Nano Res 1

Effects of dielectric stoichiometry on the photoluminescence properties of encapsulated WSe2 monolayers

Javier Martín-Sánchez1 (), Antonio Mariscal2, Marta De Luca3, Aitana Tarazaga Martín-Luengo1, Georg Gramse4, Alma Halilovic1, Rosalía Serna2, Alberta Bonanni1, Ilaria Zardo3, Rinaldo Trotta1 (), and Armando Rastelli1 Nano Res., Just Accepted Manuscript • DOI: 10.1007/s12274-017-1755-4 http://www.thenanoresearch.com on Jul. 5, 2017 © Tsinghua University Press 2017

Just Accepted

This is a “Just Accepted” manuscript, which has been examined by the peer-review process and has been accepted for publication. A “Just Accepted” manuscript is published online shortly after its acceptance, which is prior to technical editing and formatting and author proofing. Tsinghua University Press (TUP) provides “Just Accepted” as an optional and free service which allows authors to make their results available to the research community as soon as possible after acceptance. After a manuscript has been technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Please note that technical editing may introduce minor changes to the manuscript text and/or graphics which may affect the content, and all legal disclaimers that apply to the journal pertain. In no event shall TUP be held responsible for errors or consequences arising from the use of any information contained in these “Just Accepted” manuscripts. To cite this manuscript please use its Digital Object Identifier (DOI®), which is identical for all formats of publication.

Nano Research DOI 10.1007/s12274-017-1755-4

Nano Res 2

Effects of Dielectric Stoichiometry on the Photoluminescence Properties of Encapsulated WSe2 Monolayers

Javier Martín-Sánchez1*, Antonio Mariscal2, Marta De Luca3, Aitana Tarazaga

Martín-Luengo1, Georg Gramse4, Alma Halilovic1, Rosalía Serna2, Alberta Bonanni1, Ilaria

Zardo3, Rinaldo Trotta1*, and Armando Rastelli1

1 Institute of Semiconductor and Solid State Physics, Johannes Kepler University Linz,

Altenbergerstrasse 69, A-4040 Linz, Austria

2 Laser Processing Group, Instituto de Óptica, CSIC, C/Serrano 121, 28006 Madrid, Spain

3 Department of Physics, University of Basel, Klingelbergstrasse 82, 4056, Basel, Switzerland

4 Institute for Biophysics, Johannes Kepler University Linz, Gruberstrasse 40, A-4020 Linz,

Austria

KEYWORDS: 2D materials, dielectric encapsulation, transition metal dichalcogenides,

semiconductors, photoluminescence, WSe2

* e-mail: javier.martin.nano@gmail.com ; rinaldo.trotta@jku.at

ABSTRACT

Two-dimensional transition-metal-dichalcogenide semiconductors have emerged as

promising candidates for optoelectronic devices with unprecedented properties and

ultra-compact performances. However atomically thin materials are highly sensitive to

surrounding dielectric media, which imposes severe limitations to their practical applicability.

Hence for their suitable integration into devices, the development of reliable encapsulation

procedures that preserve their physical properties are required. Here, the excitonic

photoluminescence of monolayers is assessed, at room temperature and 10 K, on

mechanically exfoliated WSe2 monolayer flakes encapsulated with SiOx and AlxOy layers

employing chemical and physical deposition techniques. Conformal flakes coating on

untreated – non-functionalized – flakes is successfully demonstrated by all the techniques

2

Nano Res 3

except for atomic layer deposition, where a cluster-like oxide coating is observed. No

significant compositional or strain state changes in the flakes are detected upon encapsulation

by any of the techniques. Remarkably, our results evidence that the flakes’ optical emission is

strongly influenced by the quality of the encapsulating oxide – stoichiometry –. When the

encapsulation is carried out with slightly sub-stoichiometric oxides two remarkable

phenomena are observed. First, there is a clear electrical doping of the monolayers that is

revealed through a dominant trion – charged exciton – room-temperature

photoluminescence. Second, a strong decrease of the monolayers optical emission is

measured attributed to non-radiative recombination processes and/or carriers transfer from

the flake to the oxide. Power- and temperature-dependent photoluminescence measurements

further confirm that stoichiometric oxides obtained by physical deposition lead to a

successful encapsulation, opening a promising route for the development of integrated

two-dimensional devices.

1. Introduction

Two-dimensional (2D) semiconductor transition-metal-dichalcogenides (TMDCs) present

unique physical properties, such as a native sizeable bandgap, relatively large in-plane carrier

mobility up to 250 cm2/Vs, valley-dependent optoelectronics, strong light-matter interaction,

and indirect-to-direct bandgap transition in monolayer crystals, which make them ideal

candidates for ultra-compact optoelectronic, photonic and electronic applications [1-5].

Owing to strong spatial confinement of carriers and reduced dielectric screening of Coulomb

interactions, electron and holes in monolayer TMDCs are tightly bound, thus forming exciton

and multi-exciton complexes with binding energies up to 0.5-1 eV, more than one order of

magnitude larger than in conventional III-V 2D quantum well nanostructures [6]. The

simplest excitonic specie is the neutral exciton (X0), which consists of an electron-hole pair

bounded by Coulomb interaction. Singly charged excitons, or trions (X-), quasi-particles

3

Nano Res 4

consisting of two electrons (holes) and a hole (electron), have been reported [7,8] with

binding energies up to 20-40 meV relative to X0. Additionally, in WSe2 monolayers, broad

and relatively stable optical emission ~50 to ~200 meV below the X0 transition is typically

observed at low temperatures (< 60 K) and is associated with bound exciton states and other

multi-exciton complexes [9-11].

The optical response of 2D semiconductor materials is therefore dominated by the radiative

recombination of such exciton complexes that ultimately depends on many-body interactions

of excitons and free carriers in the crystal. In addition, the large surface-to-volume ratio

peculiar to these materials makes them highly sensitive to the surrounding environment. For

example, organic solvents commonly used in well-established cleaning protocols, such as

acetone or 2-propanol, can introduce unintentional doping in atomically-thin flakes [12], and

the photoluminescence (PL) properties of doped TMDCs are strongly affected by physisorbed

O2, O3 and/or H2O molecules, which tend to deplete the material of electrons, rendering the

neutral exciton transition dominant over negatively charged exciton complexes [13]. The

exciton dynamics in WS2, MoS2 and WSe2 monolayer flakes is also affected by the substrate

material, mainly due to strain, charge transfer or dielectric screening effects [3,14-16]. In the

case of mechanically exfoliated flakes, these effects depend on the contact area between the

exfoliated flakes and the substrate, which is affected by parameters such as substrate

preparation procedure or stiffness of the tape/stamp used to transfer the flakes [17,18]. It is

therefore not surprising to find discrepancies in published data on the characteristic optical

emission features in WSe2 monolayers at low temperatures due to uncontrollable

experimental conditions. For instance, while some works report on well-defined neutral

exciton, trion and localized states’ emissions [9,10], broad bands with no clear peaks are

found in others [11,19].

Besides fluctuations during preparation, strategies are needed to protect 2D materials against

4

Nano Res 5

environmental modifications that could produce deleterious or uncontrollable effects on their

electrical and optical properties. To this aim, different encapsulation techniques have been

explored, such as embedding between inert h-BN [20] or dielectric materials, like Al2O3,

HfO2, SiN3 or SiO2 [21-25]. It is known that a conformal uniform oxide coating of 2D

materials realized by chemical deposition techniques, such as atomic layer deposition (ALD),

is challenging and requires a surface functionalization or pre-treatment step due to the

absence of surface dangling bonds [24,26-28]. Most of the reports on Al2O3 [29], HfO2 [23,30]

or SiN3 [31,32] encapsulation effects on TMDCs 2D materials, e.g. MoS2 or WSe2, are mainly

focused on the study of their electrical response and rely on dielectric coating obtained by

chemical deposition. Systematic studies on the effects produced by encapsulation on the

flakes’ optical response are scarce and there is controversy in their interpretation [33-35].

While in some works the optical emission modification of encapsulated MoS2 flakes is

attributed to the presence of residual strain or possible electrical charging [33,34], others

relate it to dielectric screening effects [36]. In addition, possible material modifications that

could affect the flake optical emission and the role of the employed deposition techniques are

often neglected.

In this work, the optical properties of WSe2 monolayers, encapsulated by chemical and

physical deposition techniques with dielectric SiOx and AlxOy layers of different thicknesses

ranging from 5 to 30 nm, are systematically studied by PL spectroscopy. Special care is taken

to prevent possible flake degradation or unintentional doping by avoiding any treatment

procedure with organic solvents or further manipulation after the exfoliation. The

composition, strain state and morphology of the encapsulated flakes are studied by x-ray

photoelectron spectroscopy (XPS), Raman spectroscopy and atomic force microscopy (AFM),

respectively. The stoichiometry of the encapsulating oxides is measured by XPS and

correlated with the flakes’ optical emission. Our results demonstrate that the stoichiometry of

5

Nano Res 6

the oxide strongly influence the optical response of the encapsulated WSe2. Dominant trion

optical emission at room-temperature (RT) is found for flakes that were coated with

sub-stoichiometric oxides, which indicates electrical doping of the flake due to the defective

oxide. In contrast, no significant changes with respect to uncapped reference samples are

found in the PL spectra when stoichiometric oxides are employed, indicating negligible

charge transfer. Power-dependent PL at 10K and temperature-dependent PL measurements

are performed on uncapped and encapsulated monolayers to further investigate the radiative

recombination mechanisms. The suitability of each deposition technique to preserve the

optical quality of encapsulated WSe2 monolayer flakes is discussed.

2. Experimental

2.1. Samples preparation

WSe2 monolayer flakes are mechanically exfoliated on commercially available SiO2 (280

nm)/Si substrates. The substrates have been first degreased with acetone, 2-propanol and

deionized water followed by oxygen plasma treatment for 10 minutes using a commercial

system (200W) in order to enhance the flakes adhesion to the substrates [17]. The WSe2

flakes are then transferred by Scotch mechanical exfoliation on the substrates where gold

markers have been previously defined by optical lithography for flakes localization.

Monolayer flakes are identified and localized with respect to the reference markers with an

optical microscope. We limit the introduction of deleterious doping and/or contamination on

the as-exfoliated flakes by avoiding any cleaning procedure with solvents. AFM and

micro-PL are used on all the samples to confirm the monolayer thickness of the flakes. The

monolayers used in this work present optical neutral exciton emission at similar energies of

1.661 ± 0.005 eV.

