Monolayer WS 2 electro-and photo-luminescence enhancement by TFSI treatment

.

The effect of chemical passivation of 1L-TMDs on η EL combined with gated-PL emission in 1L-TMD-based LEDs was not reported to date, to the best of our knowledge.Refs.[54][55][56][57][58][59][60][61] reported PL measurements on 1L-TMDs and focused on non-gated samples, thus limiting the modulation of charge density in 1L-TMDs.Ref. [8] performed gated-PL measurements in 1L-WS 2 , finding that both TFSI treatment and electrical gating increase η P L by a factor of up to∼10 (at∼10 19 cm −2 s −1 photocarrier generation rate), because both processes reduce the n-type behaviour of 1L-WS 2 and suppress X − formation, thus enhancing X 0 radiative recombination.However, gated-PL measurements after TFSI passivation were not provided.The activation of trapping states on TFSItreated 1L-TMDs was not discussed.Ref. [67] carried out EL experiments with TFSI passivation for high-speed (MHz) modulation, but did not report PL nor EL emission tunability.Therefore, an investigation on how TFSI affects EL emission and modifies gated-PL of 1L-TMDbased devices is required.Here, we fabricate LEDs with 1L-WS 2 as active material on a metal-insulator-semiconductor (MIS) structure.We measure EL and gated-PL before and after TFSI treatment.We find that TFSI increases η EL by over one order of magnitude at RT, and PL intensity by a factor∼5.We find that X − and X 0 are present in both EL and PL before TFSI treatment, whereas X 0 dominates after.We attribute this to depletion of excess e and changes in the relaxation pathway, induced by the treatment.This paves the way to more efficient 1L-TMDs-based LEDs and excitonic devices.
Fig. 1a shows the 1L-WS 2 /hBN/SLG tunnel junction configuration used here, where the metallic electrodes provide contacts to apply a voltage (V ) between SLG and 1L-WS 2 .This is prepared as follows.
WS 2 crystals are synthesized using a two-step self-flux technique [72] using 99.9999% purity W and S powders without any transporting agents.Commercial (Alfa Aesar) sources of powders contain a number of defects and impurities (Li, O, Na, and other metals as determined by secondary ion mass spectroscopy).Before growth, W and S powders are thus purified using electrolytic [73] and H 2 [73] based techniques to reach 99.995% purity.WS 2 polycrystalline powders are created by annealing a stoichiometric ratio of powders at 900 • C for 3 weeks in a quartz ampoule sealed at 10 −7 Torr.The resulting powders are re-sealed in a different quartz ampoule under similar pressures and further annealed at 870-910 • C with thermodynamic temperature differential (hot to cold zone difference)∼40 • C. The growth process takes 5 weeks.At the end of the growth, ampoules are cooled to RT slowly (∼40 • C/hour) [74].We use this material as bulk source because our previous work [74] demonstrated that this has a point defect density∼10 9 -10 10 cm −2 , on par or better than previous reports [75].
Bulk WS 2 , hBN (grown by the temperature-gradient method [76]), and graphite (sourced from HQ Graphene) crystals are then exfoliated by micromechanical cleavage using Nitto-tape [77] on 285nm SiO 2 /Si.Optical contrast [78] is first used to identify 1L-WS 2 , SLG, FLG (3-10nm), and hBN(<5nm).The LMs are then characterized by Raman spectroscopy as discussed in Methods.After Raman characterization of all individual LMs on SiO 2 /Si, the FLG/1L-WS 2 /hBN/SLG LMH is assembled using dry-transfer as for Refs.[79,80].FLG is pickedup from SiO 2 /Si using a polycarbonate (PC) membrane on a polydimethylsiloxane (PDMS) stamp (as mechanical support) at 40 • C. We use 40 • C because this is sufficient to increase the adhesion of the PC film [81], to pick all LMs from SiO 2 /Si.Then, FLG is aligned to one edge of 1L-WS 2 on SiO 2 /Si and brought into contact using xyz micromanipulators at 40 • C, leaving the majority of 1L-WS 2 without FLG cover to be used as active area (AA).AA is the region from where light emission is expected, and it is the overlap area between 1L-WS 2 and SLG (green-shaded part in Fig. 1b).Next, FLG/1L-WS 2 is aligned to a hBN flake deposited onto SiO 2 /Si and brought into contact using xyz micromanipulators at 40 • C. Finally, FLG/1L-WS 2 /hBN is aligned to a SLG on SiO 2 /Si and brought into contact using xyz micromanipulators at 180 • C, whereby PC preferentially adheres to SiO 2 [79], allowing PDMS to be peeled away, leaving PC/FLG/1L-WS 2 /hBN/SLG on SiO 2 /Si.PC is then dissolved in chloroform for∼15mins at RT, leaving the FLG/1L-WS 2 /hBN/SLG LMH on SiO 2 /Si [79,80].After LMH assembly, Cr/Au electrodes are fabricated by electron beam lithography (EBPG 5200, Raith GMBH), followed by metallization (1:50nm) and lift-off.
