Introduction

Currently, inorganic thermoelectric ink (TE) materials are widely used for their outstanding energy conversion capability, especially in thin-film or thick-film forms.1 Several methods for the production of metal chalcogenide films have been developed.2 Among these, the chalcogennidometalat (ChaM) method is known for its rich coordination chemistry, superior semiconducting behavior, and strong binding affinity to heavy metal ions.3 The nanoscale semiconductor particles obtained using ChaM ions as precursors are widely utilized as inorganic ligands and solders. Molecular antimony telluride (Sb2Te3)-based ChaMs are used as a solder or a sintering aid for TE particles. Sb2Te3-ChaM easily fills the voids and interfaces between these TE particles.4 The performance of TE materials is related to a parameter called the dimensionless figure-of-merit: \(ZT = S^{2} T/\rho k = \left( {{\text{PF}}} \right)T/k\) where the Seebeck coefficient (S), electrical resistivity (ρ), thermal conductivity (\(\kappa\)), and absolute temperature (T).5,6,7 S2 are together typically called the power factor (PF).8 Therefore, TE materials should simultaneously possess a high Seebeck coefficient and low electrical resistivity.9 Among the numerous TE materials that have been developed10 , Sb2Te3-based compounds are predominantly used as p-type TE materials for applications at room temperature because of their excellent TE properties in those ranges.11 Sb2Te3-based TE films have been fabricated using different processes, such as printing,12 spin-coating,13 or screen-printing techniques.14 The screen-printing technique was introduced for the fabrication of TE films because of its simplicity in processing and ease in thickness control.15 Hyeongdo et al. reported the TE properties of screen-printed Bi0.5Sb1.5Te3 and Bi2Te2.7Se0.3 thick films using a post-annealing process with mechanical pressure. The Bi0.5Sb1.5Te3 thick film had a ZT of 0.89 at room temperature.16 Seong et al. reported a TE generator fabricated by multimaterial 3D printing of composition-segmented BiSbTe materials, by sequential deposition of all-inorganic viscoelastic TE inks. The maximum ZT of 3D-printed materials is 0.9 at 125°C.17 Sung et al. reported high-performance shape-engineerable TE painting using the Sb2Te3 ChaM sintering process, effective even at 350 °C, with a ZT value of 1.21.18 Nan et al. reported the use of size-controlled synthesis and transport properties of Sb2Te3 nanoplates to control the nanoparticle size, and the thickness of the nanoplates was reduced to 80 nm. The high ZT was 0.3.19

In this work, we have used Sb2Te3 ink for printing to fabricate Sb2Te3 films on a SiO2/Si substrate via the screen-printing technique. The screen-printed film thickness was controlled to 200 µm. The crystal structure, crystallite size, dislocation density, lattice strain, and particle size were studied based on x-ray diffraction (XRD; JSM-7610Fplus) results. The morphology, powder size, and chemical composition of the fabricated films were studied using field-emission scanning electron microscopy (FE-SEM; XRD6100; Shimadzu) and energy-dispersive x-ray spectroscopy (EDS), and the TE properties of the fabricated Sb2Te3 films after annealing were investigated and are discussed.

Materials and Methods

Sb2Te3 Ink Synthesis

Elemental granules of antimony (Sb), tellurium (Te), and Sb2Te3 (99.99% purity) were mixed without further purification. Ethylenediamine (>98%), acetonitrile (99.8%), glycerol (99.5%), and ethylene glycol (99.5%) were also used without further purification. Then, the stoichiometric Sb2Te3 powder was prepared by a planetary ball mill (PM 400) with a rotation speed of 300 rpm for 10, 20, or 30 h to produce small, medium, or large particles, respectively. The resulting powders were then sieved through a shaker for 30 min (with a vibratory sieve shaker; 200 basic). The formation of the Sb2Te3 powders was confirmed by XRD analysis, as shown in Fig. 1.

Fig. 1
figure 1

XRD patterns of ball-milled Sb2Te3 powder at 10, 20, and 30 h.

Sb2Te3 ink was prepared by first dissolving 0.32 g of Sb powder and 0.68 g of Te powder in 10 ml of ethylenediamine at room temperature in a fume hood 20,21 and then dissolving it after stirring for 12 h. The solution turned a light purple color. After 40 ml of acetonitrile was added to the mixed solution, the mixture was centrifuged at 3000 rpm for 2 h, after which the precipitated solution was added to a mixed solvent containing 3.6 g of glycerol and 0.4 g of ethylene glycol and stirred for 2 h. A mixture of 4 g of the Sb2Te3 powder was heated at 453 K for 5 h on a hot plate to thoroughly homogenize the ink.

Screen-Printing Processing

The Sb2Te3-ink was screen-printed on a SiO2/Si substrate with a flat painting brush using a screening machine (ZY-TB-B Coater), and the thickness was controlled to 200 µm.22 The printed substrates were sequentially dried on a hot plate at 433 K for 30 min and then annealed under vacuum at 523 K (heating step of 373 K min–1) for 30 min (Mini Lamp Annealer Mila5000; Advanced Riko). A schematic of all the processes is shown in Fig. 2.

