Intergradation of UiO‐66 Nanoparticles with Expanded Graphite for Electrocatalytic Determination of Nitrite and L‐Cysteine

In this study, metal–organic framework (MOF) nanoparticles of UiO‐66 are integrated with expanded graphite (EG) (UiO‐66/EG) by a facile solvothermal approach. The advantages of this nanocomposite UiO‐66/EG overcome the poor electronic conductivity and slow diffusion of MOFs for their electrochemical applications. Through electron microscopy, X‐ray diffraction, X‐ray photoelectron spectroscopy, and electrochemical techniques, the morphology, surface area, and physicochemical properties of this UiO‐66/EG nanocomposite are characterized. The UiO‐66/EG nanocomposite exhibits superior sensing performance over the UiO‐66 and EG when used for nitrite and L‐cysteine determination. This includes less positive oxidation potentials and enhanced oxidation currents. Using the UiO‐66/EG nanocomposite, the nitrite oxidation peak current is linear with a concentration range of 0.20 μm to 13.15 mm with the lowest limit of detection (LOD) of 0.06 μm (S/N = 3). Meanwhile, superior performance is demonstrated for L‐cysteine monitoring, where the oxidation peak current is linear over the L‐cysteine concentration in the range of 0.5–250 μm and of 0.25–3.50 mm and a LOD of 0.28 μm (S/N = 3). This UiO‐66/EG/GCE nanocomposite is successfully exploited to detect nitrite in food samples and to measure L‐cysteine in juice samples. Therefore, the proposed sensing platform enables the fabrication of high‐performance electrochemical sensors to accurately quantify nitrite and L‐cysteine in complex matrixes.


Introduction
Nitrite is often used as an antimicrobial agent and a food additive in our daily life. Its high content in the human body not only reduces the blood capacity to transport oxygen from irreversible conversion of oxyhemoglobin to methemoglobin, but also leads to methemoglobinemia. [1,2] The World Health Organization (WHO) recommended the guideline value of nitrite in drinking-water is 3 mg L −1 as nitrite ion (equivalent to 0.9 mg L −1 as nitrite-nitrogen). [3] The current acceptable daily intake level for nitrites as food additives is 0.07 milligrams per kilogram of body weight per day. [4] In addition, nitrite interacts with compounds containing amine functional groups to form toxic and carcinogenic nitrosamines, causing gastrointestinal tumors and stomach cancer. In this regard, it is very much essential to determine nitrite, especially in its ultrarace concentration level. Moreover, the existence of nitrite ions in water samples and human food product sources can cause various human diseases. On the other hand, L-cysteine, a thiol-containing amino acid that is found naturally in the human body, plays a multiple biological function in human beings. It is also a precursor in the food, pharmaceutical, medicines, or cosmetics industries. As examples of these applications, the reaction of L-cysteine with sugars yields meat flavors. [5] Note that L-cysteine is also used as a supplement to promote skin and hair health, to boost the immune system, and to combat inflammatory related problems and osteoporosis. [6] Although our body can make cysteine from the amino acids methionine and serine, the short supply of cysteine needs to be supplemented with L-cysteine to fill the gap. Unfortunately, cysteine toxicity can cause multiple disorders ranging from heart to renal (kidney) diseases.
In these contexts, different methods have been developed and employed for nitrite and L-cysteine detection, including chromatography, spectrophotometry, capillary electrophoresis, chemiluminescence, colorimetry, fluorescence spectroscopy, high-performance liquid chromatography, and electrochemical methods. [7][8][9][10] In spite of their excellent sensitivity, these methods are associated with some disadvantages, including relatively costly equipment, cumbersome sampling procedures, and toxic www.advancedsciencenews.com www.advsensorres.com reagents. On the other hand, electrochemical methods are very popular because of their simple operation conditions, fast and sensitive response/detection. [11][12][13] However, the oxidation of nitrite on bare electrodes are known to often require high potentials. The electrodes can be poisoned by substances produced during electrochemical reactions. [14] For L-cysteine, it is an electroactive amino acid because of the presence of a thiol moiety at its side chain. The oxidation states of L-cysteine on traditional electrodes are in a wide potential range. [15] The signal is generally very weak even at high overpotentials. Therefore, appropriate sensing materials or catalysts are frequently searched to efficiently lower the oxidation potentials of both nitrite and L-cysteine.
