Exciton dynamics and valley-contrasting properties in ... · Exciton dynamics and...
Transcript of Exciton dynamics and valley-contrasting properties in ... · Exciton dynamics and...
Exciton dynamics and valley-contrasting properties in heterostructures based on
atomically-thin semiconductors
Stéphane BERCIAUD IPCMS - CNRS - Université de Strasbourg - France
Workshop on Chiral Optical Modes, Aussois, Nov. 27, 2018
Tuesday March 19, 2018
Guillaume Froehlicher PhD student (2013-16)
Etienne Lorchat PhD student (2015-)
Acknowledgements
Stefano Azzini, Thibault Chervy Thomas Ebbesen, Cyriaque Genet
Cédric Robert, Delphine Lagarde, Xavier Marie
Takashi Taniguchi Kenji Watanabe
ISIS
Ajayan, Kim, Banerjee - Physics Today (2016)
2H-TMD (semiconductors) M= Mo, W X= S, Se, Te
M
X
X
C
Graphene (semimetal)
M
X
X
C1.7 1.8 1.9 2.0 2.1 2.2
0
1
360 370 380 390 400 410 420 430
0
1
Lum
inescence inte
nsity (
a.
u.)
Photon Energy (eV)
1L
3L
Bulk
E1
2g
Bulk MoS2
24.7 cm-1
3L MoS2
23.2 cm-1
Ram
an Inte
nsity (
a.
u.)
Raman Shift (cm-1)
1L MoS2
18.7 cm-1
A1g
10µm
20 µm 20 µm
a) Optical image b) Luminescence Image
MonolayerMoS2
BulkMoS2
c) d)
Entering “Flatland”
Direct bandgap emission
Helicity-dependent valley addressability
Ajayan, Kim, Banerjee - Physics Today (2016)
2H-TMD (semiconductors) M= Mo, W X= S, Se, Te
M
X
X
C
Graphene (semimetal)
M
X
X
C
Entering “Flatland”
Direct bandgap emission
Helicity-dependent valley addressability
0.9 1.0 1.1 1.2
5L
6L
7L
bulk
Ph
oto
lum
. (u
.a.)
Energy (eV)
1L
2L
3L
4L
MoTe2
PRB 2016
laser
Gr
BN
TMD
BN
Gr
WSe2
s± s±
interlayer charge and exciton transport
optoelectronics valleytronics nanomechanics
< 1 nm
|1, 𝒌
Massicotte et al., Nat Nano 2016 Mak et al., Science 2014 De Alba et al., Nat. Nano 2016
room T excitonic effects spin-valley locked properties
2D Materials: a unique toolkit for fundamental and applied physics
Nanophotonics, single photon emission
materials engineering (phase, anisotropy, strain)
electron-phonon coupling
K K’
MoX2 s+ s-
K K’
WX2 s+ s-
Spin-valley locking in monolayer TMDs
Broken inversion symmetry + strong spin orbit coupling
E
k K K’
WX2 s+ s-
K K’
MoX2 s+ s-
Valley-polarized excitons...and their coherent superpositions
G. Wang et al. RMP 2018 Z. Ye et al., Nat. Phys. 2016
s-
s+
MoS2
T=14KExc: s+
1.0
0.5
0.0
r
1.70 1.80 1.90 2.00 1.80 1.90 2.00Photon energy (eV) Photon energy (eV)
Ph
oto
lum
(arb
.)
c
2.0
MoS2 T=14 K, exc: s+, 1.96 eV
s+
s+
s-
MoS2
T=14K Exc: s-
1.0 0.5 0.0
r
1.70 1.80 1.90 2.00 1.80 1.90 2.00 Photon energy (eV) Photon energy (eV)
P
ho
tolu
m (
arb
.)
Fragile Valley Contrasts V
alle
y Po
lari
zati
on
(a.
u.)
c MoS2
s+
s-
MoS2
T=14KExc: s-
1.0
0.5
0.0
r
1.70 1.80 1.90 2.00 1.80 1.90 2.00Photon energy (eV) Photon energy (eV)
Ph
oto
lum
(arb
.)
K.F. Mak et al., Nat. Nano 2012
• r up to ~100 % at low T (MoS2) • g up to 60 % at low T (MoS2)
• Valley contrasts are lost at room T How to protect/tailor valley contrasts? C. R. Zhu et al., PRB 2014
r (g) degrees of circular (linear) polarization
𝜌 =𝜌0
1 + 2𝜏𝑋 𝜏𝑆
Outline
+_ Graphene/TMD heterostructures as a 2D optoelectronic building block
Room temperature valley polarization and coherence
in TMD/Graphene heterostructures
Room temperature Chiral coupling of valley excitons with spin-momentum locked
surface plasmons
Surface plasmons (SPs)
Strong interaction with light
+_< 1 nm
Semiconducting 2H-TMDs
M = Mo, W X = S, Se, Te
MX2
2D Matter meets 2D Light
Helicity-dependent valley addressability
M = Mo, W X = S, Se, Te
ISIS (Strasbourg) z
x
y
kz
ETM
SE
2
FIG. 2. (a) White light (WL) microscope image of the sample
and photoluminescence (PL) image of the same region under480nm excitat ion. (b) ...
tering) Under our condit ions we reach a chiral separat ion
efficiency of 15%, limit due to the finite spin-valley con-
t rast in WS2. This value could potent ially exceed 50%
by using curent ly available preparat ion schemes for the
TMD material (cite records of CPL).
