpy) 3 2+ -TprA, 6 Ru(bpy) 3 2+ -C 2 O 4 2-7 and Ru(bpy) 3 2+ -H 2 O 2 , 8,9 luminol-H 2 O 2 10 based ECL systems has more attractive glance because of its non-toxicity, cost-e cient and high quantum yield with low oxidation potential. 11,12 The luminescent property of luminol-H 2 O 2 system mostly depends on its inherent chemical or electrochemical reaction through the intermediates of oxygen species like OH • , OH and O 2 • . 13,14 The major limitation of luminol's ECL property is hampered by its poor aqueous solubility, alkaline mediated ECL dependency. To overcome this, water soluble luminol derivatives were chemically synthesized with enhanced ECL intensity in aqueous electrolyte. 15 However, the adopted synthesis involves multiple steps, desires tedious reactions without scalability. Moreover, the co-reactant like H 2 O 2 is also not quite stable at room temperature which also suppresses the e ciency of ECL signal of luminal. 16 One can overcome this problem by generating the reactive oxygen species (ROS) by using co-reactant accelerators which can produces more ROS through the reduction of dissolved oxygen and enhances the stability of ECL signal. 17 In ECL, the co-reaction or co-reactant accelerators are playing a crucial role in boosting the ECL intensity of luminophore by dissociating the co-reactant into active radicals. 18 For example, the ECL intensity of Luminol in O 2 saturated electrolyte increased by 2-fold by the co-reactant accelerator strategy. 19 Because of mass transport limitation, homogeneous phase or solution based luminol system always inhibits the e ciency of ECL emission, impeding the low molecular detection of selected analytes. [20][21][22] Use of heterogeneous or solid-state ECL has several advantages such as a minimum amount of luminophores is su cient, simpli es the experimental setup, enhance more ECL intensity. [23][24][25] In this context, luminol is an aniline monomer derivative having an ECL inert -NH 2 group facilitating the electropolymerization in acidic solution to form a stable polyluminol (PL) lm on the electrode surface. 26 However, the PL lms have less electrical conductivity in neutral or alkaline solution, which need further improvisation for highly e cient luminol system. Despite of few attempts taken in functionalizing the luminol system in solution state with metal NPs, till now no efforts were taken to hybridize PL lms particularly at solid-state, which eventually have potential scope for point-of-care/on-site application.
Mercury ion (Hg 2+ ) is a heavy metal known for individual as well as environmental toxicity. 27 The presence of Hg 2+ in human body causes brain damage and other chronic diseases. 28

Electrochemistry and ECL measurements
Commercially available glassy carbon electrode (GCE) with 0.0732 cm 2 surface area serves as working electrode, platinum foil as the counter electrode and Ag/AgCl is used as reference electrode respectively. Cyclic voltammetry (CV) and potential step experiments were performed with an Autolab electrochemical workstation (EcoChemie, The Netherlands). The ECL along with CV signals is measured simultaneously with a photomultiplier tube (PMT, Hamamatsu H9305-04). The PMT was held at −500 V with a highvoltage power supply. The photo current generated at the PMT was converted to a voltage using an electrometer system (model 6517, Keithley, Cleveland,OH) and connected to the Autolab via an analog-todigital converter (ADC).

Characterization techniques
We used Field emission scanning electron microscope (FESEM, Supra 55 VP, Carl Zeiss), Energy dispersive X-ray (EDX, Oxford Instruments X-MAX, 20 mm²) analysis and Atomic force microscope (AFM, Agilent technologies 5500 instrument) technique in order to know the morphological changes and elemental presence after modi cation of GCE. The X-ray photoelectron spectroscopic (XPS) technique used to predict the state of elements present in the outermost part of composite by using Theta Probe AR-XPS System, Thermo Fisher Scienti c (U.K).

