Content uploaded by Jaehoon Ji
Author content
All content in this area was uploaded by Jaehoon Ji on Sep 30, 2019
Content may be subject to copyright.
www.advmatinterfaces.de
FULL PAPER
1900637 (1 of 8) © 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Layer-Number-Dependent Electronic and Optoelectronic
Properties of 2D WSe2-Organic Hybrid Heterojunction
Jaehoon Ji and Jong Hyun Choi*
DOI: 10.1002/admi.201900637
light-emitting devices,[9] photovoltaics,[10,11]
and phototransistors.[12]
To fully exploit the excellent proper-
ties of 2D WSe2 for nano- and opto-
electronics, it is important to establish
heterojunctions such as p–n, n–p–n, or
p–n–p structures[13,14] and their layer-
number-dependent properties at the
heterointerfaces must be understood.
However, such studies have been limited
due to the large difficulties in manufac-
turing. Conventional approaches require
multiple complex, laborious preparation
steps to construct heterostructures such
as TMDC-graphene,[10] TMDC–TMDC,[15]
TMDC-insulating 2D material (e.g., hex-
agonal boron nitride).[16] For example,
one of the heterolayers is manually trans-
ferred onto the other flake prepared in a
separate substrate by the “pick-up and
transfer” method using an elastomeric
stamp.[17] To overcome the difficulty and
cumbersomeness of the transfer pro-
cesses, direct synthesis of heterostructures
on a single substrate has been developed,
including serial chemical vapor deposition
(CVD) and stacking of dissimilar TMDC
layers[18] and deposition of organic layers onto TMDC flakes via
thermal evaporation.[19] However, these methods still require
multiple sublimation/deposition cycles at high temperatures
(e.g., ≈1000 °C) and low pressures (e.g., ≈0 Torr). In this work,
we demonstrate a hybrid heterostructure from organic mole-
cules stacked on mechanically exfoliated WSe2 flakes of mono-
layer (1L) through 6-layer (6L) via one-step solution-phase
chemistry and systematically investigate their layer-number-
dependent electronic and optoelectronic properties (Figure 1a).
Eosin Y (EY) dye molecules are selected in this work, as they
assemble a uniform layer on TMDC flakes via van der Waals
interaction from simple dipping of the flake/substrate into
EY-containing solution. The deposited EY layer forms a type II
band heterojunction with 2D WSe2 flakes. The energy levels
of 2D WSe2 and EY are illustrated in Figure 1b. As the layer
number of WSe2 increases, its conduction band minimum
(CBM) decreases, while the valence band maximum (VBM)
increases.[2,20,21] EY’s lowest unoccupied molecular orbital
(LUMO) level is positioned lower than the CBM of 6L WSe2,
and its highest occupied molecular orbital (HOMO) level
is lower than the VBM of 1L WSe2, forming a type II align-
ment.[22–24] Additionally, EY is a water-soluble dye molecule and
can provide a p-doping on WSe2.[25,26] In a previous report, we
One of the most attractive features of 2D WSe2 is a tunability in its elec-
tronic and optoelectronic properties depending on the layer number. To
harness such unique characteristics for device applications, high quality and
easily processable heterojunctions are required and relevant layer-number-
dependent properties must be understood. Herein, a study is reported
on hybrid heterojunctions between 2D WSe2 and organic molecules from
one-step solution chemistry and their layer-number-dependent properties.
Eosin Y (EY) dye is selected as a p-dopant and uniformly stacked on mechani-
cally exfoliated WSe2 flakes via van der Waals interaction, forming a hybrid
heterojunction with a type II alignment. The EY-WSe2 heterojunction shows
significantly enhanced currents compared to pristine WSe2 with a lower bar-
rier height and a longer effective screening length. The work function of the
heterostructure is also lower than that of pristine WSe2. The efficient exciton
dissociation and doping effect by EY are confirmed by photocurrent and pho-
toluminescence measurements, where WSe2 emission is markedly quenched
by EY and exciton contribution decreases with layer number. These findings
shed critical insights into layer-number-dependent electronic and optoelec-
tronic properties of organic-WSe2 layers and also provide simple yet effective
means to construct transition metal dichalcogenide-based heterostructures,
which should be valuable for developing layered 2D devices.
J. Ji, Prof. J. H. Choi
School of Mechanical Engineering
Purdue University
West Lafayette, IN 47907, USA
E-mail: jchoi@purdue.edu
The ORCID identification number(s) for the author(s) of this article
can be found under https://doi.org/10.1002/admi.201900637.