Metal-organic framework (MOF), consisting of metal-based nodes and organic ligands, is one of the most exciting classes of chemical structures to be discovered in the past decade. The vast majority of a MOF possesses its remarkable surface area, porosity, tunable structure, controllable and fascinating morphology, and modifiable surface property. [16,17] Depending on these features, numerous applications of MOFs have been developed in different fields, such as for energy storage, [18] gas storage and separation, [19] liquid separation and purification, [20] drug delivery, [21] catalysis, [22] and sensors development. [23] In fact, MOFs can be typically considered to be promising electrochemical sensing platforms, while the sensitivity, stability, selectivity, and accuracy of these MOF sensing platforms can be determined by different factors. [17,24] The high porosity and the large surface area of a MOF are beneficial to improve the detection sensitivity. On the other hand, the selectivity of this platform/sensor can arise from the MOF derived core-shell or defect structure and the complex recognition elements contained inside MOFs. Unfortunately, the electrochemical sensing applications of MOFs are significantly hampered due to their inferior electrocatalytic abilities, chemical stability and water stability, and poor electrical conductivity. [25] To solve such challenges, pristine MOFs have been coupled with other highly conductive and mechanically durable materials (e.g., carbon materials, metal nanoparticles, and conducting polymers). Those integrated materials or composites have been sufficiently enhanced electrochemical sensing performance of MOFs. Among various MOFs, Zrbased MOFs inherit the advantages of the MOFs and demonstrate enrichment with exceptional chemical and thermal stability, enabling the application of Zr-MOFs in water, at high temperatures, and under other harsh conditions. [26] In this study, we integrate expanded graphite (EG) with a UiO-66 MOF (here nanoparticle) to form a composite that is further used as electrochemical sensors for the detection of both nitrite and L-cysteine. The UiO-66/EG composites were synthesized via an ultrafacile, cost-effective solvothermal approach. The morphology and structure of as-synthesized composites were characterized by means of electron microscopy, X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), and electrochemical techniques. Due to the synergistic effect of porous UiO-66 and loosely packed EG nanosheets, the composite shows superb electrocatalytic activity toward the oxidation of nitrite and L-cysteine. The electrochemical sensing platform based on the UiO-66/EG composite exhibits outstanding sensing performance for both nitrite and L-cysteine, namely with their ultralow limit of detections (LODs) and ultrawide linear determination ranges, fast current response, as well as high sensitivity. Mean-while, this electrode also possesses satisfactory and feasibility of monitoring nitrite and L-cysteine in real food samples.

Characterization of the UiO-66/EG Nanocomposite
A preliminary analysis of UiO-66 nanoparticles, EG, and UiO-66/EG composites was performed using XRD (Figure 1). EG exhibits a narrow and high (002) peak (black) due to the orientation of the reticulated aromatic rings carbon layers in the 3D arrangement. [27] UiO-66 and UiO-66/EG samples, on the other hand, exhibited the intense diffraction peaks at 2 = 7.3°and 8.5°( Figure 1A), corresponding to (111) and (002) planes of the UiO-66 nanoparticles, respectively. In addition, other minor diffraction peaks (as assigned in Figure 1A) also agreed well with the face-centered cubic lattice of the UiO-66 crystal. [28] From the image, UiO-66 exhibits an excellent crystallinity evidenced by the sharp diffraction peak. When compared with that of UiO-66, the UiO-66/EG composite has similar peak positions and slightly weaker peak intensities. The UiO-66/EG composite exhibits the same characteristic diffraction peak (002) as EG at ≈26.4° (Figure 1A), indicating a successful combining of UiO-66 with EG. Based on the similarity of the diffraction patterns, it appears that EG does not interfere with the crystallization of the UiO-66 sample. [29] Furthermore, the lower intensities of the peaks in the UiO-66/EG composite might be attributed to the distortion of the UiO-66 structure following the addition of EG with a high content. Consequently, the UiO-66 sample remains a key component of the UiO-66/EG composite while the UiO-66/EG composite continues to maintain its good crystal structure.
