Enhanced Average Thermoelectric Figure of Merit ... - ACS Publications

Mar 3, 2017 - A thermoelectric figure of merit as high as 1.98 and an enhancement of the average thermoelectric figure of merit by 18% are achieved fo...
0 downloads 0 Views 3MB Size
Download PDF
Research Article www.acsami.org

Enhanced Average Thermoelectric Figure of Merit of the PbTe− SrTe−MnTe Alloy Jun Luo,*,†,‡ Li You,† Jiye Zhang,† Kai Guo,† Hangtian Zhu,§ Lin Gu,§,⊥,# Zhenzhong Yang,§ Xin Li,‡ Jiong Yang,*,‡ and Wenqing Zhang†,‡ †

School of Materials Science and Engineering, Shanghai University, 99 Shangda Road, Shanghai 200444, China Materials Genome Institute, Shanghai University, 99 Shangda Road, Shanghai 200444, China § Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, Chinese Academy of Sciences, Beijing 100190, China ⊥ Collaborative Innovation Center of Quantum Matter, Beijing 100190, China # School of Physical Sciences, University of Chinese Academy of Sciences, Beijing 100190, China ‡

S Supporting Information *

ABSTRACT: Thermoelectric properties of Na-doped PbTe−SrTe system have been improved by the addition of Mn. The substitution of Mn for Pb modifies the band structure of the PbTe−SrTe alloy, which enlarges the band gap and increases the valence band degeneracy. This leads to increased thermopowers and power factors near room temperature, and the electronic contribution to the total thermal conductivity is also substantially reduced due to increased resistivity. Moreover, alloying of MnTe within the PbTe matrix introduces low angle grain boundaries, and significantly reduces the lattice thermal conductivity due to the dislocation scattering. A thermoelectric figure of merit as high as 1.98 and an enhancement of the average thermoelectric figure of merit by 18% are achieved for the sample with 4 at% Mn with respect to the Mn-free sample, which can be mainly attributed to the synergistic effects of the band structure modification and the dislocation scattering on phonon transport, both induced by alloying with MnTe. Our experimental results demonstrate the promising potential of PbTe−SrTe−MnTe system for the application of waste heat recovery. KEYWORDS: thermoelectric materials, PbTe, band gap, lattice thermal conductivity, low angle grain boundary, edge dislocation

■

routine way to improve electrical transport properties,2,3 which can be easily realized by elemental doping (for example, Nadoping for p-type PbTe and I-doping for n-type PbTe).4 Then, σ can be further increased by increasing the mobility. Comparing with σ, the increase of S is more desirable because of its square relation with ZT. The enhancement of S can be achieved through resonant doping, band engineering, and energy filtering.5−7 All these strategies focusing on the improvement of the electrical properties have been well demonstrated in PbTe-based thermoelectric materials. For example, increased carrier mobility has been reported in PbTe conanostructured with both Pb and Sb precipitates, which leads to increased ZT up to 1.4 at 673 K.8,9 Seebeck coefficient and ZT value of the Tl-doped PbTe are substantially enhanced due to the introduction of resonant states.10 Similar enhancement of Seebeck coefficient and ZT value is also observed in PbTe− CdTe,11 PbTe−MgTe,12 PbTe−MnTe,13 and PbTe−PbSe alloys,14 which is ascribed to the successful manipulation of the multiple valence bands of PbTe. The increase of Seebeck

INTRODUCTION Nowadays, the energy crisis and environmental pollution urge the search for renewable energy. Thermoelectric materials can convert waste heat directly into useful electricity and vice versa. Especially, thermoelectric materials show great potential in waste heat harvesting.1 The thermoelectric performance of a material is determined by its electrical and thermal transport properties, usually scaled by the dimensionless thermoelectric figure of merit ZT, ZT = S2σT/ktot, where S, σ, ktot, and T are the Seebeck coefficient (or thermopower), electrical conductivity, total thermal conductivity, and absolute temperature, respectively. ktot includes the contributions from both carriers (ke, electronic thermal conductivity) and phonons (kl, lattice thermal conductivity). According to the above formula, ZT can be enhanced by increasing S and σ, as well as by decreasing ke and kl. It is worth noting, for practical applications, the average ZT value in the working temperature range is even more important in comparison with the peak ZT value. PbTe is one of the best thermoelectric materials for power generation in the medium temperature range of 450−800 K. Therefore, much attention has been focused on thermoelectric performance optimization of PbTe because of its potential in waste heat recovery. Carrier concentration optimization is a © XXXX American Chemical Society

Received: December 14, 2016 Accepted: February 23, 2017

A

DOI: 10.1021/acsami.6b16060 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