2D Hybrid Materials
1. Introduction
Atomically thin tungsten diselenide or WSe2 has garnered sub-
stantial interest due to its unique electrical and optical proper-
ties, which originate from the lattice structure and confined
dimensionality.[1] Its 2D layered structure provides sizable
bandgap depending on the thickness.[2,3] It exhibits inherent
p-type characteristics and relatively high hole mobility of
around 500–700 cm2 V−1 s−1.[4] These properties may be useful
for nanoelectronics including diodes,[5] field effect transis-
tors,[6] and logic circuits[7] with a high on/off ratio and high car-
rier mobility. Additionally, its optical properties such as strong
absorption and bright photoluminescence (PL) could be har-
nessed for optoelectronic applications such as photodetectors,[8]
Adv. Mater. Interfaces 2019, 6, 1900637
www.advancedsciencenews.com
© 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1900637 (2 of 8)
www.advmatinterfaces.de
found that not only organic dopants, but also solvents where
dopant molecules are dispersed can contribute to interlayer
charge transfer to/from 2D TMDC layers, depending on the
relative electronegativity.[27] In contrast, water exhibits no sig-
nificant doping effects on thermally annealed TMDCs. Thus, by
adopting EY molecules, it is possible to study pure doping pro-
cesses at the interface between EY and WSe2. The convenience
and easiness of this facile approach allows us to investigate the
layer-number-dependent characteristics of WSe2 heterojunc-
tions, which is otherwise a significant challenge due to difficul-
ties in fabrication of heterostructures.
In this work, the electronic and optoelectronic properties of
the EY-WSe2 heterojunction are studied as a function of layer
number with conductive and photoconductive atomic force
microscopy (C-AFM and PC-AFM), Kelvin probe force micros-
copy (KPFM), and PL spectroscopy.[28–30] We find that the barrier
height for electron transport in the heterojunction is lowered
than that of pristine WSe2 by ≈350 meV via type II alignment
and decreases monotonically with increasing layer number. The
work function of the heterostructure also drops by ≈114 meV
compared to pristine samples, as the work functions of WSe2
and EY layers become aligned. Under light illumination, the
EY-WSe2 layers produce greater photocurrents than pristine
layers due to facilitated exciton dissociation at the interface. The
charge transfer at the heterostructure is confirmed by strong
PL quenching in WSe2, which indicates a p-doing effect by EY.
These observations provide insight on layer-number-dependent
properties of the WSe2-organic heterojunctions, which could
form the basis for next-generation WSe2 based optoelectronics.
2. Results and Disccussion
2.1. Layer-Number-Dependent Electronic Properties
Various WSe2 flakes were prepared by mechanical exfoliation,
and their layer numbers were determined by Raman and AFM
measurements (Figure 1c and Figures S1–S3, Supporting Infor-
mation).[31,32] The WSe2 flakes were immersed in aqueous solu-
tion containing 1 × 10−3 m EY for 24 h, subsequently rinsed
with methanol and DI water, and blow dried with air (see
the Experimental Section). The EY-WSe2 heterolayers on the
Si/SiO2 substrate show a spatially uniform height difference of
about 0.70 ± 0.06 nm throughout the entire flakes compared
to the pristine WSe2. Even after harsh washing with DI water
for more than 2 h, the stacked EY layer was not dissolved or
peeled off from the flakes (Figure 1d,e). Importantly, similar
height difference is observed regardless of the WSe2 layer num-
bers (Figure S3, Supporting Information). The uniform layer of
EY molecules allows us to study the layer-number-dependent
Adv. Mater. Interfaces 2019, 6, 1900637
Figure 1. a) A schematic of the charge transfer between 2D WSe2 and EY. Charge carriers of WSe2 can be transferred to the stacked EY layer. b) Energy
band diagram of WSe2 flakes with different layer number (1L through 6L), EY, ITO, and Pt/Ir probe. c) Raman spectra of pristine WSe2 (dotted line) and
EY treated WSe2 (solid line). There are no significant changes in Raman signatures of the WSe2 flake before and after EY adsorption. d) AFM images
of pristine WSe2 (left) and EY treated WSe2 (right) on a Si/SiO2 substrate. e) Height profiles along the lines in (d), indicating that the WSe2 flake is 4L
and there is a uniform increase of ≈0.6 nm after EY functionalization. f) Schematic of PC-AFM. Electric potential is applied on the ITO substrate, while
the Pt/Ir probe is maintained as a ground. The current flow in the sample is spatially monitored by raster-scanning with the Pt/Ir-coated probe in dark
state and under photoirradiation from a 405 nm laser.
www.advancedsciencenews.com
© 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1900637 (3 of 8)
www.advmatinterfaces.de
properties of the hybrid heterostructure by varying the WSe2
thickness. It is also worth noting that EY molecules were selec-
tively adsorbed on WSe2 and there was no noticeable EY on the
bare Si substrate.
To investigate the current flow of pristine and hybrid WSe2,
the samples were prepared on indium tin oxide (ITO)-glass sub-
strates and measured with C-AFM (Figure 1f). When a positive
potential is applied on the ITO with the Pt/Ir probe grounded,
a diode-like behavior is observed in the pristine layers (also see
Figure S4, Supporting Information for current–voltage pro-
perties under negative potentials). For example, the current on
1L WSe2 does not change significantly under low bias voltage
and then increases drastically above ≈0.67 V (Figure 2a). The
threshold voltage (determined at 1.5 nA) increases as a func-
tion of layer number, from ≈0.95 V (2L) to ≈1.7 V (6L). The
electron affinity (i.e., CBM) of 1L WSe2 (Φ1L ≈ 3.70 eV)[2,21]
is relatively smaller than the work functions of the probe tip
(Φtip ≈ 5.05 eV)[33,34] and ITO (ΦITO ≈ 4.75 eV)[35] (Figure 1b and
Figure 2a inset), rendering a Schottky contact under equilibrium
conditions. We adopt the field-assisted Fowler– Nordheim (FN)
tunneling model to elucidate the current dynamics (shown as
fitted curves in Figure 2a; also see the Supporting Information
for details).[36] Two parameters that govern the current behav-
iors are barrier height (ΦB) and screening length (d), which are
shown as a function of layer number in Figure 2c. The barrier
height between 1L WSe2 and the probe in our experiment is
determined to be ≈1.10 eV (Figure 2c), which is lower than the
reported value based on the Schottky–Mott theory (≈1.35 eV).[37]
The difference may be attributed to the Bardeen pinning effect,
a screening from the interfacial dipole,[38–40] and similar off-
sets (i.e., ≈0.25 eV) were reported in previous experimental
studies.[41,42]
As the layer number increases, the barrier height gradu-
ally decreases and reaches ≈0.80 eV for 6L WSe2 (Figure 2c).
