Engineering of layered iron vanadate nanostructure for electrocatalysis: Simultaneous detection of methotrexate and folinic acid in blood serum

,


Introduction
Methotrexate (MTX) is an antifolate therapeutic drug widely used to treat cancer and auto-immune diseases [1,2].However, MTX is cytotoxic, causing multiple organ failures and acute nephrotoxicity.It is necessary to monitor the concentration of MTX in human blood serum for timely control of dosage and guided implementation of folinic acid (FA) rescue to fit the therapeutic demand of an individual patient and prevent toxicity-related complications [3].High-performance liquid chromatography (HPLC), fluorescence polarization, chemiluminescent microparticles immunoassay, and homogenous enzyme immunoassays are routinely used to determine MTX [4][5][6].However, these assays require complex sample treatment, expensive equipment, and a long turn-around time, which limits the timely adjustment of MTX dosage and administration of FA.Therefore, a rapid, inexpensive, miniaturized analytical device for real-time monitoring of MTX is an unmet need in clinical chemotherapy.
Electrochemical voltammetric sensors displayed fascinating features for developing highly sensitive and portable bioanalytical sensing systems amenable to point-of-care measurement.The voltammetry transduction principle is based on the direct redox reaction of the biochemical analytes at the interface of a solid electrode.The slow electron transfer kinetics and high overpotential required to initiate this electrochemical process at the bare electrode limit the accuracy and sensitivity of detection.The overpotential leads to poor selectivity and causes unwanted interferences from co-existing analytes [7].Modifying transducer interfaces with nanostructured electrocatalytic materials has great potential to overcome these challenges.Electrochemical transducer interfaces modified with nanostructured ternary transition metal vanadate (M x V y O z , M = Cu, Ni, Co, Fe) [8][9][10][11][12] are of interest in various electrocatalytic applications dedicated to biochemical sensing due to their tunable electrocatalytic properties, electronic bandgap, and natural abundance [12].Among the various M x V y O z , iron vanadate (Fe x- V y O z ) has excellent electrocatalytic activity due to the synergistic strong electron-electron correlation of the multi-valent oxidation states of Fe ions (Fe 2− to Fe +6 ) and vanadium ions (V 2+ to V 5+ ) [13].Compared to amorphous FeVO 4 nanoparticles [10][11][12]14], kazakhstanite (Fe 5 V 15 O 39 (OH) 9 ⋅9H 2 O) is a hydrated layered crystalline compound with high ionic conductivity and a large surface area [15][16][17].However, the intrinsic poor electrical conductivity of Fe x V 3x O y .zH 2 O limit its electrocatalytic sensing applications.
In this study, we present the first report on engineering Fe x V 3x O y structural morphology, crystalline facet, and oxygen vacancy via calcination at different temperatures to yield oxygen-deficient dehydrated Fe x V 3x O y nanostructures.Fe x V 3x O y modified carbon screen-printed electrode was fabricated and utilized for the voltammetric determination of MTX and FA in human blood serum.The influence of changes in the physicochemical properties and electrocatalytic performance of Fe x V 3x O y towards MTX and FA was investigated.The prepared Fe x V 3x O y catalysts were characterized using various microscopic and spectroscopic techniques.The novelty of this study includes: (i) providing information on the influence of calcination temperature on the morphology, crystallinity, surface-active redox species, oxygen vacancies, and electrochemical properties of dehydrated Fe x V 3x O y , (ii) developing an ultrasensitive electrochemical sensor based on Fe x V 3x O ymodified electrodes for the simultaneous detection of MTX and FA in human clinical serum sample reported for the first time.

Experimental
All reagents and solvents were of analytical grade.The detail about the reagent supplies and instrumental analysis procedures are reported in the supporting information (SI).

Preparation of Fe x V 3x O y nanostructured electrocatalysts
The synthesis of hydrated layered Fe x V 3x O y .zH 2 O nanosheet was achieved by co-precipitation from an aqueous mixture of 90 ml of Fe (NO 3 ) 3 ⋅9H 2 O (1.2 g, 2.85 mmol) and 10 ml of NH 4 VO 3 (1.0 g, 8.55 mmol) heated at 100 • C for 24 hrs (Fig. 1a).The color of the solution changed from yellow to a brown precipitate.The brown Fe x V 3x O y .z H 2 O precipitate was separated by centrifugation, washed with deionized water and ethanol, and dried at 60 • C in a vacuum oven.The synthesized Fe x V 3x O y .z H 2 O (0.2 g) was calcined at different temperatures (T = 300, 450, 600, and 800 • C) at a heating rate of 5 • C/min for 4 hrs in a tube furnace under a flowing stream of nitrogen gas flow.The powered product obtained after calcination was generally referred to as Fe x V 3x O y -T and designated as Fe x V 3x O y -300, Fe x V 3x O y -450, Fe x V 3x O y -600, and Fe x V 3x O y -800 corresponding to calcination at (T) = 300, 450, 600 and 800 • C, respectively.