2.2. WSe2 monolayer flakes oxide encapsulation by chemical and physical deposition

6

Nano Res 7

Encapsulation by plasma-enhanced chemical vapor deposition (PECVD): the samples are left

in vacuum conditions (base pressure ~1x10-2 mbar) at 300 °C with a N2 flux of 5 sccm for 30

minutes before deposition to desorb possible adsorbates from the surface. The silicon-oxide

layers deposition is performed at 300 °C and 500 mbar by using SiH4 (425 sccm) and N2O

(710 sccm) precursors with a plasma power of 10 W (13.56 MHz RF generator). The oxide

layer thickness was controlled by the deposition time. Encapsulation by ALD: the samples are

left in vacuum conditions (base pressure ~7x10-2 mbar) at 200 °C with a N2 flux of 5 sccm for

30 minutes before deposition to desorb possible adsorbates from the surface. The

aluminum-oxide deposition takes place at 200 °C by cycling Trimethylaluminium TMA (0.8

s)/H2O (1 s) precursors with a waiting time of about 8 seconds between pulses and N2 flux of

20 sccm. The thickness of the oxide layers are controlled by the number of precursors’ pulses.

Encapsulation by electron-beam deposition (E-BEAM): the samples were annealed at 200 °C

in vacuum (base pressure of ~5x10-7 mbar) for 30 minutes to remove possible absorbents

from the surface. The deposition is performed at 200 °C and 1x10-6 mbar while rotating the

sample at 15 rpm. The oxide layer is deposited at a rate 2 Å/s. Encapsulation by pulsed laser

deposition (PLD): a home-built PLD system consisting of a UV laser ArF excimer (λ = 193

nm; 20 ns pulse duration) and a vacuum chamber equipped with a multi-target system is used.

The laser beam is focused at an angle of incidence of approximately 45° onto a ceramic Al2O3

target. PLD deposition involves the formation of a laser-generated plasma formed by high

kinetic energy species. The judicious choice of the deposition conditions allows to obtain

good quality oxides for optical applications [37]. For this experiment, the samples are placed

at 43 mm from the target surface and at 5 cm with respect to the center of the target normal,

therefore using the so called off-axis deposition configuration that hinders the arrival of the

higher kinetic energy species originated in the laser induced plasma plume and minimizes

implantation and defect formation phenomena [38,39]. The oxide deposition is performed in

7

Nano Res 8

vacuum (base pressure of 1x10-5 mbar) with the samples at RT using a laser energy density of

2.0 ± 0.2 J/cm2 on the target with a repetition rate of 20 Hz.

2.3. Characterization

The optical characterization of uncapped and encapsulated WSe2 monolayers is performed by

photoluminescence spectroscopy. We use a standard confocal micro-PL setup equipped with a

X-Y stage allowing spatial positioning of the laser spot with a resolution well below 1µm. A

continuous-wave laser is used for excitation (λ=532 nm). The laser beam is focused on the

sample by using a 50x objective with a numerical aperture value of NA=0.42. The

temperature-dependent PL experiments are performed employing a He flow cryostat. PL is

analyzed with a spectrometer with 750 mm focal length coupled to a Si CCD using a 150

grooves/mm grating. In order to perform the power-dependent measurements, we use a

half-lambda wave plate combined with a fixed linear polarizer in the excitation line, which

allows controlling reproducibly the laser excitation power. The strain state of the flakes is

studied by Raman spectroscopy. Raman measurements are performed in backscattering

geometry on isolated WSe2 monolayers at RT, with excitation wavelength of 632.8 nm

provided by a HeNe laser and power density of 14 kW/cm2 (selected after checking that no

heating effects are induced). The laser beam is focused with a 100x objective (NA=0.80). The

scattered light is collected by a T64000 triple spectrometer equipped with 1800 g/mm

gratings and liquid-nitrogen cooled multichannel charge couple device detector. Spectra are

collected by selecting scattered light polarized either parallel or perpendicular to the

polarization direction of incident light, which in the so-called Porto notation correspond to

x(zz)-x and x(zy)-x, respectively. The composition of the flakes is carried out by XPS

spectroscopy with a Thermofisher Theta Probe XPS system, operated with a monochromatic

Al-Kα x-ray beam at 15 kV, emission current of 6.7 mA and a spot size of 400 µm in diameter.

8

Nano Res 9

All spectra are acquired in constant analyzer energy mode and constant pass energy of 50 eV,

with FWHM=1.00 eV on Ag 3d5/2, and energy step size of 0.05 eV. The residual pressure of

the ultra-high vacuum analysis chamber is held in the range of 10-8 mbar. The electrical

charge distribution over the flakes is analyzed by Kelvin Force Microscopy (KFM) [40].

KFM images have been acquired at with the electrical excitation signal at the 2nd cantilever

Eigenmode (f~426 kHz) [41]. A commercial AFM (Agilent 5600, USA) and conducting

cantilevers (PtSi-FM, Nanoandmore, Germany) are used. The measured surface potential

differences ΔV corresponds to the dc potential which nullifies the oscillation at 426 kHz.

Assuming a parallel capacitor model the surface potential changes can be related to the

surface charge density by σ =ΔV ε/h, where ε~8x10-11 F/m and h=5nm are the dielectric

constant and the thickness of the oxide in Figure 5e. The morphology of the flakes is

measured by AFM employing a commercial Nanoscope system is used in tapping mode. Si

AFM tips with resonance frequency around 300 KHz are employed. Data post-processing

analysis is performed using the WSxM software [42].

3. Results and discussion

3.1. Samples preparation, structural and compositional characterization of

encapsulated WSe2 monolayer flakes.

The samples used in this work are obtained by mechanical exfoliation of a bulk WSe2 crystal

(2dsemiconductors Inc.) on thermally oxidized Si substrates. The 280-nm thick SiO2/Si layer

presents a low surface roughness with a peak-to-peak value of about 1 nm. A set of samples

consisting of uncapped (reference) and oxide-coated WSe2 monolayer flakes with oxide

thicknesses t=5, 10, 20 and 30 nm, and varying stoichiometry, is prepared. For the oxide

deposition, we have chosen well-established and representative chemical deposition

techniques – ALD and PECVD – and physical deposition methods – E-BEAM and PLD –.

The stoichiometry of the layers is determined by XPS and further details can be found in the

9

Nano Res 10

supporting information, Figure S1 and Table S1. In particular, the physical deposition

techniques led to the formation of stoichiometric Al2O3 oxides (E-BEAM and PLD). On the

other hand, sub-stoichiometric oxide layers with oxygen depletion are obtained by chemical

deposition: Al2O2.85 (ALD) and SiO1.80 (PECVD). We would like to note that the deposition

of the oxides used in this work has been performed employing standard experimental

conditions under which oxide layers are routinely prepared in our laboratory. The idea behind

our experiments is to unveil the effects of non-ideal conditions in terms of oxides

stoichiometry or high energetic species, usually obtained in the PLD case, on the optical

response of the encapsulated WSe2 monolayers.

The surface morphology of as-exfoliated and oxide-coated WSe2 is investigated by AFM.

Figure 1a shows the AFM images of a WSe2 monolayer flake coated by ALD for 50, 100 and

300 precursors [Trimethylaluminium (TMA)/H2O] cycles, that correspond to a nominal oxide

thickness tnom=5, 10 and 30 nm of aluminium-oxide on the surrounding SiO2 surface.

Although pinhole-free coating is generally expected for ALD, conformal oxide growth

requires the presence of uniformly distributed nucleation centers promoting the dissociation

and reaction of the precursors. The AFM images and corresponding line-scans show that the

growth is not conformal. Instead, nanometer-sized clusters are observed, indicating a

three-dimensional deposition with peak-to-peak roughness values up to ~15 nm. Cluster

coalescence is found to occur as the nominal thickness is increased up to 30 nm, eventually

covering the whole flake surface. This ALD growth mode is observed on freshly exfoliated

flakes, due to the low density of dangling bonds or impurities that could act as oxide

nucleation centers on the monolayer’s surface. This observation confirms the cleanness of the

flake surface after the exfoliation and deposition procedures [26]. Interestingly, a conformal

growth is observed when the flake encapsulation is carried out by PECVD, as shown in

Figure 1b. The PECVD technique also involves chemical reactions during the oxide growth

10

Nano Res 11

but, in this case, the dissociation of the precursors is enhanced by the plasma. The dissociated

species react at the surface substrate and monolayers’ surface to form a uniform oxide layer

with no need of any nucleation center. On the other hand, in the physical deposition

techniques employed here, a target material (oxide) is evaporated either by heating with an

electron beam (E-BEAM) or by a pulsed laser (PLD). One of the main advantages of these

techniques is that oxide layers with well controlled stoichiometry transfer from the target

material are expected [37]. The oxide develops on the substrate surface and on the flake

surface, by condensation. In principle, the adatom diffusivity on the substrate is determined

by the kinetic energy of arriving evaporated species and no chemical reaction/precursor

dissociation at the flake surface is required for the oxide to grow. Extra adatom mobility

energy can be added by heating the substrate. In this work, E-BEAM deposition has been

obtained for a substrate temperature of 200 ºC, whereas PLD has been performed with the

substrate at RT. The results show uniform conformal encapsulation by both E-BEAM and

PLD, as shown in Figure 1c-d. A similar surface roughness with a peak-to-peak roughness

value of ~3 nm is found for all the samples, independently of the coating oxide thickness,

except for the ALD case. A clear step of ∼0.7 nm, corresponding to the monolayer thickness

is measured in the profiles shown in Figure 1b-d.

To assess whether the flakes quality, composition or strain state are altered due to the

encapsulation process, we have performed micro-Raman spectroscopy on individual

monolayer flakes on selected samples. In particular, we focus on the vibrational modes E2g1

(in-plane) and A1g (out-of-plane), which are degenerate for monolayer WSe2, giving rise to a

single Raman peak at about ~ 250 cm-1 [43]. This peak is particularly sensitive to strain [44]

or doping in the flake [45]. Raman spectra collected in x(zz)x scattering geometry on

uncapped and encapsulated flakes, with a capping oxide layer thickness of 30 nm, are shown

in Figure 2a at wavenumbers around the E2g1/A1g modes. Solid lines represent fitting curves

11

Nano Res 12

(the procedure followed in the analysis is described in the supporting information). No shift is

observed after encapsulation within the system resolution (~0.5 cm-1), indicating that,

independent of the technique, the strain state and composition of the flakes are not

appreciably altered by the coating. We obtain consistently similar results for different oxide

thicknesses ranging from 5 to 30 nm. To further confirm our conclusions, XPS measurements

on uncapped and 5-nm-thick oxide encapsulated flakes are compared. Figure 2b shows the

XPS spectra for the 4f core-level signal of tungsten, where two main peaks attributed to the

spin-orbit components W 4f5/2 and W 4f7/2 (∆ = 2.17 eV) are resolved. Most importantly, for

all the samples but the one encapsulated by PLD, no WOx contribution to the overall spectra

at higher binding energies is detected, which confirms that no flake oxidation takes place

after encapsulation [27]. For PLD, it is found a small WOx component in the XPS spectrum

that we attribute to mixing and swallow implantation of Al2O3 species at the flake during

deposition inducing thus a slight oxidation [46,47]. By carefully analyzing the spectra for all

the studied flakes, we find asymmetries of the W 4f5/2 and W 4f7/2 peaks, suggesting the

presence of a second W 4f doublet shifted ~ 0.6 eV with respect to the main one (dashed

lines in Figure 2b). These features have been previously reported and attributed to spatial

variations in the Fermi level across the sample surface due to the presence of a mixture of van

der Waals planes and non-van der Waals planes characterized by edge planes and stepped

surfaces [48]. This is reasonable in our case, since the scanned area has a diameter around

400 µm, thus covering exfoliated flakes with different thicknesses and stepped surfaces. The

different relative spectral weight of both doublets in different samples is associated with the

uniqueness of each sample, which results from the fabrication process by mechanical

exfoliation. No relative peak shifts are found in encapsulated flakes with respect to the

uncapped flakes, ruling out significant compositional modification of the flakes upon

encapsulation.

