Introduction

Electrochemiluminescence (ECL) is the process in which electrogenerated reactive species at working electrodes undergo high energy electron transfer reactions to form excited states that emit luminescent signals [1, 2]. As an analytical technique, ECL is a promising tool that is capable of sensing and detecting a wide variety of samples as a result of the chemiluminescent analysis (low background optical signal) and the electrochemical analysis (controlled electrochemical potential) [3]. ECL has advantages over conventional electrochemical methods such as high sensitivity, wide dynamic response concentration range, and simple instrumentation [46]. ECL has been utilized for the determination of a wide variety of analytes from different background including medical diagnostics [7, 8], food [9, 10], water [11], and the immune system [12]. There are several classic ECL luminophores such as tris(2,2-bipyridyl)ruthenium(II) (Ru(bpy)3 2+), luminol, lucigenin, 9,10-diphenylanthracene (DPA), peroxide oxalate, and quantum dots. Efforts have continued to develop ECL systems based on new luminescent materials. However, ECL systems based on water-soluble porphyrin as the luminophore are rare [13].

Porphyrins are a special kind of organic compound in nature as they play an important role in biological systems [14, 15]. There are many porphyrins and their metal complexes, such as heme, chlorophyll, vitamin B12, cytochrome P-450, and catalase. Electrochemical and photochemical properties of porphyrins have been of interest for a long time because of their important roles in biological processes such as photosynthesis, oxygen transport, and biocatalysis [1618]. Porphyrins are also potentially useful in the field of photomedicine, energy conversion, and catalysis [1921]. Naturally occurring porphyrin compounds are water soluble and have special physiological activity. People use synthetic porphyrins to simulate the properties of the natural porphyrin compounds [22]; research on water-soluble porphyrins is of particular importance.

The ECL of porphyrins was first reported by Bard et al. They reported that the ECL behavior of metalloporphyrins M(TPP) in methylene chloride (CH2Cl2) solution containing 0.1 M tetra-n-butylammonium perchlorate (TBAP) was generated via annihilation [23, 24]. Vogler’s group [25] observed ECL of the tetrakis(diphosphonato)diplatinate(II) (PtTPP) using annihilation methodology. Bolin and Richter illustrated the anodic ECL of 5,10,15,20-tetraphenyl-21H,23H-porphyrin ruthenium(II) carbonyl [Ru(TPP)(CO)] in the presence of tri-n-propylamine (TPrA) as a coreactant in acetonitrile [26]. However, these experiments were carried out in organic solvents. The ECL behavior of different porphyrin derivatives and complexes has been studied and used for the sensing of various analytes. In most cases, either a porphyrin derivative or complex was fixed on the electrode surface [2729] or organic solvent was used to carry out the electrochemical experiments [30]. The ECL behavior of meso-tetra(4-carboxyphenyl)porphyrin (TCPP) in an aqueous medium was explored by our group and it has been used for quantifying pheophorbide in spinach samples [31].

Copper is an essential trace element which plays very important role in humans and other animals [32]. Copper is a widely used metal ion in industry, and its potential applications in chemistry, biomedicine, and environmental studies are well known [33]. Because of excessive use and improper treatment of industrial effluents, copper is a common pollutants. Bioaccumulation of copper results in several diseases such as Wilson’s disease, Menkes syndrome, Alzheimer’s disease, and amyotrophic lateral sclerosis [34].

It is always more convenient and easy to carry out electrochemical experiments in an aqueous medium rather than organic solvents. The ECL behavior of TSPP, a water-soluble porphyrin, has not been reported. Hence, in this work, we demonstrate the cathodic ECL behavior of the TSPP system containing K2S2O8 as a coreactant in aqueous solution, and the possible ECL mechanisms are proposed. The electroreduced TSPP can readily react with a coreactant to produce an excited state which subsequently decays back to its ground state by emitting strong luminescence. Factors affecting the ECL response were studied carefully and optimized. The ECL signal of the TSPP/K2S2O8 system was significantly quenched by Cu2+. A highly sensitive and selective ECL method was therefore developed for the sensing of Cu2+ based on its interaction with the TSPP/K2S2O8 system. The findings open new avenues to construct ECL analytical systems based on TSPP as it is stable and cost-effective.

