Trapping the carrier in the spin-locked MoS₂ atomic valley by absorption of chiral L-cysteine Open Access

In a chiral/non-centrosymmetric structure, a valley is a quantum number that defines the electronic system using energetically degenerate energy bands with non-equivalent local minima (conduction band) or maxima (valence band). In 2D hexagonal crystals, such as single layer (SL) MoS₂, where inversion symmetry is broken, two degenerate valleys can be distinguished by a pseudovector quantity, such as Berry curvature and magnetic moment taking opposite values at time reversal pair of valleys.^1,2 The valley contrasted Berry curvature and magnetic moment can couple to external electric and magnetic fields. Because of inversion symmetry breaking, optical interband transitions at time reversal pair of valleys can have valley dependent selection rules. The concept of valleytronics deals with the use of the valley index of the carriers (i.e., spin and pseudospin) that process information.³ The strong spin–orbit coupling in transition metal elements gives rise to effective interaction between the valley pseudospin and spin, helping a possible interplay between these two degrees of freedom and allowing spin manipulation via the valley phenomena.² It was found that pseudospin polarization is encoded in the corresponding wavefunction and dipole vector of the optical matrix element.⁴ The relaxation of exciton valley pseudospin can also arise from the exchange coupling between its electron and hole constituents. Therefore, this valley dependent phenomenon leads to possible manipulation of valley pseudospin by electric, magnetic, and optical means. Thus, exploring these internal quantum degrees of freedom of carriers will lead to ascertaining their potential usage for new generation electronics for storing and processing more information.

The valley depolarization phenomena in SL-MoS₂ can be viewed by photoluminescence/radiative transitions from excitons and trions as well as non-radiative phonon or defect assisted emission. Valley polarization, the selective population of one valley, can be achieved by tuning the incident photon angular momentum. This polarization can be preserved for longer than 1 ns. In recent literature, full control over the bright−dark splitting has been described as a pathway to manipulate and control the exciton valley pseudospin dynamics, thus the associated valley polarization. Designing Coulomb forces between electron and hole and their exchange interaction has also been described as a useful way for the same. In SL-MoS₂, the lifetimes of the excitons and trions become larger when immersed in high-κ dielectric environments.⁵ Coulomb potential becomes more confined within the middle layer (i.e., SL MoS₂) if the static dielectric constants of the top and bottom layers are higher. This may help to reduce the scattering between the excitons (trions) and the charged impurities at the MoS₂/dielectrics interface, thus leading to a prolonged lifetime of the excitons (trions). Modification in the interfacial interaction^6–8 between SL–Transition Metal Dichalcogenides (TMD) and preferred molecules is a familiar way to modulate carrier concentration as it does not affect the crystal structure. The spectral weight between excitons and trions of SL-TMDs can be fine-tuned by controlling carrier densities of SL-TMDs by means of electrical/chemical doping.⁹ Calculation of spin density in the presence of n-or p-type dopant clarifies that the p-orbital of the dopant plays an important role in modulating electronic and magnetic properties of layered MoS₂, although the role of the d-orbital of Mo is most significant.¹⁰ Here, it is to be noted that different surface treatments of SL-MoS₂ have been reported only to improve the photoluminescence, but their influence on the valley polarization has not yet been explored until today.

Highly polarized valley singlet and triplet interlayer excitons are found in van der Waals heterostructures¹¹ following electron–hole separation and discretized singlet and triplet transitions. Spectroscopically, a spin singlet exciton has an in-plane transition dipole.¹² It couples to $σ$+ (⁠$σ$−) polarized photon [left (right) circularly polarized] propagating in the out-of-plane (z) direction. On the contrary, the weak radiative recombination of a spin-triplet exciton, which is generally dark, can emit a z-polarized photon propagating in the in-plane direction.¹³ Brightening of dark excitons by imposing in-plane magnetic field¹⁴ was found as a possible strategy to enhance the population of excitons and valley polarization. The population decay of exciton following different radiative and non-radiative processes can be utilized to understand the extent of valley polarization and the preferred decay channels. Unexpectedly, at low carrier density, the substrate induced localized in-gap states degrade valley polarization in a 2D layered material.¹⁵ At room temperature, the decay in valley-based photoluminescence helicity is mostly governed by the electronic spin relaxation, and phonon-assisted^16,17 and defect-assisted scattering phenomena.¹⁸ At ambient conditions, the emission spectrum is mostly governed by the negative excitonic transition. The spin relaxation of negatively charged exciton becomes equivalent to intervalley relaxation,¹ and the possible spin relaxation mechanisms for a charge carrier in a semiconductor are considered to control the relaxation of valley spin polarization in SL MoS₂. Therefore, the manipulation of valley spin by coupling it with electron spin of chiral molecule of preferred symmetry will be an ingenious way to achieve efficiently enhanced valley polarization in SL MoS₂ at ambient temperatures.

