Research Article www.acsami.org
Enhanced Light Emission from Monolayer Semiconductors by Forming Heterostructures with ZnO Thin Films Min Su Kim,†,‡,# Shrawan Roy,†,§,# Jubok Lee,†,§ Byung Gu Kim,∥ Hyun Kim,†,§ Ji-Hoon Park,†,‡ Seok Joon Yun,†,§ Gang Hee Han,†,‡ Jae-Young Leem,∥ and Jeongyong Kim*,†,§ †
Center for Integrated Nanostructure Physics (CINAP), Institute for Basic Science (IBS), Suwon 16419, Republic of Korea Sungkyunkwan University (SKKU), Suwon 16419, Republic of Korea § Department of Energy Science, Sungkyunkwan University (SKKU), Suwon 16419, Republic of Korea ∥ Department of Nanoscience & Engineering, Inje University, Gimhae 621-749, Republic of Korea ‡
S Supporting Information *
ABSTRACT: Monolayer transition-metal dichalcogenides (1L-TMDs) are atomically thin direct band gap semiconductors, from which the emission of light is determined by optical transitions of exciton complexes such as neutral excitons and trions. While the quantum yields of 1L-TMDs are quite low, the ability to control the populations of exciton complexes in 1L-TMDs through various doping processes is an interesting advantage, and provides ample possibilities for engineering the optical properties of these semiconductor monolayers. Here we demonstrate a simple method of controlling the populations of excitons and trions to enhance the light emission of 1L-TMDs by having them form heterostructures with ZnO thin films (TFs). 1Ls of MoS2 or MoSe2 showed up to 17-fold increases in photoluminescence (PL) when they were placed on ∼50 nm thick ZnO TFs. This enhancement of the PL was due to charge exchanges occurring through the 1L-TMD/ZnO interface. The PL enhancements and changes in the PL spectra of the 1L-TMDs were greater when the 1LTMD/ZnO heterostructures were subjected to 355 nm wavelength laser excitation than when they were excited with a 514 nm wavelength laser, which we attributed to the onset of energy transfer by photoexcited excitons and/or the additional p-doping by photoexcited holes in ZnO. The p-doping phenomenon and the enhanced light emission of 1L-TMD/ZnO heterostructures were unambiguously visualized in spatially resolved PL and Raman spectral maps. Our approach using the 1L-TMD/ZnO TF heterostructure suggests that a rich variety of options for engineering the optical properties of 1L-TMDs may be made available by carrying out simple and intuitive manipulations of exciton complexes, and these endeavors may yield practical applications for 1L-TMDs in nanophotonic devices. KEYWORDS: molybdenum disulfide, molybdenum diselenide, tungsten diselenide, 2D semiconductors, excitons, photoluminescence, charge transfer
■
INTRODUCTION Atomically thin transition-metal dichalcogenides (TMDs) have attracted a great deal of attention with regards to their fundamental physics and various applications.1,2 These twodimensional (2D) semiconductors undergo remarkable changes in their electronic structures, depending on the number of their constituent layers, from indirect band gap bulk semiconductors to direct band gap monolayer semiconductors. Monolayer (1L) MoS2 and its analogues (MoSe2, WS2, WSe2, etc.) are examples of TMDs that have potential applications in light-emitting devices,3,4 phototransistors5,6 and other transistors,7 and sensors.8 Electron−hole pairs in 1L-MoS2 form stable exciton states even at room temperature1,2 because of the strong Coulomb interactions in such a confined 2D system. Trions are also formed, by two electrons (or one electron) and a hole (or two holes), when there is an excess of one charge or the other.9−11 © 2016 American Chemical Society
Changing the carrier density is an effective method to modulate the optical properties of 1L-TMDs.9,12−16 Excitons and trions in 1L-TMDs, in addition to being of scientific interest, have found practical applications in the electrical control of light emission.11,17 So far, manipulation of exciton and trion states by controlling the population of carriers in TMDs has been carried out in 1L-TMDs of MoS2, MoSe2, WS2, and WSe2,9,12,13,18 with electrical doping,9,12−14,17,18 or by chemical dopings,16,19 and by forming vertical heterostructures (HSs) with 1L-TMDs.20,21 In the experiments involving intrinsically n-type 1L-TMDs such as monolayers of MoS2 and MoSe2, further n-doping usually resulted in a reduction of photoluminescence (PL) while pReceived: July 1, 2016 Accepted: October 10, 2016 Published: October 10, 2016 28809
DOI: 10.1021/acsami.6b08003 ACS Appl. Mater. Interfaces 2016, 8, 28809−28815
Research Article
ACS Applied Materials & Interfaces
Figure 1. (a) X-ray diffraction pattern of the ZnO thin film (TF). The inset shows a scanning electron microscopy cross-sectional view of the ZnO TF prepared on a quartz substrate. (b) Representative PL spectra of the ZnO TF. Lasers with two different wavelengths (514 and 355 nm) were used for PL excitation. No PL peak was detected from the ZnO TF with the 514 nm line laser while a strong PL response was obtained under 355 nm wavelength laser excitation. A transmittance spectrum of the ZnO TF is displayed in the inset, indicating the absorption edge at 3.35 eV.