Materials and methods

Chemicals

All chemicals and reagents were of analytical grade and used as received without further purification. meso-Tetra(4-sulfonatophenyl)porphyrin (TSPP) (Fig. S1 Electronic Supplementary Material, ESM) was synthesized according to the reported method [35, 36]. K2S2O8 was obtained from Shuangshuang Chemical Co. (Yantai, China). CuSO4 was purchased from Shanghai Chemical Reagent Co. K2HPO4, KH2PO4, and KCl were obtained from Zhengzhou Tianzhihao Co. The 0.1 mol/L phosphate buffer (PB) was prepared by mixing 0.1 mol/L potassium dihydrogen phosphate (KH2PO4) with 0.1 mol/L dipotassium hydrogen phosphate (K2HPO4) containing 0.1 mol/L KCl as a supporting electrolyte. The pH of the buffer was adjusted to the desired value using potassium hydroxide (KOH) and phosphoric acid (H3PO4). Doubly distilled water was used throughout the experiments.

Apparatus

Cyclic voltammetry (CV) and ECL experiments were carried out on an MPI-A ECL detection system (Xi’ an Remex Analytical Instrument, China) equipped with a homemade ECL cell mentioned elsewhere [37]. A three-electrode system comprising a glassy carbon electrode (3 mm diameter, GCE) as working electrode, Pt counter electrode, and Ag/AgCl (saturated KCl) reference electrode was used to conduct electrochemical experiments. The voltage of the photomultiplier tube (PMT) was biased at 800 V during the whole process. Photoluminescence (PL) spectra measurements were performed on an LS-55 fluorescence spectrophotometer (Perkin Elmer). UV–vis absorption spectra were recorded on a UV-1102 UV–vis spectrophotometer (Tianmei Scientific Instrument, Shanghai, China). Infrared spectra were recorded on a Fourier transform infrared spectrometer (PTIR, NEXUS870, Nicolet Instrument Co. USA). The pHs of the solutions were measured using a Sartorius pH meter PB-10 (RenHe Instrument Co., Shanghai, China). An RE3000 rotary evaporation instrument (Shanghai, China) was used in the synthesis of TSPP.

Method

Prior to ECL measurements, the GCE was polished to a mirror finish successively with 0.3-μm and 0.05-μm α-Al2O3 powder on silk cloth. After each polish, the electrode was sonicated in doubly distilled water, ethanol, and distilled water then dried under nitrogen atmosphere. In this study, 0.1 M pH 7.4 PB was always employed as the supporting electrolyte. TSPP was dissolved in ultrapure water to prepare 2.0 × 10−5 M stock solution and stored at 4 °C.

ECL measurement

Typically, the ECL experiments were performed in 0.1 M PB (pH 7.4) containing 0.1 M KCl. Firstly, TSPP (100 μL of 2.0 × 10−5 M) and K2S2O8 (100 μL of 0.1 M) were added to a plastic tube, then they were diluted to 1000 μL with 0.1 M PBS. A 500-μL aliquot of the mixture was transferred to the ECL cell. A cyclic voltammetry (CV) scan was performed between the potential −0.3 V and −1.5 V at 100 mV s−1 scan rate to record the ECL signal.

ECL sensing of Cu2+

The aforementioned ECL measurement procedure was adopted for Cu2+ sensing. A known concentration of Cu2+ was added to the 500 μL mixture before running the CV to record the ECL response. The standard addition method was used to record the calibration curve.

Results and discussion

The synthesized TSPP was characterized by infrared spectroscopy using the KBr pellet method. The following signal assignments were made from the obtained IR spectrum (ESM Fig. S2): The absorption peak at 3447 cm−1 was due to porphine ring N–H stretching vibration. Peaks at 1635 and 1472 cm−1 were assigned to the benzene ring and the porphyrin backbone C=C stretching vibration. Peaks at 1190, 1126, 1040, and 638 cm−1 are due to SO3H group absorptions. Porphyrin ring C–H bending vibration resulted in the absorption peak at 1013 cm−1. The peak at 800 cm−1 was due to the phenyl ring C–H bending vibration. IR and 1H NMR (not shown) characteristics of the synthesized TSPP were found to be in good agreement with those of previous reports [3538]. Meanwhile, the PL spectrum of TSPP with the maximum emission wavelength at 413 nm was recorded as shown in ESM Fig. S3.

