An electrochemical biosensor based on glyco-conjugated Cu-BTC MOFs for voltammetric detection of bacteria

Abstract

Foodborne illnesses pose a significant public health challenge globally. According to WHO estimates, unsafe food causes approximately 600 million cases of foodborne diseases annually. Bacterial pathogens, including E. coli and P. aeruginosa, are significant contributors, causing illnesses ranging from mild gastrointestinal issues to severe, life-threatening conditions. E. coli can lead to severe gastrointestinal diseases, while P. aeruginosa poses risks in high-moisture foods due to its biofilm formation and antimicrobial resistance. Effective detection of these pathogens is vital for ensuring food safety and preventing outbreaks. This study reports the synthesis of a monosaccharide sugar-conjugated Cu-BTC bioprobe for the electrochemical detection of lectin and bacteria via classical carbohydrate-lectin interactions. Cu-BTC was drop casted onto a screen-printed carbon electrode (SPCE) and covalently linked with sugar via carbodiimide chemistry. In this easy-to-synthesize bioprobe, the Cu-BTC metal–organic framework acted as a redox mediator, while the monosaccharide sugar molecules served as bioreceptor elements. The developed 4APM@Cu-BTC/SPCE and 4APG@Cu-BTC/SPCE bioprobes exhibited significant voltammetric responses, achieving detection limits of 2461 CFU per mL and 84.68 CFU per mL towards E. coli and P. aeruginosa, respectively, with a quick response time of <15 min. At the same time, the synthesized bioprobes also proved to be effective in the detection of lectins such as concanavalin A and PA-1. Besides, the covalently bound monosaccharide sugars facilitated the selective interaction of bioprobes with the corresponding analytes while eliciting negligible responses towards common biological interferents. Moreover, the fabricated bioprobes were applied for the detection of bacterial species in spiked milk and juice samples and showed satisfactory recovery percentages of ca. 80–91% and 78–93% for E. coli and P. aeruginosa, respectively. This work provides a new approach for the advancement of a carbohydrate-based electrochemical sensing platform. By eliminating the need for an external redox mediator and utilizing a cost-effective, sensitive, and readily accessible bioreceptor, the sugar-modified Cu-BTC framework offers a promising sensing strategy. Additionally, owing to their in-built non-genetic information and involvement in host–pathogen interaction, carbohydrates can enhance their utility in sensing applications.

---

1. Introduction

Foodborne illness, commonly referred to as food poisoning, arises from the consumption of food contaminated with harmful microorganisms, toxins, or harmful chemicals. It poses significant global public health challenges, affecting millions of people annually and leading to severe health outcomes, economic burdens, and fatalities. According to the World Health Organization (WHO), these illnesses affect 600 million people and cause 420 000 deaths annually.¹ Bacterial pathogens are the leading cause of foodborne illnesses, inducing a spectrum of conditions ranging from mild gastrointestinal discomfort to severe, life-threatening infections. They are often transmitted through the consumption of contaminated foods, including raw or undercooked ground meat, unpasteurized milk, and contaminated raw vegetables.¹

Escherichia coli and Pseudomonas aeruginosa are notable for their impact on public health. E. coli, a common commensal of warm-blooded animals, is a significant contributor to foodborne illnesses. While most strains are non-pathogenic, certain variants, such as Shiga toxin-producing E. coli (STEC), can cause severe gastrointestinal illnesses, including haemorrhagic colitis and haemolytic uremic syndrome.² In contrast, P. aeruginosa, although predominantly associated with nosocomial infections, poses a food safety risk due to its ability to contaminate high-moisture, nutrient-rich foods such as milk, vegetables, and fruits. This bacterium's biofilm-forming capability and inherent antimicrobial resistance make it a critical concern, particularly for immunocompromised individuals.³ The detection of these pathogens is crucial for ensuring food safety and protecting public health. Rapid and accurate identification of these bacteria in food products can prevent outbreaks of foodborne illnesses.

Conventional diagnostic methods like cell culture, polymerase chain reaction (PCR), and immunological methods, while highly specific and sensitive, are often time-consuming, require trained personnel, and follow complex procedures, making them inappropriate for rapid detection.^4–6 Therefore, there is a pressing need to develop innovative methods to address the challenges associated with bacterial detection. Among the array of new-age diagnostic methods, electrochemical approaches have emerged as a game-changer in the biosensor industry. Their distinct advantages such as surface phenomenon sensitivity, quick response, stability, ease of deployment, and effortless operation make them surpass traditional methodologies.^4,7 These electrochemical methods harness the principles of electrochemistry to detect specific biomolecular interactions, ensuring precision and efficiency in bacterial identification.

In recent years, metal–organic frameworks (MOFs) have garnered significant attention in various biomedical applications such as drug extraction,^8,9 wound healing, chemical sensing,¹⁰ and biosensing.^11,12 MOFs are intrinsically porous materials composed of metal ions or clusters coordinated to organic ligands, resulting in structures with high surface areas, tunable porosity, and diverse surface functionalities. The incorporation of metal cations imparts redox properties to MOFs, facilitating electron movement within the framework and making them highly suitable as transducer materials in biosensors.^13,14 These characteristics have been effectively harnessed in the development of MOF-based biosensors for the detection of various biomolecules.^15,16 Additionally, MOFs have been employed in signal transduction mechanisms, including optical and electrochemical methods, to detect bacterial interactions, thereby improving the detection limits and response times of biosensors.¹⁷ These developments underscore the potential of MOFs to address current challenges in pathogen detection and contribute to the advancement of rapid and accurate diagnostic tools.¹⁸

Characteristics like ease of immobilization, high stability, and host-specificity with minimal cross-reactivity are highly desirable for any biorecognition element. Biomolecules such as aptamers, antibodies, and bacteriophages have been extensively utilized in bacterial biosensing.^7,19,20 However, due to persisting challenges in synthesis, cost, stability, yield, and storage, alternatives with greater environmental stability and lower mutation rates are gaining popularity, such as carbohydrates. Carbohydrates, the most abundant non-genome encoded organic molecules in the biosphere, possess a vast repertoire of biological information. They exhibit high specificity against their protein counterparts, known as lectins, demonstrating interactions as specific as those observed in antigen–antibody or substrate–enzyme interactions.¹⁹ Lectin–carbohydrate interaction plays a vital role in host pathogenesis, allowing the pathogen to interact with the host cell for establishing primary infection.^19,21 In light of this knowledge, carbohydrate-based sensing platforms have emerged for the detection of bacterial species (Table S2, ESI†). However, the majority of research in this domain leans towards optical methods and is primarily directed at detecting E. coli. In an innovative approach, Kaushal et al. synthesized glycoconjugate-coated Au nanorods for the photothermal ablation and optical detection of E. coli and P. aeruginosa.²² Subsequently, the same group explored the multiepitope sugar approach for the colorimetric detection and ablation of P. aeruginosa.²³ In the realm of electrochemical sensing, Zadeh et al. devised a mannose-based sensor platform on a gold-coated glassy carbon electrode for detecting E. coli.²⁴ Guo et al. combined carbohydrate recognition properties with faradiac electrochemical impedance spectroscopy to develop an impedimetric bioprobe for the detection of E. coli.²⁵

The current study aims to develop an electroactive Cu-BTC MOF for voltammetric detection of both E. coli and P. aeruginosa, utilizing sugar as a bioreceptor moiety. Previously, Cu-BTC has been utilized for electrochemical detection of E. coli; for example, Gupta et al. fabricated an immunosensor using a Cu₃(BTC)₂-PANI composite for impedimetric detection.²⁶ Herein, PANI served as a redox mediator, while Cu-BTC as a substrate for antibody linkage. However, in the present study, Cu-BTC served dual roles, acting as both a redox mediator and a substrate for the conjugation of monosaccharide sugars. The Cu-BTC MOF was hydrothermally synthesized using triethylamine (TEA) as a modulator, deposited on a screen-printed carbon electrode, and conjugated with monosaccharide sugars. The developed bioprobes, 4APM@Cu-BTC/SPCE and 4APG@Cu-BTC/SPCE, were then assessed for the voltammetric detection of E. coli and P. aeruginosa, respectively. Following the assessment, the bioprobes were evaluated on multiple analytical parameters and were also tested on juice and milk samples. Prior to bacterial spiking, juice and milk samples were diluted in phosphate-buffered saline (PBS) to minimize matrix complexity.

