Supported data and manuscript "Electrochemical insights into manganese–cobaltdoped a-Fe2 O 3 nanomaterial for cholesterol detection: a comparative approach"

Electrochemical insights into manganese–cobalt doped a-Fe 2 O 3 nanomaterial for cholesterol detection: a comparative approach Sushmitha S, a Subhasmita Ray, b Lavanya Rao, a Mahesha P. Nayak, a Karel Carva b and Badekai Ramachandra Bhat * a Herein, a self-assembled hierarchical structure of hematite (a-Fe 2 O 3 ) was synthesized via a one-pot hydrothermal method. Subsequently, the nanomaterial was doped to obtain M x Fe 2−x O 3 (M =Mn–Co; x =0.01, 0.05, and 0.1) at precise concentrations. An electrode was fabricated by coating the resulting nanocomposite onto a nickel foam (NF) substrate. Electrochemical characterization demonstrated the excellent performance of cobalt-doped a-Fe 2 O 3 , among which Co 0.05 Fe 0.95 O 3 (CF5) exhibited a superior performance, showing a two-fold increase in sensitivity of 1364.2 mAmM −1 cm −2 (±0.03, n=3) in 0.5 M KOH, a limit of detection (LOD) of ∼0.17 mM, and a limit of quantification (LOQ) of ∼0.58 mM. The Density Functional Theory (DFT) was performed to understand the doping prompting in the reduced bandgap. The fabricated electrode displayed a rapid response time of 2 s and demonstrated 95% stability, excellent reproducibility, and selectivity, as confirmed by tests with several interfering species. A comprehensive evaluation of the electrode's performance using human blood serum highlighted its robustness and reliability for cholesterol detection in clinical settings, making it a promising tool for clinical and pharmaceutical applications. 1. Introduction Cardiovascular diseases (CVDs) are the leading cause of mortality worldwide, with growing public awareness since the 1980s regarding the risks associated with elevated blood cholesterol levels. 1 Cholesterol, a unique waxy lipid, is essential for various bodily functions, necessitating its close monitoring. 2 Numerous studies have proven that elevated cholesterol levels noticeably increase the risk of CVDs. 3 In healthy individuals, the cholesterol levels are below 200 mg dL −1 (5.17 mM), and the levels around 240 mg dL −1 (6.21 mM) are associated with peripheral vascular diseases, insulin-dependent diabetes, hypertension, and cardiovascular conditions. 3–6 Regular monitoring of cholesterol levels has become essential in contemporary healthcare, prompting the development of various methods to assess cholesterol concentrations accurately. Consequently, numerous techniques have emerged for the detection of cholesterol in biological and food samples. These include spectrophotometric methods, 7 electrochemical approaches, 4 high-performance liquid chromatography (HPLC), 8 and enzymatic colorimetric techniques. 9 Although these techniques demonstrate signicant effectiveness in cholesterol sensing, they still possess certain limitations such as time-consuming processes, high cost, requirement of expensive equipment, and potential background noise. 10 To tackle the problems based on these techniques, non-enzymatic biosensors that display strong thermal and chemical stability and good sensitivity and selectivity have gained a greater interest among researchers. 11 To bridge the gap, integrating non-enzymatic biosensors and nanotechnology has resulted in the creation of nanocomposites that considerably enhance the electrochemical performance of cholesterol sensors. 12,13 These advancements offer several predominant benets, including improved selectivity, reduced cost, ease of handling, and rapid response times, rendering them highly effective for point-ofcare testing. 14 The functionality of biosensors hinges on the selection and development of nanostructured materials. 15 To develop electrochemically active substances for non-enzymatic sensing applications, various nanomaterials, including zerodimensional (0D) nanomaterials, 16 one-dimensional (1D) nanowires 17 and tubes, 18 two-dimensional (2D) metals/metal oxides, 19,20 graphenes, 21 and three-dimensional (3D) nano- owers, 22,23 cubes, 24 rods 25,26 and spheres, 27 have been analyzed. Although advanced materials remain the primary focus of research, there is an increasing interest in simpler nanocomposites that demonstrate comparable sensitivity and electrochemical performance. In recent years, transition metal a Department of Chemistry, Catalysis and Materials Chemistry Laboratory, National Institute of Technology Karnataka, Surathkal, D. K., Karnataka 575 025, India. E-mail: [email protected]u.in b Department of Condensed Matter Physics, Faculty