Influence of Synthesis Method and Electrode Geometry on GHG-Sensing Properties of 5%Gd-Doped SnO₂

Abstract

This study investigates the influence of synthesis methods and electrode geometry on the physico-chemical properties of 5%Gd-doped SnO₂. Two distinct synthesis routes, co-precipitation and hydrothermal growth, were employed, resulting in powders denoted as SnO₂: Gd 5%-CP and SnO₂: Gd 5%-HT. Morpho-structural and textural analyses reveal a uniform morphology consisting of quasi-spherical nanoparticles with dimensions of ~6 nm and mesoporosity for CP and a non-uniform morphology with larger nanoparticles of ~42 nm, with irregular shapes and macroporosity for the HT sample, respectively. The powders were deposited onto alumina substrates equipped with platinum interdigital electrodes with alternative gaps of 200 μm and 100 μm. The back-side heater allows for variation in the temperature of the layer. Sensing properties assessed under in-field-like atmospheres simulated by a computer-controlled Gas Mixing System reveal higher sensitivity to methane compared to carbon dioxide. Although the sensor signals did not differ quantitatively, they exhibited distinct saturation tendencies with an increasing methane concentration, attributed to the morpho-structure and porosity induced by the synthesis method. Differentiation was achieved by varying the interdigital gap of the electrodes, highlighting different sensor signals and conduction mechanisms, determined by the specific size of the crystallites.

1. Introduction

It has been observed in the past decades that the atmospheric abundance of greenhouse gases (GHGs) has exhibited a continuous increase [1]. In direct relation, the overall Earth’s temperature has shown a subsequent rise, with September 2023 reported as the warmest month in the past 174 years [2]. The primary GHGs of concern are carbon dioxide (CO₂), methane (CH₄), and nitrous oxide (N₂O) [3]. While the main constituents of the Earth’s atmosphere are nitrogen and oxygen, which are symmetric molecules without a dipole moment, the main GHGs, CO₂ and CH₄, exhibit a permanent dipole moment leading to IR adsorption [4,5]. Accordingly, IR technologies play crucial roles in monitoring GHGs in the atmosphere, despite the high costs of periodic maintenance and data acquisition subsystems. Thus, the challenge lies in the development of simple devices, i.e., sensors capable of selectively monitoring GHG emissions. Among various sensing techniques, metal oxide semiconductor (MOS)-based chemo-resistive sensors have shown better performance than spectroscopic ones, with high sensor response, low fabrication costs, ease of operation, and low power consumption [6]. Their problem remains the selective sensitivity to a gas compared to other potential interfering gases. The main interferent, humidity, is often neglected, although it is permanently present in the “in-field” atmosphere. Regarding CO₂ and CH₄, selective detection allows specific measures to reduce different pollution factors, although they are present simultaneously. More specifically, CO₂ emissions are mainly due to factories, combustion sources, and heavy traffic, while CH₄ emissions are due to anaerobic soils, forest fires, coal mines, oil and gas fields, the digesting processes of ruminant animals, landfills, and other waste management facilities. Furthermore, CH₄ is a long-lasting GHG with a higher warming potential than CO₂ [7].

So, the challenge of selectivity is complex; it concerns both the potential noxes and the relative humidity (%RH) of the air. In the case of chemo-resistive sensors, this challenge is addressed by examining the intrinsic properties of the MOS-sensitive element, which can be modulated by the synthesis method determining morphological aspects such as grain shape and size, by doping, or by using adaptative heterostructures [8]. Various MOS-based sensors have shown promising sensing properties for GHG detection [9,10]. However, considering the additional applicative requirements such as good sensor signal (response) and selectivity, fast transients, and simple device integration for direct use under in-field conditions, there is a continuous need for further improvements in MOS-based sensors for GHG detection [11]. It should be emphasised that although interest in GHG detection has increased, the potential interferences have generally not been considered.