To verify whether reaction occurred between TSPP and K2S2O8 without electrolysis, UV–visible absorption spectra of the system prior to electrolysis were studied as shown in Fig. 1. The absorption spectrum of the TSPP/K2S2O8 system was exactly the sum of the individual spectra, which implied that K2S2O8 did not react with TSPP. Hence this showed that there is no chemical reaction between TSPP and K2S2O8 in the ground state [39].

Fig. 1
figure 1

UV–visible absorption spectra of a K2S2O8, b TSPP, c K2S2O8+TSPP in 0.1 mol L−1 PB of pH 7.4

CV and ECL behavior of TSPP

Cyclic voltammograms were recorded for the TSPP, K2S2O8, and mixture of these two as shown in Fig. 2. For TSPP a very weak cathodic peak was observed at −0.7 V. The cathodic peak at −1.0 V was due to the electrochemical reduction of K2S2O8 [6]. For the mixture of TSPP and K2S2O8, the peak potential of K2S2O8 was shifted to −1.2 V, while the cathodic peak potential of TSPP was unaltered. This indicates that TSPP hinders the K2S2O8 electrochemical reduction under the given conditions on the GC electrode.

Fig. 2
figure 2

Cyclic voltammograms of a TSPP, b K2S2O8, c K2S2O8+TSPP on bare glassy carbon electrode in 0.1 mol L−1 PB of pH 7.4

To understand the ECL behavior of the TSPP/K2S2O8 system, potential was swept in the presence of only TSPP or K2S2O8 and both TSPP and K2S2O8. No ECL signal was observed when either TSPP or K2S2O8 alone was present in the cell. But, mixture of TSPP and K2S2O8 resulted in strong ECL signal as shown in Fig. 3. Hence, from the obtained results, it was clear that a significant ECL signal will only be observed in the presence of both TSPP and K2S2O8.

Fig. 3
figure 3

ECL intensity vs. potential curves of a TSPP, b K2S2O8, c K2S2O8 + TSPP on bare glassy carbon electrode in 0.1 mol L−1 PB of pH 7.4. Peaks 1 and 2 were due to the electron injection from the working electrode to TSPP and reduction of S2O8 2−, respectively

ECL can be explained through two main mechanisms [2, 40, 41]: one is ion annihilation ECL and the other is coreactant ECL. In the present case, when the current reaches its maximum, the ECL intensity also reaches a peak value at −1.1 V. This indicates that K2S2O8 plays an important role of coreactant in the process of generation of ECL, while TSPP act as a luminophore.

It has been reported that electrochemically oxidized or reduced porphyrins can react with coreactants to produce ECL [2326]. The possible ECL mechanism for TSPP/K2S2O8 is elucidated as below: when the potential was scanned from −0.3 V to −1.5 V two peaks were observed, one at about −0.7 V and the other at −1.2 V. The peak at −0.7 V was due to the electron injection from the working electrode to TSPP to produce the negatively charged species TSPP•− (Eq. 1). Then TSPP•− and S2O8 2− interact to produce the excited state species, i.e., TSPP*, and relaxation of TSPP* to TSPP results in the ECL signal [42]:

$$ \mathrm{TSPP}+{\mathrm{e}}^{-}\to {\mathrm{TSPP}}^{\bullet -} $$
(1)
$$ {\mathrm{S}}_2{\mathrm{O}}_8{{}^2}^{-}+{\mathrm{TSPP}}^{\bullet -}\to \mathrm{TSPP}+{\mathrm{S}\mathrm{O}}_4{{}^2}^{-}+{\mathrm{S}\mathrm{O}}_4{{}^{\bullet}}^{-} $$
(2)
$$ {\mathrm{TSPP}}^{\bullet -}+{\mathrm{SO}}_4{{}^{\bullet}}^{-}\to {\mathrm{SO}}_4{{}^2}^{-}+\mathrm{TSPP}* $$
(3)
$$ \mathrm{TSPP}*\to \mathrm{TSPP}+\mathrm{h}\upupsilon $$
(4)

