Introduction

Amorphous oxide semiconductors (AOSs) are attractive materials for applications in optoelectronics, thermoelectronics, and organic photovoltaics owing to their excellent properties, such as their high transparency, low deposition temperature, and high carrier mobility1,2,3. Several AOSs have been extensively studied, including zinc tin oxide (ZTO)4, indium zinc oxide (IZO)5, and indium gallium zinc oxide (a-IGZO)6. However, these AOSs have drawbacks, such as a high annealing temperature7,8, high off‒current9, and relatively low carrier mobility (~ 10 cm2/V-s)10,11.

In addition to the aforementioned AOS materials, amorphous indium tin zinc oxide (a-ITZO) has attracted considerable attention owing to its advantageous characteristics, including large carrier mobility and high carrier concentrations12,13. The primary factor contributing to the high electron carrier mobility within the conduction‒band minimum of a-ITZO is the increasing overlap area between the orbitals of In 5 s and Sn 5 s. These orbitals possess strong divergence, high symmetry, and an electronic configuration similar to that of (n‒1)d10n0 (n ≥ 4)14,15,16. In addition to the improved carrier mobility, the carrier concentration of a-ITZO is increased. This is achieved by substituting lattice In3+ ions with Zn2+ and Sn4+ ions, which form acceptor defects of \({\left({{{\text{Zn}}}^{2+}}_{{{\text{In}}}^{3+}}\right)}^{\mathrm{^{\prime}}}\) and donor defects of \({\left({{{\text{Sn}}}^{4+}}_{{{\text{In}}}^{3+}}\right)}^{\bullet }\). The presence of Zn2+ induces lattice distortion through the Jahn‒Teller effect, leading to the formation of oxygen vacancies (\({\left({{{\text{V}}}^{0}}_{{{\text{O}}}^{2-}}\right)}^{\bullet \bullet }\)) as donor defects. Consequently, the carrier concentration in a-ITZO increases11. Additionally, the carrier mobility of a-ITZO is approximately 20‒30 cm2/V-s13,17,18,19,20,21,22. Improvements in the electrical properties of a-ITZO are necessary for optimizing the electrical performance16. The focus should be on enhancing the carrier mobility rather than solely increasing the carrier concentration11. Dopants have been proven to be effective for enhancing the electrical properties of AOS materials. Several types of doped ITZO have been extensively investigated, including Mg-17, Al-23, Er-24, P-1, W-25, and Pr-doped ITZO26. Heterostructures combining transparent conducting oxides with metals provide another approach for improving the electrical performance. Examples include indium tin oxide/silver/indium tin oxide (ITO/Ag/ITO)27, aluminum-doped zinc oxide/silver/aluminum-doped zinc oxide (AZO/Ag/AZO)28, zinc-tin-oxide/silver/indium-tin-oxide (ZTO/Ag/ITO)29, gallium zinc oxide/silver/gallium zinc oxide (GZO/Ag/GZO)30, and indium zinc oxide/gold/indium zinc oxide (IZO/Au/IZO) heterostructures31.

Rhombohedral bismuth selenide (Bi2Se3) is a direct n-type topological insulator with a narrow band‒gap of 0.35 eV32. The bulk structure of Bi2Se3 consists of five stacked atomic layers, i.e., \({{\text{Se}}}_{1}-{{\text{Bi}}}_{1}-{{\text{Se}}}_{1}^{\mathrm{^{\prime}}}-{{\text{Bi}}}_{1}-{{\text{Se}}}_{1}\), and is referred to as a quintuple layer (QL)33. Within the QLs, covalent bonds between Se and Bi dominate, whereas van der Waals (vdW) forces govern the bonding between QLs34. Bi2Se3 is a unique material because of its insulating bulk state and gapless conducting surface state, which are attributed to the spin‒orbital coupling (SOC) and time‒reversal symmetry (TRS)35,36. These properties prevent the surface backscattering effect caused by non-magnetic impurities, resulting in efficient electron transport at the surface37,38. Bi2Se3 exhibits a high electron carrier mobility of up to 600 cm2 V−1 s−139. Consequently, Bi2Se3, with its gapless conducting surface state, has several notable features, including (a) photon-like and spin-polarized electrons, (b) a low power dissipation rate, and (c) the quantum spin Hall effect34,39,40,41.

