Thermoelectric and Topological Insulator Bismuth Chalcogenide Thin Films Grown Using Pulsed Laser Deposition

Abstract

Bismuth chalcogenides have been intensively studied for their high-performance thermoelectric properties and their novel topological surface states, which could significantly benefit novel applications in fields such as TE devices, spintronics, and quantum computing. This chapter reports recent advances in pulsed laser deposition (PLD) for the growth of bismuth chalcogenide (e.g., Bi2Te3, Bi2Se3, and Bi3Se2Te) thin films and their novel properties. It covers a wide range of fields such as thin film growth using PLD for fabricating polycrystalline and epitaxial films with different thermoelectric, nanomechanical, and magnetotransport properties as a function of the PLD processing conditions. Moreover, the proximity-induced superconductivities in Bi inclusions/bismuth chalcogenide thin films are also reported and discussed in detail.

1. Introduction

Bismuth chalcogenide thin films are of great interest because of the exciting properties of topological insulators (TIs) and their applications to thermoelectrics (TEs). These materials have been applied in integrated TE cooling devices working at near room temperature [1, 2]. TIs are exotic materials with an insulating bulk and topologically protected states on the surface which could be used in different applications, such as spintronics and quantum computing [3–6]. The topological surface states (TSSs) exhibit Dirac linear energy dispersion inside the bulk gap, spin-polarization by spin-momentum locking nature, and weak anti-localization (WAL) due to the strong spin-orbit coupling [3–6]. Thus, the WAL through magnetotransport studies has been widely used as a signature of TI materials [7–9].

For application purposes, thin film growth techniques for TE and TI materials are required. Among physical vapor deposition techniques, pulsed laser deposition (PLD) offers great versatility in growing polycrystalline and epitaxial thin films with high growth rates, multiple elements, and diverse structural morphologies for both fundamental studies and applications. The purpose of this chapter is to outline recent advances in the PLD growths of bismuth chalcogenide thin films with desired properties for TE/TI applications and fundamental studies.

2. Thin film growth using pulsed laser deposition (PLD)

Thin film growth consists of nucleation, growth, and coalescence (Figure 1a). In a physical vapor deposition, an extremely nonequilibrium process takes place at high supersaturations and at comparatively high concentrations of impure atoms [10]. Nucleation takes place at high supersaturations S (defined as S = p/p_e, where p is the vapor pressure of the deposited material evaporated from the source at temperature T and p_e is the equilibrium vapor pressure of the substrate material at temperature T_S). The incident vapor arrives at the surface of substrates and then forms small but highly mobile clusters or islands with uniform distribution. In this stage, the impinging atoms and subcritical clusters are incorporated and consequently increase their sizes, while the island density rapidly saturates. In the following stage, the islands are emerged via a coalescence phenomenon which is liquid-like for some cases, especially at high substrate temperatures. Coalescence leads to a decrease in island density and forms local denuding positions on the surface of substrates where further nucleation can then occur (Figure 1a) [11]. The sequence of film nucleation and growth events can be well appreciated in the transmission electron microscopy (TEM) images in Figure 1b–d [11].

2.1. Basic growth modes

For all phase transitions, the formation of thin films is characterized by the formation of nuclei and their growth. Depending on the interaction energies of substrate atoms and film atoms, any of three growth modes in Figure 2a–c can occur:

• 2D Frank-van der Merwe mode: layer-by-layer growth, in which the interaction between substrate and atoms of film is greater than that between adjacent atoms of film.

• 3D Volmer-Weber mode: separated islands form on the surface of substrates, in which the interaction between atoms of film is greater than that between a substrate and the adjacent atoms of film.

• Stranski-Krastanov mode: layer plus island, in which one or two monolayers form first and then grow individually.

Figure 1.

(a) Schematics of thin film growth processes: nucleation, growth, coalescence. Transmission electron microscope images of (b) nucleation, (c) growth, and (d) coalescence of Ag films on (1 1 1) NaCl substrates. Corresponding diffraction patterns are shown.

Figure 2.

Basic modes of thin film growth: (a) island in the Volmer-Weber mode, (b) layer by layer in the two-dimensional Frank-van der Merwe mode, (c) layer plus island in the Stranski-Krastanov mode. (d) Shadowing growth: a schematic of three-dimensional Monte Carlo simulations for oblique angle deposition [13].

After an initially random nucleation of islands on the surface of the substrates, the deposition on the top of the islands is faster than that in the valleys due to the oblique incident flux (the so-called shadowing effect) [12, 13]. Isolated columns are therefore formed on these islands during subsequent growths (Figure 2d) [13].

2.2. Epitaxy growth

Epitaxy refers to the growth of a single crystal film on top of a single crystal substrate. The deposited film is denoted as an epitaxial film or epitaxial layer. The growth is called homoepitaxy if the film and the substrate are the same material, and it is called heteroepitaxy if they are different materials. Epitaxial relationship is determined as: (HKL) || (hkl); [UVW] || [uvw], where (hkl) and (HKL) are the Miller indices of the overgrowth plane and substrate at the common interface. The corresponding parallel directions in the overgrowth and substrate planes, denoted by [wuw] and [UVW], respectively, must also be specified.

2.3. Factors governing the epitaxy growth

The key factors governing epitaxy growths are structural compatibility, chemical compatibility, and growth temperatures.

- Structural compatibility: The structures of a film and a substrate should have good lattice matching in terms of crystal structures (a₀, _sub) and lattice constants (a₀, _film), that is, small lattice misfit.

Lattice misfit f:

$f=\frac{a_{0,sub}-a_{0,film}}{a_{0,film}}\approx \frac{a_{0,sub}-a_{0,film}}{a_{0,sub}}\sqrt{2}.$E1

- Chemical compatibility: This includes chemical bonding and chemical diffusion.

- Growth temperatures: Good epitaxy growth is obtained at above or around the well-defined elevated substrate temperature (T_e). T_e depends on the deposition rate, particle energy, and surface contamination. Generally, a higher temperature is recommended to reduce surface contamination by desorption (or enhance surface mobility) of atoms to reach the favorable sites and also enhance the diffusivity in deposition for favoring re-crystallization and defect annihilation.

2.4. Preparation of bismuth chalcogenide films by PLD

PLD is one of the most convenient thin film growth techniques that uses a high-intensity pulsed laser beam as an external energy source to ablate a target, form a plume, and then deposit thin films onto a substrate. Figure 3 shows a typical PLD system for preparing TE and TI thin films. A substrate is heated and maintained at a desired substrate temperature (T_S) using a thermocouple and a proportional-integral-derivative temperature controller. The thermocouple was buried inside a stainless steel substrate holder, which is heated by a tungsten lamp just behind the holder. The pressure of ambient gas (He/Ar) can be fine-tuned by the needle valve. A KrF excimer laser beam (λ = 248 nm, pulsed duration of 15–20 ns, repetition rate in the range of 1–15 Hz, and fluence of 1–10 J/cm²) is guided by several UV mirrors and focused on a stoichiometric polycrystalline or a single crystal target (e.g., Bi₂Te₃, Bi₂Se₃, Bi₂Se₂Te) within the vacuum chamber by a UV lens. The target-to-substrate distance was 40 mm. During the deposition of Bi₂Se₃ films, pure (6N) He/Ar gas was introduced into the vacuum chamber, which was evacuated to a base pressure of 4 × 10⁻⁴ Pa (or 3 × 10⁻⁶ Torr) and maintained at a certain constant pressure (P), using a differential evacuation system.

