2D Zn-Porphyrin-Based Co(II)-MOF with 2-Methylimidazole Sitting Axially on the Paddle-Wheel Units: An Efficient Electrochemiluminescence Bioassay for SARS-CoV-2.

High electrocatalytic activity with tunable luminescence is crucial for the development of electrochemiluminescence (ECL) luminophores. In this study, a porphyrin-based heterobimetallic 2D metal organic framework (MOF), [(ZnTCPP)Co2(MeIm)] (1), is successfully self-assembled from the zinc(II) tetrakis(4-carboxyphenyl)porphine (ZnTCPP) linker and cobalt(II) ions in the presence of 2-methylimidazole (MeIm) by a facile one-pot reaction in methanol at room temperature. On the basis of the experimental results and the theoretical calculations, the MOF 1 contains paddle-wheel [Co2(-CO2)4] secondary building units (SBUs) axially coordinated by a MeIm ligand, which is very beneficial to the electron transfer between the Co(II) ions and oxygen. Combining the photosensitizers ZnTCPP and the electroactive [Co2(-CO2)4] SBUs, the 2D MOF 1 possesses an excellent ECL performance, and can be used as a novel ECL probe for rapid nonamplified detection of the RdRp gene of SARS-CoV-2 with an extremely low limit of detection (≈30 aM).

Introduction

Electrochemiluminescence (ECL) is an electrochemically triggered chemiluminescence reaction where excited states are generated by electron transfer between reactive intermediates and radicals produced by redox of luminophores or coreactants.[ ] In general, the luminescent efficiency can be greatly improved by attaching the coreactants to the ECL system.[ ] However, due to the molecular diffusion in the solution, the electron transfer between coreactants and luminophores is ineffective, rendering great limitations for the enhancement of ECL intensity.[ ] Therefore, the integration of the coreactant's accelerators and luminophores can serve as an effective ECL amplification strategy, which can improve the generation and collision efficiency of active intermediates originated from coreactants and luminophores.[ ]

Recently, metal‐organic frameworks (MOFs) materials have emerged as strong contenders for enhanced ECL luminophores owing to the facile tunability of metal ions and organic ligands, large surface areas, and abundant open metal sites.[ ] Among the reported systems, porphyrin‐based two‐dimensional (2D) MOFs were considered as ideal candidates for enhanced ECL performances based on the following advantages: (1) the porphyrin itself is a luminophore because of its large conjugated ring; (2) the porphyrin ring decorated with coordinated groups can act as a suitable ligand for constructing 2D MOFs; (3) both the metal ion in the center of porphyrin ring and the metal ions in the MOFs can be judiciously selected for building homo‐ or heterobimetallic 2D materials; (4) the rich open metal sites in the 2D MOFs could be used to finely regulate the ECL properties without affecting the structure of the network. Up to now, two porphyrin‐based homobimetallic 2D MOFs, [Zn2(ZnTCPP)]·3H2O·2DEF and [Co2(CoTCPP)]·2H2O·4.75DMF (ZnTCPP = zinc(II) tetrakis(4‐carboxyphenyl)porphine; DEF = N,N′‐diethylformamide; DMF = N,N′‐dimethylformamide), have been reported.[ ] However, the construction of the porphyrin‐based heterobimetallic 2D MOFs remains a big challenge.

Herein, a novel porphyrin‐based heterobimetallic 2D MOF, [(ZnTCPP)Co2(MeIm)] (1), has been successfully designed and synthesized from the self‐assembly of ZnTCPP and Co(II) salts in the presence of 2‐methylimidazole (MeIm) at room temperature (Scheme ). MOF 1 has been fully characterized by ultraviolet–visible spectroscopy (UV–vis), Fourier transform infrared (FT‐IR), powder X‐ray diffraction (PXRD), scanning electron microscopy (SEM), transmission electron microscopy (TEM), X‐ray photoelectron spectroscopy (XPS), thermogravimetric analysis (TGA), electron paramagnetic resonance (EPR) spectrum, and fluorescence spectroscopy.

Scheme 1

Design and construction of MOF 1.

