Letter pubs.acs.org/NanoLett
Formation of Nanoscale Composites of Compound Semiconductors Driven by Charge Transfer Weiwei Gao,† Roberto dos Reis,†,‡ Laura T. Schelhas,§ Vanessa L. Pool,§ Michael F. Toney,§ Kin Man Yu,†,∥ and Wladek Walukiewicz*,† †
Materials Sciences Division and ‡National Center for Electron Microscopy, Molecular Foundry, Lawrence Berkeley National Laboratory, Berkeley, California 94720, United States § Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Menlo Park, California 94025, United States ∥ Department of Physics and Materials Science, City University of Hong Kong, Kowloon, Hong Kong S Supporting Information *
ABSTRACT: Composites are a class of materials that are formed by mixing two or more components. These materials often have new functional properties compared to their constituent materials. Traditionally composites are formed by self-assembly due to structural dissimilarities or by engineering different layers or structures in the material. Here we report the synthesis of a uniform and stoichiometric composite of CdO and SnTe with a novel nanocomposite structure stabilized by the dissimilarity of the electronic band structure of the constituent materials. The composite has interesting electronic properties which range from highly n-type in CdO to semi-insulating in the intermediate composition range to highly p-type in SnTe. This can be explained by the overlap of the conduction and valence band of the constituent compounds. Ultimately, our work identifies a new class of composite semiconductors in which nanoscale self-organization is driven and stabilized by charge transfer between constituent materials. KEYWORDS: Composite, semiconductor, charge transfer, self-assembly, electronic properties
N
Formation of composites during physical mixing relies on structural dissimilarities of the constituent materials and can be driven by strain, lattice mismatch, spinodal decomposition, or entropy.19−22 So far there has been no evidence that dissimilar electronic properties of the constituent materials can play any role in the composite formation process. Here we use CdO and SnTe two semiconductor materials with drastically different electronic properties but the same crystal structure to demonstrate that the electronic dissimilarity can be a dominant driving force in the formation of a nanocomposite material. We propose a novel mechanism for creating semiconductor composites in which the structure and the scale of the
ew materials discovery often relies on the mixing of different materials to create alloys or composites with properties tailored for specific applications. Composites, in particular, are an interesting class of materials because they can exhibit new functionalities and/or improved properties that the constituents do not possess.1−3 Specifically, composites of semiconductors are widely used to create new materials with modified optoelectronic properties, band gaps and band offsets.4−8 Examples are widespread and include GaAs/GaP superlattices, InAs quantum dots, and CdS/TiO2 heterojunctions.9−11 Material composites are typically formed using controlled deposition or physical mixing. Thus, composite structures can be engineered by controlled growth of multilayers or quantum wires and dots embedded in a host matrix.10,12−16 Another approach involves spontaneous selfassembly2 that is widely used in bulk heterojunction organic photovoltaics.17,18 © XXXX American Chemical Society
Received: June 11, 2016 Revised: July 19, 2016
A
DOI: 10.1021/acs.nanolett.6b02395 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 1. Electrical and optical properties of (SnTe)x(CdO)1−x nanocomposites. (a) Electrical conductivity, carrier concentration, and carrier mobility as functions of SnTe content for (SnTe)x(CdO)1−x grown on glass substrates, determined by Hall-effect measurements at room temperature. The electrical properties vary from high n-type conducting in CdO-rich to high p-type conducting in SnTe-rich materials. A semiinsulating behavior is observed for films in the intermediate composition range. (b) The squared optical absorption coefficients, α2, as functions of photon energy. The absorption edge energies were obtained from linear fits to the squared absorption curves. (c) The uncorrected absorption edge energy as a function of SnTe content (blue solid circles). The corrected intrinsic bandgap (red solid circles) was obtained by accounting for the BM shift as well as for electron−electron and electron−ion renormalization effects. The dashed lines are linear interpolations for the Γ and L point direct gaps. The solid lines represent quadratic fits to the intrinsic gap with the bandgap bowing parameters of ∼1.75 eV for the Γ and ∼2.85 eV for the L point transitions.