This trend is consistent with conventional bandgap thinning
behavior.[3,21,43] The effective screening length of 1L WSe2 is
measured to be ≈2.5 nm (Figure 2c), which is in good agree-
ment with the theoretical value (≈3 nm).[44] As the layer number
increases, the screening length also increases from ≈3.8 nm
(2L) to ≈9.4 nm (6L). The barrier height and screening length
together (ΦB1.5⋅d) yield a linear relationship with layer number
(Figure 2d), which is consistent with gradually increasing
Adv. Mater. Interfaces 2019, 6, 1900637
Figure 2. Layer-number-dependent carrier transport properties in WSe2 and its hybrid junction. a) Current–voltage characteristics of pristine WSe2 as
a function of layer number from 1L to 6L. Inset shows a schematic of ITO-WSe2-Pt/Ir probe junction under the forward bias applied on the ITO. The
fitted lines with corresponding colors of the experimental data (filled objects) are obtained from the Fowler–Nordheim tunneling model. b) Current
versus voltage plot of hybrid WSe2-EY (inset: a schematic of the heterojunction with the forward bias). c) Effective screening length d (black) and
barrier height ΦB (red) of pristine WSe2. d) Tunneling barrier coupled with effective screening length (ΦB3/2d) demonstrates a linear relationship with
layer number. e,f) Effective screen length, barrier height, and ΦB3/2d of the hybrid junctions as a function of layer number. g) Work functions of pristine
WSe2 and WSe2-EY.
www.advancedsciencenews.com
© 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1900637 (4 of 8)
www.advmatinterfaces.de
threshold voltage. A stacked layer of EY molecules forms
a uniform junction with WSe2 flakes as we observe no
apparent spatial current variations in the EY-WSe2 samples
(Figures S8–S13, Supporting Information). The hybrid het-
erostructure demonstrates a significantly increased level of
currents under the same potential window (Figure 2b). For
example, the heterostructure from EY and 1L WSe2 shows
≈11 nA under 500 mV, whereas pristine 1L WSe2 exhibits only
about 0.38 nA under the same potential. Similarly, other het-
erosamples (with 2L through 6L WSe2) also generate far greater
amount of currents than pristine layers at given voltage. As
illustrated in the inset in Figure 2b, the EY layer may serve as an
intermediate energy level,[26] facilitating electron transfer from
the probe to the WSe2 CBM and eventually to the ITO. Inter-
estingly, the threshold voltage of the heterosamples increases
gradually with layer number, yet it does not vary significantly at
thicker layers, e.g., from ≈0.80 V (5L) to ≈0.83 V (6L).
The measured barrier heights of the heterolayers range from
≈0.65 eV (1L) to ≈0.49 eV (6L) as shown in Figure 2e. The barrier
heights are consistently lower than those of the pristine WSe2
by ≈0.35 eV, which is similar to the energy difference between
the EY LUMO and the WSe2 CBM. The deposited EY also plays
a role in screening current flow. The effective screening lengths
of the heterostructures are greater than those of the pristine
WSe2 by ≈0.50 nm regardless of WSe2 layer number. These
results indicate that the organic layer forms a robust hetero-
junction on 2D WSe2, which is also consistent with the fact that
its thickness is uniform (≈0.70 nm) on all the WSe2 layers in
our experiment. When we consider the barrier height and the
screening length together, ΦB1.5⋅d in Figure 2f shows a gradual
increase with WSe2 layer number, yet it levels off above 4L and
the increase is not significant in thicker samples. It is worth
noting that this trend corresponds to the insignificant variation
of threshold voltage in Figure 2b with thicker heterosamples
(i.e., 5L and 6L), and this may be attributed to the fact that the
CBM of thicker WSe2 layers approach the EY LUMO.
In EY-WSe2 samples, interlayer charge transfer occurs
between them and the work function is thus arranged. To deter-
mine the work function, the surface potential difference (Vme)
between the probe tip and the sample is recorded with KPFM.