MOF 1 possesses well‐known paddle–wheel [Co2(‐CO2)4] secondary building units (SBUs), which are also a kind of electroactive modules. Combining the photosensitizers ZnTCPP and the electroactive [Co2(‐CO2)4] SBUs, MOF 1 can serve as an important “two‐in‐one” platform for efficient ECL amplification. Notably, the MeIm ligand axially coordinated on the paddle–wheel [Co2(‐CO2)4] SBUs is beneficial to the electron transfer between the low‐spin Co(II) ions and oxygen. Consequently, oxygen anionic radicals (O2 •−) can be efficaciously produced by the electroactive [Co2(‐CO2)4] modules. In addition, the axially coordinated MeIm ligand can also adjust the 2D MOF 1 to a suitable distance of layers without changing the structure of layer for easy entry of O2 and enough exposure of ZnTCPP. Integration of the sufficient exposure of photosensitizers ZnTCPP and the efficient generation of O2 •− via the electroactive [Co2(‐CO2)4] SBUs in a single platform synergistically enhances the ECL performance of MOF 1. Compared with the 2D MOF [Co2(ZnTCPP)] (2) without MeIm ligand that was also prepared by us for control experiments, the MOF 1 exhibits 3.4 times higher ECL intensity than that of MOF 2. The MOF 1 with an excellent ECL performance has been applied as a novel ECL probe for rapid nonamplified detection of the RdRp gene of SARS‐CoV‐2 with an extremely low limit of detection (LOD) of ≈30 aM. Furthermore, the ECL probe shows an excellent sensitivity and stability, which will expect to liberate the detection of SARS‐CoV‐2 from the burdensome polymerase chain reaction (PCR) amplification method.

Results and Discussion

Synthesis and Characterization of MOF 1

MOF 1 was prepared by mixing ZnTCPP with MeIm, followed by addition of Co(II) salt. In the FT‐IR spectrum of MOF 1, the νas(COO−) and νs(COO−) vibrations are separated by 152 cm−1, suggesting the bidentate‐bridging coordination of the carboxylate group (Figure S1A, Supporting Information).[ ] In the solid‐state UV–vis spectrum, an additional absorption in the range of 420–475 nm is observed, which corresponds to the d–d transition of six‐coordinated Co(III) (Figure S1B, Supporting Information).[ ] The XRD patterns of MOFs 1 and 2 in the (110) plane are consistent with those simulated from the reported homobimetallic 2D MOFs [Zn2(ZnTCPP)] and [Co2(CoTCPP)] (Figure ).[ ] Therefore, we believe that MOF 1 maintains the 2D structure assembled from ZnTCPP and paddle–wheel [Co2(‐CO2)4] SBUs. The changes of the other XRD peaks may ascribe to the variable distance of layers in the c‐direction due to the axial coordination of MeIm on the paddle–wheel [Co2(‐CO2)4] SBUs. Thus, the introduction of MeIm is expected to adjust MOF 1 to a suitable layer spacing without changing the layer structure, which is beneficial to the entrance of O2 and the exposure of ZnTCPP. The TEM (Figure 1B and Figure S3, Supporting Information) and SEM images (Figure S2, Supporting Information) show that the average size of the MOF 1 is 200 nm, which is reduced compared with the MOF 2 without MeIm. The smaller particle size of the MOF 1 could be attributed to the accelerated nucleation during the self‐assembly of Co(II) salts and the deprotonated ZnTCPP with the help of MeIm. The uniform nanostructure is beneficial to the exposure of the photosensitive ZnTCPP units in the MOF 1 to oxygen radicals.

Figure 1

A) XRD patterns of MOFs 1 and 2. B) HRTEM images of MOF 1. C) TGA of MOFs 1 and 2. High‐resolution XPS of: D) Co 2p region and E) O 1s region of MOFs 1 and 2. F) EPR spectra of MOFs 1 and 2.

The existence of MeIm in MOF 1 was evidenced by TGA (Figure 1C). Compared with the two weight loss processes of MOF 2 (loss of the substituents on porphyrin ring and collapse of the porphyrin skeleton),[ ] MOF 1 shows an additional weight loss (6.2%) during 270–384 °C due to the removal of MeIm (calcd. 7.8%).[ ] Considering that the decomposition temperature of MeIm in MOF 1 is higher than that (198 °C) of free MeIm, we can reasonably speculate that the MeIm in MOF 1 is coordinated to Co(II) ions, rather than merely physically encapsulated.

