In Situ XPS Investigation of Transformations at Crystallographically

Subscriber access provided by UNIVERSITY OF CONNECTICUT

Article

In Situ XPS Investigation of Transformations at Crystallographically-Oriented MoS2 Interfaces Neha P Kondekar, Matthew G. Boebinger, Eric V Woods, and Matthew T. McDowell ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.7b10230 • Publication Date (Web): 28 Aug 2017 Downloaded from http://pubs.acs.org on September 2, 2017

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 free 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 accessible to all readers and 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.

ACS Applied Materials & Interfaces 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 45

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

ACS Applied Materials & Interfaces

In Situ XPS Investigation of Transformations at Crystallographically-Oriented MoS2 Interfaces Neha P. Kondekar†, Matthew G. Boebinger†, Eric V. Woods‡, Matthew T. McDowell ⃰ †,§

†

School of Materials Science and Engineering, Georgia Institute of Technology, 771 Ferst Drive,

Atlanta, GA 30332, USA ‡

Materials Characterization Facility, Institute for Electronics and Nanotechnology, Georgia

Institute of Technology, 345 Ferst Drive, Atlanta, GA 30332, USA §

G.W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology, 801 Ferst

Drive, Atlanta, GA 30332, USA

ACS Paragon Plus Environment

1


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

Page 2 of 45

ABSTRACT Nanoscale transition metal dichalcogenide (TMDC) materials, such as MoS2, exhibit promising behavior in next-generation electronics and energy storage devices. TMDCs have a highly anisotropic crystal structure, with edge sites and basal planes exhibiting different structural, chemical, and electronic properties. In virtually all applications, two-dimensional or bulk TMDCs must be interfaced with other materials (such as with electrical contacts in a transistor). The presence of edge sites vs. basal planes (i.e., the crystallographic orientation of the TMDC) could influence the chemical and electronic properties of these solid-state interfaces, but such effects are not well understood. Here, we use in situ x-ray photoelectron spectroscopy (XPS) to investigate how the crystallography and structure of MoS2 influence chemical transformations at solid-state interfaces with various other materials. MoS2 materials with controllably aligned crystal structures (horizontal vs. vertical orientation of basal planes) were fabricated, and in situ XPS was carried out by sputter-depositing three different materials (Li, Ge, and Ag) onto MoS2 within an XPS instrument while periodically collecting photoelectron spectra; these deposited materials are of interest due to their application in electronic devices or energy storage. The results showed that Li reacts readily with both crystallographic orientations of MoS2 to form metallic Mo and Li2S, while Ag showed very little chemical or electronic interaction with either type of MoS2. In contrast, Ge showed significant chemical interactions with MoS2 basal planes, but only minor chemical changes were observed when Ge contacted MoS2 edge sites. These findings have implications for electronic transport and band alignment at these interfaces, which is of significant interest for a variety of applications. KEYWORDS: In situ XPS, interfacial reactions, MoS2, transition metal dichalcogenide, phase transformations


2

Page 3 of 45

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


INTRODUCTION MoS2 is a member of the semiconducting layered transition metal dichalcogenide (TMDC) family of materials. It features covalently bonded Mo-S-Mo layers that are stacked and held together with relatively weak Van der Waals forces. Like graphene, ultrathin layers of MoS2 can be obtained using micromechanical cleavage,1 lithium-based intercalation and exfoliation,2,3 and chemical vapor deposition.4–8 Single- and few-layer TMDCs show different electronic properties than their bulk counterparts, potentially allowing for tunable performance in various applications. Upon decreasing thickness, for instance, MoS2 undergoes an indirect-to-direct band gap transition from 1.29 eV in the bulk to 1.90 eV as a single monolayer.9–11 This tunable band gap has stimulated interest for the use of MoS2 and other TMDCs for a range of applications, including photodetectors,12 photocatalytic systems,13 and photovoltaics.14 MoS2 has also been shown to be a superior channel material in novel field effect transistors.15–21 Beyond electronic applications, MoS2 and other TMDCs (e.g., WS2 and MoSe2) have been shown to exhibit promising performance in electrochemical energy conversion and storage systems. MoS2 exhibits low overpotentials for the hydrogen evolution reaction (HER) during water splitting;22 in addition, nanostructured TMDCs can serve as reversible electrode materials for Li-ion batteries.23 MoS2, in combination with Ni or Co additives, is also used as a catalyst for the hydrodesulfurization reaction to remove sulfur from petroleum feedstocks.24 In short, MoS2 and other TMDCs have the potential to impact a wide variety of electronic and energy devices. In virtually all applications, it is necessary to interface TMDCs with other materials to enable device operation. For instance, TMDC-based transistors require ohmic metal contacts for charge injection, and the formation of low-resistance contacts has previously been studied with metals such as Au and Pd.25,26 Ag, a similar noble metal, is another attractive contact material