Table 1. Comparison of various MOS-based gas sensors towards GHGs.

Sensing Material	Sensor Signal	Target Gas	Relative Humidity	Temperature °C	Reference
Pt/SnO2	3.6	CH4	Absent	100	[12]
Sb/Pd/SnO2	5/1.2	CH4/CO2	Absent	280	[13]
Ca/Pt/SnO2	1.02	CO2	Present	270	[14]
SnO2@CdO	10.67	CO2	Absent	30	[15]
SnO2	2.48	CO2	Present	200	[16]

presents a survey of gas sensing performances of SnO₂-based gas sensors. The sensor signals were calculated as the ratio between the reference electrical resistance value in air and the resistance value after exposure to the respective test gas.

Besides SnO₂, there is a recent focus on rare-earth oxycarbonates Ln₂O₂CO₃, which are unstable, and rare-earth oxides Ln₂O₃, which exhibit sensor signals to CO₂ ranging between 1 and 10, depending on the lanthanide element (Nd, Sm, Gd, Dy, Er, and Yb). Among these, Gd₂O₃ has demonstrated the highest sensitivity; however, its selectivity towards methane has not been analysed [17]. Our recent study reported the sensing properties of SnO₂, Gd₂O₃, and Gd-doped SnO₂ for in-field conditions with variable RH. All powders were prepared by using the wet chemical co-precipitation method. Structural analyses revealed that 5% Gd represents the doping limit, beyond which the incipient secondary phase of Gd₂O₃ appears. The selection of materials was based on the CO₂ sensor signal criterion and Gd-doped SnO₂ was excluded from the selectivity study against methane [18].

Apart from doping the base material with different ions or cations, the geometry of interdigital electrodes should be considered an important factor that significantly contributes to the overall gas sensing performances of MOS-based gas sensors [19]. During operation under field atmospheres, the role of the interdigital electrodes manifests through charge transport within the polycrystalline sensing material. Several studies report on how the width and gap of interdigital electrodes affect gas sensitivity [20]. For instance, in the case of SnO₂-based sensors, the response to NO₂ was significantly influenced by the interdigital gap; namely, sensors with wider gaps (30 μm) demonstrated good sensitivity to high concentrations of NO₂, whereas those with smaller gaps (1 μm) exhibited the opposite effect, with increased sensitivity and faster transients to lower NO₂ concentrations [21]. Gardner et al. [22] reported an exponential increase in thin film WO₃ sensitivity as the gap between electrodes decreased from 0.8 to 0.1 μm, as detailed in a study on the impact of the electrode gap on NO₂ sensing capabilities [23]. This result was attributed to the reduced number of WO₃ grains between the electrodes as the gap narrowed, leading to a decrease in grain boundary resistance during charge conduction.

The present work focuses on the possibility of modulating the sensing properties of 5%Gd-doped SnO₂ based on the following aspects: the synthesis method (co-precipitation versus hydrothermal), the operating temperature of the sensitive layer, and the interdigital electrode gap (200 μm versus 100 μm) of the layer substrate. The sensor signal is calculated from the variation in electrical resistance induced by the presence of the main GHGs (CO₂ and CH₄) in the atmosphere, considering the RH specific to in-field situations.

2. Materials and Methods

2.1. Powders Synthesis

Starting from the premise that depending on the preparation method, the materials present different morphologies [24], 5%Gd-doped SnO₂ was synthesised using two alternative methods: (i) Co-precipitation: Sn⁴⁺ and Gd³⁺ precursor solutions (containing the amounts corresponding to the formula Sn_1−xGd_xO_(4−x)/2 and the doping amount of 5 at. %Gd) were precipitated with NaOH at a temperature of 80 °C in the presence of CTAB (cetyltrimethylammonium bromide) surfactant for 3 h. The resulting precipitate was separated by centrifugation, washed repeatedly with double-distilled water, and subsequently dried at 80 °C in air. The powder was finally calcined at 550 °C for 2 h. (ii) Hydrothermal: Sn⁴⁺ and Gd³⁺ precursor solutions were mixed with NaOH at room temperature for half an hour in a Teflon vessel. The vessel was then kept at a temperature of 160 °C for 18 h in a steel enclosure. The precipitate resulting from the hydrothermal synthesis was separated by centrifugation and repeatedly washed with double-distilled water. The process was completed by drying the powder at 120 °C and heat treatment at 550 °C for 2 h.

