Introduction

Among perovskite materials, such as SrTiO3 and BaTiO3, a new class of ceramic calcium copper titanate (CaCu3Ti4O12/CCTO) has now been extensively researched due to its unique structure and extraordinary properties.1 Supercapacitors, resistors, varistors, resistive switching, sensors to detect gases, and high-frequency antennas have all benefited from this material's enormous permittivity and electrical properties.2,3,4,5,6,7,8,9,10 Despite their typical dielectric and electrical properties,11,12 they are currently being utilized extensively in the fields of energy storage and energy conversion. In addition, commercial applications such as photocatalysts for water treatment and degradation of industrial effluents13,14,15 have benefited from the bandgap of CCTO in the visible light spectrum. Based on the property and mode of synthesis and manufacturing process, CCTO can be applied in the abovementioned fields. However, several challenges exist in initial synthesis pathways and processing procedures that limit its applicability in the high-frequency range owing to significant dielectric loss.16,17,18 Several researchers have implemented new synthesis procedures and approaches to reduce dielectric loss for its practical implementation.19,20,21,22,23

In addition to intrinsic factors, it has been established that extrinsic factors such as the impact of temperature, time, pressure, and sintering conditions have a major role in creating high-quality, functional materials.24,25,26,27 Microwaves have been used commercially since 1960, and their low energy and processing time have gradually drawn the attention of the industrial world. Microwave sintering has emerged as a cost-effective alternative to conventional kiln firing for treating ceramics that traditionally demand high temperature during operation. This strategy promises to reduce production costs and attain high-purity, high-quality materials.28,29,30 Traditional sintering processes, especially for ceramics, necessitate high operating temperatures and considerable holding time to achieve the material's high density and desired characteristics. Likewise in microwave sintering, the holding time has been reduced due to the microwave's high heating rate and inverted temperature profile.31,32 Dielectric characteristics have been modified by selecting preparatory routes, processing techniques, and selection of feedstock types.

Dopants and inclusion of other elements in the structure of CCTO have further improved the material’s dielectric and electrical characteristics.33,34 This has been evident by the incorporation of rare earth elements into the lattice structure of CCTO, which yields substantial results and desirable features, expanding the material’s utility in a wide variety of real-time contexts.35,36,37 The physical characteristics and dielectric response of gadolinium-doped CCTO of varying composition were studied in this work. The microstructure and phase composition of the material were studied using x-ray diffraction (XRD) patterns, field emission scanning electron microscopy (FE-SEM), and energy-dispersive spectroscopy (EDS). X-ray photoelectron spectroscopy (XPS) was used to assess the valence states and binding energies of Cu and Ti spectra. Nontraditional sintering methods were used for the first time to process CCTO ceramics doped with the rare earth element Gd (microwave sintering). As a result of these experiments, a ceramic was created with much lower dielectric loss than pure CCTO at high frequency, indicating that this material is viable for commercial production.

Experimental

The presence of CaCu3Ti4O12 (400–800 nm, American Elements, Los Angeles, CA, USA, purity grade 99%) was identified as a single cubic pure phase CCTO referencing JCPDS file number 01-070-5808. First, the (CaCu3Ti4O12)x ceramic powder was blended and mixed well with Gd2O3 (500 μm, Sigma Aldrich, purity grade 99.9%) at x = 0.1, 0.2, and 0.3 using an agate mortar and pestle for 1 h. Next, the blended ceramic powder and gadolinium oxide were pressed using a hydraulic press (Techno Search Instruments, Model M-15) in a uniaxial direction at a pressure of 50 MPa. These green compacts or pellets of 15 mm diameter and 5 mm thickness were sintered by applying two different heating modes. The first was conventional sintering in a box furnace (VB Ceramic Consultants, Chennai, India) at 1100°C/5°C min−1 for 12 h. Secondly, the pellets were subjected to microwave irradiation (2.45 GHz) at a rapid heating rate (50°C min−1) processed using a microwave furnace (VB Ceramic Consultants, Chennai, India) at 1100°C/30 min. The schematic representation of the experimental procedure is illustrated in Fig. 1. XRD (Bruker D8 Advance, Germany) was used to investigate the shifting of Bragg’s peak due to the effect of doping with a 2.2 kW Cu anode and Ni filter at a step of 0.05 in a 2θ range of 20°–80°. XPS (ULVAC-PHI, Inc; Model: PHI5000 Version Probe III) with an Al monochromatic radiation source at 280 eV was used to investigate the chemical state of composition due to doping of gadolinium and the quantitative analysis of the elemental composition. Subsequently, FE-SEM, EDS (Thermo Fisher FEI QUANTA 250 FEG) at 20 keV, and a broadband dielectric spectrometer (Concept 80, Novocontrol Technologies, Germany) at room temperature in a frequency range of 10 Hz–20 MHz were used to investigate the effect of the dopant (Gd). To facilitate the investigation of dielectric measurement, electrodes were generated at both the ends of the sintered pellets by coating with a silver adhesive paste with sheet resistance of < 0.25 Ω/sq at 0.001 mm thickness (Thermo Fisher Scientific India Pvt Ltd.) and heat-treated at 150°C for 1 h.