12

Nano Res 13

3.2. PL on encapsulated WSe2 monolayers at RT.

In order to study the optical emission of individual encapsulated flakes, we have first

performed micro-PL measurements at RT. We emphasize that all the monolayers studied in

this work feature the same emission at =1.661 ± 0.005 eV before encapsulation, so that a

direct comparison between them is possible. Figure 3a displays the normalized PL spectra for

uncapped and encapsulated flakes. The black curve corresponds to the characteristic PL

emission for an uncapped WSe2 monolayer flake on a SiO2/Si substrate and will be used as a

reference. The spectrum is dominated by the neutral exciton (X0) transition at =1.661 eV,

marked with a dashed line.

Remarkably, no significant PL peak shift or line-width broadening is found for WSe2

monolayers encapsulated with stoichiometric Al2O3 oxides (E-BEAM and PLD). In

agreement with this observation, previous studies demonstrate that dielectric screening effects

on the optical transition energies of excitons in 2D materials surrounded by different

dielectric media are negligible, since the expected exciton binding energy reduction is nearly

compensated by a reduction of the free particle bandgap [49]. It is also remarkable that the

slight oxidation induced by the PLD coating has not a dramatic effect on the RT optical

response.

In contrast, for the flakes coated with sub-stoichiometric oxides Al2O2.85 (ALD) and SiO1.80

(PECVD), we observe a significant PL emission red-shift up to around 35 meV with respect

to the reference sample. Interestingly, for flakes coated by ALD, a gradual PL energy shift is

observed as the oxide thickness is increased (Figure 3b). A similar total shift is found for

PECVD encapsulation, but the shift is independent of the silicon-oxide layer thickness. The

shifts are accompanied by a PL peak broadening: the full-width-at-half-maximum (FWHM)

increases from ~40 meV for the uncapped sample up to ~60 and ~80 meV for PECVD and

ALD encapsulated samples, respectively.

13

Nano Res 14

It is known that sub-stoichiometric oxides with oxygen vacancies, such as SiO1.80 and

Al2O2.85 used in this work, can act as sources of carriers for the underlying flake [50,51].

Moreover, additional localized states inside the band-gap are expected for defective

amorphous oxides. Flake doping with electrons results in the gradual increase of the negative

trion (X-) emission compared to the neutral exciton, as reported for electrically gated 2D

materials, such as MoS2 [7], MoSe2 [52], WSe2 [8], or WS2 [53].

When the carrier concentration in the flake increases due to the electrical doping produced by

the sub-stoichiometric oxide, we expect a PL emission broadening due to the superposition of

both neutral and charged components. This is exactly the case for the ALD encapsulated

samples, where a gradual PL peak red-shift and broadening is observed by increasing the

flake coverage, i.e. electrical doping, as a consequence of the spectral weight change from

neutral exciton to trion (Figure S3 in the supporting information). The maximum PL energy

shift is ~35 meV, as expected for the trion relative binding energy in WSe2 monolayers

~30 meV [8,54]. A similar total shift is found for PECVD encapsulated flakes. However,

in this case, there is no smooth evolution, but rather a sharp change from a dominant neutral

exciton to a dominant trion transition independently of the capping layer thickness. This is

attributed to the uniform conformal flake encapsulation already achieved for the smallest

investigated coverage of 5 nm.

The relative integrated PL intensity at RT before and after encapsulation

( ) for all encapsulated monolayers is depicted in Figure 4 as a

function of the capping thickness. Average values ~ 0.74 ± 0.1 and ~ 0.59 ± 0.1 are

found for monolayers encapsulated by E-BEAM and PLD, respectively. In the case of

monolayers encapsulated with sub-stoichiometric oxides by ALD and PECVD, we would at

first expect a similar PL intensity for the X- peak compared to the X0 peak in undoped

monolayers, following the results obtained by gate-controlled-doping [8,52]. In contrast, we

14

Nano Res 15

find a significant PL decrease in doped monolayers. Since both Raman and XPS

characterizations demonstrate no flake damage upon encapsulation (within the resolution of

the techniques), the drop in PL intensity may be related to different mechanisms. We note that

the capping layer may also affect the excitation intensity and extraction efficiency due to

optical interference introduced by reflections at the additional interfaces. However, since the

capping layer thicknesses are much smaller than the wavelength of the emitted light, we also

expect the changes in extraction efficiency to be negligible compared to the observed

intensity changes from sample to sample (see Figure 4). One possibility is the development or

enhancement of non-radiative recombination channels for the excitons or photogenerated

carriers such as Auger processes, which are known to be particularly relevant for

heavily-doped semiconductors. On the other hand, depletion of photogenerated holes or

electrons through transfer mechanisms from the flake to traps present in the

sub-stoichiometric oxide cannot be ruled out as a possible explanation for both PL decrease

and electrical doping of the flake [55]. A qualitative correlation between oxide-induced

doping and PL intensity decrease is observed for ALD deposition. In this case the PL

intensity gradually decreases as the areal coverage of a flake (and hence the doping level)

increases. Similar results are obtained for monolayers encapsulated with stoichiometric SiO2

and sub-stoichiometric Al2O2.45 oxide layers deposited by E-BEAM as shown in the

supplementary information.

To further support the above interpretation about the gradual charging of the flake coated by

ALD, Figure 1e shows the topography and surface potential maps measured by Kelvin Force

Microscopy (KFM) [40] for a non-uniformly covered WSe2 flake by ALD with tnom=5 nm and

a reference uncapped flake. For this experiment, a 10-nm-thick Al2O2.85 layer is deposited by

ALD on the SiO2/Si substrate previous to the flakes transfer by mechanical exfoliation. The

surface potential difference between the surrounding aluminum-oxide and the uncapped

15

Nano Res 16

reference flake is |ΔV|~200 mV. Interestingly, the partially covered flake presents an evident

non-uniform surface potential distribution that closely replicates the topography of the coated

flake with oxide clusters. It should be noted that the same total difference |ΔV|~200 mV is

measured between the uncovered parts and the surrounding aluminum-oxide similarly to the

case of the uncapped flake. Most importantly, the uncovered and covered parts of the flake

show potential fluctuations of about 100 mV. This observation, together with the PL data at

RT reported above where an increasing electrical doping of the flake is revealed, suggests a

local electrical doping of the flake at the nanoscale delimited by the size of the

sub-stoichiometric oxide clusters. We point out here that the deliberate control of the

electrical doping of selected areas in 2D materials is fundamental to define p-n junctions for

applications in photovoltaics [56], light-emitting-diodes (LED) [57] or field-effect-transistors

(FET) [58]. The manipulation of the electrical doping in monolayers has been demonstrated

by different means, such as electrical gating [56,57], photoexcitation [59], self-limiting

oxidation of native WSe2 [60], photoinduced charge transfer [55,58,61], or surface

functionalization [62]. Although electrical n-type doping on WSe2 thin flakes was recently

shown by encapsulation with sub-stoichiometric SiNx layers over micrometer-sized large

areas [32], the results presented here at the nanoscale may find very interesting applications

in nanodevices.

3.3 Low temperature PL studies on encapsulated WSe2 monolayers.

Figure 5a shows characteristic power-dependent PL spectra (at 10 K) for an uncapped WSe2

monolayer. Figure 5b-e reports the spectra for flakes coated with 10 nm of oxide by the

indicated deposition methods. We begin the discussion with the uncapped reference flakes.

The overall PL emission consists of several peaks. The X0 (1.747 eV) and X- (1.714 eV) can

be well identified. There is some controversy regarding the origin of the emission at 1.696 eV.

While it has been experimentally attributed to the biexciton complex (XX) formed by two

16

Nano Res 17

bound excitons [10], recent calculations attribute this emission to an excited biexciton

complex (XX*) with relative binding energy with respect to X0 of about Eb,XX*=59 meV,

whereas a binding energy Eb,XX=20 meV is predicted for XX emission [54,63,64]. Because of

this ambiguity, we denote the transition at 1.696 eV as P1. On the other hand, relative binding

energies around 15 meV are predicted for exciton-trion (X0-X-) complexes, very close to the

predicted value for XX [63]. We denote the weak PL emission centered at 1.732 eV

(Eb~15 meV) with P0. The PL peaks L0, L1, L2 and L3 are attributed to localized exciton states

trapped in disordered potential wells that may be due to the underlying SiO2 dielectric [65,66].

The multi-peak PL spectrum is fitted by Gaussian curves (Figure S3 in supporting

information). We would like to note that very similar emission energies are found for all the

transitions on monolayers exfoliated on several SiO2/Si substrates, as shown in Figure S5a in

supporting information. The relative binding energies depicted in Figure S5b for each of the

transitions are comparable to other values reported throughout the literature. Although

additional investigations are needed for a deeper understating of the origin of the PL

transitions associated with localized states in the flakes, our observations suggest a stable

relative binding energy for these transitions with some fluctuations, such as linewidth

broadening. The latter might be attributed to an increase of the disorder in the confining

potential wells due to the presence of defects in the flakes or local fluctuations of the

underlying SiO2/Si substrate roughness.

The dependence of the PL intensity of semiconductors on the excitation power can be often

fitted with a power law [67]. The integrated intensity for each of the peaks is

calculated and fitted against the excitation power in a log-log plot to obtain the exponent

for uncapped and encapsulated flakes (Figure S6 and Table S3 in supporting information). In

the uncapped flake, the peaks X0, P0, X- and P1 show a nearly linear dependence with the

excitation power ( ) associated with the radiative recombination of free exciton complexes

17

Nano Res 18

that follows a first-order kinetic [68]. On the other hand, the peaks L0, L1, L2 and L3 present a

sub-linear dependence ( ), which is attributed to bound exciton recombination in

localized states [9]. Interestingly, a similar behavior is found for encapsulated samples with

stoichiometric oxide by E-BEAM and PLD, as reported in Figure 5b-c. It should be noted

that a broad defect-related bands are also observed for the encapsulated PLD sample labeled

as B0 and B1 in Figure 5c, consistent with the XPS data shown above, attributed to flake

oxidation and maybe to defect formation during deposition [47].