The obtained powders were labelled as SnO₂: Gd 5%-CP and SnO₂: Gd 5%-HT.

2.2. Structural and Morphological Investigations

X-ray diffraction (XRD) was performed with a Bruker D8 Advance X-ray diffractometer (Cu anode and Ni filter, λ = 1.54184 Å) in a Bragg–Brentano configuration at room temperature, in the range of 2θ from 20° to 120°. The XRD data were analysed using MAUD version 2.99 software for determining the lattice parameters and average crystallite size through Rietveld refinement.

Conventional transmission electron microscopy (CTEM), selected area electron diffraction (SAED), high-resolution transmission electron microscopy (HRTEM), and scanning TEM combined with energy-dispersive X-ray spectroscopy (STEM-EDS) were performed using a JEOL JEM-2100 transmission electron microscope. The microscope was operated at 200 kV and equipped with an X-ray detector for EDS investigations.

2.3. Textural Analysis

Textural properties were determined using the ASAP 2020 physisorption analyser (Micromeritics GmbH, Unterschleißheim, Germany). The samples were degassed at 300 °C for 10 h before nitrogen adsorption. The N₂ adsorption–desorption experiments were conducted at 77 K. The BET (Brunauer–Emmett–Teller) model was utilised to calculate the surface area of the obtained samples. Additionally, the BJH (Barrett–Joyner–Halenda) model was employed to determine the average pore size.

2.4. Layer Deposition and Sensing Investigations

SnO₂: Gd 5%-CP and SnO₂: Gd 5%-HT powders were mixed with 1,2 Propanediol and mortared for 15 min to obtain a medium-viscosity paste. This paste was then screen-printed as thick layers onto commercial alumina substrates. The resulting sensors were slowly dried at 60 °C and heat-treated at 500 °C in the air. This process allows the complete removal of the organic solvent, ensuring the layer’s porosity and promoting the adhesion of the layer to the substrate. The substrates used for the sensors are based on planar technology from Innovative Sensor Technology, Switzerland, and are equipped with platinum (Pt) electrodes and a back-side heater, with gold socket contacts. The Pt thickness deposition of both the interdigital electrodes and heater is 5 µm. The interdigital electrodes on the substrate have gaps of 200 μm and 100 μm, respectively. The heater’s electrical resistance is 15 Ohm ± 15% at 25 °C. The chip dimensions are 25 × 4 mm and 0.7 mm thick. The sensitive layer deposited above the interdigital electrodes (marked in yellow in Figure 1a) is 7 µm thick and 3.5 × 7 mm (LxW). The temperature of the sensing layer is controlled by varying the electrical power through the heater, allowing for modulation of the chemical interaction between the MOS layer and the test gases.

To simulate the in-field atmosphere, a Gas Mixing System (GMS) controlled by dedicated Keysight VEE software was employed. The GMS consists of eleven gas channels equipped with Bronkhorst El-Flow mass flow controllers, solenoid valves, and high-purity gas bottles. It operates in a dynamic regime with a total flow rate of 200 mL/min. RH control is achieved through a separate channel with a gas washing bottle filled with wet Chromosorb P-NAW. The electrical resistance changes of the sensors placed into a high-grade PTFE sensor chamber were acquired in real-time using a Keithley 6517A Electrometer (Keithley, OH, USA) (Figure 1a).