ECL spectrum
ECL spectrum is recorded using the optimized PMT voltage as 950 V and slit width is 20 nm in spectro uorimeter. Also, applied a constant potential pulse at 0. substrate. As seen in Fig. 1A, a sharp increase in peak current at 0.9 V for the 1 st cycle of CV which is due to the luminol oxidation where the radical polymerization of luminol starts and decreased in the peak current (at 0.9 V) during the subsequent electrochemical cycling. Moreover, there is a reversible redox peak growth which is due to the reduction (E Pc = 0.6 V) and oxidation (E pa = 0.7 V) of PL. The redox peaks current are gradually increases for initial few cycles (up to 10 cycles) and then exhibit stable redox response without further increase in redox peak current. Interestingly, the existence of HAuCl 4 with luminol causes incremental redox peak current density of PL during electrochemical cycling without change in inherent redox potential. In addition, a small reduction peak at 0.8 V observed which is due to the reduction of Au 3+ to Au 0 (Fig. 1B). To ensure this particular reduction reaction a control experiment with pristine HAuCl 4 was performed. To this only HAuCl 4 is dissolved in 0.5 M H 2 SO 4 without luminol yield at sharp reduction peak at exactly 0.8 V and decreased upon continuous electrochemical cycling (Fig. S1A).
After electrochemical cycling, the typical redox response of modi ed GCE (AuNPs/GCE) in 0.5 M H 2 SO 4 con rms the formation of AuNPs on GCE (Fig. S1B). The CV of typical stable redox response of PL/GCE and (PL-Au) nano /GCE clearly seen in the potential window of 0 to 1 V in 0.5 M H 2 SO 4 solution (Fig.   S2A,B), evidencing the deposition of PL. 36 However, the sweeping potential region is extended up to 1.7 V (Fig. S3) to validate the existence of typical redox characteristics of AuNPs in line with redox peaks of PL (at 0.7 V and 0.6 V). From the second cycle onwards the redox peak of PL is started to disappear with retention of AuNPs characteristic redox behavior, suggesting an over oxidation potential resulted from either leaching or deactivating the PL in 0.5 M H 2 SO 4. And also, the calculated AuNPs reduction peak charge in AuNPs/GCE (Fig. S1B) is almost 2-fold less when compared with the reduction charge of AuNPs in the (PL-Au) nano /GCE (Fig. S3). Moreover, the peak current and charge associated with oxidation (at 0.7 V) and reduction (at 0.6 V) peak of PL in (PL-Au) nano /GCE is also higher than that of pristine PL/GCE (Table S1). This is because of simultaneous growth deposition of AuNPs and PL lm on GCE.
The surface morphology of electrochemically deposited PL/GCE and (PL-Au) nano /GCE were studied by using FE-SEM (Fig. 1C-D). PL/GCE shows microstructures of polymer islands (Fig. 1C), on the other hand (PL-Au) nano /GCE exhibit a homogenous spherical nanostructures with an average size distribution of ~60 nm (Fig. 1D), evidencing the existence of nanocomposite on GCE. To elucidate the elemental compositions and surface chemistry of the nanocomposite on GCE surface an XPS study was performed.
Survey spectra of the prepared platform were presented in Fig. S4. The high resolution XPS spectra of Au 4f 7/2 and Au 4f 5/2 is denoted in Fig. 1E. lms an in-situ etching was performed using with XPS analysis. Fig. 1F illustrates the XPS spectrum of etched (PL-Au) nano /GCE with an ampli ed signal intensity of Au 4f 7/2 and 4f 5/2, revealing enhanced exposure of Au via stripping of PL. This was further supported by the existence of weaker O1s and N1s peaks (Fig. S5). To complement the surface topography and elemental structure an AFM and EDX spectral measurements were performed and the results are presented in Fig. S6. The EDX of PL/GCE is shows C, N and O elements (Fig. S6A) whereas Au along with C and O presents in (PL-Au) nano /GCE composite (Fig. S6B), which evidencing the PL and Au co-deposition on GCE surface. In addition, the AFM analysis is also displays macrostructure kind of morphology for PL/GCE (Fig. S6C), but nanostructured with spherical shapes presents in (PL-Au) nano /GCE composite (Fig. S6D) as similar with FE-SEM.
Electrical conductivity and charge transfer kinetics for the prepared (PL-Au) nano /GCE was measured using electrochemical impedance spectroscopy (EIS). The calculated charge transfer resistance (∆R ct ) derived from the Nyquist plot of PL/GCE and (PL-Au) nano /GCE is 4076 Ω and 276 Ω, respectively (Fig. S7). The lesser value of R ct clearly con rms that the (PL-Au) nano /GCE exhibit better electrical conductivity and charge transfer kinetic than the pristine PL/GCE.