Encapsulated samples with sub-stoichiometric oxides by ALD and PECVD are notably

different from the uncapped flakes (Figure 5d-e). Remarkably, free X0 emission is not

observed and mostly bound exciton emission dominates the overall PL spectrum at high

excitation power. The energy position for PL peaks corresponding to X-, P1, L0 and L1 can be

well fitted at similar energies with respect to the uncapped sample. In particular, a power law

is found for these transitions. We attribute such a super-linear dependence to the

presence of strong non-radiative decay channels, such as Auger processes, where the carrier

recombination is no longer governed by a first-order kinetic [68]. This assumption is in line

with the observed PL integrated intensity decrease at RT for encapsulated samples by ALD

and PECVD shown in Figure 4. The observed broadening in the transitions can be due to an

increase in the non-uniformities of the disorder potential upon encapsulation induced by the

top oxide [65]. At the lowest excitation power values, only broad PL bands are observed:

PECVD – C0 (1.638 eV) and C1 (1.593 eV) – and ALD – D0 (1.638 eV) and D1 (1.601 eV) –.

The integrated PL intensity of these bands follows a linear power dependence and start to

saturate at excitation power values around 200 µW (Figure S6 in supporting information).

Since the quality of the flakes is well preserved upon encapsulation, as demonstrated by the

XPS and Raman spectroscopy studies, such broad PL bands are likely due to the introduction

of intra-gap energy levels upon encapsulation with the defective oxides and additional

18

Nano Res 19

confining potential wells for excitons induced by the encapsulation with the oxide.

Additionally, non-radiative recombination channels accounting for the super-linear power

dependence of the X-, P1, L0 and L1 transitions are present, as discussed above.

Figure 6a-c shows temperature-dependent PL spectra for uncapped and encapsulated samples

with stoichiometric oxides by E-BEAM and PLD. They are very similar and both show a

blueshift for X0, P0, and X- transitions as the samples are cooled down from 300 K to 10 K,

consistent with the Varshni law (Figure S7 in supporting information), which describes the

temperature dependence of the band-gap in semiconductors:

(1)

where is the energy bandgap at 0 K, while α’ and β are material related parameters.

The fitted parameters are shown in Table S4 of the supporting information and are

comparable to values previously reported for TMDCs [53]. At temperatures around 60 K, the

bound exciton states P1, L0, L1, L2 and L3 dissociate and consequently, the number of free X0

and X- increases. The X- PL intensity decreases for temperatures above 100 K, being the X0

transition dominant up to RT. Interestingly, we found that the emission related to localized

states (bands B0 and B1 in Figure 6c) observed in the PLD sample persists up to temperatures

as high as ~220 K, pointing at a different origin – i.e. flake partial oxidation – of the emission

compared to the P1, L0, L1, L2 and L3 transitions. The peak P0 can be well resolved for the

cases of the uncapped and E-BEAM encapsulated samples in the whole temperature range.

However, it cannot be fitted for the PLD encapsulated sample, due to strong PL background

induced by the broad PL band B0 that extends from about 1.54 eV to nearly 1.70 eV. The peak

P1 is either quenching or merging with the X- emission for temperatures above 130 K.

In order to isolate and investigate the low-energy bands for encapsulated samples with

sub-stoichiometric oxides by PECVD (C0 and C1) and ALD (D0 and D1), we have performed

temperature-dependent PL experiments for a low excitation power of 10 µW. Under these

19

Nano Res 20

experimental conditions, the low-energy broad bands dominate over the X- contribution, as

shown in Figure 6d-e. As expected, the X- can be well fitted with the Varshni equation. The

fitted parameters can be found in the supporting information. The broad PL bands undergo a

redshift as the temperature increases. While the C1/D1 bands are not resolved above 160 K,

the C0/D0 bands are stable up to RT, contributing to the emission broadening reported in

Figure 3 and Figure S3. We therefore attribute the PL emission broadening observed at RT to

an overall contribution of the X- emission and localized states induced by the encapsulation

procedure. The nature of these states is clearly different from the other localized states

discussed for uncapped and encapsulated flakes by E-BEAM/PLD, and might be ascribed to

trapped excitons in disordered potential minima produced by the underlying substrate and/or

the encapsulating oxide [65]. Additionally, such a disorder can be further enhanced by the

spatially random electrostatic landscape provided by the sub-stoichiometric encapsulating

oxide layer that remains charged upon carriers exchange with the flake.

4. Conclusions

In summary, we have performed a systematic study of the effects produced by oxide

encapsulation on the photoluminescence of exfoliated WSe2 monolayer flakes. We have

shown that full conformal encapsulation on freshly exfoliated flakes can be successfully

realized by PECVD, E-BEAM and PLD techniques. In contrast, oxide encapsulation by ALD

results in inhomogeneous formation of clusters, which coalesce as the thickness is increased.

The composition and strain state of the monolayers is well preserved upon encapsulation by

all the techniques. Only encapsulated flakes by PLD show slight WOx traces detectable by

XPS. Importantly, our results demonstrate that there is a high impact on the encapsulated

flakes optical response depending on the deposited oxide stoichiometry. First, the RT PL

properties of WSe2 monoloayer flakes are preserved upon encapsulation with a stoichiometric

oxide obtained by both physical deposition techniques E-BEAM and PLD. In contrast, the

20

Nano Res 21

emission of flakes capped with sub-stoichiometric oxides by ALD and PECVD is dominated

by trion emission, due to electrical doping through carriers exchange between the oxide and

the flake. Additionally, a strong PL intensity decrease is also found, which suggests the

presence of non-radiative mechanisms in the flake and/or a possible transfer of carriers from

the flake to the oxide. Second, the power-dependent PL studies at 10K and

temperature-dependent PL measurements show that encapsulated flakes with stoichiometric

Al2O3 exhibit dominant free and bound exciton recombination processes similarly to

uncapped flakes. The flakes encapsulated by PLD display an additional broad band PL

emission that might be due to the radiative recombination through defects originated in the

flake by the arrival of energetic species during deposition. However, the flakes encapsulated

with defective sub-stoichiometric oxides exhibit an emission consistent with the development

of a high density of localized states, which leads to a broad PL band at low energies that is

stable up to RT as discussed above. The origin of these localized states might be due to local

potential wells fluctuations induced by the spatially random distribution of charges in the

sub-stoichiometric oxide. From these results, we conclude that the most important factor for

achieving a successful WSe2 monolayer flakes encapsulation in terms of preserving their

optical response is the oxide quality, i.e. the stoichiometry. As a consequence, all the

deposition techniques used in this work have the potential to achieve a satisfactory oxide

coverage of the 2D flakes, as long as the experimental deposition conditions are modified in

order to achieve the correct stoichiometry and control the kinetic energy of the deposited

species, in the PLD case. We believe that our findings can be extended to other 2D

semiconductors and oxides/dielectrics, and that they will be of crucial relevance for the

design and optimization of advanced optoelectronic devices where the design of complex

semiconductor/dielectric nanostructures is mandatory. Moreover, the encapsulation procedure

with stoichiometric oxides presented here provides a robust approach to protect 2D materials

21

Nano Res 22

while preserving their optical properties. This will be particularly relevant for the transfer of

oxide-nanomembranes embedding 2D materials on different target substrates, which can be

for example accomplished using lift-off combined with dry transfer techniques. For instance,

we envisage that these oxide membranes could be integrated onto micro-machined

piezoelectric actuators to study with unprecedented detail the effect of large anisotropic strain

fields on 2D semiconductors [69,70].

Acknowledgements

The authors would like to thank Georgios Katsaros and Tim Wehling for valuable discussions.

Stephan Bräuer, Albin Schwarz and Ursula Kainz are acknowledged for technical support.

A.M. acknowledges the financial support through BES-2013-062593. GG acknowledges

support from FWF Project P 28018-B27. IZ acknowledges financial support from the Swiss

National Science Foundation research grant (Project Grant No. 200021_165784). This work

was partially funded by the Austrian Science Fund through the projects P24471 and P26830,

and by the Spanish Ministry for Economy and Competitiveness trough the project

MINECO/FEDER TEC2015-69916-C2-1-R.

Electronic Supplementary Material: Supplementary material (X-ray photoelectron

spectroscopy analysis of encapsulated WSe2 monolayers and oxides stoichiometry; detailed

analysis of the micro-Raman spectroscopy data; detailed analysis of the micro-PL

spectroscopy data, at RT and 10K; detailed analysis of the power-dependent and

temperature-dependent micro-PL spectroscopy data) is available in the online version of this

article at http: //dx.doi.org/10.1007/***********************).

References

[1] Liu, W.; Kang, J.; Sarkar, D.; Khatami, Y.; Jena, D.; Banerjee, K. Role of Metal

Contacts in Designing High-Performance Monolayer N-Type WSe2 Field Effect

22

Nano Res 23

Transistors. Nano Lett. 2013, 13, 1983–1990.

[2] Xu, X.; Yao, W.; Xiao, D.; Heinz, T. F. Spin and Pseudospins in Layered Transition

Metal Dichalcogenides. Nat. Phys. 2014, 10, 343–350.

[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, 2–5.

[4] Kobolov, K. S.; Tominaga, J. Two-Dimensional Transition Metal Dichalcogenides;

Springer Series in Materials Science: Switzerland, 2016.

[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] He, K.; Kumar, N.; Zhao, L.; Wang, Z.; Mak, K. F.; Zhao, H.; Shan, J. Tightly Bound

Excitons in Monolayer WSe2. Phys. Rev. Lett. 2014, 113, 1–5.

[7] 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.

[8] Jones, A. M.; Yu, H.; Ghimire, N. J.; Wu, S.; Aivazian, G.; Ross, J. S.; Zhao, B.; Yan,

J.; Mandrus, D. G.; Xiao, D.; et al. Optical Generation of Excitonic Valley Coherence

in Monolayer WSe2. Nat. Nanotechnol. 2013, 8, 634–638.

[9] Huang, J.; Hoang, T. B.; Mikkelsen, M. H. Probing the Origin of Excitonic States in

Monolayer WSe2. Sci. Rep. 2016, 6, 22414.

[10] You, Y.; Zhang, X. X.; Berkelbach, T. C.; Hybertsen, M. S.; Reichman, D. R.; Heinz,

T. F. Observation of Biexcitons in Monolayer WSe2. Nat. Phys. 2015, 11, 477–482.

[11] Arora, A.; Koperski, M.; Nogajewski, K.; Marcus, J.; Faugeras, C.; Potemski, M.

Excitonic Resonances in Thin Films of WSe2 : From Monolayer to Bulk Material.

Nanoscale 2015, 7, 10421–10429.

[12] Choi, J.; Zhang, H.; Du, H.; Choi, J. H. Understanding Solvent Effects on the

23

Nano Res 24

Properties of Two-Dimensional Transition Metal Dichalcogenides. ACS Appl. Mater.

Interfaces 2016, 8, 8864–8869.

[13] 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.

[14] Yu, Y.; Yu, Y.; Xu, C.; Cai, Y. Q.; Su, L.; Zhang, Y.; Zhang, Y. W.; Gundogdu, K.; Cao,

L. Engineering Substrate Interactions for High Luminescence Efficiency of

Transition-Metal Dichalcogenide Monolayers. Adv. Funct. Mater. 2016, 26,

4733–4739.

[15] Buscema, M.; Steele, G. A.; van der Zant, H. S. J.; Castellanos-Gomez, A. The Effect

of the Substrate on the Raman and Photoluminescence Emission of Single-Layer MoS2.