2. Experimental section

2.1. Materials

All the reagents and solvents used were of analytical grade purity. Benzene-1,3,5-tricarboxylic acid (H₃BTC), 4-aminophenyl-β-D-galactopyranoside (galactose sugar), 4-aminophenyl-α-D-mannopyranoside (mannose sugar), Concanavalin A lectin, PA-1 lectin, polyvinylidene fluoride (PVDF), potassium ferrocyanide (K₄Fe(CN)₆·3H₂O), potassium bromide (KBr), ascorbic acid, dopamine, and glucose were procured from Sigma-Aldrich. Chloramphenicol, copper(II) nitrate trihydrate (Cu(NO₃)₂·3H₂O), 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC), N-hydroxysuccinimide (NHS) and 2-(N-morpholino)ethanesulfonic acid (MES) were purchased from Tokyo Chemical Industry. Phosphate buffer saline (PBS), nutrient agar, nutrient broth, cetrimide agar, bovine serum albumin (BSA), and anthrone reagent were acquired from HiMedia. Potassium nitrate (KNO₃) was procured from CDH Fine Chemicals. Triethylamine and N-methyl-2-pyrrolidone (NMP) were purchased from SRL Chemicals. Hydrogen peroxide (H₂O₂, 30% w/w) was provided by Fisher Scientific (India). Screen-printed carbon electrodes (SPCEs) were purchased from Zensor. The bacterial strains, namely, Escherichia coli, Staphylococcus aureus, and Pseudomonas aeruginosa, used during the study, were procured from CSIR-IMTECH, Chandigarh, India.

2.2. Instrumentation

A Fourier transform infrared spectrometer (FTIR, Nicolet iS10, Thermo Fischer Scientific) and a UV-visible spectrophotometer (Varian Cary 4000) were used for spectroscopic characterization. The size distribution was examined by the dynamic light scattering method using a Malvern Zetasizer Nano ZS instrument. Thermogravimetric analysis (TGA) was performed by using a METTLER TOLEDO thermal analyzer using an alumina crucible as a sample holder. The structural and morphological investigations were carried out with an X-ray diffractometer (Bruker) and a field emission-scanning electron microscope (Zeiss, Oberkochen, Germany). Electrochemical measurements were performed on an EmStat 3 Blue Potentiostat (PalmSens).

2.3. Methodology

2.3.1. Synthesis of Cu-BTC. Briefly, 1 mmol solution of Cu(NO₃)₂·3H₂O (1 : 1 v/v of H₂O : ethanol) was mixed with 2 mmol solution of H₃BTC (benzene-1,3,5-tricarboxylic acid, 1 : 1 v/v of H₂O : ethanol). To this mixture, 150 μL of triethylamine was added and stirred for 10 minutes. Subsequently, the solution was transferred into a Teflon-lined hydrothermal vessel and maintained at 100 °C for 14 hours. Following this period, the resulting product was centrifuged at 9000 rpm for 10 minutes, washed 2 times with ethanol and water, respectively, and finally vacuum-dried at 70 °C for 48 hours.

2.3.2. Deposition of Cu-BTC on a screen-printed carbon electrode (SPCE). Before deposition, electrodes were electrochemically activated to enhance their sensing performance. This involved subjecting them to 15 repetitive voltammetric cycles at a scan rate of 35 mV s⁻¹ within a potential range of −0.7 to 1.0 V, using 10 mM H₂O₂ solution in 0.1 M phosphate buffer saline (pH 7.4) as the electrolyte. Following activation, Cu-BTC was deposited onto the working area of the SPCE through drop casting. A viscous slurry containing Cu-BTC and PVDF in a ratio of 9 : 1 (w/w) was prepared in NMP solvent and drop cast onto the surface of the SPCE working electrode. Subsequently, the Cu-BTC deposited electrodes were annealed at 60 °C for 2–3 hours (Scheme 2).

2.3.3. Electrochemical characterization. Electrochemical characterization was conducted using an Emstat 3 potentiostat instrument, employing the SPCE as the substrate for material deposition. The majority of cyclic voltammetry (CV) and differential pulse voltammetry (DPV) characterization studies, including biosensing, were performed in a 0.1 M PBS (pH 7.4) electrolyte. However, certain characterization studies were carried out in Ferro electrolyte, consisting of 10 mM K₄Fe(CN)₆·3H₂O and 1 M KNO₃ in 0.1 M PBS (pH 7.4). Throughout all experiments, the potential window ranged from −0.6 to +0.4 V (vs. a reference electrode), with a scan rate of 50 mV s⁻¹.^27–29

2.3.4. Conjugation of mannose and galactose sugar on Cu-BTC/SPCE. In this study, two different sugars, 4-aminophenyl-α-D-mannopyranoside (4APM, mannose sugar) and 4-aminophenyl-β-D-galactopyranoside (4APG, galactose sugar) galactose, were employed and conjugated separately on Cu-BTC/SPCE. Initially, Cu-BTC/SPCE was incubated with an EDC-NHS mixture (2 : 5 μmol in activation buffer) for 10 minutes. EDC-NHS serves to activate the free –COOH groups present in Cu-BTC. Following this, the EDC-NHS-activated electrode was treated with 5 μL sugar solution (8 mg mL⁻¹ in coupling buffer) for 10 minutes. The resulting bioprobes (4APM@Cu-BTC/SPCE and 4APG@Cu-BTC/SPCE) were washed successively with coupling buffer (pH 7.4) and deionized water to eliminate any unbound moiety. Finally, the prepared electrodes were stored at 4 °C for further use.

For the sugar quantification, anthrone analysis was employed, which is a qualitative and quantitative test for carbohydrates. In this method, carbohydrates undergo dehydration with concentrated H₂SO₄ to form furfural, which subsequently undergoes condensation with anthrone, leading to the formation of a green-colored complex measurable spectrophotometrically at 630 nm. In the experiment, the sugar-conjugated Cu-BTC was carefully scraped from the working electrode and mixed with an ice-chilled mixture of 0.2% anthrone prepared in 95% H₂SO₄. The resulting reaction mixture was incubated for 10 minutes in a water bath, and then cooled to room temperature, and the absorbance was recorded at 630 nm.

2.3.5. Electrochemical biosensing. The synthesized bioprobes were employed for the electrochemical detection of P. aeruginosa and E. coli, along with their respective lectins. All solutions, including bacterial dilutions, were prepared in 0.1 M PBS (pH 7.4) buffer.

The mannose conjugated bioprobe (4APM@Cu-BTC/SPCE) was used for detecting E. coli and Concanavalin A (ConA) lectin. It was incubated with 10 μL of varied concentrations of ConA (ranging from 10 ng mL⁻¹ to 80 ng mL⁻¹) and E. coli (ranging from 10⁵ CFU per mL to 10¹⁰ CFU per mL) for 5 minutes at room temperature (RT), separately. Conversely, the galactose-conjugated bioprobe (4APG@Cu-BTC/SPCE) was incubated with 10 μL of varied concentrations of PA-1 lectin (ranging from 10 ng mL⁻¹ to 100 ng mL⁻¹) and P. aeruginosa (ranging from 1 CFU per mL to 10⁷ CFU per mL) for 5 min at RT, independently. Post incubation, the working electrode of SPCE was rinsed with 0.1 M PBS (pH 7.4) buffer to remove any unbound analyte and air-dried at room temperature. Finally, DPV measurements were conducted on each electrode in the presence of 0.1 M PBS (pH 7.4) electrolyte, as specified above.

2.4. Analytical performance evaluation of bioprobes

2.4.1. Reproducibility, selectivity, and cross-reactivity. Reproducibility measures the variation within a batch of electrodes. It evaluates the bioprobe's ability to yield identical sensing responses across multiple experimental setups.³⁰ In this study, batches of 5 electrodes for each bioprobe were incubated with similar concentrations of the respective bacteria and their DPV responses were measured. Selectivity and cross-reactivity are the two most critical characteristics of a biosensor. Selectivity reflects the bioprobe's ability to detect a specific analyte in a mixture containing other contaminants, while cross-reactivity measures the same ability in a mixture of similar contaminants.^30–32 The selectivity of the two bioprobes was evaluated against non-specific analytes, such as ascorbic acid, bovine serum albumin (BSA), chloramphenicol, glucose, and dopamine. Additionally, the bioprobes' cross-reactivity against available bacterial species was assessed.