Figure 2. (a) Schematic of the sample structure and the band alignment describing the transfer of electrons from 1L-MoS2 to the ZnO TF under 514 nm wavelength laser excitation. (b) 3D rendering of a PL intensity map of 1L-MoS2 and the 1L-MoS2/ZnO HS. (c) Representative PL spectra of 1LMoS2 and the 1L-MoS2/ZnO HS. The dashed and dot-dashed lines indicate the PL peak positions of 1L-MoS2 and the 1L-MoS2/ZnO HS. The inset shows a PL peak position map of 1L-MoS2 and the 1L-MoS2/ZnO HS. (d, e) Representative deconvoluted PL spectra of 1L-MoS2 (d) and the 1LMoS2/ZnO HS (e). (f) Representative Raman spectra of 1L-MoS2 and the 1L-MoS2/ZnO HS. The inset displays the peak position map of the Raman A1g mode. All PL and Raman measurements were taken with 514 nm and 532 nm wavelength laser excitations, respectively..
eV at 300 K), has a high exciton binding energy (∼60 meV at 300 K),29 and is also a good candidate for serving as a transparent conducting oxide because it extracts photogenerated electrons from active materials and transfers them to external circuits.30 Since ZnO is a semiconductor, the electronic band structures of 1L-TMD/ZnO HSs are expected to form the type II band alignment, where the charge transfer between 1L-TMDs and ZnO would be effective,25,27,28 and would enhance the PL of intrinsically n-type 1L-TMDs. ZnO thin films (TFs) and nanostructures may also be prepared using convenient processes.31 Here we report a simple yet highly effective way to p-dope and improve the light emission of 1L-TMDs by having them form HSs with ZnO TFs. We used spatially resolved PL and Raman spectroscopy to image HSs of (1L-MoS2, 1L-MoSe2,
doping or depleting excess electrons caused a PL enhancement.9,16,19,21 Substrates were found to have significant influences on the PL characteristics of 1L-MoS2, as comprehensive optical characterizations and analyses have been carried out using a variety of dielectric and metallic substrates.22−24 Most such substrates, especially SiO2, which naturally n-dopes 1L-MoS2, were reported to cause an increase of the trion population and thus a reduction of PL. These studies showed there to be a considerable upside for enhancing the PL of 1L-MoS2 by interfacing it with materials other than SiO2. Zinc oxide (ZnO) is an example of such a material that can be used to form HSs with 1L-TMDs and be advantageous for enhancing the optical, electrical, and photocatalytic properties of the TMDs.25−28 This oxide displays a wide band gap (∼3.3 28810
DOI: 10.1021/acsami.6b08003 ACS Appl. Mater. Interfaces 2016, 8, 28809−28815
Research Article
ACS Applied Materials & Interfaces
suspended 1L-MoS2 configuration as well as comparisons of the measured and normalized PL spectra are provided in Figure S2. Also note that the PL peak position was significantly blueshifted in the 1L-MoS2/ZnO HS, by 32 meV, compared to that of 1L-MoS2 (Figure 2c). As shown in the inset of Figure 2c, this blue shift was observed uniformly on the HS. We deconvoluted the representative PL spectra by fitting them with exciton peaks of the A neutral exciton (A0), A trion (A−), and B exciton, and found that upon forming the HS, the spectral weight of A0 increased from 11% to 64%, while that of the trion decreased from 89% to 36%, as shown in Figure 2d,e. This spectral modification observed in the HS was attributed to the depletion of excess electrons present in 1L-MoS2 upon it forming an HS with ZnO TF. As schematically described in Figure 2a, due to the lower lying conduction band of 1L-MoS2 than that of ZnO, the excess electrons present in intrinsically n-type 1L-MoS2 can transfer from MoS2 to ZnO.25,27 This process of charge transfer would result in the depletion of electrons, similarly to the pdoping effects by chemical doping or molecular physisorption process,14,16,18 helping converting trions to neutral excitons which will lead to a PL enhancement of 1L-MoS2. We conducted the same PL measurements using another set of CVD-grown 1L-MoS2 samples and observed the same trend of p-doping of 1L-MoS2 by ZnO TF. Furthermore, to confirm that ZnO TF in our HS is not just “less n-doping” compared to the quartz substrate that is known to n-dope 1L-TMDs prepared on it,22,23 we conducted the same PL and Raman measurements and X-ray photoemission spectroscopy (XPS) on the same-batch 1L-MoS2 prepared on charge-neutral multilayer hBN as the results are shown in Figures S3, S4, and S5. The results showed that the spectral modification and PL increase of 1L-MoS2 in HS by interfacing with ZnO TF was much larger than those of 1L-MoS2 prepared on multilayer h-BN and XPS also showed larger downshifts of Mo and S core level peaks in 1L-MoS2 in HS than in 1L-MoS2 on h-BN, which indicates that ZnO TF actually decreased the electron density of 1L-MoS2 in the HS. The observed PL increase by charge (electron) transfer from 1L-MoS2 to ZnO TF needs more discussion because usually when charge transfer is active, photoexcited excitons tend to be separated, resulting in PL quenching, as was the case for 1LWS2 prepared on graphene.34 In our HS, the charge transfer seemed to occur mostly by the excess electrons in intrinsically n-type 1L-MoS2 that would form trions while the neutral excitons are not separated. Therefore, the charge transfer in the HS acted much like the typical p-doping process by chemical treatments16 converting “light-inefficient” trions to “lightefficient” excitons, resulting in the increase of PL. We do not have direct evidence for this interpretation, but similar reasoning was used to explain the PL increase of 1L-MoSe2 in MoSe2/MoS2 heterostacked multilayers21 and also is intuitively plausible considering that trions of 1L-MoS2 are rather weakly bound with the binding energy of only ∼30 meV while binding of exciton is much stronger of which binding energy ranges 0.5−0.9 eV.35 We performed Raman spectral mapping on the same region studied above, and representative Raman spectra from pristine 1L-MoS2 and the HS are displayed in Figure 2f. The relative reduction of the Raman peak intensity in the HS was due to the optical interference effect caused by the ZnO TF (see Figure S2). The out-of-plane vibration A1g mode of such Raman spectra is known to represent the doped state of 1L-MoS2 due to strong electron−phonon coupling along the c-axis of
and 1L-WSe2)/ZnO TFs and showed that the p-doping effect occurred by way of charge transfer between the 1L-TMDs and ZnO TFs. Photoexcitation of the ZnO in the 1L-MoS2/ZnO HS using a laser with a wavelength of 355 nm resulted in a 17fold enhancement of the PL of the 1L-TMD, suggesting that not only were electrons depleted from the 1L-TMD but also that photogenerated holes may have participated in p-doping of the 1L-TMD.