Fig. 1
figure 1

Scheme of experimental procedure from compaction to sintering.

Full size image

Results

Microstructure and Phase Analysis

The XRD pattern of doped samples sintered at 1100°C/12 h in a conventional furnace is depicted in Fig. 2a. Based on the observation, the minor Bragg shift in the position of the peaks was attributed to variance in the ion radius of Ca (0.990 Å) and Gd (0.938 Å). Similarly, the prominent peak (220) shifted to a higher angle at x > 0.1 than pure CCTO for the same sintering conditions. At 33°, a minor trace of gadolinium was identified for conventionally sintered samples. For microwave-sintered (MWS) samples, a minimal shift in Bragg’s angle (2θ) to a lower position at x = 0.1 and a shift to a higher Bragg angle at x > 0.1 were observed. This shifting of Bragg’s peak was attributed to the Gd3+ into Ca2+ sites. In addition, the rapid heating rate and clean manner of heating was observed as a single phase of CCTO in the MWS samples, with no evidence of secondary phases.

Fig. 2
figure 2

X-ray diffraction (XRD) pattern of (a) conventionally sintered and (b) microwave-sintered samples at x = 0.1–0.3 with reference to pure CCTO.

Full size image

The samples were thermally etched in a muffle furnace at 900°C for 15 min before SEM examination. Figure 3a–c shows SEM images of conventionally sintered materials alongside histograms of average grain size, calculated using the ImageJ program and verified with the linear intercept technique. Evidence from the microstructure suggests a notable decrease in average grain size from ~ 20.04 ± 8.43 µm to ~ 17.52 ± 7.77 µm at x ≥ 0.1 but showed a slight reduction in grain size (~ 16.77 ± 7.39 µm) at x = 0.2 as presented in Fig. 3. The microstructure observation revealed that the average grain size inflated at x = 0.1 and attenuated more than pure CCTO (~ 19.90 ± 6.64 µm)38 as the level of Gd content increased at x ≥ 0.2.39 Figure 4 shows that the elemental composition of grains and grain boundaries can be examined by employing EDS. The research showed that no Cu phase was present along the grain boundaries of any of the Gd-doped CCTO samples sintered using a standard procedure. This reveals that the rare earth element (Gd) content in CCTO is the dominant factor in copper segregation along the grain boundaries. The difference in ionic radii between the dopant (Gd) and host species (CCTO) is responsible for the smaller grain size at x > 0.2 compared to pure CCTO.40

Fig. 3
figure 3

SEM images of conventionally sintered samples at 1100°C for a holding time of 12 h at (a) x = 0.1, (b) 0.2, and (c) 0.3 and a heating rate of 5°C min−1.

Full size image
Fig. 4
figure 4

Energy-dispersive spectroscopy (EDS) of conventionally sintered Gd-doped CCTO samples at x = 0.1, 0.2, and 0.3.

Full size image

Figure 5a–c shows that the microstructure of MWS samples is substantially finer and more refined at 1100°C/50°C min−1 for a lower holding time of 30 min. The grain size becomes finer for all Gd-doped samples at x ≥ 0.1. This contrasts with pure CCTO, where the grain size (~ 1.20 ± 0.38 µm)38 was larger than Gd-doped CCTO. It was shown that the grain size decreased from ~ 1.99 ± 0.44 µm to ~ 1.32 ± 0.27 µm as the concentration of the dopant increased. Gd ions act as a drag to prevent the grain boundaries from spreading and coarsening. According to the EDS analysis shown in Fig. 6, no Cu-rich phase was present at the grain boundaries. The prepared ceramic samples are dense, and porosity is restricted.

Fig. 5
figure 5

SEM images of microwave-sintered samples at 1100°C for a holding time of 30 min at (a) x = 0.1, (b) 0.2, and (c) 0.3 and a heating rate of 50°C min−1.

Full size image
Fig. 6
figure 6

Energy-dispersive spectroscopy (EDS) of MWS Gd-doped CCTO samples at x = 0.1, 0.2 and 0.3.