The surface potential difference is correlated with work func-
tion: ΦS = Φtip + Vme, where ΦS and Φtip denotes the work func-
tions of the sample and the tip, respectively. The measured work
function of the 1L WSe2 is ≈−4.78 eV as shown in Figure 2h
(see also Figure S5, Supporting Information), which is con-
sistent with the value reported by Wang et al. (−4.76 eV).[45] It
is noted that the work function is closer to the VBM of WSe2
(−5.35 eV), which indicates that the pristine WSe2 has inherent
p-type characteristics. The flat band potential of the EY layer is
≈−5.04 eV.[25] In EY-WSe2 heterolayers, electron transfer will
take place from WSe2 to low-lying EY to compensate the energy
level difference, and their work function should be between
the levels of WSe2 and EY. Our KPFM confirms that the work
functions of EY-WSe2 hybrids differ from those of pristine sam-
ples by ≈114 meV on average, varying from −97 meV (1L) to
−128 meV (8L). The greater adjustment of the work function in
thick layers may be induced by relatively large energy difference
between the deposited EY and thick WSe2 flake, which results
in stronger interlayer charge transfer.[46]
2.2. Layer-Number-Dependent Optoelectronic Properties
Figure 3a–f presents the current–voltage responses of the pris-
tine WSe2 and EY-WSe2 hybrids (WSe2 of 1L through 6L) with
and without light illumination. Given the limit in the CAFM
measurement (±13 nA), currents were recorded within the
limit. It is notable that asymmetric current–voltage responses
of the samples are observed depending on the direction of
applied potential. When the same magnitude of potential is
applied in different directions (e.g., +/− voltage), pristine WSe2
produces a higher level of currents under reverse bias. In con-
trast, hybrid samples exhibit an opposite behavior due to the
lowered barrier height, thus yielding more currents under for-
ward bias. For example, 2L WSe2 generates only 0.18 nA of cur-
rent under 0.75 V, but exhibits −3.6 nA under −0.75 V (blue
dots in Figure 3a–f), whereas EY-functionalized 2L WSe2 shows
12.27 and −0.64 nA under the identical potential (red dots).
Light illumination generates significantly more currents
in both pristine WSe2 and hetero layers compared to the
dark state. For example, photoirradiated 3L WSe2 produces
≈−3.6 nA under −0.5 V which is 16 times higher than that
without illumination. The illuminated EY-functionalized 3L
WSe2 demonstrates about −4.6 nA with the sample potential
which is higher than that in the dark state by a factor of more
than 30. In our experiment, a 405 nm diode laser is used for
irradiation, because this wavelength excites WSe2, but not EY
(see Figure S6, Supporting Information).[47] The photoinduced
excitons in WSe2 layers are dissociated into the current flow.
The stacked EY layer facilitates the dissociation of excitons
at the interface, generating a significantly greater amount of
currents.
The current enhancement by photoirradiation is not only
influenced by the presence of EY layer, but also depends on
the WSe2 layer number. To compare the photoresponse (E) of
the samples as a function of layer number, we define E as the
voltage difference needed to generate a fixed amount of current
(Ifix) with (VL) and without (VD) illumination, normalized by the
dark state voltage
DL
D
fix
EVV
V
I
=
−
(1)
This relation measures the ability of the sample to create photo-
currents relative to dark state. It is, however, difficult to directly
compare E of the pristine and hetero samples at identical cur-
rents. The pristine WSe2 shows strong currents at reverse bias,
whereas hetero layers exhibit an opposite trend. Thus, we com-
pare E of the pristine WSe2 under reverse bias with that of the
hybrid layers under forward bias, while maintaining the same
magnitude of currents. Figure 3g shows the photoresponse E
of both pristine and hybrid WSe2 as a function of layer number
at −10 nA for the pristine WSe2 and 10 nA for the hybrid layer.
The photoresponse of pristine WSe2 decreases with increasing
layer number from ≈0.57 (1L) to nearly 0.083 (4L), and then
slightly recovers in thicker layers (e.g., ≈0.27 with 6L). As the
layer number increases, the screening length of WSe2 also
expands. The majority of excitons in multilayer flakes may thus
be screened and cannot be dissociated to contribute to the cur-
rent flow.[48] Further, thicker flakes can absorb a greater amount
Adv. Mater. Interfaces 2019, 6, 1900637
www.advancedsciencenews.com
© 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1900637 (5 of 8)
www.advmatinterfaces.de
of photons than thinner layers,[42] which may result in a slight
recovery in E. The EY-WSe2 heterostructures show greater
photoresponses than the pristine samples, ranging between
≈0.74 (1L) and ≈0.58 (6L), yet the layer-number-dependence is
less prominent. The staggered bands between the EY layer and
WSe2 flake can promote the dissociation of excitons, thereby
leading to greater photoresponses.
We also measured the photoresponses of pristine and hetero
layers as functions of fixed current and layer number as shown
in Figure 3h,i. The pristine flakes show a less prominent layer-
number-dependent behavior at a low current (e.g., <−6 nA)
than at the high current (e.g., >−6 nA). At −1 nA, for example,
the photoresponse varies less than 0.20 between 1L and 6L.
At a high current (e.g., −10 nA), however, the photoresponse
decreases with increasing layer number and recovers moder-
ately at thicker layers, as discussed above (Figure 3g). Unlike
the low current conditions, two salient competing mechanisms
determine photoresponses at high currents. Thicker flakes
absorb more photons than thinner layers, yet their exciton
dissociation becomes less efficient due to the greater screening
effect. As a result, the photoresponse will decrease at first and
then increases with increasing layer number. On the other
hand, EY-functionalized WSe2 heterostructures show distinct
photoresponse behaviors, demonstrating not only significantly
enhanced E, but also nearly uniform responses at various cur-
rents for given layer number. The hetero samples thinner than
4L all maintain a similar level of photoresponses at ≈0.70 at var-
ious currents. The E values of thicker hetero layers (e.g., 5L and
6L) are slightly smaller, but still uniform at ≈0.50 regardless of
currents. The exciton dissociation in the hetero interfaces is
facilitated by the staggered band alignment, and may not be
severely interrupted by the screening, resulting in more uni-
form photoresponses regardless of currents and layer numbers.