The tunnel junction based on a MIS structure consists of a LMH with 1L-WS 2 as the light emitter, FL-hBN (typically from 2 to 4nm) acting as tunnel barrier, and a SLG electrode to inject holes (h) into 1L-WS 2 .We use FL-hBN<5nm so that a low (typically<5V) driving voltage is sufficient for charge injection to the 1L-WS 2 [82,83].We employ FLG (∼3-10nm) to contact 1L-WS 2 , because FLG reduces the contact resistance [84], while Cr/Au electrodes give Ohmic contacts to SLG and FLG [84].SLG could also be used to contact 1L-WS 2 , however, as the optical contrast is higher in FLG than SLG [78,85], using FLG makes it easier to align it to 1L-WS 2 during transfer.Since TFSI treatment requires direct exposure of 1L-TMDs [54], we place 1L-WS 2 on top of the stack to compare the device performance before and after treatment.We TFSI-treat 4 samples for EL and gated-PL measurements.These are immersed in a TFSI solution (0.2 mg/mL) in a closed vial for 10mins at 100 • C [54][55][56], then removed, dried by a N 2 gun, and annealed on a hot plate at 100 • C for 5mins [54][55][56].Fig. 1b is an image of the 1L-WS 2 -LEDs.The FLG electrode is placed on the side of the SLG to avoid direct tunneling of carriers from SLG to FLG, hence keeping as AA the LMH region extended over SLG and 1L-WS 2 , greenshaded in Fig. 1b.If there is a FLG/SLG overlap, tunneling through FLG-SLG may be possible, not resulting in e-h recombination into 1L-WS 2 , hence no EL [6,25,38].
Figs.1c,d sketch the band diagram of our LEDs for V =0V and V >0V, respectively.For V =0V (at thermodynamic equilibrium as indicated in Fig. 1c), the Fermi level, E F , is constant across the junction, and the net current (I ) is zero [6,21,25,28,38].For V >0V (positive potential on SLG), the SLG E F is shifted below the 1L-WS 2 valence band energy E V (Fig. 1d), and h from SLG tunnel across the hBN barrier into 1L-WS 2 , promoting EL emission by radiative recombination between the injected excess h and intrinsic e [21-24, 28, 35, 38].The EL emission is expected to increase as a function of tunneling current because of the increasing h injected into 1L-WS 2 available for e-h recombination.
The LMs are characterized by Raman, PL, EL spectroscopy using a Horiba LabRam HR Evolution.The Raman spectra are collected using a 100x objective with numerical aperture (NA)=0.9,and a 514.5nm laser with a power∼5µW to avoid damage or heating.The voltage bias dependent PL and EL are collected using a long working distance 50x objective (NA=0.45).For the PL spectra, we use a 532nm (2.33eV) laser in order to excite above the X 0 emission (∼2eV) [9,10].The power is kept∼80nW to avoid laser-induced thermal effects [2,[9][10][11].The voltage (V ) and current (I ) between source (SLG) and drain (1L-WS 2 ) electrodes are set (V ) and measured (I ) by a Keithley 2400.