In order to confirm the chemical states of each element in the composite, XPS spectra of the UiO-66/EG composite were analyzed. As shown in Figure 1B, the XPS survey of the UiO-66/EG composite confirms the presence of Zr 3d, C1s, and O 1s elements in the composite. Moreover, two signals of Zr 3d with the binding energies of 183.4 and 185.8 eV ( Figure 1C) are ascribed to its 3d 5/2 and 3d 3/2 , respectively. They indicate a Zr 4+ state in the composite. [29,30] The O 1s XPS spectrum ( Figure 1D) shows two XPS peaks at 532.0, 532.8, and 533.7 eV, which can be ascribed to Zr─O─C, C═O, and C─OH, respectively. [31] In the high-resolution C 1s XPS spectrum ( Figure 1E), the main peak is observed at 284.8 eV, which corresponds to the sp 2 carbon atom. Other weak peaks located at 285.3 and 288.5 eV stems from the epoxy/hydroxyl groups (C─O) and carbonyls (C═O), respectively. [32] These features correspond to the terephthalic acid in the UiO-66 sample. To explore the pore structure and BET surface area of the as-prepared samples, N 2 adsorption-desorption measurements were conducted ( Figure 1F). UiO-66 has a large surface area (999.0 m 2 g −1 ) and its total pore volume is 0.873 cm 3 g −1 . This evidently indicates the microporous nature of UiO-66. After the combination of EG and UiO-66, the BET (Langmuir) specific surface area, total pore volume of the UiO-66/EG decreased and the value were 218.0 m 2 g −1 and 0.275 cm 3 g −1 . The sharp reduction of BET (Langmuir) specific surface areas, total pore volume may be explained by the fact that additional high content of EG is prone to self-stacking and the mesopore volume tends to increase significantly in the composite compared with that of UiO-66.  Scanning electron microscopy (SEM) and transmission electron microscopy (TEM) images of the UiO-66/EG and UiO-66 samples were recorded. The used EG possesses an irregular porous structure composed of graphite nanosheets separated via voids and crevices (Figure 2A). In addition, the EG sheets with their wrinkled layer structure are also clearly visible. The synthesized UiO-66 particles has a round shape and exhibit a degree of agglomeration ( Figure 2B), with a particle size of ≈60-80 nm. However, in the UiO-66/EG composite, these UiO-66 nanoparticles are well dispersed ( Figure 2C) and present an amorphous polyhedron nanocrystal of UiO-66 with ≈60-100 nm particle size.
It appears that the thin layer of EG is surrounded by nanospherelike UiO-66, suggesting that the EG is decorated with UiO-66 (inset in Figure 2C). EG sheets are known to exhibit a strong coordination capability with metal ions. The dispersion of UiO-66/EG nanocomposites can be accounted for the coordination of functional groups on EG sheets with Zr 4+ metal nodes, which inhibits the aggregation of UiO-66. [33] In this way, the UiO-66 nanoparticles are densely and homogeneously grown on the surface of EG layers. The TEM image of the used EG ( Figure 2D) also shows a few EG sheets with irregular shapes and transparency, whereas these sheets resemble crumpled and paper-like structures. In the case of the TEM image of the UiO-66/EG composite, the UiO-66 nanoparticles are seen to be uniformly distributed across the layers of EG ( Figure 2F), compared to agglomerated polygonal UiO-66 nanoparticles in the absence of EG ( Figure 2E). Its ultrathin dimensions allow the interior and outline of the EG to be virtually transparent. Due to the obvious wrinkles in EG, there is a large area available to load UiO-66 (inset in Figure 2F). Based on all these results, it is evident that the UiO-66/EG composite was successfully synthesized.
