Introduction

During a baby’s early days, there’s a risk of elevated Bilirubin (BR) levels leading to hyperbilirubinemia, causing yellowing of skin and eyes1,2. BR includes water-soluble conjugated (CBR) and water-insoluble unconjugated (UBR) types. Healthy individuals have serum BR levels of 0.1–1.2 mg.dL− 1, comprising CBR (0.1–0.4 mg.dL− 1) and UBR (0.2–0.7 mg.dL− 1). Jaundice is clinically defined as blood BR levels exceeding 2.5 mg.dL[− 1 3.

The lipophilic nature of BR makes it insoluble in water, requiring hepatic conjugation in the liver for elimination5,6. Accumulation of unconjugated BR can lead to serious consequences, including hepatic dysfunction, mental disorders, brain damage, and even fatality, especially in newborns7,8. Phototherapy is employed to treat neonatal jaundice by exposing newborns to visible light, inducing the conversion of BR into lumirubin through photo-isomerization. Lumirubin, with lower toxicity and reduced albumin binding affinity, is eliminated primarily through urine, contributing to the yellow discoloration observed in jaundiced patients9,10. Due to these serious health concerns of untreated severe jaundice, the accurate and timely diagnosis and, consequently, early treatment of jaundice are arguably crucial. In recent years, numerous techniques have been investigated for the detection and measurement of BR as a diagnostic tool for biological disorders11,12,13.

Unfortunately, these approaches require long laboratory turnaround time from sample collection to publishing test results, high volumes of sample, and sophisticated laboratory equipment with trained operators, which consequently restrict their applications to laboratories and unable to meet the requirements of an effective sensor such as instantaneous response, high selectivity and sensitivity, recyclability, and high photo-stability. Several handheld portable BR devices are available for percutaneous recognition, providing results for Point-of-Care Testing. However, the accuracy of these results can be influenced by the infant’s skin color. Hemoglobin, a major light absorber in the dermis (particularly in the 400–425 nm and 500–600 nm) competes with BR absorption, impacting the outcome. As a result, percutaneous BR detections in newborns may not match the precision of blood-based procedures. A thorough literature survey reveals that there is still a need for the development of novel detection platforms that can effectively address the aforementioned challenges associated with the detection of BR14,15,16,17,18. Luminescence detection plays a very important role in analytical chemistry due to its visualization readout mode, fast response, operational simplicity, and very high sensitivity in a wide range of concentrations19.

To date, very few luminescent platforms have been utilized for the detection of BR, including gold nanocluster, graphene oxide, quantum dots, and metal–organic frameworks (MOF)20,21,22,23. Among the mentioned platforms, the luminescence probes based on metal-organic frameworks (LMOFs) are promising candidates. These compounds, with their optical properties, adjustable porosity, large surface area and high physicochemical stability, meet the criteria of an effective sensor24,25,26. Until now, LMOFs have been employed for sensing of variety molecules, but having both great sensitivity and selectivity towards the desired analyte are still rare27,28,29,30.

Moreover, the d − f heterometallic MOFs, featuring the intrinsic properties of d- and f-block, have potential applications in various fields such as luminescence. The d-block metal, like zinc with a d10 electron configuration, as the closed shell metals have a lower energy level for their triplet excited state that allows for a more favorable alignment with the excited state energy level of f-block elements, leading to enhanced efficiency in sensitizing the luminescence of lanthanide ions31,32,33. Within the Ln ion series, Ce(III) stands out as an economical element that exhibits a strong inclination towards complexation with N-heterocyclic carboxylate ligands. This preference makes it an excellent choice for the construction of d − f MOFs34,35,36,37,38,39.

Furthermore, the appealing and distinctive features of nanopaper made from bacterial cellulose (BC) have generated considerable interest in developing (bio)sensing platforms. The abundance of OH groups on steel-like nanofibres facilitates incorporation with guest materials as an excellent support host or scaffolding template40. BC nanofibers are created by cellulose-producing bacteria like Gluconacetobacter xylinus utilizing D-glucose monomers and possess an exceptionally crystalline structure with a three-dimensional (3D) network architecture of the numerous interweaved nanofibers that imparts excellent physical and chemical durability to BC41,42,43. The desirable traits of BC such as affordability, versatility, non-hazardous nature, and bio-decomposability are further complemented by the optical transparency, thermal durability, and strong mechanical and chemical integrity of BC nanopaper. These outstanding characteristics position BC as a favorably propitious matrix for the field of biosensing technology44,45,46,47,48,49,50. This research presents a novel jaundice diagnosis approach using an affordable and portable diagnostic kit with a smartphone-based color signal reader. The kit accurately monitors BR levels through a luminescent platform, which is quenched by BR, acting as an analytical signal captured by a smartphone camera and restored at specific wavelengths (Fig. 1). The study examines factors influencing the detection capability of sensing kit, evaluating attributes like specificity, sensitivity, linearity, and analysis of authentic samples for comprehensive assessment.