■
RESULTS AND DISCUSSION Prior to studying the characteristics of the 1L-TMD/ZnO HSs, the spatial uniformities of CVD-grown pristine 1L-TMDs and ZnO TFs were confirmed by examining confocal PL and Raman spectroscopy maps. Characteristic PL and Raman peaks were identified in pristine 1L-TMDs of MoS2, MoSe2 and WSe2, confirming the single-layer nature of 1L-TMDs (see Figure S1 of the Supporting Information). The intensities and the peak positions of the PL and Raman spectra of 1L-MoS2 were uniform throughout the samples, except for a small area of multilayer formation in the center of the 1L-MoSe2 sample as shown in Figure S1. As shown in Figure 1a, the X-ray diffraction pattern of the ZnO TF showed crystalline peaks, with the most intense peak in the (002) direction; this result indicated the preferred c-axis orientation, which is common in ZnO because the (001) basal plane of ZnO has the lowest surface energy.29 The average thickness of the ZnO TFs was measured to be ∼50 nm, as shown in the inset scanning electron microscopy image of Figure 1a. When excited with a 355 nm wavelength laser, the ZnO TF yielded a PL emission with two features (Figure 1b): a band edge at ∼3.3 eV and a broad defect emission centered at ∼1.8 eV originating from oxygen interstitials.32,33 With 514 nm wavelength laser excitation, no noticeable PL was detected because the energy of the excitation laser (2.41 eV) was less than the ∼3.35 eV band gap energy of ZnO which was estimated from the absorption spectra of the ZnO TFs displayed in the inset of Figure 1b. We transferred MoS2, MoSe2, and WSe2 1L-TMDs onto the ZnO TFs to fabricate the respective HSs and took confocal PL and Raman mapping measurements using 514 and 355 nm wavelength laser excitations. To directly compare pristine 1LTMDs and 1L-TMD/ZnO HSs, each 1L-TMD was placed on the scratched edge of the ZnO TF as shown in Figure 2a. The integrated PL intensity of the 1L-MoS2/ZnO HS region was much stronger than that of the pristine 1L-MoS2 region under 514 nm wavelength laser excitation, as shown in the 3D rendering of the PL intensity map of Figure 2b. Figure 2c displays representative PL spectra of 1L-MoS2 and the 1LMoS2/ZnO HS. The PL intensity of 1L-MoS2/ZnO HS was observed to be 6.6 times greater than that of pristine 1L-MoS2 on a quartz substrate. We checked the effect of optical interference due to the presence of the ZnO TF on the PL intensity by normalizing the PL spectra with respect to the suspended 1L-MoS2 configuration using the analytical process developed previously.22 The results showed that the geometric optical effect of the ZnO TF on the PL spectra was a ca. threefold reduction in the PL intensity in the wavelength range 600−700 nm relative to that of the bare quartz substrate, indicating that the actual enhancement (i.e., that excluding the optical interference effect) of the PL intensity of 1L-MoS2 by having it form a heterostructure with the ZnO TF was ca. 17fold. The calculated enhancement factors of the PL of 1L-MoS2 with quartz or ZnO TF/quartz substrates relative to that of the 28811
DOI: 10.1021/acsami.6b08003 ACS Appl. Mater. Interfaces 2016, 8, 28809−28815
Research Article
ACS Applied Materials & Interfaces
Figure 3. (a) Schematic of the sample structure and the band alignment describing the transfer of electrons from 1L-MoS2 to the ZnO TF and that of holes from the ZnO TF to 1L-MoS2 under 355 nm wavelength laser excitation. (b) Representative PL spectra of 1L-MoS2, the ZnO TF, and the 1L-MoS2/ZnO HS under the 355 nm wavelength excitation. (c) The difference (red curve) between the PL spectrum of 1L-MoS2/ZnO HS and that of the ZnO TF. The PL spectrum of 1L-MoS2 is also displayed for comparison. PL spectra were not normalized and are directly comparable in intensity. The dashed and dot-dashed lines indicate the PL peak positions of 1L-MoS2 and the 1L-MoS2/ZnO HS, respectively.