Paragraph 4: Our system consists of a hole array plas-
monic resonator covered by a monolayer of WS2 (see
Fig. 2(a)). We begin by sput tering 200nm of gold on
a glass subst rate coated by a 5nm thick Chromium ad-
hesion layer. The plasmonic hole array is milled through
the metallic layers using a Carl Zeiss Auriga FIB sys-
tem. A period ⇤ = 480nm is chosen to generate a plas-
monic mode resonant with the absorpt ion energy of the
A-exciton of WS2 at 2.01eV. We then spin-coat on the
metal a 5nm thick PMMA (polymethyl methacrylate)
layer, on top of which the mechanically exfoliated WS2
flake is t ransferred.
Paragraph 5: An SEM image of the bare plasmonic
grat ing is shown in Fig. 2(b). It consists of rotat -
ing rectangular nano-apertures (160nm x 90nm) with
an orbital period equal to 6 t imes the grat ing period.
Similar arrangements of anisotropic apertures have al-
ready been demonst rated to allow for spin-dependent
surface plasmon launching (cite Shuang Zhang, Has-
man, Capasso,...). In our case, the local orientat ion
of the anisot ropy axis of the nanoaperture is given by
φ = ⇡x/ N ⇤, where N is the number of grat ing pe-
riods needed to obtain a ⇡ -rotat ion. Therefore, light
scat tered from the array acquires a geometric phase of
φg = 2⇡σx/ N ⇤, where σ = ± 1 correspond respect ively
to right and left -handed circular polarizat ions [18,19]??.
Di↵erent iat ing thisphasewith respect to thex coordinate
gives a rotat ion-induced momentum kg = 2σ⇡ / N ⇤which
adds up to the usual grat ing law, giving the plamonic
resonant condit ion: k SP P = k / / ± nG x ± mG y + kgx, as
sketched in Fig. 2(c). The rout ing act ion of such a spin-
FIG. 3. Angle-resolved reflect ivity spect rum of the samplein circular (a) right and (b) left polarizat ion. Angle-resolved
SHG spect rum for circular (c) left and (d) right excitat ion at1.01eV.
orbit grat ing under monochromat ic excitat ion is demon-
st rated in real space using a leakage microscope [? ]. We
clearly observe the unidirect ional launching of plasmonic
waves according to the spin of the incident light beam.
... Simplify this paragraph...
Paragraph 6: The measured angle-resolved white light
reflectometry of the whole structure (Fig. 2(a)) is shown
in Fig. 3 for left and right analyzed circular polariza-
t ions. In each case, two st rongly dispersing branches are
observed, featuring a clear ant i-crossing behaviour. This
demonst rates that the system is in the st rong coupling
regime, with the two branches corresponding to upper
and lower polaritonic states. This is further confirmed
by a coupled oscillators fit to the dispersions, giving a
Rabi split t ing of 40meV. The two dispersion diagrams
show a clear mirror symmetry with respect to the normal
incident (kx = 0) axis for the two opposite handedness,
clearly demonstrat ing the capability of our st ructure as
a spin-determined polaritons launcher. ...Cooperativity
factor C = 0.8...
Paragraph 7: Whereas this represents an interest ing
plat form for general polaritonic systems, here we invest i-
gate the interplay between this opt ical chirality and the
valley-cont rast ing chirality of theWS2 monolayer. A first
demonst rat ion of such an interplay is found in the res-
onant second harmonic response of the coupled system.