2.4.2. Spiked sample analysis. Pseudomonas aeruginosa and Escherichia coli are frequently present in the atmosphere and various raw food items. Samples of raw fruit juice and milk were obtained from local stores and then diluted in 0.1 M PBS buffer. These diluted samples were spiked with 100 CFU per mL of bacteria and allowed to incubate at 37 °C for 6 h. Parallel experiments were conducted using nutrient broth media as a positive control. Each experiment was carried out in triplicate. The recovery percentage of the spiked samples was calculated relative to the positive control. Given the relatively simple composition of these matrices, additional sample preparation techniques such as magnetic extraction or sorbent-based extraction³³ were deemed unnecessary, and direct PBS dilution was sufficient to facilitate bacterial detection.

3. Results and discussion

3.1. Spectroscopic and morphological characterization

A hydrothermal method was employed for the synthesis of Cu-BTC using Cu(NO₃)₂·H₂O as a metal precursor and H₃BTC as a linker moiety (Scheme 1). To avoid the formation of copper oxide and enhance the yield of the primary product, triethylamine (TEA) was introduced as a modulator/base during the synthesis. TEA interferes with the coordination bond formation between the metal ion and linker, thereby regulating the rate of MOF's framework extension and crystal growth.³⁴ It also influences the synthesis by modulating the pH of the reaction mixture.³⁵ For instance, Tranchemontagne et al. synthesized various Zn-based MOFs, by incorporating TEA into the reaction mixture.³⁶ In the present study, TEA deprotonated the –COOH groups of the H₃BTC linker and facilitated its coordination to metal centers.^34,37 The synthesized Cu-BTC exhibited a characteristic absorption peak in the ultraviolet range of 275–300 nm, corresponding to the ligand-to-metal charge transfer (LMCT) transition from oxygen to copper ions (Cu²⁺) (Fig. 1(a)).²⁶

Scheme 1 Hydrothermal synthesis of the Cu-BTC metal–organic framework.

Fig. 1 (a) UV-Vis spectra of Cu-BTC. Inset shows the characteristic absorption peak in the range 275–300 nm; (b) FTIR spectra of Cu-BTC and sugar-conjugated Cu-BTC; (c) X-ray diffraction spectra of Cu-BTC; (d) thermal analysis spectra of Cu-BTC.

Further, FTIR studies were carried out to ascertain the presence of functional groups and chemical bonds (Fig. 1(b)). The characteristic peak from 1300 to 1600 cm⁻¹ was attributed to the symmetric and asymmetric vibrations of the carboxylate (COO–) group of the BTC ligand. The peak at 730 and 750 cm⁻¹ corresponded to the in-plane C–H bending mode,³⁸ while the peak at 470 cm⁻¹ was associated with the vibrational mode of the Cu–O bond. An additional peak was observed at 1700 cm⁻¹ in Cu-BTC, indicating the presence of free –COOH groups in the structure, suggesting partial conjugation of H₃BTC (Fig. 1(b)). These findings typically align with the coordination sphere of six-prismatic crystals,³⁹ suggesting that the H₃BTC exhibits partial participation in the six-prismatic crystal and the interaction between the linker and metal ions is rather asymmetrical.³⁹ The FTIR analysis of sugar conjugated Cu-BTC was performed by carefully scraping the material from the electrode surface. The scraped material was then ground with potassium bromide (KBr, an IR-inactive material), compressed into a transparent pellet under high pressure, and subsequently analyzed using a pellet holder accessory. In the sugar-conjugated Cu-BTC, the peak at 1700 cm⁻¹ was attenuated due to the amide bond formation between the free –COOH of the framework and –NH₂ groups of the sugar moieties. This amide linkage augments the –C–N bonds in the Cu-BTC, leading to an increment in peak intensity between 1170 and 940 cm⁻¹. Additionally, the peak became more distinct and showed broadening, indicative of an –NH loop in the structure. Furthermore, a separate peak corresponding to the amide II appeared at 1509 cm⁻¹ (ref. 19) in the carbohydrate conjugated Cu-BTC. Thus, the FTIR studies concluded the successful bond linkages and the presence of functional groups in the conjugated and non-conjugated Cu-BTC.

XRD was performed to assess the crystal structure of Cu-BTC (Fig. 1(c)). The characteristic 2θ peaks observed at 6.7°, 9.5°, 11.78°, 13.5°, 16.5°, 17.5°, 19.1°, and 26° corresponded to the (200), (220), (222), (400), (422), (333), (440), and (355) miller indices, respectively (Fig. 1(c)). The obtained XRD pattern matched with the reported COD databases (4002051 and 7107291), thereby confirming the formation of Cu-BTC.²⁹ The sharp peaks in the diffraction pattern suggested the crystallinity of the synthesized Cu-BTC, while low-intensity peaks at higher 2θ values were attributed to the incidental formation of Cu₂O. The X-ray diffraction (XRD) pattern of the sugar-conjugated Cu-BTC (Sugar@Cu-BTC), presented in Fig. S1 (ESI†), closely resembled that of the unmodified Cu-BTC. This indicated the retention of the characteristic diffraction peaks and crystallinity in sugar-conjugated Cu-BTC. Additionally, a broad hump was observed in the sugar-conjugated CuBTC, indicative of the amorphous nature of the sugar component.^40,41 This suggests its successful incorporation into the conjugate without affecting the crystalline characteristics of the framework. Moreover, FTIR analysis suggested that the sugar conjugation occurred primarily via the free COOH groups of the Cu-BTC framework and these groups do not participate in the coordination structure of the MOF, thus maintaining the structural integrity of the framework. These observations collectively confirm that the structural integrity, bonding, and crystallinity of the Cu-BTC framework were preserved upon sugar conjugation.

Next, thermal analysis of Cu-BTC and sugar conjugated Cu-BTC was carried out to evaluate the temperature stability using a thermogravimetric analyser (TGA), which predicted a two-step degradation process⁴² (Fig. 1(d)). Initially, a weight loss of 20–25% (approx.) was observed in the temperature range of 80–130 °C, attributed to the removal of physically adsorbed solvents such as water and ethanol from the pores of the MOF.⁴³ There was no notable weight loss observed thereafter until 400 °C.^39,44–46 The weight loss further increased to ∼60% around 420 °C, corresponding to the decomposition of the secondary building unit and BTC-ligand framework of MOF. The above results confirmed the structural stability of Cu-BTC up to 400 °C. However, beyond this temperature, the structure collapsed due to the decomposition of the building units. Both pristine and sugar conjugated Cu-BTC (Fig. S2, ESI†) exhibited a similar degradation pattern.

Lastly, the morphological and elemental analysis of Cu-BTC and sugar conjugated Cu-BTC was carried out using scanning electron microscopy (Fig. 2 and Fig. S3, ESI†) and EDAX, respectively. Cu-BTC exhibited a well-defined rod-like flaky 3D structure.⁴⁷ The observed morphology closely resembles that of a six-prismatic crystal, with the rod length and width ranging from 1 to 5 μm and 0.5 to 1 μm, respectively.^39,45 Such materials exhibiting a rod-like morphology offer advantages for charge carriers due to their aspect ratio, polydispersity, and orientation.⁴² EDAX analysis confirmed the presence of constituent elements, i.e., copper, oxygen, and carbon atoms, in the Cu-BTC structure (Fig. 2(e)). In the case of sugar-conjugated Cu-BTC, the material retained a morphology similar to that of the pristine MOF (Fig. S3(a) and (b), ESI†), indicating that the conjugation process did not alter the structural morphology of Cu-BTC. The conjugation was confined to the surface-exposed –COOH functional groups, resulting in a low-density surface modification. Owing to the limited extent and spatial distribution of the conjugated sugar moieties, SEM–EDX analysis did not reveal any distinct elemental differences between the conjugated and pristine Cu-BTC samples, with both exhibiting comparable elemental compositions. Additionally, DLS was utilized to determine the size distribution and diameter of Cu-BTC, revealing an average diameter of approximately 1.14 μm (Fig. S4, ESI†). The findings obtained from DLS were consistent with SEM observations.

