Literature DB >> 32548485

Enhancing the Thermoelectric Properties of Misfit Layered Sulfides (MS)1.2+q (NbS2) n (M = Gd and Dy) through Structural Evolution and Compositional Tuning.

Aleksandr V Sotnikov1, Priyanka Jood2,3, Michihiro Ohta2,3.   

Abstract

The misfit monolayered sulfides, (GdS)1.20NbS2, (DyS)1.22NbS2, (Gd0.1Dy0.9S)1.21NbS2, (Gd0.2Dy0.8S)1.21NbS2, and (Gd0.5Dy0.5S)1.21NbS2 and the misfit bilayered sulfide (GdS)0.60NbS2 were synthesized via sulfurization under flowing CS2/H2S gas and consolidated by pressure-assisted sintering. The thermoelectric properties of the monolayered and bilayered sulfides perpendicular (in-plane) and parallel (out-of-plane) to the pressing direction were investigated over a temperature range of 300-873 K. The crystal grains in all the sintered samples were preferentially oriented perpendicular to the pressing direction, which resulted in highly anisotropic electrical and thermal transport properties. All the sintered samples exhibited degenerate n-type semiconductor-like behavior, leading to a large thermoelectric power factor. The misfit layered structure yielded low lattice thermal conductivity. The evolution of the monolayered structures into bilayered structures affected their thermoelectric properties. The thermoelectric figure of merit (ZT) of monolayered (GdS)1.20NbS2 was higher than that of bilayered (GdS)0.60NbS2 due to the larger power factor and lower lattice thermal conductivity of (GdS)1.20NbS2. The lattice thermal conductivity of the monolayered sulfide was lower in (Gd x Dy1-x S)1.2+q NbS2 solid solutions. The large power factor and low lattice thermal conductivity allowed a ZT value of 0.13 at 873 K in (Gd0.5Dy0.5S)1.21NbS2 perpendicular to the pressing direction.
Copyright © 2020 American Chemical Society.

Entities:  

Year:  2020        PMID: 32548485      PMCID: PMC7288565          DOI: 10.1021/acsomega.0c00908

