1 Introduction

Accurate oxygen measurement and control are essential across various applications, from healthcare to environmental monitoring, transportation, security, medical devices, and industrial process control [1]. Several techniques can be used to identify the dissolved and gaseous O2 levels continuously and correctly. Choosing a method depends on many factors such as sensitivity, accuracy, and specific application in which oxygen detection is needed. Optical O2 sensors provide a robust and versatile solution for measuring oxygen levels across various applications, offering advantages such as high sensitivity, rapid response, long-term stability, reduced cost, adaptation to miniaturization, and ease of production and use [2, 3]. Common optical techniques used for O2 measurement rely on the changes in the luminescent properties of certain materials, such as oxygen-sensitive dyes or phosphors, in response to variations in oxygen concentration. Transition metal complexes of different types of organic dyes such as ruthenium [4, 5], palladium [6], iridium [7], platinum [8], pyrene, osmium [9, 10], and porphyrin [11] are often used as oxygen-sensitive optical sensor agents that operate on the luminescence quenching principle. Among them, tris(2,2-bipyridyl) ruthenium(II) chloride, or [Ru(bpy)3]Cl2, is a fluorescent indicator dye that is sensitive to oxygen and is frequently used for accurate and recent O2 gas monitoring. The good responsiveness of [Ru(bpy)3]Cl2 dye to O2 makes it a popular dye for use in optical-based applications; nevertheless, it has certain disadvantages that include low relative signal variation, lack of stability, and low reproducibility. Table 1 shows comparisons of the sensor characteristics of Ru(bpy)3]Cl2 dye in optical applications for O2 gas in the literature [12,13,14,15,16,17,18]. As seen in the studies, to increase the low relative signal change of [Ru(bpy)3]Cl2 dye, various additives such as metal oxide semiconductors (MOSs) [12], coordination polymers (CPs) [19], ionic liquids (ILs) [13] were added to improve the sensor sensitivity of the dye.

Table 1 Various [Ru(bpy)3]2+ complexes are employed as probes for oxygen sensors
Full size table

The enhanced gas-sensing abilities of MOSs arise from the existence of active sites, cavities, and pores on their surfaces, which serve as active centers for the adsorption of oxygen. n-type and p-type MOSs produce electrical energy when exposed to certain gases, both reducing and oxidizing gases such as O2, CO2, CO, NO2, H2S, and solvent vapors [20,21,22]. It can detect many target gases by changing their conductivity or optical properties [23, 24]. MOSs such as ZnO, CuO, SnO2, NiO, TiO2, Co3O4, Fe2O3, and WO3, when interacted with fluorescent dyes that are oxygen-sensitive can affect the diffusion of oxygen molecules into the matrix of the sensor, improving the performance, stability, selectivity of optical oxygen sensors [25,26,27,28]. Among MOSs-based gas sensors, SnO2 semiconductors stand out as a very promising material due to their affordable price, chemical stability, and sensitivity to a wide range of gases. The addition of metal heteroatoms to SnO2-based semiconductors leads to a change in their size, porosity, and specific surface area, and this change can increase gas diffusion rates by affecting the adsorption sites for gas molecules [28, 29]. In the study conducted by Kim and his colleagues, SnO2 nanowires were produced by adding Sb ions. Replacing Sb5+ ions with Sn4+ ions caused the electron concentration to change and the formation of surface defects. This alteration resulted in notable gas responses, registering values of 81.98 V and 24.99 V for concentrations of 0.05 ppm of NO2 and SO2 gases, respectively [30]. Zhu et al. In their study, they reported that the 0.45% Ho-doped SnO2 sensor gave a rapid response of 2s to 50 ppm TEA at 175 °C [31]. Chen et al. reported in their research that Ni:SnO2 hollow spheres show extraordinary sensitivity to ethanol (C2H5OH) gas, thanks to their porous structure. In this particular configuration, the detection limit against C2H5OH gas was measured to be 1 ppm [32].