Fig. 2 shows the Raman spectrum of 1L-WS 2 /hBN/SLG on Si/SiO 2 after device fabrication and before current-voltage (I-V ) measurements.The Raman modes of each LM can be identified.For 1L-WS 2 , Pos(A ) FIG. 2. 514.5nmRaman spectrum of 1L-WS2/hBN/SLG LMH after device fabrication.The SLG and hBN Raman modes are labelled on it and the modes for 1L-WS2 as for Table 1.The 1300-2900cm −1 spectral window was multiplied by a factor of 10 for better visualization 3.9±0.2cm−1 , before assembly, to∼419.8±0.2cm−1 ; 3.4±0.2cm−1 , after.All the changes in the other modes are close to our spectral resolution and errors, as for Ref. [86].Pos(A ′ 1 ) and FWHM(A ′ 1 ) are sensitive to changes in n-doping [87,88].The mechanism responsible for this effect is an enhancement of electron-phonon (e-ph) coupling when e populate the valleys at K and Q simultaneously [88].The energy of the K and Q valleys is modulated by the A ′ 1 ph [88].Since the K and Q energies are modulated out-of-phase, charge transfer between the two valleys occurs in presence of the A ′ 1 ph [87,88].When the K and Q valleys are populated by e, these are transferred back and forward from one valley to the other [88,89].This increases the e-ph coupling of out-of-plane modes, such as A ′ 1 [88].The same process does not occur for p-doping [88].The reason for this asymmetry between n-and p-doping is due to a much larger energy separation (∼230meV [88]) between the VB Γ and K valleys than that (∼100meV [88]) of the CB K and Q valleys.From the changes in Pos(A ′ 1 ) and FWHM(A ′ 1 ), and by comparison with Ref. [88], we estimate a reduction in n-doping∼ 5 × 10 12 cm −2 .
In These indicate that the SLG is p-doped, with E F ∼150±50meV [93][94][95] by taking into account the average dielectric constant (∼3.85) of the environment (ε SiO2 ∼3.8 [96] and ε hBN ∼3.9 [97]).E F ∼150meV should correspond to Pos(G)∼1584.1cm−1 for unstrained SLG [98].However, Pos(G)∼1585.1±0.2cm−1 , which implies a contribu-tion from compressive uniaxial (biaxial) strain∼0.04%(∼0.01%).The strain level for SLG and hBN are different, most likely due to the fact that the SLG is directly exfoliated onto SiO 2 /Si, while hBN is picked up and transferred by PDMS stamps, hence, this could induce a larger amount of strain on hBN.Fig. 3a plots the I-V characteristics.For V =0V the current is zero (Fig. 1c).When V is applied, an electrical rectification (i.e.diode behavior) with negligible leakage current (I <10 −11 A) for V <0 is seen.A tunneling onset, (i.e.exponential increase of I ) is seen at V ON ∼4.1V, Fig. 3a.V ON is related to the breakdown electric field (E bd ) across the junction, which depends on the voltage drop on the hBN tunnel barrier and hBN thickness (d ) accordingly to E bd =(V bd /d )∼0.7-1V/nm [82,83], where V bd is voltage breakdown V bd =qnd 2 /(ε 0 ε hBN ), q is the e charge, n is total charge concentration, ε 0 =8.854×10 −12 F/m and ε hBN ∼3.9 [82,83], so that V ON can vary between different devices.When V >V ON , h from SLG tunnel across the hBN barrier into 1L-WS 2 , promoting EL emission by radiative recombination between the injected h and majority e in 1L-WS 2 (Fig. 1c) [21-24, 35, 38].The EL intensity∼634nm (∼1.956eV) increases with tunneling current, as in Fig. 3b.No light emission is observed in reverse V < 0V and small positive (0 <V <V ON ) biases, below the tunneling condition (V ON <4.1V).A red-shift∼48meV is observed in EL emission∼634nm (∼1.956eV) with respect to the PL X 0 emission of the unbiased device (dashed black line, Fig. 3b).Fig. 3b shows a EL peak position close to X − of unbiased PL (dashed black line, Fig. 3b), implying a trionic EL emission, due to excess e in 1L-WS 2 [28,38].