3


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

Page 4 of 45

and has been shown to have better wettability and improved electrical characteristics when interfaced with MoS2 as compared to other metal contacts like Au/Ti.27 Other electronic devices require thin TMDC layers to be deposited on bulk semiconductor substrates (e.g., Si28 or Ge29), and a TMDC/Si or TMDC/Ge interface is hence inevitable. In particular, the interaction of Ge with TMDCs and other 2D materials is important to understand, as Ge-based materials have been shown to be beneficial for electronic properties and growth characteristics.30,31 In Li-ion batteries, the reaction of MoS2 with Li involves Li intercalation between the TMDC layers, followed by a “conversion reaction” to form an interspersed mixture of Li2S and Mo metal.32,33 This type of conversion reaction in chalcogenide materials is known to involve the movement of a sharp interface separating the lithium-poor and lithium-rich phases;34,35 thus, understanding the properties of such interfaces is important for the development of TMDCs for batteries. Unique electronic, physical, and chemical properties arise from the highly anisotropic crystal structure of TMDCs. As shown in Fig. 1a, covalently bonded layers are stacked to form a hexagonal crystal structure, which exposes the basal planes and the edge sites of the layers. The different bonding environment at the basal planes compared to the edge sites gives rise to different physico-chemical properties.28,36 For instance, it is known that the edge sites can be more active for catalysis,22,37–39 and that these sites act as recombination centers for photogenerated carriers in photoelectrochemical or optoelectronic devices.40 It has thus proved beneficial to engineer MoS2 nanomaterials to feature a preponderance of either edge sites or basal plane sites for specific applications.38 Since the edge sites and basal plane sites of TMDCs exhibit different structural and electronic properties, it may therefore be expected that the physico-chemical interactions of various materials interfaced with TMDCs will depend on the crystallographic orientation of the TMDC. In particular, the extent of chemical interactions or


4

Page 5 of 45

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


structural changes at the interface between a TMDC and another material may depend on whether the material is interfaced with edge sites or basal planes. While there have been a handful of studies investigating chemical and physical interactions at solid-state TMDC interfaces,41–47 there is a particular dearth of knowledge related to the effect of TMDC crystallographic orientation on interfacial interactions. Here, we use in situ x-ray photoelectron spectroscopy (XPS) to directly reveal the interfacial chemical changes when MoS2 materials with different crystallographic orientation and structural properties are interfaced with various other materials (Li, Ag, and Ge). These three materials are chosen because their interfaces with TMDCs are important for different technological applications (electronic devices and electrochemical energy storage), and they have widely different chemical properties. This study is enabled by the fabrication of MoS2 specimens with controlled crystallographic orientation, as well as by the use of in situ XPS methods to deposit these materials and probe chemical changes under vacuum conditions without air exposure. The results show that the extent of interfacial reaction depends on MoS2 orientation for some contact materials, but not for others; furthermore, the varied chemistry of different MoS2 surfaces likely contributes to these differences. This study has important ramifications for engineering TMDC interfaces and integrating TMDC materials within next-generation electronics and energy storage devices.