Fabrication of SPC-Fe x V 3x O y -t electrodes
The Fe x V 3x O y -T (10.0 mg) were dispersed in deionized water (5.0 mL) and ultrasonicated for 60 mins.Screen-printed carbon electrodes (SPCEs) were used as substrate and subjected to electrochemical cleaning by cyclic voltammetry treatment between − 0.8 and 1.1 V vs. Ag-SPE in 0.5 M sulphuric acid solution for 10 cycles.The SPCEs were rinsed with H 2 O and dried with a flowing nitrogen gas (N 2 ) stream.The Fe x V 3x O y -T inks (10.0 µL, 2.0 mg.mL − 1 ) were then dropped onto the circular working electrode area of the cleaned SPCEs and allowed to dry at room temperature.The dried Fe x V 3x O y -T inks form a stable thin film on the carbon electrode without adding a binder.The individually modified SPC/Fe x V 3x O y -T electrodes were rinsed with Milli-Q water and stored at room temperature.

Composition and structure of Fe x V 3x O y -T
The preparation process of Fe x V 3x O y -T with various nanostructures is illustrated in Fig. 1.The calcination of prepared hydrated Fe x V 3x O y .zH 2 O in an inert atmosphere could induce recrystallization, defects, and facet changes, influencing the physicochemical and catalytic properties [18].The calcined-powdered products showed different physical color appearances depending on the calcination temperatures (Fig. 1b).The brown Fe x V 3x O y .zH 2 O powder changed to varying shades of light green at calcination temperatures of 300 • C (Fe x V 3x O y -300) and 450 • C (Fe x V 3x O y -450), yellowish green at 600 • C (Fe x V 3x O y -650), and blue-black at 800 • C (Fe x V 3x O y -800).The color changes in the calcined samples are attributed to changes in the absorption band edges due to the temperature-dependent reduction of vanadium and iron to their different metastable oxidation states.The UV-Vis diffuse reflectance spectra (UV-Vis DRS) of Fe x V 3x O y .zH 2 O (Fig. 2a-i) shows reflection bands in the ultraviolet (250-330 nm), violet-blue (330-399), and orange-red (580-750 nm) regions of the visible spectra.These reflections are characteristic of the D-d transition of various V-O (V 4+ , V 5+ ) and F-O (Fe 2+ , and Fe 3+ ) species, and the mixture resulted in the brown color of Fe x V 3x O y .zH 2 O.The decrease in reflectance observed between 400 and 550 nm corresponds to the onset of the adsorption due to electronic transition across the band gap.The Fe x V 3x O y .zH 2 O has an indirect band gap of 2.2 eV (Fig S1, i).The reflectance intensity in the orange-red region decreased significantly after calcination (Fig. 2a, i-iv).Fe x V 3x O y -300, Fe x V 3x O y -450, and Fe x V 3x O y -600 showed reflection in the violet-blue and green-yellow regions due to Fe 2+ , V 3+ , and V 4+ and V 4+ ions.The combinations of reflectance from different mixes of these ions resulted in different shades of greenish color.Initially, the bandgap increased from 2.2 eV for Fe x V 3x O y .zH 2 O to 3.18 eV for Fe x V 3x O y -300 and 3.35 eV for Fe x V 3x O y -450.However, with a further increase in calcination temperature, the bandgap decreased to 2.8 eV for Fe x V 3x O y -600 and 2.3 eV for Fe x V 3x O y -800.Calcination causes the loss of hydrated water in the crystalline structure, which could result in the modification of the electronic structure and the number of the electronic state, leading to an increase in the band gap energy [19].

Elemental mapping
The EDS spectra of Fe x V 3x O y .zH 2 O before (Fig. S2, a) and after calcination (Fig. S2, b-e) confirmed the presence of Fe, V, and O. EDS elemental mapping (Fig. S3) also confirmed the homogeneous distribution of Fe, V, and O in all samples.The EDS atomic composition (inset of Fig. S3, a-e) and elemental analysis by ICP-OES (Table S1) for samples calcined below 800 • C revealed an average vanadium-to-iron atomic ratio (V/Fe) of -3.3:1, which is close to that of the ferric nitrate and ammonium vanadate precursors (3:1).The V/Fe ratio of the same calcined at 800 • C was however lowered to -2:1, indicating the volatilization of vanadium at a higher temperature.