Source DB:  PubMed          Journal:  ACS Omega        ISSN: 2470-1343


Introduction

The significant increase in global energy consumption has prompted researchers to develop new and efficient ways to convert natural and waste heat into electrical energy. Using thermoelectrics to generate electricity from waste heat, such as automotive and industrial heat emissions, is a promising way to make energy generation more sustainable.[1] The efficiency of a thermoelectric material depends on its thermoelectric figure of merit (ZT), which is equal to S2Tρ–1κtot–1. S, κtot, ρ, and T denote the Seebeck coefficient of the material, its total thermal conductivity, its electrical resistivity, and the absolute temperature, respectively. The κtot value is the sum of the lattice (κlat) and electronic (κel) thermal conductivities. Developing effective and economical thermoelectric materials with high ZT values is an ongoing challenge for researchers because realizing a high ZT is generally difficult.[2−12] The fundamental parameters are interrelated, and the requirements of high electrical conductivity and low thermal conductivity are often contradictory.[13,14] Extensive research has been carried out on various layered chalcogenides, such as SnS,[15,16] SnSe,[17−19] Ag9GaS6,[14] PbSe-Bi2Se3,[20] M(Bi/Sb)2S, or M(Bi/Sb)2Se compounds.[21] A good example of a high-performance layered chalcogenide is SnSe with different dopants, which demonstrate a state-of-the-art ZT value of ∼1.8–2.7 in a temperature range of 750–900 K. One of the greatest difficulties in achieving a high ZT is tuning electron and phonon transport in the material separately. The phonon glass–electron crystal (PGEC) concept has proven to be a key strategy for developing high-ZT thermoelectric materials.[22−24] The phonon glass region provides the disorder needed to scatter phonons and minimize κtot without affecting carrier mobility in the electron crystal region, which is necessary to achieve a high power factor (S2ρ–1). The misfit layered oxides NaCoO2 and [Ca2CoO3]CoO2 are good examples of materials that exhibit the PGEC behavior.[25−29] Because misfit layered oxides contain several subsystems, the subsystems can be tailored individually. One of the subsystems can be a good electrical conductor (large S2ρ–1), whereas the other can be a poor thermal conductor (low κtot). In other words, a layered structure makes it possible to control the phonon and electron transport properties separately. A well-designed multilayered structure will thus exhibit good thermoelectric properties. However, the high electrical resistivity of these oxides imposes a limit on ZT. Their high electrical resistivity is due to the high electronegativity of oxygen and the low covalency of the oxides.[30] The electronegativity of sulfur is lower than that of oxygen, and bonding in sulfides is highly covalent.[31] Layered sulfides can thus be expected to exhibit better electrical conductivity.[25,30,32−42] The naturally modulated structure of misfit layered sulfides makes them promising candidates as high-temperature thermoelectric materials.[43] The general formula for these materials is (MS)1.2+(NbS2), where M = Pb, Bi, Sn, Sb, or a rare earth element (T = Ti, V, Cr, Nb, and Ta). The PGEC behavior of these sulfides provides a tremendous opportunity for ZT enhancement. The TiS2 host layer has a CdI2-type structure, which contains pathways that enable high charge carrier mobility. The MS layer has an intercalated NaCl-type structure, which provides disorder to scatter phonons. The ZT of the (LaS)1.14NbS2 system was optimized by varying the La content.[33] Improved structural ordering and textured grain growth were observed when x = 1.05, and the power factor and ZT of the corresponding sample were enhanced by up to 30 and 25%, respectively. Moreover, PGEC behavior was reported in the thermoelectric properties of misfit layered selenide (SnSe)1.16NbSe2[44] and SnSe2.[45] Effective scattering of heat-carrying phonons should reduce the lattice thermal conductivity (κlat) of a thermoelectric material and increase its ZT. A reduction in the κlat of γ-Dy2S3 was achieved by introducing paramagnetic rare earth ions (Gd3+), which scatter phonons effectively without modifying the electrical transport properties of the material. The κtot of a Gd0.2Dy0.8S1.5- solid solution is 20–25% lower than that of γ-Gd2S3,[46] which suggests that the ZT of a misfit layered system can be enhanced by introducing strain into its crystal structure. With this in mind, we introduced different rare earth elements into the [LnS] subsystem. In this work, we focused on the misfit monolayered sulfides (GdDy1-S)1.2+NbS2 (x = 0, 0.1, 0.2, 0.5, and 1.0; z = 1.2 + q; q = 0.00–0.02) and the misfit bilayered sulfide (GdS)0.60NbS2. The powders were synthesized by sulfurizing the corresponding oxides under flowing CS2/H2S gas.[47] The samples were then consolidated by pressure-assisted sintering to grow highly oriented grains. The textured microstructures had highly anisotropic electrical and thermal transport properties. The effect of monolayer-to-bilayer structural evolution on the thermoelectric properties of the systems was investigated. The κlat in the monolayered sulfide was lower in (GdDy1-S)1.2+NbS2 solid solutions. The ZT value of (Gd0.5Dy0.5S)1.21NbS2 was 0.13 at 873 K perpendicular to the pressing direction.

Results and Discussion

Crystal Structures and Microstructures

X-ray diffraction (XRD) analysis was performed to investigate the crystal phases in (GdS)1.20NbS2, (DyS)1.22NbS2, (Gd0.1Dy0.9S)1.21NbS2, (Gd0.2Dy0.8S)1.21NbS2, (Gd0.5Dy0.5S)1.21NbS2, and (GdS)0.60NbS. The powder XRD patterns of the crushed sintered samples are shown in Figure a, and the out-of-plane XRD patterns of the sintered samples are shown in Figure b. The XRD patterns of the (GdS)1.20NbS2, (DyS)1.22NbS2, (Gd0.1Dy0.9S)1.21NbS2, (Gd0.2Dy0.8S)1.21NbS2, and (Gd0.5Dy0.5S)1.21NbS2 crushed powders were similar to the reported diffraction patterns of (GdS)1.20NbS2 and (DyS)1.22NbS2.[35] The XRD patterns (powders and out-of-plane sintered samples) in a larger scale are presented in Figures S1 and S2 in the Supporting Information. These results indicate that all of the synthesized sulfides have the same crystal structure. XRD data for (GdS)0.60NbS2 could not be found in the literature. No reflections corresponding to Gd or Nb binary sulfides are observed in the experimental XRD pattern. More detailed data about crystal structure analysis and the evidence of solid solutions formation is provided elsewhere.[48] The formation of (GdS)0.60NbS2 was inferred from a high-resolution transmission electron microscopy image.[48] For a more accurate determination, separate space groups in a (3 + 1)-dimensional superspace group should be considered for the LnS and NbS2 subsystems.[49,50] The [Ln2S2] (Ln = RE) subsystem has a monoclinic unit cell with C2 symmetry, whereas the [NbS2] subsystem has F2 symmetry. Both subsystems modulate each other incommensurately.[43,51]
Figure 1