Fig. 1
figure 1

Visualization of a BR assay kit with photoluminescent nanopaper and smartphone readout.

Full size image

Experimental

Methods

Blood serum samples

Blood serum samples were from Imam Khomeini Hospital. Serum samples were collected from two healthy volunteers. The ethics committee of the laboratory of Tehran University of Medical Sciences approved all experimental protocols. Informed consent was obtained from all subjects. Research protocols were conducted under the Declaration of Helsinki and relevant guidelines and regulations of the Netherlands Code of Conduct for Research Integrity and the Academic Medical Center research code.

Materials

The reagents utilized were of analytical reagent grade and were used without any further purification. The nanopaper was supplied by Nano Novin Polymer Co. located in Iran.

Apparatus

The FT-IR data was obtained using an Aligent Technology Cary 360 FT-IR spectrometer within the range of 400–4000 cm− 1. UV-visible absorption measurements were performed using a PerkinElmer Lambda 650 spectrometer equipped with a 60 mm integrated sphere to obtain the absorption spectra. Emission spectra were measured using an Agilent Cary 5000 fluorescence spectrometer equipped with a 450 W Xe lamp. The majority of the luminescence spectra were recorded at room temperature. Thermogravimetric analysis (TGA) of compound was determined by using a Q500 thermal gravimetric analyzer (TA, USA) under a N2 atmosphere at a heating rate of 50 °C/min from 40 to 900 °C. Surface area measurements (BET) were conducted using an ASAP 2420 gas adsorption instrument. The powder X-ray diffraction patterns (PXRD) of the products were analyzed within the 2-theta range of 5–50° using a D2 X-ray diffraction analysis (Brucker). Element analysis was carried out using a TM3000 scanning electron microscope (SEM) (Hitachi, Japan) with a SwiftED 3000 accessory (Oxford). Transmission electron microscopy (TEM) images and EDS spectra were provided using a JEOL JEM2100 (Japan). A single crystal of CeZn-MOF was mounted on a Xcalibur Atlas diffractometer with Mo Kα radiation (λ = 0.71073 Å). The CrysAlis PRO 1.171.42.93 program was used for data reductions and cell refinements. The crystal structures were solved using the SHELXT-2014/5 program and refined with full-matrix least-squares methods based on F2 using the SHELXL-2018/3 program51. All non-hydrogen atoms were anisotropically refined. The crystal structure of CeZn-MOF in the cif-format has been deposited with the Cambridge Crystallographic Data Center under the depository numbers CCDC-2,322,594.

Synthesis of CeZn-MOF

The hydrothermal reaction was utilized to synthesize the compound CeZn-MOF. Firstly, a mixture of Ce(NO3)3.6H2O (86 mg, 0.2 mmol), Zn(NO3)2.6H2O (60 mg, 0.2 mmol), and pzdc (67 mg, 0.4 mmol) was prepared in 8 mL water. To this 4 mL of NaOH (0.1 M) solution was added. The resulting mixture was heated at 120 °C for 24 h in a Teflon-lined stainless steel autoclave, ultimately resulting in the pale-yellow crystals. (yield; 72%) Elemental analysis (%) calcd. for C24H24Ce2N8O24Zn: C, 24.98; H, 2.10; N, 9.71. Found (%): C, 25.30; H, 2.4; N, 10.1. Selected IR (KBr pellet, cm− 1): 3325, 2360, 1652, 1587, 1384, 1355, 1168, 1130, 1118, 896, 850, 784, 746.