A crucial factor for the performance of the electrochemical sensor is the ability of the modified nanomaterial to transfer electrons onto the glass carbon electrode (GCE). The electrochemical properties of these materials-based electrodes, namely EG/GCE, UiO-66/GCE, and UiO-66/EG/GCE electrodes were investigated by use of cyclic voltammograms (CV). Among all electrodes, reversible redox peaks are observed on the CVs curves in 0.1 m KCl solution containing 5 mm [Fe(CN) 6 ] 3−/4− ( Figure 3A). The redox peak currents follow the order of GCE < EG/GCE < UiO-66/GCE < UiO-66/EG/GCE. Accordingly, this indicates that the synergistic effect of UiO-66 and EG accelerates the electron transfer rate of the modified electrode. The largest current signal of a UiO-66/EG/GCE partially results from its largest specific area. To estimate the electroactive surface area (A) of above prepared electrodes, CVs at different scan rates were accomplished. The peak currents are proportional to the square roots of the scan rates in the case of all four electrodes ( Figure 3B). Specifically, potassium ferrocyanide was used as a redox probe, and CVs were performed according to the Randles-Sevcik equation [34,35] I p = 2.69 × 10 5 n where I p is the peak current, n is the number of electrons transferred in the redox reaction, A is the electroactive surface area, D is the diffusion coefficient of the probe, C is the concentration of the reaction species in the electrolyte, and v is the scan rate. The calculated electroactive surface area of a UiO-66/EG/GCE is estimated to be 0.135 cm 2 , which is an increase of 107% over that of a GCE (0.065 cm 2 ). The electroactive surface areas of both an EG/GCE (0.074 cm 2 ) and a UiO-66/GCE (0.094 cm 2 ) are larger than that of a GCE by a factor of 15% and 45%, respectively. Moreover, it is anticipated that the electron transfer rate of a UiO-66/EG/GCE will be further accelerated, resulting in the highest redox currents on this electrode. The electrochemical impedance spectroscopy (EIS) was then employed to calculate and compare the electron transfer rates of these modified electrodes. To fit the experimental data, a Randles equivalent circuit model was used (inset to Figure 3C), where R ct , R s , C dl , and Z w represent charge transfer resistance, electrolyte resistance, double-layer capacitance, and Warburg element, respectively. As illustrated in the Nyquist plots ( Figure 3C  3 Ω) is observed when the UiO-66/EG composite is coated on a GCE. Considering the R ct values of the UiO-66/EG composite, it appears that the material exhibits an accelerated electron transfer rate, consistent with the findings from earlier studies of UiO-66-NH 2 /CNT nanocomposite [36] and UiO-66-NH 2 @PANI nanocomposite. [37] This feature partially originates from improved conductivity of the UiO-66 nanoparticles with aid of EG and/or the cooperation between UiO-66 and EG. In other words, the integration of UiO-66 and EG or the UiO-66/EG composite can serve as an electrode material that enhance electron transfer between the analytes and the electrodes.

Nitrite Sensing on the UiO-66/EG Nanocomposite
To explore the electrocatalytic activity of modified electrodes toward nitrite, the voltametric response of nitrite on these electrodes was investigated in detail by scanning the electrodes in 0.1 m pH 6.5 phosphorus buffer solution (PBS) at a scan rate of 100 mV s −1 (Figure 4A). In the PBS, no electrochemical signals but only background current can be obtained within the scanned potential range in the absence of nitrite (black lines, curve a-d).
After the injection of 1 mm nitrite, the bare GC electrode displays a broad and weak oxidation peak response (curve a"), ascribed to the oxidation of nitrite, at 1.05 V, due to poor electrocatalytic performance and a slow electron transfer rate. Differently, this oxidation potential shifts to 0.75 V on an EG/GCE (curve b"). The result indicates that EG can effectively reduce the oxidation potential of nitrite. Although it shifts back to 0.92 V the UiO-66/GCE (curve c"), its magnitude is much increased when compared with that on a GCE and an EG/GCE, probably due to an increased surface area from the modification of the UiO-66 nanoparticles onto a GCE. On a UiO-66/EG/GCE (curve d"), the highest oxidation current and an oxidation peak potential at 0.82 V are noticed. This is evidently due to the largest surface area of a UiO-66/EG/GCE among these electrodes (namely the most abundant active sites on the composite surface). It is observed on all electrodes that there are no discernable reduction peaks, indicating that the oxidation of nitrite is irreversible. The UiO-66/EG/GCE electrode exhibits noteworthy electrocatalytic activity toward nitrite oxidation and is the most promising electrode for nitrite sensing.