Figure 4. (a) 3D rendering of a PL intensity map of 1L-MoSe2 and the 1L-MoSe2/ZnO HS subjected to 514 nm wavelength laser excitation. (b) Representative PL spectra obtained from 1L-MoSe2 and the 1L-MoSe2/ZnO HS subjected to 514 nm wavelength laser excitation. The dotted and dot-dashed lines indicate the PL peak positions of 1L-MoSe2 and the 1L-MoSe2/ZnO HS, respectively. The inset shows the PL peak position map of 1L-MoSe2 and the 1L-MoSe2/ZnO HS. (c, d) Deconvoluted PL spectra of 1L-MoSe2 (c) and the 1L-MoSe2/ZnO HS (d).
TMDs.19,36−38 The A1g peak of 1L-MoS2 in the HS was observed to be blue-shifted by 1.3 cm−1 compared to the A1g peak of pristine 1L-MoS2. This relative shift of the A1g peak can be interpreted as resulting from decreased electron density caused by electron transfer,39 implying the p-doping effect of the 1L-MoS2 in the HS. We attributed the red shift of the E12g peak by 0.5 cm−1 in the HS to the tensile strain in 1L-MoS2 produced by the mismatch between its lattice and that of the ZnO TF in the HS. The spatial map of the A1g peak position, shown as an inset of Figure 2f, indicated that this trend of the A1g peak being blue-shifted in the HS was occurring uniformly. We performed PL spectral mapping on the same HS sample using 355 nm wavelength laser excitation. With this laser excitation energy, not only was 1L-MoS2 photoexcited but so also was ZnO, and this latter photoexcitation generated
excitons on the ZnO side of the HS. We can expect that such excitons in ZnO can participate in further enhancing the light emission from 1L-MoS2 in HS through Förster energy transfer and/or the transfer of photoexcited holes from ZnO to MoS2,27 as schematically described in Figure 3a. If the photogenerated holes in the ZnO TF transfer to 1L-MoS2 and participate in p-doping and the conversion of trions into neutral excitons of 1L-MoS2, the PL emission from 1L-MoS2 can be further increased. Representative PL spectra of the 1L-MoS2/ZnO HS, pristine 1L-MoS2, and the ZnO TF subjected to 355 nm wavelength laser excitation are shown in Figure 3b. In the PL spectrum of the HS, a MoS2 PL emission was clearly noted at ∼1.88 eV. This emission overlapped the broad PL emission of the ZnO TF, which originated from deep level defect states of ZnO.32,33 28812
DOI: 10.1021/acsami.6b08003 ACS Appl. Mater. Interfaces 2016, 8, 28809−28815
Research Article
ACS Applied Materials & Interfaces
Figure 5. (a) 3D rendering of a PL intensity map of 1L-WSe2 and the 1L-WSe2/ZnO HS subjected to 514 nm wavelength laser excitation. (b) Representative PL spectra obtained from 1L-WSe2 and the 1L-WSe2/ZnO HS under 514 nm wavelength laser excitation. The dotted and dot-dashed lines indicate the PL peak positions of 1L-WSe2 and the 1L-WSe2/ZnO HS, respectively. The inset shows a PL peak position map of 1L-WSe2 and the 1L-WSe2/ZnO HS. (c, d) Deconvolutions of representative PL spectra of 1L-WSe2 (c) and the 1L-WSe2/ZnO HS (d).
observed uniform blue shift of the PL peak position resulting from the HS as shown in the inset of Figure 4b; this blue shift was due to the increased neutral exciton emission and decreased trion emission. The deconvoluted PL spectra of 1L-MoSe2 and the 1L-MoSe2/ZnO HS in Figure 4c,d showed that the spectral weight of A0 increased from 73% to 82% while that of A− decreased from 27% to 18%. We also used 355 nm wavelength laser excitation for the 1L-MoSe2/ZnO HS and obtained PL spectra. The PL of 1L-MoSe2 derived from the difference between the PL of the HS and that of the ZnO film was 7.1 times stronger than the PL from the pristine 1L-MoSe2, and the peak of the PL difference was located at 1.573 eV, which was ∼10 meV higher than the PL peak of the HS obtained with the 514 nm wavelength laser excitation (Figure S6). The larger enhancement and blue shift of 1L-MoSe2 PL suggest again that the p-doping effect was stronger with the 355 nm wavelength laser excitation due to the transfer of holes in addition to the transfer of excess electrons from MoSe2 to ZnO. Since we attributed the enhancement of the PL of the n-type 1L-TMDs such as 1L-MoS2 and 1L-MoSe2 upon forming HSs with ZnO films to p-doping effects through charge exchanges across the 1L-TMD/ZnO TF interface, we may expect opposite PL behaviors such as decreased exciton emissions and increased trion emissions in HSs consisting of ZnO and p-type 1LTMDs. To test this expectation, we prepared a 1L-WSe2/ZnO TF HS since CVD-grown 1L-WSe2 has been reported to exhibit rather p-type semiconductor characteristics.40 As seen in Figure 5a, which shows an integrated PL intensity map of 1L-WSe2 and the HS, the HS yielded slightly more intense PL than did 1L-WSe2 on quartz. Their PL spectra showed more distinct differences, displaying a red shift by ∼10 meV in the PL peak position resulting from the HS compared to that from 1L-WSe2 on the quartz substrate. Here, 1L-WSe2 yielded two PL emission peaks, corresponding to neutral excitons and to trions, according to the deconvoluted PL spectra in Figure 5c,d.