(a) Powder X-ray diffraction (XRD) patterns of sintered (GdS)1.20NbS2, (DyS)1.22NbS2, (Gd0.1Dy0.9S)1.21NbS2, (Gd0.2Dy0.8S)1.21NbS2, (Gd0.5Dy0.5S)1.21NbS2, and (GdS)0.60NbS2. (b) Out-of-plane XRD patterns of sintered (GdS)1.20NbS2, (DyS)1.22NbS2, (Gd0.1Dy0.9S)1.21NbS2, (Gd0.2Dy0.8S)1.21NbS2, (Gd0.5Dy0.5S)1.21NbS2, and (GdS)0.60NbS2.

(a) Powder X-ray diffraction (XRD) patterns of sintered (GdS)1.20NbS2, (DyS)1.22NbS2, (Gd0.1Dy0.9S)1.21NbS2, (Gd0.2Dy0.8S)1.21NbS2, (Gd0.5Dy0.5S)1.21NbS2, and (GdS)0.60NbS2. (b) Out-of-plane XRD patterns of sintered (GdS)1.20NbS2, (DyS)1.22NbS2, (Gd0.1Dy0.9S)1.21NbS2, (Gd0.2Dy0.8S)1.21NbS2, (Gd0.5Dy0.5S)1.21NbS2, and (GdS)0.60NbS2. Strongly enhanced basal (00l) reflections in out-of-plane XRD are observed in the patterns of the sintered (GdS)1.20NbS2, (Gd0.1Dy0.9S)1.21NbS2, (Gd0.2Dy0.8S)1.21NbS2, (Gd0.5Dy0.5S)1.21NbS2, and (DyS)1.22NbS2 samples. This indicates that the crystalline c axis is preferentially oriented in the out-of-plane pressing direction. The degree of (00l) orientation is referred to as the Lotgering factor (f). It was calculated using the relation f = [(P – P0)/(1 – P0)], where P = ∑I(00l)/∑I(hkl). ∑I(00l) and ∑I(hkl) are the sums of the intensities of the (00l) and (hkl) reflections, respectively, and P0 is the P value of a randomized powder sample.[52] The Lotgering factors of perfectly oriented and randomly oriented samples are f = 1 and f = 0, respectively. In this study, the Bragg reflections in a range from 10° to 80° were used to calculate f. The f values determined from the out-of-plane XRD patterns of (GdS)1.20NbS2, (Gd0.1Dy0.9S)1.21NbS2, (Gd0.2Dy0.8S)1.21NbS2, (Gd0.5Dy0.5S)1.21NbS2, and (DyS)1.22NbS2 are 0.09, 0.12, 0.12, 0.11, and 0.11, respectively. The f value of (GdS)0.60NbS2 is 0.05. This indicated that the misfit monolayered sulfides have a preferential (00l) crystal orientation. Figure is the scanning electron microscopy (SEM) image of the fractured (GdS)1.20NbS2 sintered compact. SEM images of several other sintered compacts are shown in Figure S3 (Supporting Information). The platelet-like grains ranged from ∼5 to 10 μm in length. The grains are preferentially oriented in the direction perpendicular to the pressure applied during sintering. The notable layering of the grains is due to the anisotropic nature of the misfit layered sulfides.[53] The intralayer atomic bonding is much stronger than the interlayer bonding in the c direction perpendicular to the (00l) basal plane, and the pressure applied during sintering results in a preferential (00l) orientation.
Figure 2

Scanning electron microscopy image of a fractured (GdS)1.20NbS2 sintered compact.

Scanning electron microscopy image of a fractured (GdS)1.20NbS2 sintered compact.

Thermoelectric Properties

The temperature dependencies of the Seebeck coefficient (S), electrical resistivity (ρ), and thermoelectric power factor (S2ρ–1) for monolayered (GdS)1.20NbS2 and bilayered (GdS)0.60NbS2 measured perpendicular (in-plane) and parallel (out-of-plane) to the pressing direction are shown in Figure . The subscripts “in” and “out” denote the in-plane thermoelectric parameters and thermoelectric parameters along the sintering pressing direction, respectively. In both systems, Sin, Sout, ρin, and ρout increase monotonically with temperature over a range from 300 to 873 K, which is consistent with the trend of a degenerate semiconductor. The Sin value of (GdS)0.60NbS2 increases from ∼15 μV K–1 at 300 K to ∼40 μV K–1 at 873 K, and the ρin value increases from ∼8 μΩ m at 300 K to ∼17 μΩ m at 873 K. The sign of Sin is positive, which confirms the p-type carrier transport.
Figure 3

Temperature dependencies of the thermoelectric parameters of (GdS)1.20NbS2 and (GdS)0.60NbS2: (a) Seebeck coefficient (S), (b) electrical resistivity (ρ), and (c) thermoelectric power factor (S2ρ–1). In-plane values were measured perpendicular to the pressing direction, and out-of-plane values were measured parallel to the pressing direction.