The second reduction in the ECL signal at about −1.2 V was due to reduction of S2O8 2− to produce the strong oxidant species SO4 •− (Eq. 6). Then electron transfer takes place between the produced SO4 •− and TSPP•−, which is already present in the solution, to produce the excited state TSPP, i.e., TSPP* (Eq. 7). Then TSPP* will return to its ground state, TSPP, by emitting the ECL signal. The series of reactions which lead to the ECL signal are as follows:

$$ \mathrm{TSPP}+{\mathrm{e}}^{-}\to {\mathrm{TSPP}}^{\bullet -} $$
(5)
$$ {\mathrm{S}}_2{{\mathrm{O}}_8}^{2-}+{\mathrm{e}}^{-}\to {{\mathrm{S}\mathrm{O}}_4}^{2-}+{{\mathrm{S}\mathrm{O}}_4}^{\bullet -} $$
(6)
$$ {\mathrm{TSPP}}^{\bullet -}+{\mathrm{SO}}_4{{}^{\bullet}}^{-}\to {\mathrm{SO}}_4{{}^2}^{-}+\mathrm{TSPP}* $$
(7)
$$ \mathrm{TSPP}*\to \mathrm{TSPP}+\mathrm{h}\upupsilon $$
(8)

The main difference between the two mechanisms is the origin of the intermediate species SO4 •−. Because of the direct electron injection from the electrode to S2O8 2−, the reaction rate of Eq. 6 is greater than Eq. 2, which leads to the ECL intensity of the peak at −1.2 V being twice as high compared to the peak at −0.7 V. As discussed in the literature [43, 44], SO4 •− has been shown to transfer electrons in the ECL reaction sequence. Scheme 1 shows the mechanism.

Scheme 1
scheme 1

ECL reaction mechanism of the TSPP/K2S2O8 system

Another possible pathway for the generation of TSPP* is described by Eqs. 9 and 10.

$$ \mathrm{TSPP}+{\mathrm{SO}}_4{{}^{\bullet}}^{-}\to {\mathrm{TSPP}}^{+}+{\mathrm{SO}}_4{{}^2}^{-} $$
(9)
$$ {\mathrm{TSPP}}^{+}+{\mathrm{TSPP}}^{\bullet -}\to \mathrm{TSPP}*+\mathrm{TSPP} $$
(10)

However, the ECL versus potential trace in Fig. 3 shows little or no ECL activity prior to oxidation of both TSPP and S2O8 2−, indicating that light emission via Eqs. 9 and 10 is less likely to happen.

Optimization study

The ECL signal from the TSPP/K2S2O8 system can be affected by several factors such as K2S2O8 and TSPP concentration, pH value, and scan rate. Although H2O2 and K2S2O8 can act as the ECL reductive–oxidative coreactants, their efficiency is quite different. As shown in Fig. S4 (ESM), TSPP exhibits much stronger ECL in the presence of K2S2O8 than in the presence of H2O2; therefore, K2S2O8 was chosen as the ECL coreactant in the following experiment. Figure 4a depicts the effect of K2S2O8 concentration on the ECL intensity. The ECL intensity increases rapidly with the increase in K2S2O8 concentration because more TSPP* was produced from oxidation of the negatively charged TSPP by the electrogenerated SO4 •−. However, an excessive concentration of coreactants will increase the amount of reagent [45]; 0.1 M was chosen as the appropriate concentration and used throughout the following experiment. Figure 4b displays the effect of the pH of the reaction medium on the ECL intensity. The ECL intensity increases with the increase in pH from 4.5 to 7.4. Further increase in the pH resulted in the decrease of ECL signal, and the strongest ECL signal was achieved at pH 7.4, which indicated that lower alkaline conditions are suitable for TSPP ECL emission. This may be due to more active SO3 groups directly linked to the porphyrin ring being formed in lower alkaline conditions [46], leading to more enhanced ECL. At too high pH, the strong oxidant SO4 •− is consumed via the scavenging effect of OH [47]. As a result, too low or high pH causes the decrease in the ECL intensity. Thus, pH 7.4 was chosen for this system as it would produce the stronger ECL emission. Figure 4c depicts the effect of TSPP concentration on the ECL intensity. The concentration of luminophore affects the intensity to a great extent. The ECL intensity increases with the increase of TSPP concentration and then decreased with the further increase. This can be attributed to the self-absorption effect of the luminophore [48]. As a result 2 × 10−5 mol L−1 of TSPP was chosen as the optimum value. ECL efficiency significantly depends on the rate of generation/annihilation of excited species. The effect of scan rate on the ECL intensity was investigated and displayed in Fig. 4d. It can be seen that the ECL intensity increases with increasing scan rate from 0.01 to 0.1 V/s and thereafter decreases. This observation was consistent with other ECL sensors. It was due to the formation a greater number of excited-state TSPP species within a short time span as well as the faster diffusion rate of the coreactant [49]. When the scan rate was higher than 0.1 V/s, the consumption of coreactant on the electrode interface would be much faster than the diffusion of the coreactant from the bulk solution to the electrode surface. Hence, it was clear that a scan rate greater than 0.1 V/s was unfavorable to the electrochemical reaction of the TSPP/K2S2O8 system. In the following work, 0.1 V/s was selected as the optimized scan rate. The influence of different potential windows of the system on the ECL intensity was studied, and the results are shown in Fig. 4e. When the potential is too negative, hydrogen production takes place as a result of the decomposition of water and produced bubbles that can disturb the determination of the ECL. So an excitation voltage in the scanning window of −0.3 to −1.5 V was selected.

