ARTICLE

Received 20 Feb 2017 | Accepted 14 Mar 2017 | Published 4 May 2017 DOI: 10.1038/ncomms15251 OPEN

Coulomb engineering of the bandgap and excitons

in two-dimensional materials
Archana Raja1,2,3,4, Andrey Chaves2,5, Jaeeun Yu2, Ghidewon Arefe6, Heather M. Hill1,3, Albert F. Rigosi1,3,
Timothy C. Berkelbach7, Philipp Nagler8, Christian Schüller8, Tobias Korn8, Colin Nuckolls2, James Hone6,
Louis E. Brus2, Tony F. Heinz1,3,4, David R. Reichman2 & Alexey Chernikov1,8

The ability to control the size of the electronic bandgap is an integral part of solid-state
technology. Atomically thin two-dimensional crystals offer a new approach for tuning the
energies of the electronic states based on the unusual strength of the Coulomb interaction in
these materials and its environmental sensitivity. Here, we show that by engineering the
surrounding dielectric environment, one can tune the electronic bandgap and the exciton
binding energy in monolayers of WS2 and WSe2 by hundreds of meV. We exploit this
behaviour to present an in-plane dielectric heterostructure with a spatially dependent
bandgap, as an initial step towards the creation of diverse lateral junctions with nanoscale
resolution.

1 Departments of Physics and Electrical Engineering, Columbia University, New York, New York 10027, USA. 2 Department of Chemistry, Columbia University,

New York, New York 10027, USA. 3 Department of Applied Physics, Stanford University, Stanford, California 94305, USA. 4 SLAC National Accelerator
Laboratory, Menlo Park, California 94025, USA. 5 Departamento de Fı́sica, Universidade Federal do Ceará, Caixa Postal 6030, Ceará, Fortaleza 60455-760,
Brazil. 6 Department of Mechanical Engineering, Columbia University, New York, New York 10027, USA. 7 Department of Chemistry and James Franck
Institute, University of Chicago, Chicago, Illinois 60637, USA. 8 Department of Physics, University of Regensburg, Regensburg D-93040, Germany.
Correspondence and requests for materials should be addressed to A.R. (email: [email protected]) or to A.C. (email: [email protected]).

NATURE COMMUNICATIONS | 8:15251 | DOI: 10.1038/ncomms15251 | www.nature.com/naturecommunications 1

