Periodic Modulation of the Doping Level in Striped MoS2

Feb 25, 2016 - Additionally, an obvious bandgap reduction is observed in the vicinity of the domain boundary for monolayer MoS2 on Au(100). This work ...
2 downloads 0 Views 7MB Size
Download PDF
Periodic Modulation of the Doping Level in Striped MoS2 Superstructures

Xiebo Zhou,†,‡,§ Jianping Shi,†,‡,§ Yue Qi,‡ Mengxi Liu,‡ Donglin Ma,‡ Yu Zhang,†,‡ Qingqing Ji,‡ Zhepeng Zhang,‡ Cong Li,† Zhongfan Liu,*,‡ and Yanfeng Zhang*,†,‡

ACS Nano 2016.10:3461-3468. Downloaded from pubs.acs.org by KAOHSIUNG MEDICAL UNIV on 08/15/18. For personal use only.

†

Department of Materials Science and Engineering, College of Engineering and ‡Center for Nanochemistry (CNC), Beijing National Laboratory for Molecular Sciences, College of Chemistry and Molecular Engineering, Peking University, Beijing 100871, People’s Republic of China S Supporting Information *

ABSTRACT: Although the recently discovered monolayer transition metal dichalcogenides exhibit novel electronic and optical properties, fundamental physical issues such as the quasiparticle bandgap tunability and the substrate effects remain undefined. Herein, we present the report of a quasi-one-dimensional periodically striped superstructure for monolayer MoS2 on Au(100). The formation of the unique striped superstructure is found to be mainly modulated by the symmetry difference between MoS2 and Au(100) and their lattice mismatch. More intriguingly, we find that the monolayer MoS2 is heavily n-doped on the Au(100) facet with a bandgap of 1.3 eV, and the Fermi level is upshifted by ∼0.10 eV on the ridge (∼0.2 eV below the conduction band) in contrast to the valley regions (∼0.3 eV below the conduction band) of the striped patterns after high-temperature sample annealing process. This tunable doping effect is considered to be caused by the different defect densities over the ridge/valley regions of the superstructure. Additionally, an obvious bandgap reduction is observed in the vicinity of the domain boundary for monolayer MoS2 on Au(100). This work should therefore inspire intensive explorations of adlayer−substrate interactions, the defects, and their effects on band-structure engineering of monolayer MoS2. KEYWORDS: molybdenum disulfide, scanning tunneling microscope/spectroscopy, atomic structure, doping level, Au(100)

T

realized to allow a wide range of applications for TMDCs in electronic and flexible devices.5 To date, the atomic structure and optical bandgap of monolayer MX2 have mainly been characterized by highresolution transmission electron microscopy (HR-TEM) and photoluminescence (PL).21,22 However, the necessary sampletransfer process for HRTEM measurements inevitably damages the atomic structure, especially for the edges and grain boundaries.21 Furthermore, the PL spectrum reveals only information about the optical exciton and Trion states rather than the quasiparticle bandgap.7 Hence, it is a pivotal issue to develop a facile and in situ method to characterize both the atomic structure and electronic properties of MX2. In this regard, scanning tunneling microscopy/spectroscopy (STM/ STS) should be a straightforward probe for such issues.23−29 Crommie et al. demonstrated that the quasiparticle bandgap of MoSe2 deposited by ultrahigh vacuum molecular beam epitaxy (UHV-MBE) was decreased by 0.86 ± 0.08 eV, with the layer thickness variable from monolayer to trilayer on the

wo-dimensional (2D) layered materials have revolutionized materials science since the discovery of graphene, as they hold great prospect for uncovering novel physical phenomena and for exploring versatile applications.1−4 Among the large family of 2D materials, semiconducting transition-metal dichalcogenides (STMDCs), especially MX2 (e.g., M = Mo, W; X = S, Se), have recently triggered great research activity directed toward exploring their unique layer-dependent electronic and optical properties,5−8 valley-polarized carrier abilities,9,10 strain-induced bandgap variations, and more exotic properties such as charge density wave and superconductivity.11,12 It is generally known that the quasiparticle band structure is a fundamental parameter for understanding the unique transport and optical phenomena of monolayer MX2,13 and intensive research has been performed on engineering the band structure. Recently, deep in-gap states have been reported, which were induced by the structural defects of MX2, such as vacancies, lattice distortions, and antisites, etc.14,15 Concurrently, the optical bandgap of MX2 has also been found to be tunable by the layer number and the strain effect.16−20 Moreover, from the viewpoint of theoretical calculations, large bandgap variations from a few eV (semiconducting) to 0 eV (metallic) could be © 2016 American Chemical Society

Received: November 30, 2015 Accepted: February 25, 2016 Published: February 25, 2016 3461

DOI: 10.1021/acsnano.5b07545 ACS Nano 2016, 10, 3461−3468

Article

www.acsnano.org

Article

ACS Nano

Figure 1. LPCVD synthesis of monolayer MoS2 on different facets of Au foils. (a) Low-magnification SEM image of MoS2 on Au(100) and Au(111) facets, respectively. The dotted line indicates the substrate grain boundary. (b) Corresponding EBSD map of the same region in (a) showing the facet contributions (as marked using the standard EBSD color key). (c, d) High-magnification SEM images of MoS2 triangles on Au(100) and Au(111), respectively. (e) Large-scale STM image (V = 1.20 V, IT = 0.20 nA, T = 78 K; 30.00 nm × 30.00 nm) of MoS2 on Au(100). The periods of the striped superstructure pattern of MoS2 on Au(100) and of the reconstructed Au(100) surface are highlighted by the green and red lines in (e), respectively. (f) Large-scale STM image (V = 1.00 V, IT = 0.20 nA, T = 78 K; 30.00 nm × 30.00 nm) of MoS2 on Au(111). The unit cell for the evolved moiré pattern is outlined by a rhombus (∼3.20 nm in period), and the herringbone reconstruction typical for the Au(111) surface is indicated by the black arrow. (g) Raman spectra of as-grown MoS2 on Au foils and after transfer onto a SiO2/Si substrate (peak positions are marked by dashed arrows). (h) TEM image on a folded flake edge indicating the monolayer feature of MoS2. Inset: Corresponding SAED pattern of the flake recorded from a range of 200.00 nm × 200.00 nm.