Fabrication of BC@CeZn-MOF bioplatform

To fabricate the BC@CeZn-MOF biosensing platform, around 100 mg of BC nanopaper was immersed in 20 mL of aqueous solution of Ce(NO3)3.6H2O (20 mM). To react the metal ions with the O–H groups on the surface of BC, 4 ml of NaOH solution (0.1 M) was added to the above mixture and stirred for about 15 min. In the next step, 0.2 mmol of Zn(NO3).6H2O (60 mg) and 0.4 mmol of pzdc (67 mg) were added to the stirring mixture. The resulting mixture containing BC papers was heated at 120 °C for 24 h in a Teflon-lined stainless steel autoclave. The obtained BC@CeZn-MOF fabricated platform was washed three times with H2O and MeOH at room before further characterization. The mass of the synthesized MOF on the BC surface was found to be 137 mg. To optimize the amount of CeZn-MOF on BC, different concentrations of Ce(III) was utilized and in all experiments the molar ratio of 1:1:2 for Ce: Zn: pzdc and amount of BC were constant.

Recognition of BR via BC@CeZn-MOF platform

The experiment for detecting BR concentration using the smartphone-based BC@CeZn-MOF platform was carried out in a series of steps. Firstly, 5 µL of different concentrations of BR solutions in the range of 0.1–20 mg.dL− 1 using a PBS buffer (pH = 7.4) were added to the fabricated platform (circles with 1 cm diameter). Then, smartphone camera images of the test zones were captured in a dark chamber under 365 nm irradiation. Subsequently, the assay sections were exposed to blue light lamps (λ = 470 nm) for 20 min to allow for luminescence recovery. In addition, a blank sample was prepared using the same method without the addition of BR. The images were processed and analyzed using ImageJ software to determine the mean color intensity (in gray mode) of each test zone.

Real sample analysis

To evaluate the clinical performance of the newly advanced smartphone-based BC@CeZn-MOF platforms. The small quantity of human serum samples was then processed to the proposed approach for determining the amount of BR through the BC@CeZn-MOF platform. The samples were diluted 5 times using a PBS buffer solution. The standard solutions of BR, with different concentrations, were added to the diluted samples to prepare real samples and 5 µL of these samples were added to the test zone and after 5 min, the images were captured in a dark chamber under 365 nm irradiation.

Results and discussion

Description of the crystal structure

The synthesized d-f hetero-metallic CeZn-MOF in this study is isostructure to that of the praseodymium (PrIII) MOF previously synthesized in our group52. The crystal data are provided in Table S1. The asymmetric unit comprises a Ce(III) ion, a half Zn(II) ion, two pzdc2− ligands, three coordinated and one free H2O molecule (Fig. 2a). The Ce(III) ion is surrounded by a total of nine coordinating atoms, consisting of four O atoms from four pzdc2− ligands, two N atoms from two pzdc2− rings (two different coordination modes of pzdc2−), and three O atoms from three coordinated H2O, generating a capped square antiprismatic polyhedron as shown in Fig. 2b, d. The Zn(II) ion is surrounded by two N atoms originating from two pzdc2− rings, while two O atoms stem from the COO− groups in the equatorial plane, and two O atoms from the COO− groups in the axial position revealing a distorted octahedral polyhedron of ZnO6 (see Fig. 2c, e).

Fig. 2
figure 2

(a) Asymmetric unit with its atom labeling scheme, (b) Coordination environment of Ce(III), (c) Coordination environment of Zn(II), (d, e) Coordination geometry of Ce and Zn atoms, respectively.

Full size image

In polymeric construction, two O atoms from COO− group with anti – anti fashion and one N atom from the pyrazine ring of pzdc2− molecules connect the Ce ions together, the Ce···Ce distance of 6.134 (4) Å, generating a 1D sequence of Ce-pzdc. Moreover, the pzdc2− ligands through one O atom of COO− group and N atom from the pyrazine ring interconnect the 1D sequences of Ce-pzdc into 2D sheets. The Zn metalloligands as 2-connector linkers play a significant role in connecting the 2D Ce-coordination layers, the Ce···Zn separation of 6.766 (3) Å, leading to an extended 3D network coordination construction as illustrated in Fig. 3a, b.

Fig. 3
figure 3

(a) The 3D structure of CeZn-MOF and the exhibition of the Zn metallo-ligands in between 2D zigzag layers (purple), and (b) a simplified representation of the overall structure.