Indeed, WS2 valley t ransit ions have been shown to give
a high spin cont rast in circularly polarized second har-
monic generat ion (SHG), with select ion rules opposite to
those of linear absorpt ion [? ]. The angle resolved reso-
nant SHG for left and right circularly polarized excitat ion
(Fig. 3(c) and (d) respect ively) display lobes that corre-
. . . . x
Helicity-dependent directional SP launching
Surface plasmons (SPs)
Strong interaction with light
+_< 1 nm
Semiconducting 2H-TMDs
M = Mo, W X = S, Se, Te
MX2
2D Matter meets 2D Light
Helicity-dependent valley addressability
M = Mo, W X = S, Se, Te
ISIS (Strasbourg) z
x
y
kz
ETM
SE
Helicity-dependent directional SP launching
-kSP
+kSP
Surface plasmons (SPs)
Strong interaction with light
+_< 1 nm
Semiconducting 2H-TMDs
M = Mo, W X = S, Se, Te
MX2
Tailored control of the valley pseudospin using TMD-SP architectures
2D Matter meets 2D Light
Helicity-dependent valley addressability
M = Mo, W X = S, Se, Te
Helicity-dependent directional SP launching
z
x
y
kz
ETM
SE
Room temperature chiral coupling between valley excitons and helicity-momentum locked surface plasmons
T. Chervy*, S. Azzini*, et al., ACS Photonics 5, 1281 2018 (Arxiv 1701.07972) See also K. Kuipers’ “intermezzo”
-kSP
+kSP
2D Matter meets 2D Light
Coll: T. Chervy, S. Azzini, J.A. Hutchinson, T.W. Ebbesen, C. Genet (ISIS, Uni. Strasbourg) Y. Gorodetzki (Ariel, IL), S. Wang (Eindhoven, NL)
Chiral coupling: 1L-TMD exciton with spin-orbit SP mode
Angle-resolved WL absorption (1-R ~ A) – CPL analysis
Anticrossing
Strong-coupling FOM
Resonant condition TMD excitons / SP modes
2
FIG. 2. (a) White light (WL) microscope image of the sample
and photoluminescence (PL) image of the same region under480nm excitat ion. (b) ...
tering) Under our condit ions we reach a chiral separat ion
efficiency of 15%, limit due to the finite spin-valley con-
t rast in WS2. This value could potent ially exceed 50%
by using curent ly available preparat ion schemes for the
TMD material (cite records of CPL).
Paragraph 4: Our system consists of a hole array plas-
monic resonator covered by a monolayer of WS2 (see
Fig. 2(a)). We begin by sput tering 200nm of gold on
a glass subst rate coated by a 5nm thick Chromium ad-
hesion layer. The plasmonic hole array is milled through
the metallic layers using a Carl Zeiss Auriga FIB sys-
tem. A period ⇤ = 480nm is chosen to generate a plas-
monic mode resonant with the absorpt ion energy of the
A-exciton of WS2 at 2.01eV. We then spin-coat on the
metal a 5nm thick PMMA (polymethyl methacrylate)
layer, on top of which the mechanically exfoliated WS2
flake is t ransferred.
Paragraph 5: An SEM image of the bare plasmonic
grat ing is shown in Fig. 2(b). It consists of rotat -
ing rectangular nano-apertures (160nm x 90nm) with
an orbital period equal to 6 t imes the grat ing period.
Similar arrangements of anisot ropic apertures have al-
ready been demonstrated to allow for spin-dependent
surface plasmon launching (cite Shuang Zhang, Has-
man, Capasso,...). In our case, the local orientat ion
of the anisot ropy axis of the nanoaperture is given by
φ = ⇡ x/ N ⇤, where N is the number of grat ing pe-
riods needed to obtain a ⇡ -rotat ion. Therefore, light
scat tered from the array acquires a geometric phase of
φg = 2⇡σx/ N ⇤, where σ = ± 1 correspond respect ively
to right and left -handed circular polarizat ions [18,19]??.
Di↵erent iat ing thisphasewith respect to thex coordinate
gives a rotat ion-induced momentum kg = 2σ⇡ / N ⇤which
adds up to the usual grat ing law, giving the plamonic
resonant condit ion: k SP P = k / / ± nG x ± mG y + kgx, as
sketched in Fig. 2(c). The rout ing act ion of such a spin-
FIG. 3. Angle-resolved reflect ivity spect rum of the samplein circular (a) right and (b) left polarizat ion. Angle-resolved
SHG spect rum for circular (c) left and (d) right excitat ion at1.01eV.
orbit grat ing under monochromat ic excitat ion is demon-
st rated in real space using a leakage microscope [? ]. We
clearly observe the unidirect ional launching of plasmonic
waves according to the spin of the incident light beam.
... Simpli fy this paragraph...
Paragraph 6: The measured angle-resolved white light
reflectometry of the whole st ructure (Fig. 2(a)) is shown
in Fig. 3 for left and right analyzed circular polariza-
t ions. In each case, two st rongly dispersing branches are
observed, featuring a clear ant i-crossing behaviour. This
demonst rates that the system is in the st rong coupling
regime, with the two branches corresponding to upper
and lower polaritonic states. This is further confirmed
by a coupled oscillators fit to the dispersions, giving a
Rabi split t ing of 40meV. The two dispersion diagrams
show a clear mirror symmet ry with respect to the normal
incident (kx = 0) axis for the two opposite handedness,
clearly demonst rat ing the capability of our st ructure as
a spin-determined polaritons launcher. ...Cooperativity
factor C = 0.8...
Paragraph 7: Whereas this represents an interest ing
plat form for general polaritonic systems, here we invest i-
gate the interplay between this opt ical chirality and the
valley-cont rast ing chirality of theWS2 monolayer. A first
demonst rat ion of such an interplay is found in the res-
onant second harmonic response of the coupled system.