Then, coordination mode of the MeIm to Co(II) ions is further investigated. In the XPS spectra of MOF 1, the Co 2p3/2 region displays an additional characteristic peak of Co—N bond at 788.0 eV (Figure 1D),[ ] implying that MeIm acts as an axial ligand to sit on the paddle–wheel [Co2(‐CO2)4] SBUs, as also confirmed in the N 1s region (Figure S4, Supporting Information).[ ] The decreased peak for adsorbed oxygen in O 1s region of MOF 1 suggests that the water or methanol solvents on the open Co(II) sites are obviously reduced in contrast to MOF 2 (Figure 1E),[ ] revealing that MeIm and solvents are competing for axial coordination to Co(II) ions. Compared with MOF 2, the EPR spectrum of MOF 1 displays a characteristic signal at 1.87 (Figure 1F), exhibiting that the Co(II) ions in MOF 1 are five‐coordinate in a low‐spin state (t 2g 6 e g 1, S = 1/2) and MeIm is axial coordinated on the paddle–wheel [Co2(‐CO2)4] SBUs.[ ] Manipulating the electronic structures of Co(II) ions is expected to increase its binding activity for oxygen and stimulate electron transfer to generate metal‐bound superoxide (M‐O2 •−).[ ]

Effect of the Axial Coordinated MeIm

Then, we finely tuned the amount of axial‐coordinated MeIm on paddle–wheel [Co2(‐CO2)4] SBUs and investigated the effect on electrochemical activity. XRD patterns were used to monitor the results from the different amounts. When the amount of MeIm reaches 0.04 mmol (ZnTCPP:MeIm = 1:4), a new diffraction peak appears at 2θ = 5.24° (Figure S6, Supporting Information), which is due to the axial‐coordinated MeIm increasing the distance of layers in 1 along the c axis. When MeIm is 0.06 mmol (ZnTCPP:MeIm = 1:6), the diffraction peak shifted to 2θ = 5.20°, indicating that both the axial positions of paddle–wheel [Co2(‐CO2)4] SBUs are occupied by MeIm ligands, thereby further increasing the interplanar distance in the c‐direction (Figure S7, Supporting Information). The changes of XRD patterns prove that the numbers of axial‐coordinated MeIm can adjust the interlayer distance of 2D MOF 1, making the entrance of O2 easier and exposure of ZnTCPP more sufficient (Figure S8, Supporting Information). Furthermore, the effect of axial‐coordinated MeIm numbers on the electrochemical activity was investigated. Figure shows the cyclic voltammetry (CV) curves of different modified electrodes under O2‐saturated conditions. Regardless of the presence of MeIm, all samples exhibit a reduction activity toward O2 (Figure S9, Supporting Information). However, when MeIm is axially coordinated on the paddle–wheel [Co2(‐CO2)4] SBUs, a clear cathodic shift in the formal redox potential is observed, suggesting that the axial coordination of electron‐donating MeIm ligand can not only increase the electron density around Co(II) catalytic site but also decrease the barrier of electron transfer to form the metal‐bound superoxide (Co‐O2 •−). Consequently, increasing the amounts of MeIm from 0 to 0.05 mmol, the ECL intensity of MOF 1 is enhanced about 3.4 times (Figure 2B). However, the catalytic and ECL performance of MOF 1 decline sharply when the amount of MeIm is equal to 0.06 mmol (ZnTCPP:MeIm = 1:6), because if the concentration of MeIm is too high, both the axial positions of the paddle–wheel [Co2(‐CO2)4] SBUs will be occupied by MeIm ligands, so that there is no open metal sites available for combining oxygen. Therefore, the optimal amount of MeIm used for the synthesis of MOF 1 is 0.05 mmol (ZnTCPP:MeIm = 1:5). Under the optimal condition, MOF 1 exhibits an enhanced and stable ECL performance, which is much stronger than those of the homobimetallic MOFs, [(ZnTCPP)Zn2(MeIm)] and [(CoTCPP)Co2(MeIm)] (Figures S10 and S11, Supporting Information). The poor ECL performance of [(CoTCPP)Co2(MeIm)] is due to the single electron in the e g orbital in Co(II) quenching the ECL spectra when it acts as the central metal of the porphyrin ring. On the other hand, [(ZnTCPP)Zn2(MeIm)] shows poor catalytic activity toward oxygen, resulting in its ECL strength is only 17.6% of MOF 1. The decent ECL performance of MOF 1 can be ascribed to the following four reasons: (1) The axially coordinated MeIm ligand can adjust the 2D MOF 1 to a suitable distance of layers, making O2 more accessible and ZnTCPP more fully exposed; (2) The [Co2(‐CO2)4] SBU with an axially coordinated MeIm, as an electroactive site, shows a higher affinity and catalytic activity for O2; (3) MOF 1 under an oxygen‐containing environment can promote electron transfer of the low‐spin Co(II) ion (t 2g 6 e g 1) to O2; (4) The ordered arrangement of the electroactive [Co2(‐CO2)4] SBUs and the photosensitizers (ZnTCPP) effectively shortens the distance between O2 •− and porphyrin rings and accelerates the collision of free radicals.