The gas sensing characteristics of SnO2 and Ni:SnO2 nanoparticles are related to the processes of electron transfer and oxygen molecule adsorption onto the surface of metal oxide semiconductor structures. Inspired by this, we tried to improve the oxygen sensitivity and sensor properties of the commercial [Ru(bpy)3]2+ dye by adding SnO2 and Ni-doped SnO2 NPs as additions. Measurements of the dye [Ru(bpy)3]2+’s emission, kinetics, and decay time were carried out in the study throughout the concentration range of 0–100% O2, in conjunction with SnO2 and Ni:SnO2 additives immobilized into the PMMA polymeric matrix in thin film and nanofiber forms. SnO2 and Ni:SnO2 additions increased the fluorophore’s O2 sensitivity and produced a linear O2 response for the dye.

2 Materials and methods

2.1 Materials

The metal oxide powders were synthesized using chemicals obtained from Sigma-Aldrich, including tin(II) chloride dihydrate (SnCl2 × 2H2O), sodium hydroxide (NaOH; ≥ 97%), and nickel nitrate hexahydrate (Ni(NO3)2 × 6H2O) without the need for additional purification. Ethanol (EtOH; ≥ 99.8%), sodium hydroxide (NaOH, ≥ 99%), nitric acid (HNO3; 70%), and ammonia (NH3; 25%) solvents were sourced from Alfa Aesar. Additionally, methanol (CH3OH; ≥ 99.8%), acetic acid (CH3COOH, ≥ 98%), and toluene (≥ 99%) solvents were procured from Sigma-Aldrich. On the contrary, the tris(2,2-bipyridyl)ruthenium(II) chloride oxygen-sensitive dye, denoted as [Ru(bpy)3]2+, along with polymethacrylate (PMMA), dioctylphthalate (DOP), the ionic liquid (IL) 1-ethyl-3-methylimidazolium tetrafluoroborate ([EMIM+][BF4−]), and the perfluoro-compound (PFC) nonadecafluorodecanoic acid, were all provided by Sigma Aldrich. DuPont polymers supplied the polyester support (Mylar) utilized in thin-film preparation. Moreover, nitrogen (N2) and oxygen (O2) gas cylinders with a purity of ≥ 99.9% were sourced from Tinsa Gas in Izmir, Turkey.

2.2 Instrumentation

Utilizing an Empyrean X-ray diffractometer (XRD) with CuKα radiation and a scanning rate of 0.01°/min, the SnO2 and Ni-doped SnO2 NPs were exposed to structural analyses. For surface chemistry and elemental content analysis, Thermo Scientific K-Alpha X-ray photoelectron spectroscopy (XPS) was utilized. It uses a 400 nm diameter beam size and a monochromatic Al-Kα X-ray source. Fourier transform infrared spectroscopy (FT-IR, Thermoscientific Nicolet I10) was used to assess the functional groups of powders between 4000 and 600 cm−1. Microstructure imaging was carried out at various magnifications using a Zeiss Sigma 300 VP scanning electron microscope (SEM) to determine morphology. Using the FLSP920 Fluorescent Spectrometer, steady-state photoluminescence (PL) measurements were performed. Using FLSP920’s time-related single photon counting mode (TCSPC), the decay time values were obtained. O2 measurements were made using the Sonimix 7000A gas mixing machine, which blends N2 and O2 gases in a concentration range of 0–100%.

2.3 Synthesis of SnO 2 and Ni:SnO 2 structures

The sol–gel method was used in this work to synthesize powders of SnO2 and Ni:SnO2. Our other earlier study provided a detailed definition of this sol–gel method [33]. 0.63 g SnCl2 × 2H2O was dissolved in 100 ml deionized water and introduced into a 2 M ammonia solution (aqueous) to form SnO2 particles. The final gel underwent drying and subsequent calcination at 650 °C for 2 h, resulting in the formation of SnO2 particles.