Indeed, WS2 valley t ransit ions have been shown to give
a high spin cont rast in circularly polarized second har-
monic generat ion (SHG), with select ion rules opposite to
those of linear absorpt ion [? ]. The angle resolved reso-
nant SHG for left and right circularly polarized excitat ion
(Fig. 3(c) and (d) respect ively) display lobes that corre-
Grating period L=480 nm
T. Chervy*, S. Azzini* et al., ACS Photonics 2018
Room temperature chiralitons
K
K’
“Chiralitons”
• Plasmon sorter efficiency: 15% Net chiraliton flow: 6 % 40 %
Recent related results: Spin-momentum locking: TU Delft (Science 2018), UT Austin (arXiv:1801.06543) Valley polaritons: U. Sheffield, Nat. Photon. 2018, Northwestern U. Nat Photon 2018, Würzburg PRB2017
• PL polarization tomography on a linear basis
Measured as the m11 Mueller matrix element
• Chiralitonic valley coherence
Coherent superposition of counterpropagating chiralitons
No valley contrasts in bare TMD Cavity protection mechanism? Designing new 2D building blocks for chiral optics.
(𝑆1|𝑇𝑀 − 𝑆1|𝑇𝐸)/2 = 𝑚11~ 5 − 8 %
Outlook:
Outline
+_ Graphene/TMD heterostructures as a 2D optoelectronic building block
Room temperature valley polarization and coherence
in TMD/Graphene heterostructures
Room temperature Chiral coupling of valley excitons with spin-momentum locked
surface plasmons
• Graphene: 2D semi-metallic channel Quasi-transparent (~2% absorption per layer) High carrier mobility and large carrier density Gate-tunable properties
• TMD: 2D semiconducting channel Strong light-matter interaction Broadband absorption and tunable emission
De Fazio et al., ACS Nano 2016
0
1
2
3
1.5 2.0 2.50.0
0.5
1.0
PL
In
ten
sity (
arb
. u
nits)
R
/R
Energy (eV)
1L MoSe2
300 K
A B
Gr/TMD heterostructures: why the interest?
R, PL, Raman Tunable excitation
Electrical Readout
•Gr-TMD heterostructures:
Strong interlayer coupling (J. He et al., Nat Comm. 2014)
Photogating/photodetection (W. Zhang et al., Sci. Rep. 2014)
ps-range photoresponse (M. Massicotte et al., Nat Nano 2016)
Gr/TMD heterostructures: why the interest?
VB
TMD Thickness (nm)
Res
po
nse
rat
e (s
-1)
Res
po
nse
tim
e
M. Massicotte et al., Nat Nano 2016 ICFO
Strong interlayer coupling (J. He et al., Nat Comm. 2014)
Photogating/photodetection (W. Zhang et al., Sci. Rep. 2014)
ps-range photoresponse (M. Massicotte et al., Nat Nano 2016)
Low efficiency with monolayer TMD Energy transfer?
+
_
+
_
_
++
_
_
++
_
+
_
+
_
TMD Graphene
(a) Interlayer Electron Transfer
(b) Interlayer Hole Transfer
(c) Dexter-type Interlayer Energy Transfer
(d) Förster-type Interlayer Energy Transfer
Charge and energy transfer mechanisms
Probed in the steady state (Raman) Balanced electron and hole transfer
Determine exciton dynamics Relative efficiencies?
Our Experimental Approach
Understanding near-field interactions (IET vs ICT)
Implications for optoelectronic devices
PL (TMD) Raman (Gr, TMD)
Tunable excitation
+_
Gr
TMD
PL (TMD) Raman (Gr, TMD)
Tunable excitation
+_
Electrical Readout Gr
TMD
< 1 nm
Photolum. Raman
Tunable excitation
+_< 1 nm
Gr
TMD
Atomic Force Microscopy
MoSe2
Gr
Gr/MoSe2
Smooth Gr/MoSe2 domains
AFM
0.0 0.5 1.0 1.5
0
1
2
3
He
igth
(n
m)
Distance (µm)
2
0.0 0.5 1.0 1.5
0
1
2
3
0.6
He
igth
(n
m)
Distance (µm)
2
1.3 1.4 1.5 1.6 1.7 1.8 1.9
0.000
0.003
PL inte
nsity (
arb
. units)
Photon energy (eV)
1.3 1.4 1.5 1.6 1.7 1.8 1.9
0
1
PL inte
nsity (
arb
. units)
Photon energy (eV)1.3 1.4 1.5 1.6 1.7 1.8 1.9
0
1
PL
in
ten