Fig. 2
figure 2

Schematic of the screen-printing and thermal-annealing processes.

Characterization and Measurement

The microstructure of the fabricated films was characterized by FE-SEM operated at 15 kV. XRD patterns were collected using a Cu-Ka1 x-ray source (wavelength of 1.5418 Å) operating at 40 kV and 40 mA in the range of 20–70° with a scan speed of 5° min–1 and a step size of 0.01°. The crystallite size (D) of the Sb2Te3 films was evaluated by Scherrer’s equation:

$$ D = \frac{K\lambda }{{\beta \cos \theta }} $$
(1)

where K is a dimensionless shape factor (0.9), λ is the wavelength of the CuKα1 radiation (1.54186 Å), β is the full width at half-maximum (FWHM) of the diffraction peak, and θ is the Bragg's angle. The dislocation density (δ) and the lattice strain (ε) in the c-axis were calculated using23:

$$ \delta = \frac{1}{{D^{2} }} $$
(2)
$$ \varepsilon = \frac{\beta }{4\tan \theta } $$
(3)

Finally, S and ρ were measured at temperatures ranging from 325 K to 525 K using a four-probe method (ZEM-3; Advance Riko). The average size of the films was 3.0 × 15 mm2 on a SiO2/Si substrate, as described previously.24 The thermal diffusivity D was measured by laser flash analysis (LFA-457; Netzsch, Germany). The κ was calculated from:

$$ \kappa = DdC_{P} $$
(4)

The sample density d was determined by the Archimedes method, and the specific heat Cp was measured using a differential scanning calorimeter and the three-line method. The effective average particle size of the Sb2Te3 films was calculated using25:

$$ D_{{{\text{eff}}}} = 6d_{{{\text{th}}}}^{ - 1} A^{ - 1} $$
(5)

where dth  = 6.50 g cm–3 and is the theoretical density of Sb2Te3 and A is the specific surface area.

Results and Discussion

The XRD patterns of the films of small, medium, and large Sb2Te3 powders annealed under vacuum at 523 K for 30 min are shown in Fig. 3a. All the XRD pattern peaks correspond to the Sb2Te3 crystal structure. Therefore, the main peak of the (015) phase at 28° in Fig. 3b corresponds to JCPDS#15-0874 and the rhombohedral structure: a = b 0.4487, and c = 3.0873 nm.26,27 However, the 2θ angle at 28° continuously shifted toward the small-angle side, and the large-angle side showed a highly crystalline structure. The crystallite size, dislocation density, and lattice strain are shown in Table I. The crystallite size of all the films at the (015) plane decreased with increasing particle size.28 Moreover, the dislocation density and lattice strain increased with increasing particle size, as confirmed by the XRD and SEM results, and these values are affected by the TE PF.29 In larger particles, carrier-scattering mechanisms, such as impurity scattering, grain boundary scattering, and defect scattering, can be more prominent. These scattering events hinder the movement of the charge carriers, reducing the electrical conductivity and ultimately lowering the PF.

Fig. 3
figure 3

XRD patterns of (a) all the peaks and (b) the expansion of the (015) phase for small, medium, and large films.

Table I Variation in the crystalline size, dislocation density, lattice strain, and effective average particle size of the films for the different Sb2Te3 powder sizes

Figure 4 shows the SEM images and average particle sizes of the Sb2Te3 powders prepared for different durations. The average size of the particles for each sample was calculated by the log-normal distribution function30:

$$ f(D) = \left( {\frac{1}{{\sqrt {2\pi \sigma_{D} } }}} \right)\exp \left[ { - \frac{{\ln^{2} \left( {\frac{D}{{D_{0} }}} \right)}}{{2\sigma^{2} }}} \right] $$
(6)

where D corresponds to the average powder size and \(\sigma_{D}\) is the standard deviation. The average powder sizes for 10, 20, and 30 h of Sb2Te3 powder were 3.54, 2.88, and 2.36 µm, respectively, along with standard deviations of 0.30µm, 0.25µm, and 0.23 µm, respectively. The average size of the Sb2Te3 powder decreased after different curing times, as confirmed by the average sizes shown in Fig. 4b, d, and f.

Fig. 4
figure 4

SEM images (a, c, e) and powder size distributions (b, d, f) of the Sb2Te3 powder. Samples were prepared at (a, b) 10, (c, d) 20, and (e, f) 30 h.

SEM images of the morphology of the top views and cross-sections of the screen-printed Sb2Te3 films of (a, c) small size, (b, e) medium size, and (c, f) large size are shown in Fig. 5. The final thicknesses of the annealed films were 34 µm (a), 91 µm (b), and 184 µm (c). The annealed films were agglomerated particles due to ∆G = ∆HTB, where ∆G is the change in the free enthalpy, ∆H is the change in the enthalpy, and B is the Boltzmann entropy.31 The surface morphology clearly showed the densification of the films during the annealing process. Moreover, the densification of Sb2Te3 films increased the B value depending on the temperature.32 However, the particle size of the Sb2Te3 material and the chemical composition and particle size of the Sb2Te3 films are predictable PFs. In nanoscale TE materials, when the particle size is comparable to the characteristic length scale of the electrons, a phenomenon called quantum confinement occurs, which leads to discrete energy levels, affecting the electronic structure and transport properties of the material. As a result, the PF can be enhanced due to the increased carrier concentration and improved electrical conductivity.