To quantitatively compare (for the 355 nm wavelength laser excitation) the plain 1L-MoS2 spectrum and the contribution of 1L-MoS2 to the PL spectrum of the 1L-MoS2/ZnO HS, the difference between the PL spectrum resulting from the 355 nm wavelength excitation of the HS and that of the ZnO TF was calculated, and is shown in Figure 3c. The presence of MoS2 PL in the PL spectrum of the HS appeared as a peak centered at 1.876 eV in the difference spectrum plot. Note that the apparent intensity of this difference peak was observed to be about 16.7 times greater than the corresponding PL peak of pristine 1L-MoS2 under the 355 nm wavelength laser excitation. This enhancement is almost 3 times greater than that obtained with 514 nm wavelength laser excitation. In addition, the 1.876 eV peak position of the difference spectrum resulting from the 355 nm wavelength laser excitation was ∼10 meV higher than the PL peak position of the HS under 514 nm wavelength laser excitation. This larger modification of the 1L-MoS2 PL spectrum when using an excitation energy greater than the band gap of the ZnO suggests the additional p-doping effect by photogenerated holes in the ZnO TF under the 355 nm wavelength laser excitation. Here we note that while there is a distinct indication of stronger p-doping by 355 nm wavelength laser excitation, the Förster energy-transfer process may have contributed to the enhanced PL emission of 1L-MoS2 in HS. Because the photogenerated excitons in ZnO can decay through either of the charge transfer or the energy transfer, we were not able to determine which process is dominant in our HS. p-Doping and enhanced light emission was also observed in CVD-grown 1L-MoSe2/ZnO HS. As shown in the 3D-rendered PL image in Figure 4a, the 1L-MoSe2/ZnO HS yielded much stronger PL emission than did pristine 1L-MoSe2. And according to the representative PL spectra in Figure 4b, the HS yielded a 4.7-fold stronger PL peak than did pristine 1LMoSe2. An additional indication of the p-doping was the 28813
DOI: 10.1021/acsami.6b08003 ACS Appl. Mater. Interfaces 2016, 8, 28809−28815
Research Article
ACS Applied Materials & Interfaces
argon atmosphere to remove residue and to further improve the adhesion between the 1Ls and the ZnO TFs.20 Confocal PL and Raman Spectral Mapping Measurements. For the confocal PL and Raman imaging and spectroscopy, a combination of a laser confocal microscope made in house and a spectrometer was used.20,21,41 The laser light was focused with a 0.9 NA objective lens. The scattered light was collected using the same objective and guided to a 50 cm long monochromator equipped with a cooled CCD. The excitation laser was the 514 nm wavelength laser line of an argon laser for the PL measurements and the 532 nm wavelength line of a solid-state laser for the Raman measurements. Laser power was 300 μW on the sample with a typical acquisition time of 5 and 50 ms per pixel for PL and Raman spectral imaging, respectively.
Considering the p-type nature of intrinsic 1L-WSe2, we attributed the trion peak to positive trions (A+). Therefore, the p-doping effect by ZnO TF must have caused the relative increase of trion emissions in the 1L-WSe2 of the HS and thus the red shift of its PL peak compared to that of 1L-WSe2 on the quartz substrate. Indeed, upon incorporation into the HS, the spectral weight of the A0 peak was shown to decrease from 75% to 47% while that of A+ increased from 25% to 53%, as described in Figure 5c,d.
■
CONCLUSIONS We greatly enhanced the photoluminescence from n-type 1LTMDs by having them form heterostructures with ZnO thin films. Increased exciton emissions and suppressed trion emissions originating from p-doping charge exchanges through 1L-TMDs/ZnO TF interfaces were unambiguously revealed by nanoscale PL and Raman spectral mapping measurements. Our approach of stacking n-type 1L-TMDs on ZnO TFs provides a simple way to engineer the spectral emission properties of 1LTMDs without using chemical processes or external fields. Moreover, when combined with the possible patterning of ZnO TFs on arbitrary substrates, our method can help overcome the drawbacks of low quantum yields of 1L-TMDs and find diverse practical applications in optoelectronic or nanophotonic devices.