Figure 3.

Schematic illustration of a pulsed laser deposition (PLD) system. G: gauge.

The surface of substrates should be atomically clean and free from impurities because the contaminants can interact with the thin films being deposited and substantially degrade its quality and adhesion to the substrates. The presence of unwanted surface contaminants can also influence the growth and orientation of the films in an undesired manner. In the depositions for TE thin films, an approximately 300-nm-thick SiO₂ layer was thermally grown on the Si wafers (thickness 525 μm) for electrical isolation purpose. The wafers were cut into 1.5 cm × 1.5 cm substrates. The substrates were cleaned with acetone to dissolve any contaminants adhering to the surface of substrates such as grease and oils. This was followed by rinsing with methanol to remove any residues left behind after cleaning with acetone. Afterward, the substrates were rinsed in distilled water and dried with nitrogen flow. The substrates were then used for the deposition of TE thin films.

Here are some examples of PLD growth of TE films. For Bi₂Se₃ thin films, the depositions were at T_S of 200–350°C and helium ambient pressure (P) of 0.7–173 Pa. The number of laser pulses was 9000 and the deposition took 30 min. The average growth rate was approximately 0.46 Å/pulse [14]. For the growth of Bi₂Te₃ thin films, T_S was varied from room temperature (30°C) to 380°C and the Ar ambient pressure (P_Ar) was at 80 Pa. The number of laser pulses was 12,000 and the deposition took 40 min. The average growth rate was approximately 0.52 Å/pulse [15]. For the growth of Bi-Se-Te thin films, the depositions were at T_S of 200–350°C and a helium ambient pressure (P_He) of 0.027–86.7 Pa. The number of laser pulses was 9000 and the deposition took 15 min. The average growth rate was approximately 0.6 Å/pulse [16].

3. Thermoelectric bismuth chalcogenide thin films

3.1. Crystal structures of bismuth chalcogenides

The crystal structures of bismuth chalcogenides (e.g., Bi₂Te₃, Bi₂Se₃, and Bi₃Se₂Te) are usually described by a hexagonal cell consisting of 15 layers of atoms stacking along the c-axis with a sequence shown below [17], as shown in Figure 4a and b.

Figure 4.

Crystal structures of (a) Bi₂Te₃, (b) Bi₂Se₃, and (c) Bi₃Se₂Te in the form of (Bi₂)_m(Bi₂Se₂Te)_n (m = 1, n = 2) homologous series. The unit cells are marked with black thick boxes. (d) X-ray diffraction patterns of the typical TE bismuth chalcogenide thin films grown at 300°C on SiO₂/Si substrates.

A 5-atomic-layer-thick lamellae of ‐(Te⁽¹⁾‐Bi‐Te⁽²⁾‐Bi‐Te⁽¹⁾)‐ or ‐(Se⁽¹⁾‐Bi‐Se⁽²⁾‐Bi‐Se⁽¹⁾)‐ is called a quintuple layer, QL, in which the Te⁽¹⁾‐Bi and Bi‐Te⁽²⁾ or Se⁽¹⁾‐Bi and Bi‐Se⁽²⁾ are ionic-covalent bonds. Because of the weak binding (i.e., Van der Waals force) between Te or Se layers, bismuth chalcogenides could be cleaved easily along the plane perpendicular to the c-axis. It also induces the anisotropic thermal/electrical transport properties. For example, the thermal conductivity along the plane perpendicular to the c-axis (~1.5 W m⁻¹ K⁻¹) is nearly two times higher than that along the c-axis direction (~0.7 W m⁻¹ K⁻¹) [17, 18].

The crystal structures of Bi₃Se₂Te can be formed by ordered stacking of Bi₂Se₂Te and Bi₂ building blocks, that is, (Bi₂)_m(Bi₂Se₂Te)_n (m = 1, n = 2) [19, 20], in which the covalently connected double layers of bismuth (Bi–Bi) lie between two QL (Se‐Bi‐Te‐Bi‐Se) blocks (Figure 4c); the (Bi–Bi) strictly alternates with two QL (Se–Bi–Te–Bi–Se) blocks [9, 20]. Figure 4d shows the typical XRD patterns of Bi₂Te₃, Bi₂Se₃, and Bi₃Se₂Te thin films grown on SiO₂/Si substrates at T_S = 300°C. They exhibit the dominance of (0 0 1) family planes of the rhombohedral phases of Bi₂Te₃ (PDF#82-0358), Bi₂Se₃ (PDF#33-0214), and Bi₃Se₂Te (JCPDS 00-053-1190), indicating that the films are highly c-axis oriented (i.e., textured films).

3.2. Introduction to thermoelectrics and applications

Thermoelectric materials are solid-state energy converters in which the combination of thermal, electrical, and semiconducting properties allows them to be used to convert waste heat into electricity or electrical power directly into cooling and heating [21].

3.2.1. The thermoelectric figure of merit (ZT)

The performance of the thermoelectric materials is often denoted as figure of merit Z whose unit is K^–1 or ZT with the dimensionless unit [17, 22].

$ZT=\frac{{\alpha }^2\sigma }{\kappa }T=\frac{{\alpha }^2\sigma }{{\kappa }_E+{\kappa }_L}T$E2

where σ, α, κ, and T are the electrical conductivity, Seebeck coefficient, thermal conductivity, and absolute temperature, respectively. The total thermal conductivity can be split into a lattice contribution (κ_L) and an electronic contribution (κ_E). The quantity α²σ is commonly used to represent the thermoelectric power factor (PF).

The efficiency of a thermoelectric device is directly related to ZT. For power generation, the maximum efficiency (η) is expressed by [23]

$\eta =\frac{T_h-T_c}{T_h}\cdot \frac{\sqrt{1+Z\overline{T}}-1}{\sqrt{1+Z\overline{T}}+\frac{T_c}{T_h}}$E3

and for air-conditioning or refrigeration, the coefficient of performance is [23]

$COP=\frac{T_c}{T_h-T_c}\cdot \frac{\sqrt{1+Z\overline{T}}-\frac{T_h}{T_c}}{\sqrt{1+Z\overline{T}}+1}$E4

where T_h and T_c are the hot-end and cold-end temperatures of the thermoelectric materials, respectively, and $\overline{T}$ is the average temperature of T_h and T_c. Therefore, the enhanced ZT value of TE materials is important to increase the COP for practical applications.

3.2.2. Conflicting properties in thermoelectric materials

Maximizing ZT is challenging due to the interdependence of the TE parameters. An increased power factor α²σ by optimizing the carrier concentration n and/or a reduced lattice thermal conductivity κ_L by introducing the scattering centers are necessary to enhance ZT value. The dependences of these parameters with scattering factor r, carrier effective mass m^*, carrier mobility μ, and their interconnectivity limits the ZT to about 1 in large bulk materials [24].