Fe-O and V-O bonding
The FTIR spectra of Fe x V 3x O y .zH 2 O before calcination (Fig. S4, i) revealed the existence of adsorbed water molecules (O-H νs at 3575 cm − 1 ; (H 2 O) δ at 1633 cm − 1 ) and structural water of hydration (O -H νs at 3215 cm − 1 ; (H 2 O) δ at 1608 cm − 1 ).However, calcination removed adsorbed water from all samples at >100 • C (Fig. S4, ii-v), confirming the formation of dehydrated Fe x V 3x O y .The metal oxide vibrations (Fig. 2b, ii-v) confirmed the Fe-O stretching (489-424 cm − 1 ), V=O stretching, and Fe-O-V bending vibrations (1008 cm − 1 ) [20].The bands between 732 and 1031 cm − 1 were attributed to the vanadyl bonds (V-O-V, V-O) stretching and bending vibrations [21].However, the intensity of the Fe-O and O-V-O bonds decreased and shifted to lower wavenumbers as the calcination temperature increased.The lower intensity and the blue shift could be associated with shorter Fe-O and V-O bond lengths.The Fe-O and V-O interactions were resolved further by Raman spectroscopy (Fig. 1c).Peaks at the lowest vibrational frequencies (200-300 cm − 1 ) are attributed to Fe-O, O-V-O, and V-O-Fe mixed bending and torsion modes vibrations.The broad V-O…Fe and V…O…Fe bridging (420-550 cm − 1 ) [22] and terminal V-O stretching (848-930 cm − 1 ) vibrations [13] showed that oxygen atoms were mainly located at bridging and terminal positions.The main difference in the Raman spectra is the redshift and splitting of the broad band between 350 and 550 cm − 1 into well-resolved and intense spectra lines with increasing calcination temperature (Fig. 2c,i-v).The peak at 704 cm − 1 (Fe-O stretching) in all the samples before and after calcination aligns with binding with Fe 3+ ions, and the emergence of a new band at 534 cm − 1 (Fe-O stretching) after calcination at high temperatures (≥ 450 • C) confirms the reduction to Fe 2+ during the calcination process [23].Also, the blue shift and lower intensity of the V 5+ =O bonds stretching vibrations (128 cm − 1 ) with increasing calcination temperatures indicated higher oxygen deficiency (Fig. 2c, i-v).
3.1.4.Particle morphology FE-SEM (Fig. 3) and TEM images (Fig. 4) revealed differences in morphology and microstructure after calcination.The Fe x V 3x O y before calcination showed layered flake-like morphology with wrinkles and rough edges.After calcination at 300 • C, the Fe x V 3x O y -300 exhibited crumpled nanoflake-like morphology with smaller particle sizes than Fe x V 3x O y .zH 2 O (Fig. 3b).The nanoflake-like Fe x V 3x O y self-assembled into the agglomerated oval microflower structure after calcination at 450 • C (Fe x V 3x O y -450, Fig. 3c) to reduce the surface energy.The Fe x V 3x O y -600 (Fig. 3d) showed a mixture of rectangular multilayered sheets and some irregular truncated polyhedron particles, indicating Ostwald ripening [24,25] and recrystallization of Fe x V 3x O y occurred at 450 ≤ T ≤ 600 • C. The formation of truncated irregular octahedron nanoparticles with well-defined edges and smooth surfaces was observed by FE-SEM of Fe x V 3x O y -800 (Fig. 3e).TEM images (Fig. 4a-e, II) also confirmed morphological changes through the formation of two-dimensional micrometer-wide and nanometer-thick Fe x V 3x O y -300 nanoflakes, Fe x V 3x O y -450 nanoflower, and Fe x V 3x O y -600 rectangular nanosheet.HRTEM images (Fig. 4a-e, ii) and the inverse fast Fourier transform (IFFT) atomic lattice images (inset Fig. 4a-e, III) revealed well-defined lattice fringes implying a crystalline structure as observed with the XRD result.A lattice D-spacing of -0.92 nm, consistent with the interlayer spacing of the (002) crystalline plane, was predominant in the Fe x V 3x O y .zH 2 O before calcination.Two intersecting planes with a D-spacing of -0.35 nm and 0.28 nm were observed in the Fe x V 3x O y -300 and Fe x V 3x O y -450, implying that growth, respectively proceeded along the (204) and (115) crystal planes.The changes in the lattice distance to 0.56 nm (001 crystal plane) after calcination at 600 • C and 0.43 nm (001 crystal plane) confirmed recrystallization and formation of FeV 3 O 8 and after calcination at 800 • C of Fe 2 VO 4 .Also, disorder and edge dislocation defects were observed in the lattice fridges (red cycled regions in Fig. 4b-e, (IV), confirming the presence of rich crystalline oxygen vacancies in the calcined Fe x V 3x O y .