In this study, Bi2Se3 NPs were fabricated on a glass substrate via thermal chemical vapor deposition (CVD), followed by the deposition of a-ITZO via magnetron sputtering. Subsequently, the electrical properties of the ITZO/Bi2Se3 NPs, including the bulk carrier concentration, carrier mobility, resistivity, and conductivity were analyzed via Hall effect measurements at room-temperature.

Results and discussions

Crystal structures

Figure 1a,b show the XRD patterns of the glass substrate, ITZO thin films, and Bi2Se3 and ITZO/Bi2Se3 NPs before and after annealing at 250 °C. As shown in Fig. 1a, the glass substrate exhibited a broad peak at approximately 24.5°42, whereas the ITZO thin film exhibited a broad hump centered at approximately 31.58°, indicating an amorphous structure43. The dITZO-spacing at 2θ = 31.58° was approximately 0.285 nm estimated using Bragg’s law44,45. As shown in Fig. 1a, the ITZO/Bi2Se3 NPs exhibited six significant peaks at 2θ = 24.88°, 29.21°, 32.81°, 40.06°, 43.42°, and 47.56°, which corresponded to the Bi2Se3(101), Bi2Se3(015), ITZO, Bi2Se3(1010), Bi2Se3(110), and Bi2Se3(0015) planes32,43, respectively. After the annealing treatment at 250 °C, as shown in Fig. 1b, the ITZO thin film and ITZO/Bi2Se3 NPs exhibited broad peaks at approximately 31.87° and 32.7°, respectively, indicating an amorphous ITZO phase. The six significant peaks in Fig. 1b are similar to those in Fig. 1a, and correspond to the relevant Bi2Se3 and ITZO crystal planes. These results confirm that the Bi2Se3 phase was stable under the annealing at 250 °C.

Figure 1
figure 1

XRD patterns of pure glass, ITZO thin film, and ITZO/Bi2Se3 NPs (a) before and (b) after annealing at 250 °C.

Full size image

Fine structures

Figure 2a shows a low-magnitude TEM image of a 270‒nm‒thick ITZO thin film deposited on a pure glass substrate after annealing at 250 °C. The inset presents the HRTEM-selected-area electron diffraction (SAED) pattern of ITZO, indicating an amorphous structure. The SAED pattern exhibits two significant rings with inner and outer d-spacings of 0.283 and 0.165 nm, respectively. The ITZO used in this study was composed of 85 wt% In2O3, 10 wt% SnO2, and 5 wt% ZnO, indicating that the In2O3 was the host material. Sn4+ and Zn2+ ions prefer to replace In3+ ions at the b- and d-sites in the In2O3 lattice11. Hence, the dITZO-spacings of 0.283 and 0.165 nm were related to the In2O3(321) and In2O3(611) planes (JCPDS 71-2195), respectively. The former result is agreed with the XRD results as shown in Fig. 1. Figure 2b shows an HRTEM image of ITZO indicating a disordered lattice with estimated dITZO-spacings of 0.111 and 0.109 nm, which correspond to the In2O3(833) and In2O3(248) planes, respectively, according to JCPDS 71–2195. The HRTEM-energy-dispersive X-ray spectroscopy (EDS) spectrum shown in Fig. S1a confirms the presence of In, Sn, and Zn. Figure 2c presents a low-magnitude TEM image of hexagonal-shaped Bi2Se3 NPs after annealing at 250°C. The HRTEM-SAED pattern shown in Fig. 2c exhibits peaks corresponding to the Bi2Se3(1211) and Bi2Se3(110) planes. The HRTEM image of the Bi2Se3 NPs in Fig. 2d shows d-spacings of 0.211 and 0.208 nm, corresponding to the Bi2Se3(0111) and Bi2Se3(110) planes, respectively. Figure 2e shows a low-magnitude TEM image of the ITZO/Bi2Se3 NPs after annealing at 250°C. The HRTEM-SAED pattern shown in the inset of Fig. 2e exhibits peaks corresponding to the Bi2Se3 planes of (024), (012), and (0111). The HRTEM image presented in Fig. 2f shows a d-spacing of 0.319 nm, corresponding to the Bi2Se3(009) plane. In the zoomed-in HRTEM image shown in Fig. 2g, the d-spacing of dITZO was estimated to be 0.111 nm, corresponding to the In2O3(833) plane. The presence In, Sn, Zn, Bi, and Se was confirmed using HRTEM-EDS, as shown in Fig. S1b. These results indicated that the ITZO was deposited on the surface of the Bi2Se3 NPs, and formed ITZO/Bi2Se3 heterostructures.