■
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.6b08003. PL and Raman spectral maps and the representative spectra of 1L-TMDs; MoS2, MoSe2, and WSe2 (S1); calculated enhancement factors as a function of wavelength for 1L-MoS2 on a quartz substrate and 1L-MoS2 on a 50 nm ZnO TF/quartz substrate under 355 and 514 nm laser excitation compared to the suspended 1L-MoS2 (S2); PL intensity and peak positions maps and the representative spectra of 1L-MoS2, 1L-MoS2/h-BN, and 1L-MoS2/ZnO HS (S3); representative PL spectra of 1LMoS2, 1L-MoS2/h-BN, and 1L-MoS2/ZnO HS (S4); XPS spectra of 1L-MoS2/h-BN and 1L-MoS2/ZnO HS (S5); representative PL spectra of the 1L-MoSe2, ZnO thin film, and 1L-MoSe2/ZnO HS under 355 nm laser excitation (S6) (PDF)
MATERIALS AND METHODS
Monolayer TMD Synthesis and ZnO TF Growth. MoS2, MoSe2, and WSe2 1Ls were separately grown on respective SiO2/Si substrates by carrying out chemical vapor deposition (CVD). In each case, the growth was promoted by first spin-coating perylene-3,4,9,10tetracarboxylic acid tetrapotassium salt (2D semiconductor) onto the SiO2/Si substrate at 2600 rpm for 1 min. The SiO2/Si substrates were also spin-coated using solution precursors: MoO3 for MoS2 and MoSe2 1Ls and WO3 for WSe2 1Ls. For MoS2 1Ls, the upstream side with 200 mg of S (Sigma-Aldrich, St. Louis) was heated to 210 °C at a ramping rate of 42 °C/min, whereas the temperature of the downstream side with the spin-coated substrate was ramped up to 780 °C. The whole process was carried out under 500 sccm of N2 delivered over 15 min. For MoSe2 1Ls, the procedure followed was the same as that for the MoS2 1Ls except that the temperature at the upstream side with 200 mg of Se (Sigma-Aldrich) was 410 °C. The whole process was carried out under 500 sccm of N2 and 5 sccm of H2 delivered over 15 min. For WSe2 1Ls, the procedure followed was the same as that for the MoSe2 1Ls except that the temperature at the downstream side with the spincoated substrate was 800 °C. ZnO TFs were deposited on quartz substrates by using the sol−gel spin-coating method. The spin-coating solution contained zinc acetate dehydrate [Zn(CH3COO)2·2H2O] and an equivalent molar amount of monoethanolamine [NH2CH2CH2OH] dissolved in 2-methoxyethanol [CH3OCH2CH2OH]. The concentration of the zinc acetate was 0.5 mol. A detailed description of the typical spin-coating method used was provided in the previous report.26 Transfer of 1L TMDs onto ZnO TFs. 1Ls of MoS2, MoSe2, and WSe2 grown on SiO2/Si wafers by CVD were transferred to the sol− gel ZnO TFs using a typical wet transfer method.20 An aqueous HF solution was used for the transfers of the 1L-TMDs. Prior to this transfer, the sol−gel ZnO TFs were partially removed, by applying a simple scratch method using a diamond tip, to directly compare the characteristics of the 1L-TMDs and the heterostructures. And then, the poly(methyl methacrylate) (PMMA)/1L-TMD films were transferred onto the ZnO TFs. The PMMA layer covering the 1L-TMDs was removed by using acetone and the sample was baked at 80 °C for 30 min to improve the adhesion between the 1L-TMDs and the ZnO TFs. The resulting HSs were then baked at 200 °C for 2 h under an
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected] (J.K.). Author Contributions #
M.S.K. and S.R. contributed equally to this work.
Notes
The authors declare no competing financial interest.
■ ■
ACKNOWLEDGMENTS This work was supported by IBS-R011-D1. REFERENCES
(1) 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. (2) 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. (3) 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. (4) Sundaram, R. S.; Engel, M.; Lombardo, A.; Krupke, R.; Ferrari, A. C.; Avouris, P.; Steiner, M. Electroluminescence in Single Layer MoS2. Nano Lett. 2013, 13, 1416−1421. (5) Yang, S.; Wang, C.; Ataca, C.; Li, Y.; Chen, H.; Cai, H.; Suslu, A.; Grossman, J. C.; Jiang, C.; Liu, Q.; Tongay, S. Self-Driven Photodetector and Ambipolar Transistor in Atomically Thin GaTeMoS2 p-n vdW Heterostructure. ACS Appl. Mater. Interfaces 2016, 8, 2533−2539. (6) Lee, H. S.; Min, S.-W.; Chang, Y.-G.; Park, M. K.; Nam, T.; Kim, H.; Kim, J. H.; Ryu, S.; Im, S. MoS2 Nanosheet Phototransistors with 28814