Fig. 4
figure 4

Effects of (a) K2S2O8 concentration (a = 0, b = 0.02, c = 0.03, d = 0.04, e = 0.06, f = 0.08, g = 0.1), (b) pH of the reaction medium (a = 4.4, b = 6.4, c = 7.4, d = 8.4, e = 9.4, f = 10.4, g = 11.4), (c) TSPP concentration (a = 5.0 × 10−6 M, b = 8.0 × 10−6 M, c = 1.0 × 10−5 M, d = 1.5 × 10−5 M, e = 2.0 × 10−5 M, f = 2.5 × 10−5 M), (d) scan rate (a = 0.01 V/s, b = 0.05 V/s, c = 0.1 V/s, d = 0.13 V/s, e = 0.2 V/s, f = 0.4 V/s, g = 0.5 V/s), and (e) potential window (a = −0.3 to −1.6 V, b = −0.3 to −1.5 V, c = 0 to −1.5 V, d = 0.2 to −1.6 V) on the ECL intensity of the system on bare glassy carbon electrode in 0.1 mol L−1 PB

Application study

Operational stability is one of the main concerns for practical application of a sensor [50]; under the optimized conditions, a stable and strong signal was exhibited by the developed system for Cu2+ sensing. Figure 5 shows the ECL emission of TSPP on the bare glassy carbon electrode for 25 continues cycles recorded in the potential window −0.3 to −1.5 V in 0.1 M K2S2O8 (pH 7.4) at 0.1 V s−1 scan rate. The ECL signals were observed with relative standard deviation of 0.87 %. The results showed the superior stability of the ECL signal, suggesting its potential applicability as an ECL sensor.

Fig. 5
figure 5

ECL intensity vs. time plot for TSPP/K2S2O8 system under optimized conditions for continuous 25 cycles

The ECL signal response rate of the developed system to Cu2+ was then studied. The ECL intensity is quenched immediately as Cu2+ was added into the TSPP/K2S2O8 system, and remained stable for the following 1 h as shown in ESM Fig. S5. As the quenching effect of Cu2+ on the ECL signal is quite fast and stable, this suggests a broad potential application to the rapid detection of Cu2+ with no strict time control. Figure 6 presents the quenching effect of various concentrations of Cu2+ on the ECL signal of the developed probe. The ECL intensity of the TSPP in the presence of Cu2+ (I) was lower than in the absence of Cu2+ (I 0), and the ECL intensity decreased gradually with the increase in Cu2+ concentration. Even nanomolar Cu2+ could effectively quench the ECL of the TSPP/K2S2O8 system. The ECL intensity quenching and concentration of Cu2+ were correlated by plotting ln(I 0/I) against the added concentration of Cu2+, and a linear relationship was obtained from 5 to 160 nM (Fig. 6) with a detection limit of 1.56 nM (S/N = 3) (inset Fig. 6), which is lower than or comparable to other existing ECL sensors for Cu2+ [51, 52]. The linear relationship can be represented as ln(I 0/I) = 0.0055C + 0.1967 with a correlation coefficient of R 2 = 0.9986. The major characteristic of the proposed ECL sensor is that the reaction was carried out in aqueous solution, at room temperature; hence the sensor is ecofriendly and cost-effective. More importantly, it is expected that TSPP could be a new class of promising luminescent agent for ECL sensors.