Full size image

Physical measurements of BC@CeZn-MOF composite

The PXRD analysis (Fig. 4a) verifies the phase integrity and the congruence between the BC@CeZn-MOF composite and its components. The PXRD pattern reveals distinct peaks at 2θ ≈ 8.3, 10.7, 13.7, 14.7, 16.8, and 18.5° corresponding to the characteristic diffractions of CeZn-MOF crystal planes (011), (021), (031), (002), (111), and (130), respectively, which demonstrates satisfactory conformity with the simulated pattern of the highly pure crystalline CeZn-MOF. Also, in BC@CeZn-MOF composite, the diffraction peaks at 2θ ≈ 14.6, 16.2, and 22.5° correspond to the (110), (1\(\:\stackrel{-}{1}\)0), and (002) planes of BC, respectively. Therefore, the two main components of the composite, CeZn-MOF and BC, are reserved in their pristine crystalline phases, which evidences that the CeZn-MOF is embedded into the BC matrix. Additionally, to access the stability of CeZn-MOF in PBS solution, the sample was soaked in PBS for 3 days. Analysis of the PXRD patterns revealed that the framework remained stable. CeZn-MOF exhibited remarkable resistance to acids and bases, with the framework maintaining integrity within a broad pH range of 4 to 9 (refer to Fig. S1).

Figure 4b shows the FT-IR spectra of the BC@CeZn-MOF composite and its components. The FT-IR analysis of CeZn-MOF reveals the broadband at 3415–3321 cm− 1, indicating the presence of H2O molecules. The peaks observed at around 1351 (s), 1388 (s) and 1592 (s), 1658 (s) cm− 1 are attributed to the symmetric and asymmetric vibrations of the COO− groups (νsym COO and νasy COO), which in turn confirm the coordination of the ligand to Ce(III) center. Also the OH broad bands of CeZn-MOF and BC at 3415 and 3360 cm− 1, respectively, displayed a blue shift in the FT-IR spectrum of the BC@CeZn-MOF and appear at 3340 and 3285 cm− 1. The results demonstrate the existence of physical interaction between the MOF and BC and that of its contribution to the excellent compatibility of mechanical properties of the kit.

The TGA analysis of CeZn-MOF and BC@CeZn-MOF composite was utilized to investigate thermal stability by examining the correlation between percentage weight losses and temperature increments, as illustrated in Fig. 4c. Weight loss occurred in three distinct stages: the first between 100 and 300 °C, the second between 300 and 500 °C, and the third between 500 and 700 °C. These stages have been assigned to specific processes: the first stage involves the evaporation of both free and coordinated H2O molecules, the second stage corresponds to the disintegration of the pzdc2− ligands, and the third and final stage entails the decomposition of the framework, resulting in the formation of CeO2 and ZnO as the ultimate products. Furthermore, a pronounced weight loss in the second stage was noticeable at approximately 310 °C for BC@CeZn-MOF, attributed to the decomposition of the cellulose structure.

N2 sorption isotherms were measured at 77 K to investigate the surface area and porosity (Fig. 4d). The sorption profile of CeZn-MOF could be classified as type I isotherm features, which shows that the CeZn-MOF framework is micro-porous. A hysteresis loop appeared at relatively high pressure, which is ascribed to the reversible uptake of nitrogen in the pores. BET surface areas of CeZn-MOF and BC@CeZn-MOF are 653.61 m2/g and 476.28 m2/g, respectively. BET surface areas of BC@CeZn-MOF is greater than that of bacterial cellulose (73.4 m2/g) due to the loading of MOFs on BC. The calculated total pore volume of BC, CeZn-MOF and BC@CeZn-MOF are presented in Table 1.

Fig. 4
figure 4

(a) PXRD pattern of the simulated, synthesized CeZn-MOF, BC, and BC@CeZn-MOF, (b) FT-IR spectra of CeZn-MOF, BC, and BC@CeZn-MOF, (c) TGA of CeZn-MOF, and BC@CeZn-MOF, (d) BET analysis of CeZn-MOF, and BC@CeZn-MOF. (All analyses of BC@CeZn-MOF correspond to the Ce3+ concentration of 20 mM).