issues remain unclear, such as the interaction strength between MX2 and metals, the formation mechanism of the resultant superstructure, the defect state, as well as their effects on the electronic property of MX2.34 In this work, we report the STM/STS investigations of the CVD monolayer MoS2 grown on Au foils containing different facets of Au(100) and Au(111) and undergoing different annealing processes inside the UHV system. By this approach, we could probe the atomic-scale structures, the defect states, and the electronic properties of MoS2/Au(100). The unique MoS2/Au(100) system was selected with the following considerations: (1) a Au(100) substrate with 4-fold symmetry was utilized to produce different MoS2 superstructures (e.g., line-shaped superstructures), as compared with the hexagonalshaped superstructures obtained on Au(111) substrates with 6fold-symmetry; (2) the quasiparticle energy band of monolayer MoS2 was expected to be modulated by the periodically striped pattern; and (3) some novel physical and chemical phenomena would be revealed on the periodically rippled superstructures. Notably, such issues could be well addressed by using the straightforward STM/STS analysis method.

epitaxial graphene, as proven by STM/STS characterizations.23,24 This bandgap decrease was ascribed to both interlayer coupling and screening effect from the substrate. In addition, Xie et al. synthesized monolayer and bilayer MoSe2 on highly oriented pyrolytic graphite (HOPG) using the same synthetic method. They detected a novel one-dimensional midgap metallic mode in the monolayer MoSe2 as well as lineand point-like defects in bilayer MoSe2 based on STM/STS measurements.25,26 Moreover, Wee et al. fabricated monolayer MoS2 directly on graphite substrates by the chemical vapor deposition (CVD) method and presented an unexpected bandgap decrease (as large as 0.85 ± 0.05 eV, from 2.40 to 1.55 eV) at the patching boundary of two domains with a rotation angle of 18°.27 As for the fact that the band edge of MX2 is hard to precisely determine by constant Z tunneling spectroscopy due to the extremely short tunneling decay length of K-points states, repetitive measurements are usually applied to achieve and reconfirm the experimental data. Recently, Shih et al. have put forward a new measurement route of “variable Z spectroscopy”, for the accurate determination of the band edge of MX2 by probing the critical point energy locations.28 They reported that single-layer WSe2 presented an indirect quasi-particle gap with the conduction band minimum located at the Q-point (instead of K-point).28 They also detected the valence band and the conduction band offset of a type-I heterostructure junction constructed by bilayer−monolayer TMDCs.29 The metal substrate of Au(111) was also found to be a perfect candidate for the direct synthesis of MoS2.30−33 Lately, Lauritsen et al. successfully obtained submonolayer MoS2 on Au(111) using the same UHV−MBE method and observed a hexagonal superstructure pattern. Furthermore, they proved that the Mo and S orbital overlap to the Au surface could be modulated by the moiré pattern.34 Although electronic structures, such as the direct band gap and the in-gap states of MX2 ,have been studied widely on graphene, graphite, and even Au(111) substrates,23−34 some fundamental physical

RESULTS AND DISCUSSION A low-pressure CVD (LPCVD) system was selected for synthesizing MoS2 directly on Au foils on the basis of previously reported growth details.35−37 X-ray photoelectron spectroscopy (XPS) analysis was first performed to confirm the formation of MoS2 on Au foils (Supporting Information, Figure S1). Importantly, obvious shifts of the Mo 3d5/2 and S 2p peak positions to lower binding energies than those of pristine MoS2 were observed. This reflects the possible electron donation from the Au substrate to the MoS2 adlayer.36 To examine the microscopic morphology, domain size, and distribution of MoS2 on different Au facets, scanning electron microscopy (SEM) and electron backscatter diffraction (EBSD) measurements were simultaneously performed on the as-synthesized samples. The low-magnification SEM image and corresponding 3462

DOI: 10.1021/acsnano.5b07545 ACS Nano 2016, 10, 3461−3468

Article

ACS Nano

Figure 2. STM characterizations and structural simulations of the unique striped patterns for monolayer MoS2 on Au(100). (a) Large-scale STM image (1.20 V, 0.20 nA, 78 K; 50.00 nm × 50.00 nm) of the striped superstructure patterns with a typical periodicity of ∼8.00 ± 0.41 nm between adjacent stripes. Lower panel: the height profile along the green arrow. (b) Atomically resolved STM image (0.75 V, 0.20 nA, 78 K; 17.00 nm × 17.00 nm) of the striped MoS2 pattern, presenting the MoS2 atomic rows aligning well with the orientation of the striped pattern, as indicated by the blue arrows. (c) Corresponding 2D fast Fourier transform (2D-FFT) pattern from (b). (d) Corresponding atomically resolved STM image (0.81 V, 0.23 nA, 78 K; 6.00 nm × 6.00 nm). (e) Simulations of the striped patterns for monolayer MoS2 on nonreconstructed Au(100). The right two panels show the atomic models extracted from the ridges and valley regions of the striped pattern in (e). (f) Schematic diagram of slightly rippled monolayer MoS2 on Au(100).

0.315 nm) and Au(111) (a(Au(111)) = 0.288 nm), according to the published references for MoS2/Au(111).34 At this stage, it is important to characterize the thickness, crystallinity, and optical property of the obtained MoS2. To this end, complementary characterization tools such as optical microscopy (OM), Raman spectroscopy, PL, and TEM were employed, with the results shown in Figure 1g,h and the Supporting Information (Figure S2). For the as-grown MoS2 on Au foils, two typical Raman peaks (corresponding to the E1 2g and A1g modes) are observed at ∼385.50 and 404.20 cm−1, respectively, with a peak separation (Δ) of ∼18.70 cm−1 (Figure 1g). This value provides solid proof of monolayer MoS2 formation on the Au foils, according to the published reference.7 Intriguingly, a red shift (∼1.10 cm−1) of the A1g peak for MoS2/Au (∼404.20 cm−1) is observed with respect to that of MoS2 after transfer onto a SiO2/Si substrate (∼405.30 cm−1), and the intensity ratio of A1g/E1 2g is increased accordingly from 1:1 to 1.5:1, indicative of the possible electron donation from the Au substrate to the monolayer MoS2.39 Moreover, no obvious shift of the E1 2g peak for either MoS2/ Au or MoS2/SiO2/Si is detected. The HR-TEM image obtained by focusing on a folded film edge reveals a single black line contrast with a width of ∼0.52 nm (Figure 1h), which reconfirms the monolayer feature of the synthesized MoS2. In addition, the corresponding selected area electron diffraction (SAED) pattern recorded over the range of 200.00 nm × 200.00 nm typically presents only one set of hexagonally arranged diffraction spots, strongly suggestive of the high crystallinity of the obtained MoS2 (inset, Figure 1h). Thus, a high-quality MoS2 monolayer of uniform thickness was synthesized on both the Au(100) and Au(111) facets. In contrast to previous reports, however, a novel striped MoS2