To further understand the EL emission origin, we perform EL and PL spectroscopy at the same V. Fig. 4a plots PL spectra at different V.At V =0V, the PL peak is∼619.2nm(∼2.002eV), assigned to X 0 [9,69].By increasing V (i.e.increasing e density in 1L-WS 2 ), a second peak appears at longer wavelengths (∼630nm,∼1.968eV),due to X − [9][10][11]99].For V >0V, the X 0 intensity gradually decreases and nearly vanishes, while X − shifts to longer wavelengths, Fig. 4a.This is expected for trionic emission, due to e-doping induced by V [9-12, 38, 99].Similar effects were observed in 1L-MoS 2 /SiO 2 /Si [101], hBN/1L-WSe 2 /hBN/SiO 2 /Si [6], and hBN/1L-WS 2 /hBN/SiO 2 /Si [28].Therefore, for similar tunneling current, EL agrees in energy and shape with the PL emission (see, e.g., the PL and EL spectra at the bottom of Fig. 4a). Tis is confirmed by Fig. 4b, where EL and PL peak positions are plotted for 4 devices, showing EL and PL emission at very similar wavelengths.Thus, EL predominantly originates from X − [6,9,10,21,38].The variations in X − energy for different LEDs are due to changes in charge carriers density across different samples.E.g., the charge density variation in 1L-WS 2 can be due to the number of vacancies in 1L-WS 2 [41] and external impurities (PC residues and  3b adsorbed water) after LED fabrication, which may vary from sample to sample.
Next, we estimate the external quantum efficiency (EQE) of our LEDs.This is defined as the ratio between the number of emitted photons (N ph ) and that of where λ N ph−counts is the sum of the total photons collected by the spectrometer over the measured spectral range, A ef f =AA/A spot , where A spot is the microscope objective spot size (A spot =π[1.22λ/2NA] 2 ∼2.2µm 2 , with λ=618nm and NA=0.45), and N h =I ×t /q, where t is the acquisition time, and q the e charge.The efficiency factor (defined as the ratio between the photons collected by the detector and the emitted photons by EL at the sample position) of our setup, including all optical components and spectrometer, is η sys ∼0.0051, see Methods.
From Eq.1 we get EQE∼0.025%±0.021%and∼0.195%±0.324%for pristine-and TFSI treated-LEDs, respectively, corresponding to a∼8.7±1.5-foldincrease, thus demonstrating that TFSI can boost EQE by almost one order of magnitude.It was reported that, using pulsed (AC) bias, EL emission can be enhanced a factor∼4 [33] and up to∼100 in a double optical cavity (distributed Bragg reflector (DBR) with an optical mirror) [106].Therefore, AC bias and photonic cavities could be combined with TFSI treatment to achieve EQE>10% in 1L-TMDs.
We now consider the EL emission features induced by TFSI treatment.By comparing EL before and after TFSI (Figs. 8a,b), a blue-shift in EL is observed.In pristine-LEDs, the EL emission is∼641.8nm(∼1.931eV),Fig. 9a, whereas after treatment it is∼625.6nm(∼1.982eV),Fig. 9b.Fig. 9c plots the EL peak position before and after treatment in 4 devices.After treatment, the EL emission shifts to shorter wavelengths, where X 0 is expected [68,69] (dashed line in Fig. 9c).In non-biased S-based TMDs devices, this shift could be due to the depletion of excess e in n-doped 1L-WS 2 due to TFSI [54][55][56][57][58][59][62][63][64].Nevertheless, we cannot neglect the additional charge density induced by V on the MIS capacitor.E.g. the I-V characteristics in Fig. 7 show that I and V ON do not change before and after TFSI, suggesting the same tunneling condition is maintained across the 1L-WS 2 /hBN/SLG junction.In both cases a comparable electric field (and electric charge) is developed across the junction for a given V. Fig. 7 implies that, independent of TFSI treatment, the same amount of negative charge is electrostatically induced in 1L-WS 2 at V >0.However, taking into account the EL spectral shift towards X 0 emission upon bias, the expected depletion of excess  ,c.Consequently, the emission profile is not compatible with the I-V curves before and after TFSI in Fig. 7, given that the electric field across the junction should be modified by the e density change in 1L-WS 2 .
