Figure S2 presents the morphologies and size characteristics of the Zr-MOF-808 nanoparticles. The unmodified particles exhibit a well-defined morphology with distinct edges and an average size of approximately 200 nm. Its intrinsic C, O, and Zr characteristic elements are confirmed by the EDS elemental mapping. After modification with GO, high-resolution TEM is employed to clearly observe the morphologies of the composite material, as illustrated in Fig. 3. As shown in Fig. 3a, when Zr-MOF-808 is composited with GO at a 1:1 mass ratio, the host framework remains intact and exhibits decent transparency. Flocculent material, attributable to GO sheets adhering to the surface, is visible at the particle edges. With the increasing mass ratio of GO, more GO sheets accumulate on the edges and surfaces of the Zr-MOF-808. This leads to a decline in the transparency of the nanoparticles, and more C elements are detected surrounding the framework host in EDS mapping. For the [email protected] composite material, GO sheets almost completely encapsulates the Zr-MOF-808 particles, which is confirmed by the further decrease in light transmittance and the significantly increased C and O elements around.
Figure 4 presents the XRD patterns of various samples. The pristine Zr-MOF-808 exhibits three distinct diffraction peaks near 2θ = 10°, being attributed to the (311), (222), and (400) crystal planes of its intrinsic structure. In terms of the composite materials, a new characteristic peak is formed at 2θ = 11.2°, corresponding to the (001) crystal plane of GO, suggesting the successful incorporation of GO35. Notably, the intensity of this peak increases with the GO ratio, which is attributed to the progressive stacking of GO sheets on the Zr-MOF-808 surface, thereby dominating the diffraction signal. Concurrently, the attenuation of the original diffraction peaks between 20° and 30° further supports this analysis, indicating effective coverage by GO. Figure 5 displays the deconvoluted C 1s XPS spectra of various materials in FTIR spectra. For Zr-MOF-808, three characteristic peaks are observed at 284.8 eV, 286.5 eV, and 288.5 eV, corresponding to C-C, C-O, and O = C-O functional groups, respectively. With the introduction of GO, a continuous increase in the peak intensities of C-O and O = C-O is evident. This can be attributed to the epoxy groups present on the basal planes of GO, as well as the carboxyl groups located at the unsaturated sites along its edges. The enhancement in these oxygen-containing functional groups highlights the effective surface modification and chemical interaction between GO and the metal-organic framework.
Figure 6a compares the Raman spectra of various materials. The graphitic peak (G) and disordered peak (D), located at approximately 1580 cm− 1 and 1350 cm− 1 respectively, serve as key indicators for analyzing the structural characteristics in carbon-based materials. These peaks correspond to the in-plane vibration of sp2-hybridized carbon atoms and disorder-induced vibrations arising from structural defects and edges, respectively. The intensity ratio of the D to G (ID/IG) is commonly used to assess the degree of graphitization and structural order within the material: an increase in the ID/IG ratio generally suggests a higher density of defects or enhanced structural disorder, whereas a decrease in this ratio indicates improved graphitization and a more ordered structure. Specifically, the ID/IG ratio of the GO sample is measured at 0.99. In contrast, the composite materials Zr-MOF@GO-1, [email protected], and [email protected] demonstrate ID/IG values of 0.85, 0.79, and 0.97, respectively. The lower ID/IG values observed in all composites compared to pure GO suggest that the introduction of Zr-MOF-808 promotes more ordered stacking and arrangement of the GO sheets. This is primarily attributed to coordination interactions and π-π stacking between GO and Zr-MOF-808, which enhance interlayer adhesion and improve the overall structural regularity. Notably, the ID/IG reaches a minimum value of 0.79 as the mass ratio of Zr-MOF-808 to GO is 1:2 (corresponding to [email protected]), indicating an optimal structural synergy and stacking configuration. After that, as the proportion of GO further increases ([email protected]), the ID/IG value rises to 0.97, implying an excess of GO that could not effectively interact with Zr-MOF-808, leaving some GO domains in a relatively disordered state. This result further confirms that the maximum effective mass ratio for the interaction between Zr-MOF-808 and GO is 1:2.
Figure 7 compares the adsorption performance of three composite materials on four SF6 mixtures containing 10 ppm SO2 or H2S or SO2F2 or CF4, where the variation of electrical resistivity (R/R0) serves as the measurement indicator27. The introduction of GO enhances the electrical conductivity of the composite material, hence reducing the sensitive response to variations in ΔR values. In Zr-MOF@GO-1, The ΔR changes caused by different gases have been reduced to the observable range, but the response rates are still relatively close, making it difficult to effectively distinguish. Notably, a higher GO content ([email protected]) further reduces the response rate but, crucially, enables clear discrimination between gases at 60 seconds. The R/R0 values follow the order: SO2/SF6 < H2S/SF6 < SO2F2/SF6 < CF4/SF6, providing a valuable criterion for classifying discharge types. Nevertheless, an excessive amount of GO ([email protected]) renders the composite overly conductive, resulting in a loss of gas discrimination capability.