Full size image
Table 1 BET surface area and total pore volume of BC, CeZn-MOF, and BC@ CeZn-MOF.
Full size table

The Fig. 5a shows that the size of BC nanofibers and the synthesized CeZn-MOF on the surface of the BC nanofiber is less than 100 nm.The initial aim of the present study was to synthesis the CeZn-MOF on BC nanofibers with desirable uniform structure with smooth surface, highest possible surface-to-volume ratio, and highest possible available functional groups. This way, the different amount of CeZn-MOF was utilized to optimize the luminescence intensity of fabricated bio-platform. The surface SEM images of BC nanofibers and BC@CeZn-MOF platform are presented in Fig. 5. With the initial Ce(III) concentrations between 5 and 15 mM, the regular beads of MOF exhibit a uniform dispersion on the BC nanofibers as a result of group linkages between the two main components. Also, with further increasing concentrations, 20 and 25 mM, more uniform MOF beads on nanofibers with higher bead densities were observed. Upon increasing Ce(III) concentrations to 35 mM, it becomes evident that the MOF particles are accumulated on the nanofibers and the size of the particles are increased (in Fig. 5b – h).

Fig. 5
figure 5

The SEM images of (a) BC nanofibers, (b) BC with different concentrations of Ce ions: (b) 5, (c) 10, (d) 15, (e) 20, (f) 25, (g) 30, and (h) 35 mM.

Full size image

The uniform dispersion and element distribution of BC@CeZn-MOF were further studied by SEM and EDX mapping (Fig. 6a, b), respectively. Figure 6b indicates that the platform contains Ce (green), Zn (purple), N (blue), O (red), and C (yellow) elements, uniformly distributed in the composite. TEM micrographs BC@CeZn-MOF (Fig. 6c) depict that the MOFs were intimately connected to fully wrapped up BC nanofibers, which is ascribed to the electrostatic interactions and van der Waals forces induced assembly between MOFs and BC. It is worth noting that MOFs were distributed uniformly throughout the composite nanofibers based on EDX elemental mapping analysis.

Fig. 6
figure 6

(a) SEM image of BC@CeZn-MOF, (b) EDX spectrum; overlay elemental mapping images of BC@CeZn-MOF: Ce (green), Zn (purple), N (blue), O (red), and C (yellow) elements, (c) TEM-mapping of BC@CeZn-MOF.

Full size image

Luminescent properties of CeZn-MOF and BC@CeZn-MOF

The luminescent characteristics and performance of CeZn-MOF and BC@CeZn-MOF for optical sensing were assessed based on the luminescence intensity under ambient conditions. The excitation spectra and solid-state luminescence of CeZn-MOF and BC@CeZn-MOF, revealed a blue emission at 460 nm (in Fig. 7a, b). As shown in Fig. S2, the absorption spectra demonstrates a strong and extensive broadband in the 280 and 360 nm, which can be ascribed to the chromophores π→π* transition of pzdc2− ligand and intra-configuration 4f\(\:\to\:\)5d transitions of Ce(III), respectively.

Fig. 7
figure 7

(a) Excitation spectra of the main transitions in CeZn-MOF, (b) Emission spectra of CeZn-MOF and BC@CeZn-MOF, inset Fig.: blue light emission of CeZn-MOF crystal and BC@CeZn-MOF platform.

Full size image

This is while no ligand-specific emissions were detected in the emission spectrum. Moreover, according to growing MOF nanoscale on BC nanofiber, the nanodimension, narrow band gap, high surface area, and higher oscillator strength due to higher electron-hole overlap integral in nanoscale MOF compared to bulk are key factors of higher intensity of BC@CeZn-MOF compared to CeZn-MOF53,54.

The evaluation aimed to optimize the conditions and improve the efficacy of the BR assay kit that relies on smartphones. Specifically, the effect of the quantity of CeZn-MOF on BC nanofibers and how it impacts blue light illumination were investigated. Varying Ce(III) ion concentrations between 5 and 35 were used to achieve the highest emission intensity. The experimental results show that the highest intensity of the PL signal was obtained at a Ce(III) concentration of 20 mM, which was selected for the following experiments (Fig. 8).

Furthermore, the luminescence intensity of BC@CeZn-MOF at various pH values in solution was examined. The fabricated sensing platform was soaked in a series of solutions with pH from 3 to 9 for five min, subsequently, luminescence emission spectra were recorded. As depicted in Fig. S3a the emission intensity of the platform remains constant across a broad range of pH.

To assess the PL stability and sensor storage of the created BC@CeZn-MOF, they were stored at room temperature and the corresponding PL was recorded over time. The results in Fig. S3b demonstrate that the PL intensities of the produced BC@CeZn-MOF remained consistent for a long period, indicating their suitability for practical applications and long-term stability.