Nano Res. 2014, 7, 561–571.

[16] Shi, H.; Yan, R.; Bertolazzi, S.; Brivio, J.; Gao, B.; Kis, A.; Jena, D.; Xing, H. G.;

Huang, L. Exciton Dynamics in Suspended Monolayer and Few-Layer MoS2 2D

Crystals. ACS Nano 2013, 7, 1072–1080.

[17] Mag-isa, A.; Kim, J.; Lee, H.; Oh, C. A Systematic Exfoliation Technique for

Isolating Large and Pristine Samples of 2D Materials. 2D Mater. 2015, 2, 34017.

[18] Castellanos-Gomez, A.; Buscema, M.; Molenaar, R.; Singh, V.; Janssen, L.; van der

Zant, H. S. J.; Steele, G. a. Deterministic Transfer of Two-Dimensional Materials by

All-Dry Viscoelastic Stamping. 2D Mater. 2014, 1, 11002

[19] Yan, T.; Qiao, X.; Liu, X.; Tan, P.; Zhang, X. Photoluminescence Properties and

Exciton Dynamics in Monolayer WSe2. Appl. Phys. Lett. 2014, 105, 101901.

[20] Chen, X.; Wu, Y.; Wu, Z.; Han, Y.; Xu, S.; Wang, L.; Ye, W.; Han, T.; He, Y.; Cai, Y.;

et al. High-Quality Sandwiched Black Phosphorus Heterostructure and Its Quantum

24

Nano Res 25

Oscillations. Nat. Commun. 2015, 6, 7315.

[21] Wood, J. D.; Wells, S. A.; Jariwala, D.; Chen, K. S.; Cho, E.; Sangwan, V. K.; Liu, X.;

Lauhon, L. J.; Marks, T. J.; Hersam, M. C. Effective Passivation of Exfoliated Black

Phosphorus Transistors against Ambient Degradation. Nano Lett. 2014, 14,

6964–6970.

[22] Jena, D.; Konar, A. Enhancement of Carrier Mobility in Semiconductor

Nanostructures by Dielectric Engineering. Phys. Rev. Lett. 2007, 98, 136805.

[23] Kufer, D.; Konstantatos, G. Highly Sensitive, Encapsulated MoS2 Photodetector with

Gate Controllable Gain and Speed. Nano Lett. 2015, 15, 7307–7313.

[24] Park, J. H.; Fathipour, S.; Kwak, I.; Sardashti, K.; Ahles, C. F.; Wolf, S. F.; Edmonds,

M.; Vishwanath, S.; Xing, H. G.; Fullerton-Shirey, S. K.; et al. Atomic Layer

Deposition of Al2O3 on WSe2 Functionalized by Titanyl Phthalocyanine. ACS Nano

2016, 10, 6888-6896.

[25] Huber, M. A.; Mooshammer, F.; Plankl, M.; Viti, L.; Sandner, F.; Kastner, L. Z.;

Frank, T.; Fabian, J.; Vitiello, M. S.; Cocker, T. L.; et al. Femtosecond

Photo-Switching of Interface Polaritons in Black Phosphorus Heterostructures. Nat

Nanotechnol. 2016, 12, 207-211.

[26] Cheng, L.; Qin, X.; Lucero, A. T.; Azcatl, A.; Huang, J.; Wallace, R. M.; Cho, K.;

Kim, J. Atomic Layer Deposition of a High-K Dielectric on MoS2 Using

Trimethylaluminum and Ozone. ACS Appl. Mater. Interfaces 2014, 6, 11834–11838.

[27] Azcatl, A.; Kc, S.; Peng, X.; Lu, N.; McDonnell, S.; Qin, X.; de Dios, F.; Addou, R.;

Kim, J.; Kim, M. J.; et al. HfO2 on UV–O3 Exposed Transition Metal Dichalcogenides:

Interfacial Reactions Study. 2D Mater. 2015, 2, 14004.

[28] Yang, W.; Sun, Q.-Q.; Geng, Y.; Chen, L.; Zhou, P.; Ding, S.-J.; Zhang, D. W. The

25

Nano Res 26

Integration of Sub-10 Nm Gate Oxide on MoS2 with Ultra Low Leakage and Enhanced

Mobility. Sci. Rep. 2015, 5, 11921.

[29] Das, S.; Chen, H.-Y.; Penumatcha, A. V.; Appenzeller, J. High Performance

Multilayer MoS2 Transistors with Scandium Contacts. Nano Lett. 2013, 13, 100–105.

[30] Radisavljevic, B.; Radenovic, A.; Brivio, J.; Giacometti, V.; Kis, A. Single-Layer

MoS2 Transistors. Nat. Nanotechnol. 2011, 6, 147–150.

[31] Late, D. J.; Liu, B.; Matte, H. S. S. R.; Dravid, V. P.; Rao, C. N. R. Hysteresis in

Single-Layer MoS2 Field Effect Transistors. ACS Nano 2012, 6, 5635–5641.

[32] Chen, K.; Kiriya, D.; Hettick, M.; Tosun, M.; Ha, T.-J.; Madhvapathy, S. R.; Desai, S.;

Sachid, A.; Javey, A. Air Stable N-Doping of WSe2 by Silicon Nitride Thin Films with

Tunable Fixed Charge Density. APL Mater. 2014, 2, 92504.

[33] Sercombe, D.; Schwarz, S.; Del Pozo-Zamudio, O.; Liu, F.; Robinson, B. J.;

Chekhovich, E. a; Tartakovskii, I. I.; Kolosov, O.; Tartakovskii, a I. Optical

Investigation of the Natural Electron Doping in Thin MoS2 Films Deposited on

Dielectric Substrates. Sci. Rep. 2013, 3, 3489.

[34] Plechinger, G.; Schrettenbrunner, F. X.; Eroms, J.; Weiss, D.; Schüller, C.; Korn, T.

Low-Temperature Photoluminescence of Oxide-Covered Single-Layer MoS2. Phys.

Status Solidi - Rapid Res. Lett. 2012, 6, 126–128.

[35] Bhanu, U.; Islam, M. R.; Tetard, L.; Khondaker, S. I. Photoluminescence Quenching

in Gold - MoS2 Hybrid Nanoflakes. Sci. Rep. 2014, 4, 5575.

[36] Lin, Y.; Ling, X.; Yu, L.; Huang, S.; Hsu, A. L.; Lee, Y. H.; Kong, J.; Dresselhaus, M.

S.; Palacios, T. Dielectric Screening of Excitons and Trions in Single-Layer MoS2.

Nano Lett. 2014, 14, 5569–5576.

[37] Eason, R. W. Pulsed Laser Deposition of Thin Films: Applications-Led Growth of

Functional Materials; John Wiley & Sons, Inc.: Hoboken, New Jersey, 2006.

26

Nano Res 27

[38] Toftmann, B.; Schou, J.; Hansen, T.; Lunney, J. Angular Distribution of Electron

Temperature and Density in a Laser-Ablation Plume. Phys. Rev. Lett. 2000, 84,

3998–4001.

[39] Serna, R.; Nuñez-Sanchez, S.; Xu, F.; Afonso, C. N. Enhanced Photoluminescence of

Rare-Earth Doped Films Prepared by off-Axis Pulsed Laser Deposition. Applied

Surface Science 2011, 257, 5204–5207.

[40] Nonnenmacher, M.; O’Boyle, M. P.; Wickramasinghe, H. K. Kelvin Probe Force

Microscopy. Appl. Phys. Lett. 1991, 58, 2921–2923.

[41] Stark, R. W.; Naujoks, N.; Stemmer, A. Multifrequency Electrostatic Force Microscopy

in the Repulsive Regime. Nanotechnology 2007, 18, 065502.

[42] Horcas, I.; Fernández, R.; Gómez-Rodríguez, J. M.; Colchero, J.; Gómez-Herrero, J.;

Baro, A. M. WSXM: A Software for Scanning Probe Microscopy and a Tool for

Nanotechnology. Rev. Sci. Instrum. 2007, 78, 013705.

[43] Zhao, W.; Ghorannevis, Z.; Amara, K.; Pang, J. Lattice Dynamics in Mono-and

Few-Layer Sheets of WS2 and WSe2. Nanoscale 2013, 5, 9677–9683.

[44] Rice, C.; Young, R. J.; Zan, R.; Bangert, U.; Wolverson, D.; Georgiou, T.; Jalil, R.;

Novoselov, K. S. Raman-Scattering Measurements and First-Principles Calculations of

Strain-Induced Phonon Shifts in Monolayer MoS2. Phys. Rev. B - Condens. Matter

Mater. Phys. 2013, 87, 081307(R).

[45] 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(R).

[46] Barnes, J. P.; Beer, N.; Petford-Long, A. K.; Suarez-Garcia, A.; Serna, R.; Hole, D.;

Weyland, M.; Midgley, P. A. Resputtering and Morphological Changes of Au

Nanoparticles in Nanocomposites as a Function of the Deposition Conditions of the

27

Nano Res 28

Oxide Capping Layer. Nanotechnology 2005, 16, 718–723.

[47] Barnes, J. P.; Petford-Long, A. K.; Suárez-García, A.; Serna, R. Evidence for Shallow

Implantation during the Growth of Bismuth Nanocrystals by Pulsed Laser Deposition.

J. Appl. Phys. 2003, 93, 6396–6398.

[48] Klein, a.; Tomm, Y.; Schlaf, R.; Pettenkofer, C.; Jaegermann, W.; Lux-Steiner, M.;

Bucher, E. Photovoltaic Properties of WSe2 Single-Crystals Studied by Photoelectron

Spectroscopy. Sol. Energy Mater. Sol. Cells 1998, 51, 181–191.

[49] Stier, A. V; Wilson, N. P.; Clark, G.; Xu, X.; Crooker, S. A. Probing the Influence of

Dielectric Environment on Excitons in Monolayer WSe2: Insight from High Magnetic

Fields. Nano Lett. 2016, 16, 7054–7060.

[50] Griscom, D. L. Defect Structure of Glasses. J. Non. Cryst. Solids 1985, 73, 51–77.

[51] Liu, D.; Clark, S. J.; Robertson, J. Oxygen Vacancy Levels and Electron Transport in

Al2O3. Appl. Phys. Lett. 2010, 32905, 2008–2011.

[52] Ross, J. S.; Wu, S.; Yu, H.; Ghimire, N. J.; Jones, A. M.; Aivazian, G.; Yan, J.;

Mandrus, D. G.; Xiao, D.; Yao, W.; et al. Electrical Control of Neutral and Charged

Excitons in a Monolayer Semiconductor. Nat. Commun. 2013, 4, 1474.

[53] Plechinger, G.; Nagler, P.; Kraus, J.; Paradiso, N.; Strunk, C.; Schüller, C.; Korn, T.

Identification of Excitons, Trions and Biexcitons in Single-Layer WS2. Phys. Status

Solidi - Rapid Res. Lett. 2015, 9, 457–461.

[54] Zhang, D. K.; Kidd, D. W.; Varga, K. Excited Biexcitons in Transition Metal

Dichalcogenides. Nano Lett. 2015, 15, 7002–7005.