The electrical conductivity (σ) and electrical resistivity (ρ) are related to n through the carrier mobility μ:

$\text{1}/\rho =\sigma =ne\mu .$E5

The Wiedemann-Franz Law [2] states that the electronic contribution to the thermal conductivity is proportional to the electrical conductivity (σ) of the materials, with the relationship being

${\kappa }_e=L\sigma T=ne\mu LT$E6

where e is electron charge, and L is the Lorenz factor of 2.48 × 10⁻⁸ J²/K² C² for free electrons and this can vary particularly with carrier concentration [2].

The kinetic definition of α is the energy difference between the average energy of mobile carriers and the Fermi energy [25]. An increase in n leads to the increase in both the Fermi energy and the average energy, but the former increases more rapidly than the latter and thus results in a decrease in α value and a reduction factor of α²n. Thus, in attempting to increase ZT for most of the homogeneous materials, the carrier concentration (n) increases electrical conductivity (σ) but reduces the Seebeck coefficient (α). For this reason, in metals and degenerate semiconductors (energy-independent scattering approximation), the Seebeck coefficient can be expressed as [2]:

$\alpha =\frac{8{\pi }^2{k}_B^2}{3e{h}^2}{m}^{*}T{\left(\frac{\pi }{3n}\right)}^{2/3}.$E7

The high m^* causes the Seebeck coefficient to rise according to Eq. (7). High m^* materials generally possess low μ which limits the enhancement of power factor by (m^*)^3/2μ. Noticeably, the defect scatters are not only the phonons but also the electrons which lead to reduce κ_L as well as μ. Therefore, the ratio of μ/κ_L determines the improvement in ZT [17, 24]. Although the increase in the ratio is usually experimentally achieved through a greater reduction in κ_L rather than that in μ, some fundamental issues in this mechanism are not understood well [24].

Figure 5 shows the compromise between large α and high σ in thermoelectric materials that must be struck to maximize the figure of merit ZT. Meanwhile, the low carrier concentration will result in lower electrical conductivity with decreasing ZT. The ZT and PF peaks typically occur at carrier concentrations between 10¹⁹ and 10²¹ carriers per cm³ (depending on the material system), which fall in between common metals and semiconductors, that is, the concentrations found in heavily doped semiconductors [2].

Figure 5.

Maximizing the efficiency (ZT) of a thermoelectric device involves a compromise of thermal conductivity (κ, plotted on the y-axis from 0 to a top value of 10 W m⁻¹ K⁻¹) and Seebeck coefficient (α, 0–500 μV K⁻¹) with electrical conductivity (σ, 0–5000 Ω⁻¹ cm⁻¹) [2].

3.2.3. Overview of thermoelectric applications

TE devices have unique features: no moving parts, substantially less maintenance, quiet operation, high power density, low environmental impact, and high reliability [26]. Commercial use has been made mostly from Peltier thermoelectric cooling (TEC) effect, such as in small refrigerator devices used for camping and outdoor activities, automotive climate control seats, and localized cooling at the hot spots of chips. Figure 6 gives an overview of the present and potential applications of thermoelectric generators (TEGs) [27]. Indeed, TEGs have been used for the power in miniaturized autarkic sensor systems, automotive waste heat recovery systems, ventilated wood stove, heating systems, water boilers, and heat recovery in industry.

Figure 6.

Overview of the potential applications of thermoelectric generators [27].

3.3. Thermoelectric properties of polycrystalline Bi₂Te₃, Bi₂Se₃, and Bi₃Se₂Te thin films with controlled structure morphology

Some typical HRTEM images of Bi₂Te₃, Bi₂Se₃, and Bi₃Se₂Te grown using PLD are shown in Figure 7 [14–16]. HRTEM images performed on a high μ Bi₂Te₃ film with nanodisk-like morphology grown at 220°C are shown in Figure 7a. Clearly, the lower inset in Figure 7a shows the film with uniform thickness of approximately 295 nm and a SiO₂ layer with a thickness of 300 nm. It shows that projected period of 0.508 nm along the c-axis corresponds to the lattice spacing of the (0 0 6) planes. The highly (0 0 1)-orientated and crystallized structures of the film should facilitate the transport of charge carriers. The c-axis lattice constant of the Bi₂Te₃ film is 30.48 Å, which agrees closely with the value (30.44 Å) presented in JCPDS 82-0358. The other Bi₂Te₃ films grown at T_S ≥ 220°C also display similar HRTEM results.

For a Bi₂Se₃ film deposited at 300°C and 40 Pa, an HRTEM image taken at the boundary of three platelets (P1, P2, and P3) revealed the granular-polycrystalline structure of the films (Figure 7b). Moreover, P1 and P2 partly overlapped and the corresponding fast Fourier transform (FFT) of this overlapping region indexed by $\left\{003\right\}$ patterns of [0 1 0] zone axis was performed from the dashed-square area (Figure 7b, inset). The projected period along the c-axes of both P1 and P2 was 9.60 Å, corresponding to (0 0 3) planes, which was close to the reported value of 9.55 Å in Ref. [28].

HRTEM images of a Bi₃Se₂Te film deposited at 250°C and 40 Pa are shown in Figure 7c and d. Nanocrystallites with sizes of 10–20 nm are clearly observed in Figure 7c, confirming the nanocrystalline type of the Bi₃Se₂Te films. The interplanar spacing of the Bi₃Se₂Te (0 0 5) planes in the nanocrystallites is approximately 0.464 nm. Therefore, the c-axis lattice constant is determined to be 23.2 Å, closely agreeing with the value of 23.25 Å for Bi₃Se₂Te bulk (JCPDS 00-053-1190). In addition, the white lines in Figure 7c indicate the orientations of the (0 0 5) planes. It is seen that the overall orientation of the crystallites is disorganized. Intriguingly, near the interface of the film and substrate, the film has some nanoinclusions with sizes of 12–17 nm, as shown in Figure 7d. The EDS analysis shows that these are Bi semimetal nanoprecipitates (the inset in Figure 7d). The lattice spacing of Bi nanoinclusions is ~0.32 nm, corresponding to the Bi (0 1 2) planes. It has been found that Bi nanoinclusions can lead to the enhanced Seebeck coefficient and reduced lattice thermal conductivity owing to the low-energy electron filtering and phonon scattering at the nanoinclusions, respectively [29–31].

Figure 7.

(a) HRTEM images of a high carrier mobility (μ) Bi₂Te₃ film with nanodisk-like morphology grown at 220°C and P_Ar of 80 Pa. (b) HRTEM image of an optimized Bi₂Se₃ film deposited at 300°C and P_He of 40 Pa. The inset shows the FFT patterns of the dashed-square area in the HRTEM image. (c and d) HRTEM images of the Bi₃Se₂Te film grown at 250°C and P_He of 40 Pa. The white lines in (c) indicate the (0 0 5)-orientation of nanograins. Inset in (d): FFT patterns and EDS spectra performed at film and Bi nanoinclusion positions.