SnCl2 × 2H2O was dissolved in deionized water, and the resulting solution was heated to 50 °C for the production of Ni:SnO2 particles. Ni(NO3)2 × 6H2O was added to the heated solution along with 1.9 × 10−2 M NaOH and 10 ml NH3 solution. The atomic percentage of Ni atoms in the SnO2 structure was adjusted and determined to be 0.35%. The Ni(NO3)2 × 6H2O solution was added to the SnO2 mixture at 40 °C while magnetic stirring was in place. The mixture was heated to 80 °C further, forming a blue precipitate. The resulting powders were annealed for 4 h at 650 °C to yield the final gray product.

2.4 Preparation of thin film and nanofiber forms

A polymer-based sensing composite was prepared by combining a polymeric matrix (240 mg) of PMMA, a plasticizer (192 mg) of DOP, an ionic liquid (96 mg) (1-ethyl-3-methylimidazolium tetrafluoroborate ([EMIM+][BF4−]), a perfluoro compound (PFC), and a fluorophore dye (10 mg) in a CH2Cl2:EtOH (25:75) solvent. Metal oxide semiconductors (SnO2 and Ni:SnO2) were added in an amount of 10 mg to improve the O2 sensitivity of the [Ru(bpy)3]2+-based polymeric cocktails. The quantities of additives utilized in the compositions of [Ru(bpy)3]2+-based cocktails are detailed in Table 2.

Table 2 Compositions of composites based on [Ru(bpy)3]2+ with sensitivity to O2 in cocktails
Full size table

For fiber preparation, uniform cocktail solutions were injected using a 10-ml plastic syringe equipped with a metallic needle featuring an inner diameter of 40 mm. The metal point of the syringe’s needle was connected to the high-voltage power supply’s electrode, and it was affixed vertically to a syringe pump. The needle tip and collector were kept 10 cm apart while the polymer solution was dispensed at a rate of 1.0 mL/h under a 25 kV applied voltage. The surface tension of the liquid is overcome by the electric forces when the voltage is high and adequate, which allows the polymer solution to flow in extremely fine fibers. The SEM images of produced nanofibers were shown in Fig. 5a, b inset. The solvent rapidly evaporates and leaves fully coated nanofibers on spotless aluminum foil. For the thin film preparation procedure, the same composites were utilized. Using a knife coating technique, the prepared cocktails were applied on a 125 μm polyester support (Mylar TM type) and then put in a desiccator. A Tencor Alpha Step 500 Profilometer was used to measure the thickness of the thin films that were produced after the composites were made. An average thickness measurement result of 5.46 ± 0.03 µm (n = 5) was obtained. Each detection film was cut to a 1.2 cm thickness and put inside a quartz cuvette.

3 Results and discussions

3.1 Characterization of SnO 2 and Ni:SnO 2 powders

3.1.1 X-ray diffraction (XRD)

Figure 1a illustrates the X-ray diffraction (XRD) patterns corresponding to SnO2 and Ni:SnO2 particles. It is possible to determine the orthorhombic structure of SnO2 in the XRD pattern by consulting the JCPDS 41–1445 card number. Additionally, the acquired data indicate that Ni atoms are absorbed into the SnO2 lattice as expected, as evidenced by the peak shifts resulting from Ni addition to the structure [34].

Fig. 1
figure 1

The characterization of synthesized powders; SnO2 and Ni:SnO2. a XRD, b XPS, c FT-IR, and d SEM (e-e’/f-f’)

Full size image

3.1.2 X-ray photoelectron spectroscopy (XPS)

Here, XPS data were collected to confirm the existence of the pertinent elements in the SnO2 and Ni:SnO2 powders’ structures. The XPS spectra reveal prominent peaks at 486 and 494 eV, indicating the presence of Sn4+ in both SnO2 and Ni:SnO2. Moreover, the identification of the 2 + valence state for Ni in the Ni:SnO2 samples is supported by the existence of signals for Ni 2p3/2 and Ni 2p1/2 at 855 and 874 eV, together with peaks at 861 and 879 eV linked to shake-up satellite structures [35] (see Fig. 1b).