Figure 2

A) Cyclic voltamograms of MOFs 1 and 2 with different concentrations of MeIm in 10 mmol L−1 HEPES (pH 7.4) containing 0.3 mol L−1 O2‐saturated KCl solution versus Ag/AgCl reference potential. B) ECL‐potential curves of MOFs 1 and 2 with different concentrations of MeIm in aqueous solution under O2‐saturated condition. C) Free‐energy diagram of O2 adsorption on active Co(II) sites in MOFs 1 and 2. D) The free‐energy profile for the reduction of oxygen to O2 •− pathway over MOFs 1 and 2.

To further understand the role of [Co2(‐CO2)4] SBU with an axially coordinated MeIm in activating oxygen, density functional theory (DFT) calculations were performed to investigate the Gibbs free energies and electronic distributions of MOF 1 with different amounts of MeIm during the reaction. The adsorbed O2 is bound to the [Co2(‐CO2)4] SBU in the form of Co—O—O. If there is only one axially coordinated MeIm on the [Co2(‐CO2)4] SBU, both the adsorption energy (E ads) of O2 and the energy required to form O2 •− in MOF 1 are lower than those in MOF 2, which explains the higher electroreductive activity of O2 in MOF 1 (Figure 2C,D). However, if there are two axially coordinated MeIm ligands on the [Co2(‐CO2)4] SBU, it is impossible for oxygen to be adsorbed on the Co(II) ions owing to lack of open metal sites, which account for the decreased catalytic activity of O2 when the amounts of MeIm reaches 0.06 mmol. Additionally, the desorption energy of O2 •− in MOF 1 is smaller than that in MOF 2 (Figure S12, Supporting Information), indicating that the O2 •− species generated in MOF 1 are more likely to be desorbed and collide with the porphyrin anion radicals to produce singlet oxygen (1O2) intermediates.

To further study the above process, the electron transfer between Co(II) ion and oxygen in MOF 1 was investigated. MOF 1 has a higher highest occupied molecular orbital (HOMO) energy level than MOF 2 ( HOMOMOF 1= −3.86 eV, HOMOMOF 2 = −4.90 eV), which shows that the Co(II) ion in MOF 1 is easier to transfer electron to O2 (Figure ). According to the charge distribution analysis (Figure 3B), the charge density over the Co—O—O skeleton within MOF 1 is higher than that of MOF 2. Based on the charge difference of absorbed O2, it can be seen that the O—O bond is more likely to be broken in MOF 1.[ ] On the basis of the Mulliken charge, more intuitive data can be obtained. After binding oxygen, the positive charge of the Co(II) ion in MOF 1 changes from 0.48 to 0.55e (Figure 3E,F), which means the Co(II) ion loses 0.07e (only 0.01e in MOF 2, Figure 3C,D), proving that the electron transfer path of [Co2+‐O2 → Co3+‐O2 •−] is more possible to occur in MOF 1. This inference can also be confirmed by the results from the XPS spectra of Co 2p (Figure 1D), which reveal that the proportion of Co3+ in MOF 1 is higher than that in MOF 2 (Co3+/Co2+ = 5.26 and 0.88 for MOF 1 and 2, respectively; Co3+: 782.8 eV, Co2+: 786.3 eV). By the way, MOF 1 in an oxygen‐containing environment can also promote electron transfer of the low‐spin Co(II) ion (t 2g 6 e g 1) to O2.