ECL experiments
Unlike redox response of (PL-Au) nano /GCE in 0.1 M H 2 SO 4 (Fig. S2A&B), the simultaneously recorded CV showed an irreversible peak at -0.4 V and 0.6 V which is due to the dissolved oxygen reduction reaction (ORR) 37 and PL oxidation peak ( Fig. 2A.a) in O 2 gas saturated 0.1 M PBS (pH 7.4) during scanning from 0 to -0.8 V to +1 V. In contrast, the pristine PL/GCE showed oxidation peak at 0.6 V but the oxygen reduction peak observed at high cathode potential of -0.65 V ( Fig. 2A.b). The less cathode peak potential shift approximately -0.25 V and high reduction current density for dissolved O 2 reduction clearly indicates that the (PL-Au) nano /GCE is highly catalytic than pristine PL/GCE. To understand the role of AuNPs, the PL lm also deposited on polycrystalline gold surface (PL/pc-Au) which exhibit the redox response in O 2 gas saturated 0.1 M PBS (pH 7.4) ( Fig. 2A.c). Although the peak potentials of PL oxidation and ORR are similar to the (PL-Au) nano /GCE, the peak current density of PL oxidation and ORR is quite decreased. This further con rms the (PL-Au) nano /GCE shows superior electro-catalytic behavior towards ORR. The simultaneous record of ECL from these modi ed electrodes shown in Fig. 2B. As expected, high intense ECL peak was observed at 0.6 V where exactly an oxidation of PL peak exist ( Fig. 2A.a) in (PL-Au) nano /GCE. The relative intensity of observed ECL signal of (PL-Au) nano /GCE is almost three times higher than PL/GCE (Fig. 2B.b) and PL/pc-Au (Fig. 2B.c) composites. Under completely argon gas saturated 0.1 M PBS, the (PL-Au) nano /GCE shows the disappearance of ORR and ECL peak ( Fig.   2A.d&2B.d). All these results clearly con rm that, obtained ECL signals are totally dependent on the concentration of dissolved O 2 present in the electrolyte solution. The in-situ generated reactive oxygen species (ROS) from dissolved O 2 act as co-reactant which eventually react with PL anion to produce ECL signal. The ECL enhancement of (PL-Au) nano /GCE is perhaps due to the presence of AuNPs which behaves as co-reactant accelerator by producing the ROS radicals via ORR. In order to understand the signi cance of AuNPs on PL lm, the same PL lm electrodeposited in the presence of H 2 PtCl 4 and AgNO 3 with in the same experimental conditions of (PL-Au) nano /GCE (Fig. S8A,B). Interestingly, there is no ECL observed for (PL-pt) nano /GCE and (PL-Ag) nano /GCE composites (Fig. S8C,D). Owing to its well-known catalytic behavior of Pt in ORR, the reduction of O 2 to H 2 O occurs via direct 4epathway and it may not follow 2epathway which is necessary for H 2 O 2 and ROS generation. 38,39 Even though luminol and Ag has good interaction and enable excellent chemiluminescence activity in presence of H 2 O 2 . 40 The present experimental conditions doesn't show any ECL for (PL-Ag) nano /GCE suggesting (PL-Ag) nano /GCE lm could not act as ROS generator in solid-state ECL platform. Thus, the only AuNPs present in (PL-Au) nano /GCE has effective co-competence of co-reactant accelerator to generate more ROS thereby promoting the ECL activity. Further, the effect of potential window on ECL intensity was studied by varying the initial potentials. Fig. 3A&B shows the CV and corresponding ECL responses of (PL-Au) nano /GCE composite at various scan directions such as, 0 to 1 V, -0.3 to 1 V, -0.6 to 1 V and -0.8 to 1 V respectively in O 2 saturated 0.1 M PBS (pH 7.4). As illustrated in Fig. 3B, ECL intensities of (PL-Au) nano /GCE composite vary on change of potential scan direction. We observed high intense ECL signal for -0.8 to 1 V (Fig. 3B.d) scans direction, because at this particular direction maximum amount of ROS generated by reduction of O 2 (Fig. 3A.d). A small intense ECL signal is observed when we scan 0 to 1 V.
Overall, the ECL intensity of (PL-Au) nano /GCE composite gradually increases by changing the potential scan towards more negative to positive direction and concluded that, ECL of (PL-Au) nano /GCE composite is depends on the in-situ generated ROS. The ROS further oxidizes during anodic direction then reacts with poly luminol anion to emit light.

ECL stability
Since ECL signal of (PL-Au) nano /GCE composite is depends on the in-situ generated ROS, we performed a time based ECL transient experiments by changing the initial stepping potentials to study the ECL stability of ROS in O 2 saturated 0.1 M PBS (pH 7.4). Fig. 4A shows the ECL intensity vs. time transients of (PL-Au) nano /GCE composite at various initial potentials such as 0 (about 6 s) to 0.6 V (2.5 s), -0.3 (6 s) to 0.6 V (2.5 s), -0.6 (6 s) to 0.6 V (2.5 s) and -0.8 (6 s) to 0.6 V (2.5 s) respectively in O 2 saturated 0.1 M PBS (pH 7.4). The obtained ECL signals are quite stable even up to 18 consecutive cycles at various initial step potentials. Among different initial potential pulse, the potential pulse between -0.8 to 0.6 V, displays more intense and stable ECL signals which may be ascribed to the generation of more number of ROS at the interface of (PL-Au) nano /GCE (Fig. 4A).
The ECL spectrum of (PL-Au) nano /GCE (Fig. 4B) also performed to elucidate the wavelength of PL at various initial potentials in O 2 saturated 0.1 M PBS (pH 7.4). As seen from Fig. 4B, ECL signals were observed only at 430 nm which is consistent with photoluminescence spectrum. Further, the highest intensity of ECL spectrum is obtained for the pulse potential of -0.8 to 0.6 V, indicating the highest ROS generation in the speci ed potentials. This again complements the consistent of ECL transient experiment results (vide supra).