In order to improve the electrochemical performance of the sensor, the optimal conditions were determined by the recording of electrochemical signals with a UiO-66/EG/GCE in 1.0 mm nitrite in PBS solutions at diverse pH values (5.0-8.5). Upon analyzing CV curves, it appears that with an increase in pH value, the peak current of NO 2 − first increases and then decreases (Figure 4B). The peak current in this experiment reaches its maximum at pH 6.5, which is attributed to the instability of nitrite anions in an acidic medium (pH < 6.0, decomposition of NO 2 − into NO and NO 2 ). In alkaline media, on the other hand, the oxidation of nitrite is more difficult due to the absence of sufficient H + (pH > 7.0). [38,39] Meanwhile, the anodic peak potential (E pa ) of nitrite shifts negatively with an increase of the pH value from 5.0 to 8.5, which indicates that the proton concentration did not affect the peak position. It appears that the oxidation process does not follow the Nernstian relationship between E pa and pH, indicating the irreversibility of NO 2− oxidation and the fact that the electric oxidation mechanism of nitrite is driven by kinetics process. [40] www.advancedsciencenews.com www.advsensorres.com The PBS with a pH value of 6.5 was selected as the optimized pH value.
Moreover, the kinetic performance of nitrite oxidation at a UiO-66/EG/GCE was examined by cyclic voltammetry at diverse scan rates (40-200 mV s −1 ). As shown in Figure 4C, the oxidation peak potential (E pa ) shifts positively and the anodic peak current (I pa ) is regularly increasing gradually when the scan rate (v) enhanced from 40 to 200 mV s −1 . The linear behavior between the oxidation peak current with the square root of scan rate is obtained. The corresponding linear regression equation is expressed as I pa ( A) = 0.274 + 1.98 1/2 (mV s −1 ) with a correlation coefficient R 2 = 0.993 ( Figure 4D). According to the results, a diffusion-controlled electrochemical process dominates the electron transfer on the UiO-66/EG/GCE electrode, as previously reported. [41] Observations on the electrochemical oxidation of nitrite at a UiO-66/EG/GCE electrode are consistent with previous reports. [42,43] Additionally, there is a good linear relationship ( Figure 4E) between the logarithm of peak current (log I p ), and the logarithm of scan rate (log v). The linear regression equation can be expressed as log I pa ( A) = 0.417 log (mV s −1 ) + 0.509, as well as the corresponding R 2 = 0.992. The slope value of 0.417 is close to the theoretical value (0.50) of the electrochemical diffusion control process. [44] On the other hand, it can be seen that the E pa is linearly proportional to ln v within the scan rate range (Figure 4F). The linear regression equation can be given as E pa (V) = 0.0707 log v (mV s −1 ) + 0.705 and the regression coefficient R 2 = 0.997. For an irreversible electrochemical process, the correlation of E pa and log can be described by Laviron's equation: where is the transfer coefficient, n is the number of involved electrons in the rate determining step, R, T, and F are gas constant, room temperature and Faraday constant, respectively. K is a constant. According to the slope of E pa versus Iog , (1-) n was calculated to be 0.418. The value of the Tafel slope ( Figure 4F) is an indication of the number of electrons transferred in the rate-determining step for an irreversible process. In present work, the Tafel slop b value of 0.141 V decade −1 is very close to the typical range for one electron rate-limiting reaction. Taking n = 1 into account, the calculated value of is 0.59, which indicates that the transition state is more likely to form the product and favor the oxidation reaction of nitrite. In addition, the total number of electrons (n) involved in the overall nitrite oxidation can be determined from the slope of the I pa versus 1/2 plot via the Randlese Sevcik equation, which is applicable for a completely irreversible diffusion-controlled reaction. The value of n was calculated to be 2, suggesting that there are two electrons involved in the oxidation process.
The overall process of nitrite ion oxidation is then expressed by the following equations, where the nitrate (NO 3 − ) ion is suggested as the final product. [45][46][47] In the first step, UiO-66/EG interacts with nitrite to form the UiO-66/EG -NO 2 − complex, which then loses an electron to form nitrogen dioxide (NO 2 ). Following the disproportionation reaction, NO 2 is decomposed into nitrate and nitrite, followed by the electrooxidation of nitrite on UiO-66/EG/GCE to produce nitrate as a product Amperometry is generally known as a reliable technique for monitoring trace analytes because it allows a quick evaluation of the sensor and contributes to its key characteristics such as sensitivity and response time. To investigate the effect of applied potential on the current response of a UiO-66/EG/GCE toward nitrite detection, the amperometric i-t curves were recorded upon successive additions of nitrite into the 0.1 m pH 6.5 PBS at the corresponding potentials ( Figure 5A). Increases in the applied potential result in an increase in the current response and the maximum current response can be obtained at 0.9 V. A high applied potential leads to background currents and activates other electroactive substances, whereas a low applied potential results in poor electrochemical performance. As the noise signal is more pronounced at 0.90 V, a potential of 0.85 V was selected as the potential for quantitative detection of nitrite.