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms15251

T
he precise and efficient manipulation of electrons in and hexagonal boron nitride above and below monolayers (1L) of
solid-state devices has driven remarkable progress across WS2 and WSe2, we achieve tuning of the electronic quasiparticle
fields from information processing and communication bandgap, as well as of the exciton binding energy of the two
technology to sensing and renewable energy. The ability to TMDC monolayers by several 100’s of meV. We note that
engineer the electronic bandgap is crucial to these applications1. graphene is particularly well-suited to demonstrate and explore
Several methods currently exist to tune a material’s bandgap by the concept of dielectric heterostructures. It combines a high
altering, for example, its chemical composition, spatial extent dielectric screening with the possibility of adding an arbitrary
(quantum confinement), background doping or lattice constant number of additional layers as thin as only 3 Å. Furthermore, the
via mechanical strain2. Such methods are typically perturbative TMDC/graphene structures have been heavily studied recently in
in nature and not suitable for making arbitrarily shaped, a variety of contexts with potential applications in optoelectronics
atomically sharp variations in the bandgap. Consequently, there and photovoltaics39–42. Screening is found to be maximized for
is a motivation to approach this important problem from a fresh just a few layers of graphene as the surrounding dielectric,
perspective. suggesting that Coulomb-engineered bandgaps can be realized
The emerging class of atomically thin two-dimensional (2D) with a spatial resolution on the nanoscale. Moreover, an in-plane
materials derived from bulk van der Waals crystals offers an heterostructure with a spatially dependent electronic bandgap is
alternative route to bandgap engineering. Within the family shown to exhibit a potential well on the order of more than
of 2D materials, much recent research has focused on 100 meV. Our results are supported by calculations employing a
the semiconducting transition-metal dichalcogenides (TMDCs)— quantum mechanical Wannier exciton model21. The dielectric
MX2 with M ¼ Mo, W and X ¼ S, Se, Te3. In the monolayer limit, screening leading to the bandgap renormalization can be
these TMDCs are direct-bandgap semiconductors with the optical treated in a semiclassical electrostatic framework that accounts
gap in the visible and near-infrared spectral range4,5. They for the underlying substrate and the nanostructured dielectric
combine strong inter- and intraband light-matter coupling6,7 environment, which we computed using a recently developed
with intriguing spin-valley physics8–10, high charge carrier quantum electrostatic heterostructure approach30.
mobilities11,12, ready modification of the in-plane material
structure13–16 and seamless integration into a variety of van der
Waals heterostructures17. Results
Importantly, the Coulomb interactions between charge carriers Coulomb engineering of monolayer WS2. An optical micro-
in atomically thin TMDCs are remarkably strong18–21. This leads graph of a typical sample, 1L WS2 partially covered with bilayer
to a significant renormalization of the electronic energy levels and (2L) graphene, is presented in Fig. 1b. To monitor the quasi-
increase in the quasiparticle bandgap. The Coulomb interactions particle bandgap of the material, we first identify the energies of
are also reflected in the binding energies of excitons, that is, the excitonic resonances in different dielectric environments
bound electron–hole pairs2, that are more than an order of using optical reflectance spectroscopy. The relationship between
magnitude greater in TMDC monolayers than in typical exciton Rydberg states and the electronic bandgap is shown in the
inorganic semiconductors22–26. The strength of the Coulomb schematic illustration of the optical response of a 2D semi-
interaction in these materials originates from weak dielectric conductor in Fig. 1c. The Coulomb attraction between electrons
screening in the 2D limit21,27,28. For distances exceeding few and holes leads to the emergence of bound exciton states below
nanometres, the screening is determined by the immediate the quasiparticle bandgap2,43,44, which are labelled according to
surroundings of the material, which can be vacuum or air in the their principal quantum number n ¼ 1, 2, 3, y, analogous to the
ideal case of suspended samples. More generally, the interaction states of the hydrogen atom. (Throughout the rest of the
between charge carriers is highly sensitive to the local dielectric manuscript we omit the term quasiparticle for clarity of
environment23,24,26–32, as seen in measured changes of the presentation.) The difference between the bandgap Egap and the
exciton Bohr radius33 and in theoretical analysis of the exciton resonance energies defines the respective exciton binding
environmental screening34,35. Correspondingly, both the energies. In particular, the energy D12 between the exciton ground
electronic bandgap and the exciton binding energy are expected state (n ¼ 1) and the first excited state (n ¼ 2) scales with the
to be tunable by means of a deliberate change of this ground state exciton binding energy EB. Along with the
environment, as illustrated in Fig. 1a, like the influence of a experimentally determined transition energy E1 of the exciton
solvent on the properties of molecules, quantum dots, carbon ground state, we can determine the electronic bandgap via
nanotubes and other nanostructures suspended in solution36–38. Egap ¼ E1 þ EB.
In addition, passivated and chemically inert van der Waals Typical linear reflectance contrast spectra, DR/R ¼ (Rsample
surfaces allow atomically thin layers to be brought into close  Rsubstrate)/Rsubstrate, of the bare 2L graphene, 1L WS2 and the
proximity while still retaining the intrinsic properties resulting heterostructure at T ¼ 70 K are presented in Fig. 1d. For
and functionality of the individual components17. These such ultra-thin layers with moderate reflectance contrast signals
observations motivate a programme to explore the concept of on transparent substrates, the quantity DR/R is predominantly
Coulomb engineering of the bandgap by local changes in the determined by the imaginary part of the dielectric function, which
dielectric environment. This strategy offers a means of locally is proportional to the optical absorption45,46. In the spectral
tuning the energies of the electronic states in 2D materials, even region shown, the response of 1L WS2 is dominated by the
allowing in-plane heterostructures down to nanometre length creation of so-called A excitons at the fundamental optical
scales34. As a result, this approach not only effectively transition in the material, at the K and K0 points of the hexagonal
demonstrates the validity of fundamental physics with respect Brillouin zone. In particular, the ground state (n ¼ 1) excitonic
to the Coulomb interaction in atomically thin systems, but offers resonance occurs at 2.089 eV. The first excited state n ¼ 2 appears
a viable opportunity to directly harness these many-body as a smaller spectral feature at 2.245 eV, with an energy separation
phenomena for future technology. between the two states of D12 ¼ 156 meV. In addition, the first
In this report, we provide direct experimental demonstration derivatives of DR/R are presented in Fig. 1e, where the spectral
of control of the bandgap and exciton binding energy in 2D region in the range of the n ¼ 1 state is scaled by factor 0.03 for
materials using Coulomb engineering through the modification of better comparison. Here, the energies of the peaks correspond to
the local dielectric environment. By placing layers of graphene the points of inflection of the asymmetric derivative features, as

2 NATURE COMMUNICATIONS | 8:15251 | DOI: 10.1038/ncomms15251 | www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms15251 ARTICLE

a b d e
1 2 n=1
1.5 n=1
2D

1st derivative refl. contrast (offset)

+ – + – 2L
TMDC 1L WS2 &
graphene 2L graphene n=2
Egap

Reflectance contrast
1L WS2

2
S
W
2L graphene
5 µm 1.0 × 0.03 12

1L
c Exciton, n = 1
Optical
response 0.5
12
Bandgap Egap n=2
n=2 × 0.03
n = 3, 4, ...
12

+ – + – 0.0
2.0 2.1 2.2 2.3 2.0 2.1 2.2 2.3
Binding energy EB Energy
Energy (eV) Energy (eV)