Temperature dependencies of the thermoelectric parameters of (GdS)1.20NbS2 and (GdS)0.60NbS2: (a) Seebeck coefficient (S), (b) electrical resistivity (ρ), and (c) thermoelectric power factor (S2ρ–1). In-plane values were measured perpendicular to the pressing direction, and out-of-plane values were measured parallel to the pressing direction. In both systems, ρin is smaller than ρout. This is due to the anisotropic crystal structures and microstructures of the systems, which can be seen in Figure and Figure S3 in the Supporting Information. In (GdS)1.20NbS2, the ρin value is ∼18 μΩ m, whereas the ρout value is ∼40 μΩ m at 873 K. The higher values of ρout are primarily due to more pronounced scattering of the charge carriers at the interfaces between the [LnS] and [NbS2] layers and between the platelet-like particles. Moreover, the S values of all the samples are anisotropic over the entire temperature range. For example, the Sin value of (GdS)1.20NbS2 is ∼70 μV K–1 at 873 K, whereas the Sout value is ∼29 μV K–1. The anisotropy in S is due to the anisotropic band structures of the systems. An anisotropic band structure can result in larger values of m* (where m* is the effective mass of charge carrier) in the in-plane direction.[33] The anisotropic and degenerate semiconductor-like properties of the misfit sulfides prepared in this study are consistent with those reported previously for (LaS)1.14NbS2.[34] The S and ρ values of the bilayered misfit sulfide (GdS)0.60NbS2 are much less anisotropic than those of (GdS)1.20NbS2 (see Figure ). For example, the ρin values of (GdS)0.60NbS2 and (GdS)1.20NbS2 at 300 K were ∼8 and ∼16 μΩ m, respectively. There is a large difference in the Sin between the two compounds, with (GdS)0.60NbS2 having almost twice smaller values (∼15 μV K–1 at 300 K and ∼40 μV K–1 at 873 K) compared to (GdS)1.20NbS2 (∼30 μV K–1 at 300 K and ∼70 μV K–1 at 873 K). In (LnS)1.2+(NbS2) misfit compounds, the conduction is governed by the number of electrons being transferred from the LnS to NbS2 system. A hole number of q holes/Nb corresponds to a donation of (1–q) electrons/Nb from the LnS subsystem to NbS2 subsystem. In the case of a bilayered compound, which contains two NbS2 layers, a single electron is donated to two NbS2 subsystems, which reduces the number of electron/Nb and increases the number of holes/Nb.[43] This results in an overall increase in hole concentration in the bilayered system, which explains the low S of (GdS)0.60NbS2. The temperature dependencies of the in-plane and out-of-plane S, ρ, and S2ρ–1 of (GdS)1.20NbS2, (DyS)1.22NbS2, (Gd0.1Dy0.9S)1.21NbS2, (Gd0.2Dy0.8S)1.21NbS2, and (Gd0.5Dy0.5S)1.21NbS2 are shown in Figure . The S and ρ of each sample in the in-plane and out-of-plane directions are highly anisotropic. At 873 K, the ρin value of (Gd0.5Dy0.5S)1.21NbS2 is ∼17 μΩ m, whereas the ρout value is ∼36 μΩ m. Its Sin and Sout values at 873 K are ∼76 and ∼48 μV K–1, respectively. Substitution of Gd with Dy has a little effect on S and ρ. The highest Sin (∼76 μV K–1) and lowest ρin (∼36 μΩ m) are observed for (Gd0.5Dy0.5S)1.21NbS2 and (Gd0.1Dy0.9S)1.21NbS2, respectively, at 873 K. The temperature dependence of the thermoelectric power factor (S2ρ–1) of each (GdDy1 – S)1.2+NbS2 compound is shown in Figure c. The (S2ρ–1)in of each sample is much greater than its (S2ρ–1)out. The highest (S2ρ–1)in value of ∼340 μW K–2 m–1 at 873 K is observed in the sample with x = 0.5 because it has the smallest ρin. There are some investigations of thermoelectric properties in misfit layered sulfides.[33−35] We compare the thermoelectric properties of our samples with those reported in previous studies (Table ).
Figure 4

