Electrochemical studies on doped SnO₂ nanocomposite for selective detection of lung cancer biomarkers

Worldwide, cancer is the second largest cause of mortality according to the World Health Organization (WHO). Lung cancer (LC) is one of the most vulnerable types of cancer that are mostly responsible for death in people of all ages and genders. The dominant cause of lung cancer is smoking as cigarette includes 73 carcinogens such as polonium-210, NNK, 1, 3-butadiene, and pyrene. In many industrialized countries, it was observed that people who work in an environment with smoke exposure can also affected by lung cancer. The survival rates of lung cancer patients greatly depend on the nature of cancer, its stage, and how early it is being identified.¹ Therefore, detecting LC at an early stage in order to improve diagnostic accuracy and survival rates is essential. The continuous development of a cost-effective, easy, and precautionary strategy to improve early diagnosis based on the analysis of numerous volatile organic compounds (VOCs) emerging from cancer cells has recently attracted much interest.^2,3 VOCs are a wide range of compounds released from the human body that frequently represent the person's metabolic state.⁴ As a result, the detection of specific concentrations of VOCs from exhaled breath is used as cancer biomarkers since the collection and evaluation of the sample is easy.^5–7 Exhaled breath testing is painless and noninvasive, making it ideal for even children, elderly people, and severely ill patients. Hence, the development of sensors with the help of e-nose could serve as a useful tool for early diagnosis, continuous monitoring of health status, and medication.^8,9 Several studies have found that a variety of VOCs are closely linked to early detection of lung cancer.¹⁰ Particularly, acetone and toluene are the most significant biomarkers that can aid in the early diagnosis of LC in the range of 80–100 and 34.57–390.6 ppb, respectively. However, the presence of toluene and acetone in healthy humans is in the range of 20–30 and 44–531.3 ppb, respectively.^11,12 Chemoprevention and medication improvements have the potential to diminish lung cancer-related mortality.

Metal oxide semiconductor (MOS) nanomaterials are one of the promising approaches used for rapid detection of VOCs due to their high sensitivity, low cost, fast detection, good selectivity, user friendliness, and potential for point of care testing.^13–15 In addition, MOS materials have persistent chemical transduction capabilities, which may reversibly convert chemical interactions on a surface to change electrical conductivity.^16–18 Most widely used MOS materials for VOC identification are In₂O₃, SnO₂, WO₃, rGO-SnO₂,¹⁹ etc. Among all MOS materials, SnO₂ is found to be the most well-known and commercially used MOS material because of its low cost, great sensitivity to different VOCs, and compliance with microfabrication methods.^20,21 However, its VOC sensing capabilities must be developed further for sensitive and selective detection of various biomarkers. These objectives can be met by using porous SnO₂ nanostructures with unique features, such as a large specific surface area, and integrating some useful additional materials. The transition metals, such as Pt, Ni, Pd, Fe, and Co, and their oxides are well investigated dopants for SnO₂ gas sensors due to their excellent catalytic capabilities.^22–24 

In this study, we have synthesized manganese (Mn) and copper (Cu) doped SnO₂ nanocomposites to achieve high sensitivity and selectivity toward the detection of acetone and toluene. The synthesized Mn/SnO₂ and Cu/SnO₂ nanocomposites will be beneficial for achieving superior sensing toward VOCs because of the combined effects of SnO₂ (enlarged active surface area and high porosity) and Mn and Cu dopants (high conductivity and appreciable electrocatalytic activity).²⁵ The synthesized nanocomposites are characterized by various analytical techniques and utilized for VOC detection.

All the reagents are of analytical grade and were used as received without further purification. Stannous chloride dihydrate (SnCl₂·2H₂O), trisodium citrate dihydrate (Na₃C₆H₅O₇·2H₂O), copper acetate monohydrate [Cu(CH₃CO₂)₂·H₂O], manganese acetate tetrahydrate [Mn(CH₃CO₂)₂·4H₂O], potassium chloride (KCl), and potassium ferrocyanide [K₃Fe(CN)₆] were purchased from Loba Chemicals, and the solutions were prepared using deionized water (D.I. water).