3.1.3 Fourier-transformed infrared spectroscopy (FT-IR)

The broader vibration band between 580 and 680 cm−1 is caused by metal–oxygen (M–O–M) stretching vibrations [36]. When nickel is doped, the relative intensity of the peaks decreases, indicating a reversal in the coordination mode between Ni and SnO2 when compared to pure SnO2 (refer to Fig. 1c) [34].

3.1.4 Scanning electron microscope (SEM)

Observing Fig. 1d (e-e’), rod-like structures are apparent in the morphology of the synthesized SnO2 particles. In Fig. 1d (f-f’), the synthesized Ni:SnO2 powders exhibit high-density particles with aggregate formations resembling round rod-like structures [37, 38].

3.2 Oxygen-induced sensitivity of [Ru(bpy) 3 ] 2+ -based composites

Experiments have been performed to measure the oxygen sensitivity of polymeric sensing slides doped using ruthenium dye in the existence of SnO2 and Ni:SnO2 NPs. The Ru(bpy)3]2+ complex, known for its strong luminescence properties, emits light at 620 nm when excited with light at a wavelength of 468 nm. The response of Ru(bpy)3]2+-based sensing agents to oxygen gas occurs optically through the phenomenon of quenching and is widely used for optical oxygen sensing. Quenching refers to the decrease in the intensity of emitted light, such as fluorescence or phosphorescence, when a quenching agent is present in the luminescence [3]. The quenching agent of the [Ru(bpy)3]2+ fluorophore is oxygen. This phenomenon is used for optical oxygen detection because the intensity of the emitted light is directly related to the ambient oxygen concentration. The luminescence quenching of [Ru(bpy)3]2+ complexes by oxygen occurs through a dynamic quenching mechanism. When the complex is excited by incident light, it reaches an excited state. In the presence of O2 molecules, energy is radiationlessly transferred from the excited [Ru(bpy)3]2+ complex to the ground state of O2 molecules. This energy transfer depletes the excited state of the complex and reduces the intensity of the emitted light, which is observed as a decrease in luminescence.

The relationship between the concentration of O2 and the luminescence intensity of [Ru(bpy)3]2+ is frequently described by the Stern–Volmer equation. The equation relates the Stern Volmer constant (Ksv) to the concentration of oxygen [O2], allowing for the determination of oxygen concentrations in the environment based on changes in luminescence intensity [3] (Eq. (1));

$$\frac{{I}_{0}}{I}=1+{K}_{SV}[{O}_{2}]$$
(1)

Within the scope of this research, intensity-based emission spectra were calculated for [Ru(bpy)3]2+-based complexes in the form of thin film and nanofiber at different oxygen concentrations between 0 and 100% (See Fig. 2a–d). In our previous studies, we reported the signal decrease of the [Ru(bpy)3]2+ dye immobilized in PMMA matrix without additives as 2.05 times [13]. The signal change (I0/I100) of all sensor agents from C1 to C4 in an oxygen-free and fully oxygenated atmosphere was found to be 2.92, 3.59, 3.02, and 4.03, respectively (see Fig. 3 and Table 3).

Fig. 2
figure 2

Emission-excitation spectrum of a C1, b C2, c C3, and d C4 for the concentration range of 0–100% [O2]

Full size image
Fig. 3
figure 3

Comparative calibration curves of C1, C2, C3, and C4 composites for the 0–100% [O2] concentration range

Full size image
Table 3 Optical properties and O2 sensitivity of [Ru(bpy)3]2+-based composites
Full size table

The linearized calibration graphs for thin films and nanofiber composites across the range of 0–100% oxygen concentration are shown in Fig. 3. Correlation coefficients, oxygen sensitivity (I0/I100), and Stern–Volmer constants (KSV) are examples of related characteristics that are included in Table 3. In the concentration range of 0% to 100% [O2], the calibration curve for free [Ru(bpy)3]2+ shows a strong correlation coefficient of 0.9943, indicating a linear correlation represented by the equation y = 0.0119x + 1. When the oxygen-induced reactions of the several composites that were used were compared, C4 showed an especially strong linear response. Equation y = 0.0288x + 1 represents the calibration plot for this composite, resulting in a correlation coefficient (R2) value of 0.9942.