Surface area and microporosity
The Nitrogen (N 2 ) adsorption-desorption isotherms (Fig. S6) showed a type IV isotherm with a sharp knee at a relative pressure between 0.85 and 1.0.The isotherms exhibited an H3-type hysteresis loop, and no plateau was observed at the highest relative pressures, which correlates to the presence of slit-shaped mesopores according to the IUPAC classification [26].The surface area was calculated using the Brunauer-Emmett-Teller method (BET) [27] by an in-built software from Micromeritics.Detailed information is available in the supplementary (SI).The BET surface area (SA BET ) increased from 38.27 m 2 g − 1 for Fe x V 3x O y .zH 2 O to 53.74 m 2 g − 1 and 48.68 m 2 g − 1 when calcined at 300 • C (Fe x V 3x O y -300) and 450 • C (Fe x V 3x O y -450), respectively.Removing impurities and adsorbed water molecules could be responsible for the increased surface area.The significant decrease in SA BET to 5.47 m 2 g − 1 for Fe x V 3x O y -600 and 4.14 m 2 g − 1 for Fe x V 3x O y -800 (Table S2) resulted from the destruction of pore structures and increased crystallinity.The steep N 2 absorption at the beginning (P/P o < 0.1) in the isotherm of Fe x V 3x O y -600 and Fe x V 3x O y -800 could be ascribed to the presence of micropores.A t-plot analysis [28] (Fig. 5) showed that Fe x V 3x O y .zH 2 O has a microporosity of -2.3 µL.g − 1 .Calcination at 300 • C caused a slight increase to -3.7 µL.g − 1 , and this microporosity collapsed to -1.0 µL.g − 1 at higher calcination temperatures (600-800 • C).The microchannels expelling water vapor during thermal dehydration and dehydroxylation explain the slight increase in microporosity at Fe x V 3x O y -300 • C and Fe x V 3x O y -450 • C, which aligns with the loss of structural crystalline water.At higher temperatures, recrystallization collapsed this microporosity.The Fe x V 3x O y .zH 2 O, Fe x V 3x O y -300, and Fe x V 3x O y -450 had a micropore volume of 0.17, 0.31 and 0.35 cm 3 .g− 1 , respectively.The Fe x V 3x O y -600 and Fe x V 3x O y -800 had a ten-time lower pore volume of 0.015 and 0.020 cm 3 .g− 1 , respectively.The average pore diameter increased from 17.7 nm for Fe x V 3x O y .zH 2 O to 26.04 nm for Fe x V 3x O y -300.Then, a progressive decrease in the pore size diameter to 10.8 and 8.4 nm was observed with increasing calcination temperature from 600 to 800 • C. The decrease in the pore size diameter is attributed to an increase in crystallinity and changes in the structural morphology at higher calcination temperatures (≥ 450 • C), as seen in the XRD and HRTEM analysis.