The synthesis of pristine and Mn/Cu-doped SnO₂ nanocomposites was carried out using the hydrothermal method. First of all, 0.1M of stannous chloride was dissolved in 50 ml of D.I. water with vigorous stirring in a water bath at 60 °C. Then, 1M NaOH solution was added dropwise to the above solution under constant magnetic stirring. The above homogeneous solution was then transferred into a 120 ml Teflon-lined autoclave and heated at 180 °C for 24 h. The obtained precipitates were then centrifuged with D.I. water and ethanol and then dried at 60 °C for 24 h. Finally, the synthesized powders were calcined at 700 °C for 2 h. Similarly, the doping of copper (Cu) and manganese (Mn) was carried out by dropwise addition of 0.1M solution of Cu {copper acetate monohydrate [Cu(CH₃CO₂)₂·H₂O]} and Mn {manganese acetate tetrahydrate [Mn(CH₃CO₂)₂·4H₂O]} precursors into the 50 ml stannous chloride solution (0.1M).

Mn/SnO₂ and Cu/SnO₂ nanocomposite films were fabricated by dissolving an appropriate amount of nanocomposites in an organic solvent. These mixtures were then ultrasonicated to obtain a homogeneous suspension. Furthermore, Mn/SnO₂ and Cu/SnO₂ nanocomposite films were prepared on cleaned indium tin oxide coated glass slides (0.5 × 0.5 cm²) by the screen printing method. The films were further used to study electrochemical characteristics by cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS).

The structural properties of all synthesized films were studied by x-ray diffraction (XRD) analysis using a Bruker D8 Advance with a monochromator of CuKα radiation (λ = 1.541 87 Å) over a 2θ range of 5°–80°. Fourier transform infrared spectroscopy (FTIR) was used to determine the chemical functional groups of all synthesized materials using a Shimadzu FTIR-8900 with a spectral range of 4000–400 cm⁻¹. Field-emission scanning electron microscopy (FESEM) analysis was carried out using a FET Nova Nano SEM 450 to study the morphology of synthesized materials. Photoluminescence spectra were obtained to study the vacancies/defects present in samples using a Horiba iHR320 with a Syncerity CCD. The excitation wavelength is 325 nm (Kimmon IK Series He–Cd Laser).

In order to investigate the electrochemical behavior of the fabricated nanocomposite, cyclic voltammetry was carried out using a three-electrode system, an ITO glass plate (0.5 × 0.5 cm²) as a working electrode, Pt as a counter electrode, and Ag/AgCl as a reference electrode. For cyclic voltammetry, the synthesized nanocomposites were deposited on a pre-cleaned ITO glass slide. The scanning was done in the potential range of −1–1 V at a scan rate of 80 mV/s in 0.1M phosphate buffer (pH 6), acting as a supporting electrolyte. Furthermore, in order to study the fabricated nanocomposite for the detection of VOCs, the CV of the fabricated film was carried out in the presence of different concentrations of VOCs. In addition, cyclic voltammetry (CV) at different scan rates and electrochemical impedance spectroscopy (EIS) were performed using a potentiostat/galvanostat electrochemical analyzer (Metrohm) in 5 mM of K₃FeCN₆ solution to study the electrochemical properties and sensing mechanism of synthesized materials toward VOCs.