alignment has been shown to result in charge redistribution at the CdO/SnTe interface31 that, as we will argue below, is a dominant factor in the formation of the CdO/SnTe composite. The material synthesis and characterization are described in Methods. Figure 1a shows the composition dependence of the conductivity, carrier concentration, and mobility of (SnTe)x(CdO)1−x films. Pure CdO (x = 0) shows expected bulk properties for this material with an electron concentration of ∼2.2 × 1020 cm−3 and a relatively high electron mobility close to ∼95 cm2/(V s). The electron concentration rises to about 1.7 × 1021 cm−3 for (SnTe)0.05(CdO)0.95, and then rapidly decreases to the range where the material becomes semi-insulating (x > 0.34). In the same composition range, the electron mobility shows a nonmonotonic behavior with a local mobility maximum of ∼46 cm2/(V s) at x = 0.2. On the other end of the composition range, pure SnTe shows a strong p-type conductivity with a hole concentration of almost 1021 cm−3 and mobility of ∼31 cm2/(V s). Both hole concentration and mobility decrease with increasing CdO content, making the material semi-insulating for x < 0.68. In terms of electrical properties, the whole composition range can be divided into three well-defined regions: n-type for 0 ≤ x < 0.34, semiinsulating for 0.34 < x < 0.68, and p-type for 0.68 < x ≤ 1. Tauc plots (square of absorption coefficient vs photon energy) for selected samples are shown in Figure 1b. A relatively sharp onset of a strong absorption edge is observed in the whole composition range indicating a direct gap optical transition between well-defined conduction and valence band
nanocomposite are electronically controlled by a charge transfer between the two component materials. As grown, undoped CdO is intrinsically n-type with electron concentrations ranging from about 1019 to mid 1020 cm−3.23,24 The high electron concentration can be attributed to a very large electron affinity. The conduction band edge (CBE) of CdO lies 5.9 eV below the vacuum level or about 1 eV below the Fermi level stabilization energy, EFS.24,25 This extremely low CBE reduces the formation energy of the donor-like native defects that populate the conduction band with electrons. In contrast to CdO, as-grown undoped SnTe is typically p-type with hole concentrations in the 1020 to 1021 cm−3 range.26−28 It has been shown that its extreme propensity for p-type conductivity originates from a very high energy of the L−6 valence band edge (VBE).26 The VBE of SnTe is located 4.4 eV below the vacuum level, or 0.5 eV above the EFS, which is higher than the VBE in any known semiconductor.26,29 The high VBE reduces the formation energy of acceptor-like native defects resulting in p-type conductivity of the undoped material. Recently, SnTe has received significant attention as a prototypical topological crystalline insulator representing a new class of materials with gapless linear dispersion relations.28,30 The low CBE of CdO and high VBE of SnTe lead to an extreme type III band offset between those two materials, placing the CBE of CdO about 1.5 eV below the VBE of SnTe.31,32 The band offset schematics for the CdO/SnTe system are shown in the Figure S1. This unusual band B
DOI: 10.1021/acs.nanolett.6b02395 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 2. Structural characterization (TEM) of the (SnTe)x(CdO)1−x nanocomposites. (a−c) Sample with x = 0.10, (d−f) with x = 0.45 and (g−i) with x = 0.73, representing n-type, semi-insulating and p-type materials, respectively. Bright-field (BF) and dark field (DF) images of alloys with x = 0.1 (a), and 0.73 (g) confirm the columnar-like structure with columns size of about 10−15 nm (shown in detail in b and h). (d) High-resolution TEM (HRTEM) of the semi-insulating sample with x = 0.45, reveals a presence of small crystallites with a 2−3 nm size. (e) Close view of one of the crystallites from the highlighted region in panel d. The inset of e shows a fast Fourier transform (FFT) of the image. Crystal plane spacings are indicated on b,e,h. Selected area area diffraction (SAD) patterns for each of the selected samples are shown on c,f,i with the diffracted planes labeled.