Figure 1 | Engineering Coulomb interactions through environmental screening. (a) Schematic illustration of a semiconducting 2D TMDC material,
partially covered with an ultra-thin dielectric layer. The strong Coulomb interaction between charged particles in low-dimensional systems affects both the
exciton binding energy and the quasiparticle bandgap. The interaction can be strongly modified by modulating the environmental dielectric screening on
atomic length scales. (b) An optical micrograph of the heterostructure under study: monolayer WS2 covered with a bilayer of graphene. Dotted circles
indicate positions for the optical measurements. (c) Illustration of the optical response of an ideal 2D semiconductor, including exciton ground and excited
state resonances and the onset of the (quasiparticle) bandgap. (d) Reflectance contrast spectra of the bare bilayer graphene, monolayer WS2 and the
resulting WS2/graphene heterostructure at a temperature of 70 K. (e) First derivatives of the reflectance contrast spectra in d (after averaging over a
20 meV interval), offset for clarity. Peak positions of the exciton ground state (n ¼ 1) and the first excited state (n ¼ 2) resonances, roughly corresponding
to the points of inflection, are indicated by dashed lines; D12 denotes the respective energy separations. The observed decrease of D12 across the in-plane
boundary of the heterostructure is indicative of a reduction of the exciton binding energy and bandgap by more than 100 meV due to the presence of the
adjacent graphene bilayer.

indicated by dashed lines for the n ¼ 2 states. Finally, the shoulder Nanoscale sensitivity of Coulomb engineering. The spatial
on the low-energy side of the n ¼ 1 peak at 2.045 eV arises from extent of the modulation is an important aspect of our approach
charged excitons, indicating slight residual doping in the WS2 to dielectric bandgap engineering. We have been able to probe
material47,48. Overall, the 1L WS2 response matches our previous this issue spectroscopically with sub-nanometre precision. To do
observations on uncapped samples supported on fused silica24,47, so, we track the change in the WS2 bandgap for dielectric
consistent with an exciton binding energy on the order of screening when the semiconductor is capped by 1, 2 or 3 layer
300 meV. For bilayer graphene, we recover the characteristic flat graphene (Supplementary Note 3 and Supplementary Fig. 5). The
reflectance contrast over the relevant spectral range49. extracted exciton peak separation energy D12 and the corre-
In case of WS2 capped with graphene, the overall reflectance sponding evolution of the bandgap are presented in Fig. 2a,b,
contrast is offset by the graphene reflectance, similar to findings respectively. Remarkably, we observe the strongest change already
in TMDC/TMDC heterostructures50. Most importantly, however, from the first graphene layer, which is followed by rapid
we observe pronounced shifts of the WS2 exciton resonances to saturation with increasing thickness within experimental uncer-
lower energies, where the n ¼ 1 transition and the n ¼ 2 states are tainty. This result strongly suggests that the change in bandgap
now located at 2.060 eV and 2.167 eV, respectively (Fig. 1d,e). The should also occur on a similar ultra-short length scale at the in-
corresponding decrease of D12 from 156 to 107 meV is indicative plane boundary of the uncapped and graphene-capped WS2,
of a strong reduction in the exciton binding energy and bandgap. consistent with predictions from ref. 34.
In particular, the absolute shift of the n ¼ 2 state by almost For a more precise analysis of our findings we turn to a
70 meV defines the minimum expected decrease in the bandgap. Wannier-like exciton model21, where the non-local screening of
More quantitatively, by assuming a similar non-hydrogenic the electron–hole Coulomb interaction leads to environmental
scaling like that in ref. 24, that is, EB ¼ 2D12 , the reduction in sensitivity. To atomistically handle complex dielectric
exciton binding energy is estimated to be on the order of environments, we employ the recently introduced quantum
100 meV, from 312 meV in bare WS2 to 214 meV in WS2 capped electrostatic heterostructure (QEH) approach presented in ref.
by 2L graphene. From Egap ¼ E1 þ EB, we infer a bandgap for bare 30. Within this model, the electrostatic potential between
WS2 of 2.40 eV, reducing to 2.27 eV in the WS2/graphene electrons and holes confined to a 2D layer can be obtained for
heterostructure. We thus see a 130 meV decrease in the nearly arbitrary vertical heterostructures, taking into account the
bandgap energy from the presence of the capping layer. precise alignment of the individual materials and the resulting
To understand these experimental findings more intuitively, we spatially dependent dielectric response. The exciton states are
recall that although the excitons are confined to the WS2 layer, the subsequently calculated by solving the Wannier equation in the
electric field between the constituent electrons and holes permeates effective mass approximation with an exciton reduced mass of
both the material and the local surroundings (Fig. 1a). In particular, 0.16 m0 as obtained from ab initio calculations21. To account for
the screening for larger electron–hole separations is increasingly the dielectric screening from the environment, mainly through
dominated by the dielectric properties of the environment. the underlying fused silica substrate and potential adsorbates such
Therefore, the strength of the Coulomb interaction is reduced by as water, we adjust the effective dielectric screening below the
the addition of graphene layers on top of WS2, leading to a decrease 2D layer, resulting in D12 ¼ 188 meV and EB ¼ 289 meV, roughly
in both the exciton binding energy and the bandgap. matching experimental observations. Then, additional graphene

NATURE COMMUNICATIONS | 8:15251 | DOI: 10.1038/ncomms15251 | www.nature.com/naturecommunications 3

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms15251

a screening from adjacent graphene layers. The model also agrees

with a classical electrostatic screening theory for the limiting cases
200 of an uncapped WS2 monolayer on fused silica and for a layer
Energy separation 12 (meV)