In this study, the modified spin relaxation mechanism at the chiral L-cysteine–SL MoS₂ van der Waals interface at ambient conditions is investigated. The extent of valley polarization in pristine SL MoS₂, SL MoS₂ when attached with chiral/achiral molecule (Ch-SL/aCh-SL), is clarified by considering increment in photoluminescence helicity contrast (ρ). The possible depolarization pathways explored by helicity-resolved Raman spectra will further confirm the manipulation of spin/pseudospin in SL-MoS₂ valley exotic states in the presence of chiral L-cysteine molecule at ambient conditions.

Monolayers of MoS₂ were mechanically exfoliated from bulk MoS₂ (SPI supplier) and were placed on SiO₂/Si surface [details of the exfoliation procedure mentioned in Sec. S1 of the supplementary material, including the AFM image and height profile in Figs. S1(a) and S1(b)]. The location and thickness of the flake were identified by Raman spectroscopy and PL characteristics feature [Figs. 1(a) and 1(b), respectively]. The sample was then annealed at 250 °C for 1 h. For attachment of achiral cysteamine and chiral L-cysteine on the SL MoS₂ surface, the MoS₂ sample on SiO₂/Si substrate was soaked in 1/40 v/v cysteamine/methanol and L-cysteine/isopropyl alcohol (IPA)–water solution for 72 and 24 h, respectively. The samples were then rinsed thoroughly with ethanol and IPA to remove unbound molecules. As the deposition of thick organic layer degrades device performance, the above-mentioned protocol is followed to get optimized concentration of equal length chiral L-cysteine and achiral cysteamine molecules having O/S terminal localized at the S-defect sites.^9,19,20 Figure S2 of the supplementary material shows that achiral cysteamine and chiral L-cysteine molecules were tightly bound on the surface of MoS₂ (aCh-SL and Ch-SL) at their optimal concentration after repeated washing.

The helicity-resolved Photoluminescence (PL) and Raman measurements were performed with a micro-Raman spectrometer (M/S Horiba) equipped with a Peltier cooled CCD. The optical setup has been described in Fig. 1(c). The excitation laser was first guided through a vertical linear polarizer followed by a quarter-wave plate to achieve $σ$+(left) circular polarization. The circular polarization of the excitation light was confirmed at the sample position. The backscattered signal passes through the same quarter-wave plate and is collected and analyzed with a half-wave plate and a linear polarizer. Rotation of the half-wave plate at different angles enables the selection of the helicity (ρ) of the scattered light. $σ$− polarization of incident light gives similar selectivity and hence only results with $σ$+ polarization of incident light will be discussed. Two different laser sources are used for the measurements: Nd:YAG solid state laser (532 nm, 2.32 eV), and red laser (660 nm, 1.88 eV) with incident fluence of 1.7 × 10⁻² and 3.4 × 10⁻² W/m². The spectral resolutions are 0.45 cm⁻¹ (for 532 nm) and 0.35 cm⁻¹ (for 660 nm) for helicity resolved Raman experiments using 1800 lines/mm grating. In the text, the degree of PL polarization has been quantified by the helicity contrast and will be discussed later. For low temperature measurements, a Linkam THMS-350 heating and cooling stage was used to obtain temperature variation up to 78 K. All the PL and Raman spectra are collected at ambient conditions (300 K) and 78 K in two different detection configurations (parallel and cross polarization). Raman shifts were calibrated using Si Raman peak at ∼520 cm⁻¹.