EXPERIMENTAL METHODS Sample Preparation: Two types of MoS2 samples with different crystallographic orientations were fabricated for this study. Horizontally aligned MoS2 samples were produced by exfoliating


5


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

Page 6 of 45

few-layer films from bulk MoS2 single crystals (SPI Supplies). These samples were mechanically exfoliated with Scotch tape and then transferred to a sample holder for XPS experiments and attached with carbon tape. This procedure was performed inside an Ar-filled glove box (MBraun) to minimize surface oxidation. For fabrication of vertically aligned thin films of MoS2, a heavily doped, p-type Si wafer (Addison) with orientation was first cleaned by sonicating in a 1:1:1 volumetric mixture of isopropanol, acetone, and methanol. A 50 nm thick film of Mo was evaporated onto the wafer using a Denton electron beam evaporator at a pressure of 10-5 Torr. A relatively low deposition rate of 0.1 Å/s was used to ensure minimal surface roughness. This Mo thin film was then sulfurized in an inert atmosphere with an elemental sulfur source.28 In a single-zone tube furnace (Lindberg, Blue M), a ceramic boat containing 2.0 g of elemental sulfur powder (Sigma Aldrich) was placed inside a 2.54 cm diameter quartz tube. The sulfur source material was located 16.5 cm upstream from the center of the tube. Si wafer samples coated with Mo were placed in the tube at the center of the heating zone. The system was sealed and evacuated to a pressure of 25 mTorr, and it was then flushed with Ar at a flow rate of 110 sccm for 15 min before heating. The chamber was then ramped to 700°C over a period of 15 minutes, during which Ar was continuously flowed at a rate of 50 sccm. After this heating period, the furnace temperature was held at 700°C for an additional 30 minutes under the same Ar flow. It was then naturally cooled to room temperature under 110 sccm Ar flow. During this process, the sulfur source material vaporized and reacted with the Mo to yield MoS2 thin films with vertically aligned layers. Upon venting the furnace, the samples were immediately transferred to the glove box to minimize surface oxidation. Characterization of the surface of these MoS2 thin films revealed the presence of a SiOx interfacial layer, which will be discussed subsequently. This layer was removed by etching the as-synthesized material in 2 vol % HF(aq)


6

Page 7 of 45

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


for 1 min. As discussed later, this etching step did not affect the surface chemistry of the MoS2 material itself. Characterization: Transmission electron microscopy (TEM) characterization was carried out using an FEI Tecnai F30 microscope operating at 300 kV, as well as a Hitachi HD 2700 scanning TEM operating at 200 kV. For through-film (plan view) TEM imaging, MoS2 thin films were grown using the sulfurization method described above, but on custom ~40 nm thick SixN window TEM supports instead of Si wafers. Mo films were e-beam deposited directly onto the TEM supports. Cross-sectional TEM samples were fabricated via focused ion beam (FIB) milling of conventional samples using an FEI Nova Nanolab 200 FIB (30kV, 0.3 nA milling current). In Situ X-ray Photoelectron Spectroscopy and Data Analysis: The in situ XPS measurements were carried out using a Thermo K-Alpha XPS instrument. This XPS features a monochromated Al-Kα source with a 400 µm diameter spot size and 15 W x-ray gun power. The base pressure was 4.5×10-8 Torr, and the pressure within the chamber remained below 1.8×10-7 Torr throughout the experiments. The analyzer pass energy was set to 50 eV with a resolution of 0.05 eV and a dwell time of 100 ms. During these high resolution XPS measurements, the sample was flooded with slow electrons and Ar ions using the flood gun to compensate for surface charging. Since the C 1s peak underwent changes during the experiments due to deposition of overlayers, charge referencing was performed using the S 2s peak (226.3 eV)48 prior to any metal deposition on each of the bare MoS2 samples.