EBSD map in Figure 1a,b reveal that MoS2 was successfully deposited on both the Au(100) and Au(111) facets. Highmagnification SEM images on such regions present uniformly distributed triangularly shaped contrasts (Figure 1c,d) with an average edge length of ∼400 nm. Interestingly, a majority of the MoS2 triangles on the Au(111) facet are well aligned, with their relative edge orientations expressed as multiples of 60° (Figure 1d). According to the published reference, this unique growth feature is possibly due to an epitaxial relationship between MoS2 and Au(111).34 However, for MoS2/Au(100), the MoS2 triangles tend to be aligned randomly with each other (Figure 1c), probably indicative of a relatively weak interface interaction. Further STM investigations of the two typical regions reveal much different surface morphologies. As displayed in the lower left corner of Figure 1e, a hexreconstruction of Au(100) is clearly visible with a typical striped feature, having a periodicity of ∼1.40 ± 0.20 nm between adjacent stripes, which agrees well with the published report.38 This hex-reconstruction stems from a contracted quasi-hexagonal top layer on the (1 × 1) square lattice.38 In contrast, striped superstructures evolve on the remainder of the surface, showing a relatively large periodicity of ∼8.00 ± 0.41 nm between adjacent stripes (marked by the green lines in Figure 1e). The region with the large periodic superstructure pattern is considered to be covered by monolayer MoS2 since no similar reconstructions have been reported for Au(100), and this pattern appears only after the growth of MoS2 . Furthermore, on the same Au foil surface, the hexagonally shaped moiré superstructure of MoS2 also occurs occasionally, manifesting a unique period of ∼3.20 ± 0.10 nm (marked by a rhombus) (Figure 1f). This moiré superstructure is mainly derived from the lattice mismatch between MoS2 (a(MoS2) = 3463

DOI: 10.1021/acsnano.5b07545 ACS Nano 2016, 10, 3461−3468

Article

ACS Nano

Figure 3. Energy band variations for monolayer MoS2/Au(100) around the domain edge and over the periodically striped patterns. (a, b) STM image (2.00 V, 0.13 nA, 78 K; 25.00 nm × 30.00 nm) of MoS2 on Au(100) after annealing at 370 °C for 2 h and corresponding STS spectra (f = 932 Hz, It = 0.13 nA, Vrms = 10 mV, T = 78 K) captured over the striped patterns with the specific locations marked with squares (along the black dashed arrow). (c, d) STM image (1.00 V; 0.20 nA 78 K; 25.00 nm × 30.00 nm) of MoS2 on Au(100) of the same sample after further annealing at 720 °C for 2h, and the corresponding STS spectra ( f = 932 Hz, IT = 0.20 nA, Vrms =10 mV, T = 78 K) captured over the striped patterns with the locations marked with squares (along the black dashed arrow). (e) STM image (1.00 V, 0.20 nA, 78 K; 50.00 nm × 60.00 nm) obtained from the edge of a monolayer MoS2 domain. (f, g) STS spectrum ( f = 932 Hz, IT = 0.20 pA, Vrms =10 mV, T = 78K) captured around the domain edge of the striped MoS2 pattern (from the same sample of (c)). The selected locations (along the dashed arrows) are indicated by stars and circles positioned over the valley and ridge regions, respectively, with an average interval of ∼8.00 nm.

superstructure pattern was first observed for MoS2/Au(100) by high-resolution STM. This unique superstructure is likely to modulate the electronic structure of monolayer MoS2, as similarly demonstrated in striped graphene ripples synthesized on metal substrates.40,41 To gain detailed insight into the formation mechanism of the striped MoS2 pattern on Au(100), atomically resolved STM characterizations and structural simulations were then performed. The three-dimensional image in Figure 2a clearly manifests the surface undulations as well as the striped patterns of MoS2/Au(100). According to the section-view analysis, a periodicity of ∼8.00 ± 0.41 nm between adjacent stripes, along with an apparent height of ∼2.50 Å for the striped pattern, was precisely derived. Additionally, the thickness of a MoS2 flake (∼0.70 nm) on Au(100) was also obtained by measuring its height difference with regard to the underlying Au(100) plane, as shown in the Supporting Information (Figure S3). This height difference again illustrates the monolayer feature of the obtained MoS2. Further magnification of the striped superstructure pattern reveals the hexagonal honeycomb lattice of MoS2 (Figure 2b). Incidentally, both protruding and holelike shapes exist in the lattice in Figure 2b, possibly due to the different distances between the tip and sample induced by the surface roughness of the Au foil substrate.42 From successive zoom-in STM images, the MoS2 atomic rows can be observed to be aligned well with the orientation of the striped superstructure pattern, as highlighted by the blue arrows in Figure 2b. This relatively uniform orientation is possibly generated during the mild annealing process prior to STM observations in the UHV system, since random orientations of

MoS2 on Au(100) were universally observed after the CVD growth process (Figure 1c). To elucidate this specific alignment, a 2D fast Fourier transform (2D-FFT) analysis of the image was also performed, as shown in Figure 2c. Note that the lattice constant of MoS2 (∼0.315 nm) is much smaller than the periodicity of the striped MoS2 pattern (∼8.00 nm) and the MoS2 lattice possesses different symmetry with regard to the striped superstructure pattern. The 2D-FFT spots can be reliable for distinguishing the two types of configurations, as indicated by the white and blue circles in Figure 2c. Moreover, with the 2D-FFT image, the relatively high crystal quality for monolayer MoS2 on Au(100) is further confirmed by the single set of hexagonally arranged 2D-FFT spots. From the atomically resolved STM images, perfect lattices with an interatomic distance of ∼0.320 nm (Figure 2d) are observed, which is fully consistent with the lattice parameter of a MoS2(0001) plane,34 reconfirming the formation of striped MoS2 on Au(100) and its relatively high crystal quality. The origin of this unique striped MoS2 superstructure is of great interest. First, MoS2 is expected to extend over the nonreconstructed Au(100) facet, although the neighboring Au(100) surface is hex-reconstructed. Thus, the monolayer MoS2 lifts the surface reconstruction of Au(100). This is possibly due to the higher binding energy between MoS2 and the unreconstructed Au(100) facet, as compared with that of MoS2 and the hex-reconstructed Au(001) facet. Interestingly, this phenomenon was also reported for the assembly of molecules containing sulfur (α-sexithiophene and S2) on Au(100).43,44 In this case, the striped superstructure adopts a moiré pattern caused by the symmetry difference as well as the 3464