Full size image

X-Ray Photoelectron Spectroscopy

X-ray photoelectron spectroscopy was applied to examine the valence structure of Gd-doped CCTO at x =  0.1–0.3 using different heating modes. The binding energies (BE) of the elemental spectra Ti2p3/2 and Cu2p3/2 were evaluated through the Gaussian–Lorentzian peak fitting function for both CS and MWS samples as illustrated in Figs. 7a–f and 8a–f, respectively. In pure CCTO, due to the charge compensation mechanism, the Cu2+ reduces to Cu+ on heating at a higher temperature (> 1000°C). To compensate for this reduction of Cu2+ to Cu+, the excess Ti4+ occupies the Cu site owing to the reduction reaction of Cu. In Table I, the binding energies of the valence electrons have been presented. The valence states of electrons of CS pure CCTO samples were 457.73 eV and 458.97 eV for Ti3+ and Ti4+, respectively. Meanwhile, for Cu2+ and Cu+, the BE recorded were 934.02 eV and 932.67 eV.41 As the dopant concentration of Gd increased at x = 0.2, the BE of Ti3+ (458.24 eV), Ti4+ (459.45 eV) and Cu+ (932.99 eV) showed a maximum value compared to pure CCTO sintered conventionally. Meanwhile, the BE of Ti and Cu at x = 0.1 were lower at x ≥ 0.2, where Ti3+ (457.46 eV), Ti4+ (458.35 eV), and Cu2+ (935 eV) were considerably higher for Cu than in the case of undoped CCTO. In a similar pattern, the BE of Cu2+ (935.44 eV), Cu+ (933.69 eV) and Ti4+ (458.70 eV), Ti3+ (457.98 eV) were higher at x = 0.3 than pure CCTO. The oxygen deficiency of unbound oxygen drives the formation of Ti3+ and Cu+ that facilitates the charge carrier hopping between Cu2+↔Cu+ and Ti4+↔Ti3+, which results in the conductivity of grains that gives rise to the internal barrier layer capacitance (IBLC) effect42,43,44 within the CCTO lattice.45

Fig. 7
figure 7

XPS spectra of conventionally sintered samples at 1100°C/12 h: (a)–(c) Ti2p and (d)–(f) Cu2p at x = 0.1, 0.2 and 0.3, respectively.

Full size image
Fig. 8
figure 8

XPS spectra of microwave-sintered samples at 1100°C/30 min: (a)–(c) Ti2p and (d)–(f) Cu2p at x = 0.1, 0.2 and 0.3, respectively.

Full size image
Table I Binding energies of Ti and Cu spectra sintered via conventional and microwave sintering at x = 0.1–0.3.
Full size table

Correspondingly, the Cu and Ti spectra of MWS pure CCTO samples shown in Fig. 8a–f were Cu+ (933.47 eV), Cu2+ (934.97 eV) and Ti4+ (458.43 eV), Ti3+ (457.61 eV). The BE of Cu and Ti increase for MWS Gd-doped samples at x ≥ 0.1 compared to pure CCTO; this increase in BE is attributed to the presence of Gd3+ in the CCTO lattice at a different levels of composition. The oxygen vacancy and electron concentration cause the BE of Cu and Ti in the doped samples to increase. The BE of Ti3+ (457.96 eV), Ti4+ (456.49 V) at x = 0.1 was lower than those at x ≥ 0.2, where Ti3+ and Ti4+ were 457.66 eV and 458.61 eV. At x = 0.3, the measured BE was maximum where Ti3+, Ti4+ were 457.90 eV and 458.78 eV, indicating that the Gd3+ ions compensate for the oxygen loss at a higher sintering temperature (1100°C). Likewise, the BE of Cu+ (933.66 eV) was higher at x = 0.1 than x = 0.2 and 0.3, with BE of 933.49 eV and 933.42 eV, respectively. The trend reversed for Cu2+ (934.81 eV), where the BE is elevated as the level of dopant increased at x ≥ 0.2 (934.99 and 934.97 eV). It was inferred that the charge compensation phenomenon precludes the formation of oxygen vacancies due to doping of Gd3+.33,46

Dielectric Characteristics

The dielectric properties including dielectric constant (ɛ) and tanδ are plotted in Fig. 9a–d for CS and MWS Gd-doped samples at x = 0.1 to 0.3, for a frequency from 10 Hz to 20 MHz at room temperature. Silver was coated at both ends of the pellets to create a contact with the surface and act as electrodes for dielectric measurement. Pure CCTO demonstrated a high dielectric constant, but the dielectric loss was significantly reduced by adding the Gd dopant. Maximum dielectric constant ɛ > 104 at 10 Hz and ɛ > 103 at 100 kHz were recorded at x = 0.2. It also showed a low dielectric loss, tanδ of 0.5 at 1 MHz and 0.3 at 100 kHz, while at x = 0.1, the dielectric constant values were ɛ > 103 at 10 Hz and ɛ > 102 at 10 Hz–10 kHz with minimal tanδ of 0.3–0.5 at 1 kHz–1 MHz. Despite the maximum value of ɛ noted at x = 0.2, the dielectric loss was comparably low for Gd-doped samples at x = 0.3, which revealed a tanδ of 0.21 at 100 kHz and 0.42 at 1 MHz.