3. Results and Discussions

3.1. XRD Characterisation

Powder X-ray diffraction (XRD) patterns of SnO₂: Gd 5%-CP and SnO₂: Gd 5%-HT are depicted in Figure 2a,b. Both diffractograms exhibit the characteristic diffraction pattern of the SnO₂ crystalline phase, with a tetragonal structure and a symmetry space group of P4₂/mnm, according to cif no. 1526637 from the COD database.

In the case of the hydrothermal sample, two additional peaks (noted with * in Figure 2b) can be observed along with the SnO₂ diffraction peaks. The secondary phase was identified by Rietveld refinement analysis as a stannate pyrochlore corresponding to the digadolinium distannate (Gd₂Sn₂O₇) crystalline phase, with a cubic structure and a symmetry space of Fd$\overline{3}$m. Compared with the reference data in the cif used for Rietveld refinement analysis, it was observed that the lattice parameters were not significantly modified in the case of the co-precipitation sample. However, there is a slight variation for the hydrothermal sample, which can be attributed to the Gd doping effect.

The average crystallite sizes were determined to be d = 3.63 ± 0.03 nm for the SnO₂: Gd 5%-CP sample and d = 42.06 ± 0.1 nm for the SnO₂: Gd 5%-HT, as obtained by Rietveld refinement analysis (

Table 2. Lattice parameters and crystallite size—Rietveld refinement analysis.

Crystallographic Parameters/Sample	a = b	c	dhkl (nm)
SnO2: Gd 5%-CP	4.7304 ± 0.0018	3.2338 ± 0.0020	3.63 ± 0.03
SnO2: Gd 5%-HT	4.7457 ± 0.0001	3.1813 ± 0.0001	41.38 ± 0.34

).

3.2. Analytical TEM Characterisation

The low-magnification TEM image obtained for SnO₂: Gd 5%-CP (Figure 3a) reveals a uniform morphology: quasi-spherical nanoparticles with dimensions of approximately 6 nm. On the other hand, SnO₂: Gd 5%-HT is characterized by a non-uniform morphology, with larger nanoparticles of approximately 42 nm, with irregular shapes (Figure 3c). The white arrows from the SAED patterns shown in Figure 3b,d indicate the diffraction rings corresponding to the (110), (101), (211), and (200) lattice planes of tetragonal SnO₂. The patterns consist of concentric diffraction rings, typical for diffraction on polycrystalline materials.

STEM-EDS mapping was performed to observe the spatial distribution of the two elements (Sn and Gd).

Thus, the elemental maps shown in Figure 4 reveal a uniform distribution for SnO₂: Gd 5%-CP (Figure 4a) and an agglomeration tendency for Gd in the SnO₂ matrix for SnO₂: Gd 5%-HT (Figure 4b).

To confirm the Rietveld information regarding the presence of digadolinium distannate (Gd₂Sn₂O₇) as a secondary phase, additional TEM analyses were required. Figure 5a illustrates a TEM image showing a supplementary morphology, consisting in smaller nanoparticles (~7 nm). A further investigation of this area, using SAED (Figure 5b), points towards the (222), (400) and (440) crystallographic planes of Gd₂Sn₂O₇. Elemental STEM-EDS mapping (Figure 5c,d) shows an agglomeration tendency in the area of smaller nanoparticles associated with the Gd₂Sn₂O₇ secondary phase.

High-resolution TEM investigations (HRTEM) offer information regarding the porosity and the shape of the nanoparticles. Pores are visible at different magnifications for SnO₂: Gd 5%-HT (Figure 6c) compared to the sample SnO₂: Gd 5%-CP (Figure 6a).

Regarding the shape, Figure 6b reveals two SnO₂ nanoparticles oriented along the [1-1-1] axis, faceting along the (110) and (101) crystallographic planes for SnO₂: Gd 5%-CP, and Figure 6d shows an elongated SnO₂ nanoparticle oriented along the [1-1-1] axis and well-defined facets along the (110) and (101) planes, for SnO₂: Gd 5%-HT.