microscope (TEM) to examine the microstructure of samples. And energy-dispersive X-ray (EDX) spectroscopy mapping was performed to investigate elemental distribution of the sample by employing a Tecnai F20 TEM. Backscattered electron (BSE) images and EDX mapping images were also obtained by a Zeiss G300 scanning electron microscope (SEM) equipped with an Oxford EDX detector. Hall effect was measured by a Nanometrics HL5500 hall system with a constant magnetic field of 0.5 T. Seebeck coefficient and resistivity were measured simultaneously by a LINSEIS Seebeck and Electric Resistivity Unit LSR-3 under a helium atmosphere. Total thermal conductivity was calculated using ktot = λCPd, where λ, CP, and d were the thermal diffusivity, specific heat and density of the sample, respectively. λ was measured by the flash diffusivity method using a Netzsch LFA 427. CP was estimated from CP (kB/atom) = 3.07 + 0.00047(T − 300) according to literatures.13,14 Density of the sample was measured by the Archimedes drainage method. Calculation. The first-principles calculations were performed with the Vienna ab initio simulation package (VASP).33 Generalized gradient approximation functional34 and projected augmented wave (PAW) method35,36 were used. Since the experimental contents of Mn and Sr can reach up to 4%, we thus constructed a 3 × 3 × 3 supercell of the PbTe formula unit for the calculations of pure PbTe, Sr0.04Pb0.96Te4, and Sr0.04Mn0.04Pb0.92Te4. The spin−orbit coupling (SOC) is included throughout the calculations. Similar computational parameters have been successfully adopted into the study of Ybdoped PbTe.37

coefficient is also achieved in bulk nanostructured PbTe-based materials through energy filtering.15 Furthermore, the introduction of point defects, grain boundaries, and nanoscale impurities into the matrix is effective for remarkably enhanced phonon scattering, and therefore dramatically reduced the lattice thermal conductivity.16−18 For example, in the series compounds PbTe−ASbTe2 (A = Ag, Na, and K)19−26 and PbTe−MTe (M = Sr, Ca, Ba),27−29 the presence of embedded nanocrystals in the PbTe matrix results in greatly decreased lattice thermal conductivity and increased ZT value about 1.5−1.8. PbTe−PbS30,31 and Pb1−xSnxTe− PbS32 systems, which show the spinodal decomposition, also have very low lattice thermal conductivity due to their multiscale microstructures. Recently, the concept to reduce the lattice thermal conductivity through all-scale hierarchical architectures has led to significantly increased ZT value to 2.2 at 915 K in the Na-doped PbTe−SrTe system,29 which is a milestone for the bulk thermoelectric materials. In this work, we demonstrate that the average ZT values of Na-doped PbTe−SrTe can be substantially increased through the addition of a small amount of Mn. Specifically, alloying with MnTe greatly alters the band structure of PbTe−SrTe due to the antibonding d−p coupling between the magnetic Mn and the PbTe host, which enlarges the band gap and increases the valence band degeneracy. This leads to the increased thermopowers and power factors and the reduced ke due to increased resistivity. Furthermore, alloying with MnTe introduces low angle grain boundaries, testified by the scanning transmission electron microscopy, and reduce the kl by the dislocation scattering. Combining the two beneficial effects caused by Mn, a thermoelectric figure of merit as high as 1.98 and an enhancement of the average thermoelectric figure of merit by 18% are achieved for the sample with 4 atom % Mn with respect to the Mn-free sample.

■

RESULTS AND DISCUSSION Our Pb0.94−xMnxNa0.02Sr0.04Te samples have been designed on the basis of Na-doped PbTe-SrTe29 and Na-doped PbTeMnTe13 systems. Both Na-doped Pb1−xSrxTe and Na-doped Pb1−xMnxTe exhibit promising thermoelectric performance, but the ideal and concept to improve the thermoelectric properties are much different. Na-doped Pb1−xSrxTe has been reported to achieve low thermal lattice conductivity mainly through the formation of endotaxially arranged nanoscale precipitates,27,29 while the Na-doped Pb1−xMnxTe has been designed to obtain enhanced thermopower through the manipulation of the valence bands of PbTe.13 The composition of PMNST is therefore supposed to have the combined advantages of Pb1−xSrxTe and Pb1−xMnxTe, that is, low lattice thermal conductivity due to the nanoprecipitates caused by Sr, and enhanced thermopower due to the band manipulation caused by Mn. In this study, Sr content in the PMNST sample is set to be the same (0.04), and it is thus very convenient to explore the influence from MnTe. Figure 1 shows the XRD patterns of the PMNST samples, which can be exclusively indexed to PbTe phase with NaCltype structure. As displayed in the inset of Figure 1, the diffraction peak shifts to higher 2θ angle with increasing x, that is, the lattice parameter decreases with the Mn content. This indicates that Mn2+ enters the PbTe lattice and replaces Pb2+ because the radius of Mn2+ (0.8 Å) is smaller than that of Pb2+ (1.2 Å). Elemental distribution of the sample is investigated by the EDX mapping. As shown in Figure 2, the elements Pb, Te, Sr, and Mn distribute uniformly in the sample with x = 0.03, which agrees well with the structure analysis by XRD, i.e. the sample is a solid solution without any other impurities (see also Figures S1 and S2 for the BSE and EDX mapping images by a SEM). Our experimental result is different from the report by