Under all the optimal conditions, a nitrite sensing platform was constructed on a UiO-66/EG/GCE. The sensitivity, linear range, and reduced limit of detection (LOD) of the nitrite detection on a UiO-66/EG/GCE was evaluated using amperometry. Taking the i-t curves of a UiO-66/EG/GCE recorded at applied potential of 0.85 V and after successive additions of nitrite to continuously stirred 0.1 m pH 6.5 PBS (Figure 5B), one can observe that i) the response reaches the maximum steady signal (I s ) within 3 s for each addition of nitrite with an interval of 50 s, and ii) an obvious reaction occurs when NaNO 2 concentration (C) is as low as 0.2 m. After each increase of NaNO 2 , the current response increases sensitively and rapidly, which shows a nice linear dependence. The calibration curve ( Figure 5C) appears to be linearly related to nitrite concentration in the range of 0.2 m to 13.15 mm with a regression equation of I ( A) = 5.06 + 25.8C (mm) (R 2 = 0.996). The calculated LOD and the sensitivity are 0.06 m (S/N = 3) and 348 A mm −1 cm −2 , respectively. As nitrite concentrations exceed 13.15 mm, the sensitivity for nitrite detection deteriorates, probably due to the adsorption of intermediates on the active sites of a UiO-66/EG/GCE during the nitrite oxidation process. The observed LOD and linear range for nitrite detection at UiO-66/EG/GCE was superior than that of other reported modified electrodes (Table 1 [48][49][50][51][52][53] ).
A series of anti-interference experiments was conducted for the detection of nitrite on a UiO-66/EG/GCE to evaluate its selectivity. The i-t curves were collected with an initial addition of 2.5 m nitrite, followed by successive addition of 100-fold concentration of interfering species (e.g., KCl, CaCl 2 , NaNO 3 , Na 2 SO 4 , and NaCl), along with 2.5 m nitrite ( Figure 5D). Except for the apparent current responses to additions of nitrite, no conspicuous reaction is observed when the interference agents are added. The results demonstrate that a UiO-66/EG/GCE sensing platform possesses good selectivity toward nitrite determination, and it is therefore suitable for nitrite analysis in practical testing.
To assess its practical applicability, the reproducibility and storage stability of UiO-66/EG/GCE were examined. Using six parallel freshly prepared electrodes for the measurement of 1.0 mm nitrite, the reproducibility was measured using the CV technique. The relative standard deviation (RSD) was calculated to be 4.2%, indicating good reproducibility among different electrodes. Fur-  thermore, reproducibility is also assessed by recording the CV of UiO-66/EG/GCE with 1.0 mm nitrite several times with a 1 min interval between each cycling. The RSD of the current responses is 3.8% for 6 continuous cycles. The long-term stability of this sensor was tested by continuously recording the i-t curves of one UiO-66/EG/GCE toward 100 m nitrite for one week at each day intervals. An RSD of 3.9% was observed in three repeated assays, and the response of nitrite (100 m) retains about 93% of the initial. Clearly, these results demonstrate a fair reproducibility and stability of UiO-66/EG/GCE as an efficient nitrite sensor.