Figure 2
figure 2

Low-magnitude and HRTEM iamges of (a), (b) ITZO thin film, (c,d) Bi2Se3 NPs, and (e,f) ITZO/Bi2Se3 NPs, respectively. The insets are the HRTEM-SAD patterns. (g) the zoomed-in image of (f).

Full size image

Surface morphologies

Figure 3a–d show FESEM images of Bi2Se3 and ITZO/Bi2Se3 NPs before and after annealing at 250 °C. They exhibit a hexagonal shape, as observed in the HRTEM images of Fig. 2c,d. The average thicknesses of these NPs were estimated using the ImageJ software, and the results indicated that each Bi2Se3 NPs had a thicknesses of approximately 71.8 nm (divided by 20 pieces) and 86.8 nm (divided by 20 pieces) before and after annealing at 250 °C, respectively. The thickness of each QL was approximately 0.955 nm46. Thus, the average number of QLs was 83 for each pristine Bi2Se3 NPs. On average, each ITZO/Bi2Se3 NPs had thicknesses of 247.3 and 232.5 nm before and after annealing at 250 °C, respectively. These results explain why the ITZO/Bi2Se3 NPs shown in Fig. 2e is not transparent. The thickness of the covered ITZO thin films was estimated to be approximately 160 nm. Photographs of the Bi2Se3 and ITZO/Bi2Se3 NPs before annealing at 250 °C are presented in the insets of Fig. 3a,b. Their colors differed significantly; the Bi2Se3 NPs were grey, whereas the ITZO/Bi2Se3 NPs were yellowish-green. The cross-sectional SEM images in the insets of Fig. 3c,d indicate the total deposition thicknesses of the Bi2Se3 and ITZO/Bi2Se3 NPs after annealing at 250 °C, which were approximately 2.2 and 2.5 μm, respectively. The SEM–EDS results for the Bi2Se3 and ITZO/Bi2Se3 NPs are presented in Fig. S2a,b, which confirm the existence of Bi, Se, In, Sn, and Zn.

Figure 3
figure 3

FESEM images of the Bi2Se3 and ITZO/Bi2Se3 NPs of (a,b) before, and (c,d) after annealing at 250 °C, respectively. The insets in (a,b) show the photographs of the Bi2Se3 and ITZO/Bi2Se3 NPs before annealing at 250 °C, respectively. The insets in (c,d) are the cross-section images of Bi2Se3 and ITZO/Bi2Se3 NPs after annealing at 250 °C.

Full size image

Vibration modes

Figure 4a,b show the Raman spectra of the ITZO thin film and Bi2Se3 and ITZO/Bi2Se3 NPs before and after annealing at 250 °C. The Raman spectrum of ITZO before and after annealing at 250 °C exhibited a broad peak in the range of 300‒700 cm−1, which was centered at approximately 563.01 and 587.78 cm−1, respectively. The Bi2Se3 NPs exhibited three significant vibration modes of \({{\text{E}}}_{{\text{g}}}^{2}\), \({{\text{A}}}_{1{\text{g}}}^{2}\), and Se-Se bonds47,48 before annealing at 250 °C at wavenumbers of 126.69, 170.51, and 248.49 cm−1, respectively, as shown in Fig. S3a. In addition, the ITZO/Bi2Se3 NPs exhibited two significant modes of \({{\text{E}}}_{{\text{g}}}^{2}\) and \({{\text{A}}}_{1{\text{g}}}^{2}\) at 128.99 and 170.64 cm−1 before annealing at 250 °C, as shown in Fig. S3b, while the Se–Se mode was suppressed. After annealing at 250 °C, the Bi2Se3 NPs exhibited the same vibration modes (\({{\text{E}}}_{{\text{g}}}^{2}\), \({{\text{A}}}_{1{\text{g}}}^{2}\), and Se-Se at 122.37, 165.90, and 244.67 cm−1, respectively, as shown in Fig. S3c) as before the annealing at 250 °C, as did the ITZO/Bi2Se3 NPs (\({{\text{E}}}_{{\text{g}}}^{2}\) and \({{\text{A}}}_{1{\text{g}}}^{2}\) at 124.66 and 166.67 cm−1, respectively, as shown in Fig. S3d).