■

MATERIALS AND METHODS Sample Preparation. In our PbTe−SrTe−MnTe system, Pb0.94−xMnxNa0.02Sr0.04Te (denoted as PMNST) samples with x = 0, 0.01, 0.02, 0.03, and 0.04 were prepared by a combined process of melt-annealing and spark plasma sintering (SPS). The starting materials were weighted and sealed in a glovebox under an argon atmosphere. Pb (99.99%), Mn (99.3%), Na (99.5%), Sr (99%), and Te (99.99%) were weighted according to the nominal composition of the sample, which were mixed in a graphite crucible and then sealed into an evacuated quartz tube (∼10−2 Pa). The vertically placed quartz tube was heated to 1323 K in the well type muffle furnace and held at this temperature for 24 h. Then the sample was quickly cooled to 1073 K by stopping heating (turn off the power supply and cool to 1073 K at the furnace) and annealed at 1073 K for 120 h. After that, the sample was naturally cooled to room temperature in the furnace. The as-melted ingots were ground in a mortar by hand for 30 min in order to obtain fine powders for SPS. Finally, the sample was sintered by SPS at 773 K for 5 min under an axial pressure of 50 MPa. Relative densities of all SPS-treated samples were above 95%. Characterization. All the sample characterizations were performed on SPS-treated samples. Phase identification and structure analysis were carried out by X-ray powder diffraction (XRD) using a PANalytical X’Pert Pro diffractometer with Cu Kα1 (λ = 1.54056 Å) radiation. Spherical aberration-corrected scanning transmission electron microscopy (STEM) was performed using a JEM-ARM 200F transmission electron B

DOI: 10.1021/acsami.6b16060 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 3a and b show the temperature dependent Seebeck coefficient and resistivity of the PMNST sample, respectively. The literature reported Pb0.94Na0.02Sr0.04Te sample,29 corresponding to the Mn content x = 0, is also included for comparison. The Seebeck coefficient increases monotonously with the temperature, and it agrees with the Hall measurement for p-type transport behavior. At low temperature, the sample with higher Mn content has also a larger Seebeck coefficient. However, the difference is decreased with the increase of temperature. For instance, the Seebeck coefficient of the sample with x = 0.04 is ∼40% larger than that of the sample with x = 0.01 at 340 K (Figure 3c), but this difference reduced to only ∼2% at 784 K. The trend of the ρ versus T curve is basically in agreement with that of the S versus T curve. And it is expectable that the sample with higher Seebeck coefficient has also higher resistivity because of the lower mobility shown in Table 1. Comparing with the literature reported Pb0.94Na0.02Sr0.04Te sample,29 the Seebeck coefficient of the PMNST sample is substantially enhanced, from 100 to 160 μV K−1, as shown in Figure 3c. The greatly increased Seebeck coefficient can result in enhanced thermoelectric power factors, from 13.6 to 16.4 μW cm−1 K−2 near room temperature (Figure 3c), as well as the decreased ke due to the increased resistivity. It is to note that the alloyed SrTe also contributes to the enhanced Seebeck coefficient and power factor of our sample because its effect on modifying the band structure. 38 Interestingly, the Seebeck coefficient and power factor also increase with the Sr contration in the PbSe−SrSe alloy because of the similar band structure modification effect.40 First-principles band structure calculations have been carried out to rationalize the variations on the electrical transport properties. Figure 4a−c shows the calculated energy band structures of pure PbTe and its alloys with SrTe and MnTe. For pure PbTe, there are two energy extrema at the valence band, i.e. the principal light hole band (L band) with the band degeneracy Nv = 4 and a secondary heavy hole band (Σ band) with Nv = 12, as shown in the inset of Figure 4a. The Σ band has lower energy than the L band, which is 0.16 eV in our calculation (Figure 4a). The energy difference between Σ band and L band evolves with the temperature: as the temperature increases, the L band reduces its energy gradually while the Σ band has a roughly constant energy.37 This leads to the convergence of the L and Σ bands at about 700 K.41,42 The convergence of these two bands results in a larger band valley number Nv of 16. Since the density of states effective mass m*d is related to Nv by md* = Nv2/3mb*, where mb* is the band mass of a single valley,6 md* is thus increased due to the band convergence. Therefore, the Seebeck coefficient is enhanced accordingly owing to the increase of the band degeneracy. The addition of either Sr or Mn in the PbTe matrix can modify the electronic structure of PbTe, though in a different manner, as shown in Figure 4b and c. For the PbTe-SrTe alloy, the replacement of Sr for Pb results in not only the enlarged band gap at the L point, from 0.15 eV in pristine PbTe (Figure 4a) to 0.30 eV in Sr0.04Pb0.96Te (Figure 4b), but also decreased energy difference between L and Σ bands down to 0.09 eV in

Figure 1. XRD patterns of Pb0.94−xMnxNa0.02Sr0.04Te (PMNST) samples.

Figure 2. STEM image of the Pb0.94−xMnxNa0.02Sr0.04Te sample with x = 0.03 and EDX mapping images corresponding to the rectangular area.

Biswas et al.,29 in which SrTe appears as the second phase in the PbTe matrix because of its extremely small solubility. But our result agrees well with the very recent work by Tan et al.,38 which confirms that at least 5 atom % Sr can substitute for Pb in the PbTe-SrTe alloy by quenching the sample at high temperature. Moreover, we have proved that PbSe-SrSe can form continuous solid solution in the entire composition range.39 These recent findings combined with the above XRD and EDX analysis demonstrate that our PbTe−SrTe−MnTe samples with the Mn content from x = 0 to 0.04 are all singlephase solid solutions. Hall effect measurements indicate p-type transport behavior of the PMNST sample. As shown in Table 1, the room temperature carrier concentration (nH) is in the range of 6.4− 6.9 × 1019 cm−3, indicating that 1 mol % Na doping is effective in obtaining heavily doped p-type PbTe based materials.2 The carrier concentration is almost identical for our samples, and the mobility (μH) decreases nearly linearly with the Mn content. The carrier concentration of our samples is lower than that of literature reported Pb0.94Na0.02Sr0.04Te sample (9.2 × 1019 cm−3).29 The lower carrier concentration of our samples as well as the decrease of Hall mobility with the Mn content can only be ascribed to the substitution of Mn for Pb, implying that the addition of Mn might modify the electronic structure of PbTe.