Figure 3

A) Schematic diagrams of energy level distributions of HOMO and lowest unoccupied molecular orbital (LUMO) for MOFs 1 and 2. B) The charge difference distributions of MOFs 1 and 2 after oxygen binding obtained by DFT calculation, the isosurfaces are set to 0.001 e bohr−3 (Yellow, charge accumulation; Purple, charge depletion). Mulliken charge distributions of: C,D) MOF 2 and E,F) MOF 1 before and after oxygen binding.

ECL Amplification Mechanism

To gain more insight into the mechanism of ECL amplification, the ECL and photoluminescence (PL) spectra of MOF 1 were measured. In contrast to the typical porphyrin PL peaks located at 610 and 660 nm, the ECL emission peak of MOF 1 locates around 634 nm, which is consistent with the emission band of 1O2 (Figure S14B, Supporting Information).[ ] To elucidate further which kind of intermediates is involved in the ECL amplification, we introduced benzoquinone (quencher of O2 •−)[ ] and sodium azide (NaN3, quencher of 1O2),[ ] respectively, into the ECL system. The results revealed that the ECL intensity is remarkably reduced in the presence of both 10 mM benzoquinone and NaN3 (Figure S15, Supporting Information). In the EPR spectra (Figure ), 12 hyperfine splitting peaks (corresponding to the superimposed spectra of HOO• and O2 •−) are captured at the O2 reduction potential (−0.6 V), while a group of three congruous Zeeman splits (1O2‐entrapped signal) is observed at the emission potential (−1.6 V).[ ] Based on the EPR analysis and ECL quenching study, we believe that O2 is first reduced to HOO• and O2 •− (Figure S16, Supporting Information), which is the basis of the ECL reaction. Subsequently, the porphyrin is gradually reduced to anionic radical (ZnTCPP•−) and acted as a photosensitizer to produce 1O2 by collision with O2 •−.

Figure 4

A) EPR spectra of MOF 1 with DMPO and O2 at −0.6 V (top); and with TEMP and O2 at −1.6 V (bottom). B) SWVs of MOF 1 in aqueous solution under O2‐saturated condition. C) Proposed mechanism for ECL amplification of MOF 1.

The redox processes of MOF 1 between 0 and −1.7 V were further explored by square wave voltammetry (SWV). MOF 1 undergoes three reduction processes at potential of −0.52, −1.20, and −1.51 V (Figure 4B), corresponding to the generation of HOO• and O2 •− (0.52 V) and ZnTCPP anion radical (ZnTCPP•− = −1.20 V, ZnTCPP2− = −1.51 V), respectively.[ ] Considering the low starting potential (−1.4 V) required for an ECL system, the ZnTCPP•− is an active intermediate that can collide with O2 •− to form 1O2 molecules. Due to the strong oxidizing property of HOO•when pH 7; φ(ZnTCPP/ZnTCPP•‐) = ‐1.00 V versus SHE, we speculate that HOO• may help the accumulation of ZnTCPP•− to achieve a cyclic amplification of ECL signals. Specifically, considering the efficiency of reduction reaction on the electrode is always not 100%, the unreduced ZnTCPP•− can be oxidized to ZnTCPP by highly oxidizing HOO•, thereby changing the equilibrium constant of the reaction (ZnTCPP + e‐ → ZnTCPP•−) and increasing the amount of ZnTCPP•− produced in this reaction. After the cyclic amplification process, the fully accumulated ZnTCPP•− will collide with O2 •− to generate 1O2 molecules.

Based on the above results, we propose a possible ECL reaction mechanism as shown in Figure 4C and the following Equations (1)–(8). First, O2 is reduced to O2 •−, while the low‐spin Co2+ ion is oxidized to Co3+ ion (Equation 1). Then, Co3+ is immediately electroreduced to Co2+ (Equation 2). The resulting O2 •− species, on the one hand, abstracts proton from water to generate HOO• and OH− (Equation 3), and on the other hand, collides with the ZnTCPP•− from the electrochemical reduction of ZnTCPP (Equation 4) to produce 1O2 (Equation 6).[ ] In addition, the ZnTCPP•− can also undergo an annihilation reaction with HOO• to give ZnTCPP and HO2 − (Equation 5). Finally, the dimerization of 1O2 produces the (1O2)2* species in an excited state (Equation 7), which returns to the ground state with a light emission (Equation 8).[ , ]

Nonamplified Detection of SARS‐CoV‐2

Recently, a few groups have worked to develop ECL biosensors for early diagnosis and timely detection of SARS‐CoV‐2.[ ] However, the constant mutation of SARS‐CoV‐2 has caused higher viral loads with a faster spread and greater infectivity of the mutant strains. Therefore, it is urgent to develop a rapid and accurate detection method for SARS‐CoV‐2.