L-Cysteine Sensing on the UiO-66/EG Nanocomposite
The as-fabricated UiO-66/EG/GCE was further applied to monitor L-cysteine. We measured the CVs of a GCE, an EG/GCE, a UiO-66/GCE, and a UiO-66 /EG/GCE in the absence and presence of 1.0 mm of L-cysteine in 0.1 m pH 6.5 PBS at a scan rate of 100 mV s −1 (Figure 6A). Within the scan potential range, there seem to be no observable peaks in the CVs of the electrodes in the blank solution, that is, in the absence of L-cysteine (curve a-d). After the addition of 0.1 mm L-cysteine, a broad oxidation wave is found at around 0.9 V. On a bare GCE, its peak current is 2.4 A (curve a"). When an EG/GCE or a UiO-66/GCE is used, the current is increased to 11.2 and 8.5 A (curve b", c"), respectively. Here, the onset potential for L-cysteine oxidation is relatively positive, which demonstrates that it is more difficult to initialize electrochemical redox reaction of L-cysteine at these three electrodes. Interestingly, when a UiO-66 /EG/GCE is employed (curve d"), a remarkable anodic peak is observed only with an onset potential of 0.4 V. The oxidation current is even enhanced to 16.5 A. The result also indicates that a UiO-66/GCE features good electrochemical activity toward L-cysteine oxidation. Overall, L-cysteine has been transformed into L-cystine during the electrocatalytic sensing process when two electrons and protons are released (Equation (7)) [54] 2L − cysteine → L − cystine + 2e − + 2H + The electrochemical performance of UiO-66/GCE modified electrode at increasing pH range from 5.0 to 8.5 in 0.1 m PBS containing 1 mm L-cysteine were recorded. A plot of oxidation peak current density versus pH ( Figure 6B) depicts the variation in peak potentials as function of pH. Within the pH value range from 5.0 to 8.5, a maximum oxidation current observed at pH 6.5. A linear relationship between E pa and pH yields a linear regression equation: E pa (V) = −0.068 pH + 1.12 (R 2 = 0.99). A slope of −68 mV pH −1 confirms that the number of electrons transferred is equal to that of protons involved. The CVs at different scan rates were further recorded ( Figure 6C). The oxidation peak current intensities increase progressively as the scanning rates increases. The observed anodic peak currents are proportional to the scanning rates with the correlation coefficient of 0.998 ( Figure 6D), indicating that the electrochemical oxidation of L-cysteine at a UiO-66 /EG/GCE is controlled by an adsorption process.
The optimization of applied potential for amperometric detection of L-cysteine on a UiO-66/EG/GCE was conducted ( Figure 7A). The applied potential applied at a UiO-66/EG/GCE was varied from 0.50 to 0.70 V. During this process, 100 m Lcysteine was successively added into the stirred 0.1 m pH 6.5 PBS. Considering to achieve a large and stable oxidation current as well as to avoid interference, an applied potential of 0.60 V was selected. Under these optimal conditions, amperometric i-t curves of L-cysteine at a UiO-66 /EG/GCE (holding at 0.60 V) were recorded ( Figure 7B), where the stable response is increased step wisely with the successive introduction of L-cysteine The UiO-66 /EG/GCE displays a sensitive and fast amperometric response to L-cysteine and the response time is only 3 s. The oxidization steady-state current (I s ) is linearly dependent on L-cysteine concentration, where two linear sections are found: in the lower range of 0.5-250.0 m as well as higher range of 250.0 m to 3.50 mm ( Figure 7C). In this study, the L-cysteine sensing performance of a UiO-66 /EG/GCE was evaluated with (Table 2 [55][56][57][58][59][60] ) with that of those published in the literature, specifically those sensors based on MOFs. The UiO-66 /EG/GCE depicts wider linear ranges and lower LODs for L-cysteine. Based on S/N = 3, the calculated LOD is 0.28 m.