Figure 8a shows the T_S-dependent α, σ, and PF (= α²σ) of some nanostructured Bi₂Te₃ films [15]. The σ value gradually increased from 34.5 ± 0.1 to 814.3 ± 1.5 S/cm when T_S was increased from 30 to 300°C, and then sharply decreased to 647.3 ± 0.4 S/cm at 340°C and 414.0 ± 1.2 S/cm at 380°C. The enhanced σ (= 647.3 – 814.3 S/cm) of the films grown at 220–340°C originated from the substantially enhanced μ because the n exhibited a slight decrease [15]. Although the coupled relationship between σ (= neμ) and |α| (~n^−2/3) generally constrains the concurrent enhancement of σ and |α|, a reduction in n and a substantial increase in μ in the same optimal range of T_S (= 220–340°C) could lead to high values of both σ and |α|. Consequently, the PF of the stoichiometric Bi₂Te₃ films grown in the range of 220–340°C reached remarkably high values, ranging between 18.2 ± 0.25 and 24.3 ± 0.44 μW cm⁻¹ K⁻², whereas the PF was low (≤0.44 μW cm⁻¹ K⁻²) in the case of nonstoichiometric films deposited at T_S ≤ 120 or 380°C (Figure 8a).

Figure 8.

(a) Substrate temperature (T_S) dependence of room temperature Seebeck coefficient α (red circles), electrical conductivity σ (blue triangulars), and power factor (PF = α²σ, black squares) of the Bi₂Te₃ and Bi₄Te₅ (for “PH” point) films. The morphology abbreviations: CNP, columnar nanoparticle; CNF, columnar nanoflower; ND, nanodisk; CP, compact polycrystalline; LTP, layered triangular platelet; PH, polyhedral. (b) Contour plot of the Bi₂Se₃ film’s PF as a function of P_He and T_S. The morphology abbreviations: SC, smooth and compact; RG, rice grain; TP, triangular polygonal; S-LFs, super-layered flakes; L-HPs, layered hexagonal platelets. (c) Contour plot of the film’s PF as a function of T_S from 200 to 350°C and P_He from 0.027 to 86.7 Pa. (d) |S| vs. σ of the films in this study and the relevant novel TE materials in the literature, listed in Table 1. Solid curves denote different PFs from 1 to 50 μW cm⁻¹ K⁻².

In order to check the evolution of the PF (= α²σ) as a function of P_He and T_S, the contour plot is illustrated Figure 8b. The PF of Bi₂Se₃ films increased with increasing T_S from 200 to 300°C because σ became considerably larger but the Seebeck coefficient diminished only slightly. However, for films deposited at 350°C, PF was lowered primarily because of the reduction in S and not the increase in σ. At intermediate pressures (40–93 Pa), the Bi₂Se₃ films remained stoichiometric or slightly Se-rich compositions, which in turn led to the reduced carrier concentrations and significantly enhanced the α values [14]. Thus, the PF of Bi₂Se₃ films grown at intermediate pressure was typically higher than the PF of films grown at a low or high pressure. The optimal value of PF was 5.54 ± 0.34 μW cm⁻¹ K⁻² for the layered hexagonal platelet Bi₂Se₃ films deposited at 300°C and 40 Pa [14].

System	Type	Method	n; (10^19; cm^−3)	μ; (cm^2/Vs)	σ (S/cm)	α (μV/K)	PF (μW; cm^−1 K^−2)	References
Bi2Te3	Layered smooth film	PLD	10.1	90.6	1464	−186	50.6	[39]
Bi2Te3	Layered compact polycrystalline	PLD	5.0	102	814.3	−172.8	24.3	[15]
Bi2Te3	Nanoparticle film	PLD	9.7	14.8	230	−91	1.9	[34]
Bi2Te3	Super-assembled film	PLD	4.0	12.4	79	−113	1.0	[33]
Bi2Se3	Layered hexagonal platelets	PLD	7.4	81.4	963.8	−75.8	5.5	[14]
Bi3Se2Te	Nanocrystalline film	PLD	35.5	34.4	1747.5	−68.8	8.3	[16]
Bi2Te3	Layered structure	Sputtering	95	12.1	1840	−70	8.8	[35]
Bi2Se3	Bulk	Melting and; hot-pressing	–	–	251.9	−175	7.7	[32]
Bi2Se0.3Te2.7	Bulk	Ball milling hot pressing	–	–	892	−190	32.2	[36]
Bi2Se1.5Te1.5	Bulk	Zone melting	1.2	230	441.6	−193	16.5	[37]
Bi2Se1.8Te1.2	Nanoplatelet bulk	Polyol method	–	–	199.6	−80.9	1.3	[38]
Bi2Se2Te	Bulk	Ball milling hot pressing	–	–	1613	−60	5.8	[36]

Table 1.

Material, type, method, carrier concentration (n), mobility (μ), electrical conductivity (σ), Seebeck coefficient (S), power factor (PF = S²σ) of the optimal bismuth chalcogenide films in this study as compared to properties of Bi₂Te₃, Bi₂Se₃, Bi₂Se_xTe_1−x bulk and films reported in the literature. All the selected values were recorded at room temperature.

The T_S- and P_He-dependent PF of nanocrystalline Bi₃Se₂Te films is further shown in Figure 8c. The films grown at 200°C only have PFs of 1.0–2.8 μW cm⁻¹ K⁻². The PFs of the films grown at higher T_S are significantly enhanced because of their high σ values. Around T_S = 250–350°C and P_He = 40 Pa, a window for high PF is clearly observed. An optimal PF of 8.3 μW cm⁻¹ K⁻² is achieved for a Bi₃Se₂Te film deposited at 250°C and 40 Pa.

Table 1 summarizes the transport and room-temperature TE properties of bismuth chalcogenides in the literature [14, 15, 32–39]. For PLD growths, the highly (0 0 1)-oriented layered Bi₂Te₃ films achieved a PF of 50.6 μW cm⁻¹ K⁻² [39], and the layered compact polycrystalline film possessed a PF value of 24.3 μW cm⁻¹ K⁻² [15]. The Bi₂Se₃ films generally have lower TE properties than those of Bi₂Te₃ films. For example, the optimal PF of the Bi₂Se₃ films grown by PLD was 5.5 μW cm⁻¹ K⁻² [14], which was slightly lower than the PF of Bi₂Se₃ bulk (PF ≈ 7.7 μW cm⁻¹ K⁻² ) [32]. The nanocrystalline Bi₃Se₂Te films had an optimal PF of 8.3 μW cm⁻¹ K⁻² [16]. Further, PLD growth allows fabrication of nanostructured TE films with different morphologies of nanoparticle Bi₂Te₃ film (PF = 1.9 μW cm⁻¹ K⁻²) [34] and super-assembled Bi₂Te₃ film (PF = 1.0 μW cm⁻¹ K⁻²) [33]. The Bi₂Te₃ film deposited by the sputtering technique had PF of 8.8 μW cm⁻¹ K⁻² [35]. There are some reports of TE properties for bulk materials of bismuth chalcogenides, such as Bi₂Se_1.8Te_1.2 nanoplatelet (PF ≈ 1.3 μW cm⁻¹ K⁻² ) [38], Bi₂Se₂Te (PF ≈ 5.8 μW cm⁻¹ K⁻² ), Bi₂Se_1.5Te_1.5 (PF ≈ 16.5 μW cm⁻¹ K⁻² ) [37], and Bi₂Se_0.3Te_2.7 (PF ≈ 32.2 μW cm⁻¹ K⁻² ) [36]. Unfortunately, the thermal conductivity κ of the films is missed in the reports to fully evaluate the TE performance of the films. Nevertheless, the κ of polycrystalline films with small grain sizes should be reduced thanks to the extensive phonon scattering at interfaces and grain boundaries.