X-ray diffraction (XRD) was used to examine the phase composition and crystal structure of pristine and metal-doped SnO₂. The XRD graph of all the as-synthesized nanomaterials is shown in Fig. 1. According to JCPDS Card No. 01-077-0447, the diffraction peaks of synthesized pristine SnO₂ nanoparticles at 2θ = 26.6°, 33.8°, 51.8°, 54.8°, 58°, 61.9°, 64.7°, 66°, 71.3°, and 78.79° are indexed to the tetragonal rutile structure (a = 4.735 Å and c = 3.181 Å). The broad diffraction peaks of the synthesized materials indicate that all samples are nano-crystalline in nature. However, in the present investigation, it was observed that Cu-doped nanocomposites and Mn-doped nanocomposites show a tetragonal rutile structure without any extra secondary phase/impurity. However, in the case of doping, a slight shift in the peak position is observed as compared to pure SnO₂ [shown in Fig. 1(b)]. These changes in the peak position may attributed be tensile/compressive strain arising on the lattice after the incorporation of metals into the pure SnO₂ nanoparticles. Furthermore, lattice parameters (a = b and c) and volume for tetragonal SnO₂ nanoparticles are calculated using the following equations and are summarized in Table I:²⁶ 

where h, k, and l are the Miller indices, “a” and “c” are the lattice parameters, and d is the spacing between the planes of definite Miller indices. The decreases in lattice parameters (a and c) and crystal structure volume with Mn and Cu doping might be attributed to the shorter bond lengths of Mn–O and Cu–O than Sn–O as shown in Table I. This bond length was used to study the effect of transition metal (Mn and Cu) doping on the Sn–O bond length (L) of synthesized materials, which was calculated using the following equation:

where δ is the oxygen shift parameter, and its equation is given as follows: $δ=1/4×1−a2/c2$⁠.²⁶ Moreover, the type of strain occurred in the case of doping can be estimated using a Williamson–Hall recommended mathematical expression in Eq. (4) by assuming that the contributions of particle size and strain to line broadening are unrelated to one another, and both have a Cauchy-like profile,

where λ is the wavelength of x-rays (1.54 Å), β is the FWHM of the synthesized nanomaterials, θ is the Bragg angle, D is the crystallite size, and ε is the internal lattice strain. For all the significant peaks with comparably higher intensity, the D and ε values are determined using the least squares fit of β cos θ/λ vs sin θ/λ plots, as shown in Figs. 2(a) and 2(b). The magnitude of lattice strain is determined using the slope of this linear fit, and the crystallite size is determined using the intercept on the β_hkl cos θ axis. The intercept has moved to a lower value, indicating that the crystallite size has increased, and the negative slope indicated that compressive strain is obtained due to transition metal ion doping (bond length reduction).²⁷ This demonstrates that Mn/Cu binds to the Sn site in SnO₂ host materials.²⁸ 

Furthermore, the crystallite size of synthesized pure SnO₂ nanoparticles and metal-doped nanocomposites can be calculated using Scherer’s equation as follows:

Pure SnO₂ has a crystallite size of 15.67 nm, while the doping of metals into SnO₂ increases the average crystalline size of doped materials (Mn = 21.44 nm and Cu = 24.88 nm). It may be due to the large grain growth of Mn- and Cu-doped SnO₂ as compared to pristine SnO₂ nanoparticles.²⁹ 

Fourier transform infrared spectroscopy (FTIR) is performed to investigate the functional groups involved in the interaction of Mn/SnO₂ and Cu/SnO₂ nanocomposites. Figure 3 shows the FTIR spectra in the range from 4000 to 400 cm⁻¹ for all sintered undoped and doped nanoparticles. Three prominent peaks were observed at 3396, 1695, and 1373 cm⁻¹ corresponding to the O–H stretching, C=O stretching, and O–H bending, respectively. Furthermore, the anti-symmetric vibration of the Sn–O–Sn fingerprint of SnO₂ is observed in the spectrum at 600 cm⁻¹ and confirms the formation of metal oxide.