DOI: 10.1021/acsnano.5b07545 ACS Nano 2016, 10, 3461−3468

Article

ACS Nano

The influence of the monolayer MoS2 domain edge/ boundary on the device performance is a very fundamental issue. Several TEM characterizations of monolayer MoS2 triangles revealed either a Mo-terminated edge with a straight shape or an S-terminated edge with a crooked shape.22 Further, theoretical calculations have predicted that both Mo- and Sterminated edges may introduce metallic states inside the bulk bandgap of monolayer MoS2.15 However, more straightforward experimental observation of the edge state of monolayer MoS2 has rarely been reported.46 A large-scale STM image obtained from the edge of a monolayer MoS2 domain on Au(100) at 720 °C is shown in Figure 3e. The spatial dependence of the electronic structures is then measured from the middle to the edge of a domain and on the ridge and the valley regions, respectively (Figure 3f,g). The STS spectra from the valleys are displayed in Figure 3f, where each spectrum is labeled with a letter according to the marked spatial position in Figure 3e. Clearly, for the STS spectra from A−D in the valleys of the striped pattern, the band profile remains relatively flat, with the valence band maximum (VBM) and conduction band minimum (CBM) located at ∼−1.00 and ∼+0.30 V, respectively, corresponding to a quasiparticle bandgap (Eg) of ∼1.30 eV (Figure 3f). The reduced Eg and n-doping phenomena of MoS2/Au(100) as compared with that of the theoretical case (with Eg ∼ 1.80 eV)7 are considered to be mainly mediated by the electron donation from the Au substrate and the strong adlayer−substrate interaction.34 Intriguingly, for the STS spectrum obtained from the edge of the MoS2 flake, as indicated by E (Figure 3e), the VBM of MoS2 shifts upward to ∼−0.50 V, and the CBM shifts downward to ∼+0.10 V, corresponding to an Eg of ∼0.60 eV. Notably, this electronic feature seems in sharp contrast with that of the bare Au(100) substrate, as shown in Figure 3f (indicated by F, G). This means that at the domain edge of monolayer MoS2/Au(100) the band gap is greatly decreased by ∼0.70 eV. This tendency of the band-gap variations was also observed along the ridge regions (Figure 3g). For the measured positions far from the edge of the monolayer MoS2 domain (as indicated by A′−D′ in Figure 3e), the VBM and CBM are located at ∼−1.10 and ∼+0.20 V, respectively, corresponding to an Eg of ∼1.30 eV. However, at the edge (indicated by E′ in Figure 3e), the VBM and CBM of monolayer MoS2 are located at ∼−0.50 and ∼+0.10 V, respectively, corresponding to an Eg of ∼0.60 eV. In general, the Fermi energy of the ridge region is upshifted by ∼0.10 eV with respect to that of the valley region (the same as that from Figure 3d). Along the ridge regions, the band gap was similarly modulated by about ∼0.70 eV at the domain boundary. The distinct reduction of Eg at the edges of the MoS2 domains is possibly due to the appearance of metallic edge states, as demonstrated in the MoS2/graphite system.46 According to previous aberration-corrected scanning transmission electron microscopy (STEM) characterizations, the macroscale straight edge of MoS2 domain was not sharp at the nanometer scale and contained many atomic steps. The step edges provided additional Mo and S sites with unsaturated bonds, which possibly contributed some metallic states.15 In short, an obvious band gap reduction is observed in the vicinity of the domain edge for monolayer MoS2 on Au(100), and a periodic modulation of the doping level was detected on the periodically striped superstructures possessing valley- and ridge-type regions after annealing at 720 °C. Such results are

lattice mismatch between MoS2 and the nonreconstructed Au(100) lattice. To address this issue, simulated patterns generated from monolayer MoS2 stacking on nonreconstructed Au(100) with a relative rotation angle of 0° are depicted in Figure 2e. A periodically striped modulation of ∼5.76 nm between adjacent stripes results for monolayer MoS2 on Au(100). Obviously, the striped configuration of the simulated pattern is in good agreement with that of the experimental data, although the period seems slightly smaller. For more details, atomic models obtained from the ridge and valley regions of the striped pattern are also displayed in the right panels of Figure 2e. Here, the different placement of MoS2 with regard to the nonreconstructed Au(100) surface is clearly observed, with S atoms positioned nearly on top of the Au atoms and Mo on the Au hollow sites for the ridge regions (upper right panel). However, both Mo and S are placed nearly on the Au hollow sites for the valley regions (lower right panel). This phenomenon possibly induces either weak or strong coupling between the adlayer and substrate.34,45 Finally, the schematic diagram for monolayer MoS2 on Au(100) is also displayed in Figure 2f to show the periodic undulation of the superstructure pattern. Notably, to explain the deviation between the experimentally observed periodicity (∼8.00 nm) in the striped pattern and the simulated data (∼5.76 nm), the strain effect derived from the different thermal expansion coefficients between the adlayer and substrate should be considered. Furthermore, other periods (∼6.80 to ∼7.20 nm) were occasionally observed, as shown in the Supporting Information (Figure S4). Another two simulations of the origins of the different moiré periods are also displayed, where monolayer MoS2 samples are stretched by ∼1.3% with a(MoS2) = 0.319 nm and by ∼1.6% with a(MoS2) = 0.320 nm on Au(100), in line with two type periods of ∼7.20 and ∼8.00 nm, respectively. To explore the effect of the striped pattern on the energy band of monolayer MoS2/Au, comparative STM/STS characterizations were performed with the samples undergoing different annealing processes (Figure 3a−d). The large-scale STM image in Figure 3a reveals the morphology of the striped MoS2 pattern on Au(100) after thermal annealing at ∼370 °C for 2 h. The positions for the STS measurements are marked by the letters a−f along the vertical direction of three bright lineshaped stripes. As shown in Figure 3b, the valence band maximum (VBM) and conduction band minimum (CBM) of monolayer MoS2 are located at ∼−1.20 V and ∼+0.30 V for ridge and valley positions, respectively, corresponding to a quasiparticle bandgap (Eg) of ∼1.50 eV. Intriguingly, after hightemperature annealing at ∼720 °C for 2 h (Figure 3c), the VBM and CBM of MoS2 from the ridge positions are located at ∼−1.10 V and ∼+0.20 V, respectively. However, the VBM and CBM from the valley positions appear at ∼−1.00 V and ∼+0.30 V, respectively (Figure 3d). In this regard, the high-temperature-annealed MoS2/Au(100) presents a tunable doping level (0.1 eV shift) regarding the ridge and valley regions. This differs from that of the low-temperature annealed sample with a fixed doping level. The overall different n-doping effects for the ridge/valley regions after high temperature annealing is proposed to be induced by their different defects densities, as will be discussed in the next part. Moreover, a minor reduction of the band gap (from 1.50 to 1.30 eV) with increasing annealing temperature can also be obtained from the STS data, which is probably mediated by an increased interface interaction between CVD MoS2 and Au foils. 3465