ECL study at various compositions
To validate the ECL experimental conditions and its associated effect of ECL intensity of (PL-Au) nano /GCE, the different concentrations of luminol and HAuCl 4 .3H 2 O were taken for electro-deposition (see in supporting information, section 2). From the experimental observation it is found that 1.5 mM of HAuCl 4 .3H 2 O and 1 mM luminol is su cient for deposition of (PL-Au) nano /GCE composite, which apparently enabled a high intense ECL signal (Fig. S9, S10).

Effect of Hg 2+ ion
It has been identi ed in the literature that, the AuNPs and Hg 2+ ion has strong speci c metallophilic interaction and form spontaneous Au-Hg amalgam which tunes the catalytic properties of AuNPs and accelerate the rate of decomposition of H 2 O 2 . 41 Such study can be escalated for various applications particularly in environmental pollutant monitoring, and detection of Hg 2+ as preservatives in vaccines.
Inspired from that herein, the effect of Hg 2+ ion with (PL-Au) nano /GCE by exploring the various Hg 2+ ion concentrations from 10 to 150 nM in O 2 saturated 0.1 M PBS (pH 7.4) at 0.1 V/s (Fig. S11A). Interestingly, the ECL intensity of (PL-Au) nano /GCE linearly increases with each addition of Hg 2+ ion and reaches maximum limit up to 150 nM, after that there is no much enhancement was observed. The observed enhancement of ECL intensity (signal on mechanism) during the addition of Hg 2+ ion is due to the a nity between AuNPs and Hg 2+ ion, leading to the formation of Au-Hg amalgam via aurophilic interactions. 41 As expected, the surface properties of AuNPs has been further changed upon interaction with Hg 2+ , and accelerate the in-situ generation of highly reactive ROS through catalytic reduction of dissolved oxygen. 42 The detailed ECL mechanism of (PL-Au) nano /GCE before and after Hg 2+ addition is shown in the scheme 1B. The calibration curve of ECL intensity versus Hg 2+ ion concentrations were represented in  (Fig. 5B). The detection limit or limit of detection (LOD) is obtained by using 3xstandard deviation/slope and the value is observed to be 0.1 nM and LOD is comparable with previous methods ( Table 1). It is worthy to note that the similar experiments were also performed by using PL/GCE in the presence of Hg 2+ ion at different concentrations and there is no change in ECL intensity of PL/GCE during the Hg 2+ addition (Fig. S12). This results undoubtedly con rms that (PL-Au) nano /GCE composite only capable of producing more ECL in the presence of Hg 2+ ion. Table 1 Comparison of Hg 2+ linear range and limit of detection with previous reported methods. The recovery of sample were also calculated in both tap water and serum indicated in table 2. From the observed results we can say that the proposed strategy to detect Hg 2+ in real sample analysis could be possible. The percentage recoveries of the analyte were in the range of 93 to 104 suitable for practical application (Table 2). Further, we compared the Hg 2+ recovery in tap water and serum samples by using Atomic absorption spectroscope (AAS) technique. The standard Hg 2+ ion solutions of 25, 50, 100, 150, 250, 300 and 500 nM were calibrated, the linear plot shown in Fig. S13 and the unknown concentrations of real samples were analyzed and shown in table S2. The obtained recovery (%) by AAS is consistent with adopted ECL method.

Conclusions
In conclusion, we developed a co-reactant free solid-state ECL strategy by electro deposition of polyluminol-gold on GCE surface and observed a stable ECL signal in O 2 saturated neutral buffer solutions. The observed ECL intensity effectively increases with addition of Hg 2+ ion even at picomolar range with good linear relationship. Even though the adopted methodology is simple and one pot synthesis procedure involved to prepare the self-enhanced solid state ECL platform. This is the rst time we utilized luminol-gold probe to detect Hg 2+ with signal-on ECL platform rather than quenching. Moreover, this co-reactant free novel solid-state self-enhanced ECL offers good recovery in the real sample analysis of Hg 2+ ion sensing and obtained results were comparable with standard spectroscopic technique of atomic absorption spectroscope.