2.3. Layer-Number-Dependent Doping Effect
We investigate the interlayer doping effect between the stacked
EY and WSe2 flakes by using PL spectroscopy. Figure 4a–f
Adv. Mater. Interfaces 2019, 6, 1900637
Figure 3. Photoconductivity of pristine WSe2 flakes and hybrid junctions for various layer numbers. a–f) Current–voltage characteristics of pristine
WSe2 (blue) and EY-WSe2 (red) in the dark (filled) and under illumination (hollow). g) A plot of photoresponse of pristine WSe2 (dark blue) and EY-
treated WSe2 (pink) to obtain −10 and 10 nA of current, respectively. To compare how effectively the sample transport photoexcited carriers, the photo-
response of the samples is defined as the required voltage difference to obtain a same amount of current (I ) with ( VL) and without (VD) illumination
normalized by the dark voltage itself. Under light illumination, even though only small amount of the voltage (0.15 V) is applied on the 1L heterojunc-
tion, it can generate 10 nA of current which requires 0.57 V in the dark. Mapping of photoresponse with respect to the layer number and current:
h) pristine WSe2 i) hybrid junction. The photoresponse of hybrid junction is uniform, even the current level varies.
www.advancedsciencenews.com
© 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1900637 (6 of 8)
www.advmatinterfaces.de
presents the PL spectra of the pristine (blue line) and the
hybrid (red line) samples for various WSe2 thickness. A 633 nm
laser is used for excitation which can only excite the WSe2, thus
the PL spectra are solely from the WSe2 emission. Pristine 1L
exhibits a dominant peak around 748 nm (1.66 eV). As the layer
number increases up to 4L, the PL spectra become broader,
with asymmetric tails at longer wavelength. Thicker flakes (5L
and 6L) display distinct two peaks around 768 nm (1.61 eV) and
848 nm (1.46 eV), which are much weaker in the intensity and
longer in the wavelength than those of 4L. EY-functionalized
WSe2 samples show drastically weaker PL spectra compared
to the pristine flakes. The strong PL quenching indicates that
p-doping take place on WSe2 by the EY layer, which results in
more positive trions and suppresses the excitonic recombina-
tion. The emission peaks of the hetero flakes are red-shifted
from those of pristine samples. For example, the 1L hybrid
has its maximum peak at ≈765 nm (1.62 eV) which is ≈17 nm
longer than that of pristine WSe2. Thicker hybrid samples
(5L and 6L) also show their emission features at longer wave-
lengths ≈774 nm (1.60 eV) and ≈862 nm (1.44 eV). The PL
shifts are caused by the dielectric screening by the EY layer (see
the Supporting Information for details).
To investigate the layer-number-dependent PL character-
istics of the samples, the PL spectra are deconvoluted with
multiple peaks (see Figure S7, Supporting Information for
details). The 1L WSe2 PL consists of two components: exciton
and positive trion.[49] Besides excitons (≈748 nm for pristine
1L WSe2 and ≈765 nm for EY-treated 1L sample), the hole-rich
characteristics of WSe2 causes positive trions to appear as the
shoulder at longer wavelengths around 757 nm (1.64 eV) for
the pristine 1L WSe2 and 772 nm (1.61 eV) for the hybrid 1L
samples (Figure 4g).[50] The PL signals of multi-layer samples
Adv. Mater. Interfaces 2019, 6, 1900637
Figure 4. a–f) PL spectra of pristine WSe2 (blue) and EY-WSe2 (red) for various layer numbers from 1L to 6L. The PL of each WSe2 flake is markedly
quenched by the EY treatment due to charge transfer. g) Spectral intensities and h) positions of excitons X (dark blue: pristine, red: hybrid junction),
trions X+ (cyan: pristine, magenta: hybrid junction), and indirect transitions i (purple: pristine, pink: hybrid junction) as a function of layer number.
i) Exciton contributions in PL signatures of pristine (dark blue) and EY (pink) treated WSe2 samples for various layer numbers. Exciton contribution
increases with layer number in both WSe2 and EY-WSe2 samples. Trion contribution is consistently greater in EY-WSe2 than pristine WSe2, due to a
p-doping effect by EY.
www.advancedsciencenews.com
© 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1900637 (7 of 8)
www.advmatinterfaces.de
are deconvoluted into direct band transitions (i.e., exciton and
trion) and also indirect band transition which arises from
interlayer interactions in multi-layer WSe2 (see the Supporting
Information for details).[43] The indirect transition becomes
prominent in thicker layers, displaying the distinct emission
peaks at long wavelengths at ≈848 nm for pristine WSe2 and
≈862 nm for EY-WSe2.
The PL intensity varies significantly with the layer number.