sity (
arb
. u
nits)
Photon energy (eV)
1.3 1.4 1.5 1.6 1.7 1.8 1.9
0
1
PL inte
nsity (
arb
. units)
Photon energy (eV)
5 µm
Bare MoSe2
2 nm 0.6 nm
Decoupled Gr/MoSe2 Coupled Gr/MoSe2
Strong PL Quenching ~ 300
Photoluminescence mapping
1020 1021 1022 1023
0.01
0.1
1
I PL/
ph
oto
ns (
arb
units)
photons (cm-2 s-1)
Exciton: dynamics: PL vs photons
• PL saturation on bare and decoupled MoSe2
Bare MoSe2
Decoupled Gr/MoSe2
Normalized PL: IPL
photons
• PL saturation on bare and decoupled MoSe2: Exciton-Exciton Annihilation (EEA)
EEA: N. Kumar et al., PRB 89, 125427 (2014), S. Mouri et al., PRB 90, 155449 (2014), D. Sun et al., Nano Lett. 14, 5625 (2014)
S. Mouri et al., PRB 2014
1020 1021 1022 1023
0.01
0.1
1
I PL/
pho
ton
s (
arb
units)
photons (cm-2 s-1)
Bare MoSe2
Decoupled Gr/MoSe2
Exciton dynamics: PL vs photons
1020
1021
1022
1023
0.01
0.1
1
I PL/
photo
ns (
arb
units)
photons
(cm-2 s
-1)
Exciton dynamics: PL vs photons
• No PL saturation on Gr/MoSe2 and IPL(B) ~ IPL(A)
• PL saturation on bare and decoupled MoSe2
Exciton-Exciton Annihilation (EEA)
Bare MoSe2
Decoupled Gr/MoSe2
Coupled Gr/MoSe2
1020
1021
1022
1023
0.01
0.1
1
I PL/
ph
oto
ns (
arb
units)
photons
(cm-2 s
-1)
• No PL saturation on Gr/MoSe2
Bare MoSe2
Decoupled Gr/MoSe2
Coupled Gr/MoSe2
0 2 4 6 8 10 1210
-3
10-2
10-1
100
Bare MoSe2
Decoupled SLG/MoSe2
Coupled SLG/MoSe2
PL C
ounts
(arb
. units)
Time Delay (ns)
IRF
• PL saturation on bare and decoupled MoSe2
• No PL saturation on Gr/MoSe2 and IPL(B) ~ IPL(A)
Drastic reduction of the excitonic lifetime (< 1 ps)
Fast interlayer Energy Transfer?
Exciton dynamics: PL vs photons
G. Froehlicher, E. Lorchat, S. Berciaud, Phys. Rev. X 8, 011007 (2018) Fast TRPL at INSA Toulouse
1020
1021
1022
1023
0.01
0.1
1
I PL/
ph
oto
ns (
arb
units)
photons
(cm-2 s
-1)
• No PL saturation on Gr/MoSe2
Bare MoSe2
Decoupled Gr/MoSe2
Coupled Gr/MoSe2
• PL saturation on bare and decoupled MoSe2
• No PL saturation on Gr/MoSe2 and IPL(B) ~ IPL(A)
Drastic reduction of the excitonic lifetime (< 1 ps)
Fast interlayer Energy Transfer?
Exciton dynamics: PL vs photons
G. Froehlicher, E. Lorchat, S. Berciaud, Phys. Rev. X 8, 011007 (2018)
-10 0 10 20 30 40
10-2
10-1
100 Data
Fit
IRF
PL c
ounts
(arb
. units)
Time Delay (ps)
Fast TRPL at INSA Toulouse
Outline
+_ Graphene/TMD heterostructures as a 2D optoelectronic building block
Room temperature valley polarization and coherence
in TMD/Graphene heterostructures
Room temperature Chiral coupling of valley excitons with spin-momentum locked
surface plasmons
SOC
+K −K
Back to valley contrasts
Photonic state Excitonic state
𝜎± |±1
1
2 𝜎+ ± |𝜎−
1
2 +1 ± |−1
ΓTMD ≪ ΓS ⇒ 0% valley polarization
SOC
+K −K
Photonic state Excitonic state
𝜎± |±1
1
2 𝜎+ ± |𝜎−
1
2 +1 ± |−1
ΓTMD ≪ ΓS ⇒ 0 % valley polarization
ΓTMD/Gr ≪ ΓS ⇒ finite valley polarization
Back to valley contrasts
Our experimental approach
BN-encapsulated WS2-Gr heterostructure
Mueller polarimetry setup
Mueller Polarimetry
m11,22 = valley coherence m33 = valley polarization
Degrees of linear and circular polarization
mi0: birefringence
m0i: dichroisim
Experimental “sanity check” : m11 = m22 and mij,i j= 0
Mueller Polarimetry in WS2/Gr
Laser @600 nm
Mapping valley contrasts
45% valley polarization, 30% valley coherence @Room Temperature
E.Lorchat, S.Azzini, T.Chervy et al., ArXiv:1804.06725 (doi: 10.1021/acsphotonics.8b01306)
2.07 2.06 2.05 2.042.07 2.06 2.05 2.04
(b)
*
s+s+
s+s-
PL In
ten
sity (
arb
. u
nits)