Table 1. Room-Temperature Carrier Concentration (nH) and Mobility (μH) of the Pb0.94−xMnxNa0.02Sr0.04Te Sample nH (cm−3) μH (cm2 V−1 s−1)

x = 0.01

x = 0.02

x = 0.03

x = 0.04

6.93 × 1019 101.0

6.43 × 1019 86.8

6.89 × 1019 77.9

6.75 × 1019 61.6

C

DOI: 10.1021/acsami.6b16060 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 3. (a) Temperature-dependent Seebeck coefficient and (b) resistivity of the PMNST sample. (c) Mn-content-dependent Seebeck coefficient and power factor at 340 K. The data with Mn content x = 0 is taken from the literature.29 The solid line is a guide to the eye.

enhanced. Our calculated result is consisitent with the report that the band gap is enlarged by alloying of PbTe with MnTe.13 Figure 4d shows the density of states (DOS) around the VBM (set as the zero energy point) for the aforementioned samples. It can be clearly seen that the DOS continuously increase from pure PbTe to Sr0.04Pb0.96Te and Sr0.04Mn0.04Pb0.92Te. This is the origin for the nearly 65% enhancement of the Seebeck coefficient and 20% enhancement of the power factor for the samples alloyed with MnTe at similar carrier concentrations (Figure 3c). Interestingly, the contribution from MnTe in Figure 4d is relatively small, indicating that the d states from Mn interact with the host, which is consistent with k-point dependent band projection shown in Figure 4c, rather than forming the noninteracting resonant levels. According to the literatures,13,38 alloying PbTe with either MnTe13 or SrTe38 results in not only the enlarged band gap but also the convergence of the L and Σ bands at lower temperature, leading to the enhanced Seebeck coefficient at low and medium temperature. However, as the temperature increases, the energy of L band decreases continuously and becomes lower than that of Σ band at higher temperature. Therefore, this multiband conduction effect shall be reduced with increasing temperature. For our sample with fixed Sr concentration but increased Mn content, the increase trend of the Seebeck coefficient with the Mn content is identical to the PbTe−MnTe alloy. Thus, the temperature dependent Seebeck coefficient of our sample (Figure 3a) can be well explained by this two-valence band model. The high-temperature Seebeck coefficient for the sample with different Mn content is almost the same, which can be attributed to the reduced multiband conduction effect. Alloying with MnTe also affects the transport of lattice heat, as shown in Figure 5 for the temperature dependent thermal transport properties of the PMNST samples. Thermal diffusivity (Figure 5a) has been measured directly using the flash diffusivity method, while the total thermal conductivity has been calculated by ktot = λCPd. As shown in Figure 5b, the total thermal conductivity decreases with the Mn content, owing to the both reduced ke and kl. The electronic contribution to the thermal conductivity is decreased due to the increased resistivity with the Mn content. On the other hand, the lattice thermal conductivity is effectively reduced through the alloying effect of MnTe with the PbTe matrix. The electronic thermal conductivity ke has been estimated using the Wiedemann− Franz Law by ke = LT/ρ, where L is the Lorentz number.

Figure 4. Calculated energy band structures of (a) pristine PbTe (inset shows the calculated Fermi surface containing both L and Σ bands), (b) Sr0.04Pb0.96Te alloy, and (c) Sr0.04Mn0.04Pb0.92Te alloy. The band projections on Mn d electrons have been provided in (c), with the content of d electrons on bands denoted by the error bars. The density of states around the valence band maximum for these samples are shown in panel d.

the latter. These influences by SrTe are consistent with other ionic divalent dopants in PbTe, and lower the temperature when the two energy pockets converge. For the PbTe−SrTe solid solution, the increase of the band gap with the Sr concentration has been confirmed by measuring the band gap with a optical diffuse reflectance method,38 which agrees well with our calculation. Mn atoms in PbTe, on the other hand, change the valence band significantly. In our calculation, the magnetic moment of Sr0.04Mn0.04Pb0.92Te is 5 μB per Mn atom, showing that all the spin up d states of the Mn atom are fully occupied. As shown in Figure 4c, the contributions from the d states of Mn are k-point dependent, indicating a strong interaction with the host at the energy range around valence band maximum (VBM), that is, the d-p antibonding according to the literature.43 The interaction between Mn and the host introduces complexity on the electronic structures. For example, the band gap is further enlarged w.r.t. Sr0.04Pb0.96Te, and the density of states at VBM are also significantly D

DOI: 10.1021/acsami.6b16060 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

mechanism of the low angle grain boundary is still not very clear, but it should be related to the lattice strain originated from the point defect in the solid solution. In our sample, the lattice strain increases with the Mn concentration because of the substitution of Mn for Pb. In this case, more edge dislocations could be formed to release the lattice strain. Consequently, in our sample with higher Mn content, these edge dislocations aggregate to form low angle grain boundaries. This kind of crystal defects are desirable for themoelectric materials because these defects can effectively scatter phonons but do not prohibit the transport of carriers.45 Comparing with the literature reported Na-doped PbTe-SrTe system,29 a reduction of 40% in the total thermal conductivity is finally achieved for the PMNST sample with x = 0.04. To further understand the reduced lattice thermal conductivity of the PMNST sample, the Callway’s model is used to calculate the lattice thermal conductivity46 Figure 5. (a) Thermal diffusivity, (b) total thermal conductivity, (c) electronic thermal conductivity, and (d) lattice thermal conductivity of the PMNST sample. The corresponding data of the Pb0.94Na0.02Sr0.04Te sample is reproduced from the literature.29