Finally, Figure 8d shows the |S| vs. σ plot for the list in Table 1. The solid curves denote different values of PFs (= S²σ). It can be found that TE nanomaterials usually possess low σ values due to the separating or voided structure morphology, but bulk and thin films have superior σ. Note, the significant reduction in thermal conductivity κ is the key factor for employing nanostructured materials in the TE field.

4. Nanomechanical properties of Bi₂Te₃ and Bi₃Se₂Te thin films

Effects of helium ambient pressure (in PLD) on the nanomechanical properties of Bi₂Te₃ and Bi₃Se₂Te thin films have been investigated [9, 40]. The Bi₂Te₃ thin films were grown at T_S of 250°C on c-plane sapphire substrates using excimer laser PLD with a power density of 5 J/cm², at a repetition rate of 2 Hz. The helium pressures in PLD growth varied from 2 × 10⁻⁵ to 6.5 × 10⁻³ Torr. Similarly, Bi₃Se₂Te thin films were deposited on Al₂O₃ (0 0 0 1) substrates at a fixed substrate temperature of 250°C and P_He ranging from 2 × 10⁻⁵ to 6.5 × 10⁻¹ Torr through PLD. The light source of the PLD system was a KrF excimer laser with λ = 248 nm, pulse duration of 20 ns, fluence of 7.0 J/cm², and repetition rate of 2 Hz. The target-to-substrate distance, the number of laser pulses and deposition time were 40 mm, 3000, and 25 min, respectively. The grown films had the average thickness of 200 nm (the average growth rate was ~0.67 Å/pulse). Nanoindentation experiments were performed on a MTS Nano Indenter^® XP system with a three-sided pyramidal Berkovich indenter tip using the continuous stiffness measurement (CSM) technique [41]. The hardness and Young’s modulus of the Bi₂Te₃ and Bi₃Se₂Te thin films were determined from the load-displacement results through the analytical method proposed by Oliver and Pharr [42].

As shown in Figure 9a, the hardness monotonically increased from 2.92 ± 0.12 to 4.02 ± 0.14 GPa for Bi₂Te₃ films, and from 2.5 ± 0.2 to 3.0 ± 0.1 GPa for Bi₃Se₂Te films when P_He was increased from 2.0 × 10⁻⁵ to 2.0 × 10⁻³ Torr. Similarly, the Young’s modulus of Bi₂Te₃ thin films was 106.31 ± 0.63, 115.51 ± 1.92, and 127.46 ± 9.21 GPa for P_He at 2.0 × 10⁻⁵, 2.0 × 10⁻⁴, and 2.0 × 10⁻³ Torr, respectively. For P_He > 2.0 × 10⁻³ Torr, the hardness (Young’s modulus) of Bi₃Se₂Te films continues to increase with increasing P_He, namely 3.2 ± 0.1 GPa (105.2 ± 10.2 GPa) at 2.0 × 10⁻¹ Torr and 5.8 ± 0.2 GPa (188.5 ± 4.3 GPa) at 6.5 × 10⁻¹ Torr (Figure 9b).

Figure 9.

Material and nanomechanical properties of Bi₂Te₃ and Bi₃Se₂Te thin films grown on Al₂O₃ (0 0 0 1) substrates at T_S of 250°C and various helium pressures (P_He) between 2.0 × 10⁻⁵ and 6.5 × 10⁻¹ Torr [9, 40]: (a and b) the hardness and Young’s modulus, (c) grain size (D), (d) the Hall-Petch behavior observed on the Bi₃Se₂Te thin films, in which the hardness is observed to increase approximately with D^−1/2 (D is grain size).

The crystallite sizes (D) of the films were estimated using the Scherrer equation D = 0.9λ/Bcosθ, where λ, B, and θ are the X-ray wavelength, full width at the half maximum of the Bi₂Te₃ (0 0 15) peak or Bi₃Se₂Te (0 0 5) peak, and Bragg diffraction angle, respectively. The P_He-dependent D of the Bi₂Te₃ and Bi₃Se₂Te films is shown in Figure 9c. The grain size increases monotonically from 11.0 to 20.0 nm for Bi₂Te₃ and from 16.1 to 20.5 nm for Bi₃Se₂Te with increasing P_He from 2.0 × 10⁻⁵ to 2.0 × 10⁻³ Torr. In the nanoscale, grain size can affect significantly the mechanical properties of materials. The dislocation activities can be drastically suppressed in a polycrystalline material when the grain size is decreased, and thus the grain boundary sliding and/or grain rotations become the dominant deformation behavior, which in turn would lead to the manifestations of the inverse Hall-Petch effect [43]. Softening caused by grain boundary sliding is mainly attributed to large amount of defects in grain boundaries, which allow rapid diffusion of atoms and vacancies under stress [44]. Consequently, the plastic deformation of Bi₂Te₃ films should be dominated by the grain boundary sliding and/or grain rotation rather than the dislocation activity because of D ≤ 20 nm [40], which is consistent with the results in Refs. [45–48].

In contrast, the 200-nm-thick Bi₃Se₂Te films with D of 16.1–25.1 nm grown at a larger P_He range of 2.0 × 10⁻⁵ to 6.5 × 10⁻¹ Torr exhibited the nanomechanical followed Hall-Petch relationship [44, 49]. The hardness and Young’s modulus of the Bi₃Se₂Te thin films monotonically increased with increasing P_He because of a corresponding decrease in grain sizes (Figure 9a–c). Figure 9d shows that hardness (H) increased linearly with D^−1/2 (where D is the grain size of the Bi₃Se₂Te films in the nanoscale regime) which is the typical Hall-Petch relationship [44, 49]. This is because the multiplication and mobility of dislocations are hindered by reducing the grain size [44]. It is reasonable for the observed phenomenon when the present grain sizes ranged between 25.1 and 16.1 nm which is larger than the typical critical D_c of 10 nm [44, 49]. It is demonstrated that the hardness and Young’s modulus of the Bi₂Te₃ and Bi₃Se₂Te thin films can be enhanced by proper selection of the ambient pressure in PLD growths.

5. Topological insulator bismuth chalcogenide thin films and their novel properties

5.1. The epitaxial growths of bismuth chalcogenide thin films

Figure 10.

Bi₂Te₃ films grown on SrTiO₃ (1 0 0) substrates [53]: (a) XRD φ-scan patterns of the 200-nm-thick Bi₂Te₃ (0 1 5) plane and the SrTiO₃ (1 1 1) plane; (b) schematics of the in-plane arrangement of h-Bi₂Te₃ / SrTiO₃ (1 0 0). Bi₂Se₃ films grown on c-plane sapphire substrates [64]; (c) in-plane φ-scan $\left\{01\overline{1}5\right\}$ planes of the film and $\left\{01\overline{1}2\right\}$ planes of the substrate; (d) the schematic depicting twin/domain growth: in one case, the basal plane of Bi₂Se₃ is aligned with that of Al₂O_3, and in the other, there is a rotation of 60°/180°. Bi₃Se₂Te films grown on c-plane sapphire substrates [9]; (e) φ-scan patterns of hexagonal h-Bi₃Se₂Te thin films grown on Al₂O₃ (0 0 0 1) substrates; and (f) schematics of the in-plane arrangement of hexagonal h-Bi₃Se₂Te/h-Al₂O₃ (0 0 0 1).