To further elucidate the underlying mechanism of the differences in resistivity variations after Zr-MOF@GO adsorbs different gases, the adsorption characteristics and charge transfer processes of various composite materials are systematically explored using DFT analyses. The corresponding composite molecular models of Zr-MOF@GO with CF4, SF6, H2S, SO2, and SO2F2 are constructed, and the calculated results are demonstrated in Figure S3.
Figure 8a displays the differential charge density of these gas adsorption models, where yellow and blue regions represent the tendency of electron loss and gain, respectively. During gas adsorption, Zr atoms tend to lose electrons. Notably, the SO2 and SO2F2 models exhibit significantly larger blue regions, indicating their stronger electron-accepting ability. Based on molecular orbital theory, we further analyzed the lowest unoccupied molecular orbital (LUMO) and the highest occupied molecular orbital (HOMO), as shown in Fig. 8b. The LUMO and HOMO energy levels of the pristine Zr-MOF@GO are − 3.95 eV and − 5.43 eV, respectively, with a band gap of 1.48 eV. After gas adsorption, the band gap of the composite material decreases, facilitating electron transition from HOMO to LUMO and thereby enhancing electrical conductivity. In particular, the band gap of the SO2-adsorbed composite decreases significantly to 1.09 eV, indicating a notable improvement in conductivity. To investigate the mechanism behind the band gap reduction in the Zr-MOF@GO/SO2 system, we further analyzed the density of states (DOS), as shown in Figure S4. The results reveal that the introduction of SO2 leads to the emergence of new orbital occupations near − 4.27 eV after adsorption, while no significant changes are observed at other energy positions. By comparing with the total orbital occupations, the changes near − 4.29 eV, 0.86 eV, and 1.24 eV in the system are attributed to the contribution of SO2 after adsorption. Further analysis of the occupations of the Zr 4d and 5s orbitals in the adsorption substrate and the O 2s and 2p orbitals in the SO2 molecule clearly shows an overlap between the O 2p and Zr 4d orbitals near − 4.29 eV and − 5.57 eV (Figure S4). This overlap suggests the presence of bonded adsorption between Zr-MOF@GO and SO2 gas molecules. Similarly, as shown in Figures S5–S8, bonded adsorption also occurs between Zr-MOF@GO and SF6, H2S, and SO2F2, but no such effect is observed for CF4. In addition, Fig. 8c compares the adsorption energies between Zr-MOF@GO and various gases. The results indicate higher adsorption energies for H2S, SO2, and SO2F2, which may be attributed to their stronger molecular polarity compared to SF6 and CF4, thereby enhancing intermolecular interactions with Zr-MOF@GO. In summary, the Zr-MOF@GO composite exhibits strong bonded adsorption with H2S, SO2F2 and SO2, weak interaction with SF6, and primarily van der Waals adsorption with CF4.
By comparing Fig. 7 and Figure S1, it is observed that the introduction of GO effectively slows down the gas adsorption rate of MOFs. In the MOFs@GO system, distinct resistance variation trends can be clearly observed upon exposure to different gas components. Moreover, as the GO content increases, this attenuation effect becomes more pronounced. Combined with XRD, TEM, Raman, XPS and SEM (Figure S2), it is confirmed that the reduction in adsorption effectiveness results from the wrapping of MOFs by GO, which inhibits MOFs adsorption efficiency. This finding overcomes the limitations of intrinsic MOFs materials in discriminative gas detection and demonstrates the effectiveness of the composite modification of Zr-MOF-808 with GO for the selective detection of SF6 decomposition products.
After gas introduction, the resistance of the MOFs@GO system increases. Using the response curve of MOFs@GO to pure SF6 gas as a control, the change in material resistance is significantly smaller when the gas mixture contains SO2, H2S, or SO2F2 than that of the pure SF6 and CF4-containing systems. Combined with simulation results, it is evident that gas adsorption in MOFs@GO involves both van der Waals interactions and chemical adsorption, with van der Waals forces playing the dominant role. The overall resistivity shows an increasing trend; however, chemical adsorption causes a delay in the resistance rise. Due to the lowest adsorption energy of MOFs@GO for CF4, van der Waals adsorption predominates entirely, which instead enhances the resistance increase after adsorption. In the case of SO2F2, chemical adsorption promotes the charge transfer process, leading to the emergence of a resistivity hysteresis effect. Notably, after SO2 adsorption on Zr-MOF-808, new energy states form near the LUMO level of the composite system, accompanied by molecular orbital overlap, which reduces the bandgap to a minimum of 0.86 eV (Figure S4). This facilitates electron excitation and transfer, resulting in the most pronounced hysteresis effect.
Based on the differences in the hysteresis effect of the increase in resistance of MOFS@GO caused by the adsorption of different gases, it can be effectively used to identify various components of SF6 decomposition products after internal discharge in GIS equipment at room temperature, with a detection accuracy of up to 10 ppm. This makes Zr-MOF-808@GO a potential application in sensing and detection in this field.