Fig. 5
figure 5

SEM images of the surfaces and cross-sections of the (a, d) small, (b, e) medium, and (c, f) large Sb2Te3 powders.

The chemical composition of the Sb2Te3 films was confirmed to be Sb and Te by EDS, and the elemental mappings are shown in Fig. 6. The atomic ratios of the Sb:Te films were 34:65, 35:65, and 35:65, as shown in Table II. Therefore, the small powder Sb2Te3 films have a good ratio of 2:3, because this ratio results in the highest PF. The significant peaks include Sb and Te, indicating the composition of the films, and Si and O are from the SiO2/Si-wafer substrate.33

Fig. 6
figure 6

EDS spectra and surface maps of (a) small, (b) medium, and (c) large-size Sb2Te3 films.

Table II Atomic ratio and thickness of films of small, medium, and large Sb2Te3 powders measured by EDS

The S, ρ, and PF of films of small, medium, and large sizes of the Sb2Te3 powders depend on temperature, as shown in Fig. 7. The obtained ρ values of the films for small, medium, and large films are 0.26, 0.29, and 0.23 m Ωm, respectively, at 515 K. The ρ of all the films decreased with increasing temperature, as shown in Fig. 7a. Moreover, the ρ of all the films decreased with increasing crystallite size, which indicates semiconductor behavior.34,35 In fact, the ρ of the semiconductor can be determined by:

$$ \rho = \frac{1}{ne\mu } $$
(7)

where n is the carrier concentration, e is the charging unit, and µ is the mobility in carrier concentration.36 Therefore, the small films had the lowest dislocation density and lattice strain.

Fig. 7
figure 7

Relationships of temperature with respect to (a) ρ, (b) S, (c) κ, (d) PF, and (e) ZT for small, medium, and large films.

The S of all the films increased with increasing temperature from 325 K to 425 K and decreased with increasing temperature from 425 K to 525 K, as shown in Fig. 7b. The increased S values are due to decreasing carrier concentration, and the decreased values are due to increasing carrier concentration,37 indicating a change in the TE type. However, all S values were positive, indicating p-type TE behavior, which decreased with increasing particle size. In fact, the variation in S values as a function of the Sb2Te3 film size is inversely related to the carrier concentration:

$$ S(T) = \frac{{8\pi^{2} k_{{\text{B}}}^{2} }}{{3eh^{2} }}m^{*} T\left( {\frac{\pi }{3n}} \right)^{\frac{2}{3}} $$
(8)

where kB is Boltzmann’s constant, h is Planck’s constant, m* is the effective mass of the charge carrier, and n is the carrier concentration.38 The S increased with decreasing powder size due to the decrease in carrier concentration. It can be assumed that small grains have a greater density of grain boundaries than medium and large grains, which exhibit high S values together with low ρ values, which was confirmed by the SEM images of the surface film.39 The PF of the small films increases with increasing temperature from 325 K to 425 K and decreases with increasing temperature from 425 K to 525 K, as shown in Fig. 7c. However, the PF of the film increased with increasing temperature from 325 K to 525 K. Therefore, the maximum PF was 11.6 × 10–4 W m–1 K–2 at 425 K. It is important to note that there is an optimal particle size range for maximizing the PF. A too-small particle size may lead to increased phonon scattering, which can detrimentally affect the electrical conductivity. Conversely, excessively large particle sizes may result in increased carrier scattering and reduced PF. Compared with those of bulk TE materials, the κ of films of small, medium, and large sizes of Sb2Te3 powders are significantly lower.40 These κ values are 0.78  W m–1 K–2, 0.93  W m–1 K–2, and 1 W m–1 K–2 at 515 K for small, medium, and large particles, respectively, as shown in Fig. 7d. Consequently, the maximum ZT value was achieved for films of small Sb2Te3 powders, which were 1.04 at 468 K, as shown in Fig. 7e.

Conclusions

Sb2Te3 films on SiO2/Si-wafer substrates were fabricated from different particle sizes (small, medium, and large) of Sb2Te3 powder mixed with ChaM solution for the screen-printing technique, with a fixed thickness of 200 µm and annealing at 523 K under a vacuum. All the films exhibited rhombohedral crystal structures and increased crystallite sizes with increasing particle size after annealing. The effective average particle size of the Sb2Te3 films increased with increasing powder size. When ρ decreased with increasing temperature, S increased and the PF increased. The maximum ZT was 1.04 at 468 K for the small-size Sb2Te3 film. Finding the right balance is crucial for optimizing the thermoelectric performance. In some cases, S can increase due to enhanced energy filtering and quantum confinement effects, resulting in improved thermoelectric performance.