Fig. 2 (a)–(d) Scanning electron microscopy images of Cu-BTC at different magnifications (200 nm, 500 nm, 1 μm, and 2 μm). (e) Elemental analysis data of Cu-BTC.

3.2. Electrochemical characterization

Cu-BTC frameworks are composed of Cu²⁺ ions as the metal center and trimesic acid as a linker. They exhibit electroactivity due to the presence of partially occupied d-orbitals in Cu atoms, resulting in metallic states at the Fermi level.⁴⁸ The electrochemical behavior of Cu-BTC was studied by cyclic voltammetry (CV) using the SPCE as a substrate. Cyclic voltammetry is a frequently used method for the determination of redox processes and electron transfer mechanisms. Through the current profile, information about the material's redox potential and the electrochemical reaction rates on the electrode surface was obtained.⁴⁹

Screen-printed carbon electrodes (SPCEs) often require activation before being used in biosensing experiments to ensure that they perform optimally. Prior to Cu-BTC deposition, the SPCEs were electrochemically activated with H₂O₂. This process typically enhanced the electrode's electrochemical properties by improving the surface smoothness and increasing the surface area for efficient electron transfer. Consequently, it facilitated better material adherence and improved the sensor's performance in sensing applications.^50,51 Following activation, Cu-BTC was deposited onto the working electrode of the SPCE using NMP as the solvent and PVDF as the binder (Scheme 2). NMP, being an organic solvent, facilitated the dispersion of the Cu-BTC and PVDF mixture, while its low volatility provided sufficient time for deposition. PVDF assisted in the adhesion of Cu-BTC to the electrode surface. However, being insulative in nature, its concentration in the mixture was maintained at only 10%.

Scheme 2 Deposition of Cu-BTC on the working electrode of SPCE using the drop casting method.

The Cu-BTC worked as an electrochemically conductive transducer element. CV of the bare SPCE and Cu-BTC-modified SPCE was recorded in 0.1 M PBS (pH 7.4) and K₄Fe(CN)₆·3H₂O + KNO₃ in PBS (pH 7.4) at a scan rate of 50 mV s⁻¹ (ref. 27–29) (Fig. 3 and Fig. S6, ESI†). The bare SPCE exhibited no significant faradiac redox activity. However, Cu-BTC/SPCE displayed well-defined redox behavior, showing two redox pairs with high faradiac current and peak-to-peak separations (ΔE1_p and ΔE2_p) of approximately 380.2 mV and 228.6 mV, respectively (Fig. 3(a)). The redox pair of O₁–R₁ (Fig. 3(a)) and O_1′–R_1′ (Fig. S6(a), ESI†) represents the oxidation and reduction of Cu ions from Cu(0) to Cu(I) and vice versa. On the other hand, the redox pair of O₂–R₂ (Fig. 3(a)) and O_2′–R_2′ (Fig. S6(a), ESI†) shows the conversion from the Cu(I) to Cu(II) oxidation state and vice versa, in PBS and Ferro electrolyte, respectively. Briefly, the CV profiles of Cu-BTC/SPCE in PBS buffer were attributed to the following reactions,²⁸ demonstrating the electrochemical activity of Cu-BTC on the SPCE substrate.

Cu(0) ↔ Cu(I) + e⁻ (−9.5 mV, E1_pa/−389.7 mV, E1_pc)

Cu(I) ↔ Cu(II) + e⁻ (135.7 mV E2_pa,/−92.9 mV, E2_pc)

Fig. 3 Cyclic voltammogram (CV) in 0.1 M PBS (pH 7.4) electrolyte. (a) Graph showing CV of bare SPCE, activated SPCE, and Cu-BTC at 50 mV s⁻¹ scan rate. (b) Graph showing CV of Cu-BTC at different scan rates. (c) Graph showing the relationship between the square root of scan rate and oxidation/reduction peak current. O₁, R₁: oxidation–reduction pair of Cu(0) → Cu(I); O₂, R₂: oxidation–reduction pair of Cu(I) → Cu(II).

The electrochemical reaction mechanism was investigated by studying the effect of scan rate on the Cu-BTC. In both electrolytes, oxidation (O_1/1′, O_2/2′) and reduction peak currents (R_1/1′, R_2/2′) increased proportionally with the scan rate, displaying a linear correlation with the square root of the scan rate (Fig. 3(b), (c) and Fig. S6(b), (c), ESI†). This observed electrochemical behavior (viz. linear correlation between scan rate and reduction/oxidation peak currents) can be attributed to the diffusion-controlled faradiac response occurring within the metal–organic framework via the redox hopping process. Redox hopping, a Cottrell-like behavior, is a dominant charge transfer mechanism found in several polymeric structures, including MOFs. In this process, electronic movement within the framework occurs via self-exchange reactions between redox centers coupled with the motion of counter-balancing ions.^29,52 The diffusion-controlled kinetics, inferred from the linear relationship, followed the following oxidative regression equations in respective electrolytes:

PBS electrolyte: i_pa = 9.16v + 68.55 (R² = 0.9881) (Fig. 3(c))

Ferro electrolyte: i_pa = 52.2v + 49.43 (R² = 0.9848) (Fig. S3(c), ESI†)

Evident from the I_pa/I_pc ratio of more than 1, the observed electrochemical reaction displayed the characteristics of a quasi-reversible process. Unlike fully reversible reactions, where the forward and backward reactions occur at identical rates, in quasi-reversible reactions, there exists an imbalance in the rates of oxidation and reduction. This discrepancy results in a noticeable asymmetry in the redox peak behavior of Cu-BTC. Additionally, the electroactive surface area (ESA) of Cu-BTC/SPCE was calculated to be 0.014 cm², as determined from the Randles Sevick equation.

On the other hand, a shift in peak potential was noticed in the ferro electrolyte for the second redox pair (O_2′–R_2′) of Cu-BTC (Fig. S6(a), ESI†). This was attributed to the interference of the ferro electrolyte with the inherent conductivity of the MOF. The redox characteristics of the ferro electrolyte overlapped with those of the second redox pair of Cu-BTC, and due to its high conductivity, it amplified the second oxidation and reduction peaks (O_2′–R_2′) of Cu-BTC. Due to the above reasons as well as the degradative effect of the ferro electrolyte towards biomolecules, all biosensing experiments were carried out in 0.1 M PBS (pH 7.4).

3.3. Electrochemical biosensing

The interaction between lectins and carbohydrates plays a pivotal role in the pathogenesis of various bacteria. During host–pathogen recognition, bacterial surface lectins recognize their carbohydrate counterparts on host cells and establish themselves within the host body. Taking advantage of lectin-carbohydrate interactions in bacterial species, we have prudently chosen 4-aminophenyl-α-D-mannopyranoside (4APM) and 4-aminophenyl-β-D-galactopyranoside (4APG) sugars for modification of Cu-BTC. Prior to conjugation, the electrochemical stability of Cu-BTC-modified screen-printed carbon electrodes (Cu-BTC/SPCE) across a pH range of 3 to 11 was assessed, as the conjugation process necessitates a brief incubation in a pH 5 solution. As shown in Fig. S5 (ESI†), a decrease in peak potential was observed with an increase in pH from 3 to 11, suggesting protonation in the electrochemical reactions. This decline in peak potential can also be attributed to the presence of anionic charges on the electrode material, which facilitate rapid electron transfer under acidic conditions due to the enhanced diffusion of H⁺ ions towards the electrode surface.^27,53 Despite the drop in peak potential, the cycling stability of Cu-BTC was maintained across the entire pH range, demonstrating stable electrochemical behavior. Notably, while a substantial drop in peak potential occurred between pH 3 and pH 11, the retention of significant peak potential at pH 7 and the stable voltammograms at pH 5 validated the suitability of conducting the conjugation and sensing experiments at both the pH values (vide infra). Carbohydrates were conjugated with the Cu-BTC/SPCE by using EDC:NHS chemistry. The EDC/NHS conjugation chemistry involved coupling the free –NH₂ group of carbohydrates to the free –COOH group of Cu-BTC through the formation of an NHS ester. This NHS ester was more stable than the O-acylisourea intermediate.⁵⁴ Initially, at a mild acidic pH (pH 5.0), NHS was coupled to the –COOH group by EDC to form the NHS ester. Subsequently, this ester reacted with the primary amine (–NH₂) at a physiological pH (pH 7.4), leading to the completion of the conjugation through the formation of an amide linkage between the two groups. The extent of carbohydrate conjugation was quantified using the anthrone assay, which revealed successful conjugation of approximately 13 μg of 4-aminophenyl-α-D-mannopyranoside (4APM) and 17 μg of 4-aminophenyl-β-D-galactopyranoside (4APG) per mg of Cu-BTC. As the conjugation occurred via the free –COOH groups, which are not involved in the framework coordination of the MOF, the structural integrity and crystallinity of Cu-BTC remained unaffected. Furthermore, while a reduction in current response was observed—attributed to the inherently resistive nature of the sugar moieties—this did not compromise the electrochemical functionality of the MOF. Collectively, these findings along with the IR and X-ray observations demonstrate that the Cu-BTC framework retained its structural and electrochemical stability post-conjugation, thereby supporting its effective application in subsequent biosensing experiments.