κl =

(1)

where kB is the Boltzmann constant, e the electron charge, η the reduced chemical potential, and Fn(η) the Fermi integral, respectively. The reduced chemical potential η has been obtained by fitting the experimental Seebeck coefficient using the equation29,31 S=

⎤ kB ⎡ 2F1(η) − η⎥ ⎢ e ⎣ F0(η) ⎦

∫0

θD/ T

τCx 4e x (e x − 1)2

dx

(3)

where v is the averaged sound velocity, T the absolute temperature, ℏ the Planck constant, θD the Debye temperature, τC the relaxation time, respectively, and x = ℏω/kBT, where ω is the phonon frequency. The overall relaxation time τC is a combined term determined by various phonon scattering processes. We first consider our sample as a solid solution and the dominant scattering mechanisms are Umklapp phonon− phonon scattering, point-defect scattering and boundary scattering (see Supporting Information for the detailed calculation procedure). As shown in Figure 7, the lattice

Lorenz number as a function of temperature shown in Figure S3 has been calculated by29,31 ⎛ k ⎞2 3F (η)F2(η) − 4F1(η)2 L = ⎜ B⎟ 0 ⎝e ⎠ F0(η)2

3 kB ⎛ kBT ⎞ ⎜ ⎟ 2π 2v ⎝ ℏ ⎠

(2)

As shown in Figure 5c, the electronic contribution to the total lattice thermal conductivity is less than 30%, which can be ascribed to the high resistivity due to the addition of Mn. The alloying of MnTe with the PbTe matrix leads to an effectively reduced lattice thermal conductivity (Figure 5d). It is to note that low angle grain boundaries are commonly observed in our sample with high Mn contents (Figure 6). Very recently, similar low angle grain boundaries have also been observed in the Pb1−xMgxTe0.2Se0.2 solid solution.44 The formation

Figure 7. Experimental and calculated lattice thermal conductivities. The inset shows that the lattice conductivity of the sample with x = 0.02, 0.03, and 0.04 can not be reproduced by the simple alloy model. The scatters are the experimental data, and the solid line is the calculated result.

thermal conductivity of the sample with x = 0 and 0.01 can be well reproduced by this “alloy model”. We also find that the contribution from the boundary scattering can be neglected since our samples are solid solutions with good crystallinity. Thus, the dominant scattering mechanisms for the sample with x = 0 and 0.01 are only Umklapp scattering and point-defect scattering, which can be described as an alloy model. However, with the increase of the Mn content, the calculated lattice thermal conductivity predicted by the simple alloy model is obviously much higher than the experimental data (inset of Figure 7). It is reasonable to deduce that we should take into account the dislocation scattering for the sample with high Mn

Figure 6. STEM image of the PMNST sample with x = 0.03. E

DOI: 10.1021/acsami.6b16060 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

increased valence band DOS, leading to the increased thermopowers and power factors of the Pb0.94‑xMnxNa0.02Sr0.04Te sample at low and medium temperature. Meanwhile, the total thermal conductivity is effectively reduced by the alloying of MnTe with the PbTe matrix. The electronic contribution to the thermal conductivity is reduced owing to the increased resistivity. The lattice thermal conductivity is decreased because of enhanced phonon scattering by mass and strain field fluctuations as well as the presence of low angle grain boundaries with edge dislocation arrays. The combined effect of band structure modification and alloy/dislocation scattering leads to not only a ZT value as high as 1.98 but also an enhancement of the average ZT value around 18% for the sample with x = 0.04, which is promising for the commercial application in waste heat harvesting.

content since the low angle grain boundaries with edge dislocation arrays are observed in the sample (Figure 6). The edge dislocation arrays at the grain boundaries can be simplified as many single-dislocations inside a grain. Thus, we can calculate the contribution of the edge dislocation arrays using the corresponding relaxation time of the normal dislocation scattering (see Supporting Information for the details), which has been successfully adopted into the study of Bi0.5Sb1.5Te3.45 As shown in Figure 7, the calculated lattice thermal conductivities for the sample with x = 0.02, 0.03, and 0.04 agree well with the experimental ones if the dislocation scattering is taken into account. The combination of the theoretical analysis and experimental data reveals clearly that the addition of Mn leads to not only the point defect in the PMNST sample but also low angle grain boundaries with edge dislocation arrays at higher Mn content. Especially, these edge dislocation arrays play an important role in further reducing the lattice thermal conductivity. Figure 8 shows the temperature dependent ZT value of the PMNST sample. The highest ZT value is 1.98 at 783 K for the

■

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.6b16060. BSE and elemental mapping images of the Pb0.91Mn0.03Na0.02Sr0.04Te sample, temperature-dependent Lorentz number, and detailed calculation of lattice thermal conductivity (PDF)

■

AUTHOR INFORMATION

Corresponding Authors

*Tel.: +86 21 66138036. Fax.: +86 21 66138036. E-mail: [email protected]. *E-mail: [email protected]. ORCID