Fig. 8
figure 8

PL intensities at different concentrations of Ce(III) ions: 5–35 mM, inset Fig.: blue emission intensity of the platforms at different concentrations of Ce(III) ions.

Full size image

Performance platform for BR sensing

The effectiveness of the luminescent BC@CeZn-MOF platform for visually detecting BR was examined. Fig. S4a and S4b represent the emission intensity of the platform before and after the sensing process, respectively. Once different concentrations of BR were used, the fabricated platform displayed color change from blue to dark under 365 nm (Fig. 9a). The smartphone captured the corresponding luminescence photos before and after adding the different concentrations of BR. Then, the color changes were quantified using ImageJ software. Based on Fig. 9b, as the concentration of BR increases, there is a notable decrease in the intensity of the emission spectrum of BC@CeZn-MOF. There was an excellent linear relationship between the gray intensity of sensing platform and BR concentration in the range of 0.1–20 mg.dL− 1 (with a correlation coefficient (r) of R2 = 0.9949), as demonstrated in Fig. 9c. The linear regression equation for BR is ΔS = 535.23x + 370.84, where ΔS (ΔS = S-S0 which S0 and S are signal intensities before and after adding BR, respectively) is the analytical signal and C is the BR concentration in mg. dL− 1. Thus, the value of gray intensity could be used for quantitative detection of BR. The LOD is estimated to be 0.027 mg.dL− 1 (0.46 µM), which is much lower than the minimum requirement of human serum.

Fig. 9
figure 9

(a) The fabricated BC@CeZn-MOF platforms under UV light, λ = 365 nm through increasing concentrations of BR (0.1 − 20 mg.dL− 1), (b) Luminescence emission intensity of BC@CeZn-MOF platform toward various BR concentrations, (c) Calibration plot of BR sensing using the developed smartphone assay kit in the range of 0.1 − 20 mg.dL− 1 of BR.

Full size image

Interference study

To evaluate the selectivity of the fabricated BC@CeZn-MOF platform for the BR determination, competitive experiments were carried out at 10 mg.dL− 1 of BR in the presence of some biomolecules and metal ions. The substances including (Cl−, Na+, K+, Ca2+, Zn2+, Mg2+,), glucose (Glu), uric acid (UA), urea, L-alanine (Ala), L-ascorbic acid (AA), albumin (Alb), and cyanocobalamin were selected to record the influence, according to the recommended procedure. As shown in Fig. 10, the results demonstrate that the presence of interfering chemicals had negligible effects on the signals. This finding highlights the excellent selectivity of the smartphone-based assay kit developed for BR detection. The presence of various quenchers in blood is a common challenge when it comes to detecting specific substances. However, in this case, only BR exhibited a response to the blue light emission and underwent photo-isomerization, resulting in rearrangement. Consequently, the BC@CeZn-MOF analytical response remained largely unaffected by the presence of other interference species. These measurements effectively demonstrate the selectivity and robust anti-interference capability of the assay.

Fig. 10
figure 10

Effect of some coexisting (bio)chemicals on the analytical response of the fabricated BC@CeZn-MOF toward BR sensing.

Full size image

Application of the method in real samples assay

To evaluate the actual applicability of the BC@CeZn-MOF assay kit in real samples, the platform was utilized as a sensing platform for the detection of BR in human serum samples of 5 infants and comparing the results of our developed sensor with the Audit Total BR Assay Kit (form MAN Company) as a clinical reference method for BR level detection. Fig. S5 shows the results of both methods that have been plotted against each other to compare the methods. The regression coefficient of this comparison (r = 0.9955) reveals the results, and the performance of our developed smartphone-based BR assay kit is comparable to reference methods that are routinely used for determination of BR level in blood samples and jaundice diagnosis in clinical laboratories. The obtained results indicate the usability of the developed smartphone-based assay kit for BR level detection at the point-of-care and also clinical laboratories.

To evaluate the practical efficacy of the luminescence platform, the standard method was employed to detect BR in human serum. The luminescence intensity was measured after diluting the real samples 5 times with PBS. The spiked BR exhibited a recovery ranging from 98 to 104%. The CeZn-MOF luminescence system exhibits good accuracy for the recognition of total BR, as illustrated by the relative standard deviation (RSD) being less than 3.3%. This indicates that the system holds promise as a reliable method. Table 2 presents the experimental results. Also, Table S2 shows a comparison of the performance of BC@CeZn-MOF platform in BR detection relative to other synthetic platforms.