Temperature dependencies of the thermoelectric parameters of (GdS)1.20NbS2, (DyS)1.22NbS2, (Gd0.1Dy0.9S)1.21NbS2, (Gd0.2Dy0.8S)1.21NbS2, and (Gd0.5Dy0.5S)1.21NbS2 in the (GdDy1-S)1.2+NbS2 system: (a) Seebeck coefficient (S), (b) electrical resistivity (ρ), and (c) thermoelectric power factor (S2ρ–1). The in-plane values were measured perpendicular to the pressing direction, and the out-of-plane values were measured parallel to the pressing direction.

Table 1

Thermoelectric Properties of (GdS)1.20NbS2, (Gd0.5Dy0.5S)1.21NbS2, (DyS)1.22NbS2, and (GdS)0.60NbS2 and Important Milestones Achieved in Previous Studies

sampledirectionT (K)ρ (μΩ·m)S (μV·K–1)κtot (W·m–1·K–1)κlat(W·m–1·K–1)S2·ρ–1(μW·m–1·K–2)ZTref
(GdS)1.20NbS2in-plane3007303.32.21280.01 
87318702.71.52810.09 
out-of-plane3001611.91.410.00 
87340291.71.1180.01 
(Gd0.5Dy0.5S)1.21NbS2in-plane3006293.11.91460.02 
87317762.31.13400.13 
out-of-plane3001452.11.650.00 
87336481.91.2600.03 
(DyS)1.22NbS2in-plane3006322.91.61860.02 
87315712.51.13490.12 
out-of-plane3001611.81.330.00 
87342351.71.2260.01 
(GdS)0.60NbS2in-plane3008153.32.4260.00 
87317403.11.9970.03 
out-of-plane3001172.61.850.00 
87321332.41.4540.02 
(LaS)1.14NbS2in-plane95022832.00.93160.15(33)
(La2S2)0.62NbS2in-plane3001222  50 (35)
(Yb2S2)0.62NbS2in-plane30019600.800.42000.1(35)
Cu0.1TiS2in-plane8002–1421.800.810600.47(36)
TiS2–4%AgSnSe2in-plane70012–2302.181.015500.8(54)
Temperature dependencies of the thermoelectric parameters of (GdS)1.20NbS2, (DyS)1.22NbS2, (Gd0.1Dy0.9S)1.21NbS2, (Gd0.2Dy0.8S)1.21NbS2, and (Gd0.5Dy0.5S)1.21NbS2 in the (GdDy1-S)1.2+NbS2 system: (a) Seebeck coefficient (S), (b) electrical resistivity (ρ), and (c) thermoelectric power factor (S2ρ–1). The in-plane values were measured perpendicular to the pressing direction, and the out-of-plane values were measured parallel to the pressing direction. The temperature dependencies of the total (κtot) and lattice (κlat) thermal conductivities of monolayered (GdS)1.20NbS2 and bilayered (GdS)0.60NbS2 in the in-plane and out-of-pane directions are shown in Figure a. κtot is the sum of κlat and the electronic thermal conductivity (κel). κel can be calculated using the Wiedemann–Franz law κel = LTρ–1, where L is the Lorentz number (2.44 × 10–8 W Ω K–2).[33,34] The in-plane and out-of-plane κlat values of (GdS)0.60NbS2 are 21 and 24% higher, respectively, than those of (GdS)1.20NbS2 at 873 K. The higher value of κlat in bilayered (GdS)0.60NbS2 could be attributed to less numbers of GdS/NbS2 interfaces, which play an important role in efficiently scattering phonons. Similar results have been reported for homologous layered compounds, such as (PbSe)5(Bi2Se3)3 (m = 1, 2, and 3).[55] The in-plane κlat of each sample is higher than the out-of-plane κlat at 300 K (Figure c). This is due to a higher degree of phonon scattering at the interfaces between the [LnS] and [NbS2] layers and between the platelet-like particles.
Figure 5

(a) Comparison of κtot and κlat in (GdS)1.20NbS2 and (GdS)0.60NbS2. Temperature dependencies of the (b) total (κtot) and (c) lattice (κlat) thermal conductivities of the (GdDy1-S)1.2+NbS2 solid solutions (GdS)1.20NbS2, (DyS)1.22NbS2, (Gd0.1Dy0.9S)1.21NbS2, (Gd0.2Dy0.8S)1.21NbS2, and (Gd0.5Dy0.5S)1.21NbS2.