edges. The absorption edge energies obtained from linear fits to the squared absorption curves are shown in Figure 1c. The absorption edge energy varies from about 0.8 eV for x = 1 to about 3.3 eV for x = 0.05. It is important to note that because of the large electron concentration in the samples on the CdO-rich side and large hole concentration in the samples on the SnTe-rich side, the measured absorption edge energies do not represent the intrinsic band gaps and have to be corrected for the Burstein− Moss (BM) effect.33,34 In addition, the absorption edge energy is affected by bandgap renormalization effects including electron−electron and electron−ion interactions.35,36 The Burstein−Moss shift was calculated using a standard method described in Supporting Information.24,35,36 The calculated BM shift, as a function of SnTe content, is shown in the Figure S2. The intrinsic bandgap represented by red dots in Figure 1c was obtained by subtracting the BM shift as well as the shifts resulting from electron−electron and electron−ion interactions from the absorption edge energy. The intrinsic band gaps show an unusual composition dependence with the largest band gap of ∼2.7 eV at x ∼ 0.3. This behavior is different from that found for semiconductor alloys, where the band gaps always show negative bowing.
The unusual evolution of the electrical and optical properties of the (SnTe)x(CdO)1−x raises questions about the structural properties of this material. In principle, the four elements in the targets can form a large variety of compounds and alloys including four simple binary semiconductors, CdO, CdTe, SnTe, and SnO. All the potential compounds could also form composite materials. In order to investigate the overall microstructure and the range of crystalline order within the films, we have performed extensive studies of the structural properties of these materials using Rutherford backscattering spectrometry (RBS), X-ray diffraction (XRD), and transmission electron microscopy (TEM). For the TEM study, we have selected samples with x = 0.1, 0.45, and 0.73 representing the ntype, semi-insulating and p-type material, respectively. Brightfield (BF), dark-field (DF), and high-resolution micrographs of these three investigated samples are displayed in Figure 2 panels a and b, d and e, and g and h, respectively. Conductive samples (with x = 0.1 and 0.73) present continuous crystalline regions forming columnar-like ordered films with columns size of about 10−15 nm (Figure 2a,b and g,h). For the semiinsulating sample with 45% of SnTe (Figure 2d,e), small crystallites distributed within the layer are observed. These crystallites (with a 2−3 nm size) appear to be attached to each other forming a nanoscale composite film. C
DOI: 10.1021/acs.nanolett.6b02395 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 3. Crystal structure of (SnTe)x(CdO)1−x nanocomposites. (a) XRD patterns for (SnTe)x(CdO)1−x in the whole composition range. The crystal structure of the composite is dominated by the majority phase and no obvious diffraction peaks corresponding to any minority phase are seen. (b) Calculated lattice constants obtained from Synchrotron XRD as a function of SnTe content x. The orange and blue dots are for films grown on glass and Si, respectively. The red dots represent SAD results. The error bars are estimated by considering different samples and diffraction peaks width. The error bars for the SnTe content are as large as symbol size.