In summary, the TEM investigations confirm the Rietveld refinement analysis of the XRD patterns and highlight the main differences between the two samples. SnO₂: Gd 5%-CP exhibits a uniform morphology characterised by ~6 nm quasi-spherical nanoparticles and a uniform distribution of Gd in the SnO₂ matrix. SnO₂: Gd 5%-HT, on the other hand, has larger quasi-spherical and elongated nanoparticles (~42 nm), an uneven distribution of Gd in the Sn matrix, and higher porosity, which is easily visible at different magnifications.

3.3. Textural Characterisation

The specific surface area (SBET) and porosity of SnO₂: Gd 5%-CP and SnO₂: Gd 5%-HT samples were determined using N₂ adsorption/desorption isotherms, confirming the porous nature of the samples (Figure 7a,b). According to the International Union for Pure and Applied Chemistry (IUPAC) classification, these isotherms exhibit type IV behaviour for both samples.

The N₂ adsorption/desorption isotherms of the SnO₂: Gd 5%-CP sample suggest mesoporosity, while those of the SnO₂: Gd 5%-HT sample indicate monolayer adsorption at low pressure and multilayer adsorption at high pressure [25], suggesting macroporosity. This behaviour is consistent with previously reported studies [26]. The fitting curves of the BET surface area of samples CP and HT were linearly correlated (Figure 8a,b), with correlation coefficients of R_CP² = 0.9999 for SnO₂: Gd 5%-CP and R_HT² = 0.9997 for SnO₂: Gd 5%-HT.

The specific surface area (S_BET) was calculated using the BET theory [27]. Thus, S_BET was 91.32 m²·g⁻¹ for sample CP, while for sample HT it was 12.48 m²·g⁻¹. The average pore diameter was determined using the Barrett–Joyner–Halenda (BJH) model. The average pore diameter was 10.93 nm for sample CP and 18.53 for sample HT.

The BET surface area, total pore volumes, average, and BJH pore diameter are summarised in

Table 3. BET surface area, total volumes, average, and BJH pore size.

Sample	BET Surface Area (m2/g)	Total Pore Volume (cm3/g)	Average Pore Diameter (nm)	BJH Pore Diameter (nm)
SnO2: Gd 5%-CP	91.32	0.25	10.93	8.35
SnO2: Gd 5%-HT	12.48	0.058	18.53	13.25

.

Along with the increase in the pore size, the specific surface area decreased in the case of sample HT compared to sample CP, where a decrease in the pore size was observed. This behaviour could be due to the blocking of several micropores as a result of the presence of very small nanoparticles. The results obtained agree with those previously presented by Cai et al. [28]. According to Pandey et al. [29], the size/crystallinity of the particles is inversely proportional to the surface area. The XRD and TEM studies confirm that the particle size is smaller in the case of sample SnO₂: Gd 5%-CP. The broadening of the diffraction maxima in the case of sample CP is due to the smaller size of the particles.

3.4. Sensing Characterisation

Commencing from the widely acknowledged premise that temperature modulates chemical interactions involving charge exchange between the sensitive material and the gases in the ambient atmosphere [30], the first step in material selection was related to the operating temperature. Thus, the SnO₂: Gd 5%-CP and SnO₂: Gd 5%-HT sensors were consecutively exposed to 50%RH, 5000 ppm CO₂, and 5000 ppm CH₄, across a range of temperatures from 200 to 450 °C (Figure 9).

The sensor signals were calculated as the ratio between the initial reference electrical resistance value (measured in dry synthetic air with 5.0 purity grade) and the resistance value after exposure to the test gases. Specifically, for each temperature, water vapours were dosed to achieve an average in-field relative humidity of 50% RH (Figure 9a,b) and the change in resistance in moist air compared to dry air was monitored. CO₂ and CH₄ were also dosed at concentrations specific to their detection limits (Figure 9c,d).