To confirm the quality of our pristine sample, Raman and PL measurements using excitation of 2.32 eV were employed, as shown in Fig. 1 [panels (a) and (b)]. SL MoS₂ is a direct bandgap semiconductor with energy gap located at K and K′ points of the Brillouin zone where the highest and the lowest parts of the corresponding valence and conduction bands have significant contributions from Mo d-orbitals and S p-orbitals. The valence band is split by ∼150 meV due to the spin–orbit interaction. Raman active A_1g (405 cm⁻¹) and E_2g¹ (386 cm⁻¹) modes in backscattering geometry are widely used to measure the layer thicknesses and crystal quality. The difference of ∼19 cm⁻¹ in the Raman shift for the E_2g and A_1g modes and PL characteristics spectra for SL-MoS₂, as shown in Figs. 1(a) and 1(b), confirm the high quality of the monolayer. The peak positions marked by gray color lines in the PL spectrum of Fig. 1(b) signify an intense peak at 1.84 eV related to trion (A1) due to unintentional n-doping of SL-MoS₂, exciton A0 at 1.90 eV, and B-exciton at 2.0 eV at 300 K (assignment of peak related to excitonic transitions matches previous experimental observations as described in Refs. 5 and 21). The peak position of a possible exciton bound to tiny unintentional defect²² induced in gap state in ambient conditions, X_D, is also marked by a gray vertical line. Upon changing the temperature to 78 K, the excitonic peak shifts to 1.96 eV and trion peak shifts to 1.87 eV. Our excitation energy, 1.88 eV (marked by a red arrow), is in resonance with the neutral exciton A0 at 300 K and with the trion at 78 K (data not shown).

Figure 1(c) describes a schematic diagram of our experimental setup. SL-MoS₂ on Si/SiO₂ substrate is excited with left circularly polarized light (⁠$σ$+) of excitation energy 1.88 eV (in resonance with A0) and detected separately for $σ$+ (left) and $σ$− (right) emission configuration for all three types of samples. Table SI of the supplementary material describes the detection configuration. Figure 1(d) schematically describes the radiative recombination of neutral and charged excitons via intravalley and intervalley relaxations in steady state, respectively. The valence and conduction bands in SL MoS₂ are split into four bands, although the conduction band (CB) splitting is much smaller than the valence band. For the singlet exciton, the hole in the lower state has the same spin as the electron in the CB, while, for the triplet exciton, it is the opposite. Thus, in SL MoS₂, the excitonic transition is normally singlet in nature, although triplet transitions are induced by conduction band splitting.

The modifications in the SL-MoS₂ by adsorption of chiral and achiral molecules are clarified by off-resonance (2.32 eV) PL [(a) and (b)] and Raman spectrum [(c) and (d)] as shown in Fig. S2. A red-shift and decrease in luminescence signify n-doping for a aCh-SL system, while an enhancement in luminescence for Ch-SL suggests p-doping. In Figs. S2(c) and S2(d), the observed shift in A_1g and E¹_2g for attachment of both the molecules is found to be negligible (∼0.5 cm⁻¹), implying the doping level to be ∼10¹²/cm². The unpolarized emission with off-resonance 2.32 eV excitation for all the three samples, (a) SL-MoS₂, (b) Ch-SL, and (c) aCh-SL, are shown in Fig. S3 and has been interpreted as an indication of loss of valley polarization due to simultaneous population of both the valleys.¹ 

The connection between valley degree of freedom and photon helicity in SL MoS₂ is understood based on angular momentum conservation.²³ Thus, the valley-selective excitation near the K (K′) point is also the spin-selective excitation when the excitation laser energy matches the energy gap. This fact makes it possible to detect the valley polarized electrons by the direction of spin. The spin relaxation of a singlet negatively charged exciton (i.e., trion) reflects the relaxation of a hole. In all the three samples, SL, aCh-SL, and Ch-SL, hole spin relaxation is a measure of depolarization. Here, it is considered that, for holes, valley and spin indices are locked at the valence band Bloch states. The $σ$+ excitation creates excitons with electron spin up and hole spin down at the K point. In the literature, it is shown that the valley depolarization originates from the electron/hole spin relaxation due to the D’yakonov–Perel’ (DP), Elliott–Yafet (EY),²⁴ and BAP mechanisms. PL depolarization also includes intervalley scattering, including the electron–phonon and/or short-range impurity scatterings. The spin relaxation of the electron and hole can even cause the bright exciton transition between the K and K′ valleys. Intravalley scattering is forbidden due to the spin degeneracy near the valence band edge lifted by ∼150 meV. Again, intervalley scattering from the K to K′ point involving simultaneous spin flip requires coupling with both the atomic scale scatterers and with magnetic defects.¹² Figure 2 describes the optical control of valley spin polarization in SL MoS₂, nearly in resonance with A0 (exciton) and A1(trion) at ambient and low temperatures. A steady state, helicity-resolved photoluminescence spectra with $σ$+ excitation for pristine SL in ambient conditions (300 K) is shown in Fig. 2(a). The dark blue and orange lines correspond to parallel and cross detection configurations of the spectrum, respectively. The area under the curve describes the population of excitons and trions in the corresponding configuration. The observed steady state PL spectra is mainly governed by the valley polarization during the valley carrier initialization process. In the same detection configuration (⁠$σ$+$σ$+), valley selective carrier relaxation is observed, while in cross configuration (⁠$σ$+$σ$−), the valley selectivity is lost. The extent of valley polarization quantified as helicity contrast (ρ) is defined as