Surface composition and (Fe, V) oxidation states
XPS analysis revealed the Fe x V 3x O y .zH 2 O surface atom composition, elemental binding sites, and changes in oxidation states before and after calcination at different temperatures.The survey scan in Fig. S7 confirmed the Fe, V, O, C, and N atoms in all the samples.The highresolution Fe2p core-level spectra (Fig. 6a) showed a broad split spinorbit doublet with characteristic satellite peaks attributed to the photoemission of Fe 2p 3/2 and Fe 2p 1/2 species.The spin-orbit constant (ΔBE) between the Fe 2p 3/2 and Fe 2p 1/2 (13.5 eV) was close to the standard value for iron oxide (13.0 eV) species [29].The broad Fe 2p 3/2 spectra of Fe x V 3x O y before and after calcination at different temperatures were deconvoluted into three components with repeated patterns of half intensity for the Fe 2p 1/2 spectra.The BE at 710.7 eV confirms the existence of Fe 2+ ions, while the peak at 712.5 and 715.3 eV were due to the octahedral and tetrahedral coordinated Fe 3+ ions.The BE of Fe 2+ and Fe 3+ ions shifted towards higher values with increasing calcination temperatures from 300 to 800 • C. The higher BE shifts imply an increased shielding of the Fe cation by electronegative (V-O) ligands [30], implying an increase in the degree of the ionic bond character of the Fe-V-O bond.The high-resolution V 2p spectrum of Fe x V 3x O y before calcination showed a single spin-orbit doublet at 517.5 eV (V 5+ 2p 3/2 ) and 525.1 eV (V 5+ 2p 1/2 ) of V + 5 (Fig. 6c, i).Before calcination, the V atoms predominately exist in the (+5)-oxidation state in the Fe x V 3x O y .zH 2 O. Notably, after calcination, the V2p spectrum of FeV 3 O y -300 and FeV 3 O y -450 showed an additional pair of doublets attributed to the V 4+ ions at 516.4 eV (V 4+ 2p 3/2 ) and 523.6 eV (V 4+ 2p 1/2 ) (Fig. 6c, ii and iii).The Fe x V 3x O y -600 and Fe x V 3x O y -800 showed a third component attributed to V 3+ ions at 515.4 eV (V 3+ 2p 3/2 ) and 523.4 eV (V 3+ 2p 1/2 ).The lower oxidation state indicated that more electrons were injected into the 3d-orbital of vanadium during the oxygen-deficient calcination process.The O1s spectra of Fe x V 3x O y .zH 2 O and Fe x V 3x O y -300 were deconvoluted into three components, while the Fe x V 3x O y -450, A slight shift of O 1s towards lower BE with increasing temperature indicates a new oxygen chemical environment.The value of (V 4+ + V 3+ )/V 5+ and oxygen vacancy (O ads /(O lat + O ads )) influences the catalytic activity [32].Where oxygen (O ads ) is the adsorbed oxygen species.
The relative atomic percentage ratio of (O ads / (O lat + O ads )) indicated the concentration of the oxygen vacancies.The Fe x V 3x O y .zH 2 O presented a relative oxygen vacancy (O ads / (O lat + O ads )) concentration of 9.7%, which increased to 11.7% for Fe x V 3x O y -300 and Fe x V 3x O y -450, 8.7% for Fe x V 3x O y -600 and 20.3% for Fe x V 3x O y -800.The higher concentrations of (O ads / (O lat + O ads )) confirmed the formation of oxygen vacancies with increasing calcination temperature.The relative atomic ratio of V/Fe at the surface of Fe x V 3x O y .zH 2 O (2.2: 1) was lower than the bulk atomic V/Fe ratio (3.2:1) from ICP-OES analysis.The lower ratio showed that Fe relative to V atoms is abundant on the surface of the FeV 3 O y after calcination .The increase in V/Fe ratios after calcination is attributed to the enrichment of reduced V-O at the surface of the FeV 3 O y .The value of (V 4+ + V 3+ )/V 5+ ratios increased with increasing calcination temperature from 0.18 for Fe x V 3x O y -300 to 0.31 for Fe x V 3x O y -800.This increase indicated that more V 5+ were reduced to V 4+ and V 3+ with increasing calcination temperature under the nitrogen atmosphere.

Electrochemical properties of SPC/Fe x V 3x O y -T electrodes
The SPC/Fe x V 3x O y .zH 2 O and SPC/Fe x V 3x O y -T electrodes were characterized using cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) in aqueous [Fe(CN) 6 ] 3-/4− redox electrolyte containing 0.1 M KCl (Fig. S8).The cyclic voltammograms in Fig. S8a showed good reversibility for the redox process of [Fe(CN) 6 ] 3-/4-on the SPC/Fe x V 3x O y -T.The redox peak current density was higher at the modified SPC/Fe x V 3x O y -T electrodes than at the unmodified SPC electrode (Fig. S8a, ii-iv).The intensity of the redox peak current is proportional to the electroactive surface area (E SA ) of the different SPC/ Fe x V 3x O y -T electrodes and was calculated from the Randle-Sevick relationship (Eq.( 1)) [33].
The SPC/Fe x V 3x O y -300 and SPC /Fe x V 3x O y -450 had higher electroactive surface area than the SPC/Fe x V 3x O y .zH 2 O, SPC/Fe x V 3x O y -650, and SPC/Fe x V 3x O y -800.The SPC/Fe x V 3x O y -450 and SPC/Fe x V 3x O y -300 showed excellent electrochemical properties due to the synergistic effect of their two-dimensional morphology, high electroactive surface area, high pore volume, and pore diameter that endowed them with faster ion diffusion and rapid charge transfer properties.
Electrochemical impedance spectroscopy was used to evaluate the electrochemical reaction kinetics and charge transfer properties of the various SPC/Fe x V 3x O y -T electrode.Fig. S8b, ii-iv compares the Nyquist plots of the bare SPC, SPC/Fe x V 3x O y .zH 2 O, and various SPC/Fe x V 3x O y -T.The fitted EIS data were summarized in Table S4.The semicircle diameter of the Nyquist plot corresponding to the interface charge transfer resistance (R CT ) of the SPC/Fe x V 3x O y -T and uncalcined Fe x V 3- x O y .zH 2 O were higher than the SPC electrode, indicating the semiconducting properties of the Fe x V 3x O y -T.Compared to the uncalcined Fe x V 3x O y .zH 2 O, the R CT decreased significantly after calcination at 300 to 450 • C (Table S4) and increased at higher temperatures (≥ 600 • C).The lowest R CT of the SPC/Fe x V 3x O y -300 and SPC/Fe x V 3x O y -450 confirms the higher efficient charge transfer process.