Benefiting from the excellent ECL performance, the MOF 1 has been applied to the detection of RdRp gene of SARS‐CoV‐2 with nonamplifing mode. Specifically, MOF 1 was coated by chitosan (CS) and activated with glutaraldehyde (GA) to react with pDNA‐1 (P1). The RdRp gene (tDNA) and bovine serum albumin (BSA)‐modified pDNA‐2 (P2‐BSA) were then ligated by complementary base pairing (Scheme ). Since BSA blocked the interfacial electron transfer from the proteins, the RdRp gene was detected by signal off mode (Figure S18, Supporting Information).

Scheme 2

ECL biosensor for the detection of RdRp‐COVID DNA sequence (CS: chitosan; GA: glutaraldehyde; P1: pDNA‐1; P2‐BSA: bovine serum albumin (BSA)‐modified pDNA‐2; tDNA: RdRp gene).

Electrochemical impedance spectroscopy (EIS) confirms the successful construction of the proposed ECL biosensor (Figure ). In the concentration range of 100 pM–100 aM, the ECL intensity has a linear relationship with the logarithm of the target DNA concentration (Figure 5B,C). The linear equation is y = 141.86 lg(C Target) – 47.31 with a correlation coefficient of 0.998 and the LOD calculated by the 3σ method is 30 aM, which is the lowest value reported for the ECL methods without amplification, and is also the top 3 as compared with the ECL methods with amplification (Table S1, Supporting Information).[ ] In addition, the specificity for DNA assay was proved by different mismatched and random DNAs (Figure 5D and Figure S20, Supporting Information). Based on the high sensitivity and specificity, the ECL system of MOF 1 is expected to be a rapid and sensitive probe for the detection of SARS‐CoV‐2, and may be an ideal alternative to time‐consuming PCR amplification methods for rapid screening SARS‐CoV‐2 in the initial stage of the epidemic to prevent further spread.

Figure 5

A) EIS spectra for the assembly of the biosensor: a) GCE/MOF 1, b) GCE/MOF 1/CS, c) GCE/MOF 1/CS/pDNA‐1, d) GCE/MOF 1/CS/pDNA‐1/peptides, e) GCE/MOF 1/CS/pDNA‐1/tDNA, f) GCE/MOF 1/CS/pDNA‐1/tDNA/pDNA‐2‐BSA. B) ECL responses of the developed biosensor toward RdRp gene with different concentrations. C) Calibration curve for the RdRp gene detection. D) Specificity of biosensor with various mismatched DNAs (SM: single‐base mismatched, TM: three‐base mismatched, random: noncomplementary).

Conclusion

In summary, a novel porphyrin‐based heterobimetallic 2D MOF, [(ZnTCPP)Co2(MeIm)] (1), has been successfully designed and built from the electroactive [Co2(‐CO2)4] SBUs with an axially coordinated MeIm and photosensitizers (ZnTCPP) for self‐enhanced ECL. By rationally regulating the amounts of MeIm, the electron transfer between the low‐spin Co(II) ions and oxygen can be improved while preserving the oxygen adsorption sites, which is beneficial to the production of O2 •−. In addition, the axially coordinated MeIm can also adjust the 2D MOF 1 to a suitable distance of layers without changing the structure of layer, making the entrance of O2 easier and exposure of ZnTCPP more sufficient. Integration of the sufficient exposure of photosensitizers ZnTCPP and the efficient generation of O2 •− via the electroactive [Co2(‐CO2)4] SBUs in the platform synergistically enhances the ECL performance of MOF 1. Benefiting from the excellent ECL performance, MOF 1 as a nonamplified ECL biosensor for the RdRp gene of SARS‐CoV‐2 shows a high sensitivity, which will have a great application potential for rapid detection of SARS‐CoV‐2 in the initial stage of the epidemic to prevent further spread.