DOI: 10.1021/acsami.6b08003 ACS Appl. Mater. Interfaces 2016, 8, 28809−28815
Research Article
ACS Applied Materials & Interfaces Thickness-Modulated Optical Energy Gap. Nano Lett. 2012, 12, 3695−3700. (7) Radisavljevic, B.; Radenovic, A.; Brivio, J.; Giacometti, V.; Kis, A. Single-Layer MoS2 Transistors. Nat. Nanotechnol. 2011, 6, 147−150. (8) Cho, B.; Yoon, J.; Lim, S. K.; Kim, A. R.; Kim, D.-H.; Park, S.-G.; Kwon, J.-D.; Lee, Y.-J.; Lee, K.-H.; Lee, B. H.; Ko, H. C.; Hahm, M. G. Chemical Sensing of 2D Graphene/MoS2 Heterostructure Device. ACS Appl. Mater. Interfaces 2015, 7, 16775−16780. (9) Mak, K. F.; He, K.; Lee, C.; Lee, G. H.; Hone, J.; Heinz, T. F.; Shan, J. Tightly Bound Trions in Monolayer MoS2. Nat. Mater. 2013, 12, 207−211. (10) Berkelbach, T. C.; Hybertsen, M. S.; Reichman, D. R. Theory of Neutral and Charged Excitons in Monolayer Transition Metal Dichalcogenides. Phys. Rev. B: Condens. Matter Mater. Phys. 2013, 88, 045318. (11) Lui, C. H.; Frenzel, A. J.; Pilon, D. V.; Lee, Y.-H.; Ling, X.; Akselrod, G. M.; Kong, J.; Gedik, N. Trion-Induced Negative Photoconductivity in Monolayer MoS2. Phys. Rev. Lett. 2014, 113, 166801. (12) Ross, J. S.; Wu, S.; Yu, H.; Ghimire, N. J.; Jones, A. M.; Aivazian, G.; Yan, J.; Mandrus, D. G.; Xiao, D.; Yao, W.; Xu, X. Electrical Control of Neutral and Charged Excitons in a Monolayer Semiconductor. Nat. Commun. 2013, 4, 1474. (13) Jones, A. M.; Yu, H.; Ghimire, N. J.; Wu, S.; Aivazian, G.; Ross, J. S.; Zhao, B.; Yan, J.; Mandrus, D. G.; Xiao, D.; Yao, W.; Xu, X. Optical Generation of Excitonic Valley Coherence in Monolayer WSe2. Nat. Nanotechnol. 2013, 8, 634−638. (14) Tongay, S.; Zhou, J.; Ataca, C.; Liu, J.; Kang, J. S.; Matthews, T. S.; You, L.; Li, J.; Grossman, J. C.; Wu, J. Broad-Range Modulation of Light Emission in Two-Dimensional Semiconductors by Molecular Physisorption Gating. Nano Lett. 2013, 13, 2831−2836. (15) Lin, J. D.; Han, C.; Wang, F.; Wang, R.; Xiang, D.; Qin, S.; Zhang, X.-A.; Wang, L.; Zhang, H.; Wee, A. T. S.; Chen, W. ElectronDoping-Enhanced Trion Formation in Monolayer Molybdenum Disulfide Functionalized with Cesium Carbonate. ACS Nano 2014, 8, 5323−5329. (16) Mouri, S.; Miyauchi, Y.; Matsuda, K. Tunable Photoluminescence of Monolayer MoS2 via Chemical Doping. Nano Lett. 2013, 13, 5944−5948. (17) Ross, J. S.; Klement, P.; Jones, A. M.; Ghimire, N. J.; Yan, J.; Mandrus, D. G.; Taniguchi, T.; Watanabe, K.; Kitamura, K.; Yao, W.; Cobden, D. H.; Xu, X. Electrically Tunable Excitonic Light-Emitting Diodes Based on Monolayer WSe2 p-n Junctions. Nat. Nanotechnol. 2014, 9, 268−272. (18) Shang, J.; Shen, X.; Cong, C.; Peimyoo, N.; Cao, B.; Eginligil, M.; Yu, T. Observation of Excitonic Fine Structure in a 2D TransitionMetal Dichalcogenide Semiconductor. ACS Nano 2015, 9, 647−655. (19) Dhakal, K. P.; Duong, D. L.; Lee, J.; Nam, H.; Kim, M. S.; Kan, M.; Lee, Y. H.; Kim, J. Confocal Absorption Spectral Imaging of MoS2: Optical Transitions Depending on the Atomic Thickness of Intrinsic and Chemically Doped MoS2. Nanoscale 2014, 6, 13028−13035. (20) Park, S.; Kim, M. S.; Kim, H.; Lee, J.; Han, G. H.; Jung, J.; Kim, J. Spectroscopic Visualization of Grain Boundaries of Monolayer Molybdenum Disulfide by Stacking Bilayers. ACS Nano 2015, 9, 11042−11048. (21) Kim, M. S.; Seo, C.; Kim, H.; Lee, J.; Luong, D. H.; Park, J.-H.; Han, G. H.; Kim, J. Simultaneous Hosting of Positive and Negative Trions and the Enhanced Direct Band Emission in MoSe2/MoS2 Heterostacked Multilayers. ACS Nano 2016, 10, 6211−6219. (22) Buscema, M.; Steele, G. A.; van der Zant, H. S. J.; CastellanosGomez, A. The Effects of the Substrate on the Raman and Photoluminescence Emission of Single-Layer MoS2. Nano Res. 2014, 7, 561−571. (23) Li, Y.; Qi, Z.; Liu, M.; Wang, Y.; Cheng, X.; Zhang, G.; Sheng, L. Photoluminescence of Monolayer MoS2 on LaAlO3 and SrTiO3 Substrates. Nanoscale 2014, 6, 15248−15254. (24) Lien, D.-H.; Kang, J. S.; Amani, M.; Chen, K.; Tosun, M.; Wang, H.-P.; Roy, T.; Eggleston, M. S.; Wu, M. C.; Dubey, M.; Lee, S.-C.;