of Mathematics and Physics, Charles University, Ke Karlovu 3, Prague 12116, Czech Republic Cite this: RSC Adv.,2025,15,34176 Received 19th June 2025 Accepted 4th September 2025 DOI: 10.1039/d5ra04373e rsc.li/rsc-advances 34176 |RSC Adv.,2025,15,34176–34190 © 2025 The Author(s). Published by the Royal Society of Chemistry RSC Advances PAPER Open Access Article. Published on 18 September 2025. Downloaded on 10/2/2025 5:21:14 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. View Article Online View Journal | View Issue oxides including ZnO, CuO, SnO 2 ,Fe 2 O 3 ,Ag 2 O, WO 3 , NiO 2 , and V 2 O 5 have been extensively investigated for their potential in non-enzymatic sensing applications. 6,28–30 These materials exhibit high efficacy owing to their metal centers and greater surface areas, thereby enhancing the density of electrochemically active sites in alkaline environments. 16,31–36 In the past few years, doping metal oxides has been shown to boost electrochemical performance and photocatalytic activity and improve electrical and photoelectrochemical features by boosting the charge carrier density and conductivity. 37 Currently, activated carbon derived from Piper nigrum is being synthesized and combined with a-Fe 2 O 3 , followed by modication with a carbon paste electrode (APC-Fe 2 O 3 /CPE) 38 exhibiting high sensitivity with a linear range of 25 nM to 300 nM, an LOD of 8 nM, and an LOQ of 26 nM. Additionally, electrolyte-gated transistor-based biosensors utilizing a-Fe 2 O 3 decorated with ZnO nanorods demonstrated a broad linear range of 0.1 to 60 mM and a sensitivity of 37.34 mAmM −1 cm −2 . 39 Moreover, the phase transition of Fe 3 O 4 to a-Fe 2 O 3 was achieved via electrophoretic lm deposition on ITO-coated glass plates, which exhibited a sensitivity of 193 nA mg −1 dl cm −2 with a linear range and response time of 25–500 mg dl −1 and 60 s, respectively. 40 Another study on the electrochemical behavior of citrate-modied b-cyclodextrin (CIT-BCD) and Fe 3 O 4 synthesized via the coprecipitation method (CITBCD@Fe 3 O 4 ) reported a linear range of 0 to 100 mM and an LOD of 3.93 mM. 41 A bimetallic nanocomposite glassy carbon electrode combined with a Cu 2 O/MoS 2 nanohybrid, which demonstrated a sensitivity of 111.74 mAmM −1 cm −2 with an LOD of 2.18 mM and a linear range spanning from 0.1 to 180 mM, was showcased as an effective alternative for cholesterol sensing. 42 Additionally, utilizing low-cost galvanic deposition, a ZnO/WO 3 composite demonstrated a sensitivity of 176.6 mAcm −2 mM −1 with a linear range of 0–320 mM and an LOD of 5.5 nM. 43 Another study on the electrochemical study of NiO/CuO nanocomposites synthesized by an electrospinning method reported a sensitivity of 10.27 mAmM −1 cm −2 , a linear range of 0.8 to 6.5 mM and an LOD of 5.9 mM. 44 Despite these advancements, challenges such as cost-effectiveness, lifespan, stability, and sensitivity to pH and temperature variations complicate the use of complex molecular sensors. To address these issues, our study aims to develop a transition metal-doped metal oxide cholesterol sensor to enhance selectivity, sensitivity, and stability while reducing costs. This study clearly shows the strategic comparative synthesis of bimetallic M x Fe 2−x O 3 (M =Mn–Co), and to the best of our knowledge, this is the rst comparative study investigating a nonenzymatic cholesterol sensor. The primary goal of this work was to synthesize a-Fe 2 O 3 via a hydrothermal approach and subsequently implement doping M x Fe 2−x O 3 (M =Mn–Co, and x=0.01, 0.05, and 0.1). The doping concentrations of 1%, 5%, and 10% were selected to systematically investigate the variations in electronic structure, redox activity, and conductivity induced by doping. The obtained materials were fabricated on NF, and the electrodes were evaluated for sensitivity, LOD, LOQ, response time, and linear range. The comparison of the doped materials was intended to choose the most promising dopants, which were then subjected to a comparison analysis of their properties for biosensing applications. Electrochemical characteristics were assessed by Cyclic Voltammetry (CV), chronoamperometry (CA), Differential Pulse Voltammetry (DPV), Electrochemical Impedance Spectroscopy (EIS), and the evaluation of electrochemical active surface area (ECSA). The selected material was further analyzed for its application in blood serum cholesterol detection. 