where I₊ and I₋ are the intensities of the PL signals corresponding to the same (⁠$σ$+$σ$+) or cross (⁠$σ$+$σ$−) circular polarization with respect to incident polarization (σ+),¹ while n_k and n_k′ are the population of excitons in K and K′ i.e., hole spin up and hole spin down valleys, respectively. Here, it is considered that, for holes, valley and spin indices are locked at the valence band Bloch states. Figure 2(b) describes the spectral variation of helicity contrast (ρ) at 300 K for all the three samples, SL (red line), Ch-SL (blue line), and aCh-SL (black line). For SL, the spectrum consists of equal dominance of luminescence of A0 and A1 peaks at 1.86 eV in the $σ$+ polarization detection configuration, but the extent of polarization is found to be more for Ch-SL sample (ρ ∼ 0.25). A very weak emission from defect trapped exciton, X_D, at 1.77 eV²⁵ is observed for $σ$− polarization detection configuration SL sample. For aCh-SL, the helicity contrast peak is lower and red shifted to 1.84 eV due to more contribution from the trion A1. Here, we note that proximity to surface defects, dangling bonds, dielectric disorder, surface roughness of the Si/SiO₂ substrate,¹⁵ and temperature degrade the valley polarization of the MoS₂ sample. Figure 2(c) describes the helicity resolved PL spectra for SL at 78 K, and subsequent helicity contrast spectra for all three samples are shown in Fig. 2(d). For comparison between helicity contrast (ρ) for valley exciton A0, trion A1, and defect exciton X_D observed at two temperatures (300, 78 K), a blue dashed line is drawn in Figs. 2(b) and 2(d) as a guide to the eye to estimate the extent of polarization. It can be seen that at 78 K, ρ value for A exciton is higher for Ch-SL sample at 1.86 eV compared to the SL-MoS₂. The spectrum for aCh-SL portrays an enhanced trion helicity (ρ ∼ 0.3 compared to pristine A exciton ρ ∼ 0.25) at 1.84 eV. As with the decrease in temperature, the photoluminescence peak blue shifts and the maximum extent ρ was at an energy higher than the excitation energy of 1.88 eV. With decreasing temperature, as described in Fig. 2(d), ρ is maximized for Ch-SL at scattered photon energy of 1.86 eV. However, it is observed that defect passivated aCh-SL leads to small enhanced helicity contrast (ρ) around 1.84 eV (trion A1) at 78 K, as shown by black curve in Fig. 2(d).

According to the BAP mechanism, spin relaxations through the e–h exchange interaction are prominent for 2D materials as compared to other mechanisms. The presence of such an interaction allows a valence hole spin in lightly n-doped MoS₂ monolayer to relax through a simultaneous valley- and spin-flip scattering with the conduction band electrons, i.e., intervalley scattering. It does not include any intravalley phenomena or spin-flip phenomena for the same valley. It includes an exchange splitting of ∼1 meV, and an unintentional doping level of 5 × 10¹² cm⁻², which are met for all three samples. Thus, the BAP mechanism appears to be an effective relaxation process in monolayer MoS₂ even at 78 K for all three samples for favoring the excitonic singlet transition. The other mechanisms, such as DP and phonon mediated scattering, are less effective at low temperatures. This claim will be further clarified by helicity resolved Raman scattering. Other than the above-mentioned mechanism, the substrate induced effect also plays a role in valley depolarization. In SL MoS₂, the precise 2D motion of the charge carriers and the presence of mirror symmetry plane in the D_3h point group suggest that the effective magnetic field felt by the spin of a charge carrier has no in-plane component. Mirror symmetry in SL MoS₂ can be broken in substrate-supported samples or in field-effect transistor structures in which electric fields may be present. It could lead to spins of charge carriers with different crystal momenta to precess at different rates between scattering events with possible connection between acoustic phonons LA at K, K′ points.