DOI: 10.1021/acsnano.5b07545 ACS Nano 2016, 10, 3461−3468

Article

ACS Nano

Figure 4. Comparison of defect densities on the ridges and valley regions of striped MoS2 superstructures on Au(100) after a mild annealing process. (a) Atomically resolved STM images (0.75 V, 0.20 nA, 78 K; 20.00 nm × 20.00 nm) of monolayer MoS2 on Au(100) after annealing at ∼370 °C for 2 h. (b) Atomically resolved STM image (0.75 V, 0.20 nA, 78 K; 20.00 nm × 20.00 nm) of MoS2 on Au(100) after annealing at ∼720 °C for 2 h. (c, d) Zoom-in STM image of a single S vacancies defect. (e) Experimental statistics of the defect densities on the ridges and valley regions of striped MoS2 patterns from different annealing temperatures. (f) Raman spectra of monolayer MoS2 triangles on Au foils before and after annealing in the UHV system at ∼720 °C for 2 h.

DV = 0.030 ± 0.009/nm2 for 720 °C, and DR = 0.100 ± 0.02/ nm2 and DV = 0.034 ± 0.007/nm2 for 820 °C annealing, respectively (also see Figure S6 for more STM/STS data of 500 and 820 °C annealed samples). These results strongly suggest that defects (vacancies of S) are preferentially evolved on ridges rather than on valley regions of the striped MoS2 patterns on Au(100), especially after high-temperature annealing processes. Additionally, a more obvious increase of the defect density can be noticed on the ridge regions compared with that of the valley regions with increasing annealing temperature. This probably indicates much weaker interface interaction of the ridge/substrate than that of the valley/substrate, in parallel with unstable and stable surfaces. However, an in-depth theoretical analysis of this interesting system should be desirable to ascertain the specific stacking order as well as the disparate binding feature for the different regions of striped MoS2 on Au(100). Moreover, the Raman spectra of MoS2/Au foils before and after annealing (∼720 °C, ∼2 h) in the UHV system were also acquired to examine the effects of the annealing processes (Figure 4f). An obvious red shift (∼1.20 cm−1) of the A1g peak for the annealed sample (∼403.00 cm−1) is observable with respect to that of the as-grown sample (∼404.20 cm−1), and the intensity ratio of A1g/E1 2g is decreased from 1:1 to 1:1.5 throughout the annealing process. This reflects an enhanced n-doping effect to the monolayer MoS2/Au(100) after the thermal annealing process under UHV conditions.39,49

essential for understanding the electronic structures of MX2 materials and advancing their further applications in electronics and photonics. In order to achieve an in-depth understanding of the origin of the different n-doping effects from the ridge and valley regions on monolayer MoS2, the CVD monolayer MoS2/Au was thermally annealed in the UHV system. Samples with triangular MoS2 domains (with an average edge length of ∼400 nm) were selected as candidates. After even high-temperature annealing at ∼720 °C under UHV for 2 h, irregular MoS2 flakes with edge sizes of ∼100 nm were obtained, as presented in the Supporting Information (Figure S5). This domain size reduction is probably mediated by the surface reconstruction of the Au(100) facet, as well as the substrate-facet formation effect during the annealing process.47 Atomically resolved STM images of MoS2/Au(100) after thermal annealing at different temperatures are presented in Figure 4a−d. For MoS2/Au(100) annealed at ∼370°C for 2 h, the STM image reveals honeycomb lattices with an interatomic distance of ∼0.32 nm, suggesting the preservation of the MoS2 lattice (Figure 4a). A few MoS2 defects are also visible as indicated by the blue arrows. However, after even high temperature annealing (∼720 °C, ∼2 h), relative high density defects are generated on both ridge and valley regions of the striped MoS2 patterns (Figure 4b). Moreover, a majority of defects can be identified as using S vacancies (1S) (Figure 4c, d). It is worth mentioning that this 1S vacancy defect has been reported to be capable of inducing an n-doping effect to monolayer MoS2.48 In this regard, the current annealing process for as-grown samples is useful for tuning the doping level of single-layer MoS2. Figure 4e shows the statistics of the defects densities of MoS2/Au(100) after 370, 500, 720, and 820 °C annealing processes, which are DR = 0.024 ± 0.005/nm2 (defect density for the ridge) and DV = 0.022 ± 0.006/nm2 (defect density for the valley) for 370 °C; DR = 0.041 ± 0.010/nm2 and DV = 0.027 ± 0.005/nm2 for 500 °C; DR = 0.055 ± 0.006/nm2 and

CONCLUSION In summary, we observed novel periodically striped superstructure patterns of monolayer MoS2 on the Au(100) facet of Au foils, most of which exhibits a large periodicity of ∼8.00 nm. By virtue of low-temperature STM/STS characterizations, we demonstrated that the doping level of monolayer MoS2 on Au(100) can be periodically modulated by the striped superstructure pattern after a high temperature annealing process, with a Fermi level upshift of ∼0.1 eV on the ridge compared with that on the valley of the striped pattern. The 3466

DOI: 10.1021/acsnano.5b07545 ACS Nano 2016, 10, 3461−3468

Article

ACS Nano

Municipal Science and Technology Planning Project (No. Z151100003315013).

different n-doping effect is considered to be modulated by the relative high/low defect densities on the ridge/valley regions, in line with weak/strong interface interactions for MoS2/Au(100). Moreover, an obvious bandgap reduction is also observed in the vicinity of the domain boundary for monolayer MoS2 on Au(100). Briefly, this work presents profound insights toward understanding the electronic structure of MX2 and the substrate effect and should inspire intensive investigations on the novel physical and chemical properties of these unique long-change superstructure patterns.