Using the methodology reported by Wenzel et al.,49 an L-

shaped sample holder was used to allow for the sputter deposition of the three different target materials (Li, Ag, and Ge) onto the MoS2 samples within the XPS, as illustrated in Fig. 1b. The holder is designed so that the Ar sputter gun within the XPS chamber can be used to deposit the


7


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

Page 8 of 45

target material. The holder features an 85° angle between the upright portion and the horizontal surface. The target material was attached to the upright end of the holder using conductive carbon tape, and a MoS2 sample was affixed to the horizontal surface of the sample holder 0.5 mm away from the target material (Fig. 1b). The holder assembly was then placed within a vacuum transfer stage, which was evacuated inside the antechamber of the glove box. It was then transferred directly into the XPS analysis chamber, which prevented exposure to atmosphere. This procedure is useful for minimizing further oxidation of the MoS2, and it is necessary for preventing oxidation/nitridation of the Li metal during experiments that utilize Li. Each of the target materials to be deposited were prepared in different ways. For Li experiments, a piece of Li foil (Sigma Aldrich, 99.9 % purity) was mechanically cleaned with Teflon inside the glove box to remove surface oxide before being attached to the L-shaped sample holder. For Ag experiments, a ~300 nm thick Ag film was thermally evaporated (Denton Explorer) onto a piece of Si wafer substrate using Ag source material (99.99 % purity). For Ge experiments, Czochralski-grown, undoped, Ge single crystals (MTI Corp.) were cleaned with deionized water to dissolve the surface oxide,50 then transferred to the glove box and attached to the sample holder. Once the sample holder was inside the XPS instrument, sputter deposition of the target material was carried out with Ar+ ions using an acceleration voltage of 3.0 keV and an ion current of 9.0 µA. The ion beam was rastered over a rectangular area of 4 mm × 2 mm. The angle between the horizontal sample and the sputter gun was 32˚, and the sputter gun was directed at a position ~2 mm above the surface of the MoS2 sample. This resulted in sputtering of the target material in a cone-shaped plume, with the plume oriented to partially deposit onto the MoS2 sample. Due to the nature of this sputtering process, the thickness of the sputtered thin


8

Page 9 of 45

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


films varied with position on the surface of the MoS2 samples. For repeatable measurements, it was necessary to always collect XPS spectra from the same position on the sample. This position was chosen to be that at which the maximum amount of material was deposited (Pmax). To find Pmax for each target material, a set of XPS scans was performed over a wide area after sputtering. The point at which the integrated area of a given XPS peak from the deposited material was maximized was assumed to correspond to the thickest section of the deposited film. For these calibration experiments, sputtering was carried out onto a Si wafer instead of a MoS2 sample to ensure minimal roughness. During collection of in situ data, high resolution photoelectron spectra were first obtained at the position Pmax on the bare MoS2 sample before any material was deposited. For Li deposition, a series of four sequential three-minute sputtering steps was performed, and a high resolution photoelectron spectrum was collected after each sputtering step. For Ag, four sequential one-minute sputtering steps were performed, and for Ge, three one-minute sputtering steps and one three-minute sputtering step were used; high resolution photoelectron spectra were similarly obtained. Deposition of ~1-3 nm of the target material was found to allow for in situ determination of the interfacial reactions at the MoS2 surface without fully attenuating the XPS signal from the MoS2 buried under the deposited material. Since each deposited material attenuates MoS2 photoelectrons to different degrees, the allowable deposition thickness is different for Li, Ge, and Ag. To determine the thickness of each material as a function of sputtering time, the three materials were sputtered in situ onto an MgO substrate, followed by acquisition of Mg core level spectra. MgO substrates were used because MgO has been found to be unreactive in contact with Li.49,51 The MgO substrates (MTI Corp.) were single crystals with a orientation, and they