Experiment fully covered with bulk graphite on top, the results of which are
QEH calculation indicated by dashed lines in Fig. 2a (see Supplementary Note 2 for
On SiO2 details). The calculated exciton binding energy changes from
160
On graphite
290 meV for uncapped WS2 to 120 meV for the case of a trilayer
graphene heterostructure. As previously discussed, the binding
energies together with the absolute energies of the exciton ground
120 state resonances can be used to infer the size of the bandgap. The
evolution of the bandgap and the corresponding n ¼ 1 and n ¼ 2
exciton transition energies are presented in Fig. 2b. The binding
80 energies obtained from the QEH model are compared with
experimentally determined limits from the relation EBpD12 by
0 1 2 3
assuming a non-hydrogenic scaling EB ¼ 2D12 as was observed
Number of graphene layers on WS2
for a single WS2 layer on SiO2 (ref. 24) or conventional
b 2D hydrogenic scaling with EB ¼ 9=8D12 for a homogeneous
dielectric. These two relations provide, respectively, boundaries
Bandgap: Exciton states: for the scaling in generic heterostructures of 1L TMDCs
from EB = 212 n=1 embedded in a dielectric environment with higher dielectric
2.4
from EB = 9/812 n=2 screening than the SiO2 support and lower dielectric screening
QEH calc. than the corresponding bulk crystals. In general, the scaling of
EB with D12 converges towards the 2D-hydrogen model as the
2.3 screening of the surroundings approaches that of the bulk
Energy (eV)

TMDC. For the case of trilayer graphene, this simple estimate

implies a bandgap reduction of at least 150 meV and at most
230 meV.
2.2

Flexibility of material systems and configurations. In addition

to the graphene-capped WS2 samples, a variety of hetero-
2.1 structures were investigated in a similar manner. These include
1L WS2 encapsulated between two graphene layers, graphene-
capped 1L WSe2, graphene-supported 1L WSe2 and 1L WSe2 on
an 8 nm layer of hexagonal boron nitride (hBN). In all cases, a
0 1 2 3
decrease in D12 separation was observed with increasing dielectric
Number of graphene layers on WS2
screening of the environment (see Supplementary Notes 4
Figure 2 | Out-of-plane spatial sensitivity of environmental screening. and 5 including Supplementary Figs 6 and 7 as well as the
(a) Experimentally and theoretically obtained energy separation D12 Supplementary Table 1 for individual reflectance spectra and
between the n ¼ 1 and n ¼ 2 exciton states as a function of the number of additional sample details). A summary of the results is presented
layers of capping graphene. Dashed lines indicate D12 values from the in Fig. 3a, including experimentally obtained n ¼ 1 and n ¼ 2
solution of the electrostatic model for uncapped WS2 supported by fused transition energies, as well as the corresponding shifts of the
silica substrate (grey) and covered with bulk graphite (red), representing bandgap, estimated as above (see also Supplementary Fig. 8 for
two ideal limiting cases. (b) Absolute energies of the experimentally the shift of the B exciton states in WSe2). The bandgap of WSe2
measured exciton ground state (n ¼ 1) and the first excited state (n ¼ 2) can be thus tuned by more than 100 meV and the largest shift of
resonances, as well as the estimated positions of the bandgap obtained by almost 300 meV is observed for graphene-encapsulated WS2, the
adding the exciton binding energy to the energy of the n ¼ 1 state. The structure with the highest dielectric screening. For comparison,
binding energy scales with D12, where the limiting cases are an the influence of an arbitrary dielectric environment is presented
experimentally determined non-hydrogenic scaling for WS2 on SiO2 in Fig. 3b, which shows the calculated exciton binding energy of
substrate from ref. 24 ðEB ¼2D12 Þ and the 2D-hydrogen model in a 1L WS2 encapsulated between two thick layers of varying
homogeneous dielectric ðEB ¼9=8D12 Þ. These are compared to the bandgap dielectric constants. As we have shown, the change in the
energies deduced from the calculated exciton binding energies using the bandgap is roughly the same as the change in the binding energy
QEH model. The solid lines are guides to the eye. and thus can be as high as 500 meV (corresponding to the
intrinsic value of the exciton binding energy for a sample
suspended in vacuum).
layers are added on top of the WS2 monolayer with all parameters
being fixed. The interlayer separation between WS2 and graphene In-plane dielectric heterostructure. Finally, we demonstrate an
is set to 0.5 nm, corresponding to the average of the interlayer in-plane 2D semiconductor heterostructure with a spatially
separations for the materials in literature51,52. dependent bandgap profile by constructing a spatially varying
The theoretically predicted energy separation D12 is plotted in dielectric environment surrounding the semiconductor. We scan
Fig. 2a as a function of the number of graphene layers and across the structure (cf. Fig. 1b) through regions of bare WS2 and
compared to experiment. The calculations reproduce both the WS2 covered by a bilayer of graphene. The corresponding path is
abrupt change and the subsequent saturation of D12 with illustrated schematically in Fig. 4c and in the inset. First-order
graphene thickness. Furthermore, the absolute energy values are derivatives of the reflectance contrast spectra are presented in
in semi-quantitative agreement with the measurements, support- Fig. 4a,b in the spectral range of the WS2 exciton n ¼ 1 and n ¼ 2
ing the attribution of the measured change of D12 to the dielectric resonances, respectively. Each spectral trace corresponds to a

4 NATURE COMMUNICATIONS | 8:15251 | DOI: 10.1038/ncomms15251 | www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms15251 ARTICLE

a a b
Bandgap: from EB = 212 Exciton n=1
2.1 n=1 n=2 Pos.
2.4 from EB = 9/812 states: n=2
(µm)

Reflectance contrast derivative (a.u.)