Pristine 1L shows very strong PL emission (Figure 4h). The
intensities of exciton (dark blue) and positive trion (cyan)
decreases monotonously about two orders of magnitude with
increasing the layer number up to 5L and then level off beyond
5L. Compared to exciton and trion transitions, the intensity of
the indirect peaks of few layers does not change significantly
(shown in purple).[42] The overall PL signal becomes weak and
the indirect transition becomes more prominent. EY-WSe2
heterolayers demonstrate more than an order of magnitude
smaller intensities than pristine samples (Figure 4h). Both
exciton (red) and trion (magenta) features decrease in their
intensities about tenfold with increasing layer number up to 4L.
Thicker layers (5L and 6L) do not show considerable difference
in the intensity. Similar to the pristine samples, the decrease of
the indirect peak intensity depending on the layer number is
relatively small.
The doping effect is manifested by comparing the intensity
ratio between exciton and trion peaks.[51] Figure 4i presents
the proportion of exciton (ΔXex) compared to positive trion,
−
+
()
XX
X
, as a function of layer number. Here X and X+ denote
the intensity of exciton and trion recombination, respectively.
ΔXex of pristine WSe2 is ≈57% with 1L and gradually increases
to ≈75% with 6L. As the layer thickens, the hole-rich proper-
ties of WSe2 becomes more balanced with decreasing trion[52,53]
and increasing exciton contributions to the PL emission.[51] The
hetero flakes show considerably reduced exciton contributions,
ranging from ≈40% (1L) to ≈59% (6L). Nearly 20% of exciton
contribution is reduced compared to the pristine WSe2. The
reduced exciton contribution strongly supports the p-doping
effect of the stacked EY layer on WSe2 flakes.
3. Conlcusions
In summary, we have systemically investigated the layer-
number-dependent properties of the WSe2-organic hetero-
junctions from simple solution-based chemistry. Currents,
photocurrents, and work function all vary as a function of the
layer number of WSe2. EY serves as an intermediate layer to
facilitate the dissociation of the photoexcited energies into cur-
rents and is also effective in p-doping on WSe2 flakes. These
behaviors agree well with the designed type II alignment of het-
erojunction. It is evident that unique layer-number-dependent
properties of WSe2 are well retained after organic function-
alization. Given the simplicity and the effectiveness in the
preparation, our approach may be adopted to engineer the
properties of TMDCs for electronics and optoelectronics. This
work also offers a means to study the physics of 2D materials
and enhance the versatility and performance of the WSe2 based
devices.
4. Experimental Section
Sample Preparation: Microsized flakes were mechanically exfoliated
from WSe2 crystals (2D materials supplies) using the scotch tape
method.[54] The flakes were deposited onto the 285 nm thick Si/SiO2
substrates and their layer numbers were examined under a Renishaw
inVia confocal Raman microscope with a 100× objective lens. For the
electrical measurement, the exfoliated flakes were deposited on glass
substrates that had ITO with a thickness of 180 nm and resistivity of
8–10 Ω cm−2. Before deposition, the substrates were cleaned by bath
sonication with acetone, methanol, and DI water, and also treated with
oxygen plasma for 10 min with 250 watts to improve the contact between
flakes and substrates. The deposited samples were subsequently
annealed at ≈250 °C for 1 h with Ar gas, which removed tape residues.
To construct hybrid heterojunctions, the deposited WSe2 flakes were
immersed in aqueous solution containing 1 × 10−3 m EY (Sigma-Aldrich)
at 4 °C for 24 h. The samples were then washed with methanol and DI
water to remove unbound and weakly bound EY molecules, and finally
dried with air.
Sample Characterization: Raman and PL signatures of WSe2 flakes
were measured under the Renishaw confocal microscope with a 633 nm
HeNe laser at room temperature. The low laser power (0.2 mW)
was used to avoid unexpected heating of the flakes. Both AFM and
PC-AFM measurements were conducted using a Bruker Dimension
Icon under the ambient conditions. AFM images were obtained using
a SCANASYST-AIR probe, while a SCM-PIC-V2 probe with a Pt/Ir coated
and antimony doped low resistive tip was used in a PF-TUNA module
for current measurement. In C-AFM, the current limit in the preamplifier
is ±13 nA and the current sensitivity is ≈1 nA V−1. A 405 nm laser
(Laserglow Technologies) with ≈1 mW cm−2 at the sample was used to
excite the flakes using a backside illumination optics assembly (Bruker).
Supporting Information
Supporting Information is available from the Wiley Online Library or
from the author.
Acknowledgements
This work was supported by National Science Foundation.
Conflict of Interest
The authors declare no conflict of interest.
Keywords
electronic properties, heterostructure, layer-number-dependence,
optoelectronics, transition metal dichalcogenide, WSe2
Received: April 10, 2019
Revised: June 9, 2019
Published online: July 3, 2019
[1] M. Zeng, Y. Xiao, J. Liu, K. Yang, L. Fu, Chem. Rev. 2018, 118, 6236.
[2] Y. Liu, P. Stradins, S.-H. Wei, Sci. Adv. 2016, 2, e1600069.