Figure 4
figure 4

Raman spectra of ITZO thin film, and Bi2Se3 and ITZO/Bi2Se3 NPs (a) before and (b) after annealing at 250 °C. The inset shows the typical layered structure of Bi2Se347.

Full size image

Bi2Se3 has a layered crystal structure, as shown in the inset of Fig. 4a49, where each layer comprises five monoatomic layers, i.e., \({{\text{Se}}}_{1}-{{\text{Bi}}}_{1}-{{\text{Se}}}_{1}^{\mathrm{^{\prime}}}-{{\text{Bi}}}_{1}-{{\text{Se}}}_{1}\); therefore, it is called a quintuple layer (QL). Covalent bonds dominate the bonding within the QL, and the vdW force connects the QLs32. \({{\text{E}}}_{{\text{g}}}^{2}\) is a Raman active mode, i.e., the in-plane symmetric bending mode associated with the shearing of the upper/lower \({{\text{Se}}}_{1}-{{\text{Bi}}}_{1}\) bond in the opposite vibration direction. \({{\text{A}}}_{1{\text{g}}}^{2}\) is a Raman active mode similar to \({{\text{E}}}_{{\text{g}}}^{2}\) and represents the out-of-plane symmetric stretching of the upper/lower \({{\text{Se}}}_{1}-{{\text{Bi}}}_{1}\) bond in the opposite vibration direction34,47. The Se–Se vibration mode is assigned to the in-plane vibration of the topmost hexagonal network of Se atoms in Bi2Se3 layered structures48,50. Therefore, the Se-Se vibration mode was observed in the Bi2Se3 NPs at 248.49 and 244.67 cm−1 before and after annealing at 250 °C, respectively. The Se–Se vibration mode of the Bi2Se3 NPs was suppressed after the NPs were covered with the ITZO thin film, as shown in Fig. 4a (before annealing) and b (after annealing), implying that the topmost \({{\text{Se}}}_{1}\) atoms in the Bi2Se3 layered structure bonded with the ITZO thin film, suppressing the in-plane Se-Se vibration. These results confirmed that ITZO/Bi2Se3 NPs were successfully fabricated.

Binding energies

Figure 5a–d show the X-ray photoelectron spectra (XPS) of the ITZO thin film for the In 3d, Sn 3d, Zn 2p, and O 1 s orbitals, respectively. In Fig. 5a, the spectrum is split into two peaks of In 3d5/2 and In 3d3/2 at 443.63 and 451.18 eV with an energy difference of 7.55 eV, indicating that In mainly existed in a trivalent form (In3+) in the In2O3 lattice51. In Fig. 5b, the spectrum for the Sn 3d orbital is split to 485.29 and 493.74 eV peaks, which correspond to Sn 3d5/2 and Sn 3d3/2, respectively, indicating the presence of tetravalent Sn (Sn4+) in the SnO2 lattice52. The broad peak centered at 496.72 eV near Sn 3d3/2 is related to the Sn-loss signal (Snloss)52, implying that the ITZO thin film had a high conductivity53. In Fig. 5c, the splitting of spin‒orbit doublets of Zn 2p3/2 and Zn 2p1/2 is observed at energy of 1021.08 and 1044.25 eV, which is assigned to the divalent zinc (Zn2+) in the ZnO lattice54. In Fig. 5d, the O1s orbital is deconvoluted into two peaks at approximately 528.89 and 530.73 eV. The former is related to the oxygen bonded with the metal forming the metal–oxygen (M–O) bonds in the metallic oxides, whereas the latter is attributed to the chemisorbed oxygen (Ochemi: O2−, O− etc.) on the nanostructure surface52,54,55. These results confirmed that Sn4+ and Zn2+ replaced In3+ in the lattice.

Figure 5
figure 5

XPS spectra of ITZO thin film of (a) In 3d, (b) Sn 3d, (c) Zn 2p, and (d) O 1 s before annealing at 250 °C.

Full size image

Figure 6a–c show the XPS spectra of the Bi2Se3 NPs. In Fig. 6a, four peaks are observed at 158.12, 163.43, 159.68, and 164.06 eV, respectively. The first two peaks are attributed to the Bi 4f7/2 and Bi 4f5/2 orbitals of Bi3+ in the Bi2Se3 lattice, whereas the last two peaks are assigned to the Bi 4f7/2 and Bi 4f5/2 orbitals of Bi3+ in the Bi2O3 lattice34. In Fig. 6b, a broad peak is deconvoluted into peaks at 53.51 and 54.18 eV, respectively. The former peak is related to the divalent Se (Se2-) of Se 3d5/2 in the Bi2Se3 lattice, whereas the latter is ascribed to Se 3d3/256. In Fig. 6c, O 1 s peaks are observed at 530.23 and 532.52 eV. The former peak is related to the oxygen binding energy of the metal‒oxygen bonds in the metal‒oxide lattice, and the latter peak corresponds to the chemisorbed oxygen on the nanostructure surface.