Using UV–vis spectra, the Tauc plot of (αhν)² vs bandgap energy is also plotted as shown in Figs. 4(a)–4(c) to estimate the value of bandgap. The bandgap values calculated for SnO₂, Cu/SnO₂, and Mn/SnO₂ are 3.68 eV, 2.9 eV, and 3.4 eV, respectively. The absorption band edge gives birth to redshift phenomena. This bandgap narrowing of the transition metal phenomenon is reported due to the Moss–Burstein effect in which the doping of transition metals shifts the Fermi level toward the conduction band because of sp–d spin exchange interactions.³⁰ 

Photoluminescence is a widely used technique for investigating electron–hole surface interactions for determining bandgap energy in semiconductor materials and capable of detecting specific defects and impurity levels.³¹ As a result, photoluminescence spectra were acquired, retaining the excitation wavelength at 325 nm, in order to examine the existence of defects in the as-synthesized pure and transition metal-doped SnO₂ samples, as shown in Figs. 5(a)–5(c). The synthesized nanomaterials have a prominent peak at 368 nm, but generally, the near band-edge emission of SnO₂ is ∼390 nm.³² This blueshift of the peak may be attributed to electron transition, mediated by defect levels in the bandgap.³² Furthermore, the occurrence of double ionization oxygen vacancies causes emissions in the visible region around 400–480 nm.^33,34 The emission band recorded at 550–600 nm has been linked to crystalline defects including Sn interstitials or Sn vacancies, as well as oxygen vacancies, which generate a large number of trapped states inside the bandgap of SnO₂ nanostructures.³⁵ The red/orange emission band (600–700 nm) was ascribed to deep traps that generate defect energy levels inside the SnO₂ bandgap.^36,37 This visible emission occurred when a conduction band electron recombines with the “$Vo++$” centers, while the “$Vo++$” center is generated by recombination of an active hole with an electron in the deep trap $(Vo+)$⁠.³⁸ 

The photoluminescence investigations also show that when transition metals are doped, the photoluminescence intensity reduces significantly. This could be described by a concentration quenching effect caused by an energy interchange between two Mn/Cu ions. This nonradiative mechanism is often defined as a cross-relaxation process, wherein two adjacent ions exchange energy through the process of excitation energy migration.³⁹ 

The FESEM technique was then used to obtain the information about the surface morphology, homogeneity, and particle size. The FESEM images of sintered pure SnO₂, Cu–SnO₂, and Mn–SnO₂ are displayed in Figs. 6(a)–6(c). The FESEM analysis indicates the spherical shaped morphology of the pristine SnO₂ nanomaterial with a particle size of 20–38 nm. However, in the case of Cu- and Mn-doped SnO₂ nanocomposites, high agglomeration was observed. Further, the energy dispersive x-ray spectroscopy (EDS) images shown in Figs. 6(d) and 6(e) indicate the presence of Sn, O, Mn, and Cu, which confirms the formation of Cu/SnO₂ and Mn/SnO₂ nanocomposites. The atomic percentage of the major constituents (Sn, O, Mn, and Cu) of the nanocomposite materials along with the average atomic ratio (Mn:Sn and Cu:Sn) is shown in Table II.

The surface properties of the different electrodes were characterized by CV in 5 mM [Fe(CN)₆]^3−/4− solution prepared in 0.1M KCl solution at a scan rate of 100 mV. Figure 7 shows the CVs of the bare ITO electrode and ITO modified with SnO₂, Cu/doped SnO₂, and Mn/doped SnO₂ nanocomposites. The electrochemical response of ferricyanide as a redox probe is found to be a reversible process. In the case of bare ITO, a pair of well-separated redox peaks with a peak separation (ΔEp) of 0.15 V was observed. After modification of ITO with SnO₂, the redox peak currents increase, which may be due to the presence of semiconducting materials. Furthermore, Cu- and Mn-doped SnO₂ nanocomposites showed highly enhanced peak current as compared to SnO₂, which is attributed to the large electrical conductivity of transition metal ions. In addition, a shift in ΔEp is observed in the case of Cu (0.34 V) and Mn (0.22 V) doping compared to pristine SnO₂ (0.18 V). This demonstrates that the combination of the metal-doped SnO₂ can improve the catalytic electrochemical behavior and the required conduction paths on the electrode surface to promote the transfer of electrons on the modified electrode surface.