598 601 604 607
0.0
0.5
1.0(d)
*
m3
3
Wavelength (nm)
*
*
XX
XY
PL In
ten
sity (
arb
. u
nits)
(a)
598 601 604 607
0.0
0.5
1.0(c)
m1
1
Wavelength (nm)
Energy (eV)Energy (eV)
Record valley coherence up to 60 % (10 %) in BN-capped WS2/Gr (BN capped WS2)
Large valley contrasts in WS2/Gr (20 K)
Val
ley
Co
he
ren
ce
Val
ley
Pola
riza
tio
n
F. Cadiz et al. PRX 2017 (valley coherence of 50% in BN-capped MoS2)
E.Lorchat, S.Azzini, T.Chervy et al., ArXiv:1804.06725 (doi: 10.1021/acsphotonics.8b01306)
*Raman features
Conclusion and outlook
2D matter meets 2D light: many games to play at the interface between (chiral) nanophotonics, condensed matter physics and materials science Interfacing 2D materials with chiral plasmonic resonators Probing dark excitonic states using surface plasmons Electrical and electromechanical control of chiral coupling Charge and energy transfer mechanism(s) Intervalley scattering mechanisms in TMD/Gr vs bare TMD
Acknowledgements
Stefano Azzini Thibault Chervy
Yuri Gorodetzki (Ariel, IL) Shaojun Wang (Eindhoven, NL)
James Hutchinson Thomas Ebbesen Cyriaque Genet
Cédric Robert, Delphine Lagarde, Xavier Marie
Takashi Taniguchi Kenji Watanabe
ISIS
Funding:
Etienne Lorchat Guillaume Froehlicher Luis Parra-Lopez Florentin Fabre StNano Clean room
Supplementary Slides Net charge transfer in TMD-Gr heterostructures General slides on heterostructures and on energy transfer
How about photoinduced charge transfer?
Raman Scattering
Tunable excitation
+_
…study of net (and slow) eletron transfer from TMD to Graohene (cf PRX 2018). This effect is likely extrinsic (not observable if we replace SiO2 by hBN) but is of interest for photodetection/photogating (see next slide).
Gr/TMD heterostructures: why the interest?
W. Zhang et al., Sci. Rep. 2014 KAUST
Strong interlayer coupling (J. He et al., Nat Comm. 2014)
Photogating/photodetection (W. Zhang et al., Sci. Rep. 2014)
ps-range photoresponse (M. Massicotte et al., Nat Nano 2016)
1200 1400 1600 2600 28000
1
2
3
4
5
(no) D mode
2L sus
1L sus
G mode
C
ou
nts
(a
rb.
un
its)
Raman Shift (cm-1)
2D mode
Raman spectroscopy: a quantitative probe of doping
0 5 10 15 20 25
0
5
10
15 electrons
holes
Sample 1
Sample 2
Sample 3
Sample 4
Sample 5
G (
cm
-1)
G (cm
-1)
0 5 10 15 20 25
-15
-10
-5
0
5
10
15
Sample 1
Sample 2
Sample 3
Sample 4
Sample 5
2
D -
0 2D(c
m-1)
G-
0
G(cm
-1)
strain
2.22
D
G
55.02
D
G
h+
2.02
D
G
e-
2D
Data: G. Froehicher & SB, PRB 2015. See also: A. Das et al., Nat. Nano 2008 S. Pisana et al., Nat Mater 2007, J. Yan et al., PRL 2007,…
1590 2640 2680 2720
0
1
2
3
4
52D
Ra
ma
n in
ten
sity (
arb
. u
nits)
Raman shift (cm-1)
G
1590 2640 2680 2720
0
1
2
3
Ra
ma
n in
ten
sity (
arb
. u
nits)
Raman shift (cm-1)
2DG
1590 2640 2680 2720
0
1
2
3
4
5
Ra
ma
n in
ten
sity (
arb
. u
nits)
Raman shift (cm-1)
G 2D
1590 2640 2680 2720
0
1
Ra
ma
n in
ten
sity (
arb
. u
nits)
Raman shift (cm-1)
5 µm Reference on SiO2
Bare MoSe2
Coupled Gr/MoSe2
Decoupled Gr/MoSe2
ph increases
Eex = 2.33 eV
Raman response vs photon flux (1)
1020
1021
1022
1023
456789
13141516171819
G (
cm
-1)
ph
(cm-2
s-1)
1020
1021
1022
1023
1584
1586
1588
1590
1592 SLG/SiO2
Decoupled SLG/MoSe2
Coupled SLG/MoSe2
G (
cm
-1)
ph
(cm-2
s-1)
G-mode frequency G-mode FWHM
Clear signatures of a net photoinduced charge transfer
Raman response vs photon flux (2)
Clear signatures of a net photoinduced charge transfer
Raman response vs photon flux (2)
1584 1586 1588 1590 15924
6
8
10
12
14
16