[55] Choi, J.; Zhang, H.; Choi, J. H. Modulating Optoelectronic Properties of

Two-Dimensional Transition Metal Dichalcogenide Semiconductors by Photoinduced

Charge Transfer. ACS Nano 2015, 10, 1671-1680.

[56] Pospischil, A.; Furchi, M. M.; Mueller, T. Solar-Energy Conversion and Light

28

Nano Res 29

Emission in an Atomic Monolayer P-N Diode. Nat Nanotechnol. 2014, 9, 257–261.

[57] Ross, J. S.; Klement, P.; Jones, A. M.; Ghimire, N. J.; Yan, J.; G., M.; Taniguchi, T.;

Watanabe, K.; Kitamura, K.; Yao, W.; et al. Electrically Tunable Excitonic

Light-Emitting Diodes Based on Monolayer WSe2 P-N Junctions. Nat Nano 2014, 9,

268–272.

[58] Fang, H.; Chuang, S.; Chang, T. C.; Takei, K.; Takahashi, T.; Javey, A.

High-Performance Single Layered WSe2 P-FETs with Chemically Doped Contacts.

Nano Lett. 2012, 12, 3788–3792.

[59] Mitioglu, A. A.; Plochocka, P.; Jadczak, J. N.; Escoffier, W.; Rikken, G. L. J. A.;

Kulyuk, L.; Maude, D. K. Optical Manipulation of the Exciton Charge State in

Single-Layer Tungsten Disulfide. Phys. Rev. B 2013, 88, 245403.

[60] Yamamoto, M.; Nakaharai, S.; Ueno, K.; Tsukagoshi, K. Self-Limiting Oxides on

WSe2 as Controlled Surface Acceptors and Low-Resistance Hole Contacts. Nano Lett.

2016, 16, 2720-2727.

[61] Mouri, S.; Miyauchi, Y.; Matsuda, K. Tunable Photoluminescence of Monolayer

MoS2 via Chemical Doping. Nano Lett. 2013, 13, 5944–5948.

[62] Lin, J. D.; Han, C.; Wang, F.; Wang, R.; Xiang, D.; Qin, S.; Zhang, X. A.; Wang, L.;

Zhang, H.; Wee, A. T. S.; et al. Electron-Doping-Enhanced Trion Formation in

Monolayer Molybdenum Disulfide Functionalized with Cesium Carbonate. ACS Nano

2014, 8, 5323–5329.

[63] Kylänpää, I.; Komsa, H. P. Binding Energies of Exciton Complexes in Transition

Metal Dichalcogenide Monolayers and Effect of Dielectric Environment. Phys. Rev. B

- Condens. Matter Mater. Phys. 2015, 92, 205418.

[64] Mayers, M. Z.; Berkelbach, T. C.; Hybertsen, M. S.; Reichman, D. R. Binding

Energies and Spatial Structures of Small Carrier Complexes in Monolayer

29

Nano Res 30

Transition-Metal Dichalcogenides via Diffusion Monte Carlo. Phys. Rev. B - Condens.

Matter Mater. Phys. 2015, 92, 161404(R).

[65] Hichri, A.; Amara, I. Ben; Ayari, S.; Jaziri, S. Exciton, Trion and Localized Exciton in

Monolayer Tungsten Disulfide. 2016, arXiv:1609.05634v1

[66] Wang, G.; Bouet, L.; Lagarde, D.; Vidal, M.; Balocchi, A.; Amand, T.; Marie, X.;

Urbaszek, B. Valley Dynamics Probed through Charged and Neutral Exciton Emission

in Monolayer WSe2. Phys. Rev. B - Condens. Matter Mater. Phys. 2014, 90, 075413.

[67] Schmidt, T.; Lischka, K.; Zulehner, W. Excitation-Power Dependence of the

near-Band-Edge Photoluminescence of Semiconductors. Phys. Rev. B 1992, 45,

8989–8994.

[68] Chiari, A.; Colocci, M.; Fermi, F.; Li, Y. H.; Querzoli, R.; Vinattieri, A.; Zhuang, W.

H. Temperature-Dependence of the Photoluminescence in GaAs-GaAlAs Multiple

Quantum Well Structures. Phys. Status Solidi 1988, 147, 421–429.

[69] Martín-Sánchez, J.; Trotta, R.; Piredda, G.; Schimpf, C.; Trevisi, G.; Seravalli, L.;

Frigeri, P.; Stroj, S.; Lettner, T.; Reindl, M.; et al. Reversible Control of In-Plane

Elastic Stress Tensor in Nanomembranes. Adv. Opt. Mater. 2016, 4, 682–687.

[70] Trotta, R.; Martín-Sánchez, J.; Wildmann, J. S.; Piredda, G.; Reindl, M.; Schimpf, C.;

Zallo, E.; Stroj, S.; Edlinger, J.; Rastelli, A. Wavelength-Tunable Sources of Entangled

Photons Interfaced with Atomic Vapours. Nat. Commun. 2016, 7, 10375.

Figure captions

Figure 1.- (a) AFM pictures of encapsulated WSe2 monolayer flakes by ALD (Al2O2.85) for a

nominal oxide thickness t=5 nm, t=10 nm and t=30 nm. The line-scans over the AFM pictures

are also plotted. The observed morphology corresponds to a cluster-like oxide coating of the

30

Nano Res 31

flake due the lack of dangling bonds on the flakes' surface. (b-d) AFM pictures of

encapsulated monolayers for an oxide thickness t=10 nm by PECVD (SiO1.80), E-BEAM

(Al2O3) and PLD (Al2O3). A uniform and conformal encapsulation is evidenced.

Representative topography profiles are drawn over the images, where a step of about 0.7 nm

corresponding to a monolayer thickness is measured for conformally encapsulated flakes.

Full oxide coating of the monolayer is observed for a thickness t=30 nm in the case of ALD.

(e) Comparative topography and surface potential maps by KFM on a 5-nm-thick

encapsulated monolayer by ALD and uncapped reference flake. A clear relation is found

between the sub-stoichiometric Al2O2.85 oxide clusters and the static charges distribution over

the monolayer.

Figure 2.- (a) Stacked Raman spectra (open symbols) collected in the x(zz)-x scattering geometry on

uncapped (squares) and encapsulated WSe2 monolayers (t=30 nm) by E-BEAM (circles), ALD (triangles),

PECVD (diamonds) and PLD (stars). The dashed line marks the frequency of the degenerate E12g and A1g

modes distinctive of monolayers. Solid lines represent the best fits to the spectra. (b) XPS spectra of

uncapped and encapsulated monolayers (t=5 nm) by E-BEAM, ALD, PECVD and PLD. The W 4f

core-level signal with W 4f5/2 and 4f7/2 components are resolved together with a W 5p level contribution at

a lower binding energy. An additional WOx component is found for the encapsulated flake by PLD,

indicating partial flake oxidation.

Figure 3.- (a) Normalized micro-PL spectra at RT of uncapped (black curve) and encapsulated WSe2

monolayers for different capping layer thicknesses and varying oxide stoichiometry. Oxides are produced

by ALD (Al2O2.85), PECVD (SiO1.80), PLD (Al2O3) and E-BEAM (Al2O3). (b) Corresponding FWHM and

relative PL peak energy shift values for all the samples. The values for the uncapped reference monolayer

are marked with a horizontal dashed line.

Figure 4.- Relative integrated PL intensity at RT of encapsulated WSe2 monolayers for different capping

31

Nano Res 32

layer thicknesses (t=5, 10, 20 and 30 nm) with respect to uncapped monolayers

( ). Oxides are produced by ALD (Al2O2.85), PECVD (SiO1.80), PLD (Al2O3) and

E-BEAM (Al2O3). An excitation power of 10 µW is used.

Figure 5.- Power-dependent PL spectra at 10 K for an uncapped WSe2 monolayer (a), and encapsulated

(t=10 nm) monolayers by E-BEAM (b), PLD (c), PECVD (d) and ALD (e). The energy position for X0, P0,

P1 and X- free exciton, and L0, L1, L2 and L3 bound exciton recombinations are marked with dashed

vertical lines. Additional PL bands are indicated for encapsulated monolayers by PLD (B0 and B1), PECVD

(C0 and C1) and ALD (D0 and D1).

Figure 6.- Temperature-dependent PL spectra at an excitation power of 10 µW, for uncapped (a) and

encapsulated WSe2 monolayers with stoichiometric oxides by E-BEAM (b) and PLD (c). The

temperature-dependent PL spectra for encapsulated monolayers with sub-stoichiometric oxides by ALD

and PECVD are shown in (d) and (e), respectively. The dashed lines serve as a guide for the eye to follow

the energy shift of the PL components with temperature.

32

Nano Res 33

Figures

33

Nano Res 34

Figure 1

Figure 2

34

Nano Res 35

Figure 3

35

Nano Res 36

Figure 4

Figure 5

36

Nano Res 37

Figure 6

Supplementary Information

Effects of Dielectric Stoichiometry on the Photoluminescence Properties of Encapsulated WSe2

Monolayers

Javier Martín-Sánchez1*, Antonio Mariscal2, Marta De Luca3, Aitana Tarazaga

Martín-Luengo1, Georg Gramse4, Alma Halilovic1, Rosalía Serna2, Alberta Bonanni1, Ilaria

Zardo3, Rinaldo Trotta1*, and Armando Rastelli1

37

Nano Res 38

1 Institute of Semiconductor and Solid State Physics, Johannes Kepler University, Altenbergerstrasse

69, A-4040 Linz, Austria

2 Laser Processing Group, Instituto de Óptica, CSIC, C/Serrano 121, 28006 Madrid, Spain

3Department of Physics, University of Basel, Klingelbergstrasse 82, 4056, Basel, Switzerland

4Johannes Kepler University of Linz, Institute for Biophysics, Gruberstrasse 40, A-4020 Linz, Austria

KEYWORDS: dielectric encapsulation, 2D materials, transition metal dichalcogenides,

semiconductors, photoluminescence, WSe2

* e-mail: javier.martin.nano@gmail.com ; rinaldo.trotta@jku.at

Encapsulating oxides characterization by x-ray photoelectron spectroscopy (XPS).

The stoichiometry of the encapsulating oxides AlxOy and SiOx deposited by electron-beam (E-BEAM),

pulsed laser deposition (PLD), atomic layer deposition (ALD) and plasma-enhanced chemical vapor

deposition (PECVD) is evaluated by x-ray photoelectron spectroscopy (XPS). All binding energies (BEs)

are calibrated with respect to the conventional C 1s peak at 284.8 eV. The corresponding XPS experimental

data are fitted by using 30% Lorentzian and 70% Gaussian curves, and the smart background provided by

the Avantage software (Thermo Scientific™), based on the Shirley background with the additional

constraint that the background should not be of a higher intensity than the actual data at any point in the

region, as shown in Figure S1.