(a) Comparison of κtot and κlat in (GdS)1.20NbS2 and (GdS)0.60NbS2. Temperature dependencies of the (b) total (κtot) and (c) lattice (κlat) thermal conductivities of the (GdDy1-S)1.2+NbS2 solid solutions (GdS)1.20NbS2, (DyS)1.22NbS2, (Gd0.1Dy0.9S)1.21NbS2, (Gd0.2Dy0.8S)1.21NbS2, and (Gd0.5Dy0.5S)1.21NbS2. The temperature dependencies of the total (κtot) and lattice (κlat) thermal conductivities of the monolayered misfit samples (GdS)1.20NbS2, (DyS)1.22NbS2, (Gd0.1Dy0.9S)1.21NbS2, (Gd0.2Dy0.8S)1.21NbS2, and (Gd0.5Dy0.5S)1.21NbS2 in the in-plane and out-of-plane directions are shown in Figure , respectively. The misfit layered structure of each sample affords a low κlat, which results in a low κtot. For example, the out-of-plane κlat value in (Gd0.5Dy0.5S)1.21NbS2 is ∼1.6 W m–1 K–1, and the out-of-plane κtot value is ∼2.1 W m–1 K–1 at 300 K. The κlat is lower in solid solutions of the (GdDy1-S)1.2+NbS2 monolayered sulfides. The lowest in-plane κlat value of ∼1.1 W m–1 K–1 is observed in (Gd0.5Dy0.5S)1.21NbS2 at 873 K.

Thermoelectric Figure of Merit

The dependence of the figure of merit (ZT) on temperature is illustrated in Figure . The ZT values of all the samples increases monotonically with temperature, and ZT in the in-plane direction is much larger than ZT in the out-of-plane direction due to larger (S2ρ–1)in values. For example, the in-plane ZT value of (Gd0.2Dy0.8S)1.21NbS2 at 873 K is ∼0.12, whereas its ZTout value is ∼0.02. The ZT value of (GdS)1.20NbS2 in the in-plane direction at 873 K is 71% higher than that of (GdS)0.60NbS2 (Figure a). This is because the in-plane κtot is lower in (GdS)1.20NbS2 (Figure a), and its in-plane power factor (S2ρ–1) is larger (Figure c). The (Gd0.5Dy0.5S)1.21NbS2 sample shows the highest in-plane ZT value of ∼0.13 (Figure b), because it has the lowest in-plane κtot (Figure b) and a large in-plane S2ρ–1 value (Figure c). Our ZT is close to that of other misfit sulfides (LaS)1+NbS2 and (LaS)1+CrS2 at a temperature range of 300–873 K (Table ).[33,34]
Figure 6

Temperature dependence of the thermoelectric figure of merit (ZT): (a) (GdS)1.20NbS2 and (GdS)0.60NbS2; (b) (GdS)1.20NbS2, (DyS)1.22NbS2, (Gd0.1Dy0.9S)1.21NbS2, (Gd0.2Dy0.8S)1.21NbS2, and (Gd0.5Dy0.5S)1.21NbS2.

Temperature dependence of the thermoelectric figure of merit (ZT): (a) (GdS)1.20NbS2 and (GdS)0.60NbS2; (b) (GdS)1.20NbS2, (DyS)1.22NbS2, (Gd0.1Dy0.9S)1.21NbS2, (Gd0.2Dy0.8S)1.21NbS2, and (Gd0.5Dy0.5S)1.21NbS2.

Conclusions

Misfit sulfides (GdDy1-S)1.2+(NbS2) (x = 0, 0.1, 0.2, 0.5, and 1.0; z = 1.2 + q; q = 0.00–0.02; n = 1 and 2) were synthesized via sulfurization under flowing CS2/H2S gas. Crystal grains in all of the sintered samples are preferentially oriented perpendicular to the pressing direction, which results in highly anisotropic electrical and thermal transport properties. The in-plane power factors and lattice thermal conductivity of both the monolayered and bilayered sulfides are higher than the out-of-plane values over the entire experimental temperature range. This was because charge carrier and phonon scattering are more pronounced at the interfaces between the [LnS] and [NbS2] layers and between the platelet-like particles. The ZT of monolayered (GdS)1.20NbS2 in the in-plane direction is larger than that of bilayered (GdS)0.60NbS2. This is because (GdS)1.20NbS2 has a larger power factor and lower lattice thermal conductivity (κlat). The lowest in-plane κlat value (∼1.1 W m–1 K–1) is observed in (Gd0.5Dy0.5S)1.21NbS2 at 873 K. The (Gd0.5Dy0.5S)1.21NbS2 solid solution shows the highest ZT value (0.13) at 873 K in the in-plane direction.