18
20
G (
cm
-1)
G (cm
-1)
increasing ph
Width-Frequency Correlation
Correlation between the 2D- and G-mode frequencies
• Comparison with Gr/MoS2
W. Zhang et al., Scientific Reports 4, 3826 (2014)
electrical measurements
𝑒− transfer
0 5 10 15 20 25
-15
-10
-5
0
5
10
15
Sample 1
Sample 2
Sample 3
Sample 4
Sample 5
2D
-
0 2D(c
m-1)
G-
0
G(cm
-1)
55.02
G
D
h+
2.02
G
D
e-
holes
electrons Evidence for TMD Gr electron transfer
1584 1586 1588 1590 1592
2674
2676
2692
2
D (
cm
-1)
G (cm-1)
0.11
increasing ph
𝑒− transfer
-300 -200 -100 0 100 200 3001582
1584
1586
1588
1590
1592
1594
G (
cm
-1)
EF (meV)
-300 -200 -100 0 100 200 300 400
-5.5 -2.4 -0.6 0.0 0.6 2.4 5.5 9.8
1584
1586
1588
1590
1592
1594
1596
1598
1600
1602
n (x1012
cm-2)
EF (meV)
wG (
cm
-1)
-2
0
2
4
6
8
10
12
14
DG
G (
cm
-1)
G. Froehlicher and S. Berciaud, PRB 91, 205413 (2015) G. Froehlicher, E. Lorchat, S. Berciaud, Phys. Rev. X 8, 011007 (2018)
Transfered electron density nG
Quantifying photoinduced doping
-300 -200 -100 0 100 200 300
4
6
8
10
12
14
16
G (
cm
-1)
EF (meV)
Evidence for net photoinduced electrpon transfer Extrinsic effect – slow dynamics
0 2 4 6 8 10 12
-200
-100
0
100
200
300
sample 1
sample 2
sample 3
EF (
me
V)
ph (x1023 cm-2 s-1)
1020 1021 1022 1023 1024
100
150
200
250
300
0.6
1.4
2.4
3.8
5.5
n (
x1
01
2 c
m-2
)
S5
EG
rF
(m
ev)
ph (cm-2 s-1)
S4
0 1 2 3 4 5 6100
150
200
250
300
0.6
1.4
2.4
3.8
5.5
nG (
x10
12 c
m-2
)
Air
EF (
meV
)
ph (x1023 cm-2 s-1)
Reproducibility and environmental effects
0 1 2 3 4 5 6100
150
200
250
300
0.6
1.4
2.4
3.8
5.5
nG (
x10
12 c
m-2
)
Air
EF (
meV
)
ph (x1023 cm-2 s-1)
Vacuum
Electron trapping by molecular adsorbates in air Direct saturation under vacuum
1020 1021 1022 1023 1024
-200
-100
0
100
200
300
-2.4
-0.6
0
0.6
2.4
5.5
n (
x1
01
2 c
m-2
)
S3
S2
EG
rF
(m
eV
)
ph (cm-2 s-1)
S1
1020 1021 1022 1023 1024
100
150
200
250
300
0.6
1.4
2.4
3.8
5.5
n (
x1
01
2 c
m-2
)
S5
EG
rF
(m
ev)
ph (cm-2 s-1)
S4
Several Gr/MoSe2/SiO2 samples Ambient AIR
Gr/MoSe2/SiO2 vs MoSe2/Gr/SiO2
Air (full) vs vacuum (open)
Microscopic mechanism (in vacuum)
MoSe2 graphene
van der Waals gap
−
+
e−
h+
𝐸F,eM
𝐸F,hM
𝐸FiM
Light on, t=0
MoSe2 graphene
van der Waals gap
−
+
e−
h+
𝐸FiM
Light on, t∞
• n+nG=p • Saturation for n Γe = p Γh
• Independent of fphoton
𝐸FGr
𝐸FGr
+ +
Balanced electron and hole currents at saturation Intrinsic mechanism? Optical determination of band offsets?
Recap: Fermi level and PL intensity
Exciton dynamics in MoSe2 is largely independent of the electron and hole transfer efficiencies
Strong hint for dominant energy transfer
Recap: Fermi level and PL intensity
Exciton dynamics in MoSe2 is largely independent of the electron and hole transfer efficiencies
Strong hint for dominant energy transfer
Partial conclusion
Strong interlayer coupling in Gr/TMD heterostructures
Saturation of the net electron transfer
Highly efficient (sub)-picosecond energy transfer
Förster vs Dexter energy transfer?
Electrical control of charge and energy transfer
Implications for optoelectronics and optospintronics
More info: G. Froehlicher, E. Lorchat, S. Berciaud, Phys. Rev. X 8, 011007 (2018)
van der Waals Heterostructures
Haigh, Gorbachev et al., Nature Materials 2012 Manchester Group
Castellanos-Gomez et al. 2D Materials 1 011002 (2014)
No dangling bounds No lattice mismatch issues Rotational degree of freedom • 2010 : Graphene on hBN • 2017 : wet or dry transfer, pick up and lift,… • Numerous possibilities!