Experimental Section

Materials and Reagents

Cobalt(II) nitrate hexahydrate (Co(NO3)2⋅6H2O, AR, 99%), potassium chloride (KCl), sodium hydroxide (NaOH), 2‐[4‐(2‐hydroxyethyl)‐1‐piperazinyl] ethanesulfonic acid (HEPES), potassium ferricyanide (K3[Fe(CN)6]), potassium ferrocyanide (K4[Fe(CN)6]), ethanol, and methanol were purchased from Aladdin (Shanghai, China). Tetraoctylammonium bromide (TOAB) was obtained from Alfa Aesar Chemicals Co. Ltd. (Shanghai, China). 2‐Methylimidazole (MeIm, AR, 99%) was bought from Maclin (Shanghai, China). Zinc(II) tetrakis(4‐carboxyphenyl)porphine (ZnTCPP) was commercially available from J & K Chemical Ltd. (Shanghai, China). Ultrapure water obtained from a Millipore water purification system (≥18 MΩ, Milli‐Q, Millipore) was used for the whole experiment. ECL measurements were conducted in 10 mM HEPES buffer solution at pH 7.0 containing 0.1 m KCl as the electrolyte.

Characterization

The morphologies of the synthesized hybrid materials were investigated by using FEI Quanta250 Field Emission Scanning Electron Microscope (FESEM). Fourier‐transformation infrared (FT‐IR) spectra were recorded with an IR‐Prestige‐21 FT‐IR spectrometer (Shimadzu Co., Japan). Powder X‐ray diffraction patterns (PXRD) analysis was carried out on a Bruker D8‐Focus Bragg–Brentano X‐ray Powder diffractometer equipped with a Cu‐sealed tube (λ = 1.54178 Å) at room temperature. X‐ray photoelectron spectroscopy (XPS) was performed by K‐Alpha X‐ray photoelectron spectroscopy (PHI Quantera II ESCA System). The fluorescence spectra were recorded on a fluorescence spectrophotometer (Edinburgh Analytical Instruments, FLS920). Cyclic voltammetries (CVs) were measured using a CHI 660D electrochemical workstation (CHI Co., USA). ECL measurements were carried out on a MPI‐E multifunctional electrochemical and chemiluminescent analytical system (Xi'an Remex Analytical Instrument Co., Ltd., China). Electrochemical impedance spectroscopy (EIS) measurements were conducted using an Autolab PGSTAT30 (Eco Chemie) controlled by NOVA 1.10 software. All electrochemical studies were carried out with a conventional three electrode system. A Pt wire electrode and an Ag/AgCl electrode were used as counter and reference electrodes, respectively. Modified glassy carbon electrodes (GCE, 5 mm in diameter) were used as working electrodes.

Synthesis of 2D MOF 1

MeIm (3.3 mg, 0.04 mmol, 2 mL methanol) and ZnTCPP (8.5 mg, 0.01 mmol, 10 mL methanol) solutions were mixed under stirring. Then, the solution of Co(NO3)2⋅6H2O (5.8 mg, 0.02 mmol) in 3 mL methanol was added into the above mixed solution. The resulting mixture was stirred vigorously at room temperature for 24 h and the precipitated green–purple powder was then collected by centrifugation (4000 rpm, 10 min), washed with methanol (3 × 10 mL), and dried in vacuum at 60 °C for 12 h to give the MOF 1 material. Yield: 9.9 mg, 56.2%.

To monitor the MeIm axial coordination number, based on the above protocol, the addition amount of MeIm was changed to 0.05 mmol (4.1 mg) and 0.06 mmol (5.0 mg) to synthesize MOF 1 with different amounts of MeIm.

Synthesis of MOF 2

A solution of Co(NO3)2⋅6H2O (5.8 mg, 0.02 mmol) in 3 mL methanol and a solution of ZnTCPP (8.5 mg, 0.01 mmol) in 10 mL methanol were mixed with stirring. The resulting mixture was stirred vigorously at 60 °C for 24 h and the precipitated green–purple powder was then collected by centrifugation (4000 rpm, 10 min), washed with methanol (3 × 10 mL), and dried in vacuum at 60 °C for 12 h to give the MOF 2 material.

Synthesis of [(ZnTCPP)Zn2(MeIm)]

MeIm (4.1 mg, 0.05 mmol, 2 mL methanol) and TCPP (8.5 mg, 0.01 mmol, 10 mL methanol) solutions were mixed under stirring. Then, the solution of Zn(NO3)2⋅6H2O (8.9 mg, 0.03 mmol) in 3 mL methanol was added into the above mixed solution. The resulting mixture was stirred vigorously at room temperature for 24 h and the precipitated green–purple powder was then collected by centrifugation (4000 rpm, 10 min), washed with methanol (3 × 10 mL), and dried in vacuum at 60 °C for 12 h to give the [(ZnTCPP)Zn2(MeIm)] material.