Topological insulator (TI), a new class of quantum matter, possesses insulating in bulk and robust gapless topological surface states (TSSs) in which the spin of the electron is locked perpendicular to its momentum by strong spin-orbit interaction [6, 50, 51]. TIs have been identified as promising materials for exploiting exciting physics and developing potential applications in optoelectronics, spintronics, and quantum computations [50–54]. Dirac fermions in TIs have also been studied by angle-resolved photoemission spectroscopy [55–57] or scanning tunneling microscopy [58, 59]. In magnetotransport studies, TSS can be studied by weak antilocalization (WAL) [4, 9, 60, 61] and Shubnikov-de Haas oscillations [3, 62].

Topological insulator bismuth chalcogenide thin films have been grown epitaxially on various substrates using PLD. Onose et al. reported the epitaxial growth of Bi₂Se₃ thin films on InP (1 1 1) substrates (the lattice mismatch of 0.2%) [7]. A designed Se-rich target with an atomic ratio of Bi:Se = 2:8 was used to compensate for the issue of high doping carriers and to avoid unwanted Se-deficient phases. The pulsed laser power and repetition were 140 mJ and 10–20 Hz, respectively. The Bi₂Se₃ films obtained a small full-width at half-maximum (FWHM) for the XRD rocking curve of (0 0 0 6) peak. The surfaces of the films are composed of triangular pyramids with step-and-terrace structures, reflecting the hexagonal symmetry of Bi₂Se₃. The epitaxial relationship is Bi₂Se₃ (0 0 1) || InP (1 1 1) and Bi₂Se₃ $[11\overline{2}0]\left|\right|\text{InP}[00\overline{1}]$. Le et al. reported the epitaxial growth of Bi₂Se₃ films on SrTiO₃ (STO) (1 1 1) substrates using PLD at T_S of 300 and 350°C [63]. The PLD conditions were the pulse fluence of 3.7 J/cm², helium pressure of 40 Pa (300 mTorr), and repetition rate of 2 Hz. The laser source was KrF excimer laser (λ = 248 nm, duration 20 ns). By comparing the Bi₂Se₃ {0 1 5} and STO {2 0 0} diffraction peaks, the epitaxial relationship between the film and substrate was determined to be Bi₂Se₃ (0 0 1) || STO (1 1 1) and Bi₂Se₃ [1 1 0] || STO $[\stackrel{-}{1}10]$ [63].

Figure 10 presents the PLD epitaxial growths of Bi₂Te₃, Bi₂Se₃, and Bi₃Se₂Te thin films on large-misfit substrates [9, 53, 64]. The PLD conditions for growing Bi₂Te₃ films on STO (1 0 0) were as follows: substrate temperature of 300°C; helium ambient pressure of 40 Pa; repetition rate of 2 Hz; pulsed fluence of approximately 3.4 J/cm². As shown in Figure 10a, a φ-scan was conducted on the (0 1 5) plane of a 200-nm-thick Bi₂Te₃ film and the (1 1 1) plane of an STO (1 0 0) substrate in skew symmetric geometry by tilting the samples. The in-plane orientation of a hexagonal h-Bi₂Te₃/STO (1 0 0) film displayed a 12-fold symmetry instead of the expected six-fold symmetry of the (0 1 5) plane in Bi₂Te₃. Figure 10b shows a schematic drawing of the in-plane atomic arrangement between an h-Bi₂Te₃ film and an STO (1 0 0) substrate. Since the principal crystallographic orientations of h-Bi₂Te₃ films grown on STO (1 0 0) substrates can be aligned along either the STO [1 0 0] or STO [0 1 0] directions, the in-plane arrangements result in an observed 12-fold symmetry. The angle differences between STO [0 1 0] and the two orientations of h-Bi₂Te₃ [1 1 0] were 30° and 0°, respectively, as shown in Figure 10b. In other words, the in-plane relationships were Bi₂Te₃ [1 1 0] || STO [0 1 0] and Bi₂Te₃ [1 0 0] || STO [1 0 0]. Lee et al. reported epitaxial growth via domain matching epitaxy of Bi₂Se₃ thin films on Al₂O₃ (0 0 0 1) substrates with over 13% lattice misfit and a critical thickness of less than one monolayer [64]. A relatively low repetition rate of 0.2 Hz and low T_S of 250°C are key parameters of the PLD growth to achieving high-quality Bi₂Se₃ epitaxial films. Figure 10c shows φ-scan XRD results performed on $\left\{01\overline{1}5\right\}$ planes of the film and $\left\{01\overline{1}2\right\}$ planes of the substrate. Clearly, the presence of six peaks corresponding to the Bi₂Se₃ $\left\{01\overline{1}5\right\}$ planes confirmed the epitaxy. A schematic depicting twin/domain growth and direction in the study is illustrated in Figure 10d [64]. In one case, the Bi₂Se₃ basal plane is aligned with that of Al₂O₃, and in the other, there is a rotation of 60°/180°. The epitaxial relationships are written as (0 0 0 1) Bi₂Se₃ || (0 0 0 1) Al₂O₃ (out-of-plane) and $[2\overline{1}\overline{1}0]$ Bi₂Se₃ || $[2\overline{1}\overline{1}0]$ Al₂O₃ or $[2\overline{1}\overline{1}0]$ Bi₂Se₃ || $[11\overline{2}0]$ Al₂O₃ (in-plane). Figure 10e shows the typical φ-scan patterns of hexagonal h-Bi₃Se₂Te/h-Al₂O₃ (0 0 0 1) [9]. With the skew symmetric geometry, the φ-scan measurements were performed on the (1 1 6) plane of Al₂O₃ substrates and the (1 1 12) plane of Bi₃Se₂Te films. As shown in Figure 10e, the in-plane orientations of both Al₂O₃ (1 1 6) and Bi₃Se₂Te (1 1 12) exhibited six-fold symmetries with a 30° difference. Figure 10f illustrates the in-plane atomic arrangement between h-Bi₃Se₂Te and h-Al₂O₃ (0 0 0 1). The epitaxial relationship between the films and substrates is (0 0 0 1) Bi₃Se₂Te || (0 0 0 1) Al₂O₃ and $[110]$ Bi₃Se₂Te || $[210]$ Al₂O₃ [9]. This in-plane orientation was established to obtain the optimal lattice matching.

5.2. Magnetotransport properties of bismuth chalcogenide thin films

The weak antilocalization (WAL) which is a negative quantum correction to classical MR caused by the wave nature of electrons is used as a signature of TSS. In TIs, WAL is induced by both the helicity of the surface state and the spin-orbit coupling of bulk [4, 61, 65, 66]. In a low B field, the 2D WAL MR of a system with strong spin-orbit interaction can be described using the Hikami-Larkin-Nagaoka model [65], which is [4, 60, 61]

$\frac{\Delta R_{\square }(B)}{{[R_{\square }(0)]}^2}=-\alpha \frac{e^2}{2{\pi }^2\hslash }\left[\Psi \left(\frac{1}{2}+\frac{B_{\phi }}{B}\right)-\ln \left(\frac{B_{\phi }}{B}\right)\right]$E8

where $R_W$ is the sheet resistance, $\Delta R_W=R_W(B)-R_W(0)$, $\Psi (x)$ is the digamma function, $B_{\varphi }=\hslash /(4e{L}_{\varphi }^2)$ is a magnetic field varying with the coherence length $L_{\varphi }$, α is a parameter and reflects the number of conduction channels. In a 3D TI, α = −1/2 for a single coherent transport channel in the 2D surface states, and α = −1 for two independent coherent transport channels with similar L_φ in the 2D surface states [60, 65].