REFERENCES (1) Zhu, F.-F.; Chen, W.-J.; Xu, Y.; Gao, C.-L.; Guan, D.-D.; Liu, C.H.; Qian, D.; Zhang, S.-C.; Jia, J.-F. Epitaxial Growth of Twodimensional Stanene. Nat. Mater. 2015, 14, 1020−1025. (2) Song, C.-L.; Wang, Y.-L.; Cheng, P.; Jiang, Y.-P.; Li, W.; Zhang, T.; Li, Z.; He, K.; Wang, L.; Jia, J.-F.; Hung, H.-H.; Wu, C.; Ma, X.; Chen, X.; Xue, Qi-K. Direct Observation of Nodes and Twofold Symmetry in FeSe Superconductor. Science 2011, 332, 1410−1413. (3) Mak, K. F.; Lee, C.; Hone, J.; Shan, J.; Heinz, T. F. Atomically thin MoS2: A New Direct-Gap Semiconductor. Phys. Rev. Lett. 2010, 105, 136805. (4) Li, L.; Yu, Y.; Ye, G. J.; Ge, Q.; Ou, X.; Wu, H.; Feng, D.; Chen, X. H.; Zhang, Y. Black Phosphorus Field-Effect Transistors. Nat. Nanotechnol. 2014, 9, 372−377. (5) Wang, Q. H.; Kalantar-Zadeh, K.; Kis, A.; Coleman, J. N.; Strano, M. S. Electronics and Optoelectronics of Two Dimensional Transition Metal Dichalcogenides. Nat. Nanotechnol. 2012, 7, 699−712. (6) Chhowalla, M.; Shin, H. S.; Eda, G.; Li, L.-J.; Loh, K. P.; Zhang, H. The Chemistry of Two-Dimensional Layered Transition Metal Dichalcogenide Nanosheets. Nat. Chem. 2013, 5, 263−275. (7) Splendiani, A.; Sun, L.; Zhang, Y.; Li, T.; Kim, J.; Chim, C.-Y.; Galli, G.; Wang, F. Emerging Photoluminescence in Monolayer MoS2. Nano Lett. 2010, 10, 1271−1275. (8) Shi, J.; Ma, D.; Han, G.-F.; Zhang, Y.; Ji, Q.; Gao, T.; Sun, J.; Song, X.; Li, C.; Zhang, Y.; Lang, X.-Y.; Zhang, Y.; Liu, Z. Controllable Growth and Transfer of Monolayer MoS2 on Au Foils and Its Potential Application in Hydrogen Evolution Reaction. ACS Nano 2014, 8, 10196−10204. (9) Xiao, D.; Liu, G.-B.; Feng, W.; Xu, X.; Yao, W. Coupled Spin and Valley Physics in Monolayers of MoS2 and Other Group-VI Dichalcogenides. Phys. Rev. Lett. 2012, 108, 196802. (10) Mak, K. F.; He, K.; Shan, J.; Heinz, T. F. Control of Valley Polarization in Monolayer MoS2 by Optical Helicity. Nat. Nanotechnol. 2012, 7, 494−498. (11) Jo, S.; Costanzo, D.; Berger, H.; Morpurgo, A. F. Electrostatically Induced Superconductivity at the Surface of WS2. Nano Lett. 2015, 15, 1197−1202. (12) Britnell, L.; Ribeiro, R.; Eckmann, A.; Jalil, R.; Belle, B.; Mishchenko, A.; Kim, Y.-J.; Gorbachev, R.; Georgiou, T.; Morozov, S.; Grigorenko, A.; Geim, A. K.; Casiraghi, C.; Neto, A.; Novoselov, K. Strong Light-Matter Interactions in Heterostructures of Atomically Thin Films. Science 2013, 340, 1311−1314. (13) Li, H.; Zhang, Q.; Duan, X.; Wu, X.; Fan, X.; Zhu, X.; Zhuang, X.; Hu, W.; Zhou, H.; Pan, A.; Duan, X. Lateral Growth of Composition Graded Atomic Layer MoS2(1−x)Se2x Nanosheets. J. Am. Chem. Soc. 2015, 137, 5284−5287. (14) Zou, X.; Liu, Y.; Yakobson, B. I. Predicting Dislocations and Grain Boundaries in Two-Dimensional Metal-Disulfides from the First Principles. Nano Lett. 2013, 13, 253−258. (15) Zhou, W.; Zou, X.; Najmaei, S.; Liu, Z.; Shi, Y.; Kong, J.; Lou, J.; Ajayan, P. M.; Yakobson, B. I.; Idrobo, J.-C. Intrinsic structural defects in monolayer molybdenum disulfide. Nano Lett. 2013, 13, 2615−2622. (16) Jin, W.; Yeh, P.-C.; Zaki, N.; Zhang, D.; Sadowski, J. T.; AlMahboob, A.; van Der Zande, A. M.; Chenet, D. A.; Dadap, J. I.; Herman, I. P.; Sutter, P.; Hone, J.; Osgood, R. M. Direct Measurement of the Thickness-Dependent Electronic Band Structure of MoS2 Using Angle-Resolved Photoemission Spectroscopy. Phys. Rev. Lett. 2013, 111, 106801. (17) Zhang, Y.; Chang, T.-R.; Zhou, B.; Cui, Y.-T.; Yan, H.; Liu, Z.; Schmitt, F.; Lee, J.; Moore, R.; Chen, Y.; Lin, H.; Jeng, H.-T.; Mo, S.K.; Hussain, Z.; Bansil, A.; Shen, Z. Direct Observation of the Transition from Indirect to Direct Bandgap in Atomically Thin Epitaxial MoSe2. Nat. Nanotechnol. 2013, 9, 111−115. (18) Feng, J.; Qian, X.; Huang, C.-W; Li, J. Strain-Engineered Artificial Atom as a Broad-Spectrum Solar Energy Funnel. Nat. Photonics 2012, 6, 866−872.