Jun Luo: 0000-0002-8235-2338 Lin Gu: 0000-0002-7504-031X

Figure 8. Temperature dependent ZT values of PMNST samples. The inset shows the corresponding average ZT values in our measure temperature range. The ZT value of the Pb0.94Na0.02Sr0.04Te sample is reproduced from the literature.29

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS J.Y. acknowledges the fruitful discussion with Prof. Guoqiang Liu from Ningbo Institute of Materials Technology and Engineering, Chinese Academy of Sciences, on the band structure calculations. The work was financially supported by the National Natural Science Foundation of China (Grant No. 51371194, 51632005, 11674211, 21501118, 51522212, 51421002, and 51672307), National Program on Key Basic Research Project (Grant No. 2013CB632500 and 2014CB921002), the Program for Professor of Special Appointment (Eastern Scholar) at Shanghai Institutions of Higher Learning (Grant No. TP2015041), Young Eastern Scholar Project of Shanghai Municipal Education Commission (Grant No. QD2015031), and the Strategic Priority Research Program of Chinese Academy of Sciences (Grant No. XDB07030200).

sample with x = 0.04, which is slightly higher than that of the literature reported Pb0.94Na0.02Sr0.04Te sample at the same temperature.29 As shown in Figure 8, comparing with the Pb0.94Na0.02Sr0.04Te sample,29 the PMNST sample has comparable or slightly higher ZT value at high temperature, but a substantial enhancement of ZT is attained for the PMNST sample in the temperature range from 340 to 700 K. This results in the average ZT value of 1.14, 1.24, 1.25, and 1.31 in our measured temperature range for the PMNST sample with x = 0.01, 0.02, 0.03, and 0.04, respectively. However, the average ZT value in the same temperature range is only 1.11 for the Pb0.94Na0.02Sr0.04Te sample. Thus, an increase about 18% in the average ZT value is achieved for the PMNST sample with x = 0.04, which can be ascribed to the modified band structure as well as the reduced thermal conductivity originated from the dislocation scattering.

■

■

CONCLUSIONS In summary, Mn has been added to the Na-doped PbTe-SrTe system to improve the thermoelectric properties through the modification of the band structures and introduction of low angle grain boundaries. Structure analysis indicates that Mn enters into the crystal lattice and occupies the Pb site. The substitution of Mn for Pb results in enlarged band gap and

REFERENCES

(1) Bell, L. E. Cooling, Heating, Generating Power, and Recovering Waste Heat with Thermoelectric Systems. Science 2008, 321, 1457− 1461. (2) Pei, Y. Z.; LaLonde, A.; Iwanaga, S.; Snyder, G. J. High Thermoelectric Figure of Merit in Heavy Hole Dominated PbTe. Energy Environ. Sci. 2011, 4, 2085−2089. (3) Pei, Y. Z.; Lensch-Falk, J.; Toberer, E. S.; Medlin, D. L.; Snyder, G. J. High Thermoelectric Performance in PbTe Due to Large F