Figure 11.

Magnetoresistance (MR) results of the Bi₂Te₃, Bi₂Se₃, and Bi₃Se₂Te thin films grown by PLD [9, 64, 67]. (a and e) The MR (B = ±1 T) of a 27-nm-thick Bi₂Te₃ and a 200-nm-thick Bi₃Se₂Te films grown on Al₂O₃ (0 0 0 1) substrates. (b and f) Variation in the extracted electron dephasing length L_φ and parameter −α of the films as a function of temperatures. (c and d) MR of the epitaxial Bi₂Se₃ films grown on Al₂O₃ (0 0 0 1) substrates at T_S = 250°C as a function of temperatures [64]. The solid green lines in (a) and (e) in low B are the theoretical predictions of 2D weak antilocalization (WAL) using Eq. (8).

The typical magnetoresistance (MR) results of some bismuth chalcogenides (i.e., Bi₂Te₃, Bi₂Se₃, and Bi₃Se₂Te) thin films grown by PLD are presented in Figure 11 [9, 64, 67]. The Bi₂Te₃, Bi₂Se₃, and Bi₃Se₂Te thin films were grown on Al₂O₃ (0 0 0 1) substrates using PLD at T_S of 225, 250, and 250°C, respectively [9, 64, 67]. The WAL effect which presents as a sharp increase in resistance when B increases in the low magnetic field B regime is clearly observed on the films. Figure 11a, b, e and f show MR curves at several temperatures (T) and the extracted α(T) and $L_{\varphi }(T)$ values using Eq. (8) for the 27-QL (~27 nm)-thick Bi₂Te₃ and 200-nm-thick Bi₃Se₂Te thin films [9, 67]. At 2 K, L_φ are 158.1 and 195.2 nm, and L_φ decreases monotonically with increasing T, obeying the power laws, as L_φ ~ T^−0.50 and L_φ ~ T^−0.79 for the Bi₂Te₃ and Bi₃Se₂Te films, respectively (Figure 11b and f). Theoretically, the result of L_φ ~ T^−0.50 observed in the 27-QL-thick Bi₂Te₃ thin film indicates the predominant electron-electron scattering in 2D weakly disordered systems. The L_φ ~ T^−0.79 is closed to the L_φ ~ T^−0.75 for 3D systems, if e-e scattering is the dominant dephasing source [9]. The electron screening effect in 3D system is more effective than that in 2D systems. However, the e-e scattering is strongly weakened with high carrier densities (n = 10²⁰ cm⁻³ for the Bi₃Se₂Te films [9]). In 3D disordered conductors (namely, in the bulk state), dephasing by electron-phonon (e-ph) scattering would be significant and dominant. The e-ph scattering also causes a faster decay rate on L_φ, that is, L_φ ~ T^−1.0 [68]. Additionally, the combination of 2D e-e scattering in the TSS thin layer and e-ph scattering in the bulk would further result in the L_φ ~ T^−0.79. Consequently, e-ph scattering could be the force of channel separation as temperatures increase. It is worthy of mentioning that the MR result is a signature (not conclusive evidence) for the presence of TSS on Bi₃Se₂Te films. Further theoretical calculations and experiments are needed for reaching a final conclusion.

Figure 11b and f also present the –α(T) results of the Bi₂Te₃ and Bi₃Se₂Te films. The –α values of the Bi₂Te₃ film increase with increasing T from 0.43 at 2 K to 0.45 at 10 K, indicating the existence of a single coherent transport channel (i.e., likely a single surface state) [61]. Meanwhile, the –α values of the 200-nm-thick Bi₃Se₂Te film increase with increasing T from 0.5 at 2 K to 0.85 at 10 K and to 0.71 at 15 K, suggesting increased channel separation with T [69]. In WAL, the independent phase-coherent channels occur when the carriers in one channel lose phase coherence before being scattered into the other channel.

In Figure 11c, the MR (B,T) of a 50-nm-thick Bi₂Se₃ film was measured with B perpendicular to the film plane and ranging from −9 T to +9 T. At low B and T, the distinctive dips of WAL are clearly observed. The WAL effect results from strong spin-orbit coupling, showing the absence of backscattering giving rise to the destructive interference between the two time reversal symmetry loops when there is no magnetic field [64]. Resistance increases sharply with increasing B because the quantum interference is destroyed and backscattering increases (Figure 11c). Figure 11d is the enlarged MR at low B (–2 T to +2 T), and it reveals the existence of WAL as a function of T in a discernible way [64]. WAL gradually weakens as temperature increases, eventually disappearing entirely at T = 48 K; thus, the dependence on B is quadratic-like at low field. The MR cusp feature at low B is broadened and finally disappears with increasing temperature owing to the decrease in the phase coherence length. In addition to the WAL effect, Figure 11c shows a 2D, non-saturating linear MR at high B, which usually occurs with several TI materials of Bi₂Se₃ [70], Bi₂Te₃ [71], Bi₂Te₂Se [72], and Bi₃Se₂Te [9]. Theoretical models propose that the linear MR can appear in the gapless linear-dispersive energy spectrum when only the first landau level is filled [73, 74], or in the presence of both the gapless linear spectrum and Landau level overlaps [75]. Noticeably, the WAL and linear MR simultaneously reflect the 3D contribution of spin-orbit coupling in bulk and the Dirac nature of the 2D surface states. Because the magneto transport is a bulk sensitive measurement, it remains a major challenge to directly probe the topological nature [64].

Figure 12.

(a) Temperature-dependent normalized ab resistivities (ρ/ρ_{300 K}) between 1.8 and 300 K of 46- and 200-nm-thick Bi₂Te₃ films. Upper inset: an EDS mapping image of a typical Bi-rich cluster. Lower inset: zoomed-in view of the ρ/ρ_{300 K} in the low temperature range. (b) ρ(T) in 1.75–6.0 K of the 200 nm film at various H_||c from 0 to 1 T. Inset: the onset T_c of the two superconducting transitions as a function of magnetic field. (c) Auger electron spectroscopy (AES) elemental depth profiling of a non-superconducting (46-nm-thick) and a superconducting (200-nm-thick) Bi₂Te₃ films. (d and e) The size distribution of Bi-rich clusters and Bi nanoclusters inside the clusters. (f) Schematics of the surface characteristics and a suggested superconducting mechanism in the Bi₂Te₃ films [53].