Furthermore, scan rate studies were carried out to understand the electrochemical behavior of the modified electrode as shown in Fig. 8. The linear relationship was observed for oxidation and reduction peak currents of all synthesized materials as a function of the square root of scan rates in the range of 5–100 mV/s. The redox peak potential is found to be shifted (i.e., anodic peak potential toward positive and cathodic peak potential toward negative) as the scan rate increases, indicating that the process is a diffusion control process. In Nernstian or reversible systems, the peak current is obtained by the Randles–Sevcik method as follows:⁴⁰ 

where I_p is the peak current, A is the area of the electrode (0.25 cm²), n is the number of electrons transferred in the redox reaction (1), C is the concentration (5 mM), D is the diffusion coefficient of the transferred species, and is the scan rate (100 mV/s). The magnitude of the diffusion coefficient with SnO₂, Mn/SnO₂, and Cu/SnO₂ films was found to be 20 × 10⁻⁶, 54 × 10⁻⁶, and 81 × 10⁻⁶ cm²/s, respectively. These results indicated that doped nanocomposites have a high diffusion coefficient than pristine nanocomposites due to the high electrocatalytic activity of doped transition metal ions (Mn and Cu).⁴¹ 

In order to determine the selectivity of the proposed nanocomposite for the detection of VOCs, the response of all the synthesized samples is also examined. The results demonstrated that pristine SnO₂ is sensitive to all types of VOCs [Fig. 9(a)]. However, Figs. 9(b) and 9(c) indicate that the high degree of selectivity is obtained in the presence of copper and manganese doping toward acetone and toluene, respectively, as compared with other VOCs and pristine SnO₂. Furthermore, the selectivity factor of Cu/SnO₂ nanocomposites for acetone is 12, 3.6, 2.6, 2.8, and 1.8 times higher than that of benzene, isopropyl alcohol (IPA), methanol, toluene, and ethanol, respectively. However, in Mn/SnO₂ nanocomposites, the selectivity factor for toluene is 6.6, 4, 9.3, 11.2, and 10 times higher than that for benzene, isopropyl alcohol (IPA), acetone, methanol, and ethanol, respectively. This high selectivity of Mn and Cu loaded SnO₂ sensors toward toluene and acetone may be promoted by the catalytic action of metal ions, i.e., spill-over effects.^41,42

To estimate the quantitative detection of acetone and toluene using the synthesized Cu/SnO₂ and Mn/SnO₂ nanocomposites, the electrochemical sensing characteristics are studied by cyclic voltammetry (CV). The CV curve of Cu/SnO₂ nanocomposites in 0.1 M phosphate buffer solution (pH 6) indicated well-defined peaks proportional to the concentration of toluene in the range of 20–100 ppb [at room temperature (RT)] as shown in Figs. 10(a) and 10(b). The linearization equation obtained is I (μA) = 0.80 + 0.011 c/μM with the correlation coefficient of 0.9974. However, the CV curve of Mn/SnO₂ nanocomposites demonstrated a good linear range from 1 to 1000 ppb toward acetone (at RT) as shown in Figs. 10(a) and 10(b). The piecewise linearization equations are I (μA) = 17.41 + 0.22 c/μM for 1–100 ppb and I (μA) = 35.67 + 0.023 c/μM for 200–1000 ppb with the correlation coefficients of 0.9981 and 0.9915, respectively. The limit of detection for toluene is 1 ppb, and for acetone, it is 0.1 ppb. These results completely satisfy the WHO limits for acetone and toluene for lung cancer detection.^11,12 In addition, in comparison to most electrochemical sensors, the detection limit and the linear range for the proposed sensor show potential advantages (Table III). Furthermore, the obtained results are also comparable with the commercially available sensors for VOC detection, such as Unitec SRL, Citytech, Environmental Sensors CO, and applied sensor (TGS 8100).^43,44