Experimental Section

Sulfurization and Sintering

We prepared the misfit layered sulfides (GdS)1.20NbS2, (DyS)1.22NbS2, (Gd0.1Dy0.9S)1.21NbS2, (Gd0.2Dy0.8S)1.21NbS2, (Gd0.5Dy0.5S)1.21NbS2, and (GdS)0.60NbS2 for the study. Commercially available 99.99% pure Gd2O3, Dy2O3, and Nb2O5 (Sibmetalltorg, Russia) were used as starting materials for sulfurization. Stoichiometric quantities of the oxide powders were thoroughly mixed to obtain binary oxide powders. For example, 3.30 g of Dy2O3 and 1.96 g of Nb2O5 were used to prepare ∼5 g of a binary oxide powder. The powders were placed in quartz boats and set in a quartz reaction tube after purging with Ar gas. The powders were heated to 1073 K at a rate of 10 K min–1 under flowing 1:1 CS2/H2S in Ar gas and then sulfurized at 1073 K for 6 h. They were then cooled to room temperature at a rate of 10 K min–1 under flowing Ar. The flow rate of Ar gas was fixed at 10 mL min–1. The obtained powders were ground well and sulfurized again under the same conditions at 1073 K for an additional 6 h to improve their homogeneity. The sulfurized powders were then placed in quartz tubes. The tubes were sealed, evacuated to 4 × 10–3 Pa, then placed in a furnace, and heated to 1323 K at a rate of 10 K min–1. The powders were annealed for 24 h and then cooled at a rate of 10 K min–1 to further homogenize them. The sample powders were placed in graphite dies of 10 mm in diameter, which were then inserted into an FUT-17000 sintering apparatus (Tokyo Vacuum, Japan) and heated at a rate of 10 K min–1. The samples were sintered for 2 h at 1223 K with 70 MPa uniaxial pressure under vacuum (7 × 10–3 Pa). They were cooled at a rate of 20 K min–1 to obtain high-density oriented sintered compacts, which were cut into bars and plates for electrical and thermal transport measurements.

Chemical Analysis

The starting oxides and obtained products were analyzed via inductively coupled plasma atomic emission spectroscopy (ICP-AES) to assess their chemical purity. The impure contents in the samples were determined using a PGS-2 spectrometer (Carl Zeiss Jena, Germany) with a direct current arc excitation source (13 A). The modernized PGS-2 spectrograph was equipped with 900 pcs mm–1 grating and a photoelectric spectra recorder. The ICP-AES results showed that the purity of the obtained samples was high, which enabled us to exclude the influence of impurities on their thermoelectric properties (Table S1, Supporting Information).

X-ray Diffraction (XRD) Analysis

The crystal structures of the powders and the sintered compacts were examined using an XRD-7000 diffractometer (Shimadzu, Japan) equipped with a Cu Kα radiation source from (2θ) 10° to 80°. The crystal orientations of the sintered samples were analyzed at room temperature from (2θ) 10° to 80° using a MiniFlex 600 powder XRD (Rigaku, Japan) equipped with a Cu Kα radiation source.

Scanning Electron Microscopy (SEM)

The microstructures of the sintered compacts were observed using a JSM-6610LV scanning electron microscope (JEOL, Japan) at 20 kV.

Electrical Transport Measurements

The Seebeck coefficients and electrical resistivities of the sintered compacts were measured simultaneously using temperature differential and four-probe methods, respectively, on a ZEM-3 system (Ulvac-Riko, Japan). Measurements were performed perpendicular (in-plane) and parallel (out-of-plane) to the pressing direction in a He atmosphere in a temperature range of 300–973 K. The dimensions of the bars used for in-plane measurements were typically ∼3 × 2 × 10 mm3, whereas those of the bars used for out-of-plane measurements were ∼3 × 2 × 7 mm3. The heating and cooling cycles provided reproducible Seebeck coefficients and electrical resistivity values for all of the sintered compacts. The uncertainties of both the Seebeck coefficient and electrical conductivity measurements were ∼5%.

