Supplementary Materials for
Control of trion-to-exciton conversion in monolayer WS
2
by orbital angular
momentum of light
Rahul Kesarwani, Kristan Bryan Simbulan, Teng-De Huang, Yu-Fan Chiang, Nai-Chang Yeh*,
Yann-Wen Lan*, Ting-Hua Lu*
*Corresponding author. Email: ncyeh@caltech.edu (N.-C.Y.); ywlan@ntnu.edu.tw (Y.-W.L.); thlu@ntnu.edu.tw
(T.-H.L.)
Published 1 April 2022,
Sci. Adv.
8
, eabm0100 (2022)
DOI: 10.1126/sciadv.abm0100
This PDF file includes:
Figs. S1 to S11
Tables S1 and S2
Note S1
1.
Surface morphology and Spectroscopic characterization of WS
2
.
Figure S1.
AFM image of single-layer WS
2
flake
. Monolayer deposition of a CVD-grown WS
2
flake onto SiO
2
/Si substrate confirmed by its (a) AFM image and (b) Raman spectrum.
2.
Deconvolution of the exciton, trion and defects peaks in photoluminescence (PL) spectra
excited by different orbital angular momentum (OAM) light at 300 K.
Figure S2.
OAM light dependence of PL spectra at room temperature (RT)
. Deconvoluted
RT (300 K) PL spectra of monolayer WS
2
excited by different value of OAM (ℓ) at 30 μW laser
power. Each PL spectrum has been fitted by the superposition of three components as indicated
by the blue, red, and gray curves, which correspond to the exciton, trion, and defect states,
respectively. It is evident that the exciton intensity increases with the OAM value ℓ while the
trion intensity decreases. Similar procedure has been taken to analyze the PL spectra obtained by
using other laser powers. These PL spectral analyses are used to estimate the trion-to-exciton
conversion efficiency for a given OAM value by considering the trion-to-exciton intensity ratio.
3.
Deconvolution of the exciton, trion and defects peaks in PL spectra excited by different
OAM light at 77 K.
Figure S3.
OAM light dependence of PL at 77 K
. Deconvolution of the low temperature (77 K)
PL spectra of monolayer WS
2
excited by different values of OAM at 30 μW laser power. Each
PL spectrum has been fitted by the superposition of three components as indicated by the blue,
red, and gray curves, which correspond to the exciton, trion, and defect states, respectively. It is
found that the trion-to-exciton intensity ratio is independent of the topological charge of the
OAM light.
4.
Fitting details of deconvoluted PL spectra taken at 300 K.
We used an origin software for fitting the PL spectra. The wavelengths for each peak are fixed
within reasonable range from ~600 nm to ~650 nm (2.067 to 1.908 eV). For example, the fitting
Adj. R-square value of PL spectra at 30 μW laser power for all values of ℓ are listed in Table S1.
The full-width half maximum (FWHM) fitting error is listed which has been obtained after the
deconvolution of PL spectra. However, the error bar putting in the reported figures is estimated
via taking the average of repeating the PL measurements three times (each time deconvoluted the
PL spectra).
Table S1.
Coefficient of determination and fitting error at 300 K
. Fitting parameters for the
PL spectra of 30 μW laser power under different OAM values ℓ of light at 300 K.
OAM value, ℓ
@30μW
Adj. R-square
Exciton FWHM
fitting error (nm)
Trion FWHM
fitting error (nm)
0
0.99977
0.216
1.390
1
0.99975
0.957
3.734
2
0.99969
0.051
0.762
3
0.99975
0.053
0.497
4
0.99967
0.059
0.663
5
0.99979
0.045
0.407
6
0.99974
0.045
0.459
7
0.99976
0.038
0.408
8
0.9998
0.035
0.359
5.
Fitting details of deconvoluted PL spectra taken at 77 K.
We used an origin software for fitting the PL spectra. The wavelengths for each peak are fixed
within reasonable range from ~600 nm to ~630 nm (2.067 to 1.968 eV). For example, the fitting
Adj. R-square value of PL spectra at 30 μW laser power for all values of ℓ are listed in Table S1.
The FWHM fitting error is listed which has been obtained after the deconvolution of PL spectra.
Table S2.
Coefficient of determination and fitting error at 77 K
. Fitting parameters for the
PL spectra of 30 μW laser power under different OAM values ℓ of light at 77 K.
OAM value, ℓ
@30μW
Adj. R-square
Exciton FWHM
fitting error (nm)
Trion FWHM
fitting error (nm)
0
0.99983
0.107
0.105
1
0.99963
0.193
0.176
2
0.99947
0.243
0.199
3
0.9986
0.285
0.207
4
0.99771
0.267
0.157
5
0.998
0.292
0.126
6
0.99536
0.455
0.160
7
0.99864
0.486
0.284
8
0.99407
0.299
0.267
6.