Resonance Raman scattering of different phonon modes with polarized incident radiation reveals information of electronic states with specific symmetries present at the band edge that cannot be extracted with non-resonance excitation. When circularly polarized light is incident, resonant Raman intensity of the first order bands of pristine MoS₂ depends on electron–photon and electron–phonon matrix elements related to the initial and final states.^16,17 Contrary to a first order Raman band, a second order Raman band can be expressed as a convolution of multiple two phonon processes across the Brillouin zone and, very often, the scattering processes are found to be doubly degenerate.²⁶ 

Figures 3(a)–3(c) describe the resonance Raman spectra for the spectral range 350–550 cm⁻¹ of SL, Ch-SL, and aCh-SL at ambient temperatures (300 K), while Figs. 4(a)–4(c) describe the same collected at 78 K when excited with 1.88 eV excitation after subtracting the excitonic background. The Raman spectra for the same range, including excitonic background, are shown in Figs. S4 and S5, respectively. For incident light $σ$+ polarization and detection in $σ$+/$σ$− (same and cross) configuration, the observed features are compared with the selection rules in Table S1 (Ref. 16). The A_1g mode is present for all three samples, with its maximum intensity for $σ$+ polarization configuration in resonance (1.88 eV) as well as off-resonance excitation (data not shown), and it agrees with Raman selection rules for all the three samples at room temperature. For the E_2g mode, the behavior is different for excitation with 1.88 eV compared to the off-resonance condition.²⁷ The crystal symmetry forbids the E_2g mode to be present in parallel configuration and allows in cross configuration with maximum intensity. Contrary to this, E_2g mode is found to be present for all the three samples for same incident and scattered polarization at ambient conditions (300 K), as shown in Figs. 3(a)–3(c). In comparison at 78 K, the intensity of the same mode is lower even in cross polarization configuration, as shown in Figs. 4(a)–4(c). This observation can be interpreted as reduced coupling of the in-plane E_2g mode with the A exciton.^28,29 An excitonic transition connects lower part of the conduction band composed of Mo d_z² (for the electron) and upper part of the valence band of d_xy character for the hole. Thus, stronger electron–phonon coupling is expected for out-of- plane A_1g vibration, and hence, under resonance, A_1g feature is enhanced in comparison to the in-plane E_2g mode. Another interesting feature is that the band around 450 cm⁻¹, associated with the second order acoustic phonon mode, shows a prominent change in line shape in cross polarization at 300 K. At low temperature, 78 K, the band line shape remains unaltered for both the detection configurations for all three samples.

The resonant excitation reduces the number of possible spin relaxation mechanisms, such as coupling to the excited states. It is known that a threshold of twice the zone edge LA phonon energy above the bandgap is required above which phonon-assisted intervalley scattering can cause depolarization.¹⁷ There are two possible mechanisms responsible for the electron or hole spin-flip during this phonon-mediated intervalley scattering event: (i) spin-flip mediated by short range scattering from impurities, the presence of a background carrier population could enhance the probability of such a process²³ and (ii) intervalley scattering process through the nearly spin-degenerate $Γ$ valley of the Brillouin zone (BZ).³⁰ Following this, large-momentum low-energy exciton states can provide relaxation channels for bright excitons and reduce photoluminescence quantum efficiency. Following this, for SL MoS₂, first-order Raman bands are used for estimation of polarization^16,23 and the second order double resonance Raman band of SL-MoS₂ are used for estimation of valley depolarisation as they originate due to intervalley scattering by acoustic phonons, a mechanism responsible for the destruction of valley polarization.¹⁷ 