Thermal Transport Measurements

The total thermal conductivity (κtot) of each sintered compact was calculated using κtot= DCPd, where d, D, and CP are the density, thermal diffusivity, and heat capacity of the sintered compact, respectively. The thermal diffusivity was measured directly using an LFA 457 MicroFlash laser flash apparatus (Netzsch, Germany) from 300 to 973 K under an Ar flow of 100 mL min–1. The heat capacities of the compacts were determined indirectly using an LFA 457 with a Pyroceram 9606 reference standard. The coins used for the out-of-plane measurements were typically ∼10 mm in diameter and ∼2 mm thick. Those used for the in-plane measurements were ∼6 × 6 mm2 square plates that were 2 mm thick. The densities of the sintered compacts were determined by Archimedes’ method using an AccuPyc II 1340 pycnometer (Shimadzu, Japan). The estimated relative uncertainty of the thermal conductivity measurements was within 6%. The combined relative uncertainty for all of the measurements to determine ZT was approximately 11%.
  18 in total

1.  Complex thermoelectric materials.

Authors:  G Jeffrey Snyder; Eric S Toberer
Journal:  Nat Mater       Date:  2008-02       Impact factor: 43.841

2.  Cooling, heating, generating power, and recovering waste heat with thermoelectric systems.

Authors:  Lon E Bell
Journal:  Science       Date:  2008-09-12       Impact factor: 47.728

3.  Correction to Pb7Bi4Se13: a lillianite homologue with promising thermoelectric properties.

Authors:  Alan Olvera; Guangsha Shi; Honore Djieutedjeu; Alexander Page; Ctirad Uher; Emmanouil Kioupakis; Pierre F P Poudeu
Journal:  Inorg Chem       Date:  2015-04-01       Impact factor: 5.165

4.  Thermoelectric power generation: from new materials to devices.

Authors:  Gangjian Tan; Michihiro Ohta; Mercouri G Kanatzidis
Journal:  Philos Trans A Math Phys Eng Sci       Date:  2019-07-08       Impact factor: 4.226

Review 5.  Compromise and Synergy in High-Efficiency Thermoelectric Materials.

Authors:  Tiejun Zhu; Yintu Liu; Chenguang Fu; Joseph P Heremans; Jeffrey G Snyder; Xinbing Zhao
Journal:  Adv Mater       Date:  2017-03-06       Impact factor: 30.849

Review 6.  Multi-Scale Microstructural Thermoelectric Materials: Transport Behavior, Non-Equilibrium Preparation, and Applications.

Authors:  Xianli Su; Ping Wei; Han Li; Wei Liu; Yonggao Yan; Peng Li; Chuqi Su; Changjun Xie; Wenyu Zhao; Pengcheng Zhai; Qingjie Zhang; Xinfeng Tang; Ctirad Uher
Journal:  Adv Mater       Date:  2017-01-23       Impact factor: 30.849

7.  Ultralow thermal conductivity and high thermoelectric figure of merit in SnSe crystals.

Authors:  Li-Dong Zhao; Shih-Han Lo; Yongsheng Zhang; Hui Sun; Gangjian Tan; Ctirad Uher; C Wolverton; Vinayak P Dravid; Mercouri G Kanatzidis
Journal:  Nature       Date:  2014-04-17       Impact factor: 49.962

Review 8.  Thinking Like a Chemist: Intuition in Thermoelectric Materials.

Authors:  Wolfgang G Zeier; Alex Zevalkink; Zachary M Gibbs; Geoffroy Hautier; Mercouri G Kanatzidis; G Jeffrey Snyder
Journal:  Angew Chem Int Ed Engl       Date:  2016-04-25       Impact factor: 15.336

Review 9.  High-Performance Thermoelectric SnSe: Aqueous Synthesis, Innovations, and Challenges.

Authors:  Xiao-Lei Shi; Xinyong Tao; Jin Zou; Zhi-Gang Chen
Journal:  Adv Sci (Weinh)       Date:  2020-02-13       Impact factor: 16.806

10.  High thermoelectric performance in low-cost SnS0.91Se0.09 crystals.

Authors:  Wenke He; Dongyang Wang; Haijun Wu; Yu Xiao; Yang Zhang; Dongsheng He; Yue Feng; Yu-Jie Hao; Jin-Feng Dong; Raju Chetty; Lijie Hao; Dongfeng Chen; Jianfei Qin; Qiang Yang; Xin Li; Jian-Ming Song; Yingcai Zhu; Wei Xu; Changlei Niu; Xin Li; Guangtao Wang; Chang Liu; Michihiro Ohta; Stephen J Pennycook; Jiaqing He; Jing-Feng Li; Li-Dong Zhao
Journal:  Science       Date:  2019-09-27       Impact factor: 47.728

View more

北京卡尤迪生物科技股份有限公司 © 2022-2023.