DOI: 10.1021/acsami.6b16060 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces Nanoscale Ag2Te Precipitates and La Doping. Adv. Funct. Mater. 2011, 21, 241−249. (4) Dughaish, Z. H. Lead Telluride as a Thermoelectric Material for Thermoelectric Power Generation. Phys. B 2002, 322, 205−223. (5) Minnich, A. J.; Dresselhaus, M. S.; Ren, Z. F.; Chen, G. Bulk Nanostructured Thermoelectric Materials: Current Research and Future Prospects. Energy Environ. Sci. 2009, 2, 466−479. (6) Pei, Y. Z.; Wang, H.; Snyder, G. J. Band Engineering of Thermoelectric Materials. Adv. Mater. 2012, 24, 6125−6135. (7) Yang, J.; Xi, L. L.; Qiu, W. J.; Wu, L. H.; Shi, X.; Chen, L. D.; Yang, J. H.; Zhang, W. Q.; Uher, C.; Singh, D. J. On the Tuning of Electrical and Thermal Transport in Thermoelectrics: an Integrated Theory-Experiment Perspective. NPJ. Comput. Mater. 2016, 2, 15015. (8) Sootsman, J. R.; Kong, H.; Uher, C.; D’Angelo, J. J.; Wu, C. I.; Hogan, T. P.; Caillat, T.; Kanatzidis, M. G. Large Enhancements in the Thermoelectric Power Factor of Bulk PbTe at High Temperature by Synergistic Nanostructuring. Angew. Chem., Int. Ed. 2008, 47, 8618− 8622. (9) He, J. Q.; Sootsman, J. R.; Xu, L. Q.; Girard, S. N.; Zheng, J. C.; Kanatzidis, M. G.; Dravid, V. P. Anomalous Electronic Transport in Dual-Nanostructured Lead Telluride. J. Am. Chem. Soc. 2011, 133, 8786−8789. (10) Heremans, J. P.; Jovovic, V.; Toberer, E. S.; Saramat, A.; Kurosaki, K.; Charoenphakdee, A.; Yamanaka, S.; Snyder, G. J. Enhancement of Thermoelectric Efficiency in PbTe by Distortion of the Electronic Density of States. Science 2008, 321, 554−557. (11) Pei, Y. Z.; LaLonde, A. D.; Heinz, N. A.; Snyder, G. J. High Thermoelectric Figure of Merit in PbTe Alloys Demonstrated in PbTe-CdTe. Adv. Energy Mater. 2012, 2, 670−675. (12) Pei, Y. Z.; LaLonde, A. D.; Heinz, N. A.; Shi, X. Y.; Iwanaga, S.; Wang, H.; Chen, L. D.; Snyder, G. J. Stabilizing the Optimal Carrier Concentration for High Thermoelectric Efficiency. Adv. Mater. 2011, 23, 5674−5678. (13) Pei, Y. Z.; Wang, H.; Gibbs, Z. M.; LaLonde, A. D.; Snyder, G. J. Thermopower Enhancement in Pb1‑xMnxTe Alloys and its Effect on Thermoelectric Efficiency. NPG Asia Mater. 2012, 4, e28. (14) Pei, Y. Z.; Shi, X. Y.; LaLonde, A.; Wang, H.; Chen, L. D.; Snyder, G. J. Convergence of Electronic Bands for High Performance Bulk Thermoelectrics. Nature 2011, 473, 66−69. (15) Heremans, J. P.; Thrush, C. M.; Morelli, D. T. Thermopower Enhancement in Lead Telluride Nanostructures. Phys. Rev. B: Condens. Matter Mater. Phys. 2004, 70, 115334. (16) Poudel, B.; Hao, Q.; Ma, Y.; Lan, Y. C.; Minnich, A.; Yu, B.; Yan, X. A.; Wang, D. Z.; Muto, A.; Vashaee, D.; Chen, X. Y.; Liu, J. M.; Dresselhaus, M. S.; Chen, G.; Ren, Z. F. High-Thermoelectric Performance of Nanostructured Bismuth Antimony Telluride Bulk Alloys. Science 2008, 320, 634−638. (17) Lan, Y. C.; Minnich, A. J.; Chen, G.; Ren, Z. F. Enhancement of Thermoelectric Figure-of-Merit by a Bulk Nanostructuring Approach. Adv. Funct. Mater. 2010, 20, 357−376. (18) Pichanusakorn, P.; Bandaru, P. Nanostructured Thermoelectrics. Mater. Sci. Eng., R 2010, 67, 19−63. (19) Hsu, K. F.; Loo, S.; Guo, F.; Chen, W.; Dyck, J. S.; Uher, C.; Hogan, T.; Polychroniadis, E. K.; Kanatzidis, M. G. Cubic AgPbmSbTe2+m: Bulk Thermoelectric Materials with High Figure of Merit. Science 2004, 303, 818−821. (20) Poudeu, P. F. R.; D’Angelo, J.; Downey, A. D.; Short, J. L.; Hogan, T. P.; Kanatzidis, M. G. High Thermoelectric Figure of Merit and Nanostructuring in Bulk p-Type Na1‑xPbmSbyTem+2. Angew. Chem., Int. Ed. 2006, 45, 3835−3839. (21) Androulakis, J.; Hsu, K. F.; Pcionek, R.; Kong, H.; Uher, C.; D’Angelo, J. J.; Downey, A.; Hogan, T.; Kanatzidis, M. G. Nanostructuring and High Thermoelectric Efficiency in p-Type Ag(Pb1−‑ySny)mSbTe2+m. Adv. Mater. 2006, 18, 1170−1173. (22) Zhou, M.; Li, J. F.; Kita, T. Nanostructured AgPbmSbTem+2 System Bulk Materials with Enhanced Thermoelectric Performance. J. Am. Chem. Soc. 2008, 130, 4527−4532. (23) Ke, X. Z.; Chen, C. F.; Yang, J. H.; Wu, L. J.; Zhou, J.; Li, Q.; Zhu, Y. M.; Kent, P. R. C. Microstructure and a Nucleation