To estimate the extent of polarization and the proper decay channel responsible for valley depolarization, the spectral range of 355 to 490 cm⁻¹ has been deconvoluted with a sum of Lorentzian functions after subtracting the excitonic background. Figures 3(a)–3(c) describe the deconvoluted spectra for SL-MoS₂, Ch-SL, and aCh-SL where the contribution of E_2g (380 cm⁻¹), A_1g (406 cm⁻¹), 2LA (419 cm⁻¹) and three other second order bands are considered. Under resonance conditions, higher order combination bands and zone edge acoustic phonons near M and K points of the Brillouin zone, such as 2LA (K/M) ∼ 450 cm⁻¹, are seen along with the luminescence background. Three Lorentzian functions (notation P1 ∼ 436 cm⁻¹, P2 ∼ 452 cm⁻¹, P3 ∼ 462 cm⁻¹, width ∼ 14 cm⁻¹) are used to describe the broad spectral feature around 450 cm⁻¹. The non-dispersive second order P2 mode is related to K and M points in BZ, while the dispersive P3 mode is related to scattering process from K to K′. Si peak at 520 cm⁻¹ is used as internal standard to calibrate the frequency shift. During the deconvolution procedure, the intensity, width, and peak position are considered as free fitting parameters as mentioned in Ref. 17.

At 78 K, the same fitting procedure is followed, considering the hardening of Raman mode. As E_2g peak height remains unaltered in two detection configurations at 1.88 eV, it has been used for intensity calibration.

In a recent study by Lee et al.,²⁶ the appearance of E_2g mode in parallel polarization was related to a defect-assisted process that involves phonons in the transverse optical E branch slightly away from the $Γ$ point. This process can be enhanced by the selective coupling of valley pseudospin to photon helicity. Here in our observation, presence of prominent E_2g mode in parallel detection configuration is also observed for all three samples (Fig. 3) at ambient condition. Thus, we can confirm that in our system, the valley pseudospin, in addition to the crystal symmetry, plays a significant role in understanding the resonance Raman scattering spectra for excitations close to valley exciton resonances (A exciton here). At 78 K, intensity of the E_2g mode compared to the A_1g mode is lower for all three samples, as shown in Figs. 4(a)–4(c), due to the modified resonance condition. Following this, the intensity of A_1g mode will be considered as the extent of valley polarization marker band i.e., the coupling of the out-of-plane A_1g mode with exciton/trion and depolarization pathways, non-dispersive second order P2 mode connecting K and M points in BZ; dispersive P3, related to scattering process from K to K′ at ambient (300 K) and low temperature (78 K) will be discussed.

Figures 5(a) and 5(b) describe the change in intensity of the A_1g mode in two detection configurations (open symbol for parallel and filled symbol for crossed configuration) at 300 and 78 K, respectively. The parallel detection configuration is considered to define intravalley relaxation mechanism, while the cross-detection configuration defines intervalley spin relaxation mechanism. It gives an idea about the light coupling efficiency for the extent of valley polarization that can be reached for the mode upon 1.88 eV excitation. At 300 K, the intensity of out-of- plane A_1g mode at $Γ$ point is lower in cross configuration for the SL-MoS₂ as compared to the Ch-SL and aCh-SL samples. Adsorption of achiral cysteamine/chiral cysteine on MoS₂ modifies the fast relaxation via optical phonon near K point at ambient conditions, and as a result, the A′ or A_1g mode (other than $Γ$ points in BZ) contributes in the cross-polarization configuration. For aCh-SL, the drop in A_1g mode is minimum. At 78 K [Fig. 5(b)], as the laser excitation energy is in resonance with the A1 trion peak, the excess electron–phonon contribution becomes prominent for all three samples. In Fig. 5(b), the enhanced A_1g intensity is observed for all three samples, while drop in the A_1g mode intensity is minimum for aCh-SL in cross polarization.

Next, we look into the relative intensity ratio of P2 and P3 mode (with respect to the E_2g mode for each of the samples) in Figs. 5(c) and 5(d) for 300 and 78 K. At 300 K, in parallel configuration, higher value for P2 mode compared to P3 mode is observed as shown in Fig. 5(c). In cross configuration, where intervalley scattering is probed, P3 mode is found to be dominated over P2 mode for SL, aCh-SL. On the contrary, for Ch-SL, the P2 and P3 modes are found to have similar intensity in cross configuration. Thus, the effective depolarization from K to K′ or via other points in the BZ are decreased in the case of Ch-SL sample at 300 K.