Electrochemical behavior of MTX and FA at SPC/Fe x V 3x O y -T electrodes
The redox behavior of MTX and FA in phosphate buffer solution (PBS) at SPC/Fe x V 3x O y -T electrodes was investigated using cyclic voltammetry (Fig. 7).The CV of the bare SPC, SPC/Fe x V 3x O y .zH 2 O, and SPC/Fe x V 3x O y -T electrodes in the PBS alone did not show redox peaks within the studied potential window.An unresolved broad anodic peak between +0.6 and +1.1 V was observed for both FA and MTX at the bare SPC electrode (Fig. 7a), indicating the slow kinetics of MTX and FA electrooxidation.However, the CVs of SPC/Fe x V 3x O y .zH 2 O and all the SPC/Fe x V 3x O y -T electrodes showed a well-resolved anodic peak at -0.75 V for FA and -0.90 V for MTX (Fig. 7, b-f).No cathodic peak was observed, confirming the irreversibility of FA and MTX oxidation at the bare SPC and SPC/Fe x V 3x O y -T modified electrodes.The well-resolved redox peaks and higher current response confirmed the electrocatalytic activity of Fe x V 3x O y -T towards the oxidation of MTX and FA.The anodic peak currents density for FA and MTX electrooxidation at all calcined Fe x V 3x O y -T modified SPC electrodes was significantly higher than the Fe x V 3x O y .zH 2 O modified SPC electrodes (Fig. S9).It could be inferred that the formation of low valency state (Fe 2+ , V +3 , and V 4+ ) species and oxygen vacancies defects during calcination facilitated MTX and FA electrooxidation.The defect sites could facilitate the adsorption of MTX and FA onto the surface of the dehydrated Fe x V 3x O y -T.The MTX and FA anodic current density varied with the Fe x V 3x O y -T calcination temperatures.
The MTX and FA anodic current density varied with the Fe x V 3x O y -T calcination temperatures.The Fe x V 3x O y -300 and Fe x V 3x O y -450 had the highest peak current density for FA and MTX (Fig. S9).This is attributed to the higher electroactive surface area, mesoporosity, and higher electron transfer kinetics of Fe x V 3x O y -450.Based on these findings, Fe x V 3x O y -450 was chosen for further electrochemical studies.The effects of pH on the electrocatalytic oxidation of FA (pK 1 = 3.1, carboxylic acid; pK 2 = 4.8, pK 3 = 10.4,amines) and MTX (pK 1 = 4.7, carboxylic acid; pK 2 = 8.8, pK 3 = 9.2, amine) at the SPC/Fe x V 3x O y -450 was investigated.The anodic peak potential of MTX shifted towards a more negative value with increasing pH from 3.0 to 5.0 as MTX species underwent deprotonation (Fig. S10, a).In contrast, at pH ≥ 6.0, the oxidation of MTX was not dependent on the pH as the MTX became fully deprotonated.The electrooxidation of MTX at the SPC/Fe x V 3x O y -450 electrode involves the transfer of electrons and a protons abstraction process [34,35].The plot of the anodic potential against the pH of MTX is linear in the ranges of 3.0 < pH < 6.0 with a regression equation E pa (V) = 1.04 -0.028 pH (R 2 =0.985),Fig. S10 (a').The obtained slope of 28 mV per pH unit was about half of the theoretical Nernst value (59 mV/pH), indicating that the ratio of electron transfers to proton  x O y -450 electrode was 0.5.The CVs of FA at the SPC/Fe x V 3x O y -450 showed two anodic peaks (I a and I a' ) in the pH range 3.0 < pH ≤ 6.0 (Fig. S10, b).The second anodic peak (I a' ) was not observed in the CVs of FA at neutral and basic pH (≥ 7.0).The first anodic peak potential (I a = 0.72 V) at pH 3.0 shifted linearly to a more negative potential (I a = 0.65 V) at pH 6.0, while there was a shift to a more positive value at pH ≥ 7.0.The linear regression equation of potential against pH was E pIa = 0.79 -0.023 pH at pH ≤ 6.0 and E pIa = − 0.70 -0.012 pH at pH ≥ 7.0 (Fig. S10,  b').The ratio of proton to electron transfer involved in the oxidation of FA at the SPC/Fe x V 3x O y -450 is 0.5 at pH ≤ 6.0.The mechanism of electrooxidation of FA at acidic and alkali pH has previously been documented in the literature [36,37].It could be seen that the second anodic peak (I a' ) of FA overlaps with the anodic peak (I a ) of MTX in the pH ranges of 3.0 ≤ pH ≤ 5.0.The overlap of the (I a' ) peak of FA with the anodic peak of MTX would complicate the accurate determination of MTX in the presence of FA at pH ≤ 5.0.Therefore, pH 7.0, where the second anodic peak (I a' ) of FA is absent, was selected to quantitatively determine MTX and FA at the SPC/Fe x V 3x O y -450 electrodes.
The electrooxidation of MTX and FA at the SPC/Fe x V 3x O y -450 electrode in PBS (pH 7.0) was investigated by changing the CV scan rate from 5 to 150 mV/s.The anodic current density of MTX and FA increases, and the potential shifts to more positive values with an increasing scan rate from 10 to 150 mVs − 1 (Fig. S11).The anodic peak current of FA and MTX were linearly proportional to the scan rate in the studied ranges (Fig. S11, a' and b').The linearity indicated that the oxidation of FA and MTX at the SPC/Fe x V 3x O y -450 is an adsorptioncontrolled process, demonstrating that MTX and FA are adsorbed onto the active sites of Fe x V 3x O y -450 and are electrooxidized.The differential pulse voltammetry (DPV) of SPC/Fe x V 3x O y -450 for repeated measurement of MTX (Fig. S12a) showed a decrease in the current response with an increasing number of DPV repetitive scans.The observed current density decrease is attributed to the adsorption of oxidized MTX product onto the active sites of the SPC/Fe x V 3x O y -450 electrode.This is a characteristic behavior of the adsorption-controlled oxidation mechanism [38].Fig. S12 (b) showed that a simple rinsing step with PBS solution between each successive run of MTX could remove the adsorbed byproduct and restore the baseline signal (Fig. S12b) to achieve repeated use.There were no differences in the current response to a repetitive measurement of 50.0 µM of MTX even after the 10th measurement, confirming good repeatability and reusability.The reaction time of SPC/Fe x V 3x O y -450 with MTX was optimized using 50 µM of MTX.The current density increases with increasing reaction time up to 5 min, after which there was no significant increase in the peak current response.Hence, 5 min without applying a preconcentration potential was used for the measurements.