C-H Lee et al. Nat. Nano (2014) (Columbia)
Atomically thin p-n junctions
52
Graphène
hBN
MoS2
WSe2
-3
-4
-5
-6
--7
Gra
ph
en
e
Mo
S 2
Mo
Se2
Mo
Te2
WS 2
WSe
2
TMD: Y. Liang et al., APL 103, 42106 (2013), M. Ugeda et al., Nat. Mater. 5, 1091 (2014) Graphene: Y.-J. Yu et al., Nano Lett. 9, 3430 (2008)
-4.57
-3.89 -3.74
-4.09
-3.61
-4.29
Ene
rgy
(eV
)
TMD/TMD: Type II band alignment TMD/Graphene: Photoinduced TMD Gr e and h transfer
Band offsets in 2D materials
Microscopic Mechanism
• No PL saturation on Gr/MoSe2
MoSe2 graphene
Vacuum level
𝜒M
𝑊G
van der Waals gap
𝐸FGr
𝑒−
𝐸FM
MoSe2 graphene
van der Waals gap
𝐸FM
In the dark
Before contact: n-doped MoSe2
Weakly doped graphene
After contact (in the dark): Neutral MoSe2
n-doped graphene
𝐸FGr
X
E
Outlook: low-temperature photoluminescence
1.4 1.5 1.6 1.7
0
1
2
3
4
1.4 1.5 1.6 1.7
1.4 1.5 1.6 1.710-1
100
101
102
103
104
1.4 1.5 1.6 1.7
PL
In
ten
sity (
arb
. u
nits)
Defect states Trion
ExcitonMoSe2 MoSe2/Gr
Exciton
4K4K
x10
PL
In
ten
sity (
arb
. u
nits)
Energy (eV)
Defect states Trion
Exciton
MoSe2 MoSe2/Gr
Exciton
4K4K
Energy (eV)
: Exciton
: Trion
No trion emission!
Förster energy transfer: near field dipole-diople interaction
𝐄𝐷 =1
4𝜋𝜀0𝑘2 𝐫˄𝝁𝐷 ˄𝝁𝐷
𝑒𝑖𝑘𝑟
𝑟2+
3𝐫 𝐫. 𝝁𝐷
𝑟2− 𝝁𝐷
1
𝑟3−
𝑖𝑘
𝑟2𝑒𝑖𝑘𝑟
Th. Förster Annalen der Physik 437, 55 (1948) Novotny, Hecht Principles of Nano Optics, Ch 8 R0 : Förster radius
Förster and Dexter energy transfer
Coulomb (FRET) term ‘Long’ range (power law) Implies spectral overlap
Exchange (Dexter) term Short range (exponential, idem CT) Implies overlap of molecular orbitals
1 2 Förster 1946-51
Dexter 1953
A Govorov, PL Hernandez Martinez, HV Demir Understanding and Modeling Förster-type Resonance Energy Transfer (FRET), Springer 2016
5 µm0
1
Ph
oto
n E
mis
sio
n R
ate
(arb
.)
Quartz
Graphene
Elaser Elum
GrapheneNPL
Elaser Elum
Quartz
Energy transfer in low-dimensional heterostructures
Z. Chen et al., ACS Nano 4, 2964 (2010) (QD-Gr) F. Federspiel et al., Nano Lett. 15, 1252 (2015) (QD-Gr, NPL-Gr) L. Gaudreau et al., Nano Lett. 13, 2030 (2013) (Dye-Gr)
Graphene
Quartz
D. Kozawa et al., Nano Lett. 16, 4087 (2016)
Nanoscale emitter/graphene
C. Roquelet et al., APL 97, 141918 (2010)
Molecules/nanotubes
TMD/TMD
nanoplatelets
Highly Efficient Förster-type Energy transfer
Interlayer energy transfer has been largely overlooked in TMD/Gr
© F. Vialla
quartz
0 1 10 100
0.1
1
gT~1/d
4
2D-graphene model
MgO thickness (nm)
Decay R
ate
(n
s-1)
CVD graphene
Quartz
MgO
0 1 10 100
0.1
1
gT~1/d
4
2D-graphene model
MgO thickness (nm)
De
cay R
ate
(n
s-1)
CVD graphene
Quartz
MgO
2D Materials at IPCMS
• Optical spectroscopy
• Phonons, excitons and their coupling(s)
IPCMS-Strasbourg
-50 0 50 100150200250300
0
12L MoTe
2oMX
iMX
oX
iX
Breathing
Ram
an In
ten
sity (
arb
. un
its)
Raman Shift (cm-1)
Shear
168 172 176
2L
5L
4L
3L
1L oX modeR
am
an Inte
nsity (
arb
. units)
Raman Shift (cm-1)
0 2 4 6 8 10 12
169
170
171
172
173
174
Fre
qu
en
cy (
cm
-1)
Number of Layers
oX
Mo Te
Interlayer interactions: Davydov splitting and unified description of the phonon modes
2H TMD - Froehlicher et al., Nano Lett. 2015, Miranda et al., Nano Lett. 2017. Coll: L. Wirtz group, Uni. Luxembourg
1T’ TMD - Lorchat et al. ACS Nano 2016 (ReS2 and ReSe2)
See also Froehlicher et al., J. Raman Spec. 2018 (special issue on 2D materials)
+_