METHODS Growth Procedures. Monolayer MoS2 was grown with MoO3 (Alfa Aesar, purity 99.9%) and S (Alfa Aesar, purity 99.5%) serving as precursors by using a low pressure CVD method. All of the sample growth was finished in a three-zone furnace (Lindberg/Blue M HTF55347c) equipped with a 1 in. diameter quartz tube. The typical temperatures of Au, MoO3, and S powder were 530, 530, and 102 °C, respectively. Ar (50 sccm) and H2 (5 sccm) were used as the carrier gases. After growth, the MoS2/Au samples were detached from Au foils and transferred onto target substrates using the typical PMMAmediated transfer-printing technique for TEM, PL, and Raman measurements. Characterizations. The samples were characterized by OM (Olympus BX51), SEM (Hitachi S-4800, 2 kV), XPS (Kratos Analytical AXIS-Ultra with monochromatic Al Kα X-ray), Raman and PL (Renishaw, Invia Reflex, excitation light of 514 nm in wavelength), and TEM (FEI Tecnai G2 F20, acceleration voltage 200 kV). EBSD was collected using a JEOLJSM-6500F analytical SEM with the Oxford Technology EBSD system. During EBSD collection, the probe current was 5 nA, the accelerating voltage was 20 kV, and the angle of incidence was 70°. UHV LT-STM/STS systems were utilized under a base pressure better than 10−10 mbar. The STS spectra were measured at ∼78 K by recording the output of a lock-in system with the manually disabled feedback loop. A modulation signal of 10 mV at 932 Hz was selected under a tunneling condition of 1.20 V, 0.20 nA. Prior to STM observations, the CVD MoS2/Au foils samples were loaded into the UHV systems and degassed at about 600 K for several hours with the base pressure greater than 10−9 mbar.

ASSOCIATED CONTENT S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsnano.5b07545. Additional XPS, PL, TEM, and STM/STS of as-grown and transferred MoS2 (PDF)

AUTHOR INFORMATION Corresponding Authors

*E-mail: zfl[email protected]. *E-mail: [email protected]. Author Contributions §

X.Z. and J.S. contributed equally to this work.

Notes

The authors declare no competing financial interest.

ACKNOWLEDGMENTS This work was financially supported by the National Natural Science Foundation of China (Grants Nos. 51290272, 51472008, 51222201, 21201012, 51121091, 51072004, and 51201069), the Ministry of Science and Technology of China (Grants Nos. 2012CB921404, 2011CB921903, 2012CB933404, and 2013CB932603), and the Beijing 3467

DOI: 10.1021/acsnano.5b07545 ACS Nano 2016, 10, 3461−3468

Article

ACS Nano (19) Pan, H.; Zhang, Y.-W. Edge-Dependent Structural, Electronic and Magnetic Properties of MoS2 Nanoribbons. J. Mater. Chem. 2012, 22, 7280. (20) Lin, Y.-C.; Dumcenco, D. O.; Huang, Y.-S.; Suenaga, K. Atomic Mechanism of the Semiconducting-to-Metallic Phase Transition in Single-Layered MoS2. Nat. Nanotechnol. 2014, 9, 391−396. (21) Najmaei, S.; Liu, Z.; Zhou, W.; Zou, X.; Shi, G.; Lei, S.; Yakobson, B. I.; Idrobo, J.-C.; Ajayan, P. M.; Lou, L. Vapour Phase Growth and Grain Boundary Structure of Molybdenum Disulphide Atomic Layers. Nat. Mater. 2013, 12, 754−759. (22) van der Zande, A. M.; Huang, P. Y.; Chenet, D. A.; Berkelbach, T. C.; You, Y.; Lee, G.-H.; Heinz, T. F.; Reichman, D. R.; Muller, D. A.; Hone, J. C. Grains and Grain Boundaries in Highly Crystalline Monolayer Molybdenum Disulphide. Nat. Mater. 2013, 12, 554−561. (23) Ugeda, M. M.; Bradley, A. J.; Shi, S.-F.; Felipe, H.; Zhang, Y.; Qiu, D. Y.; Ruan, W.; Mo, S.-K.; Hussain, Z.; Shen, Z.-X.; Wang, F.; Louie, S.; Crommie, M. F. Giant Bandgap Renormalization and Excitonic Effects in a Monolayer Transition Metal Dichalcogenide Semiconductor. Nat. Mater. 2014, 13, 1091−1095. (24) Bradley, A. J.; M. Ugeda, M.; da Jornada, F. H.; Qiu, D. Y.; Ruan, W.; Zhang, Y.; Wickenburg, S.; Riss, A.; Lu, J.; Mo, S.-K.; Hussain, Z.; Shen, Z.; Louie, S.; Crommie, M. F. Probing the Role of Interlayer Coupling and Coulomb Interactions on Electronic Structure in Few-Layer MoSe2 Nanostructures. Nano Lett. 2015, 15, 2594−2599. (25) Liu, H.; Jiao, L.; Yang, F.; Cai, Y.; Wu, X.; Ho, W.; Gao, C.; Jia, J.; Wang, N.; Fan, H.; Yao, W.; Xie, M. Dense Network of OneDimensional Midgap Metallic Modes in Monolayer MoSe2 and Their Spatial Undulations. Phys. Rev. Lett. 2014, 113, 066105. (26) Liu, H.; Zheng, H.; Yang, F.; Jiao, L.; Chen, J.; Ho, W.; Gao, C.; Jia, J.; Xie, M. Line and Point Defects in MoSe2 Bilayer Studied by Scanning Tunneling Microscopy and Spectroscopy. ACS Nano 2015, 9, 6619−6625. (27) Huang, Y. L.; Chen, Y.; Zhang, W.; Quek, S. Y.; Chen, C.-H.; Li, L.-J.; Hsu, W.-T.; Chang, W.-H.; Zheng, Y. J.; Chen, W.; Wee, A. Bandgap Tunability at single-layer Molybdenum Disulphide Grain Boundaries. Nat. Commun. 2015, 6, 6298. (28) Zhang, C.; Chen, Y.; Johnson, A.; Li, M.-Y.; Li, L.-J.; Mende, P. C.; Feenstra, R. M.; Shih, C.-K. Probing Critical Point Energies of Transition Metal Dichalcogenides: Surprising Indirect Gap of Single Layer WSe2. Nano Lett. 2015, 15, 6494−6500. (29) Zhang, C.; Chen, Y.; Huang, J.-K.; Wu, X.; Li, L.-J.; Yao, W.; Tersoff, J.; Shih, C.-K. Visualizing Band Offsets and Edge States in Bilayer-Monolayer Transition Metal dDichalcogenides Lateral Heterojunction. Nat. Commun. 2016, 7, 10349. (30) Jaramillo, T. F.; Jorgensen, K. P.; Bonde, J.; Nielsen, J. H.; Horch, S.; Chorkendorff, I. Identification of Active Edge Sites for Electrochemical H2 Evolution from MoS2 Nanocatalysts. Science 2007, 317, 100−102. (31) Nielsen, J. H.; Bech, L.; Nielsen, K.; Tison, Y.; Jørgensen, K. P.; Bonde, J. L.; Horch, S.; Jaramillo, T. F.; Chorkendorff, I. Combined Spectroscopy and Microscopy of Supported MoS2 Nanoparticles. Surf. Sci. 2009, 603, 1182−1189. (32) Helveg, S.; Lauristen, J. V.; Lægsgaard, E.; Stensgaard, I.; Nørskov, J. K.; Clausen, B. S.; Topsøe, H.; Besenbacher, F. AtomicScale Structure of Single-Layer MoS2 Nanoclusters. Phys. Rev. Lett. 2000, 84, 951. (33) Bollinger, M. V.; Lauristen, J. V.; Jacobsen, K. W.; N, J. K.; Helveg, S.; Besenbacher, F. One-Dimensional Metallic Edge States in MoS2. Phys. Rev. Lett. 2001, 87, 196803. (34) Sørensen, S. G.; Füchtbauer, H. G.; Tuxen, A. K.; Walton, A. S.; Lauritsen, J. V. Structure and Electronic Properties of In Situ Synthesized Single-Layer MoS2 on a Gold Surface. ACS Nano 2014, 8, 6788−6796. (35) Ling, X.; Lee, Y.-H.; Lin, Y.; Fang, W.; Yu, L.; Dresselhaus, M. S.; Kong, J. Role of the Seeding Promoter in MoS2 Growth by Chemical Vapor Deposition. Nano Lett. 2014, 14, 464−472. (36) Shi, J.; Yang, Y.; Zhang, Y.; Ma, D.; Wei, W.; Ji, Q.; Zhang, Y.; Song, X.; Gao, T.; Li, C.; Bao, X.; Liu, Z.; Fu, Q.; Zhang, Y. Monolayer