Voltammetric detection of MTX and FA using SPC/Fe x V 3x O y -450
Differential pulse voltammetry (DPV) was used to determine the level of MTX and FA in PBS and human blood serum samples under the optimized conditions.The anodic current density increased with increasing concentrations of FA (Fig. 8a) and MTX (Fig. 8b).The analytical calibration curve was established from the plot of change in the current density (ΔI = I a -I o ) before (I o ) and after MTX or FA interaction (I a ).The linear concentration range was from 0.05 to 200 µM with a regression equation of ΔI = 0.473[FA] (µM) + 16.84 (R 2 = 0.982) for FA (Fig. 8c) and from 0.005 to 200 µM with a regression equation ΔI = 1.249 [MTX] (µM) + 14.42 (R 2 = 0.995) for MTX (Fig. 8d).The detection limit (LoD) and quantification (LoQ) were estimated from a multiplier of 3σ/sensitivity and 10σ/sensitivity, respectively.There was no appreciable current increase without MTX and FA, and the background signal's standard deviation (σ) was 4.04 (n = 6).The LoD and LoQ were 2.85 nM and 9.83 nM, respectively, for MTX and 7.79 and 25.9 nM, respectively, for FA.The comparison of the linear range and LoD of SPC/Fe x V 3x O y -450 with previously reported MTX and FA sensing systems are summarized in Table 1.The SPC/Fe x V 3x O y -450 showed the limit of detections for FA and MTX in the low nM range, comparable to the other reported sensing systems.However, the SPC/Fe x V 3x O y -450 had the advantage of detecting MTX in a wider linear range (0.005 to 200 µM) compared to previously reported voltammetric sensing systems.Although the linear range of our previously reported AuE/anti-MTX Ab (0.0002 -270) in ref. [4] and GCE-IPA/anti-MTX Ab (0.0003 -300) in ref. [5] are considerably wider than SPC/Fe x V 3x O y -450 (0.005 to 200 µM), it should be noted that these methods are based on very expensive and easily denatured antibodies, and simultaneous determination of MTX and FA has not been possible.The linear range and the LoD of the SPC/Fe x V 3x O y -450 are within the clinically targeted MTX concentrations of 5.0 to 10.0 µM at 24 h, 0.10 µM at 72 h, and 0.05 µM at 96 h after intravenous MTX administration [39], which makes it possible to measure MTX and FA in blood serum without dilution.