Fig. 9
figure 9

Dielectric properties of Gd-doped CCTO (a), (c) dielectric constant at x = 0.1–0.3 and (b), (d) tanδ as a function of frequency from 10 Hz to 20 MHz for CS and MWS with reference to pure CCTO.

Full size image

All MWS samples displayed very low tanδ, with the most negligible value of 0.08 at x = 0.2 for a frequency of 100 kHz. Although pure CCTO presented a higher dielectric constant than Gd-doped CCTO, all samples revealed low tanδ for the frequency range from 1 kHz to 20 MHz at room temperature. At low dopant content (x ≤ 0.2), the ɛ > 102 at 10 Hz shifted down as the level of dopant increased. In contrast, the tanδ was very low, presenting a minimum value at x ≥ 0.2 for the frequency range from 1 kHz to 20 MHz. To confirm, the presence of Gd3+ in the lattice structure of CaCu3Ti4O12 prevented the dielectric loss at a high operating frequency range at room temperature.

The relaxation peaks and Cole–Cole plots of CS and MWS samples in Fig. 10a–f demonstrate that the relaxation effect for high frequency was suppressed by the effect of Gd3+, which generated lower dielectric loss than pure CCTO.

Fig. 10
figure 10

Dielectric relaxation peaks and Cole–Cole plots of (a)– (c) conventionally sintered and (d)– (f) MWS Gd-doped samples as a function of frequency at room temperature.

Full size image

Discussion

Finally, the XRD patterns of the CS and MWS samples showed that CCTO sintered in a microwave revealed a single phase, while CCTO sintered in a conventional oven revealed secondary phases due to the high-temperature sintering and the addition of dopant. Grain size decreased for both conventionally sintered and MWS Gd-doped CCTO samples as dopant concentration increased, as shown by microstructural analysis. This was caused by the restriction of grain boundary movement caused by the dragging action of the dopant (Gd3+). EDS studies later confirmed the absence of a Cu-rich phase along the grain boundaries. The reduction in oxygen vacancies by adding Gd3+ in the CCTO lattice resulted in the variation in the BE of Cu and Ti spectra. Findings from the dielectric studies revealed a lower dielectric constant than pure CCTO and a shallow dielectric loss at a high operating frequency (100 kHz–20 MHz). Evidence for this includes the planar defect promoted by the presence of twinning parallel to planes (100), (001), and (010)18,47 that resulted in distortion of (TiO6)−4 due to the substitution of Ca with Gd. This modification in the structure leads to a change in dielectric constant and tanδ.48,49 It is important to note that the dielectric characteristics of a material, such as its dielectric constant and dielectric relaxation, are influenced in part by the nature and contact of electrodes over the surface.50,51 In addition, the selection of suitable dopant elements impacts the dielectric characteristics of the material.52

Conclusion

The addition of dopant (Gd) to the CaCu3Ti4O12 (CCTO) structure at x =  0.1–0.3 processed using two different heating modes at 1100°C imparted significant changes in the dielectric response compared to pure CCTO. Based on the results acquired utilizing various characterization tools, the following conclusions were drawn:

  1. 1.

    XRD patterns confirmed the presence of a single pure phase CCTO for MWS Gd-doped samples. A minor shift in Bragg’s angle (2θ) to a lower position was observed at x > 0.1 for conventionally sintered Gd-doped CCTO samples at 1100°C/12 h due to the inclusion of Gd3+ into the lattice.

  2. 2.

    The microstructural analysis revealed grain size and morphology where the grain size decreased from 20.04 ± 8.43 µm to ~ 17.52 ± 7.77 µm at x ≥ 0.1 but showed a slight reduction in grain size (~ 16.77 ± 7.39 µm) at x = 0.2 for CS samples. As the level of dopant concentration increased, the grain size of MWS samples decreased from ~ 1.99 ± 0.44 µm to ~ 1.32 ± 0.27 µm owing to the inhibition of the grain boundary mobility.

  3. 3.

    XPS studies showed the charge carrier hopping and charge compensation mechanism of Cu and Ti spectra that holds valid for the IBLC effect within the CCTO structure.

  4. 4.

    The dielectric response recorded at room temperature for the frequency range from 10 Hz to 20 MHz showed a lower dielectric constant than pure CCTO, which was related to the ionic radius of gadolinium that reduces the segregation of the Cu-rich phase along the grain boundaries. However, a very minimal tanδ was observed for all MWS samples at x ≥ 0.1, and the dielectric constant was > 103 at 10 Hz, along with a minimal tanδ of 0.3–0.5 at 1 kHz–1 MHz at x = 0.1 for CS Gd-doped samples, which was facilitated by the resistance offered by the grain boundary.