Figure 6
figure 6

XPS spectra of Bi2Se3 NPs of (a) Bi 4f, (b) Se 3d, and (c) O 1 s.

Full size image

Figure 7 presents the In 3d (Fig. 7a), Sn 3d (Fig. 7b), Zn 2p (Fig. 7c), O 1 s (Fig. 7d), Bi 4f. (Fig. 7e), and Se 3d (Fig. 7f) XPS spectra of the ITZO/Bi2Se3 NPs. As shown in Fig. 7a, the In 3d spectrum is split into two peaks at approximately 443.85 and 451.41 eV, implying the splitting of In3+ into the In 3d5/2 and In 3d3/2 orbitals in the In2O3 phase. The Sn 3d orbital presented in Fig. 7b shows three peaks at 485.62, 493.99, and 496.91 eV. The first two peaks are attributed to Sn 3d5/2 and Sn 3d3/2, indicating the presence of Sn3+ in the SnO2 lattice. The last peak is assigned to Snloss, which is easily detected in highly conductive materials such as ITZO. The Zn 2p peak is observed two peaks located at 1021.04 and 1044.13 eV, as shown in Fig. 7c. The former peak was ascribed to Zn 2p3/2, and the latter peak corresponded to Zn 2p1/2, implying the presence of Zn2+ in the ZnO lattice. The O 1 s peak was deconvoluted into two peaks at 529.20 and 530.22 eV, which were related to the metal‒oxygen bonds in the metal-oxide lattices and the chemisorbed oxygen on the nanostructure surface, respectively, as shown in Fig. 7d. The signal intensities of Bi 4f. and Se 3d were reduced owing to the ITZO covering, as shown in Fig. 7e,f. Relevant Bi2Se3 lattice peaks of Bi 4f7/2 (159.11 eV), Bi 4f5/2 (163.58 eV), and Se 3d5/2 (53.67 eV) were still detected. In addition, the related Bi2O3 peaks of Bi 4f7/2 at 160.45 eV and Bi 4f5/2 at 165.56 eV were observed, even when the Bi2Se3 NPs were covered with the ITZO thin film. The samples were stored in an ambient environment; therefore, the Bi2O3 phase was formed on the surface of the Bi2Se3 NPs. Thus, the orbital signal of Bi 4f. corresponding to the Bi2O3 lattice was detected. These results indicated that the ITZO covered the surface of Bi2Se3 NPs.

Figure 7
figure 7

XPS spectra of ITZO/Bi2Se3 NPs of (a) In 3d, (b) Sn 3d, (c) Zn 2p, (d) O 1 s, (e) Bi 4f, and (f) Se 3d.

Full size image

Electrical properties

Figure 8 presents the results for the bulk carrier concentration, resistivity, carrier mobility, and conductivity of the ITZO thin film and Bi2Se3 and ITZO/Bi2Se3 NPs after annealing at 250 °C. Table S1 presents the corresponding values, including bulk carrier concentration (− 1 × 10–19 cm−3), carrier mobility (1 × 102 cm2/V-s), resistivity (1 × 10–4 Ω-cm), and conductivity (1 × 103 Ω−1-cm−1), respectively. The bulk carrier concentrations of the ITZO thin film and ITZO/Bi2Se3 NPs were similar, i.e., approximately − 8 × 10–19 cm−3, whereas that of the Bi2Se3 NPs was − 11.04 × 10–19 cm−3. This suggests that the bulk charged carriers of the Bi2Se3 NPs had no significant effect on the ITZO/Bi2Se3 NPs. The carrier mobility (\(\mu \)) of the ITZO/Bi2Se3 NPs was 1.2 × 102 cm2/V-s, which was 263.6% higher than that of the ITZO thin film. The resistivity (ρ) of the ITZO thin film was 23.12 × 10–4 Ω-cm, and that of the ITZO/Bi2Se3 was 6.11 × 10–4 Ω-cm, a decrease of 73.57%. For n-type semiconductors, the resistivity (ρ) can be described as \(\rho =\frac{1}{qn\mu }\), where \(q\) represents the electric charge (1.6 × 10–19 C), and \(n\) represents the carrier concentration, which is similar between ITZO, Bi2Se3, and ITZO/Bi2Se3 in this work. Hence, the decrease in resistivity of ITZO/Bi2Se3 NPs is due to the increase in carrier mobility (\(\mu \)). The conductivity (σ) of ITZO/Bi2Se3 NPs was 1.64 × 103 (1/Ω-cm), which was 281.4% higher than that of the ITZO thin film. Because conductivity (σ) is inversely proportional to resistivity (ρ), the ITZO/Bi2Se3 NPs had higher conductivity than the ITZO thin film. The Hall measurement results indicate that the electrical properties of ITZO can be improved through the formation of a heterostructure with Bi2Se3.