Simultaneous determination of MTX in the presence of FA
To test the applicability of the SPC/Fe x V 3x O y -450 electrode for monitoring MTX and FA in human clinical samples, human serum was spiked with concentrations of MTX (0.05 -5.0 µM) and FA (0.5 -40 µM) within the clinically relevant concentration ranges.The SPC/Fe x V 3x O y -450 electrode did not show an appreciable current response in the serum samples, indicating endogenous electroactive serum constituents are not oxidized within the studied potential window.The DPV of the mixed FA and MTX spiked blood serum in Fig. 9(a) showed two well-resolved anodic peaks at 550 and 798 mV corresponding to the irreversible electrooxidation of FA and MTX, respectively.The peak separation between the FA and MTX signal was 248 mV, demonstrating that the SPC/ Fe x V 3x O y -450 electrode can detect MTX and FA in a mixed solution without signal overlap.The current response increases simultaneously with increasing concentrations of both MTX and FA in the spiked blood serum samples (Fig. 9a).An appreciable current response was observed at low concentrations of MTX (≤ 0.05) in the serum samples.However, the response was recorded at higher concentrations of FA (≥0. .This confirmed that the SPC/Fe x V 3x O y -450 sensor could detect lower concentrations of MTX in the presence of a higher concentration of FA without a signal crosstalk effect.The precision of the SPC/Fe x V 3x O y -450 was evaluated using an inter-assay of human serum samples spiked with 0.05, 0.10, and 0.20 µM of MTX and 0.5, 1.0, and 2.0 µM of FA.The recovery concentrations were evaluated from the calibration plot in Fig. 9(b) and 9(c).The percentage recoveries (Table S5) were between 93.5 and 104.6% for MTX and 99.5 to 105% for FA, demonstrating that the SPC/Fe x V 3x O y -450 sensor has good precision for monitoring MTX  and FA in blood serum samples.The SPC/Fe x V 3x O y -450 electrode could selectively and accurately detect MTX in a mixed solution of both analytes without interference.The selectivity of the SPC/Fe x V 3x O y -450 for MTX and FA in the presence of other electroactive endogenous serological substances, such as dopamine (DOP), ascorbic acid (AA), glucose (Glu), tetrahydrofolic acid (HFA) and uric acid (UA).The interferent substances showed no voltammetric response within the examined potential range (0.5-1.0 V), Fig. S14.The peak current of the mix of FA, MTX, and the interfering substances did not change significantly, demonstrating the good antiinterference properties of SPC/Fe x V 3x O y -450.

Fig. 1 .
Fig. 1.Schematic illustration of the preparation of different Fe x V 3x O y -T electrocatalyst.

Fig. 8 .
Fig. 8. DPV of (a) MTX and (b) FA at the SPC/Fe x V 3x O y -450 electrode and the corresponding linear calibration plot of current density against (c) MTX concentrations and (d) FA concentrations in PBS (pH 7.0).
5), probably due to the higher sensitivity of SPC/Fe x V 3x O y -450 for MTX than FA.The linear regression equation for the simultaneous quantification was ΔIp = 1.279[MTX] (µM) + 1.55 (R 2 = 0.998), within a linear dynamic range of 0.05 to 5.0 µM of MTX and ΔIp = 0.118[FA] (µM) + 1.117 (R 2 = 0.992) for the linear range of 0.50 to 40 µM of FA.The DPV of the mixed solutions of 10.0 µM of FA with different concentrations of MTX from 0.5 to 3.0 µM (Fig S13) showed the current response due to the MTX increases with increasing MTX concentrations, while the current response due to FA did not change significantly (Fig S13)

Fig. 9 .
Fig. 9. (a) DPV of different standard concentrations of MTX and FA in blood serum and the corresponding calibration plot of change in current density against different (b) MTX and (c) FA concentrations.

Table 1
Comparison of SPC/Fe x V 3x O y -450 with other electroanalytical methods.MTX and b FA sensors.(TWthis work), EIS: Electrochemical impedance spectroscopy-multivariate data analysis; DPV: differential pulse voltammetry. a