5.3. Proximity-induced superconductivities in Bi inclusions/bismuth chalcogenide thin films

Recent studies have shown a two-dimensional interface state between TIs and superconductors resulting from the superconducting proximity effect that supports Majorana fermions [76, 77]. Majorana fermions, novel particles which are their own antiparticles, can potentially be applied to topological quantum computing, which has motivated intense interest in TIs [53]. Koren et al. observed the local superconductivity in Bi₂Te₂Se and Bi₂Se₃ films below 2–3 K, which was naturally induced by small amounts of superconducting Bi inclusions or precipitations on the surface [78]. Moreover, Le et al. reported superconductivity at an onset critical temperature of approximately 3.1 K in a topological insulator 200-nm-thick Bi₂Te₃ thin film grown by pulsed laser deposition [53]. Indeed, Figure 12a shows the normalized resistivity ρ/ρ_{300 K} of a 46-nm-thick Bi₂Te₃ film (S₁) and a 200-nm-thick Bi₂Te₃ film (S₂) as functions of temperatures (T) from 1.8 to 300 K. Both films show a decrease in resistivity (ρ) with decreasing T in the range of 20–300 K, implying that the films exhibit weak metallic properties commonly seen in narrow band-gap semiconductors with high carrier concentrations [53]. Below 20 K (Figure 12a, the lower inset), the ρ/ρ_{300 K} of S₁ shows a gentle upturn because of the weak localization of electrons [7], whereas the ρ/ρ_{300 K} of S₂ reaches a plateau before dropping slightly at T_c1 ≈ 5.8 K and then sharply at T_c2 ≈ 3.1 K. Figure 12b further shows the H_||c-dependent ρ(T) of S₂ in the low T regime, where H_||c is the applied magnetic field along the c-axis of the film. At H_||c = 0, ρ drops abruptly by 8% below $T_{c2}^{onset}$ but does not go down to zero, even at T = 1.8 K. This nonzero ρ at low T indicates that the superconducting volume ratio is not 100%. The inset in Figure 12b shows that the $T_{c2}^{onset}$($T_{c1}^{onset}$) decreases from 3.1 (5.8) to 1.8 (5.4) K with increasing H_||c from 0 to 0.2 T, strongly indicating that both transitions are superconducting in nature.

The detailed investigations of S₂ strongly suggest the existence of superconducting Bi nanoclusters on the surface that induce the T_c1 ~ 5.8 K. EDS lateral elemental mapping revealed that the distributions of Te and Bi were not uniform, and many Bi-rich (47–54 at.%) clusters were visible (green color), as shown in the upper inset in Figure 12a, differing substantially from the uniform distribution and cluster-free surface observed in film S₁. The size distribution and the most probable size of Bi-rich clusters are in the range of 400–2400 nm and 560–772 nm, respectively. Intriguingly, a closer inspection reveals that Bi-rich clusters are composed by some Bi nanoclusters (or nanograins) with a size of 20–62 nm.

The Bi-rich environment on the film surface is confirmed by AES analysis (Figure 12c) [53]. This is because the vapor pressure of Te (at T_S = 300°C) is approximately 10⁵ times higher than that of Bi, and therefore, more Te atoms are re-evaporated from the 300°C substrates [79]. The loss of Te is more severe in film S₂ than in film S₁ (14.3% in S₂ and 4.5% in S₁ at depth Z = 0, Figure 12c) because of the six times longer deposition time of S₂ (60 min) compared to S₁ (10 min). In addition, the nonstoichiometric effect is strongly depth-dependent (Figure 12c). The Te/Bi ratio gradually increases toward the stoichiometric ratio of 3/2 in 200-nm-thick films or slightly exceeds it (Te-rich) in 46-nm-thick films when the depth (Z) of the films increases. Under such a sufficiently high surface concentration of Bi atoms, the Bi clusters precipitate and segregate readily on the S₂ surface to minimize overall free energy, as long as the substrate temperature of 300°C is higher than the 271°C melting point of Bi, as demonstrated in Figure 12f. Notably, the Bi clusters can only be observed in highly Bi-rich (14.3% at Z = 0) films (S₂) and not in low Bi-rich (4.5% at Z = 0) films (S₁), suggesting a critical Bi-rich concentration for Bi precipitation (separating a Bi phase) in a Bi₂Te₃ film [53]. The T_c1 at ~5.8 K found in our samples should be induced by the superconducting transition of the Bi nanoclusters, which is closely consistent with the T_c of 6.3 K for the surface Bi islands observed in Bi₂Te₂Se films [78]. The tiny resistivity drop at T_c1 = 5.8 K (by approximately 0.5%, Figure 12b) indicates that the amount of superconducting Bi nanoclusters in S₂ is likely small and, therefore, the Josephson coupling between these islands is extremely weak. Since the superconductivity of Bi nanoclusters survived until H_||c = 1.0 T (Figure 12b), the critical field of Bi nanoclusters is in between 0.3 and 1.0 T. This section demonstrates that natural defects generated during PLD growths, namely superconducting Bi nanoclusters or Bi inclusions, can substantially induce non-superconducting TI thin films (i.e., Bi₂Te₃, Bi₂Se₃, and Bi₂Te₂Se) into superconducting states at low temperatures.

6. Conclusion

This chapter provides the effects of ambient pressures and substrate temperatures in PLD growths on the structural-morphology, thermoelectric, nanomechanical, and magnetoresistance properties of bismuth chalcogenide thin films. The thermoelectric power factor of the stoichiometric Bi₂Te₃ films grown in the range of 220–340°C and P_Ar of 80 Pa reached remarkably high values, ranging between 18.2 ± 0.25 and 24.3 ± 0.44 μW cm⁻¹ K⁻². The optimal PF values were 5.54 ± 0.34 μW cm⁻¹ K⁻² for the layered hexagonal platelet Bi₂Se₃ films deposited at 300°C and P_He of 40 Pa and 8.3 μW cm⁻¹ K⁻² for the nanocrystalline Bi₃Se₂Te films deposited at 250°C and P_He of 40 Pa. We also reported the effects of P_He in PLD on nanomechanical properties (i.e., hardness and Young’s modulus) of Bi₂Te₃ and Bi₃Se₂Te thin films. It was observed that the hardness and Young’s modulus increased with increasing P_He, depending on the grain sizes following the inverse Hall-Petch effect for Bi₂Te₃ films grown at P_He ≤ 2.0 × 10⁻³ Torr and following the Hall-Petch relationship for Bi₃Se₂Te grown at P_He of 2.0 × 10⁻⁵ to 6.5 × 10⁻¹ Torr. PLD has been successfully employed to grow epitaxially bismuth chalcogenide thin films on large-misfit substrates, for example, Bi₂Te₃/SrTiO₃ (1 0 0), Bi₂Se₃/Al₂O₃ (0 0 0 1), and Bi₃Se₂Te/Al₂O₃ (0 0 0 1). The magnetotransport studies show that the bismuth chalcogenide thin films such as Bi₂Te₃, Bi₂Se₃, and Bi₃Se₂Te films present a two-dimensional weak antilocalization effect in a low magnetic field (B) regime and linear magnetoresistance in a high B regime, which could be attributed to the topological insulator surface states. Furthermore, proximity-induced superconductivities in Bi₂Te₃ thin films have an onset T_c of approximately 3.1 K, evidently induced by Bi inclusions (nanoclusters with onset T_c at 5.8 K) segregated on the surface of films.

Acknowledgments

Financial support from Vietnam National Foundation for Science and Technology Development (NAFOSTED) under Grant number 103.99–2015.17, the Ministry of Science and Technology, Taiwan under Contract Nos.: 103-2923-M-009-001-MY3, 103-2628-M-009-002-MY3, 103-2119-M-009-004-MY3, and the Ministry of Education (MOE-ATU plan at National Chiao Tung University) are gratefully acknowledged.