Figure 8
figure 8

Bulk carrier concentration, resistivity, carrier mobility, and Conductivity of the ITZO thin film, and Bi2Se3 and ITZO/Bi2Se3 NPs after annealing at 250 °C.

Full size image

Proposed mechanism for the enhanced carrier mobility in ITZO/Bi2Se3 NPs

Bi2Se3 is identified as an n-type semiconductor with a narrow band-gap of 0.35 eV, and its Fermi level resides within its conduction band32,57. On the other hand, ITZO is characterized as an n-type semiconductor with a broad band-gap of 3.40 eV58. The band structure at the interface undergoes bending upon the formation of the heterostructure between ITZO and Bi2Se3. The band diagram of the ITZO/Bi2Se3 NPs is depicted in Fig. 9, with Fig. 9a illustrating the band diagram before the contact of ITZO and Bi2Se3 in a thermal equilibrium state. Here, EVAC represents the vacuum level, EC is the conduction band level, EF is the Fermi energy level, and EV is the valence band level. Figure 9b shows the band bending after the contact of ITZO and Bi2Se3 in a thermal equilibrium state, leading to the migration of electrons (e‒) and holes (h+) from ITZO to Bi2Se3. The primary carrier (e‒) concentrations progressively increase from the ITZO surface to the interface between ITZO and Bi2Se3 due to the migration of electrons from ITZO to Bi2Se3. Additionally, defects such as dislocations and impurities at the ITZO/Bi2Se3 interface could contribute to decreased carrier concentrations at the ITZO surface of the ITZO/Bi2Se3 NPs59. The carrier mobility, determined through Hall effect measurements, can be estimated using the formula60, \({\mu }_{H}=\frac{\sigma }{e{n}_{H}}\), where \({\mu }_{H}\) is the carrier mobility, \(e\) is the free electron charge, \({n}_{H}\) is the carrier concentration. Therefore, \(\sigma \) and \({n}_{H}\) directly affect the \({\mu }_{H}\). The higher \(\sigma \) corresponds to a the larger \({\mu }_{H}\), while a lower \({n}_{H}\) results in a larger \({\mu }_{H}\). As discussed above, the \(\sigma \) in ITZO/Bi2Se3 is significantly higher than that in pure ITZO. Moreover, the extrinsic properties following the contact of ITZO with Bi2Se3 lead to a decrease in the concentration of primary charged carriers (e‒) from the ITZO/Bi2Se3 interface to the ITZO surface. Simultaneously, the minor charged carriers (h+) in ITZO migrate to Bi2Se3, indicating that e‒ in ITZO/Bi2Se3 NPs exhibits a longer lifetime than that in pure ITZO due to the low recombination rate between e‒ and h+. Therefore, the synergetic effect after the contact between ITZO and Bi2Se3 enhances the electrical properties of the ITZO/Bi2Se3 NPs.

Figure 9
figure 9

Proposed band diagram of ITZO/Bi2Se3 NPs (a) before and (b) after contact.