To get a better insight on the effects of TFSI on 1L-WS 2 based LEDs, Figs.9d,e plot normalized PL spec-FIG.9. EL spectra of (a) pristine and (b)TFSI-treated LEDs at similar tunneling current∼12nA, fitted with Lorentzians.c) Position of EL emission for different LEDs before (black) and after (red) TFSI.Color-plot of the gated-PL of (d) pristine and (e) TFSI-treated LED at similar laser excitation power and integration time tra as a function of V before and after TFSI.In the pristine case (Fig. 9d), the PL map shows an evolution in emission spectra from∼620nm (∼2.000eV) to∼638nm (∼1.943eV), corresponding to a spectral shift from X 0 to X − due to excess e in 1L-WS 2 induced by V.After TFSI treatment (Fig. 9e), the PL exhibits only a minor shift from∼618nm (∼2.006eV) to∼622nm (∼1.993eV), implying that the induced e-charge in 1L-WS 2 does not contribute to the X − emission pathway.Therefore, similar to Figs. 9a,b, PL also indicates that the emission after TFSI treatment predominantly originates from radiative recombination of X 0 , independent of V. Refs.[54,56,[58][59][60][61] claimed that TFSI treatment reduces the extent of n-type behavior in S-based 1L-TMDs due to S vacancies passivation, consistent with the suppression of X − formation in Refs.[57,[62][63][64][65]. Ref. [8] reported that TFSI acts as a Lewis acid, i.e. it can accept an e pair from a donor [53], suppressing X − formation.Whereas Refs.[50][51][52] claimed that TFSI may activate sub-gap states and reduce the n-type behavior in S-based TMDs, as well as reducing X − formation.Our I-V, EL and gated-PL results suggest that TFSI treatment i) depletes the excess e in 1L-WS 2 , acting as a Lewis acid [8] and ii) favours the radiative recombination of X 0 independent of bias, due to the activation of trapping states [50,52] in 1L-WS 2 caused by the treatment.One would expect changes in the excitonic emission at such trapping states at RT, where the thermal energy can assist carrier de-trapping, and radiative recombination from excitons [64].Therefore, the modification from non-radiative to radiative recombination by activation of trapping states could be further engineered to achieve more efficient optoelectronic devices.

CONCLUSIONS
We demonstrated a one order of magnitude enhancement in EL emission of 1L-WS 2 -LEDs by performing TFSI treatment.EL predominantly originates from trions in pristine devices, while neutral excitons dominate in treated ones.The neutral excitonic emission is also restored in 1L-WS 2 gated-PL measurements.We attribute these changes to a reduction of n-doping of 1L-WS 2 , as well as changes in the relaxation and recombination pathways within 1L-WS 2 .This paves the way to more efficient 1L-TMDs-based LEDs, and shed light into tunability of the excitonic emission of these devices.

Raman characterization of LMH individual constituents
Raman spectroscopy allows us to monitor LMs at every step of device fabrication.This should always be FIG.10.(a) Low-and (b) high-frequency 514.5nmRaman spectra of 1L-WS2 (red) and bulk-WS2 (black) on Si/SiO2, normalized to the Si peak, with labels as for Table 1 performed on individual LMs before and after assembly in LMHs and devices.This is an essential step to ensure reproducibility of the results, but, unfortunately, this is often neglected in literature.
The first order Raman modes, i.e.The 2LA(M) mode originates from a second-order double resonant process [118][119][120], where momentum conservation is satisfied by two LA ph with opposite momenta around K-and M-points [119], therefore sensitive to differences in band structure between bulk and 1L-WS 2 [68,121].