13.8
Energy (eV)

2.0

Energy (eV)
2.3 ×100
13.2
12.5
2.2 1.9
12.2
11.6
2.1 1.8 10.6
9.7
70 K 5K 1.7 8.4
2.0
7.1
5.7
4.8
3.0
1.8
0.3
0
1L: WS2 WSe2 Graphene
2.00 2.05 2.10 2.15 2.15 2.20 2.25
Thick: SiO2 hBN Energy (eV) Energy (eV)

EB (meV) 1 2 1
b c 2L graphene
20 <25
Top 1L WS2
100
1L WS2 2.35

Bandgap (eV)
10 Bottom
8 2.30
200
6 2.25
Top

300 2.20
4
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Position (µm)
400
2
Figure 4 | In-plane heterostructure via Coulomb engineering of
500 monolayer WS2. (a) First-order derivatives of the reflectance contrast of a
1L WS2 sample for varying spatial positions across the lateral 1L WS2/2L
1
1 2 4 6 8 10 20 graphene boundary. The data are shown in the spectral range of the exciton
Bottom ground state (n ¼ 1) resonance and vertically offset for clarity. (b) For the
spectral range of the excited state (n ¼ 2) of the exciton with the vertical
Figure 3 | Influence of the choice and configuration of materials on the axis scaled by factor of 100 for direct comparison. Full circles in a,b indicate
dielectric tuning of the bandgap. (a) Experimentally measured exciton peak energies of the resonances, corresponding to the points of inflection of
ground state (n ¼ 1) and the first excited state (n ¼ 2) transition energies, the derivatives. (c) Spatially dependent bandgap energy extracted from the
as well as the estimated shifts of the bandgap for a variety of exciton peak positions along the profile of the lateral WS2/graphene
heterostructures. Their respective stacking configurations are indicated heterostructure, as illustrated in the schematic representation and marked
along the horizontal axis. The bandgap is obtained by adding the exciton by the dashed line in the optical micrograph. The shaded areas indicate the
binding energy to the measured transition energy of the n ¼ 1 state. To diffraction limit corresponding to the spatial resolution of our measurement
estimate the binding energy from the energy separation of the exciton and the solid line is a guide to the eye.
states D12, we considered the limiting cases of a non-hydrogenic scaling
from ref. 24 ðEB ¼2D12 Þ and the 2D-hydrogen model ðEB ¼9=8D12 Þ. (b) An
overview of predicted changes in the exciton binding energy in 1L WS2, The spatial dependence is presented in Fig. 4c along the path
encapsulated between two thick layers of dielectrics. The binding energy EB marked in the optical micrograph (inset), which includes two
is calculated by using the electrostatic approach in the effective mass WS2/graphene in-plane junctions. As previously discussed, the
approximation and presented in a 2D false-colour plot as a function of the induced energy shifts result in an overall decrease of the relative
top and bottom dielectric constants. The changes in the magnitude of EB are energy separation D12 from about 160 meV, down to 105 meV.
roughly equal to the corresponding shifts of the bandgap and can reach Here, the binding energy is extracted by multiplying D12 with the
500 meV. scaling factor deduced from the QEH calculations presented in
Fig. 2 (1.54 and 1.40 for the bare and 2L graphene-covered
sample, respectively) to obtain the bandgap at each point.
different spatial position x on the sample; the bilayer graphene The resulting bandgap profile is representative of a potential well
flake covers the WS2 monolayer between 5 and 12 mm on the x (graphene-covered area) surrounded by two adjacent barriers at
axis. Like the data shown in Fig. 1d,e, both the ground and excited higher energies (bare sample). Model self-energy calculations on
state resonances of the WS2 excitons shift to lower energies in the monolayer TMDCs in structured dielectric environments34
presence of graphene. The peak energies are extracted from the suggest that the interface between the uncapped and capped
points of inflection of the derivative, indicated by circles in regions should yield an in-plane type-II heterostructure. In
Fig. 4a,b. The appearance of multiple transitions in the same particular, the areas capped by graphene are expected to have a
spectrum reflects the limited spatial resolution (1 mm) and a small higher local valence band that acts as a potential well for holes.
amount of the WS2 monolayer not being in close contact with The dielectric effect on the conduction band is predicted to be
graphene (see Supplementary Note 1 and Supplementary Fig. 2 weaker, with a slightly higher energy for the capped regions
for details). leading to a small barrier for electron flow from the bare to

NATURE COMMUNICATIONS | 8:15251 | DOI: 10.1038/ncomms15251 | www.nature.com/naturecommunications 5