composition deposited on glass are shown in Figure 3a. The data were integrated from 2D images (Figure S8). Patterns can be matched to CdO for x < 0.13 and SnTe for x > 0.73. These results show that the crystal structure of the material is dominated by the majority phase and no obvious diffraction peaks corresponding to any minority phase are seen. Broad diffuse peaks, seen in the intermediate composition range, are indicative of nanoscale crystallites. This is consistent with the nanodiffraction and HRTEM that show very small size crystallites for the sample with x = 0.45. To determine if the diffuse peaks in the intermediate composition region are due to the amorphous substrate or the film itself, a set of samples were grown on single crystal Si substrates to reduce any signal related to the substrate. On the single crystal Si substrates the pattern matched to SnTe for x > 0.55, and the intermediate (0.13 < x < 0.55) region shows broad peaks confirming the existence of nanocrystalline film (Figure S9). To identify the phase progression in these films lattice parameters were calculated from both the electron diffraction and XRD data (Figure 3b). For x < 0.13 the lattice parameter is constant, independent of x, and corresponds to CdO. Conversely, at x > 0.73/0.55 (glass/Si) the lattice parameter is again constant but now matches SnTe (to within experimental error). Both the SAD and the XRD results agree in the composition range where both methods were used. Closer investigation of each majority phase region shows that there is no obvious shift in the lattice parameter in these regions. A lattice parameter contraction on the SnTe-rich side or expansion on the CdOrich side is expected for isolated Cd or Sn substituting into the crystal lattice. Because this change is not seen and there are no obvious diffraction peaks from the minority phase we can conclude that the films can be described as a composite structure, comprised of a crystalline majority phase with small inclusions of a crystalline. Additionally, in the intermediate region the TEM data shown in Figure 2e indicates that small (∼2 nm diameter) sized crystallites of either phase can coexist
Selected area diffraction (SAD) patterns for the same selected samples are presented in Figure 2c,f,i. Cubic lattice parameters of ∼4.97 Å (±0.05 Å), ∼5.265 Å (±0.05 Å), and ∼6.32 Å (±0.05 Å), respectively, for samples with 10%, 45% and 73% of SnTe content were extracted by analyzing the SAD patterns (see details in the Figures S3 and S4 and Table S1) and high-resolution TEM (HRTEM) images. The presence of diffused ring in the background of SAD pattern for the sample with x = 0.45 (Figure 2f) might suggest the presence of an amorphous-like phase reminiscent in the film. This observation is in agreement with the broad peaks in the XRD for the semiinsulating region and will be discussed below. Both, XRD and SAD show the overall crystallographic phase within the film with different spatial resolutions, by collecting an average of the scattering points within the film. In order to demonstrate that the local constituents of the semi-insulating film are indeed fine grains dispersed within a large region, we need a more localized way to probe the local crystallinity. To accomplish that, we performed nanobeam electron diffraction (in STEM mode) by acquiring a series of diffraction patterns at each pixel of the scanning area (see Figure S5). The speckled pattern observed for each of the scanning positions clearly demonstrates that the intermediate composition film consists of small size misoriented crystallites that are also evident in the HRTEM images shown in Figures 2d,e. The diffuse ring, amorphous-like SAD pattern shown in Figure 2f originates from averaging the collected nanodiffraction patterns. The compositional uniformity of the films was studied using energy dispersive X-ray (EDX) mapping. A complete EDX analysis of the elemental distribution is shown in the Figure S6. The maps show a uniform average distribution of all elements over the ∼200 nm range corresponding to the entire sample thickness. This is in a good agreement with the local macroscopic composition obtained by RBS and shown in the Figure S7. Further structural analysis was performed by Synchrotron 2D wide-angle X-ray diffraction. XRD of films with varying D
DOI: 10.1021/acs.nanolett.6b02395 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 4. Band offsets and Fermi level stabilization energy in (SnTe)x(CdO)1−x nanocomposites. (a) CBE and VBE energies at the Γ and L point of the Brillouin zone as functions of SnTe content x. The energies were calculated from the composition dependent energy gaps shown in Figure 1c and from the known positions of the CBE at Γ point in CdO and VBE at L point in SnTe with respect to vacuum level. Direct gap transitions are shown by orange (Γ point) and purple (L point) arrows. (b) Fermi level energy relative to vacuum level as a function of SnTe content x. The red and blue lines represent Γ-CBE and L-VBE energies, respectively. The Fermi level tends to converge on the Fermi level stabilization energy EFS at ∼4.9 eV below vacuum level on both sides and enters the band gap for an intermediate composition range.
the structure of the composite. For the CdO-rich or SnTe-rich compositions, the structure consists of a majority and minority component where the majority phase is crystalline and the minority phase is crystalline with crystallites of a few nm in size. As a result, the structure and the lattice parameter of the composite are determined by the majority component, which is consistent with the XRD results shown in Figure 3b. In the intermediate composition range, the material is a composite of both components with crystallites of a few nanometers in size. As is shown in Figure 2d,e, such small crystallites can be seen in TEM but not in XRD. In order to understand the composition dependence of the optical and electrical properties of the material, we consider the electronic band structure of the end point compounds, CdO and SnTe, placed on the absolute energy scale (shown in the Figure S1). The composition dependence of the band gaps, as well as CBE and VBE energies, can be obtained from an interpolation between those two compounds. The results of the linear interpolations for the band gaps are shown in Figure 1c as dashed lines. The Γ point energy gap of about 2.2 eV in CdO increases with increasing SnTe content reaching the value of ∼4.6 eV in SnTe. For SnTe-rich composites, the small direct gap of ∼0.3 eV at L point in SnTe is increasing with increasing CdO content to ∼5.7 eV in CdO.29,40 Although this linear interpolation of the two direct gaps is consistent with the trends observed in the composition dependence of the intrinsic band gap of (SnTe)x(CdO)1−x (shown in Figure 1c), a better agreement is obtained assuming band gap bowing parameters of ∼1.75 eV for the Γ point, and ∼2.85 eV for the L point band gap. Results in Figure 1c show that the calculated band gap represented by solid lines is in good agreement with the experiment, and that a transition from the direct Γ point band gap to the direct L point band gap occurs at x = 0.45. The material around this composition represents an unusual case of a semiconductor with two similar direct gaps at different points of the Brillouin zone. Such a well-defined scale of the structural features can be attributed to the fact that the formation of the nanocomposite is solely driven by a short distance charge transfer. This is in contrast to previous reports on formation of composites of compound semiconductors where competing driving forces
and indeed the XRD data in this region is consistent with nanocrystalline composite. The results of the structural studies confirm that although the material is segregated on a localized nanometer scale, it still exhibits compositional uniformity on a larger scale. Thus, it can be considered as a composite of two mostly stoichiometric compounds. Such a nanoscale composite is distinctly different from multicomponent random substitutional semiconductor alloys in which atoms occupy lattice sites with a probability determined by the alloy composition. Although composite materials are well-known and studied, to our knowledge this is the first case of a nanocomposite of compound semiconductors that exhibits a gradual and predictable evolution of electronic properties over the entire composition range in a manner similar to what one would expect for an alloyed material. It could be argued that the reason for this exceptionally uniform mixing originates from drastically different electronic properties of the component materials. It is now well established that because of extremely large electron affinity of 5.9 eV of CdO a large density of electrons (∼4 × 1013 cm−2) accumulate on the CdO surface or at interfaces between CdO and other materials.23,31,32,37 An even larger density of holes (∼1 × 1015 cm−2) is formed at SnTe surface/interface.26 As is illustrated in the Figure S8 this leads to a large charge transfer at CdO/SnTe interface. It is important to emphasize that the accumulated charge is confined to a narrow, 2−3 nm surface/interface region.31,32 Consequently, the electron transfer and an associated energy gain stabilizes the CdO/SnTe interfaces and enhances the growth of separate phases of CdO and SnTe. Such a nanoscale composite is distinctly different from multicomponent random substitutional semiconductor alloys in which atoms occupy lattice sites with a probability determined by the alloy composition38 and the content of each element is restricted by stoichiometry requirements. As an example, the properties of the extensively studied Zn 1−y Cd y Se 1−x Te x quaternary alloy system are determined by a composition weighted averages of four possible binary compounds.39 However, because the charge transfer and the energy gain are limited to 2−3 nm region from the interface, it introduces limits on the preferential size of the nanocrystals and thus also E
DOI: 10.1021/acs.nanolett.6b02395 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
∼1.8 eV, and the conductivity changes from highly p-type conducting to semi-insulating. These results appear to indicate that the minority compound is uniformly distributed within the majority phase forming a fine nanocomposite material. The properties of the material are fully determined by the relative content of the two compounds. The proposed structure of the nanocomposite provides a justification for the interpolation scheme used to determine composition dependence of the electronic band structure. Because of the large overlap between the conduction band of CdO and the valence band of SnTe, there are no barriers at the interfaces of the different phases, and electrons and holes are completely delocalized, experiencing an average potential of the nanocomposite. As a result, the electronic band structure is given by the composition-weighted average of the band structures of the component compounds in the way similar to what would be expected for random alloys. In conclusion, we have synthesized (SnTe)x(CdO)1−x, a nanocomposite of two compound semiconductors, from x = 0 to x = 1. The electronic dissimilarity and a large charge transfer between CdO and SnTe is a dominant driving force for the composite formation. The scale of the nanocomposite phases is directly related to the thicknesses of the charge accumulation regions at CdO/SnTe interface. The optical and electrical properties of the composite vary from a highly n-type, through a semi-insulating, to a highly p-type semiconductor. This behavior can be understood through a simple interpolation of the electronic band structure of the end-point compounds. This research points at the existence of a new class of semiconductor materials with a nanocomposite-like structure and alloy-like electronic properties. Methods. Synthesis of (SnTe)x(CdO)1−x Composite. A series of nominally (SnTe)x(CdO)1−x thin films were grown on soda lime microscope glass slides and single crystal Si substrate with (001) orientation by using a dual-gun radio frequency magnetron sputtering system with separated CdO and SnTe targets. The chamber was pumped down to a base pressure 0.34. On the other composition end, as has been shown previously26 the high hole concentration in SnTe originates from the location of the VBE at about 0.5 eV above EFS. The L point VBE shifts downward with increasing CdO content resulting in decreasing hole concentration and mobility. The material becomes semi-insulating for x < 0.68, as is shown in Figure 4b, when the VBE falls below EFS. In the intermediate composition range 0.3 < x < 0.68, the EFS pins the Fermi energy in the indirect band gap between Γ point CBE and L point VBE, making the material semi-insulating. The above analysis shows that the optical and electrical properties of (SnTe)x(CdO)1−x materials can be explained by the electronic band structure obtained from a simple interpolation of the relevant band edges between the end point compounds. The results of structural characterization indicate that for CdO-rich or SnTe-rich material, the lattice parameter is controlled by the majority compound. However, the optical and electrical properties are significantly affected by even a small content of the minority compound. Thus, as seen in Figure 3b, although the lattice parameter remains almost constant for 0.64 < x ≤ 1, the band gap changes from ∼0.3 to F
DOI: 10.1021/acs.nanolett.6b02395 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
(4) Rajeshwar, K.; de Tacconi, N. R.; Chenthamarakshan, C. R. Chem. Mater. 2001, 13, 2765−2782. (5) Esaki, L.; Tsu, R. IBM J. Res. Dev. 1970, 14, 61−65. (6) Jaros, M. Rep. Prog. Phys. 1985, 48, 1091−1154. (7) Bonilla, L. L.; Grahn, H. T. Rep. Prog. Phys. 2005, 68, 577−683. (8) Ahn, C. H.; Senthil, K.; Cho, H. K.; Lee, S. Y. Sci. Rep. 2013, 3, 2737. (9) Gudiksen, M. S.; Lauhon, L. J.; Wang, J.; Smith, D. C.; Lieber, C. M. Nature 2002, 415, 617−620. (10) Wu, J.; Jin, P. Front. of Phys. 2015, 10, 7. (11) Marschall, R. Adv. Funct. Mater. 2014, 24, 2421−2440. (12) Cargnello, M.; Johnston-Peck, A. C.; Diroll, B. T.; Wong, E.; Datta, B.; Damodhar, D.; Doan-Nguyen, V. V.; Herzing, A. A.; Kagan, C. R.; Murray, C. B. Nature 2015, 524, 450−453. (13) Abeles, B.; Tiedje, T. Phys. Rev. Lett. 1983, 51, 2003−2006. (14) Vouilloz, F.; Oberli, D. Y.; Dupertuis, M. A.; Gustafsson, A.; Reinhardt, F.; Kapon, E. Phys. Rev. Lett. 1997, 78, 1580−1583. (15) Claudon, J.; Bleuse, J.; Malik, N. S.; Bazin, M.; Jaffrennou, P.; Gregersen, N.; Sauvan, C.; Lalanne, P.; Gerard, J. M. Nat. Photonics 2010, 4, 174−177. (16) Pernot, G.; Stoffel, M.; Savic, I.; Pezzoli, F.; Chen, P.; Savelli, G.; Jacquot, A.; Schumann, J.; Denker, U.; Mönch, I.; Deneke, C.; Schmidt, O. G.; Rampnoux, J. M.; Wang, S.; Plissonnier, M.; Rastelli, A.; Dilhaire, S.; Mingo, N. Nat. Mater. 2010, 9, 491−495. (17) Chen, D.; Liu, F.; Wang, C.; Nakahara, A.; Russell, T. P. Nano Lett. 2011, 11, 2071−2078. (18) Heeger, A. J. Adv. Mater. 2014, 26, 10−27. (19) MacManus-Driscoll, J. L.; Zerrer, P.; Wang, H.; Yang, H.; Yoon, J.; Fouchet, A.; Yu, R.; Blamire, M. G.; Jia, Q. Nat. Mater. 2008, 7, 314−320. (20) Evers, W. H.; De Nijs, B.; Filion, L.; Castillo, S.; Dijkstra, M.; Vanmaekelbergh, D. Nano Lett. 2010, 10, 4235−4241. (21) MacManus-Driscoll, J. L. Adv. Funct. Mater. 2010, 20, 2035− 2045. (22) Nguyen, M. T.; Diaz, A. F. Adv. Mater. 1994, 6, 858−860. (23) King, P. D. C.; Veal, T. D.; Jefferson, P. H.; Zúñiga-Pérez, J.; Muñoz-Sanjosé, V.; McConville, C. F. Phys. Rev. B: Condens. Matter Mater. Phys. 2009, 79, 035203. (24) Speaks, D. T.; Mayer, M. A.; Yu, K. M.; Mao, S. S.; Haller, E. E.; Walukiewicz, W. J. Appl. Phys. 2010, 107, 113706. (25) Walukiewicz, W. Phys. B 2001, 302-303, 123−134. (26) Nishitani, J.; Detert, D.; Beeman, J.; Yu, K. M.; Walukiewicz, W. Appl. Phys. Express 2014, 7, 091201. (27) Brebrick, R. F.; Strauss, A. J. Phys. Rev. 1963, 131, 104−110. (28) Tanaka, Y.; Ren, Z.; Sato, T.; Nakayama, K.; Souma, S.; Takahashi, T.; Segawa, K.; Ando, Y. Nat. Phys. 2012, 8, 800−803. (29) Littlewood, P. B.; Mihaila, B.; Schulze, R. K.; Safarik, D. J.; Gubernatis, J. E.; Bostwick, A.; Rotenberg, E.; Opeil, C. P.; Durakiewicz, T.; Smith, J. L.; Lashley, J. C. Phys. Rev. Lett. 2010, 105, 086404. (30) Hsieh, T. H.; Lin, H.; Liu, J.; Duan, W.; Bansi, A.; Fu, L. Nat. Commun. 2012, 3, 982. (31) Nishitani, J.; Yu, K. M.; Walukiewicz, W. Appl. Phys. Lett. 2014, 105, 132103. (32) Wahila, M. J.; Lebens-Higgins, Z. W.; Quackenbush, N. F.; Nishitani, J.; Walukiewicz, W.; Glans, P. A.; Guo, J. H.; Woicik, J. C.; Yu, K. M.; Piper, L. F. J. Phys. Rev. B: Condens. Matter Mater. Phys. 2015, 91, 205307. (33) Burstein, E. Phys. Rev. 1954, 93, 632−633. (34) Moss, T. S.; Hawkins, T. D. F. Infrared Phys. 1961, 1, 111−115. (35) Berggren, K. F.; Sernelius, B. E. Phys. Rev. B: Condens. Matter Mater. Phys. 1981, 24, 1971−1986. (36) Walukiewicz, W. Phys. Rev. B: Condens. Matter Mater. Phys. 1990, 41, 10218−10220. (37) King, P. D. C.; Veal, T. D.; McConville, C. F.; Zúñiga-Pérez, J.; Muñoz-Sanjosé, V.; Hopkinson, M.; Rienks, E. D. L.; Jensen, M. F.; Hofmann, P. Phys. Rev. Lett. 2010, 104, 256803. (38) Adachi, S. Physical Properties of III−V Semiconductor Compounds; Wiley: New York, 1992.