2. Experimental details 2.1 Chemicals and materials Nickel foam (NF; thickness 0.5 mm, 99.9% purity) was purchased from Global Nanotech Mumbai. Iron(III) chloride tetrahydrate (FeCl 3 $4H 2 O, 99% Purity), dopamine hydrochloride (DA), poly(vinylidene uoride) (PVDF), Triton™X-100, and urea (CO(NH 2 ) 2 , 98% purity) were procured from Sigma-Aldrich, Germany. Manganese(II) acetate tetrahydrate ((CH 3 COO) 2 - Mn$4H 2 O, 98.5% Extra Pure), cobalt(II) acetate tetrahydrate ((CH 3 COO) 2 Co$4H 2 O, 98.5% Extra Pure), potassium hydroxide pellets (KOH, 85% Extra Pure), uric acid ((UA), C 5 H 4 N 4 O 3 , 99% AR), cholesterol (C 27 H 46 O, 97% extra pure), L-ascorbic Acid ((AA), C 6 H 8 O 6 , 99% extra pure), N-methyl-2-pyrrolidone (NMP) (C 5 H 9 NO, 98% Purity), potassium chloride (KCl, 99% Purity), and sodium chloride (NaCl, 99.5% Purity) were purchased from LOBA Chemie Pvt. Ltd. Ultra-pure Milli-Q water (Elga Veolia) was utilized throughout the experiment. All chemicals were of analytical grade and used without ltering. 2.2 Synthesis of M x Fe 2−x O 3 (M =Mn–Co and x=0.01, 0.05, and 0.1) a-Fe 2 O 3 and M x Fe 2−x O 3 nanostructures were synthesized by preparing a homogeneous solution of 0.1 M FeCl 3 $4H 2 O and 0.1 M CO(NH 2 ) 2 dissolved in 80 mL of Milli-Q water for 30 minutes. To this solution, calculated quantities of Mn and Co precursors (with x=0.01, 0.05, and 0.1) were added. The resulting mixture was loaded into a 100 mL autoclave and subjected to hydrothermal processing at 120 °C for 10 hours. Once the reaction was completed, the mixture was allowed to reach room temperature and collected by centrifugation. The precipitate was washed twice with Milli-Q water, followed by a single wash with ethanol. Once washed, the products were dried in an oven at 90 °C overnight and then annealed at 625 °C for 3 hours. A schematic representation of the synthesis process of M x Fe 2−x O 3 is presented in Fig. 1. The resulting products were designated with specic names based on their composition as Co 0.01 Fe 1.99 O 3 (CF1), Co 0.05 Fe 1.95 O 3 (CF5), Co 0.1 Fe 1.90 O 3 (CF10), Mn 0.01 Fe 1.99 O 3 (MF1), Mn 0.05 Fe 1.95 O 3 (MF5), and Mn 0.1 Fe 1.90 O 3 (MF10). 2.3 Fabrication of the electrode The synthesized nanocomposite comprising a-Fe 2 O 3 and M x - Fe 2−x O 3 (M =Mn–Co, and x=0.01, 0.05, and 0.1) was fabricated on the NF electrode. Initially, accurately measured product samples were blended with PVDF in a proportion of 9 : 1 in a mortar and pestle. Subsequently, NMP was gradually © 2025 The Author(s). Published by the Royal Society of Chemistry RSC Adv.,2025,15,34176–34190 | 34177 Paper RSC Advances Open Access Article. Published on 18 September 2025. Downloaded on 10/2/2025 5:21:14 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. View Article Online added drop by drop till the mixture achieved paste consistency. The resulting paste was uniformly coated on a 1 ×1cm 2 area of priorly treated NF and dried in a vacuum oven for an extended period of 24 hours at 60 °C, and the quantity of product loaded was 10 ±0.2 mg. 2.4 Material and electrochemical characterisations The structural properties that enabled the identication and characterization of distinct phases and crystalline compositions of the material under research were analyzed using a Rigaku Miniex 600 powder X-ray diffraction (XRD) device. The measurement covers a range of 5° to 90° and a scanning rate of 3° per minute with monochromatic Cu-Karadiation of wavelength 0.154 nm. The crystallite size (D) was calculated based on the Debye–Scherrer equation (eqn (1)) 45 : D¼Kl bcos q(1) where Kis the Scherrer constant, lis the Cu-Karadiation wavelength, bis the full width at half maximum (FWHM) of the peak, and qis the Bragg angle. The phonon vibration modes were examined using a confocal Raman microscope integrated with a compact Raman spectrometer (Renishaw, UK) with an objective lens magnication of ×50. The optical absorption spectra were recorded using an ultraviolet-near infrared spectrophotometer (UV-vis-NIR, Lambda 950, PerkinElmer, Singapore). The chemical composition of the synthesized sample was determined using a Thermo Fisher Scientic ESCALAB Xi + Xray Photoelectron Spectrophotometer (XPS) with an Al KaXray source (1486.7 eV) for the analysis. Morphological imaging was performed using a eld emission scanning electron microscope (FESEM) (7610FPLUS, Jeol, Japan) equipped with an energy-dispersive X-ray spectrometer (EDAX), enabling a thorough investigation of the morphological characteristics. The electrochemical characterization was conducted using an Autolab PGSTAT204 electrochemical workstation. CV experiments employed a three-electrode conguration, with the synthesized material coated on a NF serving as the working electrode (WE), covering an area of 1 ×1cm 2 . A saturated Ag/ AgCl electrode (with potassium chloride) served as the reference electrode (RE), while a platinum electrode functioned as the counter electrode (CE), both operating within a 0.5 M KOH electrolyte solution. CV analysis was performed within a potential window of 0 to 0.75 V, at scan rates ranging from 5 to 120 mV s −1 . CA studies involved sequentially adding a cholesterol solution into the electrolyte under continuous stirring, maintained at a constant applied potential of +0.55 V vs. Ag/ AgCl. DPV was executed over a potential range of 0.35 to 0.65 V vs. Ag/AgCl at a scan rate of 10 mV s −1 and a pulse amplitude of 50 mV. The ECSA was evaluated by varying the scan rates from 5 to 75 mV s −1 within a potential range. The double-layer capacitance (C dl ) was determined by plotting Dj (where Dj=j a −j c ) against the scan rate. The accurate surface area was then calculated using the formula C dl /C s , 46 where C s is the capacitance of an atomically smooth surface, taken as 40 mF cm −2 . 47,48 The sensitivity was evaluated from the calibration curve obtained by plotting current versus concentration, as expressed by the corresponding equation (eqn (2)): Sensitivity ¼slope of the calibration curve Surface area (2) The formulas used to calculate the LOD and LOQ are as follows (eqn (3) and (4)): LOD ¼3X standard deviation slope of the calibration curve (3) Fig. 1 Schematic representation of the synthesis of M x Fe 2−x O 3 (M =Mn–Co and x=0.01, 0.05, and 0.1). 34178 |RSC Adv.,2025,15,34176–34190 © 2025 The Author(s). Published by the Royal Society of Chemistry RSC Advances Paper Open Access Article. Published on 18 September 2025. Downloaded on 10/2/2025 5:21:14 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. View Article Online LOQ ¼10X standard deviation slope of the calibration curve (4) In order to ensure reproducibility, each test was carried out three times (n=3), and the mean ±standard deviation was employed to depict the results. Statistical signicance was regarded as a variation with p< 0.03 at the level of condence set at 97%. 3. Results and discussion The phase purity and composition of a-Fe 2 O 3 and M x Fe 2−x O 3 (M =Mn–Co, and x=0.01, 0.05, and 0.1) were analyzed using a powder XRD. Fig. S1(a–c) present the XRD patterns of a-Fe 2 O 3 , and M x Fe 2−x O 3 , revealing a clear match with the JCPDS standard (No. 33-0664) and conrming its rhombohedral hexagonal phase. The non-appearance of additional impurity diffraction peaks in the XRD spectra conrms the pristine nature of aFe 2 O 3 . The diffraction peaks at 2qangles of 23.83°, 32.99°, 35.46°, 40.69°, 49.47°, 54.08°, 57.14°, 62.57°, and 63.87° correspond to the (012), (104), (110), (113), (024), (116), (018), (214), and (300) planes of a-Fe 2 O 3 ,. 49 Fig. 2(a and b) present the XRD spectra of a-Fe 2 O 3 , CF5, and MF1 materials under investigation. A notable shiin the diffraction peaks toward lower angles, along with changes in peak intensity, conrms successful doping in the respective host structures. This shiis ascribed to the lattice distortion induced by the incorporation of Co 3+ (0.61 _ A) and Mn 3+ (0.645 _ A), which possess smaller ionic radii compared to Fe 3+ (0.645 _ A), into the a-Fe 2 O 3 lattice. The crystallite size, estimated using the Debye–Scherrer equation (eqn (1)) concerning the (104) diffraction plane, revealed an initial size of 16.18 nm for pure aFe 2 O 3 . In contrast, the doped samples MF1 and CF5 exhibited signicantly larger crystallite sizes of 32.69 nm and 37.39 nm, correspondingly. This increase in crystallinity introduces defects in the a-Fe 2 O 3 structure, enhancing the electron transfer kinetics, which accounts for the superior electrochemical performance of CF5/NF. 50 The Raman spectra analysis depicted in Fig. 2(c) provides valuable insights into the structural characteristics of nanoparticles for a-Fe 2 O 3 , MF1, and CF5. Additional Raman spectra data for the remaining dopants are presented in Fig. S4(a and b). The vibrational spectra of pure a-Fe 2 O 3 exhibit two distinct peaks at 223 and 495 cm −1 corresponding to A 1g modes and ve peaks at 243, 290, 407, 607, and 815 cm −1 corresponding to E g modes. These seven peaks collectively indicate the characteristic vibrational modes of the a-Fe 2 O 3 structure. 