Mechanism for Nanoprecipitates in PbTe-AgSbTe2. Phys. Rev. Lett. 2009, 103, 145502. (24) Cook, B. A.; Kramer, M. J.; Harringa, J. L.; Han, M. K.; Chung, D. Y.; Kanatzidis, M. G. Analysis of Nanostructuring in High Figureof-Merit Ag1‑xPbmSbTe2+m Thermoelectric Materials. Adv. Funct. Mater. 2009, 19, 1254−1259. (25) Guéguen, A.; Poudeu, P. F. P.; Li, C. P.; Moses, S.; Uher, C.; He, J. Q.; Dravid, V.; Paraskevopoulos, K. A.; Kanatzidis, M. G. Thermoelectric Properties and Nanostructuring in the p-Type Materials NaPb18‑xSnxMTe20 (M = Sb, Bi). Chem. Mater. 2009, 21, 1683−1694. (26) Kanatzidis, M. G. Nanostructured Thermoelectrics: The New Paradigm? Chem. Mater. 2010, 22, 648−659. (27) Biswas, K.; He, J. Q.; Zhang, Q. C.; Wang, G. Y.; Uher, C.; Dravid, V. P.; Kanatzidis, M. G. Strained Endotaxial Nanostructures with High Thermoelectric Figure of Merit. Nat. Chem. 2011, 3, 160− 166. (28) Biswas, K.; He, J. Q.; Wang, G. Y.; Lo, S. H.; Uher, C.; Dravid, V. P.; Kanatzidis, M. G. High Thermoelectric Figure of Merit in Nanostructured p-Type PbTe−MTe (M = Ca, Ba). Energy Environ. Sci. 2011, 4, 4675−4684. (29) Biswas, K.; He, J. Q.; Blum, I. D.; Wu, C. I.; Hogan, T. P.; Seidman, D. N.; Dravid, V. P.; Kanatzidis, M. G. High-Performance Bulk Thermoelectrics with All-Scale Hierarchical Architectures. Nature 2012, 489, 414−418. (30) He, J. Q.; Girard, S. N.; Kanatzidis, M. G.; Dravid, V. P. Microstructure-Lattice Thermal Conductivity Correlation in Nanostructured PbTe0.7S0.3 Thermoelectric Materials. Adv. Funct. Mater. 2010, 20, 764−772. (31) Johnsen, S.; He, J. Q.; Androulakis, J.; Dravid, V. P.; Todorov, I.; Chung, D. Y.; Kanatzidis, M. G. Nanostructures Boost the Thermoelectric Performance of PbS. J. Am. Chem. Soc. 2011, 133, 3460−3470. (32) Androulakis, J.; Lin, C. H.; Kong, H. J.; Uher, C.; Wu, C. I.; Hogan, T.; Cook, B. A.; Caillat, T.; Paraskevopoulos, K. M.; Kanatzidis, M. G. Spinodal Decomposition and Nucleation and Growth as a Means to Bulk Nanostructured Thermoelectrics: Enhanced Performance in Pb1‑xSnxTe−PbS. J. Am. Chem. Soc. 2007, 129, 9780−9788. (33) Kresse, G.; Furthmüller, J. Efficient Iterative Schemes for Ab Initio Total-Energy Calculations Using a Plane-Wave Basis Set. Phys. Rev. B: Condens. Matter Mater. Phys. 1996, 54, 11169−11186. (34) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, 3865−3868. (35) Kresse, G.; Joubert, D. From Ultrasoft Pseudopotentials to the Projector Augmented-Wave Method. Phys. Rev. B: Condens. Matter Mater. Phys. 1999, 59, 1758−1775. (36) Blöchl, P. E. Projector Augmented-Wave Method. Phys. Rev. B: Condens. Matter Mater. Phys. 1994, 50, 17953−17979. (37) Jian, Z. Z.; Chen, Z. W.; Li, W.; Yang, J.; Zhang, W. Q.; Pei, Y. Z. Significant Band Engineering Effect of YbTe for High Performance Thermoelectric PbTe. J. Mater. Chem. C 2015, 3, 12410−12417. (38) Tan, G. J.; Shi, F. Y.; Hao, S. Q.; Zhao, L. D.; Chi, H.; Zhang, X. M.; Uher, C.; Wolverton, C.; Dravid, V. P.; Kanatzidis, M. G. NonEquilibrium Processing Leads to Record High Thermoelectric Figure of Merit in PbTe-SrTe. Nat. Commun. 2016, 7, 12167. (39) Zeng, L. J. K.; Zhang, J. Y.; You, L.; Zheng, H. Y.; Liu, Y. F.; Ouyang, L.; Huang, P.; Xing, J. J.; Luo, J. Enhanced Thermoelectric Performance in PbSe-SrSe Solid Solution by Mn Substitution. J. Alloys Compd. 2016, 687, 765−772. (40) Wang, H.; Gibbs, Z. M.; Takagiwa, Y.; Snyder, G. J. Tuning Bands of PbSe for Better Thermoelectric Efficiency. Energy Environ. Sci. 2014, 7, 804−811. (41) Gibbs, Z. M.; Kim, H.; Wang, H.; White, R. L.; Drymiotis, F.; Kaviany, M.; Snyder, G. J. Temperature Dependent Band Gap in PbX (X = S, Se, Te). Appl. Phys. Lett. 2013, 103, 262109. (42) Jaworski, C. M.; Nielsen, M. D.; Wang, H.; Girard, S. N.; Cai, W.; Porter, W. D.; Kanatzidis, M. G.; Heremans, J. P. Valence-Band Structure of Highly Efficient p-Type Thermoelectric PbTe-PbS Alloys. Phys. Rev. B: Condens. Matter Mater. Phys. 2013, 87, 045203. G

DOI: 10.1021/acsami.6b16060 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces (43) Tan, X. J.; Shao, H. Z.; Hu, T. Q.; Liu, G. Q.; Ren, S. F. Theoretical Understanding on Band Engineering of Mn-Doped Lead Chalcogenides PbX (X = Te, Se, S). J. Phys.: Condens. Matter 2015, 27, 095501. (44) Fu, T. Z.; Yue, X. Q.; Wu, H. J.; Fu, C. G.; Zhu, T. J.; Liu, X. H.; Hu, L. P.; Ying, P. J.; He, J. Q.; Zhao, X. B. Thermoelectric Performance of PbTe Bulk Materials with Figure of Meirit zT > 2 by Multi-functional Alloying. J. Materiomics 2016, 2, 141−149. (45) Kim, S. I.; Lee, K. H.; Mun, H. A.; Kim, H. S.; Hwang, S. W.; Roh, J. W.; Yang, D. J.; Shin, W. H.; Li, X. S.; Lee, Y. H.; Snyder, G. J.; Kim, S. W. Dense Dislocation Arrays Embedded in Grain Boundaries for High-Performance Bulk Thermoelectrics. Science 2015, 348, 109− 114. (46) Callaway, J. Model for Lattice Thermal Conductivity at Low Temperatures. Phys. Rev. 1959, 113, 1046−1051.

H

DOI: 10.1021/acsami.6b16060 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX