Understanding the Chemical Nature of the Buried Nanostructures in

2 hours ago - The valence band features near the edge do not change significantly with Sb doping. This understanding of chemical nature of low lattice...
0 downloads 0 Views 924KB Size
Subscriber access provided by UNIV OF SOUTHERN INDIANA

C: Energy Conversion and Storage; Energy and Charge Transport

Understanding the Chemical Nature of the Buried Nanostructures in Low Thermal Conductive Sb-Doped SnTe by Variable Energy Photoelectron Spectroscopy Anamul Haque, Ananya Banik, Rahul Mahavir Varma, Indranil Sarkar, Kanishka Biswas, and Pralay K. Santra J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.9b01081 • Publication Date (Web): 29 Mar 2019 Downloaded from http://pubs.acs.org on March 29, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Understanding the Chemical Nature of the Buried Nanostructures in Low Thermal Conductive Sb-doped SnTe by Variable Energy Photoelectron Spectroscopy Anamul Haque,‡,# Ananya Banik,¶ Rahul Mahavir Varma,† Indranil Sarkar, ⊥,§ Kanishka Biswas,*,¶ Pralay K. Santra*,‡ ‡

Centre for Nano and Soft Matter Sciences (CeNS), Jalahalli, Bengaluru-560013, India. #



Manipal Academy of Higher Education (MAHE), Manipal-576104, India.

New Chemistry Unit, Jawaharlal Nehru Centre for Advanced Scientific Research (JNCASR), Jakkur, Bengaluru-560064, India.



Solid State and Structural Chemistry Unit, Indian Institute of Science (IISc), Bengaluru-560012, India. ⊥DeutschesElektronen-Synchrotron §Presently

DESY, Notkestrasse 85, D-22607 Hamburg, Germany

at Institute of Nano Science and Technology, Phase-10, Sector-64, Mohali, India *E-mail: [email protected]; [email protected]

1 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract. Nanoprecipitates embedded in a matrix of thermoelectric materials decrease the lattice thermal conductivity significantly by extensive heat carrying phonon scattering. Recently, two dimensional (2D) layered intergrowth nanostructures of SnmSb2nTe3n+m embedded in SnTe matrix provided record low lattice thermal conductivity in SnTe, but an understanding of chemical nature of these layered nanostructures is still not clear. Herein, we studied the chemical nature the intergrowth nanostructures of a series Sb-doped SnTe by variable energy X-ray photoelectron spectroscopy (XPS) at synchrotron, that is well known to probe buried interfaces and embedded nanostructures. The primary oxidation states of Sb, Sn and Te in these intergrowth structures is found to be in +3, +2 and -2 respectively, which is expected from the composition. However, both the Sn and Sb are found to be slightly oxidized in the surface. From the intensity variation with photon energy, we have found a thin layer of SnO2 (~ 4.5 nm) on the sample surfaces and the thickness decreases with Sb doping. Te is also found in 0 oxidation states which corroborates with the variation of Sn vacancies with Sb doping. The valence band features near the edge do not change significantly with Sb doping. This understanding of chemical nature of low lattice thermal conductive Sb doped SnTe will help in further to design the thermoelectric materials with their surface phenomenon.

2 ACS Paragon Plus Environment

Page 2 of 25

Page 3 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Introduction Solids with low thermal conductivity are crucial to the advancement of various fields like thermoelectrics, thermal barrier coatings, etc. Thermal transport is considered to be a limiting factor to realize efficient thermoelectric materials.1-2 Thermal conductivity of any material can be divided into two part, electronic thermal conductivity (κel) and lattice thermal conductivity (κlat).1 Since, electronic thermal conductivity (κel) is directly proportional to the electrical conductivity via Widemann-Franz law, reduction of κlat via scattering of phonons propagation has been the main focuses of recent thermoelectric research.3 Extrinsic strategies such as solid solution alloying,4-5 all-scale hierarchical nano-mesostructuring3 and embedding endotaxial nanostructures in the bulk matrix6 have been proven effective to suppress the lattice thermal conductivity (κlat) to the minimum. On the other hand, solids with intrinsically low κlat are attractive being capable of offering nearly independent control over electrical transport.7 Investigations on minimal κlat in certain thermoelectric solids have unveiled non-traditional phonon-scattering mechanisms such as complex crystal structures,8 part-crystalline part-liquid state,9 bonding asymmetry,10-11 superionic substructure with liquid-like cation disordering,12-14 lone-pair induced bond anharmonicity,15 ferroelectric instability induced phonon softening16 and anisotropic layered crystal structure17-18. Stabilization of two dimensional (2D) layered intergrowth heterostructure compounds in the form of nanodomian is an effective strategy to reduce lattice thermal conductivity, where additional phonon scattering occurs from heterostructured layers and interfaces.19 Tin telluride (SnTe) has recently attracted huge attention as an environment friendly alternative of the champion thermoelectric material, lead telluride (PbTe).20-21 Sb-alloyed SnTe shows ultralow κlat (~0.67 W/mK) resulting from the formation of the layered 2D nanostructures of homologous SnmSb2nTe3n+m in the SnTe matrix.22 SnmSb2nTe3n+m originates from the layered intergrowth 3 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

family, AmIVB2nVTem+3n (AIV = Ge/Sn/Pb; BV = Sb/Bi), where m and n represent stoichiometry of ATe and B2Te3 respectively.23 These compounds crystallize in tetradymite Bi2Te2S-type structures with the large unit cell, which can also be visualized as intergrowths of SnTe-type (rocksalt) and Sb2Te3-type (hexagonal) phases. The intergrowth family accommodates a broad diversity of mixed-layered structures. However, it is difficult to synthesize these phase-pure intergrowth compounds via conventional high-temperature solid-state melting reactions due to the incongruent melting point of most of them.22-23 Stabilization of nanostructured SnmSb2nTe3n+m in SnTe matrix and its impact on the thermal conductivity of SnTe have motivated us to study the Sb-alloyed SnTe system further to realize the the chemical nature of the nanostructures, local environment and its interaction with the SnTe matrix. The detailed microscopic understanding from high resolution transmission electron microscopy revealed the formation of endotaxial nanoprecipitates in lower Sb concentration (~4%) in SnTe, whereas a higher concentration of Sb (~15 %) forms nanodomains of layered intergrowth Sb-rich SnmSb2nTe3n+m compounds.22 The formation of such nanostructures within the SnTe matrix significantly enhances the phonon scattering, which ultimately leads to achieving the theoretical minima of lattice thermal conductivity, κmin. Although HRTEM studies revealed nanostructures formation, the chemical nature of these buried nanoprecipitates is still not clear to the research community. In this context, we recall that X-ray photoelectron spectroscopy (XPS) has been extensively used to study the electronic and chemical nature of bulk materials as well as surfaces and thus also known as electron spectroscopy for chemical analysis (ESCA).24 Along with the determination of chemical composition, variable energy XPS has also been used to determine the internal buried heterostructures of various nano structures, e.g., core-shell,25 quantum dot-quantum well,26 cesium 4 ACS Paragon Plus Environment

Page 4 of 25

Page 5 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

lead halide perovskites27-28 etc. This was possible as the dimension of the heterostructures are comparable to the mean free path of the photoelectron. One can easily vary the mean free path of the photoelectrons by changing the incident photon energy.29 This method is not only restricted to nanoparticles and recently Mukherjee et. al,30 have probed the boron diffusion in CoFeB/MgO magnetic tunnel junctions using variable energy photoelectron spectroscopy. Thus, we though to explore variable energy XPS as a tool to investigate the embedded nanostructures in SnTe. Herein, we have studied the chemical nature of pristine SnTe, 4% Sb-doped SnTe exhibiting the nanoprecipitates and 15 % Sb-doped SnTe, exhibiting the nanodomains of layered intergrowth using variable energy (1487 - 6000 eV) photoelectron spectroscopy. The results suggest that both Sn and Sb (in the case of doped samples) are present primarily in +2 (SnTe) and +3 (Sb2T3) oxidation states respectively. Additionally, a thin layer of oxidized SnO2 and Sb2O3 are present on the sample surface. The photon energy dependent intensity variation suggests that the oxidized layers are only a few nanometers thick and the thickness decreases with increase in Sb doping. With Sb doping, there are hardly any changes in the valence band edge, however, there is a slight shift in the valence band edge which corroborates the effect of Sb doping in SnTe, that probably decreases the excess p-type carrier and optimize its electronic transport.

5 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Experimental Methods: 2.1 Reagents. We used tin (Sn, Alfa Aesar 99.99+ %), tellurium (Te, Alfa Aesar 99.999+ %), and antimony (Sb, Alfa Aesar 99.999+ %) for synthesis of Sn1-xSbxTe. 2.2 Synthesis. High-quality ingots (~7 g) of Sn1-xSbxTe (x = 0-0.15) were synthesized by the solidstate melting reaction of appropriate stoichiometric ratios of high-purity starting materials of Sn, Sb, and Te in a quartz tube. The tubes were sealed under high-vacuum (10−5 Torr) and slowly heated to 900 oC over 12 hrs, then kept for 10 hrs, and slowly cooled to room temperature. 2.3 Powder X-ray diffraction. Powder X-ray diffraction was recorded on a Bruker D8 diffractometer using a Cu Kα (λ = 1.5406 Å) radiation. 2.4 TEM measurements. TEM imaging was carried out using an aberration-corrected FEI TITAN cubed 80-300 kV transmission electron microscope operating at 300 kV. We prepared TEM samples by conventional mechanical thinning followed by Ar ion milling to perforation to generate large electron-transparent thin area. 2.5 Hard X-ray photoemission spectroscopy. The hard X-ray photoemission spectroscopy (HAXPES) measurement was performed on freshly prepared series of Sb-doped SnTe sample at HAXPES end station of P09 beamline at PETRA-III synchrotron, DESY, Hamburg, Germany using 3700, 4750, and 5946 eV photon energies. The low energy XPS measurements were performed in a home-built spectrometer using Al-Ka (hn = 1486.7 eV) as the photon source and hemispherical analyzer with a constant pass energy of 20 eV. For these measurements, the sample surfaces were scraped using a diamond file inside the ultra-high vacuum (UHV) chamber. Ultraviolet photoelectron spectroscopy (UPS) was carried out in the same instrument with He-I as the photon source. All spectra were calibrated with respect to the binding energy of Au 4f7/2 at 84.0 eV and deconvoluted by using X-ray Photoelectron Spectroscopy Tool (XPST) program 6 ACS Paragon Plus Environment

Page 6 of 25

Page 7 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

package which has been developed for IGOR Pro by Dr. Martin Schmid, Philipps University, Marburg. The background was considered as the combination of Shirley and parabolic functions. The core level peaks were considered to have pseudo-Voigt functional form. In fitting, the Gaussian-Lorentzian ratio was varied from 0.05 to 0.95 and the intensity ratio of the spin orbit split peaks were fixed to 0.667 for d orbital core levels and 0.75 for f orbitals core levels.

7 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Results and Discussion Polycrystalline ingots of Sn1-xSbxTe (x = 0 – 0.15) are prepared in vacuum sealed quartz tube at 900 °C. The room temperature powder X-ray diffraction (PXRD) patterns of Sn1-xSbxTe (x = 0 – 0.15) (Figure 1a) confirm the face centered cubic (Fm-3m) structure for low Sbconcentration. However, for the 15 mol% Sb-doped SnTe, small intensity secondary peaks appear (as shown in Figure 1b), which indicates to the formation of Sb-rich layered intergrowth

Figure 1. (a) PXRD patterns of Sn1-xSbxTe (x = 0-0.15) samples. (b) Zoomed PXRD patterns of Sn0.85Sb0.15Te with Sn-based layered intergrowth compound, SnmSb2nTe3n+m. (c) Low-magnification TEM image of sample Sn0.96Sb0.04Te showing nanoscale precipitates. (d) HRTEM image of Sn0.96Sb0.04Te sample, confirming coherent matrix-precipitate interface. Precipitate is marked by the circle. (e) HRTEM image of Sn0.85Sb0.15Te with the nanodomains of layered intergrowth nanostructures.

8 ACS Paragon Plus Environment

Page 8 of 25

Page 9 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

compounds originating from the homologous series of quasi-binary (SnTe)m(Sb2Te3)n systems, i.e. SnmSb2nTe3n+m. Figure 1c-e present TEM image of Sn1-xSbxTe. Although the PXRD pattern of Sn1-xSbxTe for x=0.04 indicate “single phase” nature, TEM images confirm the formation of coherent nanoscale precipitates (Figure 1c and d). With increasing Sb concentration, nanoprecipitate nature transforms from nanodots to layered nanodomains, which are visible in HRTEM image of Sn0.85Sb0.15Te sample (see Figure 1e). Previously, electron diffraction patterns showed the presence of superstructure spot along direction of cubic Sn0.85Sb0.15Te sample indicated local Sb segregation along the directions of cubic SnTe and formation of nanostructured layered intergrowth compound, SnmSb2nTe3n+m.22 Hence, the presence of superlattice spots along direction and extra peak in PXRD pattern of Sn0.85Sb0.15Te indicate the formation of layered intergrowth nanostructures of Sb-rich SnmSb2nTe3n+m compounds. To understand, the chemical nature of these thermoelectric materials, we have performed variable energy hard X-ray photoemission spectroscopy (HAXPES) of 4% and 15% Sb-doped SnTe and compared with undoped SnTe and Sb2Te3. These two dopant concentrations were chosen based on different heterostructures that we had observed under TEM, namely conventional nanoprecipitate and layered nanostructured intergrowth for 4% and 15% Sb dopant concentrations respectively. The survey scans of 4% and 15% Sb-doped SnTe along with undoped SnTe are shown in Figure S1 in Supporting Information (SI) containing all the expected core level photoemission peaks. These spectra were collected at 3700 eV photon energy and normalized with Te 3d5/2 signal at 572.3 eV. We found a small amount of C 1s signal at 285 eV from all the samples, which comes from the adsorbed carbonaceous species on the sample surface. A significant amount of O 1s was also observed in all the samples. To verify whether the surface got oxidized, we have

9 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

collected the survey scans before and after scraping the sample surface with a diamond file (insitu scraping) inside the ultra high vacuum (UHV) chamber in a lab source (Al – Ka; hn = 1486.7 eV) instrument. Ion sputtering was not used to clean the surface as sputtering causes Sn vacancies in SnTe in an uncontrolled manner.31-32 The survey scan spectra at lower photon energy are shown in Figure S2 in SI. Both the C 1s and O 1s signal from the sample surface diminishes with scraping which suggest that only the sample surfaces are contaminated or oxidized. In order to understand the chemical nature of different species, we have collected the specific core levels at three different photon energies – 3700 eV, 4750 eV, and 5946 eV. High-resolution Sn 3d core level spectra for each sample are shown in Figure 2. We have collected both 3d5/2 and 3d3/2 core levels that shows a spin-orbit splitting of 8.4 eV as reported

Figure 2. Core-level photoemission spectra of Sn 3d5/2 from (a) undoped SnTe, (b) 4 % Sb-doped SnTe and (c) 15 % Sb-doped SnTe collected at different photon energies. The experimental data are shown by black circles, and the solid red line represents the total fit, which is the sum of Sn2+ from SnTe (highlighted by blue shade) and Sn4+ from SnO2 (highlighted in green shade) and Shirley background (not sown in the figure).

earlier.33 For better presentation, only the 3d5/2 core levels are shown here. For all samples, we observed two types of Sn with binding energies 487.0 eV and 485.3 eV. Based on previous 10 ACS Paragon Plus Environment

Page 10 of 25

Page 11 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

reports,31 we assigned the higher binding energy peak to Sn4+ that could originate from SnO2, whereas the low binding energy peak corresponds to Sn2+ from SnTe.33-34 To further confirm we collected Sn 3d5/2 core level with Al – Kα photon energy without and with in-situ scraping the sample surface, as shown in Figure S3 in SI. With unscraped sample, only one type of Sn core level corresponds to Sn4+ is observed; whereas the same binding energy peak vanishes completely after scraping the sample surface. Instead we observed only the low binding energy peak that corresponds to Sn2+. The sample surfaces were not scraped for HAXPES measurements to determine the thickness of the oxide layer as discussed later in the manuscript. To understand these results quantitatively, we deconvoluted the Sn 3d5/2 spectrum into two components corresponding to Sn2+ from SnTe, and Sn4+ from SnO2 as shown by the blue and green shades respectively. We have taken utmost care in deconvoluting these core levels in a selfconsistent manner as mentioned in details in the experimental section. The details of the fitting parameters are mentioned in Table S1 in SI. The systematic changes in intensity variation as a function of photon energy as well as doping concentration are mentioned in Table S2 in SI, and we have discussed later in the manuscript. Figure 3 shows core-level photoemission spectra of Te 3d5/2 from three samples at different photon energies. In this case, Te 3d core level exhibit a spin-orbit coupling of 10.4 eV. Here for the better presentation, we have shown only the 3d5/2 core levels. Two different species of Te with the binding energies of 572.3 eV and 573.0 eV are present in all three samples. These two oxidation states can be assigned to “-2” (SnTe; shown in blue shade) and “0” (metallic Te; shown in magenta shade) based on previous reports.33 The higher oxidation state of Te (i.e., metallic) is due to Sn vacancies as the Te next to these vacancies will have non-bonding orbital which increases its binding energy. We have not observed any peak corresponding to TeO2 which 11 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

has typically binding energy of 576.6 eV. The variation in their relative intensities with photon energy and doping concentrations are mentioned later. As we have observed oxidized surface species, SnO2 in Sn 3d core level spectra, the higher resolution O 1s core level was collected for all three samples as a function of photon energy. For undoped SnTe, two distinct peaks are observed with binding energy at 530.7 eV and 532.5 eV shown by yellow and orange color in Figure 4a. Similar O 1s core levels with different species at

Figure 3. Core level photoemission spectra of Te 3d5/2 from (a) undoped SnTe, (b) 4%Sb doped SnTe, and (c) 15% Sb doped SnTe sample at 3700, 4750, and 5946 eV photon energies. The black circles represent the experimental data. Total fit is shown by the solid red line. Two components, Te2- and Te0 are shown in blue and magenta shades, which are originated from SnTe and Te next to Sn vacancies.

530.8 eV and 532.3 eV have also been reported in ZnO.35-36 It was established that the higher binding energy oxygen species is due to the formation of metal hydroxide on the surface of the samples.36-37 We also collected the core level O 1s at lower photon energy (Al-Ka) as shown in Figure S4 in SI. The unscraped sample shows two types of oxygen species as observed at higher photon energies. However, the relative intensity of the metal hydroxide species is higher compared

12 ACS Paragon Plus Environment

Page 12 of 25

Page 13 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

to the metal oxide species. After scraping the sample surfaces, we could not observe any detectable amount of O 1s – which supports that both the oxygen species are present on the sample surface. The O 1s core level spectra for 4% and 15% Sb-doped SnTe are shown in Figure 4b and c respectively. But, O 1s and Sb 3d5/2 core levels have similar binding energies, and they overlap with each other and not straightforward to analyze. To facilitate this we have collected the Sb 3d3/2 core level as they do not overlap with any other core level peaks. We carefully deconvoluted the

Figure 4. Core level photoemission spectra of (a) O 1s from undoped SnTe. Two different O 1s species correspond to metal oxide (yellow) and metal hydroxide (orange) species. O 1s and Sb 3d5/2 and Sb 3d3/2 from (b) 4% Sbdoped SnTe and (c) 15% Sb-doped SnTe which comprises of two oxygen species and two Sb species corresponds to Sb2Te3 (blue) and Sb2O3 (green).

core level spectra with two types of O 1s and two types of Sb 3d evident from Sb 3d3/2 core level. The binding energy for O 1s for both the dopant samples matches with undoped SnTe sample as shown in yellow and orange color shade. The spin-orbit splitting of Sb 3d core level is found to be 9.4 eV.38 For 4% Sb doped sample, the binding energy of Sb 3d5/2 is 528.6 eV that corresponds to the Sb2Te3. On careful observation, we found a minute quantity of oxidized Sb 3d5/2at 530.4 eV. This oxidized species is evident in Sb 3d3/2 peak. With increase in dopant concentration, the relative 13 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

intensity of Sb 3d core level has gone up, however, the binding energies of different species remain unchanged. The calculated relative intensities are mentioned in Table S2 in SI. We find that with change in photon energy as well as change in dopant concentrations, the relative intensity of different species changes and we have explained in the next section. The relative intensity ratios of different species are shown in Figure 5 as a function of photon energy for undoped and Sb-doped SnTe samples. From Sn 3d5/2 core level analysis, it was found that with increase in photon energy the relative intensity of Sn4+ with respect to Sn2+ decreases as shown in Figure 5a. We also noticed that with increase in Sb doping the relative oxidation of SnTe decreases. The relative intensity variation of Te0 with respect to Te2- with photon energies are plotted in Figure 5b. We find that there is a slight decrease in the intensity ratio of Te0/Te2- at 4750 eV; however, it increases with increases in photon energy as mentioned in Table S2. Considering the small changes compared to other species (e.g., Sn4+/Sn2+ and metal hydroxide/metal oxide), we viewed the variation is almost uniform throughout the sample, which in turn, indicates that the change in Te0/Te2- is intrinsic to the sample and not a surface oxidation phenomena. This further supports the origin of Te0 is due to Sn vacancies which are present in SnTe. With an increase in Sb doping, the metallic Te intensity is decreasing slightly that is due to decrease in Sn vacancy with increase Sb doping in SnTe as reported earlier.22 The metal hydroxide to metal oxide intensity variation is shown in Figure 5c. In this case, we found a significant change in relative intensities with photon energy. The low photon energy experiment with scraped samples suggested both these species are on the surface, however the relative concentrations of the metal hydroxides are more on the surface of the samples compared to the metal oxide. The relative intensity ratio of Sb2O3 with respect to Sb2Te3 is shown in Figure 5d for Sb-doped samples and there are hardly any differences in the relative intensity ratio at two different dopant 14 ACS Paragon Plus Environment

Page 14 of 25

Page 15 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Figure 5. (a) Experimentally calculated intensity ratio of Sn4+/Sn2+ as a function of photon energy for three samples (shown by symbols) along with the calculated variations in the intensity ratio obtained for different oxide layer thickness (solid lines). With an increase in Sb doping, the relative concentration of Sn4+ decreases. Relative intensity variation of (b) Te0/Te2-, (c) O-/O2- and (d) Sb3+ (Sb2O3) /Sb3+ (Sb2Te3) as a function of photon energy for different samples. The intensity ratio scale is kept same for all the species to promptly identify the variations as a function of photon energy. The data with expanded scale is shown in Figure S5 in the supporting information. (e) The schematic representation of SnO2 overcoated on top of SnTe. The thickness of the SnO2 layer (t) was calculated from the intensity variation of Sn4+/Sn2+ at different photon energies as described in the main text.

concentrations. However, with the increase in photon energy, the relative ratio decreases suggesting that Sb2O3 layer is more on the surface of samples as expected. To obtain quantitative information on the layer thickness of SnO2 on top of SnTe, we followed the procedures described earlier39 using the core level peak intensities measured at three different photon energies. The differential photoemission intensity (dI) of a specific core level from a depth z from the surface within a volume element dv can be expressed as 𝑑𝐼 = 1

𝜎𝐼) 𝑛(𝑧)𝑒 /023456 7 𝑑𝑣, where λ is the mean free path of the photoelectron, σ is the photoemission cross section of the particular core level, n(z) is the number density of the element. Since we have not observed any SnTe from the unscraped samples at lower photon energy, we assumed that the SnO2 layer is only on the surface and forms a sharp interface as shown schematically in Figure 5e. If the thickness of SnO2 is “t”, on top of SnTe, the photoemission intensity of Sn4+ can be expressed as: 15 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

>

𝐼9:;< = 𝜎𝐼) ∫) 𝑛9:;< 𝑒 / 2 𝑑𝑧

Equation 1

and for Sn2+ can be expressed as 1



𝐼9:?< = 𝜎𝐼) ∫> 𝑛9:?< 𝑒 / 2 𝑑𝑧

Equation 2

where, 𝑛9:;< = 1, 0 < 𝑧 ≤ 𝑡 and 𝑛9:;< = 0, 𝑧 > 𝑡 𝑛9:?< = 0, 0 < 𝑧 ≤ 𝑡 and 𝑛9:?< = 1, 𝑧 > 𝑡 With these conditions, the intensity ratio can be expressed as FGH;< FGH?