49 Notably, the Fig. 2 (a) XRD spectra of a-Fe 2 O 3, CF5, and MF1; (b) shift of the (104) and (110) XRD peaks to lower diffraction angles due to the incorporation of substituents; (c) Raman spectra analysis of a-Fe 2 O 3, MF1, and CF5, respectively; and (d) UV-Vis absorption profiles of a-Fe 2 O 3, CF5, and MF1, respectively. © 2025 The Author(s). Published by the Royal Society of Chemistry RSC Adv.,2025,15,34176–34190 | 34179 Paper RSC Advances Open Access Article. Published on 18 September 2025. Downloaded on 10/2/2025 5:21:14 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. View Article Online absence of extra peaks signies purity and conrms that aFe 2 O 3 nanoparticles are produced without contaminants. The Raman spectra of MF1 and CF5 closely resemble those of pure a-Fe 2 O 3 , underscoring that the doped nanoparticles maintain their unique hematite structure. Despite the incorporation of Mn and Co dopants, the structural characteristics of a-Fe 2 O 3 remain unchanged, as evidenced by the consistent presence of these features across the doped samples. The UV-vis spectra of a-Fe 2 O 3 , CF5, and MF1 were measured in terms of absorbance, as illustrated in Fig. 2(d). The band edge absorption for a-Fe 2 O 3 typically occurs in the 520–565 nm wavelength range. In this study, the compound exhibited strong absorption at 542 nm, with CF5 and MF1 also demonstrating prominent peaks within this range. The band gap energies of the materials were evaluated through the Kubelka–Munk function f(R) 2 plotted against energy (eV) by utilizing reectance spectra depicted in Fig. S4(c). Linear extrapolation at f(R) 2 = 0 estimated the resulting band gap values: a-Fe 2 O 3 (2.05 eV) > MF1 (2.02 eV) > CF5 (1.9 eV) (Fig. S4(d)). The observed reduction in the bandgap can be attributed to the formation of additional energy levels proximate to the valence band edge, which reduces the energy requirement for electronic transitions from valence to conduction bands. 51,52 The electrocatalytic performance of bare NF, a-Fe 2 O 3 /NF, and M x Fe 2−x O 3 /NF (M =Mn–Co, and x=0.01, 0.05, and 0.1) was evaluated by maintaining a scan rate of 50 mV s −1 over a potential window of 0 V to +0.75 V vs. Ag/AgCl using CV measurements in 0.5 M KOH, as depicted in Fig. S2(a–c). Among the doped materials, CF5/NF and MF1/NF demonstrated superior performance, which is demonstrated in Fig. 3(a). Notably, among the three electrodes, CF5/NF proved superior electrocatalytic activity, which is due to cobalt doping endowing more active sites to the a-Fe 2 O 3 surface, making it easier for reactant molecules to adsorb and activate. Furthermore, cobalt ions can change the electronic structure of a-Fe 2 O 3 , resulting in improved charge transfer kinetics and catalytic activity. 53 This is evident from the anodic and cathodic peaks observed during redox reactions. To elucidate the surface-active sites linked to the number of electrons exchanged primarily in the oxidation process, a detailed study of the CV proles of a-Fe 2 O 3 /NF, CF5/NF, and MF1/NF, are illustrated in Fig. S3(a–c). The number of electron transfers for the three primary materials was determined to be 0.6 ×10 19 for aFe 2 O 3 /NF, 4.53 ×10 19 for MF1/NF, and 12.19 ×10 19 for CF5/ NF. 54,55 The present study highlights that within the three materials, CF5/NF exhibits the highest electron transfer rate during the oxidation process, illustrating the better electrocatalytic performance of CF5/NF in alkaline electrolytes. Unlike the BET method, the ECSA analysis entails the complete immersion of the material electrode in the electrolyte, enabling a precise assessment of its surface area activity, and the ECSA results of all the synthesised materials are depicted in Fig. S5(a–g). 48,54 Fig. 3(b) reveals that a-Fe 2 O 3 /NF, MF1/NF, and CF5/NF possessed 2C dl values of 14.54 mFcm −2 , 14.69 mFcm −2 , and 18.58 mFcm −2 , accordingly. The ECSA results correlate with the genuine surface areas of each electrode as 0.182 cm 2 , 0.184 cm 2 , and 0.232 cm 2 in a respective manner. Furthermore, a detailed evaluation of the surface area calibration plot and calculation are enclosed in Fig. S5 (h, i) and Table S1. The electrocatalytic performance results demonstrate that the CF5/ NF electrode possesses the highest ECSA activity, indicating its superior potential for electrocatalytic applications due to an increased number of active sites. To assess the efficiency of electrical charge transfer, EIS was performed, and the results were interpreted through Nyquist plots, as presented in Fig. 3(c), to evaluate the resistance and capacitance of the material. For a more detailed interpretation, the Nyquist plots were modeled with the equivalent circuit R s (Q ct (R ct (Q ce R ce ))), with the tted parameters provided in Table S2. The electrolyte resistance between the working and counter electrodes, represented by solution resistance (R s ), was determined from the x-intercept of the semicircle at high frequencies. Additionally, the electrode–electrolyte interface impedance, comprising parallel elements Q ct and R ct , was included in the model. The internal resistance of the catalyst was represented by parallel elements R ir and Q ir in series with R ct . Notably, the semicircle for CF5/NF displayed a smaller radius, indicating a reduction in internal and charge transfer impedance. This is also conrmed by the Mott–Schottky plot in Fig. 3 (a) Comparative analysis of CVs of bare NF, a-Fe 2 O 3 /NF, CF5/NF, and MF1/NF in 0.5 M KOH at a scan rate of 50 mV s −1 ; (b) ECSA plot of a-Fe 2 O 3 /NF, CF5/NF, and MF1/NF; and (c) Nyquist plots of a-Fe 2 O 3 /NF, CF5/NF, and MF1/NF, respectively. 34180 |RSC Adv.,2025,15,34176–34190 © 2025 The Author(s). Published by the Royal Society of Chemistry RSC Advances Paper Open Access Article. Published on 18 September 2025. Downloaded on 10/2/2025 5:21:14 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. View Article Online Fig. S4(e). This decrease in impedance and solution resistance correlates with the improved electrocatalytic performance observed for CF5/NF. 56 The nanocomposites were morphologically characterized by FESEM and EDAX. The FESEM images of a-Fe 2 O 3 , CF5, and MF1 nanoparticles are depicted in Fig. 4(a–f), respectively, illustrating their rhombus-like structure. The surface morphology is rough and porous, indicating the occurrence of a-Fe 2 O 3 formation (Fig. 4(a, and b)). In CF5 and MF1, the incorporation of dopants into the lattice structure of a-Fe 2 O 3 results in the formation of small granule-like structures on the surface of the framework; however, no surface defects are present within the lattice. Signicantly, this observation implies that the a-Fe 2 O 3 structure is not altered in CF5 and MF1. In Fig. S6(c–e), the EDAX spectra reveal the images showcasing a relatively uniform distribution of Fe and O components throughout the sample, conrming the homogeneity of the synthesized materials. 57 The low-intensity proles observed for the Co and Mn dopants, in conjunction with the base elements, validate the effective incorporation of Co and Mn into the aFe 2 O 3 matrix. Furthermore, EDAX analysis shows the existence of carbon (C) peaks at around 0.2 keV, which is attributed to the carbon tape used for sample handling during SEM and EDAX measurements. XPS spectra was used to validate the surface chemical environment and oxidation state of CF5. Fe, O, Co, and C elements were identied on the surface of CF5, based on the survey scan (Fig. 5a) carried out over an energy range of 100–1000 eV, with no observable traces of contaminants. Two distinguished peaks, individually splitting into doublets, were observed in the highresolution Fe 2p XPS spectra (Fig. 5b), demonstrating spin– orbit coupling in a-Fe 2 O 3 . In particular, the peaks at 723.27 eV and 726.06 eV were found to correspond to Fe 2+ 2p 1/2 and Fe 3+ 2p 1/2 , whereas the peaks at 709 eV and 710.92 eV correlated with the binding energies of Fe 2+ 2p 3/2 and Fe 3+ 2p 3/2 , respectively. The combined presence of Fe 2+ and Fe 3+ oxidation states points to a complex surface chemistry that may be impacted by dopant ion electron donation and oxygen vacancies from hightemperature annealing, among other factors. 58 The clear peaks for the O 1s spectra of CF5 in Fig. 5c indicate the appearance of three distinct forms of surface oxygen species, which are represented by peaks at 530.87 eV involved with iron–oxygen bonding represented by the lowest energy peak, which is associated with lattice oxygen (O L ), 531.63 eV at the intermediate energy level, oxygen vacancies (O v ) appear as chemisorbed oxygen species, and 532.87 eV the nal peak with the highest energy related to adsorbed oxygen species (O c ). The obtained results were consistent with the ndings previously reported in the literature. 59,60 Similarly, the Co 2p XPS spectrum in Fig. 5d shows an obvious peak at 780 eV, which implies the presence of Co in the CF5 lattice. This peak exhibits additional subdivision, with Co 2p 3/2 demonstrating three distinct peaks that show changes in the cobalt oxidation state. In particular, the Co 2+ peak appears at around 777.88 eV, while the Co 3+ peak is detected at 780.94 eV. Furthermore, a satellite peak of Co 2+ is seen at 785.61 eV, which could be ascribed to plasmons, energy loss processes, or shake-ups. In addition, the cobalt 2p 1/2 peak is observed at 792.16 eV, which adds to our understanding of the entire span of cobalt's oxidation states in the CF5 lattice. 61,62 In this study, the Vienna Ab initio Simulation Package (VASP) was used to perform rst-principles DFT simulations in order to Fig. 4 FESEM images of (a and b) a-Fe 2 O 3 ; (c and d) CF5; and (e and f) MF1. © 2025 The Author(s). Published by the Royal Society of Chemistry RSC Adv.,2025,15,34176–34190 | 34181 Paper RSC Advances Open Access Article. Published on 18 September 2025. Downloaded on 10/2/2025 5:21:14 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. View Article Online examine the electronic structure and geometrical optimizations of both pristine and CF5 structures. 63–65 Exchange-correlation interactions were described using the Perdew–Burke–Ernzerhof (PBE) functional in the generalized gradient approximation (GGA), while electron–ion interactions were correctly represented using the projector augmented wave (PAW) approach. 66,67 To maintain a compromise between accuracy and computing economy, a plane-wave basis set with a kinetic energy cutoffof 520 eV was taken into consideration. For selfconsistent computations, a G-centered 3 ×3×1k-point mesh was used to sample the Brillouin zone. Using a U value of 5.0 eV based on the Dudarev and Botton technique, 68 the Hubbard U correction was implemented within the GGA + U framework to account for strong electron correlations due to the presence of localized d-electrons in Fe and Co atoms. To minimize the force on each atom to less than 0.01 eV _ A −1 , structural optimizations were performed. The energy convergence criterion between subsequent self-consistent stages was set at 1 ×10 −5 eV. Aer building the bulk haematite structure to analyze its structural and electrical characteristics, two Fe atoms were swapped out for Co atoms in a supercell structure consisting of 80 atoms to study the effects of Co doping. A thorough examination of the structural and electrical properties of both pristine and CF5 structures was made possible by these computational methods. 3.1. Structural properties DFT computations were used to examine the structural characteristics of pristine a-Fe 2 O 3 and CF5, as shown in Fig. 6(a and b). Fe atoms occupy octahedral positions inside a hexagonal close-packed lattice of oxygen (O) atoms in a-hematite's rhombohedral structure (R-3c). For pristine a-Fe 2 O 3 , the optimized lattice parameters were found to be a=5.44 _ A and a=55.270°, which are values that closely match previous theoretical research and experimental results. The haematite structure's intrinsic cation–anion interactions are reected in the Fe–O bond lengths within the octahedral units, which vary from 1.96 _ A to 2.10 _ A. Minimal structural changes are brought about by the addition of 5% cobalt substitution, in which Co 3+ ions take the place of Fe 3+ ions. Optimal values of a=5.42 _ A and a=55.270° were obtained by a modest contraction of the lattice parameters because of the lesser ionic radius of Co 3+ (roughly 0.61 _ A) compared to Fe 3+ (about 0.645 _ A). The unit cell volume decreases as a result of this little drop in lattice characteristics, indicating that cobalt was successfully incorporated into the haematite lattice without causing signicant structural deformities. Additionally, there are minor variations in the bond lengths between the Co-substituted Fe–O octahedra and pure Fe–O Fig. 5 XPS spectrum of CF5: (a) survey; (b) Fe 2p; (c) O 1s; and (d) Co 2p. 34182 |RSC Adv.,2025,15,34176–34190 © 2025 The Author(s). Published by the Royal Society of Chemistry RSC Advances Paper Open Access Article. Published on 18 September 2025. Downloaded on 10/2/2025 5:21:14 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. View Article Online bonds. The Co–O distances, which range from 1.95 _ A to 1.97 _ A, are marginally shorter than the comparable Fe–O bonds. The stronger Co–O interactions brought about by the more localized electronic character of the Co 3d states are responsible for this change. 3.2 Electronic properties of a-Fe 2 O 3 and CF5 The electronic structure of a sensing material plays a key role in dening its conductivity, charge transfer efficiency, and surface reactivity, all of which directly inuence its sensing ability. In biosensing applications, a-Fe 2 O 3 is a promising material because of its surface activity, tunable bandgap, 69–71 and a stable semiconducting nature. However, its efficiency is limited by its weak electrical conductivity and comparatively high bandgap. In order to alter the electrical structure and possibly lower the bandgap, introduce mid-gap states, and improve charge carrier mobility, Co doping was investigated. Co-incorporation's effects on the material's electronic conductivity and surface interaction potential can be understood by examining the density of states (DOS) and band structure of both pure and CF5. Since enhanced conductivity and charge transfer capacities can improve the sensor's responsiveness, this information is crucial for optimizing hematite-based materials for cholesterol detection. The creation of more effective biosensors is guided by the fundamental insights into the material's potential as an active sensing element that the electronic structure analysis offers. Fig. 6 Optimized geometry of (a) Fe 2 O 3 and (b) CF5. Fig. 7 (a) Electronic band structure of a-Fe 2 O 3 in the Brillouin zone of the rhombohedral cell, emphasizing the band gap along the highsymmetry path G-T-H 2 jH 0 -L-G-S 0 jS 2 -F-G. (b) a-Fe 2 O 3 density of states (DOS), showing the band gap and the distribution of electronic states. (c) Electronic density of states (DOS) of CF5, demonstrating the effect of Co doping on electronic characteristics and band gaps. © 2025 The Author(s). Published by the Royal Society of Chemistry RSC Adv.,2025,15,34176–34190 | 34183 Paper RSC Advances Open Access Article. Published on 18 September 2025. Downloaded on 10/2/2025 5:21:14 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. View Article Online 3.3 Band structure analysis Pristine a-Fe 2 O 3 is an indirect bandgap semiconductor, according to the band structure calculations, with the conduction band minimum (CBM) at a separate k-point and the valence band maximum (VBM) at the G-point. The calculated bandgap, which is shown in Fig. 7(a), is 2.3 eV, which is the same value as previously published. The Fe 3d orbitals predominate in the conduction band, whilst the O 2p orbitals contribute to the majority of the valence band. The electrical structure is significantly inuenced by the Fe 3d-O 2p hybridization. The band structure changes noticeably when 5% Co is doped at Fe sites. The bandgap drops to 1.93 eV, suggesting the formation of impurity states close to the conduction band, as shown in Fig. 7(c). The Co 3d orbitals, which add more electronic channels and may improve electrical conductivity, are the source of these states. Moreover, a shiin the Fe 3d bands results from a minor perturbation of the Fe electronic states caused by Co inclusion. Moreover, compared to normal LDA/GGA, which frequently underestimates the band gap, the LDA + U technique tends to increase it. 72,73 Compared to the conventional DFT approaches, the use of U helps to correct the electron–electron interactions, especially in the Fe 3d orbitals, which results in a bigger bandgap and a more accurate representation of the electronic structure. 74 3.4 DOS analysis As shown in Fig. 7(b and c), the atom projected density of states (PDOS) offers vital information on the electrical structure of both pure and CF5. The O 2p states contribute to the majority of the valence band in pure a-Fe 2 O 3 , with Fe 3d states contributing close to the VBM. Fe 3d orbitals contribute to the conduction band, which has a wide band gap and restricts electrical conductivity. The DOS undergoes notable changes at 5% Co doping, especially in the vicinity of the CBM. By adding Co 3d states, more electronic states are produced close to the Fermi level, which lowers the bandgap and increases the concentration of carriers. This change suggests increased Fe–Co hybridization-induced electrical conductivity. Additionally, the doped system is better suited for applications needing effective charge transfer because the presence of Co states close to the CBM promotes electron transport. Fig. 8 (a) Comparison of CVs for the influence of 2 mM cholesterol in 0.5 M KOH at a scan rate of 50 mV s −1 for a-Fe 2 O 3 /NF, CF5/NF, and MF1/ NF; (b) scan rate study of CF5/NF in 0.5 M KOH from 5 to 120 mV s −1 . The anodic and cathodic peak currents are inserted as a function of the square root of the scan rate (inset); (c) successive addition analysis for CF5/NF at a scan rate of 50 mV s −1 ; (d) CA study of CF5/NF conducted in a 0.5 M KOH solution, and the measurements performed under stirring conditions at an applied potential of +0.55 V; (e) calibration plot of CF5/ NF to estimate the cholesterol levels; and (f) DPV study of CF5/NF to analyze the electrochemical response to successive cholesterol additions. 34184 |RSC Adv.,2025,15,34176–34190 © 2025 The Author(s). Published by the Royal Society of Chemistry RSC Advances Paper Open Access Article. Published on 18 September 2025. Downloaded on 10/2/2025 5:21:14 PM. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. View Article Online