and 73% of SnTe) by mechanical thinning and dimpling followed by Ar+ ion polish using a Gatan Precision Ion Polishing System operated at 3.0 keV with a 6° milling angle. A final polish using 0.5 keV at shallow angles (2°) was performed before the samples were analyzed. Conventional TEM Micrographs were acquired with a Gatan UltraScan 1000 (2k × 2k) camera, while SAD and nanodiffraction patterns were recorded using a Gatan Orius 830 (2k × 2k). For STEM images, the scattered electrons were captured by a Fischione Model 3000 ADF detector. EDX spectra were collected by using an FEI SuperX quad windowless detector based on silicon drift technology with a solid angle of 0.7 steradians for about 15 min. Synchrotron X-ray Diffraction. The crystalline structure of the films was analyzed with wide-angle X-ray scattering (WAXS) on Beamline 11-3 at the Stanford Synchrotron Radiation Lightsource (SSRL). Two-dimensional scattering was collected with an MAR345 image plate detector and X-ray energy of 12.7 keV. Spectra were integrated between 5° < χ < 175° (χ is the polar angle) using the WxDiff software package.45 The grazing incidence angle was 2°.
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.6b02395. Determination of the intrinsic band gap, microstructure analysis, composition of (SnTe)x(CdO)1−x composites, Synchrotron X-ray diffraction (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was performed at the EMAT, National Center for Electron Microscopy/Molecular Foundry and was supported by the Director, Office of Science, Office of Basic Energy Sciences, Materials Sciences and Engineering Division, of the U.S. Department of Energy under Contract No. DE-AC0205CH11231. Use of the Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, is supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences under Contract No. DE-AC02-76SF00515. W.W.G., V.L.P., and L.T.S. acknowledge partial support from the U.S. Department of Energy through the Bay Area Photovoltaic Consortium under Award No. DEEE0004946. R.R. acknowledges support from CAPES/BR, BEX process number 12047-13-9. We acknowledge C. Gammer, Z. Anderson, and P. Ercius who made contribution to the development of the code used to acquire the nanobeam diffraction datasets.
■
REFERENCES
(1) Dagotto, E. Science 2005, 309, 257−262. (2) Treat, N. D.; Brady, M. A.; Smith, G.; Toney, M. F.; Kramer, E. J.; Hawker, C. J.; Chabinyc, M. L. Adv. Energy Mater. 2011, 1, 82−89. (3) Balazs, A. C.; Emrick, T.; Russell, T. P. Science 2006, 314, 1107− 1110. G
DOI: 10.1021/acs.nanolett.6b02395 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters (39) Brasil, M. J. S. P.; Tamargo, M. C.; Nahory, R. E.; Gilchrist, H. L.; Martin, R. J. Appl. Phys. Lett. 1991, 59, 1206−1208. (40) Burbano, M.; Scanlon, D. O.; Watson, G. W. J. Am. Chem. Soc. 2011, 133, 15065−15072. (41) 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. J. Am. Chem. Soc. 2007, 129, 9780−9788. (42) Walukiewicz, W. Phys. Rev. B: Condens. Matter Mater. Phys. 1988, 37, 4760−4763. (43) Walukiewicz, W. Appl. Phys. Lett. 1989, 54, 2094. (44) Yu, K. M.; Mayer, M. A.; Speaks, D. T.; He, H.; Zhao, R.; Hsu, L.; Mao, S. S.; Haller, E. E.; Walukiewicz, W. J. Appl. Phys. 2012, 111, 123505. (45) Mannsfeld, S. C.; Tang, M. L.; Bao, Z. Adv. Mater. 2011, 23, 127−131.
H
DOI: 10.1021/acs.nanolett.6b02395 Nano Lett. XXXX, XXX, XXX−XXX