The resultant bioprobe facilitated the electrochemical biosensing of bacterial species and their lectin proteins via specific binding interactions. The mechanism behind biosensing lies in the lectin–carbohydrate interaction. Lectins are non-immune proteins with specific saccharide recognition sites that recognize and bind carbohydrates protruding from the surface of other cells. This interaction, akin to antigen–antibody or enzyme–substrate interactions, plays a crucial role in microbial pathogenesis. Structural studies have identified carbohydrate recognition domains (CRDs) in lectins, which are responsible for their carbohydrate-binding activity and can also discriminate between anomeric isomers of carbohydrates based on their specificities.⁵⁵ For example, the Con A lectin specifically binds to the a-anomer of glucose and mannose, but not to the b-anomer of either.⁵⁶

In this study, the model organisms E. coli and P. aeruginosa bind to the mannose and galactose bio-probes through the FimH and PA-1 lectins present on their surface, respectively.¹⁹ The carbohydrates were conjugated on the Cu-BTC/SPCE through carbodiimide chemistry, leading to the formation of two abovementioned bacterial species-specific bioprobes: 4APM@Cu-BTC/SPCE and 4APG@Cu-BTC/SPCE (Scheme 3). Further, these bioprobes were utilized for the voltammetric (DPV) detection of bacterial species (E. coli and P. aeruginosa) and lectins (Con A and PA-1) under the optimized conditions. In this procedure, the bioprobes were exposed to the respective analytes for 5 min and subsequently rinsed with PBS buffer to eliminate any loosely bound analyte. Following this step, the electrolyte was introduced, and after 10 min, a consistent response was observed. Both bioprobes exhibited current saturation within 10 min of adding the electrolyte, indicating a stable and repeatable behavior (Fig. S7, ESI†). The data also indicated that a minimum of 5 min was necessary for the electrolyte to reach equilibrium and produce a stable current response. Taking this into consideration, along with the incubation time for bacteria, it was determined that the bioprobes required only 10–15 minutes to generate a reliable sensing response. The peak current of 4APM@Cu-BTC/SPCE and 4APG@Cu-BTC/SPCE decreased linearly with the increase in the concentration of E. coli (Fig. 4(a) and (c)) and P. aeruginosa (Fig. 4(b) and (d)), respectively. The mannose bioprobe showed a detection range of 10⁵ to 10¹⁰ CFU per mL with a LOD of 2461 CFU per mL, whereas 4APG@Cu-BTC/SPCE demonstrated a detection range from 1 to 10⁷ CFU per mL with a LOD of 84.68 CFU per mL (Table S2, ESI†). A similar response was obtained for the detection of Con A and PA-1 lectins (Fig. S8 and Table S2, ESI†). The differential pulse voltammetry (DPV) technique is preferred for analytical purposes owing to its sensitivity and ability to distinguish between faradiac and non-faradiac currents. Its differential nature and shorter pulse time enhance measured currents and provide a high signal-to-noise ratio, even with minimal analyte volume.^29,57

Scheme 3 Conjugation of monosaccharide sugars on the Cu-BTC deposited SPCE using carbodiimide chemistry.

Fig. 4 (a) and (b) Differential pulse voltammetric (DPV) response of 4APM@Cu-BTC and 4APG@Cu-BTC bioprobes as a function of the concentration of E. coli and P. aeruginosa. (c) and (d) Corresponding calibration curves for E. coli and P. aeruginosa.

Post-sugar conjugation, the DPV response of the Cu-BTC decreased due to the hindrance of electronic flow. A similar response was obtained when the bioprobes interacted with the respective analyte: bacteria or lectin protein. The binding of the analyte resulted in the formation of an additional biomolecular layer over the bioprobe, leading to a decrement in the current signal due to the suppression of ionic transport across the electrode–electrolyte interface. The sensing mechanism is predominantly a surface-based resistive phenomenon, wherein Cu-BTC functions as a transducer. The modifications occurring on the Cu-BTC surface due to sugar conjugation and subsequent analyte binding resulted in the reduction of conductivity within the Cu-BTC framework. Cu-BTC not only provided the functional carboxyl (–COOH) groups for carbohydrate conjugation but also demonstrated intrinsic electrochemical activity due to the presence of Cu²⁺ ions. Its electrochemical activity enabled the transduction of carbohydrate–lectin and carbohydrate–bacteria interactions into measurable signals. Furthermore, its stability across different pH conditions ensured that the electrochemical response remained unaffected, with no disruption in the oxidation peak. Biomolecules, being non-conductive (∼10⁻⁷ S m⁻¹ for the bacterial cell membrane⁵⁸), cause steric hindrance to electron flow on the electrode surface, leading to a decrease in faradaic currents.^26,59 As their concentration increases, the effective area of the electrode decreases, leading to a decrease in current, attributed to the hindrance in interfacial electron transfer kinetics.⁵⁸

3.4. Sugar@Cu-BTC bioprobes’ performance evaluation

The performance of the synthesized bioprobes was evaluated for reproducibility, selectivity, cross-reactivity, and real sample analysis. These performance evaluation tests are crucial as they ensure the reliability and consistency of results across different experiments and in complex biological or environmental samples. These evaluations are essential for establishing the credibility and utility of the biosensing technique, which is essential for its successful application in various biosensing fields. The reproducibility study involved preparing five identical electrodes for each bacterium and recording their DPV response at a fixed bacterial concentration under identical conditions (Fig. S9, ESI†). The results showed a relative standard deviation (RSD) of 0.83% and 1.06% for E. coli and P. aeruginosa, respectively, indicating excellent reproducibility of the bioprobes. Selectivity studies were performed against related interferents such as ascorbic acid, bovine serum albumin (BSA), chloramphenicol, glucose, and dopamine. Additionally, cross-reactivity against other lectins and bacteria was assessed. In the case of 4APM@Cu-BTC/SPCE (Fig. 5(a)), significant changes in anodic peak current were observed in the presence of Con A lectin and E. coli, compared to other interferents and cross-reactive lectin proteins or bacterial species. Similarly, the 4APG@Cu-BTC/SPCE bioprobe (Fig. 5(b)) displayed a change in anodic peak current in the presence of PA-1 lectin, P. aeruginosa, and biomolecule mixtures containing them. Both bioprobes exhibited minimal current changes for interfering species, indicating their specificity towards their respective analytes. This specificity behavior can be attributed to the interaction between the mannose and galactose sugars of the bioprobes and their corresponding lectins.

Fig. 5 Selectivity tests of bioprobes against non-specific analytes and cross-reactivity with lectin and bacteria. (a) 4APM@Cu-BTC/SPCE towards E. coli; (b) 4APG@Cu-BTC/SPCE for P. aeruginosa. The bacteria mixture consisted of P. aeruginosa, S. aureus, and E. coli. The protein mixture consisted of Con A, PA-1, and bovine serum albumin (BSA). Bacterial concentrations were 10¹⁰ CFU per mL for 4APM@Cu-BTC/SPCE and 10⁶ CFU per mL for the 4APG@Cu-BTC/SPCE bioprobe. Protein and interferent concentration was 80 ng mL⁻¹.

Most bacterial sensing probes have been evaluated primarily for their selectivity against other bacterial species, with limited studies assessing their specificity against non-bacterial biomolecules. In contrast, the present study not only demonstrates selectivity against non-target bacterial species but also evaluates the probes’ response to other biomolecules, providing a more comprehensive assessment of specificity. When compared to existing studies, the developed bioprobes exhibit a comparable level of selectivity against non-target bacterial species, even at high analyte concentrations, while maintaining minimal and consistent standard deviation in recorded responses. For instance, the aptasensor developed by Shahrokhian and Ranjbar for Escherichia coli detection demonstrated selectivity against four bacterial species, including Staphylococcus aureus and Pseudomonas aeruginosa, though with variable standard deviations. Furthermore, while significant selectivity was observed, the response varied among different E. coli strains, with the lowest signal for E. coli DH5α and the highest for E. coli E23.⁶ Similarly, in another study from the same research group, an aptasensor developed for P. aeruginosa detection exhibited selective recognition of four bacterial strains with a comparable specificity pattern, albeit with variable standard deviations.⁶⁰

Comparable selectivity trends have been reported in other studies for the detection of E. coli and P. aeruginosa. While some studies, including the present work, have assessed selectivity against two or three bacterial species,^26,61,62 others have evaluated a broader range. For instance, Viswanath et al. developed a ZIF-8-based immunosensor for P. aeruginosa detection and validated its selectivity against eight bacterial species.⁶³ Despite variations in the number of bacterial strains tested across studies, all reported similar selectivity trends. However, the present study uniquely demonstrates not only high selectivity but also significantly lower and more consistent standard deviation in the selectivity data—an aspect not commonly observed in previous reports. In prior studies, either high selectivity was achieved with considerable standard deviation, or vice versa. In contrast, the developed bioprobes exhibit both robust selectivity and minimal variability, highlighting their reliability and reproducibility in bacterial detection applications.

The sensing and selectivity studies were conducted for non-pathogenic bacterial strains and did not include evaluations across different strains of each bacterium. The strain specificity of a bacterium toward a carbohydrate largely depends on the expression of its respective surface lectins. These lectins, being critical for pathogenesis, are typically expressed at higher levels in pathogenic strains compared to non-pathogenic ones.^64,65 Accordingly, it was anticipated that the synthesized bioprobes will exhibit higher sensitivity towards pathogenic strains. While the bioprobes demonstrate specificity in their interactions, there remains the potential for cross-reactivity with bacteria expressing similar types of lectins. For example, FimH, a mannose-specific lectin present in Gram-negative bacteria such as Escherichia coli, Klebsiella pneumoniae, and Salmonella enterica, interacts with mannose carbohydrates.^65,66E. coli exhibits strong binding to single mannose residues at the terminal position of glycan chains, whereas K. pneumoniae preferentially binds to mannose residues in complex glycan structures. Similarly, S. enterica interacts with mannose in a manner akin to E. coli but with reduced binding affinity.^64,65 Given that the 4APM@Cu-BTC/SPCE bioprobe possesses mannose as the recognition element, it is capable of detecting bacteria expressing FimH lectins. However, due to the relatively weaker interaction of FimH in K. pneumoniae and S. enterica with single mannose residues, the sensing response for these species may be less pronounced compared to that for E. coli. Furthermore, the weaker binding may result in bacterial loss during washing steps, further reducing sensitivity. In conclusion, while it remains uncertain whether the bioprobes demonstrate strain specificity or bacterial cross-reactivity, the results indicate a measurable affinity for the targeted bacterial species.

The practicality of the prepared bioprobes was demonstrated by evaluating them for the detection of respective bacterial species in fresh juice and milk samples. Initially, the samples were diluted tenfold with 0.1 M PBS buffer and then spiked with bacterial concentrations of 100 CFU per mL, respectively. Post-incubation, the recovery percentage relative to the positive control was calculated. E. coli showed recovery percentages of 79.8% and 90.43%, while P. aeruginosa exhibited recovery percentages of 77.72% and 92.7% in spiked juice and milk samples, respectively (Table 1). The CFU per mL count of spiked samples was validated through standard UV-visible spectroscopy, resulting in a recovery percentage that was consistent with the results of the electrochemical technique (Table 2).

Table 1 Recovery percentage of the bioprobes in sensing the respective bacteria in spiked real samples

Bioprobe	Sample	Bacteria	Spiked value (CFU per mL)	Obtained value^a (CFU per mL)	Recovery^b (%)	Validation recovery (%)
a Bacterial concentration after incubation at 37 °C. b Recovery percentage relative to the final incubated concentration of the positive sample.
4APM@Cu-BTC/SPCE	Nutrient broth	E. coli	100	21.0 × 10^4
	Milk		100	19.1 × 10^4	90.43	92.17
	Fresh juice		100	16.7 × 10^4	79.8	80.72
4APG@Cu-BTC/SPCE	Nutrient broth	P. aeruginosa	100	8.17 × 10^3
	Milk		100	7.57 × 10^3	92.7	93.77
	Fresh juice		100	6.35 × 10^3	77.72	78.71

---

Table 2 Metal–organic framework-based biosensing probes for the electrochemical detection of E. coli and P. aeruginosa bacteria

Sr. no.	Bioprobe	Sensing technique	Limit of detection	Linear range	Incubation time	Ref.
Apt = aptamer, Ab = antibody, PEDOT = poly(3,4-ethylenedioxythiophene), COF = covalent organic framework, PANI = polyaniline, MWCNTs = multi-walled carbon nanotubes, DNA = deoxyribonucleic acid, AgNP = silver nanoparticle, ZIF = zeolitic imidazolate framework, BTC = benzene tricarboxylic acid, PEI = polyethylenimine, MIL = materials of institute lavoisier, MPBA = 4-mercaptophenylboronic acid, EIS = electrochemical impedance spectroscopy, ECL = electrochemiluminescence, DPV = differential pulse voltammetry, SWV = square wave voltammetry, AuNP = gold nanoparticle, MOF = metal organic framework, and CFU = colony forming unit.
E. coli
1	Apt@PANI/Cu3(BTC)2	DPV	2 CFU per mL	2.1 × (10^1 to 10^7) CFU per mL	20 min	6
2	Ru-ConA@NH2-MIL-53(Al)	ECL	16 cells per mL	(50–5.0) × 10^4 cells per mL	60 min	68
3	Apt@Ti3C2Tx and SZr-FcMOF/AuNPs/4-MPBA	DPV	3 CFU per mL	10–10^5 CFU per mL	1.5 hours	67
4	Ab@PEDOT/MIL-53(Fe)	EIS	4 CFU per mL	2.1 × (10^2 to 10^8) CFU per mL	10 min	61
5	Apt@polyMn-MOF	DPV	3.5 CFU per mL	10–10^8 CFU per mL	40 min	69
6	Ab@m-COF	SWV	3 CFU per mL	10–10^8 CFU per mL	20 min	70
7	Ab@PEI/CdS/ZIF-8	DPV	3 CFU per mL	10–10^8 CFU per mL	1 hour	62
8	Ab@PANI/Cu3(BTC)2	EIS	2 CFU per mL	2 × (10^0 to 10^8) CFU per mL	10 min	26
9	4APM@Cu-BTC	DPV	2461 CFU per mL	10 ^5 to 10 ^10 CFU per mL	10 min	This work
P. aeruginosa
10	Fc-GO/Apt@ZIF-8	DPV	1 CFU per mL	1.2 × (10^1 to 10^7) CFU per mL	1.5 hours	60
11	Cu-ZrMOF@Apt@DNA	DPV	2 CFU per mL	10–10^6 CFU per mL	<2 hours	71
12	Apt@MIL-101(Cr)/MWCNT and Apt/AgNPs/c-g-C3N4	DPV	1 CFU per mL	10–10^7 CFU per mL	<1.5 hours	72
13	Ab@ZIF-8/AuNPs	SWV	3.53 CFU per mL	10–10^5 CFU per mL	30 min	63
14	4APG@Cu-BTC	DPV	84.68 CFU per mL	1 to 10 ^7 CFU per mL	10 min	This work