Fig. 6
figure 6

ECL intensity vs. potential curves of TSPP/K2S2O8 system under optimized conditions in the presence of various concentration of Cu2+. Inset plot of ln(I 0/I) against the concentration of Cu2+, where I 0 and I are the ECL intensities in the absence and presence of Cu2+, respectively

In addition to sensitivity, selectivity is a very important issue when evaluating the performance of a sensing system. Hence, to assess the selectivity of the developed ECL sensor, we examined the ECL intensity changes in the presence of common metal ions as potential interfering agents. Figure 7 shows the ECL response of the TSPP/K2S2O8 system on the bare carbon electrode in the presence of various metal ions under the optimized conditions. Only Cu2+ dramatically quenches the ECL intensity of the TSPP/K2S2O8 system, whereas Pb2+, Fe2+, Ni2+, Co2+, Al3+, Zn2+, Mg2+, Ba2+, Cr3+, and Cd2+ ions showed negligible quenching effects. The TSPP/K2S2O8 system is therefore highly selective towards Cu2+ in the presence of other common metal ions. Though the reason for the selectivity of the TSPP system is not clear, the selectivity of porphyrin-based systems towards Cu2+ is not new [53, 54]. A probable reason for this behavior can be explained by two different mechanisms. One is that Cu2+ will form a covalent bond with the nitrogen atoms of pyrrole rings and dative bonds with the other nitrogen atoms of the porphyrin [55, 56]. It is well known that the photophysical properties of metalloporphyrins are entirely different from those of porphyrin; in other words, metalloporphyrins will alter the florescence of the original system [57]. Paramagnetism of the Cu2+ is the other possible reason that explains the selectivity. Paramagnetic Cu2+ may cause the reversible electron transfer in the system which may result in the quenching of fluorescence, or Cu2+ may cause electron spin-orbit coupling and transform the excited singlet state to a triplet state which quenches the fluorescence of the system by internal conversion [54]. Maybe any one or all the aforementioned pathways are true for the developed system, which explains both the mechanism and selectivity of the TSPP/K2S2O8 system through which it senses Cu2+.

Fig. 7
figure 7

Comparison of the quenching effect of various metal ions on the TSPP/K2S2O8 system. The quenching coefficient is determined in terms of (I 0 − I)/I 0, where I 0 and I are the ECL intensities in the absence and presence of metal ions, respectively

To demonstrate the applicability of the developed Cu2+ sensor, the performance of the proposed method was tested to detect the concentration of Cu2+ in different real-life water samples obtained from Yellow River at Lan Zhou, Gansu province, China and tap water from our laboratory. The Yellow River water samples were filtered through 0.22-μm membranes and then centrifuged at 12,000 rpm for 20 min to remove sediment and suspended solids before use [48]. The resultant water samples were further used to estimate the Cu2+ using the developed method. For the Yellow River water samples, experimental results showed the decrease in the ECL intensity, and its comparison with the standard calibration curve suggested the presence of Cu2+ ions. As shown in Table 1, the Yellow river water sample contains 9.43 ± 1.21 nM Cu2+, but Cu2+ ions are not detected in the tap water. The average recoveries of the spiked samples were in the range of 97.7 % to 102.7 %. These results indicate that the proposed method based on ECL is efficient and applicable for real sample analysis.

Table 1 Determination of Cu2+ ion in real-life water samples

Conclusion

The cathodic ECL behavior of TSPP with K2S2O8 as the coreactant in aqueous solution was explored. An ECL signal is generated upon sweeping the potential from −0.3 V to −1.5 V. Two cathodic peaks appeared at −0.7 V and −1.2 V as a result of the reduction of TSPP and K2S2O8, respectively. On the basis of the experimental studies undertaken, we proposed possible mechanisms. The chemical reaction between the electrochemically generated SO4 •− and TSPP•− will produce a highly excited species, i.e., TSPP*, which will return to its ground state by emitting its excess energy in the form of an ECL signal. The ECL quenching property of Cu2+ was explored and used to develop a highly sensitive and selective method. A possible mechanism for the quenching and selectivity of the system towards Cu2+ has been explained. The developed method was found to be applicable for the analysis of Cu2+ in tap and river water samples.