Synthesis of [(CoTCPP)Co2(MeIm)]

MeIm (4.1 mg, 0.05 mmol, 2 mL methanol) and TCPP (8.5 mg, 0.01 mmol, 10 mL methanol) solutions were mixed under stirring. Then, the solution of Co(NO3)2⋅6H2O (8.7 mg, 0.03 mmol) in 3 mL methanol was added into the above mixed solution. The resulting mixture was stirred vigorously at room temperature for 24 h and the precipitated green–purple powder was then collected by centrifugation (4000 rpm, 10 min), washed with methanol (3 × 10 mL), and dried in vacuum at 60 °C for 12 h to give the [(CoTCPP)Co2(MeIm)] material.

Preparation of BSA Modified pDNA‐2 (P2‐BSA)

Solution A consisted of 200 µL pDNA‐2 (1 × 10−6 mol L−1) and 100 µL 12.5% glutaraldehyde (GA), which was incubated at 37 °C for 1 h with shaking. Then, 100 µL 1% BSA was added to solution A, and the mixture was shaken at 37 °C for 2 h. Finally, the resulting solution was washed with ultrapure water (3 × 4 mL) through an ultrafiltration tube (MWCO = 30 KD) to remove unreacted pDNA‐2 and glutaraldehyde. The resulting pDNA‐2 for the RdRp gene with BSA modified was named P2‐BSA.

Fabrication of the ECL Biosensor

The GCE (d = 5 mm) was polished carefully with 0.3 and 0.05 µm of Al2O3 powder and sonicated sequentially in ethanol and rinsed by deionized water for further use. A total of 5 µL MOF 1 (1 mg mL−1 in aqueous solution) was spread on the GCE and coated with 10 µL chitosan (0.1% w/w in 2% acetic acid solution) and dried at room temperature to obtain the CS/MOF 1/GCE. Subsequently, 20 µL GA (12.5% w/w in water) was dropped onto the CS/MOF 1/GCE and incubated at 37 °C for 1 h. After the reaction, the electrodes were washed with ultrapure water (3 × 2 mL), and 20 µL pDNA‐1 (1 × 10−6 mol L−1) was then covered onto the above‐mentioned modified GCE and left for 2 h at 37 °C. After that, 20 µL peptides solution (1% w/w in water) was distributed onto the surface to block the nonspecific binding sites. After being washed thoroughly with ultrapure water (3 × 2 mL), 20 µL of the tDNA was dropped onto the surface of the electrode and incubated at 37 °C for 2 h. After rinsing thoroughly with ultrapure water (3 × 2 mL), 20 µL P2‐BSA was added to the electrode and incubated at 37 °C for 2 h. Finally, the electrode would generate ECL signal in the O2‐saturated aqueous electrolyte solution and gave the quantitative criteria for the proposed DNA assay.

Calculation Method

All density function theory (DFT) calculations were performed based on the projector‐augmented‐wave method as implemented in the Vienna Ab initio simulation package (VASP). The Blöchl's all‐electron‐like projector‐augmented plane wave (PAW) method was used to describe the interactions between ion cores and valence electrons. The electronic exchange‐correlation potential was treated within the functional of Perdew–Burke–Ernzerhof (PBE) function for generalized gradient approximation (GGA). The plane‐wave cutoff was set to 400 eV and the convergence threshold was set as 1 × 10−3 eV in energy and 0.02 eVÅ−1 in force. The Monkhorst–Pack k‐point was set with 3 × 3 × 1 grid meshes for structure optimization. The DFT‐D3 method is adopted to describe the weak van der Waals (vdW) interactions between the adsorbate and the substrate. The adsorption energy (E ads) of O2 is calculated by the following formula: where E sa, E ad, and E s represent the total energies of the adsorbate plus substrate system, isolated adsorbate, and pure substrate, respectively.

In addition, HOMO, LUMO, and Mulliken charge distributions are calculated by Dmol3 module in the Materials studio program based on the GGA‐PBE method.

Conflict of Interest

The authors declare no conflict of interest.

Supporting Information

Click here for additional data file.