---

The analytical performance of the presented bioprobes was compared with previously reported methods. It is evident from Table 2 that antibodies and aptamers have been mostly explored as biorecognition elements for MOF-based electrochemical biosensing. For instance, Shahrokhian and Ranjbar reported an electrochemical aptasensor for E. coli using a MOF/PANI hybrid nanocomposite. Despite employing electrochemically active polyaniline, methylene blue was utilized as an electroactive indicator for bacterial detection.⁶ A similar approach was adopted for P. aeruginosa detection, where ferrocene-graphene oxide (Fc-GO) and zeolitic imidazolate framework-8 (ZIF-8) served as the electroactive indicator and immobilization platform, respectively.⁶⁰ Some reports have described sandwich assay-based biosensing, where the analyte is sandwiched between two different transducer molecules, with one working as electrode modifier and the other as an electroactive indicator. For example, Dai et al. developed a Faraday-cage-type aptasensor by sandwiching E. coli between a MXene and Fc-MOF.⁶⁷

The sensing systems reported previously entail numerous components and require multistep synthesis, rendering them laborious, expensive, and time-consuming. Furthermore, traditional bio-receptors like aptamers and antibodies present challenges in terms of synthesis, compound purity, environmental stability, and storage and usage conditions, adding to the complexity and labor intensiveness. On the contrary, the transducer material synthesized in this study offers inherent functionality and electroactivity. In addition, the bioprobes developed in this study are simpler to synthesize and easy to operate and possess shorter incubation and detection times with reasonably low and manageable detection limits.

4. Conclusion

Metal–organic frameworks (MOFs) have shown promise in bacterial detection, although research in this area is still in its infancy, especially related to the development of sugar-based bacterial detection methods (Table S3, ESI†). This study focused on MOF-sugar-based electrochemical biosensing of Escherichia coli and Pseudomonas aeruginosa. The hydrothermally synthesized Cu-BTC exhibited significant electrochemical activity due to the partially occupied d-orbitals in Cu atoms. Following synthesis, sugar moieties were covalently attached to the free –COOH groups of Cu-BTC using carbodiimide chemistry, enabling selective detection of the respective analytes. The Cu-BTC functioned as a comprehensive transducer, serving both as a conjugation platform for bioreceptor attachment and as an electroactive material facilitating analytical signal transduction for biosensing. Under optimal experimental conditions, the developed bioprobes (4APM@Cu-BTC/SPCE and 4APG@Cu-BTC/SPCE) demonstrated a linear response within the specified sensing range and achieved detection limits (LOD) of 2461 and 84.68 CFU per mL against E. coli and P. aeruginosa, respectively. Compared to conventional detection techniques, electrochemical sensing offers distinct advantages, including ease of fabrication, suitability for point-of-care applications, and rapid response times. Both biosensors exhibited high selectivity against common interferents and demonstrated excellent analytical performance, with significant recovery rates in milk and juice samples. Furthermore, they were cost-effective, environmentally sustainable, and simple to operate, as their synthesis involved minimal use of hazardous chemicals, reduced reagent consumption, and a non-laborious process, thereby minimizing environmental impact. To further assess the environmental sustainability of this analytical approach, green chemistry evaluation tools such as the green analytical procedure index (GAPI) or modified GAPI (MoGAPI) could be employed to quantify the greenness of individual methodological steps.^73,74 Although the limits of detection (LODs) were slightly higher than the reported values, the approach demonstrated significant advantages due to its simplicity, ease of synthesis, and reduced detection time. Additionally, the stability and analytical versatility of metal–organic frameworks (MOFs), coupled with the cost-effectiveness of carbohydrates, their versatile non-genetic biological information, and their role in pathogenesis, provide a robust foundation for the development of MOF-sugar-based sensing bioprobes.

Author contributions

Deepanshu Bhatt: conceptualization, experimentation, methodology, investigation, and writing – original draft; Deepak Kumar: experimentation and methodology; Abhay Sachdev: supervision, resources, writing – review & editing and validation; Akash Deep: resources, reviewing, and editing.

Ethics approval

This is an observational study; no ethical approval is required.

Data availability

The data supporting this article have been included as part of the ESI.†

Conflicts of interest

The authors have no relevant financial or non-financial interests to disclose.

Acknowledgements

This work was supported by the Department of Biotechnology, India (BT/PR36154/NNT/28/2020), Council of Scientific & Industrial Research (CSIR-HCP0057) and CSIO in-house research grant (OLP-0309). The authors also appreciate the support from CSIR-CSIO, Chandigarh and IIT Roorkee for providing access to various analytical facilities needed for this study. DB is thankful to CSIR India for providing the SRF fellowship for pursuing PhD.