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms15251

capped regions. Since the overall energy shifts of the bandgap are 2. Klingshirn, C. Semiconductor Optics 3rd edn (Springer, 2007).
larger than thermal energy at room temperature, our results 3. Novoselov, K. S. et al. Two-dimensional atomic crystals. Proc. Natl Acad. Sci.
render the observed phenomenon technologically promising for USA 102, 10451–10453 (2005).
applications under ambient or even high-temperature conditions. 4. Splendiani, A. et al. Emerging photoluminescence in monolayer MoS2. Nano
Lett. 10, 1271–1275 (2010).
5. Mak, K. F., Lee, C., Hone, J., Shan, J. & Heinz, T. F. Atomically thin MoS2:
Discussion a new direct-gap semiconductor. Phys. Rev. Lett. 105, 136805 (2010).
We have demonstrated a new approach to the engineering of 6. Zhang, C., Wang, H., Chan, W., Manolatou, C. & Rana, F. Absorption of light
by excitons and trions in monolayers of metal dichalcogenide MoS2:
electronic properties through local dielectric screening of the
experiments and theory. Phys. Rev. Lett. 89, 205436 (2014).
Coulomb interaction in 2D heterostructures. We have shown 7. Poellmann, C. et al. Resonant internal quantum transitions and femtosecond
tuning of the bandgap and exciton binding energy in monolayers radiative decay of excitons in monolayer WSe2. Nat. Mater. 14, 889–893
of WS2 and WSe2 for a variety of combinations with graphene (2015).
and hBN layers. The overall shift of the bandgap ranged from 100 8. Xu, X., Yao, W., Xiao, D. & Heinz, T. F. Spin and pseudospins in layered
to 300 meV, with an estimated theoretical limit of about 500 meV. transition metal dichalcogenides. Nat. Phys. 10, 343–350 (2014).
9. Yu, H., Cui, X., Xu, X. & Yao, W. Valley excitons in two-dimensional
In addition, the saturation of the screening effect with the semiconductors. Natl Sci. Rev. 2, 57–70 (2015).
thickness of the dielectric layer is found both in theory and 10. Stier, A. V., McCreary, K. M., Jonker, B. T., Kono, J. & Crooker, S. A. Exciton
experiment to occur on a nanometre length scale. We have diamagnetic shifts and valley Zeeman effects in monolayer WS2 and MoS2 to
demonstrated the flexibility of the technique by examining a 65 Tesla. Nat. Commun. 7, 10643 (2016).
variety of material combinations including WS2, WSe2, graphene 11. Jariwala, D., Sangwan, V. K., Lauhon, L. J., Marks, T. J. & Hersam, M. C.
Emerging device applications for semiconducting two-dimensional transition
and hBN in several distinct configurations, with top and bottom metal dichalcogenides. ACS Nano 8, 1102–1120 (2014).
alignment as well as in a sandwich-type structure. We emphasize 12. Cui, X. et al. Multi-terminal transport measurements of MoS2 using a
that the screening effect is not restricted to any particular choice van der Waals heterostructure device platform. Nat. Nanotechnol. 10, 534–540
of a capping material. Finally, we demonstrated Coulomb (2015).
engineering of a prototypical in-plane dielectric heterostructure, 13. Gong, Y. et al. Vertical and in-plane heterostructures from WS2/MoS2
monolayers. Nat. Mater. 13, 1135–1142 (2014).
illustrating the feasibility of our approach. As a consequence, we
14. Huang, C. et al. Lateral heterojunctions within monolayer MoSe2-WSe2
expect that patterning of dielectric layers on top of these ultra- semiconductors. Nat. Mater. 13, 1096–1101 (2014).
thin semiconductors or placing the latter on a prefabricated 15. Kappera, R. et al. Phase-engineered low-resistance contacts for ultrathin MoS2
substrate will allow us to explore a variety of novel devices in the transistors. Nat. Mater. 13, 1128–1134 (2014).
2D plane. In addition to the impact for more conventional 16. Guo, Y. et al. Probing the dynamics of the metallic-to-semiconducting
optoelectronic devices—such as transistors, light emitters and structural phase transformation in MoS2 crystals. Nano Lett. 15, 5081–5088
(2015).
detectors— one can envision custom-made superstructures for 17. Geim, A. K. & Grigorieva, I. V. Van der Waals heterostructures. Nature 499,
2D layers that permit integration with photonic cavities, 419–425 (2013).
plasmonic nanomaterials and quantum emitters for the creation 18. Cheiwchanchamnangij, T. & Lambrecht, W. R. L. Quasiparticle band structure
of new hybrid technologies. The considerable strength of the calculation of monolayer, bilayer, and bulk MoS2. Phys. Rev. B 85, 205302