[3] A. Arora, M. Koperski, K. Nogajewski, J. Marcus, C. Faugeras,
M. Potemski, Nanoscale 2015, 7, 10421.
[4] H. C. P. Movva, A. Rai, S. Kang, K. Kim, B. Fallahazad, T. Taniguchi,
K. Watanabe, E. Tutuc, S. K. Banerjee, ACS Nano 2015, 9, 10402.
Adv. Mater. Interfaces 2019, 6, 1900637
www.advancedsciencenews.com
© 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1900637 (8 of 8)
www.advmatinterfaces.de
[5] A. Ahmet, M. Kolyo, M. E. Gonzalez, I. Giuseppe, W. Kenji,
T. Takashi, F. Gianluca, K. Andras, Adv. Mater. 2018, 30, 1707200.
[6] P. Pushpa Raj, G. S. Michael, O. Akinola, T. W. Anthony, N. H. Anna,
P. B. Dayrl, X. Kai, G. M. David, Z. W. Thomas, D. R. Philip, Nano-
technology 2017, 28, 475202.
[7] M. Tosun, S. Chuang, H. Fang, A. B. Sachid, M. Hettick, Y. Lin,
Y. Zeng, A. Javey, ACS Nano 2014, 8, 4948.
[8] S.-H. Jo, D.-H. Kang, J. Shim, J. Jeon, M. H. Jeon, G. Yoo, J. Kim,
J. Lee, G. Y. Yeom, S. Lee, H.-Y. Yu, C. Choi, J.-H. Park, Adv. Mater.
2016, 28, 4824.
[9] J. S. Ross, P. Klement, A. M. Jones, N. J. Ghimire, J. Yan,
D. G. Mandrus, T. Taniguchi, K. Watanabe, K. Kitamura, W. Yao,
D. H. Cobden, X. Xu, Nat. Nanotechnol. 2014, 9, 268.
[10] D. Jariwala, V. K. Sangwan, L. J. Lauhon, T. J. Marks, M. C. Hersam,
ACS Nano 2014, 8, 1102.
[11] H. Zhang, J. Choi, A. Ramani, D. Voiry, S. N. Natoli, M. Chhowalla,
D. R. McMillin, J. H. Choi, ChemPhysChem 2016, 17, 2854.
[12] N. Guo, F. Gong, J. Liu, Y. Jia, S. Zhao, L. Liao, M. Su, Z. Fan,
X. Chen, W. Lu, L. Xiao, W. Hu, ACS Appl. Mater. Interfaces 2017, 9,
34489.
[13] a) H.-M. Li, D. Lee, D. Qu, X. Liu, J. Ryu, A. Seabaugh, W. J. Yoo,
Nat. Commun. 2015, 6, 6564; b) J. I. J. Wang, Y. Yang, Y.-A. Chen,
K. Watanabe, T. Taniguchi, H. O. H. Churchill, P. Jarillo-Herrero,
Nano Lett. 2015, 15, 1898.
[14] J. Choi, H. Chen, F. Li, L. Yang, S. S. Kim, R. R. Naik, P. D. Ye,
J. H. Choi, Small 2015, 11, 5520.
[15] R. Cheng, D. Li, H. Zhou, C. Wang, A. Yin, S. Jiang, Y. Liu, Y. Chen,
Y. Huang, X. Duan, Nano Lett. 2014, 14, 5590.
[16] J. Wierzbowski, J. Klein, F. Sigger, C. Straubinger, M. Kremser,
T. Taniguchi, K. Watanabe, U. Wurstbauer, A. W. Holleitner,
M. Kaniber, K. Müller, J. J. Finley, Sci. Rep. 2017, 7, 12383.
[17] F. Pizzocchero, L. Gammelgaard, B. S. Jessen, J. M. Caridad,
L. Wang, J. Hone, P. Bøggild, T. J. Booth, Nat. Commun. 2016, 7, 11894.
[18] Y. Gong, J. Lin, X. Wang, G. Shi, S. Lei, Z. Lin, X. Zou, G. Ye,
R. Vajtai, B. I. Yakobson, H. Terrones, M. Terrones, B. K. Tay, J. Lou,
S. T. Pantelides, Z. Liu, W. Zhou, P. M. Ajayan, Nat. Mater. 2014, 13,
1135.
[19] D. Jariwala, S. L. Howell, K.-S. Chen, J. Kang, V. K. Sangwan,
S. A. Filippone, R. Turrisi, T. J. Marks, L. J. Lauhon, M. C. Hersam,
Nano Lett. 2016, 16, 497.
[20] Z. Wenjun, H. Yi, M. Lianbo, Z. Guoyin, W. Yanrong, X. Xiaolan,
C. Renpeng, Y. Songyuan, J. Zhong, Adv. Sci. 2018, 5, 1700275.
[21] Z. Changjian, Z. Yuda, R. Salahuddin, W. Yi, L. Ziyuan, C. Mansun,
C. Yang, Adv. Funct. Mater. 2016, 26, 4223.
[22] V. S. Manikandan, A. K. Palai, S. Mohanty, S. K. Nayak, J. Photo-
chem. Photobiol., B 2018, 183, 397.
[23] E. L. Lim, C. C. Yap, M. Yahaya, M. M. Salleh, Semicond. Sci. Technol.
2013, 28, 045009.
[24] R. Sharma, A. Kamal, R. K. Mahajan, RSC Adv. 2016, 6, 71692.