References

1. H. Lee and Y. Yoon, Food Sci. Anim. Resour., 2021, 41, 1.
2. J. M. Hunt, Clin. Lab. Med., 2010, 30, 21–45.
3. X. Li, N. Gu, T. Y. Huang, F. Zhong and G. Peng, Front. Microbiol., 2023, 13, 1114199.
4. A. A. Karbelkar and A. L. Furst, ACS Infect. Dis., 2020, 6, 1567–1571.
5. O. Simoska and K. J. Stevenson, Analyst, 2019, 144, 6461–6478.
6. S. Shahrokhian and S. Ranjbar, Analyst, 2018, 143, 3191–3201.
7. S. Madhu, S. Ramasamy and J. Choi, Pharmaceuticals, 2022, 15, 1488.
8. R. M. Abdelhameed, S. F. Hammad, I. A. Abdallah, A. Bedair, M. Locatelli and F. R. Mansour, J. Pharm. Biomed. Anal., 2023, 235, 115609.
9. F. R. Mansour, R. M. Abdelhameed, S. F. Hammad, I. A. Abdallah, A. Bedair and M. Locatelli, Carbohydr. Polym. Technol. Appl., 2024, 7, 100453.
10. S. F. Hammad, I. A. Abdallah, A. Bedair, R. M. Abdelhameed, M. Locatelli and F. R. Mansour, TrAC, Trends Anal. Chem., 2024, 170, 117425.
11. E. Cesewski and B. N. Johnson, Biosens. Bioelectron., 2020, 159, 112214.
12. N. Kajal, V. Singh, R. Gupta and S. Gautam, Environ. Res., 2022, 204, 112320.
13. L. S. Xie, G. Skorupskii and M. Dinca, Chem. Rev., 2020, 120, 8536–8580.
14. R. Sakthivel, L.-Y. Lin, Y.-F. Duann, H.-H. Chen, C. Su, X. Liu, J.-H. He and R.-J. Chung, ACS Appl. Mater. Interfaces, 2022, 14, 28639–28650.
15. A. Bedair, R. M. Abdelhameed, S. F. Hammad, I. A. Abdallah, M. Locatelli and F. R. Mansour, Microchem. J., 2024, 204, 111132.
16. A. Bedair, R. M. Abdelhameed, S. F. Hammad, I. A. Abdallah, M. Locatelli and F. R. Mansour, Microchem. J., 2024, 207, 112198.
17. H. M. Bedair, A. Bdair, M. Hamed, M. Locatelli and F. R. Mansour, Talanta Open, 2024, 100374.
18. F. R. Mansour, S. F. Hammad, I. A. Abdallah, A. Bedair, R. M. Abdelhameed and M. Locatelli, TrAC, Trends Anal. Chem., 2024, 172, 117596.
19. D. Bhatt, S. Singh, N. Singhal, N. Bhardwaj and A. Deep, Anal. Bioanal. Chem., 2023, 415, 659–667.
20. L. M. Castle, D. A. Schuh, E. E. Reynolds and A. L. Furst, ACS Sens., 2021, 6, 1717–1730.
21. E. Stansell and R. C. Desrosiers, Yale J. Biol. Med., 2010, 83, 201–208.
22. S. Kaushal, N. Priyadarshi, A. K. Pinnaka, S. Soni, A. Deep and N. K. Singhal, Sens. Actuators, B, 2019, 289, 207–215.
23. P. Garg, N. Priyadarshi, M. D. Ambule, G. Kaur, S. Kaul, R. Gupta, P. Sagar, G. Bajaj, B. Yadav and V. Rishi, Nanoscale, 2023, 15, 15179–15195.
24. S. Hargol Zadeh, S. Kashanian and M. Nazari, Biosensors, 2023, 13, 619.
25. X. Guo, A. Kulkarni, A. Doepke, H. B. Halsall, S. Iyer and W. R. Heineman, Anal. Chem., 2012, 84, 241–246.
26. A. Gupta, S. K. Bhardwaj, A. L. Sharma, K.-H. Kim and A. Deep, Environ. Res., 2019, 171, 395–402.
27. A. A. Gill, S. Singh, N. Agrawal, Z. Nate, T. E. Chiwunze, N. B. Thapliyal, R. Chauhan and R. Karpoormath, Microchim. Acta, 2020, 187, 1–9.
28. N. T. Dat, N. N. Tien, N. T. T. Ngan and V. T. Thu, Analyst, 2023, 148, 1777–1785.
29. R. Rani, A. Deep, B. Mizaikoff and S. Singh, Electroanalysis, 2020, 32, 2442–2451.
30. P. Bhalla and N. Singh, Eur. Phys. J. B, 2016, 89, 1–8.
31. S. Ahmed, N. Shaikh, N. Pathak, A. Sonawane, V. Pandey and S. Maratkar, Tools, Tech. Protoc. Monit. Environ. Contam., 2019, 53–73.
32. Y. Liu, H. Yu, O. Alkhamis, J. Moliver and Y. Xiao, Anal. Chem., 2020, 92, 5041–5047.
33. M. Locatelli, A. Kabir, M. Perrucci, H. I. Ulusoy, S. Ulusoy, N. Manousi, V. Samanidou, I. Ali, S. I. Kaya and F. R. Mansour, Microchem. J., 2024, 111903.
34. T. Tsuruoka, S. Furukawa, Y. Takashima, K. Yoshida, S. Isoda and S. Kitagawa, Angew. Chem., 2009, 121, 4833–4837.
35. M. H. Rosnes, F. S. Nesse, M. Opitz and P. D. Dietzel, Microporous Mesoporous Mater., 2019, 275, 207–213.
36. D. J. Tranchemontagne, J. R. Hunt and O. M. Yaghi, Tetrahedron, 2008, 64, 8553–8557.
37. W. Zhang, M. J. Bojdys and N. Pinna, Angew. Chem., Int. Ed., 2023, 62, e202301021.
38. M. Kanno, T. Kitao, T. Ito and K. Terashima, RSC Adv., 2021, 11, 22756–22760.
39. Y. Song, X. Li, C. Wei, J. Fu, F. Xu, H. Tan, J. Tang and L. Wang, Sci. Rep., 2015, 5, 8401.
40. R. Niknam, M. Mousavi and H. Kiani, Food Bioprocess Technol., 2020, 13, 882–900.
41. F. Pei, Y. Lv, X. Cao, X. Wang, Y. Ren and J. Ge, Fermentation, 2022, 8, 422.
42. G. Gizer, M. Sahiner, Y. Yildirim, S. Demirci, M. Can and N. Sahiner, Curr. Res. Green Sustainable Chem., 2021, 4, 100110.
43. M. Heydari, M. Gharagozlou, M. Ghahari and S. Naghibi, Appl. Organomet. Chem., 2020, 34, e5994.
44. A. Asghar, N. Iqbal, T. Noor, M. Ali and T. L. Easun, Nanomaterials, 2019, 9, 1063.
45. B. Omkaramurthy, G. Krishnamurthy and S. Foro, SN Appl. Sci., 2020, 2, 342.
46. K. Yang, Y. Yan, W. Chen, H. Kang, Y. Han, W. Zhang, Y. Fan and Z. Li, RSC Adv., 2018, 8, 23671–23678.
47. A. R. Bagheri and M. Ghaedi, Arabian J. Chem., 2020, 13, 5218–5228.
48. B. Lukose, B. Supronowicz, P. St Petkov, J. Frenzel, A. B. Kuc, G. Seifert, G. N. Vayssilov and T. Heine, Phys. Status Solidi, 2012, 249, 335–342.
49. N. Elgrishi, K. J. Rountree, B. D. McCarthy, E. S. Rountree, T. T. Eisenhart and J. L. Dempsey, J. Chem. Educ., 2018, 95, 197–206.
50. M. I. González-Sánchez, B. Gómez-Monedero, J. Agrisuelas, J. Iniesta and E. Valero, J. Electroanal. Chem., 2019, 839, 75–82.
51. M. I. González-Sánchez, B. Gómez-Monedero, J. Agrisuelas, J. Iniesta and E. Valero, Electrochem. Commun., 2018, 91, 36–40.
52. S. Lin, P. M. Usov and A. J. Morris, Chem. Commun., 2018, 54, 6965–6974.
53. S. Jose, M. Ghosh and A. Varghese, Mater. Adv., 2024, 5, 3812–3823.
54. M. H. Jazayeri, H. Amani, A. A. Pourfatollah, H. Pazoki-Toroudi and B. Sedighimoghaddam, Sensing Bio-Sensing Res., 2016, 9, 17–22.
55. N. Sharon and H. Lis, Glycobiology, 2004, 14, 53R–62R.
56. H. Ghazarian, B. Idoni and S. B. Oppenheimer, Acta Histochem., 2011, 113, 236–247.
57. Garima, A. Kumar and A. Sachdev, Instrum. Sci. Technol., 2023, 1–18.
58. S. Balser, M. Röhrl, C. Spormann, T. K. Lindhorst and A. Terfort, ACS Appl. Mater. Interfaces, 2024, 16(11), 14243–14251.
59. P. Raj, M. H. Oh, K. Han and T. Y. Lee, Chemosensors, 2021, 9, 49.
60. S. Shahrokhian and S. Ranjbar, ACS Sustainable Chem. Eng., 2019, 7, 12760–12769.
61. A. Gupta, A. L. Sharma and A. Deep, J. Environ. Chem. Eng., 2021, 9, 104925.
62. M. Zhong, L. Yang, H. Yang, C. Cheng, W. Deng, Y. Tan, Q. Xie and S. Yao, Biosens. Bioelectron., 2019, 126, 493–500.
63. K. Balaji Viswanath, N. Krithiga, A. Jayachitra, A. K. Sheik Mideen, A. J. Amali and V. S. Vasantha, ACS Omega, 2018, 3, 17010–17022.
64. D. A. Rosen, J. S. Pinkner, J. N. Walker, J. S. Elam, J. M. Jones and S. J. Hultgren, Infect. Immun., 2008, 76, 3346–3356.
65. S. G. Stahlhut, V. Tchesnokova, C. Struve, S. J. Weissman, S. Chattopadhyay, O. Yakovenko, P. Aprikian, E. V. Sokurenko and K. A. Krogfelt, J. Bacteriol., 2009, 191, 6592–6601.
66. S. A. Zeiner, B. E. Dwyer and S. Clegg, Infect. Immun., 2012, 80, 3289–3296.
67. G. Dai, Y. Li, Z. Li, J. Zhang, X. Geng, F. Zhang, Q. Wang and P. He, ACS Appl. Nano Mater., 2022, 5, 9201–9208.
68. L. Sun, Y. Chen, Y. Duan and F. Ma, ACS Appl. Mater. Interfaces, 2021, 13, 38923–38930.
69. Y. Lou, Q. Jia, F. Rong, S. Zhang, Z. Zhang and M. Du, Food Chem., 2022, 395, 133618.
70. S. Xiao, X. Yang, J. Wu, Q. Liu, D. Li, S. Huang, H. Xie, Z. Yu and N. Gan, Sens. Actuators, B, 2022, 369, 132320.
71. X. Zhang, G. Xie, D. Gou, P. Luo, Y. Yao and H. Chen, Biosens. Bioelectron., 2019, 142, 111486.
72. R. Abedi, J. B. Raoof, M. Mohseni and A. B. Hashkavayi, Bioelectrochemistry, 2023, 150, 108332.
73. M. Locatelli, A. Kabir, M. Perrucci, S. Ulusoy, H. I. Ulusoy and I. Ali, Adv. Sample Preparation, 2023, 6, 100068.
74. F. R. Mansour, J. Płotka-Wasylka and M. Locatelli, Analytica, 2024, 5, 451–457.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4nj04605f

---