The results highlight the decrease in the influence of water with the increase in temperature, different maximum sensitivity to CH₄ depending on the operating temperature, and better sensitivity to CO₂ for SnO₂: Gd 5%-HT only at 200 °C.

The second step in material selection was related to selectivity. Corroborating the individual dependencies from Figure 9, one can notice that the sensitivity to CH₄ is higher relative to both CO₂ and RH at the operating temperature of 400 °C. Therefore, the simultaneous impact of CH₄ (within a concentration range of 1000 to 5000 ppm) and RH (within 0 to 50%) was monitored for both samples at this temperature (Figure 10).

As can be observed, there are differences between dry and humid air but the RH level (10–50%) does not affect the sensor signals for CH₄. This observation holds significant applicative relevance, as long as in-field humidity varies around the average value of 50%. The power law dependence of the signals on CH₄ concentration reveals varying exponents, indicating a more pronounced tendency toward saturation for SnO₂: Gd 5%-CP. The exponent n is correlated with surface band-bending; the higher the sensor signal for CH₄, the higher the value of n [31]. Although the differences between signals are subtle, they become more pronounced when the curves are extrapolated to the lower explosion limit of 5 vol% in air (equivalent to 50,000 ppm CH₄). It is important to emphasise that, apart from its explosive potential, high methane levels reduce the amount of oxygen in the air, impacting human breathing and heart rate [32,33]. Lastly, methane has a higher heat trapping capacity than CO₂, contributing significantly to a more detrimental greenhouse effect [6,34].

Referring back to Figure 10a,b, the slight differences between the sensor signals can be explained based on the morpho-structural properties induced by the chemical synthesis method. In the case of SnO₂: Gd 5%-CP, the saturation tendency observed during CH₄ exposure might be attributed to the crystallite characteristics. The thick polycrystalline sensing layer consists of successive planes of small crystallites in close contact, obstructing the micropores and hindering gas diffusion. This results in the limitation of chemisorption, a phenomenon known as the Weisz limitation [35,36]. On the other hand, for SnO₂: Gd 5%-HT, the gradual effect of CH₄ is due to the un-localised surface chemisorption [37], facilitated by the irregular shapes (e.g., more surface defects) and the porosity of the crystallites, as revealed by morpho-structural and textural investigations.

To improve the methane sensitivity, the focus was oriented towards the influence of the electrode geometry. According to the literature, the overall electrical resistance and the sensitivity can be tuned by changing the aspect ratio of the interdigital electrodes [38]. In our study, commercial planar substrates with an interdigital electrode gap of 100 μm were used as an alternative to the previous samples obtained using an interdigital electrode gap of 200 μm. The results for both materials, SnO₂: Gd 5%-CP and SnO₂: Gd 5%-HT, are presented in Figure 11a,b.

The differences are obvious; as the interdigitated distance decreases, the signal for CH₄ decreases for SnO₂: Gd 5%-CP and increases for SnO₂: Gd 5%-HT sensors. For simplicity of exposition,

Table 4. Comparison of SnO₂: Gd 5%-based sensors towards CH₄ relative to CO₂ for in-field conditions.

Sensing Material/Electrode Gap	Sensor Signal	Target Gas	Relative Humidity	Temperature
SnO2: Gd 5%-CP/200 μm	3.01/1.14	CH4/CO2	50%	400 °C
SnO2: Gd 5%-CP/100 μm	2.37/1.05
SnO2: Gd 5%-HT/200 μm	3.39/1
SnO2: Gd 5%-HT/100 μm	4.75/1.5

shows the sensor signals for both 5000 ppm CH₄ and 5000 ppm CO₂. For comparison, the presentation is similar to Table 1.

Analysis of these behaviours begins with the assumption of a linear arrangement of grains between the interdigital electrodes. Their gap can be approximated as d ≈ N × g, where N represents the number of grains and g denotes the average grain size.