The derived Stern–Volmer constants (KSV) offer valuable insights into the sensitivity of the quenching process. In our previous study, the KSV value for the composite containing free [Ru(bpy)3]2+ was reported as 1.40 × 10–2. The Stern–Volmer values of C1, C2, C3, and C4 determined from the slopes of the calibration graphs were measured as 1.81 × 10–2, 2.36 × 10–2, 1.87 × 10–2 and 2.88 × 10–2, respectively. In particular, it is seen that the KSV value for the C4 composite is approximately two-fold higher compared to the free form. These experimental findings conclusively demonstrate that the C4 nanofiber form exhibits superior oxygen sensitivity. This superiority is ascribed to both a significantly greater relative signal change and higher Stern–Volmer constant (KSV) values in comparison to alternative composite materials.

3.3 Interactions between the [Ru(bpy) 3 ] 2+ and MOSs

The quenching of emission-based intensities of [Ru(bpy)3]2+ in the presence of MOSs additives can be explained by an energy transfer mechanism upon exposure to O2 gas, as well as electron transfer to the semiconductor. These mechanisms collectively result in a decrease in the emission signal of the indicator dye, offering a sensitive response to the presence of oxygen in the environment. In the first mechanism, electron transfer occurs from the excited state of [Ru(bpy)3]2+ to the metal oxide semiconductors (MOSs). In the presence of oxygen, the semiconductor actively interacts with the excited electrons generated by the indicator dye. This interaction results in the transfer of electrons from the dye’s excited state to the semiconductor, causing a reduction in the emission intensity, a phenomenon commonly referred to as quenching. Essentially, electrons from the excited state of [Ru(bpy)3]2+ are transferred to the MOSs, causing a reduction in the emission signal. The second mechanism is a process of energy transfer caused by the overlap of the indicator dye’s emission spectrum and the MOSs excitation band. Energy transfer occurs when the excitation spectrum of one component coincides with the emission spectrum of another. In this instance, the emission spectrum of the indicator dye and the excitation spectrum of MOSs overlap. Upon excitation of MOSs, energy is transferred to the [Ru(bpy)3]2+ dye through this spectral overlap, leading to a decrease in the emission intensity of [Ru(bpy)3]2+ [1].

To elucidate the factors contributing to the augmented oxygen-sensing capability, individual measurements of the excitation and emission spectra for [Ru(bpy)3]2+ and MOSs were conducted. The powders, incorporated into a thin film within a PMMA matrix, were individually subjected to excitation at their respective absorption wavelengths. The superimposed spectra of the normalized emission spectrum of [Ru(bpy)3]2+ and the normalized fluorescence spectra of SnO2 and Ni:SnO2 particles are shown in Fig. 4a and b. With an excitation spanning a broad wavelength range from 300 to 450 nm, the MOSs additives exhibit a wide emission band ranging from 400 to 620 nm. The enhancement of [Ru(bpy)3]2+’s oxygen-sensing ability in the presence of metal oxide particles can be ascribed to possible energy transfer, which is made possible by the effective overlap between the fluorophore’s absorption spectra and the emission spectra of the semiconductors, which serve as light collectors.

Fig. 4
figure 4

The absorption and emission spectra of the PMMA embedded [Ru(bpy)3]2+; a SnO2, and b Ni:SnO2, individually

Full size image

Alongside these transitions, the gas-sensing efficacy of MOSs is intimately associated with both electron transfer processes and the adsorption of oxygen. The presence of active pores and cavities further enhances the capability of sensor by providing ample surface area and active centers for gas adsorption. This combination of factors contributes to the high gas sensing capacity exhibited by MOSs gas sensors. the gas sensing capabilities of MOSs exhibit complexity through a range of processes, encompassing oxygen adsorption, the generation of electronic transitions, oxygen vacancies involving interstitial oxygen antacids and metal ions, as well as direct electron transfer (ET) between the conduction band (CB) and valence band (VB), along with interactions with interstitial metal ions. The formation of defect states and electronic transitions contributes to the sensitivity of MOSs gas sensors to different gases in the environment. The incorporation of p-type NiO into n-type SnO2-rich semiconductors serves various functions, such as functioning as a recombination barrier, establishing a p–n heterojunction for electron transfer, elevating electrical resistance, and generating active holes capable of capturing oxygen. These mechanisms collectively contribute to the improved gas-sensing performance of the semiconductor, especially in terms of oxygen absorption from the environment [39].