Full size image

Conclusions

ITZO/Bi2Se3 NPs were synthesized on a pure glass substrate through a thermal CVD at 600°C and magnetron sputtering at room-temperature. XRD and HRTEM analyses confirmed the formation of crystalline Bi2Se3 and disordered ITZO phases. FESEM images indicated that the average thicknesses of the Bi2Se3 and ITZO/Bi2Se3 NPs were approximately 79.3 and 239.9 nm, respectively. The Raman spectra indicated that the ITZO coverage suppressed the Se–Se vibration mode of the Bi2Se3 NPs. XPS measurements revealed the elemental binding energies of the ITZO and Bi2Se3 lattices, confirming that ITZO covered the Bi2Se3 NPs, which was consistent with the HRTEM image of the ITZO/Bi2Se3 NPs. Hall effect measurements of the electrical properties revealed that the bulk carrier concentration of ITZO did not affect through the heterostructure with Bi2Se3. However, the formation of the heterostructure with Bi2Se3 increased the carrier mobility and conductivity of ITZO by 263.6% and 281.4%, respectively, and reduced the resistivity of ITZO by 73.57%. These results indicate that the electrical properties of ITZO can be significantly improved through the formation of a heterostructure with Bi2Se3 owing to its gapless surface state, expanding the potential applications in electronic and thermoelectric devices of ITZO/Bi2Se3 heterostructures.

Experimental

Fabrication of ITZO thin films

The ITZO thin films were deposited on a glass substrate (20 × 20 × 7 mm3) via magnetron sputtering under a base pressure of 6.0 × 10−6 Torr. The substrate‒target distance was 17 cm. The 3-inch ITZO target was composed of 85 wt% In2O3, 15 wt% SnO2, and 5 wt% ZnO. The deposition was performed at a power of 200 W for 750 s, with the introduction of Ar gas at 25 sccm under a working pressure of 3.0 × 10−3 Torr.

Synthesis of Bi2Se3 and ITZO/Bi2Se3 NPs and annealing treatments

Pristine Bi2Se3 NPs were synthesized on a glass substrate (20 × 20 × 7 mm3) via a catalyst-free vapor–solid mechanism using a thermally evaporated deposition process in a horizontal quartz tube furnace. 0.1 g of Bi powder (purity = 99%, 4.78 × 10−4 mol, Merck, Darmstadt, Germany) and 0.1 g of Se powder (purity = 99%, 1.27 × 10−3 mol, Alfa Aesar, Ward Hill, MA, USA) were used as mixture precursors. They were placed in an alumina boat, which was put in the central heating zone of the quartz tube. The evaporating temperature was 600 °C, the heating rate was 10 °C/min, and the heating was performed at the pressure of 1.0 × 10−2 Torr. These conditions were maintained for 20 min to synthesize the Bi2Se3 NPs. The glass substrate was placed vertically upstream in a quartz tube at 140 °C, 21 cm from the precursor mixture. Pristine Bi2Se3 NPs were grown on glass substrates. After 60-min, the system was cooled to room-temperature. ITZO thin films were deposited on the Bi2Se3 NPs for 750 s at a working distance of 17 cm via magnetron sputtering (25 sccm Ar, 200 W) using a 3-inch ITZO target at room-temperature under a working pressure of 3.0 × 10−3 Torr. The ITZO thin films, Bi2Se3 NPs, and ITZO/Bi2Se3 NPs were placed in the heating zone of a horizontal quartz tube furnace. The furnace was heated to 250 °C at a heating rate of 10 °C/min under 2.0 × 10−2 Torr, and then annealing was performed at this temperature for 90 min. Subsequently, the system was cooled down to room-temperature.

Characterization of ITZO, Bi2Se3 NPs, and ITZO/Bi2Se3 NPs

The crystallographic orientations and fine structures of the ITZO thin films and Bi2Se3 and ITZO/Bi2Se3 NPs were examined using high-resolution transmission electron microscopy (HRTEM; JEOL JEM-2010) and X-ray diffraction (XRD) spectroscopy (Bruker D2 PHASER, Cu Kα radiation, λ = 1.5405 Å, operating at 40 kV and 30 mA). The sample morphologies were examined using field-emission scanning electron microscopy (FESEM; JOEL JSM-6335F). The binding vibration modes were analyzed via Raman spectroscopy (3D Nanometer-scale Raman PL microspectrometer) using a semiconductor laser with an excitation energy of 2.54 eV. The chemical binding energies and valence states of the elements were analyzed using X-ray photoelectron spectroscopy (XPS; Perkin-Elmer model PHI 1600, operating at 250 W) with Mg Kα X-rays (1253.6 eV). The electrical properties, i.e., the bulk carrier concentration, carrier mobility, resistivity, and conductivity, were evaluated via Hall effect measurements (Ecopia HMS-3000 Hall Measurement System) at room-temperature under an input current of 5 mA and a magnetic flux density of 0535 T.