In the case of SnO₂: Gd 5%-CP, due to the small average crystallite size of ~6 nm and to the operating temperature of 400 °C, a complete grain depletion of free charge carriers can be considered, according to Barsan et al. [31]. The flat band conditions determine the transport of electrons along the surface, whose length is represented by d. Therefore, as d increases, both N and the sensor signal increase similarly. This explains the behaviour observed for SnO₂: Gd 5%-CP (Figure 11a). In terms of transients, this is reflected in short response time (~10 s), and recovery time (~6 min), where the response time is the interval required to reach 90% of the equilibrium resistance value after exposure to CH₄ and the recovery time is the interval required to return to 90% of the resistance value before exposure to CH₄.

In the case of SnO₂: Gd 5%-HT, featuring an average crystallite size of ~42 nm, charge exchange chemical interactions take place on the surface of the grains, resulting in the formation of a double potential barrier between adjacent grains [39]. As a result of the applied electric field, the electrons move from one grain to another, leaping over the intergranular barriers. Subsequent interactions induce changes in the surface energetic band-bending, leading to variations in overall electrical conductance (resistance). Using the Schottky approximation, one can write the overall surface band-bending as:

N_gb × qΔV_s = −k_BT × lnS_N

(1)

where N_gb is the number of grain boundaries (barriers) = N − 1 ≈ d/g − 1; q is the electron charge; V_s is the band-bending changes; k_B is the Boltzmann constant; T is the operating temperature; and S_N is the overall sensor signal defined as R_air/R_CH4. Thus, the smaller d is, the higher the sensor signal is. In other words, the electrical signal is collected more efficiently with smaller-gap electrodes because barrier-limited conduction lessens its impact on the electrical current. This explains the behaviour of SnO₂: Gd 5%-HT (Figure 11b). In terms of transients, this is reflected in longer response times (~40 s) and longer recovery times (~19 min).

The obtained results create the premises for the development of applications, with the miniaturization of the sensor being the future technological objective.

4. Conclusions

The samples SnO₂: Gd 5%-CP and SnO₂: Gd 5%-HT were obtained by alternative synthesis methods, co-precipitation and hydrothermal growth. XRD patterns corresponded to the SnO₂ crystalline phase, with a tetragonal structure. The broadening of the diffraction maxima of sample CP is attributable to the smaller size of the particles.

TEM investigations confirmed the Rietveld refinement analysis of the XRD patterns and highlighted the uniform morphology observed in SnO₂: Gd 5%-CP, characterised by quasi-spherical nanoparticles with a consistent distribution of Gd within the SnO₂ matrix. In contrast, SnO₂: Gd 5%-HT displayed quasi-spherical and elongated nanoparticles, an uneven distribution of Gd within the Sn matrix, and a higher porosity.

Textural analysis associated the reduction in pore size in sample CP with micropore blocking caused by the presence of very small nanoparticles. To highlight the sensitive selectivity, electrical resistance measurements were conducted in the temperature range of 200–450 °C, under conditions of 50% RH, 5000 ppm CO₂, and 5000 ppm CH₄.

Analysis of the sensor signals indicated a temperature of 400 °C for further investigations at different CH₄ concentrations. The distinct exponents observed in the power law dependence explain the more pronounced tendency towards saturation for SnO₂: Gd 5%-CP, following Weisz’s limitation. Conversely, for SnO₂: Gd 5%-HT, the impact of CH₄ is promoted by size, irregular shapes, and porosity of the crystallites.

If the synthesis method does not cause significant changes in the sensor signal, the electrode technology, i.e., the way of reading the sensor signal, does. This was demonstrated using commercial substrates with electrode interdigital gaps of 200 and 100 μm. The different conduction mechanisms, mediated by the crystallite sizes determined by the chemical synthesis method, explained the obvious differences in the sensor signal.