Comparison of the trion/exciton intensity ratio at fixed value of OAM light.
Figure S4.
Intensity ratio between trion and exciton as function of OAM light at a fixed
laser power
. Comparison of the trion/exciton intensity ratio of monolayer WS
2
vs.
the OAM
value ℓ at 300 K under constant laser powers of 30 μW and 100 μW. It is found that the
trion/exciton intensity ratio increases with increasing laser power for a given ℓ. These results
suggest that the OAM of light can effectively suppress the trion/exciton emission, and a lower
laser power leads to a higher conversion efficiency.
7.
PL spectra for different topological charges of OAM and laser powers at 300 K.
Figure S5.
Power and OAM light dependence of PL spectra at 300 K
. PL spectra of
monolayer WS
2
at RT (300 K) measured via the modified in-house PL setup with different OAM
of light (ℓ= 0 to 8) and laser powers: (a) 30 μW, (b) 50 μW, (c) 70 μW, and (d) 100 μW.
8.
PL spectra for different topological charges of OAM and laser powers at 77 K.
Figure S6.
Power and OAM light dependence of PL spectra at 77 K
. PL spectra of
monolayer WS
2
at low temperature (77 K) in vacuum using the same PL setup with different
OAM of light and laser powers: (a) 30 μW, (b) 50 μW, and (c) 100 μW. The PL spectra of
monolayer WS
2
at 77 K exhibit showing an exciton-rich behavior for all values of OAM and
laser powers. Notably, the intensity ratio at low temperature (77 K) is nearly independent of
OAM.
9.
Dependence of the peak position and intensity ratio of excitons and trions on the OAM at 300
K for laser powers ≤ 100 μW.
Figure S7.
Data analysis of trion and exciton from PL spectra at 300 K
. (a) and (b) are the
PL peak positions of excitons and trions, respectively, of the monolayer WS
2
taken at 300 K for
laser powers of 30 μW, 50 μW, 70 μW and 100 μW. (c) Variation of trion-to-exciton intensity
ratio extracted from the monolayer WS
2
PL spectra at different OAM of light, which signifies
that the conversion efficiency from trion-to-exciton is nearly the same for all laser powers.
10.
Dependence of the peak position and intensity ratio of excitons and trions on the OAM at 77
K for laser powers ≤ 100 μW.
Figure S8.
Data analysis of trion and exciton from PL spectra at 77 K.
PL peak positions of
(a) excitons and (b) trions, respectively, of the monolayer WS
2
at 77 K for 30 μW, 50 μW and
100 μW. (c) Variation of trion to exciton intensity ratio extracted from the monolayer WS
2
PL
spectra at different OAM of light. The blue shifts in exciton and trion peaks around 3 - 4 nm (for
30 μW and 50 μW) have been observed when increasing the topological charge of OAM light
from 0 to 8. We further note that the small and nearly constant trion-to-exciton intensity ratio
with increasing OAM and laser power, signifies that the PL spectra monolayer WS
2
at 77 K are
exciton-rich and independent of OAM.
11.
Deconvolution of the exciton, trion and defects peaks in PL spectra under different gate
voltages excited by odd values of OAM light for 100 μW laser power at 300 K.
Figure S9.
Gate voltage and OAM light dependence of deconvoluted PL spectra at 300 K
.
Deconvoluted PL spectra as a function of energy/wavelength of a monolayer WS
2
for an odd
number of topological charges of OAM light (x-axis) under the excitation of 100 μW laser power
at the gate voltage V
g
of -50 V (top row), 0 V (middle row) and +50 V (bottom row) respectively.
Each PL spectrum has been fitted by the superposition of three components as indicated by the
blue, red, and gray curves, which correspond to the exciton, trion, and defect states, respectively.
It found that the trion-to-exciton conversion with the topological charge of the OAM light is
higher for positive gate voltage compare to that for negative voltage.
12.
Effect of gate voltage on the intensity ratio and the conversion efficiency excited by odd
values of OAM light for 100 μW laser power at 300 K.
Figure S10.
Gating effect on trion/exciton intensity ratio as function of OAM light
. (a)
Optical microscopy image of monolayer WS
2
flake on SiO
2
/Si substrate with electrodes.
Variations of the trion/exciton intensity ratio with the OAM value ℓ under a constant laser power
of 100 μW at the gate voltage V
g
of (b) -50 V, (c) 0 V, and (d) +50 V. (e) Trion-to-exciton
conversion efficiency at 300 K as a function of gate voltage at a constant laser power of 100 μW.
The trion-to-exciton conversion efficiency is smaller for a negative gate voltage (-50 V) and
becomes larger for a positive gate voltage (+50 V). It is because positive gate voltage can induce
more electrons to form trions, which is in contrast to the case of negative gate voltage for n-type
semiconductor.