9


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

Page 10 of 45

were plasma-cleaned (PE-25, Plasma Etch Inc.) for ten minutes prior to deposition. The thickness of the deposited material was found by using the attenuation of the integrated intensity of the Mg 1s peak from the MgO substrate in conjunction with the electron attenuation length (EAL) for the deposited material. The EAL values for each deposited material were obtained from the NIST standard database.48 A Shirley background was used for fitting XPS peaks in these experiments, and the peak shapes used were 70% Gaussian/30% Lorentzian for all Mo 3d, S 2p, Li 1s, and Ge 3d peaks, while a Donjiac-Sunjic lineshape was used for the asymmetric Ag 3d peak. For each high resolution spectrum, the best fit for each dataset was obtained as that with the minimum chisquare value. Further details regarding the peak-fitting procedure can be found in Section 2 of the SI.

RESULTS AND DISCUSSION To understand the impact of crystallographic orientation on the nature of interfacial transformations, MoS2 films with either exposed edge sites (“vertically aligned”) or exposed basal planes (“horizontally aligned”) were fabricated. As detailed in the methods section, horizontally aligned specimens were fabricated by exfoliating thin MoS2 sections from single crystals. Figure 2a shows a plan-view high resolution TEM (HRTEM) image and a selected area electron diffraction (SAED) pattern of a horizontally aligned sample; the electron beam is oriented through the basal planes, as shown in the schematic below. The SAED pattern shows that this is a single crystal with the hexagonal 2H MoS2 crystal structure.2 In contrast, vertically aligned MoS2 materials were synthesized via sulfidation of thin Mo metal films on Si substrates (see methods).28 Figure 2b shows a plan-view HRTEM image and SAED pattern of a vertically


10

Page 11 of 45

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


aligned MoS2 specimen. The image shows that the material is polycrystalline and features an abundance of widely spaced lattice fringes; these fringes correspond to the {002} planes with a spacing of 6.2 Å. These fringes are visible here because the MoS2 layers are oriented parallel to the viewing direction, as shown in the schematic below the image. The vertical alignment of the {002} lattice planes is further confirmed by the cross-sectional scanning TEM image of a similar sulfidized MoS2 sample in Fig. 2c. This image shows that the {002} planes are aligned perpendicular to the surface of the film; in addition, the surface of the MoS2 film is somewhat rough. High resolution XPS was performed to characterize the pristine MoS2 materials. Figure 3a and c show Mo 3d and S 2p spectra obtained from a horizontally aligned MoS2 sample prior to any deposition. The binding energy (BE) values for the most intense Mo 3d and S 2p doublets were found to be 229.1 ± 0.1 eV (Mo 3d5/2) and 161.9 ± 0.1 eV (S 2p3/2). These values correspond to MoS2 peaks reported in previous work (Mo 3d5/2 BE = 229.3 eV, S 2p3/2 BE = 162.1 eV).52 A small amount of non-stoichiometric MoxSy with intermediate oxidation states may also be present,52 as evidenced by the smaller doublets at lower BE in the Mo 3d and S 2p spectra at 228.2 ± 0.2 eV (Mo 3d5/2) and 162.1 ± 0.1 eV (S 2p3/2). This could be due to the presence of surface defects introduced during mechanical exfoliation. These defect sites have been shown to be a combination of S vacancies and metallic clusters of Mo;53 since there are fewer S atoms around the Mo atoms at such sites, the electronic environment is different than stoichiometric MoS2.52 For the as-synthesized vertically aligned films, strong Si 2p and O 1s peaks were observed, indicating the presence of SiOx (see Fig. S1 in the supporting information (SI)). Since the synthesized MoS2 films were conformal and without obvious pores or pinholes, the Si signal


11


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

Page 12 of 45

was presumed to arise from a thin interfacial film on the surface of the MoS2 instead of from the underlying Si substrate. This hypothesis was further confirmed by etching the vertically aligned MoS2 in HF, which removed the Si 2p and most of the O 1s signal from the sample (Fig. S1). Since the Mo and S core levels were still readily detectable even with the SiOx film present, the SiOx was likely

In Situ XPS Investigation of Transformations at Crystallographically

Recommend Documents