Table 2 Recognition of total BR in real samples.
Full size table

Reusability

To assess the potential reusability of BC@CeZn-MOF in detecting BR, the PL of the developed BR sensor is quenched in the presence of BR and then restored upon blue light (λ = 470 nm) exposure during 20 min. As shown inset Fig. 11, the blue emission light of BC@CeZn-MOF was quenched in the presence of 10 mg.dL− 1 of BR under λ = 365 nm irradiation, while the recovered BR selectively undergoes photo-isomerization and photo-oxidation upon exposure to UV light (λ = 470 nm), converting it into colorless oxidation products, lumirubins, acting as non-PL quencher species (as shown in Fig. S6). Furthermore, the assay kit demonstrated remarkable restoration of its initial luminescence characteristics even after undergoing three cycles of repeated luminescence detection experiments, as depicted in Fig. 11.

Fig. 11
figure 11

The correlation between emission intensity ratio of the sensing platform and cycles. Inset: quenching and recovering luminescence intensity of BC@CeZn-MOF composite under λ = 365 and 470 nm, respectively, in presence of BR.

Full size image

Possible mechanism of the sensing system

Having confirmed the exceptional detection capability of BC@CeZn-MOF towards BR, delving into a comprehensive investigation of its sensing mechanism becomes imperative. This mechanism can be corroborated through the examination of spectral overlaps between MOF and BR. The emission band of CeZn-MOF exhibit a significant overlap with the absorbance spectrum of BR (Fig. 12a) indicative of IFE, the phenomenon in which an excited donor transfers energy to an acceptor group through a non-radiative process, occurring between the two species55,56,57. To provide a comprehensive illustration of the quenching process, an analysis of time-resolved luminescence decay was conducted both in the absence and presence of BR as depicted in Fig. 12b, the luminescence lifetime of MOF (τCeZn−MOF) exhibited a reduction from 0.38 µs to 0.21 µs upon the introduction of BR (τMOF + BR). The outcome verifies that the luminescence quenching observed in CeZn-MOF upon BR interaction is primarily attributed to a static quenching effect58,59,60. Moreover, the restoration of luminescence can be attributed to the intrinsic photo-isomerization of BR under specific light conditions with a wavelength of λ = 470 nm. In this process, the bridging exocyclic double bond in each of the two dipyrrinones of BR adopts a Z configuration, resulting in the formation of lumirubins structure as a non-photoluminescence quencher. This leads to a distinct enhancement of the PL signal.

Fig. 12
figure 12

(a) Overlap of emission spectra of CeZn-MOF and absorbance spectra of BR, (b) Time-resolved decay curves of MOF and MOF + BR in the absence and presence of BR.

Full size image

Conclusions

To summarize, this contribution has successfully introduced a novel smartphone-based diagnostic bio-platform that is user-friendly, affordable, safe, and portable. This kit allows for the diagnosis of jaundice by visually determining the levels of BR in real samples. An innovative luminescent assay kit using BC@CeZn-MOF composite coupled with smartphone technology was successfully designed, enabling highly sensitive and specific detection of BR. Furthermore, the luminescence sensing probe can be effectively quenched in the presence of BR due to the overlapping absorption band of BR and PL of the platform via IFE quenching mechanisms. It is important that the developed assay kit can also be considered as a green recognition platform because of uses BC nanopaper as non-venomous, affordable, environmentally material, and reduces the need for excessive sample collection. Moreover, the present platform demonstrates notable characteristics such as remarkable sensitivity, and excellent selectivity in detecting BR and displayed ultra-fast response (0.21 µs) with a low detection limit (0.027 mg.dL− 1). It stands out as the luminescent platform based on lanthanide MOF that overcomes interference from other biomolecules and metal ions in human serum samples. Additionally, the usability strategy for detecting BR in the developed smartphone-based assay kit relied on selectively recovering the quenched PL of BC@CeZn-MOF platform when exposed to blue light (λ = 470 nm). This recovery is attributed to the photo-isomerization of BR to lumirubins isomer. The finding reveals that the successfully fabricated lanthanide-based diagnostic bioplatform has demonstrated significant potential as a robust sensing tool for detecting BR in point-of-care settings and routine clinical medicine.