The properties of the fabricated electrode interface are further analyzed using electrochemical impedance spectroscopy (EIS). The semicircle section of a typical Nyquist plot corresponds to the electron-transfer resistance (R_ct) at higher frequencies, while the linear part at lower frequencies shows the diffusion limiting process. Figures 11(a)–11(c) show the charge transport processes in undoped and doped SnO₂ modified electrodes with the appropriate Randles equivalent circuit. In the Randles circuit, it is assumed that both the resistance to charge transfer (R_ct) and the Warburg impedance (W) are in parallel to the constant phase element (CPE) and double layer capacitance C_dl. At the electrode–electrolyte interface, the charge transfer resistance (R_ct) is calculated to study the electron transfer process. The value of R_ct depends on the dielectric and insulating features of the electrode. Thus, R_ct is an important parameter in stating the effectiveness of the coated electrode material for the charge transfer from the solution to the electrode. It is observed that Cu- and Mn-doped SnO₂ nanocomposites have a low R_ct value (∼1739.4 Ω for Cu and ∼2220.0 Ω for Mn) due to excellent ionic conduction offered by dopants on the surface of pure SnO₂ (∼6734.7 Ω). This implies that the presence of metal transition elements with SnO₂ is suitable for achieving a superior sensor response as metal doping enhances the charge transfer kinetics due to their higher conductivity. The EIS results are accompanied by the CV data, which showed that the metal-doped nanocomposite exhibited higher anodic and cathodic peak currents than pure SnO₂. Furthermore, in the presence of VOCs, the R_ct value is noticeably greater than the buffer. This can be associated with the chemical texture of VOCs, which can readily release the proton and oxidize itself, resulting in the extraction of the electron from the sensor surface.

A possible sensing mechanism is proposed based on the CV and EIS studies, which indicated that during the blank CV (buffer solution) analysis of the synthesized nanocomposite, oxygen molecules are released by electrolysis in buffer solution. These oxygen molecules get attach to the surface of SnO₂ nanoparticles and collect electrons from the conduction band, yielding oxygen species, such as O²⁻ and O⁻. Furthermore, it is discovered that in the presence of VOCs (acetone and toluene), the surface oxygen species of SnO₂ will react with VOCs and allow the trapped electron to return to the SnO₂ conduction band; this reduces the depletion layer width and therefore lowers the height of the potential barrier. This improves the conductivity of the electrode, and thus, increases in current were seen while sensing VOCs, as illustrated in Figs. 10(a) and 10(c). Furthermore, doping the transition metal in SnO₂ usually resulted in an abundant surface oxygen vacancy, which adsorbs more oxygen molecules than ideal sites due to the reduced adsorption energy of oxygen molecules in the vacancy site.^48,54 As a result, there were more active sites for VOC adsorption on the surface of metal-doped SnO₂. Furthermore, in the case of nanocomposites, a large number of electrons transferred lead to a significant reduction in surface charge density, resulting in an increase in oxidation current. As a result, in the case of metal-doped SnO₂, excellent VOC sensing capability is obtained.

In this work, pristine and metal-doped SnO₂ nanocomposites have been successfully synthesized by the hydrothermal method and applied as sensing materials for the determination of lung cancer biomarkers. The CV and EIS analyses performed on Cu/SnO₂ and Mn/SnO₂ nanocomposites showed the best selectivity toward toluene and acetone along with a linear dynamic range from 20 to 100 and 1 to 1000 ppb, respectively. In addition, the lower detection limit obtained for toluene (1 ppb) and acetone (0.1 ppb) was well below the permissible limit of cancer biomarkers. This improved sensing response is observed due to the synergistic effect of SnO₂ and metal dopants (copper and manganese) as these materials have high electrocatalytic activity, a large surface area, and high conductivity. The findings of this work suggest that doped SnO₂ could be employed as a promising sensing material for the identification of different VOCs in an effort to diagnose lung cancer quickly and accurately.

The authors would like to acknowledge the “Board of Research in Nuclear Science (BRNS)” for funding under the DAE YSRA scheme (Grant No. 59/20/02/2020-BRNS/59001).

The authors do not have any conflicts of interest.

The data that support the findings of this study are available from the corresponding author upon reasonable request.