Coulomb forces in atomically thin materials is thus not only of (2012).
19. Ramasubramaniam, A. Large excitonic effects in monolayers of molybdenum
fundamental importance, but also offers a strategy towards and tungsten dichalcogenides. Phys. Rev. B 86, 115409 (2012).
deterministic engineering of bandgaps in the 2D plane. 20. Qiu, D. Y., da Jornada, F. H. & Louie, S. G. Optical spectrum of MoS2:
many-body effects and diversity of exciton states. Phys. Rev. Lett. 111, 216805
Methods (2013).
Sample preparation. Monolayers of WS2 and WSe2, mono- and few-layer gra- 21. Berkelbach, T. C., Hybertsen, M. S. & Reichman, D. R. Theory of neutral and
phene, and hBN samples were produced by mechanical exfoliation of bulk crystals charged excitons in monolayer transition metal dichalcogenides. Phys. Rev. B
(2Dsemiconductors and HQgraphene). The thickness of the layers was confirmed 88, 045318 (2013).
by optical contrast spectroscopy. The heterostructures were fabricated using 22. Zhang, C., Johnson, A., Hsu, C.-L., Li, L.-J. & Shih, C.-K. Direct imaging of
well-established polymer-stamp transfer techniques described in refs 50,53 for the band profile in single layer MoS2 on graphite: quasiparticle energy gap, metallic
WS2 based samples and ref. 54 for the WSe2 samples (see Supplementary Note 2 edge states, and edge band bending. Nano Lett. 14, 2443–2447 (2014).
and Supplementary Fig. 1 for additional details). 23. He, K. et al. Tightly bound excitons in monolayer WSe2. Phys. Rev. Lett. 113,
026803 (2014).
24. Chernikov, A. et al. Exciton binding energy and nonhydrogenic Rydberg series
Optical spectroscopy. To study the exciton states, we performed optical in monolayer WS2. Phys. Rev. Lett. 113, 076802 (2014).
reflectance measurements using a tungsten-halogen white-light source. The light 25. Ye, Z. et al. Probing excitonic dark states in single-layer tungsten disulphide.
was focused to a 1–2 mm spot on the sample for the measurements on WS2, and to Nature 513, 214–218 (2014).
a 5–10 mm spot for the measurements on WSe2 due to larger sample sizes. The 26. Ugeda, M. M. et al. Giant bandgap renormalization and excitonic effects in a
samples were kept in an optical cryostat at temperatures around 70 K and 4 K for
monolayer transition metal dichalcogenide semiconductor. Nat. Mater. 13,
the WS2 and WSe2 samples, respectively. The reflected light was spectrally resolved
1091–1095 (2014).
in a grating spectrometer and subsequently detected by a CCD (see Supplementary
27. Keldysh, L. V. Coulomb interaction in thin semiconductor and semimetal films.
Notes 1 and 3 and Supplementary Fig. 5 for additional details of the experimental
measurements and analysis procedures). JETP Lett. 29, 658–661 (1979).
28. Cudazzo, P., Tokatly, I. V. & Rubio, A. Dielectric screening in two-dimensional
insulators: implications for excitonic and impurity states in graphane. Phys.
Theoretical methods. Exciton binding energies were calculated within the Rev. B 84, 085406 (2011).
Wannier–Mott model, with an exciton reduced mass obtained from density 29. Lin, Y. et al. Dielectric screening of excitons and trions in single-layer MoS2.
functional theory (DFT) calculations21. The electron–hole-screened Coulomb Nano Lett. 14, 5569–5576 (2014).
interaction was obtained from the quantum electrostatic heterostructure 30. Andersen, K., Latini, S. & Thygesen, K. S. Dielectric genome of van der Waals
approach30 (Supplementary Note 2 and Supplementary Figs 3 and 4 for additional heterostructures. Nano Lett. 15, 4616–4621 (2015).
details). 31. Latini, S., Olsen, T. & Thygesen, K. S. Excitons in van der Waals
heterostructures: the important role of dielectric screening. Phys. Rev. B 92,
Data availability. The datasets generated during and/or analysed during the 245123 (2015).
current study are available from the corresponding author on reasonable request. 32. Kylänpää, I. & Komsa, H. P. Binding energies of exciton complexes in
transition metal dichalcogenide monolayers and effect of dielectric
environment. Phys. Rev. B 92, 205418 (2015).
References 33. Stier, A. V., Willson, N. P., Clark, G., Xu, X. & Crooker, S. A. Probing the
1. Capasso, F. Band-gap engineering: from physics and materials to new influence of dielectric environment on excitons in monolayer WSe2: insight
semiconductor devices. Science 235, 172–176 (1987). from high magnetic fields. Nano Lett. 16, 7054–7060 (2016).