The selectivity of a UiO-66/EG/GCE toward the detection of L-cysteine was also investigated against normally co-existed interfering species with L-cysteine such as uric acid (UA), glucose, Cysteine (Cys), Glycine (Gly), and L-lysine (L-lys). In the recorded i-t curves ( Figure 7D), insignificant current responses are detected for the addition of interferants with same concentration and an obvious current response can be detected only after the addition of 100 m L-Cys. These results suggest good selectivity of a UiO-66/EG/GCE electrode toward the monitoring of Lcysteine. Five different glassy carbon electrodes were modified with the UiO-66/EG composite; the amperometric response towards the oxidation of 50 m L-cysteine was recorded. The oxidation current values were similar with a relative standard deviation of 2.41%, revealing that the fabricated L-cysteine sensor is very stable and reproducible. The electrode displayed 91.7% recovery  with an RSD of 3.26% for the detection of L-cysteine after 7 days, thereby confirming the good stability of the sensor. It is the objective of preparing an electrochemical sensor that it can be used in real life application. The designed sensor was then tested to determine nitrite and L-cysteine in commercially available samples. The amperometric response of the total analyte concentration (the sum of initial and added concentration) was recorded on UiO-66/EG/GCE the designed sensor. A commercially available sample of pork skin aspic was used as a source of nitrite, where the following steps were taken: the pork skin aspic was weighed and immersed in 100 mL of water for 1 h. This mixture was then filtered to obtain the immersion liquid. A further centrifugation was required to obtain the supernatant/sample from this immersion liquid. Additionally, the concentration of Lcysteine in peach juice (directly available in the market) was also determined. We extracted and filtered the juice using Whatman No.40 filter paper, and then centrifuged the filtrate at 10 000 rpm for 10 min. For the analysis, supernatant liquid was used. From the obtained results (Table 3), one can conclude that the UiO-66 /EG/GCE shows good analytical performance since the spike recovery values are between 98.0% and 100.0% with satisfactory RSD values ranging from 6.1% to 8.7%. Furthermore, reasonable recovery rates ranging from 102% to 108% are achieved for L-cysteine analysis in peach juice, and the RSD values are between 2.2% and 5.8%. Based on these results, it is evident that the present UiO-66 EG/GCE is suitable for monitoring nitrite and L-cysteine in real food samples.

Conclusions
The present work demonstrates the fabrication of a UiO-66 /EG composite as well as its sensitive and selective determination of  In summary, such excellent analytical performance of the UiO-66/EG composites make them potential to be integrated into next-generation electrochemical sensing devices.
Preparation of UiO-66/EG Composites: From natural graphite flake as an initial material, expanded graphite (EG) was prepared through chemical oxidation and microwave heating. [61] The concentrated nitric acid/acetic anhydride and chlorate were acted as the oxidizing and inserting reagents, respectively. Under ambient conditions, EG was obtained by transferring the product into a microwave oven at 700 W for 30 s.
In this study, a modified version of the reported method [62] for the preparation of UiO-66 composites is followed. In the first step, 29.12 mg ZrCl 4 and 20.75 mg BDC were dissolved in 15 mL DMF. Then, 0.225 mL concentrated HCl and 0.156 mL CH 3 COOH were added slowly into the mixed solution with stirring for 5 min. The UiO-66 nanoparticles were synthesized.
Subsequently, 50 mg EG was added to the mixture solution with ultrasonic treatment for 10 min. After that, the mixture was transferred into a 25 mL Teflon lined stainless-steel autoclave for a homogeneous reaction, which was maintained at 120°C for 24 h. The UiO-66/EG composite was collected by vacuum filtration, repeatedly washed with ethanol for three times and deionized water for at least five times. Finally, the solid product was placed in a vacuum oven at 80°C for 24 h.
Materials Characterization: The surface morphologies of the UiO-66 and UiO-66/EG composites were investigated using field-emission SEM (Germini SEM 300 microscope, Zeiss, German). The TEM images were recorded on a JEM-2100 microscope (JEOL Ltd., Tokyo, Japan). The composition and crystal structure of the samples were checked by XRD (Bruker D8 Advanced with DaVinci design instrument. XPS measurements were performed on a ESCALAB Xl+ (Thermo Fisher Scientific). Nitrogen (N 2 ) adsorption-desorption measurements of the MOF samples were carried out by using a Micromeritics ASAP 2020 at 77 K.
Electrochemical Measurement: Electrochemical experiments were performed on a CHI 660D electrochemical workstation (Shanghai CH Instruments, China) with a conventional three-electrode system, consisting of a glass carbon electrode (GCE, 3 mm in diameter) as the working electrode, a saturated calomel electrode (SCE) as the reference electrode and a platinum wire as the counter electrode. Prior to electrochemical tests, the GCE was properly polished with alumina powders, sonicated with ethanol/water (v/v = 1/1) for 15 min, and dried by nitrogen. The asprepared MOF materials were dispersed in DMF to form a UiO-66/EG suspension (2.0 mg mL −1 ). Then, 5 L of the sonicated UiO-66/EG suspension was cast onto the GCE surface (UiO-66/EG/GCE) and dried at room temperature. For comparison, UiO-66/GCE and EG/GCE were prepared via the same procedure. The electrochemical properties of these modified electrodes were evaluated using cyclic voltammetry and EIS in a 5.