[25] N. Zhang, J. Shi, F. Niu, J. Wang, L. Guo, Phys. Chem. Chem. Phys.
2015, 17, 21397.
[26] F. Zhang, F. Shi, W. Ma, F. Gao, Y. Jiao, H. Li, J. Wang, X. Shan,
X. Lu, S. Meng, J. Phys. Chem. C 2013, 117, 14659.
[27] J. Choi, H. Zhang, H. Du, J. H. Choi, ACS Appl. Mater. Interfaces
2016, 8, 8864.
[28] J. Choi, H. Zhang, J. H. Choi, ACS Nano 2016, 10, 1671.
[29] H. Zhang, J. Huang, Y. Wang, R. Liu, X. Huai, J. Jiang, C. Anfuso,
Opt. Commun. 2018, 406, 3.
[30] Y. Li, C.-Y. Xu, L. Zhen, Appl. Phys. Lett. 2013, 102, 143110.
[31] C. Lee, H. Yan, L. E. Brus, T. F. Heinz, J. Hone, S. Ryu, ACS Nano
2010, 4, 2695.
[32] P. Nemes-Incze, Z. Osváth, K. Kamarás, L. P. Biró, Carbon 2008, 46,
1435.
[33] Y. Lv, J. Cui, Z. M. Jiang, X. Yang, Nanoscale Res. Lett. 2012, 7, 659.
[34] R. B. Narsimha, K. P. Naresh, D. Melepurath, ChemPhysChem 2015,
16, 377.
[35] L. Yan, Y. Song, Y. Zhou, B. Song, Y. Li, Org. Electron. 2015, 17, 94.
[36] N. M. Ravindra, J. Zhao, Smart Mater. Struct. 1992, 1, 197.
[37] W. Jaegermann, C. Pettenkofer, B. A. Parkinson, Phys. Rev. B 1990,
42, 7487.
[38] W. Chen, E. J. G. Santos, W. Zhu, E. Kaxiras, Z. Zhang, Nano Lett.
2013, 13, 509.
[39] I. Popov, G. Seifert, D. Tománek, Phys. Rev. Lett. 2012, 108,
156802.
[40] S. G. Louie, M. L. Cohen, Phys. Rev. Lett. 1975, 35, 866.
[41] C. Gong, L. Colombo, R. M. Wallace, K. Cho, Nano Lett. 2014, 14,
1714.
[42] Y. Son, Q. H. Wang, J. A. Paulson, C.-J. Shih, A. G. Rajan, K. Tvrdy,
S. Kim, B. Alfeeli, R. D. Braatz, M. S. Strano, ACS Nano 2015, 9,
2843.
[43] Y. Li, X. Li, T. Yu, G. Yang, H. Chen, C. Zhang, Q. Feng, J. Ma,
W. Liu, H. Xu, Y. Liu, X. Liu, Nanotechnology 2018, 29, 124001.
[44] A. V. Stier, N. P. Wilson, G. Clark, X. Xu, S. A. Crooker, Nano Lett.
2016, 16, 7054.
[45] Z. Wang, Q. Li, Y. Chen, B. Cui, Y. Li, F. Besenbacher, M. Dong,
NPG Asia Mater. 2018, 10, 703.
[46] Q. Peng, Z. Wang, B. Sa, B. Wu, Z. Sun, Sci. Rep. 2016, 6, 31994.
[47] P. Chakraborty, B. Roy, P. Bairi, A. K. Nandi, J. Mater. Chem. 2012,
22, 20291.
[48] T. Akama, W. Okita, R. Nagai, C. Li, T. Kaneko, T. Kato, Sci. Rep.
2017, 7, 11967.
[49] S. B. Desai, G. Seol, J. S. Kang, H. Fang, C. Battaglia, R. Kapadia,
J. W. Ager, J. Guo, A. Javey, Nano Lett. 2014, 14, 4592.
[50] E. Courtade, M. Semina, M. Manca, M. M. Glazov, C. Robert,
F. Cadiz, G. Wang, T. Taniguchi, K. Watanabe, M. Pierre,
W. Escoffier, E. L. Ivchenko, P. Renucci, X. Marie, T. Amand,
B. Urbaszek, Phys. Rev. B 2017, 96, 085302.
[51] S. Mouri, Y. Miyauchi, K. Matsuda, Nano Lett. 2013, 13, 5944.
[52] W. Huang, X. Luo, C. K. Gan, S. Y. Quek, G. Liang, Phys. Chem.
Chem. Phys. 2014, 16, 10866.
[53] N. R. Wilson, P. V. Nguyen, K. Seyler, P. Rivera, A. J. Marsden,
Z. P. L. Laker, G. C. Constantinescu, V. Kandyba, A. Barinov,
N. D. M. Hine, X. Xu, D. H. Cobden, Sci. Adv. 2017, 3,
e1601832.
[54] H. Li, G. Lu, Y. Wang, Z. Yin, C. Cong, Q. He, L. Wang, F. Ding,
T. Yu, H. Zhang, Small 2013, 9, 1974.
Adv. Mater. Interfaces 2019, 6, 1900637
A preview of this full-text is provided by Wiley.
Content available from Advanced Materials Interfaces
This content is subject to copyright. Terms and conditions apply.