6 NATURE COMMUNICATIONS | 8:15251 | DOI: 10.1038/ncomms15251 | www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms15251 ARTICLE

34. Rösner, M. et al. Two-dimensional heterojunctions from nonlocal Superstratic and Superatomic Solids, an NSF MRSEC (Award Number DMR-1420634),
manipulations of the interactions. Nano Lett. 16, 2322–2327 (2016). by FAME, one of six centers of STARnet, a Semiconductor Research Corporation
35. Trolle, M. L., Pedersen, T. G. & Véniard, V. Model dielectric function for 2D programme sponsored by MARCO and DARPA, as well as through the AMOS
semiconductors including substrate screening. Sci. Rep. 7, 39844 (2017). programme at SLAC National Accelerator Laboratory within the Chemical Sciences,
36. Brus, L. Size, dimensionality, and strong electron correlation in nanoscience. Geosciences and Biosciences Division for data analysis. Use of the Shared Materials
Acc. Chem. Res. 47, 2951 (2014). Characterization Laboratory (SMCL) was made possible by funding from Columbia
37. Walsh, A. G. et al. Screening of excitons in single, suspended carbon nanotubes. University. Device fabrication was supported by the Nanoelectronics and Beyond pro-
Nano Lett. 7, 1485–1488 (2007). gramme of the National Science Foundation (Grant DMR-1124894) and the Nanoe-
38. Malapanis, A., Jones, D. A., Comfort, E. & Lee, J. U. Measuring carbon lectronics Research Initiative of the Semiconductor Research Corporation. An.C.
nanotube band gaps through leakage current and excitonic transitions of acknowledges support from the Science Without Borders programme of the Brazilian
nanotube diodes. Nano Lett. 11, 1946–1951 (2011). National Research Council (CNPq) and the Lemann Foundation. J.Y. thanks the
39. Roy, K. et al. Graphene-MoS2 hybrid structures for multifunctional Kwanjeong Educational Foundation for support. H.M.H. and A.F.R. acknowledge
photoresponsive memory devices. Nat. Nanotechnol. 8, 826–830 (2013). funding from the National Science Foundation through the Integrated Graduate Edu-
40. Georgiou, T. et al. Vertical field-effect transistor based on graphene-WS2 cation and Research Training Fellowship (DGE-1069240) and the Graduate Research
heterostructures for flexible and transparent electronics. Nat. Nanotechnol. 8, Fellowship Programme (DGE-1144155), respectively. Al.C., T.K., P.N. and C.S. gratefully
100–103 (2013). acknowledge funding from the Deutsche Forschungsgemeinschaft through the Emmy
41. Bertolazzi, S., Krasnozhon, D. & Kis, A. Nonvolatile memory cells based on Noether Programme (CH 1672/1-1) and via GRK1570.
MoS2/grapheme heterostructures. ACS Nano 7, 3246–3252 (2013).
42. Bernardi, M., Palummo, M. & Grossman, J. C. Extraordinary sunlight
absorption and one nanometer thick photovoltaics using two-dimensional Author contributions
monolayer materials. Nano Lett. 13, 3664–3670 (2013). A.R. and Al.C. conceived the project and performed the experiments. A.R., Al.C. and
43. Haug, H. & Koch, S. W. Quantum Theory of the Optical and Electronic T.F.H. wrote the manuscript. P.N. and H.M.H. contributed towards optical spectroscopy;
Properties of Semiconductors 5th edn (World Scientific, 2009). A.R., P.N., J.Y., G.A., H.M.H. and A.F.R. prepared the samples. Theoretical calculations
44. Kazimierczuk, T., Fröhlich, D., Scheel, S., Stolz, H. & Bayer, M. Giant Rydberg were provided by An.C., T.C.B. and D.R.R. All authors commented on the manuscript
excitons in the copper oxide Cu2O. Nature 514, 343–347 (2014). and discussed the results.
45. Mak, K. F. et al. Measurement of the optical conductivity of graphene. Phys.
Rev. Lett. 101, 196405 (2008).
46. Li, Y. et al. Measurement of the optical dielectric function of monolayer
transition-metal dichalcogenides: MoS2, MoSe2, WS2, and WSe2. Phys. Rev. B Additional information
90, 205422 (2014). Supplementary Information accompanies this paper at http://www.nature.com/
47. Chernikov, A. et al. Electrical tuning of exciton binding energies in monolayer naturecommunications
WS2. Phys. Rev. Lett. 115, 126802 (2015).
48. Plechinger, G. et al. Identification of excitons, trions and biexcitons in Competing interests: The authors declare no competing financial interests.
single-layer WS2. Phys. Stat. Sol. RRL 9, 457–461 (2015). Reprints and permission information is available online at http://npg.nature.com/
49. Mak, K. F., Shan, J. & Heinz, T. F. Seeing many-body effects in single- and reprintsandpermissions/
few-layer graphene: observation of two-dimensional saddle-point excitons.
Phys. Rev. Lett. 106, 046401 (2011). How to cite this article: Raja, A. et al. Coulomb engineering of the bandgap and excitons
50. Rigosi, A. F., Hill, H. M., Li, Y., Chernikov, A. & Heinz, T. F. Probing interlayer in two-dimensional materials. Nat. Commun. 8, 15251 doi: 10.1038/ncomms15251
interactions in transition metal dichalcogenide heterostructures by optical (2017).
spectroscopy: MoS2/WS2 and MoSe2/WSe2. Nano Lett. 15, 5033–5038 (2015).
51. Gutiérrez, H. R. et al. Extraordinary room-temperature photoluminescence in Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in
triangular WS2 monolayers. Nano Lett. 13, 3447–3454 (2013). published maps and institutional affiliations.
52. Baskin, Y. & Meyer, L. Lattice constants of graphite at low temperatures. Phys.
Rev. 100, 544–544 (1955). This work is licensed under a Creative Commons Attribution 4.0
53. Lee, C.-H. et al. Atomically thin p-n junctions with van der Waals International License. The images or other third party material in this
heterointerfaces. Nat. Nanotechnol. 9, 676–681 (2014).
article are included in the article’s Creative Commons license, unless indicated otherwise
54. Castellanos-Gomez, A. et al. Deterministic transfer of two-dimensional
in the credit line; if the material is not included under the Creative Commons license,
materials by all-dry viscoelastic stamping. 2D Mater. 1, 011002 (2014).
users will need to obtain permission from the license holder to reproduce the material.
To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/
Acknowledgements
We would like to thank Simone Latini and Mark S. Hybertsen for fruitful discussions.
Funding for this research was provided in part by the Center for Precision Assembly of r The Author(s) 2017

NATURE COMMUNICATIONS | 8:15251 | DOI: 10.1038/ncomms15251 | www.nature.com/naturecommunications 7