He, J.-H.; Javey, A. Engineering Light Outcoupling in 2D Materials. Nano Lett. 2015, 15, 1356−1361. (25) Chen, L.; Xue, F.; Li, X.; Huang, X.; Wang, L.; Kou, J.; Wang, Z. L. Strain-Gated Field Effect Transistor of a MoS2-ZnO 2D-1D Hybrid Structure. ACS Nano 2016, 10, 1546−1551. (26) Liu, Y.; Xie, S.; Li, H.; Wang, X. A Highly Efficient Sunlight Driven ZnO Nanoshhet Photocatalyst: Synergetic Effect of P-Doping and MoS2 Atomic Layer Loading. ChemCatChem 2014, 6, 2522−2526. (27) Tan, Y.-H.; Yu, K.; Li, J.-Z.; Fu, H.; Zhu, Z.-Q. MoS2@ZnO Nano-Heterojunctions with Enhanced Photocatalysis and Field Emission Properties. J. Appl. Phys. 2014, 116, 064305. (28) Zhang, K.; Zhang, Y.; Zhang, T.; Dong, W.; Wei, T.; Sun, Y.; Chen, X.; Shen, G.; Dai, N. Vertically Coupled ZnO Nanorods on MoS2 Monolayers with Enhanced Raman and Photoluminescence Emission. Nano Res. 2015, 8, 743−750. (29) Ö zgür, Ü .; Alivov, I.; Liu, C.; Teke, A.; Reshchikov, M. A.; Doğan, S.; Avrutin, V.; Cho, S.-J.; Morkoç, H. A Comprehensive Review of ZnO Materials and Devices. J. Appl. Phys. 2005, 98, 041301. (30) Lee, S.-H.; Han, S.-H.; Jung, H. S.; Shin, H.; Lee, J.; Noh, J.-H.; Lee, S.; Cho, I.-S.; Lee, J.-K.; Kim, J.; Shin, H. Al-Doped ZnO TF: a New Transparent Conducting Layer for ZnO Nanowire-Based DyeSensitized Solar Cells. J. Phys. Chem. C 2010, 114, 7185−7189. (31) Seong, K.; Kim, K.; Park, S. Y.; Kim, Y. S. Micro-Patterned ZnO Semiconductors for High Performance Thin Film Transistors via Chemical Imprinting with A PDMS Stamp. Chem. Commun. 2013, 49, 2783−2785. (32) Kim, M. S.; Kim, S.; Leem, J.-Y. Laser-Assisted Sol-Gel Growth and Characteristics of ZnO Thin Films. Appl. Phys. Lett. 2012, 100, 252108. (33) Ahn, C. H.; Kim, Y. Y.; Kim, D. C.; Mohanta, S. K.; Cho, H. K. A Comparative Analysis of Deep Level Emission in ZnO Layers Deposited by Various Methods. J. Appl. Phys. 2009, 105, 013502. (34) Tan, H.; Fan, Y.; Rong, Y.; Porter, B.; Lau, C. S.; Zhou, Y.; He, Z.; Wang, S.; Bhaskaran, H.; Warner, J. H. Doping Graphene Transistors Using Vertical Stacked Monolayer MoS2 Heterostructures Grown by Chemical Vapor Deposition. ACS Appl. Mater. Interfaces 2016, 8, 1644−1652. (35) Shi, H.; Pan, H.; Zhang, Y.-W.; Yakobson, B. I. Quasiparticle Band Structure and Optical Properties of Strained Monolayer MoS2 and WS2. Phys. Rev. B: Condens. Matter Mater. Phys. 2013, 87, 155304. (36) Tonndorf, P.; Schmidt, R.; Böttger, P.; Zhang, X.; Börner, J.; Liebig, A.; Albrecht, M.; Kloc, C.; Gordan, O.; Zahn, D. R. T.; de Vasconcellos, S. M.; Bratschitsch, R. Photoluminescence Emission and Raman Response of Monolayer MoS2, MoSe2, and WSe2. Opt. Express 2013, 21, 4908−4916. (37) Zhao, W.; Ghorannevis, Z.; Amara, K. K.; Pang, J. R.; Toh, M.; Zhang, X.; Kloc, C.; Tan, P. H.; Eda, G. Lattice Dynamics in Monoand Few-Layer Sheets of WS2 and WSe2. Nanoscale 2013, 5, 9677− 9683. (38) Li, H.; Zhang, Q.; Yap, C. C. R.; Tay, B. K.; Edwin, T. H. T.; Olivier, A.; Baillargeat, D. From Bulk to Monolayer MoS2: Evolution of Raman Scattering. Adv. Funct. Mater. 2012, 22, 1385−1390. (39) Chakraborty, B.; Bera, A.; Muthu, D. V. S.; Bhowmick, S.; Waghmare, U. V.; Sood, A. K. Symmetry-Dependent Phonon Renormalization in Monolayer MoS2 Transistor. Phys. Rev. B: Condens. Matter Mater. Phys. 2012, 85, 161403. (40) Zhou, H.; Wang, C.; Shaw, J. C.; Cheng, R.; Chen, Y.; Huang, X.; Liu, Y.; Weiss, N. O.; Lin, Z.; Huang, Y.; Duan, X. Large Area Growth and Electrical Properties of p-type WSe2 Atomic Layers. Nano Lett. 2015, 15, 709−713. (41) Kim, M. S.; Yun, S. J.; Lee, Y.; Seo, C.; Han, G. H.; Kim, K. K.; Lee, Y. H.; Kim, J. Biexciton Emission from Edges and Grain Boundaries of Triangular WS2 Monolayers. ACS Nano 2016, 10, 2399−2405.
28815
DOI: 10.1021/acsami.6b08003 ACS Appl. Mater. Interfaces 2016, 8, 28809−28815