This can be explained considering that the main first-order (E ′ , A ′ 1 ) and secondorder (2LA(M)) Raman modes are enhanced for 2.41eV excitation, due to exciton-ph coupling effects involving B exciton transitions [116,122].These depend on mode symmetry (i.e.differ between out-of-plane and in-plane modes) as well as N [118].In bulk-WS 2 , the out-ofplane A 1g is resonant with the B exciton, unlike E 1 2g [118].The enhancement of A 1g decreases with decreasing N due to the dependence of the lifetime of the intermediate excitonic states on N [118].The difference between I(2LA)/I(E ′ 1 ) in 1L-WS 2 and I(2LA)/I(E 1 2g ) in bulk-WS 2 is due to a change in band structure from direct bandgap in 1L to indirect in bulk-WS 2 [68][69][70][71], which changes the double resonance conditions [118][119][120].
Since LA(M) and LA(K) and E 2 2g (M) are one-ph processes from the edge of the BZ (q =0) [68][69][70][71], they should not be seen in the Raman spectra since, due to the Raman fundamental selection rule [123], one-ph processes are Raman active only for ph with q∼0, whereas for multi-ph scattering the sum of ph momenta needs to be∼0 [118][119][120][121]. However these modes can be activated in presence of defects, as these can exchange momentum with ph, Fig. 11 plots the Raman spectra of a∼3nm hBN flake (black curves) and bulk-hBN (red curves).The latter has 2 Raman-active modes [124,125], C and E 2g .
The red curves in Figs.12a,b are the Raman spectra of SLG on SiO 2 /Si before LMH assembly.
Thus, the presence (or coexistence) of biaxial strain cannot be ruled out.

Spectrometer efficiency
The η sys of our spectrometer is derived as follows.We use a 50x objective (NA=0.45).Hence, the solid angle is θ=(1-cosθ)×2π, where θ=arcsin(NA/n), and n is the refractive index.
Therefore, the calculated overall collection+Horiba efficiency is: M 50x−ef f ×(M ef f ) 7 ×S ef f ×G ef f ×CCD ef f ∼0.0067.To experimentally validate the calculation, we use a 0.5pW laser at 632.8nm and measure the counts at the CCD detector N counts =149748.The photon energy at 632.8nm is E ph =(1.24/0.638)×1.6e−19 =3.13e −19 J.The laser power is P opt =0.5e −12 J/s.As a result, if the system efficiency is 100% we expect to get 0.5e −12 /3.13e −19 =1597444 counts.

FIG. 1
FIG. 1. a) Schematic of LED.Cr/Au electrodes, SLG, FLG, hBN, and 1L-WS2 are indicated.b) Optical image of device.Scale bar 4µm.The dotted lines highlight the footprint of SLG, FLG, hBN, 1L-WS2.The green-shaded part corresponds to the active area∼23µm 2 .Cr/Au contacts the bottom SLG; FLG contacts the top 1L-WS2.Band diagram for (c)V =0V and (d) V >0V.Tuning the SLG EF (gray dotted line) across the 1L-WS2 valence band edge, EV , allows h tunneling from SLG to 1L-WS2, resulting in current onset and light emission via radiative recombination with e from the ntype 1L-WS2.The blue circles represent e accumulated on 1L-WS2 due to the MIS structure, while the red circles are h injected into 1L-WS2 through the hBN barrier

FIG. 3
FIG. 3. a) I as a function of V for 1L-WS2-LED.b) EL spectra for different tunneling currents without TFSI treatment.The dashed black line is the PL spectrum collected at V =0 and normalized to the maximum EL intensity.

FIG. 4
FIG. 4. a) Evolution of PL as a function of V.For comparison, an EL spectrum for I∼16nA is shown (red).The dashed lines are guides to the eye for the X 0 and X − positions.In all PL measurements up to 3V, I <10 −11 A. At 4V, I ∼10nA, indicating h tunneling through hBN into 1L-WS2.b) EL and PL positions from 4 different devices.The dashed line plots the unbiased PL position of X 0 measured in Fig.3b

FIG. 7 .
FIG. 7. I -V curves of 3 LEDs before (solid black lines) and after (dashed red lines) TFSI treatment

TABLE I .
Pos and (FWHM) in cm −1 of WS2 Raman peaks, before and after LMH assembly, and TFSI treatment