MoS2 Growth on Au Foils and On-Site Domain Boundary Imaging. Adv. Funct. Mater. 2015, 25, 842−849. (37) Shi, J.; Zhang, X.; Ma, D.; Zhu, J.; Zhang, Y.; Guo, Z.; Yao, Y.; Ji, Q.; Song, X.; Zhang, Y.; Li, C.; Liu, Z.; Zhu, W.; Zhang, Y. Substrate Facet Effect on the Growth of Monolayer MoS2 on Au Foils. ACS Nano 2015, 9, 4017−4025. (38) Hammer, R.; Sander, A.; Förster, S.; Kiel, M.; Meinel, K.; Widdra, W. Surface Reconstruction of Au(001): High-Resolution Real-Space and Reciprocal-Space Inspection. Phys. Rev. B: Condens. Matter Mater. Phys. 2014, 90, 035446. (39) Shi, Y.; Huang, J.-K.; Jin, L.; Hsu, Y.-T.; Yu, S. F.; Li, L.-J.; Yang, H. Selective Decoration of Au Nanoparticles on Monolayer MoS2 Single Crystal. Sci. Rep. 2013, 3, 1839. (40) Park, C. H.; Yang, L.; Son, Y. W.; Cohen, M. L.; Louie, S. G. Anisotropic Behaviours of Massless Dirac Fermions in Graphene under Periodic Potentials. Nat. Phys. 2008, 4, 213−217. (41) Park, C.-H.; Yang, L.; Son, Y.-W.; Cohen, M. L.; Louie, S. G. New Generation of Massless Dirac Fermions in Graphene under External Periodic Potentials. Phys. Rev. Lett. 2008, 101, 126804. (42) Perrot, E.; Humbert, A.; Piednoir, A.; Chapon, C.; Henry, C. R. STM and TEM studies of A Model Catalyst: Pd/MoS2(0001). Surf. Sci. 2000, 445, 407−419. (43) Jiang, Y.; Liang, X.; Ren, S.; Chen, C.-L.; Fan, L.-J.; Yang, Y.-W.; Tang, J.-M.; Luh, D.-A. The Growth of Sulfur Adlayers on Au(100). J. Chem. Phys. 2015, 142, 064708. (44) Höfer, A.; Duncker, K.; Kiel, M.; Wedekind, S.; Widdra, W. Adsorption of α-sexithiophene on Au(001): Molecule-Induced Partial Lifting of the Substrate Reconstruction. Phys. Rev. B: Condens. Matter Mater. Phys. 2011, 83, 075414. (45) Popov, I.; Seifert, G.; Tománek, D. Designing Electrical Contacts to MoS2 Monolayers: A Computational Study. Phys. Rev. Lett. 2012, 108, 156802. (46) Zhang, C.; Johnson, A.; Hsu, C.-L.; Li, L.-J.; Shih, C.-K. Direct Imaging of Band Profile in Single Layer MoS2 on Graphite: Quasiparticle Energy Gap, Metallic Edge States, and Edge Band Bending. Nano Lett. 2014, 14, 2443−2447. (47) Eda, G.; Yamaguchi, H.; Voiry, D.; Fujita, T.; Chen, M.; Chhowalla, M. Photoluminescence from Chemically Exfoliated MoS2. Nano Lett. 2011, 11, 5111−5116. (48) Komsa, H.-P.; Kotakoski, J.; Kurasch, S.; Lehtinen, O.; Kaiser, U.; Krasheninnikov, A. V. Two-Dimensional Transition Metal Dichalcogenides under Electron Irradiation: Defect Production and Doping. Phys. Rev. Lett. 2012, 109, 035503. (49) Chakraborty, B.; Bera, A.; Muthu, D.; Bhowmick, S.; Waghmare, U. V.; Sood, A. Symmetry−Dependent Phonon Renormalization in Monolayer MoS2 Transistor. Phys. Rev. B: Condens. Matter Mater. Phys. 2012, 85, 161403(R).

3468

DOI: 10.1021/acsnano.5b07545 ACS Nano 2016, 10, 3461−3468