Figure 5(d) shows that, for all three samples, P2 mode is dominant over P3 at 78 K even in cross polarization configuration. It implies that the intensity of K to M transition (intravalley) is much higher than the K–K′ intervalley phonon mediated transition. The P3 mode disperses rapidly away from K point, and the probability of intervalley transition becomes much lower at 78 K. Here, it should be mentioned that, at 78 K, A₀ exciton selectivity is lost, which is being reflected in the corresponding Raman features as well. The excitation laser line is in resonance with the A1 trion peak at 78 K. The DP mechanism (intervalley) fades out and only BAP mechanism (intravalley) becomes responsible for spin relaxation. For proper estimation of depolarization mechanism at 78 K, the samples need to be probed at A₀ excitation wavelength. Thus, for pristine SL-MoS₂ on Si/SiO₂ substrate, spin-locked intravalley transitions are lowered in the presence of intervalley (possibly K–K′ transitions, transitions connecting M and K points in the BZ, the presence of in-gap states due to unintentional defects and substrate effect) transitions following relaxation channels via multiple acoustic phonons. In aCh-SL, the sulfur-deficient states are removed, but the chemical coordination of the molecule with SL-MoS₂ induces a small n-doping effect (∼10¹² cm⁻²), leading to generation of trion weighted spectral feature. While radiatively decaying, trions eject an electron and the recoil electron carries away the trion’s momentum, allowing all trions to decay radiatively. Therefore, including the energy of the recoil electron is essential when determining the PL spectral shape from the trion momentum distribution.¹⁹ In aCh-SL placed on the Si/SiO₂ substrate, the momenta of charge carriers are randomized by scattering and result in unbalanced in-plane component of magnetic field and spin lifetimes for holes, and hence the enhancement in emission is not observed at 300 K.¹⁵ It induces more channels for spin relaxation (as observed from enhanced 2LA mode) and reduces the valley polarization. The enhanced valley contrast (ρ) in the presence of chiral L-cysteine molecules is mostly governed by the formation of L-cysteine dimer at pH 10. The formation of dimer is driven by optimization of each cysteine molecule on MoS₂ similar to the case of (110) Au surface.³¹ It introduces selective adsorption of oxygen (O) at the sulfur defect site as well as chemically absorbed on top of S atom present at the MoS₂ basal plane.³² It can be inferred that adsorption of O at the basal plane of MoS₂ balances the substrate induced magnetic field, thus reducing depolarization due to in-gap states. This adsorption modifies the behavior of the confined charge carriers in Ch-SL-MoS₂ at the Si/SiO₂ substrate and reduces the unbalanced K dependent components. A schematic illustration of the absorption phenomena is shown in Figs. 6(a) and 6(b), and a comparison of valley based transitions in Figs. 6(c) and 6(d) at the interface of aCh-SL MoS₂ and Ch-SL MoS₂ is depicted. For Ch-SL MoS₂, the K to K′ intervalley transition and transition connecting M and K points are lowered by preferred locking of spins at the VB states. As a result, an enhanced valley contrast transition in Ch-SL MoS₂ sample is observed. Hence, following this simple and effective method, controlling the population of spin-locked states of confined carrier in 2D SL-MoS₂ by adsorption of chiral L-cysteine is possible. A further estimation of exciton/trion g-factor temperature dependence from magneto PL measurement^33,34 will give precise estimation of the spin locked states in the modified atomic valley.

In summary, we find that the defect passivation with achiral cysteamine molecule induces n-doping, and cannot remove in-gap states formed due to the Si/SiO₂ substrate. The helicity dependent photoluminescence and resonance Raman scattering measurements highlight that balanced in-plane magnetic field due to selective absorption of chiral molecule with spin–orbit effective field component in Ch-SL MoS₂. The spin relaxation in the system is modified by lowering the intervalley transitions. This gives rise to enhanced valley polarization in Ch-SL MoS₂ at ambient conditions and, thus, Ch-SL MoS₂ interface can be utilized as a model system to maximize valley polarization. We believe that our experimental study can be applicable for room temperature valley information-based LEDs and photonic logical devices.

See the supplementary material for additional information on proposed symmetry, assignment of Raman modes for SL-MoS₂, microscopic and spectral characteristics of SL-MoS₂, and helicity resolved PL when excited with 2.32 and 1.88 eV excitation, respectively.

S.B. acknowledges DSKPDF, UGC for a fellowship. A.K.S. acknowledges the Department of Science and Technology for financial support under the Nanomission Project and Year of Science Professorship.

The author have no conflicts to disclose.

The data that support the findings of this study are available within the article and its supplementary material.