38

Nano Res 39

Figure S1.- XPS spectra with the corresponding fits for the encapsulating oxides deposited by E-BEAM (a), PLD (b), ALD (c) and PECVD (d). The data fitting and analysis indicate that oxides deposited by E-BEAM and PLD are stoichiometric (Al2O3). Non-stoichiometric compositions are deduced for oxides deposited by ALD (Al2O2.85) and PECVD (SiO1.80), respectively.

The stoichiometry of the oxides is calculated from the peak areas of the fitted curves from the core levels,

with sensitivity factors taken from the tabulated libraries in the Avantage software. The quantification

routine automatically applies the correct instrument transmission function to the data. In Table S1,

estimated atomic percentage, binding energy, and full-width-at-half-maximum (FWHM) values are listed

for all the oxides under study.

Oxide layer Peak

assignment

At.% Peak

BE

(eV)

FWHM

(eV)

E-BEAM / Al2O3

O 1s (Al-oxide) 53.42 531.29 1.98

O 1s (OH) 11.23 532.96 1.84

Al 2p 35.36 74.62 1.61

39

Nano Res 40

PLD / Al2O3

O 1s (Al-oxide) 39.47 531.35 1.52

O 1s (OH) 16.46 532.42 1.32

O 1s (WOx) 17.67 530.20 1.5

Al 2p 26.40 74.67 1.59

ALD / Al2O2.85

O 1s (Al-oxide) 52.37 531.27 1.94

O 1s (OH) 10.94 532.99 1.93

Al 2p 36.68 74.38 1.6

PECVD / SiO1.80

O 1s 64.32 533.11 1.57

Si 2p 35.68 103.56 1.55

Table S1.- Atomic percentage, binding energy and FWHM values for each of the fitted curves in Figure S1 for uncapped and encapsulated flakes by E-BEAM, PLD, ALD and PECVD.

For all the aluminum-oxide layers, a second peak at BEs higher than the main Al-oxide peak in the O 1s

spectrum is found, assigned to –OH species due to atmospheric exposure of the films, in accordance with

literature [1–3]. A third oxygen peak, at lower BEs than the main Al-oxide peak, is found for the PLD

grown oxide layer, which is assigned to oxygen bonds with tungsten (WOx), in agreement with the WOx

contribution found for the same deposition technique in the W 4f high resolution spectra (Figure 2b in the

main text). The measurements reveal two stoichiometric O:Al ratios (~1.5) for the oxides deposited by

E-BEAM and PLD, and two non-stoichiometric oxides, one corresponding to the ALD technique, with

O:Al ratio of about 1.43, and a second one for the SiOx layer grown by PECVD, with a O:Si ratio of about

1.80.

Micro-Raman spectroscopy of uncapped and encapsulated WSe2 monolayers.

Micro-Raman measurements on uncapped and selected encapsulated WSe2 monolayers are shown in

40

Nano Res 41

Figure 2a in the main text. We have measured all the samples in x(zz)-x and x(zy)-x configurations and we

have found in both configurations a single Raman peak at about ~ 250 cm-1 arising from the degenerate

E2g1 and A1g modes, as expected for monolayers. All measurements are performed under the same

conditions, i.e. same power and polarization. The multi-peak Raman spectra are fitted with Lorentzian

curves, as displayed in Figure S2a for the case of a representative sample (ALD, t=30 nm). This spectrum

is the same shown in Figure 2a, but here it is displayed in a broader range in order to highlight the fit

procedure. The broad band peaked at about 261 cm-1 is usually attributed to a 2LA (M) mode [4]. The thick

solid line represents the fit curve obtained by deconvoluting the whole spectrum with Lorentzian curves

(thin lines). We have performed the characterization on at least two different points on the flakes with the

same results.

In Figure S2b, we plot the frequency of the E2g1/A1g peak as obtained by Lorentzian fits as a function of the

capping thickness for two representative samples, E-BEAM (circles) and PLD (stars), and the uncapped

reference sample (squares). The green rectangle is 0.5 cm-1 high and indicated the spectral resolution of the

employed setup. No clear trend with the capping thickness is found but a fluctuation of the peak frequency

independently of the sample. Most importantly, such fluctuation is within the system spectral resolution.

Therefore, Raman measurements indicate no flake damage, oxidation, or strain induced by the

encapsulation procedure within the resolution of the technique.

0 5 10 20 30250.0

250.5

251.0

251.5

245 250 255 260 265 270

(b)

Inte

nsity

(arb

. uni

ts)

Raman shift (cm-1)

A1g/E12g

(a) uncapped EBEAM PLD

Freq

uenc

y (c

m-1)

Capping layer thickness (nm)

Figure S2.- (a) Raman spectrum (open circles) of ALD-encapsulated monolayer deconvoluted with

41

Nano Res 42

Lorentzian curves (thin lines). Fit result is shown by a thick line. Dashed line marks the degenerate E2g

1/A1g modes. (b) Frequency of the degenerate E2g1/A1g modes (as determined by Lorentzian fits to

Raman spectra) as a function of the oxide capping layer thickness. Colored area (0.5 cm-1 high) indicates the spectral resolution.

Micro-photoluminescence (micro-PL) at room-temperature (RT) of WSe2 monolayers coated by

ALD.

Figure S3 shows the normalized micro-PL spectra of WSe2 monolayers encapsulated by ALD with

non-stoichiometric Al2O2.85 oxide for different encapsulating oxide thicknesses: 0, 5, 7.5, 10 and 30 nm. A

gradual spectral weight change in the PL spectra from dominant neutral exciton ( =1.661 eV, green

curve) to trion ( =1.626 eV, blue curve) is observed as the capping oxide layer thickness is increased. As

mentioned in the main text of the manuscript, this is related to a gradual charging effect in the monolayer

flake due to electrons transfer from the overlying non-stoichiometric oxide layer. The gradually increasing

charging of the flake is attributed to the cluster-like growth of the capping oxide, as shown in the AFM

pictures in Figure 1 of the main manuscript, i.e. increasing charging as the clusters coalescence. An

additional low-energy band attributed to the development of localized states due to the sub-stoichiometric

oxide is resolved for the thicker oxide layers with nominal thicknesses from 7.5 to 30 nm.

42

Nano Res 43

Figure S3.- Micro-PL spectra (black lines) and fits (red lines) for encapsulated WSe2 monolayers with non-stoichiometric Al2O2.85 oxide by ALD with different oxide layer thicknesses: 0,5, 7.5, 10 and 30 nm. The PL spectra are fitted with Gaussian curves corresponding to: neutral exciton X0 (green), trion (blue) and localized-states-related band (magenta).

Micro-PL of WSe2 monolayers at 10 K: multi-peak Gaussian fitting.

The multi-peak PL spectrum of a WSe2 monolayer flake with and without oxide coating is fitted with

Gaussian curves as shown in Figure S4. We identify the X0 and X- transitions at energies around 1.747 ±

0.001 eV and 1.715 ± 0.001 eV, respectively. The peaks P0 and P1 are fitted at 1.734 ± 0.003 eV and 1.695

± 0.002 eV, respectively. The other transitions corresponding to localized states are at energies 1.683 ±

0.002 eV (L0), 1.666 ± 0.02 eV (L1), 1.646 ± 0.03 eV (L2) and 1.633 ± 0.005 eV (L3). A similar fitting is

performed for the encapsulated flakes by E-BEAM. In the case of the monolayer coated by PLD, the peaks

X0, X-, P1 , L0 and L1 are well fitted with two additional broad PL bands B0 (1.637 eV) and B1 (1.560 eV).

43

Nano Res 44

The fitting to the PL multi-peak emission for flakes encapsulated by ALD and PECVD is obtained with

similar peak positions for X0, P0, X-, P1, L0 and L1 with respect to the uncapped flake. The transitions L2

and L3 cannot be resolved for the PLD, ALD and PECVD cases due to the evolution of broad PL bands at

similar energies. In the case of ALD encapsulation, two additional PL bands D0 (1.638 eV) and D1 (1.601

eV) are included in the fit. In the case of PECVD encapsulation, two additional PL bands C0 (1.638 eV)

and C1 (1.593 eV) are included in the fit. A significant broadening of the X-, P1, L0 and L1 transitions is

observed for monolayers encapsulated by ALD and PECVD with respect to the uncapped monolayer. The

corresponding peak positions and FWHM values for the fitted Gaussian curves are reported in Table S2.

Uncapped E-BEAM PLD ALD PECVD

Peak

linewidth

(meV)

Peak

position

(eV)

Peak

linewidth

(meV)

Peak

position

(eV)

Peak

linewidth

(meV)

Peak

position

(eV)

Peak

linewidth

(meV)

Peak

position

(eV)

Peak

linewidth

(meV)

Peak

position

(eV)

X0 11.4 1.747 11.4 1.748 16.2 1.745 -- --

P0 11.0 1.732 11.7 1.735 -- -- 37.3 1.731 22.7 1.730

X- 13.8 1.714 15.0 1.716 16.3 1.715 26.9 1.715 20.3 1.715

P1 11.8 1.696 13.8 1.696 12.9 1.683 20.0 1.670 20.9 1.699

L0 9.6 1.683 10.6 1.684 11.5 1.696 21.4 1.681 22.9 1.683

L1 19.4 1.664 24.2 1.664 15.7 1.662 25.6 1.665 33.1 1.667

L2 14.2 1.642 11.3 1.639 -- -- -- -- -- --

L3 33.7 1.631 33.0 1.626 -- -- -- -- -- --

B0 -- -- -- -- 76.1 1.637 -- -- -- --

44

Nano Res 45

B1 -- -- -- -- 90.8 1.560 -- -- -- --

C0 -- -- -- -- -- -- -- -- 45.0 1.638

C1 -- -- -- -- -- -- -- -- 56.0 1.593

D0 -- -- -- -- -- -- 39.6 1.638 -- --

D1 -- -- -- -- -- -- 55.0 1.601 -- --

Table S2.- Energy positions and linewidth for X0, P0, X-, P1, L0, L1, L2 and L3 transitions, and B0/B1, C0/C1 and D0/D1 PL bands, extracted from the fitted Gaussian curves to the PL spectra in Figure S4.

45

Nano Res 46

Figure S4.- Low-temperature micro-PL spectra for uncapped WSe2 monolayers (a) and encapsulated monolayers by: (b) E-BEAM; (c) PLD; (d) PECVD; (e) ALD. The PL emission is fitted using multiple Gaussian curves.

Micro-PL of uncapped WSe2 monolayers at 10K.

The micro-PL spectra for several as-exfoliated WSe2 monolayers are reported in Figure S5a. Interestingly,

although there is some scatter in the values for the linewidth of each of the transitions, similar relative

binding energies with respect to X0 are found for all the transitions as can be seen in Figure S5b. In Figure

S5b, we point out that similar values are obtained in comparison with reported data in references [5-9].

46

Nano Res 47

Figure S5.- (a) Micro-PL spectra for uncapped WSe2 monolayers fitted with Gaussian curves. (b) Relative binding energies for all the transitions with respect to the X0 peak for the monolayers reported on (a). Relative binding energy values extracted from references [[5-9]] are also included for comparison.

Power-dependent micro-PL of uncapped WSe2 monolayers at 10 K.