3.4 Decay time measurements

To gain a deeper understanding of the improved O2 sensitivity exhibited by the utilized ruthenium dye in the presence of SnO2 and Ni:SnO2 nanoparticles, we performed an evaluation centered on oxygen-induced decay times. This assessment was conducted by examining the response of doped [Ru(bpy)3]2+ in both thin film and nanofiber composite sensors (refer to Table 4). The nanofiber composites, characterized by their highly porous structures and hollow spaces, exhibited notably shorter fluorescence decay times when exposed to O2. In contrast, thin films, with their less porous nature, demonstrated comparatively longer fluorescence decay times due to limited O2 diffusion within their less permeable structures. As depicted in Table 4, the ruthenium dye displayed a multi-exponential decay pattern across all employed matrices. Particularly noteworthy is the reduction in decay time values (τavr: average decay time) for C2 and C4, decreasing from 180 to 151 ns and from 244 to 184 ns, respectively, under fully oxygenated conditions. These findings imply an interpretation in which increased rotational motions of the corresponding ruthenium complex take place, subsequently leading to the relaxation of the excited state. This contributes to the observed reduction in decay times.

Table 4  Decay time values of C1, C2, C3 and C4 sensing agents
Full size table

3.5 Reproducibility of the sensor slides

Response and recovery times are crucial features of optical gas sensors. Photoluminescence-based gas sensing is generally a reversible process, allowing the material to revert to its original photoluminescence state once the gas is eliminated, making it apt for continuous monitoring. Hence, we conducted thorough measurements of response performance and temporal characteristics. Our focus was particularly on the C2 and C4 composites, varying quencher (O2) concentrations. The outcomes of this investigation are depicted in Fig. 5. Based on the results, it is evident that nanofiber-based composites outperformed other sensing formats. Since nanofiber composites are more sensitive to oxygen than thin films, their kinetic changes were examined. When exposed alternately to 0% and 100% O2 atmosphere, both response and recovery times of C2 and C4 composites were determined as 5s 10s and 3s 9s, respectively. The C4 composite yielded 17 cycles over 1800s. In particular, the relative standard deviation for signal levels at both the upper and lower ends of the scale remained consistently below 3% for both C2 and C4. Based on these impressive results, it is clear that sensor fibers containing these MOSs additives show a consistent and reproducible response in the context of oxygen measurements.

Fig. 5
figure 5

Kinetic responses based on the intensity of a C2 and b C4 composites between the oxygen concentrations ranging from 0 to 100%. Inset: The SEM characterization of produced nanofiber forms

Full size image

4 Conclusion

The purpose of this work was to examine how the ruthenium dye’s oxygen sensitivity in PMMA polymeric matrices is enhanced by SnO2 and Ni:SnO2 additions. FT-IR, SEM, XPS, and XRD techniques were used to characterize the produced MOS particles. Agents with O2-sensitive sensors have been produced in both thin-film and nanofiber forms. All of the sensing slides were investigated in terms of oxygen-induced intensity, decay time kinetics, and intensity-based kinetic response for the concentration range of 0–100% [O2]. The results showed that MOS-doped [Ru(bpy)3]2+-based nanofiber composites had higher Stern Volmer (Ksv) values, larger response range, faster response time, and more reproducible measurements compared to thin film forms. When exposed to a fully oxygenated atmosphere, the intensity-based signal drops of the C2 and C4 were measured as 76% and 80%, respectively. These results make [Ru(bpy)3]2+ together with MOSs dopants become promising candidates as oxygen probes.