The integrated PL intensity for X0, P0, X-, P1, L0, L1 and L2 peaks is calculated and fitted against the

excitation power for each of the samples in a log-log plot to obtain the exponent of the power law

as depicted in Figure S6. Since the L2 peak is only properly identified for uncapped and coated

flakes by E-BEAM, it will be fitted only for these samples. For the rest of the cases, the broad low energy

PL bands are fitted with two convoluted Gaussian curves and the total integrated intensity of both bands is

calculated: PLD (B0 and B1 bands), ALD (C0 and C1 bands) and PECVD (D0 and D1 bands). It should be

noted that the saturation regime is reached for the L1 transition and broad bands at about 200 µW for the

encapsulated flakes by ALD and PECVD. For these cases, the rest of the transitions become visible at

around 50 µW. The power law exponents for each of the fitted curves are given in Table S3. Structural

damage and associated non-radiative recombination channels could account for the slightly smaller values

of α with respect to uncapped and E-BEAM samples.

47

Nano Res 48

Figure S6.- Power-dependent PL integrated intensity for transitions X0, P0, X-, P1, L0, L1, L2 and bands B=B0+B1, C=C0+C1 and D=D0+D1 for uncapped and encapsulated flakes by E-BEAM, PLD, PECVD and ALD techniques.

Uncapped E-BEAM PLD ALD PECVD

X 1.18 ± 0.03 1.18 ± 0.04 0.91 ± 0.01 -- --

P0 1.1 ± 0.03 1.11 ± 0.04 -- -- --

X- 1.06 ± 0.05 1.09 ± 0.02 1.02 ± 0.04 1.8 ± 0.2 1.7 ± 0.2

P1 1.11 ± 0.08 0.96 ± 0.05 0.86 ± 0.03 1.9 ± 0.2 1.7 ± 0.3

L0 0.8 ± 0.2 0.9 ± 0.1 0.75 ± 0.08 2.3 ± 0.2 1.9 ± 0.1

L1 0.65 ± 0.07 0.58 ± 0.03 0.48 ± 0.04 2.0 ± 0.1 1.9 ± 0.2

48

Nano Res 49

L2 0.7 ± 0.1 0.77 ± 0.07 -- -- --

Bands B, C, D -- -- 0.32 ± 0.08 0.8 ± 0.1 1.1 ± 0.1

Table S3.- Power law exponents for each of the fitted curves in Figure S5 for uncapped and encapsulated flakes by E-BEAM, PLD, ALD and PECVD. Temperature-dependent micro-PL of uncapped and encapsulated WSe2 monolayers. The temperature-dependent PL spectra for uncapped and encapsulated samples by E-BEAM and PLD

shown in Figure 6 (main text of the manuscript) were fitted with Gaussian curves corresponding to

transitions X0, P0, X-, P1, L0 and L1. Assuming temperature-independent binding energies for X0, P0, and X-,

the peaks energy shift can be fitted with the Varshni equation as shown in Figure S7a-c as a function of the

temperature. The fitted parameters are shown in Table S4 and are comparable to other values previously

reported for TMDCs [10]. The fitted energy shifts for the cases of encapsulated flakes by ALD and

PECVD are reported on Figure S7d-e (fitting parameters values can be found in the main text of the

manuscript). The energy shifts for the rest of the transitions for each of the cases are shown in Figure S7f-j.

The fitted parameters for the ALD and PECVD cases are: Eg(0)=1694 meV, α=3.5x10-4eV/K (Eg(0)=1705

meV, α=4.0x10-4eV/K) for ALD (PECVD) samples, and β=170 K for both.

Figure S7.- PL peaks energy shift for X0, P0 and X- free exciton emissions fitted with the Varshni equation

49

Nano Res 50

for uncapped (a) and encapsulated flakes by E-BEAM (b), PLD (c), ALD (d) and PECVD (e). PL peaks energy shift for P1, bound/localized exciton emissions L0, L1 and L2 and B0 band for the PLD case (f-h). PL bands D0/D1 for the ALD case (i) and C0/C1 for the PECVD case (j).

Uncapped E-BEAM PLD

X P0 X- X P0 X- X P0 X-

Eg(0) (meV) 1744 1727 1712 1748 1736 1715 1750 -- 1718

α’ (x10-4 eV/K) 4.2 4.3 4.2 4.3 4.5 4.3 4.3 -- 4.5

β (K) 170 170 170 170 170 170 170 170 170

Table S4.- Fitted parameters to the temperature-dependent experimental data using the Varshni equation for uncapped and encapsulated monolayers by E-BEAM and PLD techniques. Optical characterization by spectroscopic ellipsometry of the encapsulating oxide layer. All the samples presented in the PL studies were fabricated by exfoliating flakes on the same commercially

available substrates which consist of a 280-nm-thick SiO2 layer grown on a Si substrate. Therefore, the

main difference between different samples is the encapsulating oxide layer, whose thickness and

stoichiometry have been varied as explained throughout the main text of the manuscript. We have

measured the optical constants (refractive index and extinction coefficient) of the encapsulating oxides by

spectroscopic ellipsometry, in the energy range of interest where the monolayers present the optical

emission (from 1.45 to 1.75 eV), for Al2O3 deposited by E-BEAM (n=1.61) and PLD (n=1.65), and

non-stoichiometric SiO1.80 deposited by PECVD (n=1.48). As expected, no absorption (negligible

extinction coefficient k~0) was found. At this point, it should be noted that the refractive index values for

the encapsulating oxides are very close to each other and, therefore, the encapsulating oxide layer cannot

account for the dramatic decrease of the PL intensity observed for the case of encapsulation with

sub-stoichiometric oxides, especially when performing comparative studies between the different samples.

Room-temperature PL characterization of encapsulated WSe2 monolayers with SiO2 and Al2O2.45.

Figure S8a-b show the corresponding XPS spectra obtained for encapsulating oxides SiO2 and Al2O2.45.

The analysis of the data is performed as indicated above in the first section “Encapsulating oxides

characterization by x-ray photoelectron spectroscopy (XPS)”. In Table S5, estimated atomic percentage,

50

Nano Res 51

binding energy, and full-width-at-half-maximum (FWHM) values are listed for the oxides under study.

The oxide encapsulation is performed by electron-beam deposition (E-BEAM). The samples are first

annealed at 200 °C in vacuum (base pressure of ~5x10-7 mbar) for 30 minutes to remove possible

adsorbates from the surface. Afterwards, the deposition is carried out while rotating the sample at 15 rpm.

The SiO2 deposition was done at a substrate temperature of 200 ºC, whereas the Al2O2.45 oxide deposition

was performed at room-temperature.

Figure S8c shows the PL spectra for an uncapped reference WSe2 monolayer and encapsulated monolayers

with a 30-nm-thick stoichiometric SiO2 and sub-stoichiometric Al2O2.45 oxide layers. In agreement with the

results reported in Figure 3 and 4 of the manuscript, there is a clear transition from a PL spectrum

dominated by the neutral exciton (sotiochiometric SiO2 oxide) to a PL spectrum dominated by charged

exciton (trion) recombination (sub-stoichiometric Al2O2.45 oxide). There is a PL peak shift of about 35 meV

as expected for the trion relative binding energy in WSe2 monolayers. This shift is similar to the values

reported for encapsulating sub-stoichiometric oxides in Figure 3 of the manuscript.

51

Nano Res 52

Figure S8.- XPS spectra with the corresponding fits for the encapsulating oxides deposited by E-BEAM: SiO2 (a) and Al2O2.45 (b). Normalized PL spectra for uncapped reference WSe2 monolayer and encapsulated WSe2 monolayers with SiO2 and Al2O2.45 are reported in (c).

Oxide layer Peak At.% Peak FWHM

52

Nano Res 53

assignment BE

(eV)

(eV)

E-BEAM /

Al2O2.45

O 1s

(Al-oxide) 45.05 531.09 1.92

O 1s (OH) 18.21 532.59 1.78

Al 2p 36.73 74.35 1.6

E-BEAM /

SiO2

O 1s 66.36 532.97 1.5

Si 2p 33.64 103.55 1.49

Table S5.- Atomic percentage, binding energy and FWHM values for each of the fitted curves in Figure S8a-b for encapsulated flakes by E-BEAM.

References. [1] Haeberle, J.; Henkel, K.; Gargouri, H.; Naumann, F.; Gruska, B.; Arens, M.; Tallarida, M.;

Schmeißer, D. Ellipsometry and XPS Comparative Studies of Thermal and Plasma Enhanced Atomic Layer Deposited Al2O3-Films. J. Nanotechnol. 2013, 4, 732–742.

[2] Naumann, V.; Otto, M.; Wehrspohn, R. B.; Werner, M.; Hagendorf, C. Interface and Material Characterization of Thin ALD-Al2O3 Layers on Crystalline Silicon. Energy Procedia 2012, 27, 312–318.

[3] Reddy, N.; Bera, P.; Reddy, V. R.; Sridhara, N.; Dey, A.; Anandan, C.; Sharma, A. K. XPS Study of Sputtered Alumina Thin Films. Ceram. Int. 2014, 40, 11099–11107.

[4] Zhao, W.; Ghorannevis, Z.; Amara, K.; Pang, J. Lattice Dynamics in Mono-and Few-Layer Sheets of WS2 and WSe2. Nanoscale 2013, 5, 9677–9683.

[5] Wang, G.; Bouet, L.; Lagarde, D.; Vidal, M.; Balocchi, A.; Amand, T.; Marie, X.; Urbaszek, B. Valley Dynamics Probed through Charged and Neutral Exciton Emission in Monolayer WSe2. Phys. Rev. B - Condens. Matter Mater. Phys. 2014, 90, 075413.

[6] Huang, J.; Hoang, T. B.; Mikkelsen, M. H. Probing the Origin of Excitonic States in Monolayer WSe2. Sci. Rep. 2016, 6, 22414.

[7] Srivastava, A.; Sidler, M.; Allain, A. V.; Lembke, D. S.; Kis, A.; Imamoğlu, A. Optically Active Quantum Dots in Monolayer WSe2. Nat. Nanotechnol. 2015, 10, 491–496.

[8] Koperski, M.; Nogajewski, K.; Arora, A.; Cherkez, V.; Mallet, P.; Veuillen, J.-Y.; Marcus, J.; Kossacki, P.; Potemski, M. Single Photon Emitters in Exfoliated WSe2 Structures. Nat. Nanotechnol. 2015, 10, 503–506.

[9] You, Y.; Zhang, X. X.; Berkelbach, T. C.; Hybertsen, M. S.; Reichman, D. R.; Heinz, T. F. Observation of Biexcitons in Monolayer WSe2. Nat. Phys. 2015, 11, 477–482.

[10] Plechinger, G.; Nagler, P.; Kraus, J.; Paradiso, N.; Strunk, C.; Schüller, C.; Korn, T. Identification of Excitons, Trions and Biexcitons in Single-Layer WS2. Phys. Status Solidi - Rapid Res. Lett. 2015, 9, 457–461.

53

Nano Res 54

54