Competitive Effects of Oxidation and Quantum Confinement on Modulation of the Photophysical Properties of Metallic-Phase Tungsten Dichalcogenide Quantum Dots

Metallic-phase transition metal dichalcogenide quantum dots (TMDs-mQDs) have been reported in recent years. However, a dominant mechanism for modulating their intrinsic exciton behaviors has not been determined yet as their size is close to the Bohr radius. Herein, we demonstrate that the oxidation effect prevails over quantum confinement on metallic-phase tungsten dichalcogenide QDs (WX2-mQDs; X = S, Se) when the QD size becomes larger than the exciton Bohr radius. WX2-mQDs with a diameter of ~12 nm show an obvious change in their photophysical properties when the pH of the solution changes from 2 to 11 compared to changing the size from ~3 nm. Meanwhile, we found that quantum confinement is the dominant function for the optical spectroscopic results in the WX2-mQDs with a size of ~3 nm. This is because the oxidation of the larger WX2-mQDs induces sub-energy states, thus enabling excitons to migrate into the lower defect energy states, whereas in WX2-mQDs with a size comparable to the exciton Bohr radius, protonation enhances the quantum confinement.


Introduction
Two-dimensional layered transition metal dichalcogenides (TMDs) are known to have two different phases, semiconducting and metallic (1T/1T'), which induce miscellaneous changes in their electronic and optical properties [1][2][3]. Unlike semiconducting TMDs, metallic-phase (1T/1T') TMDs and their QDs show a feature of high charge transfer efficiency for photocatalytic and electrocatalytic abilities [4][5][6]. Recently, two sizes of metallic-phase TMDs-QDs (TMDs-mQDs) were demonstrated to have the photophysical properties of different charge excitation and decay pathways, which was explained by defect states and valance band splitting [3]. The variation in the optical energy structure in QDs is generally ascribed to quantum confinement and edge functionality [7][8][9]. However, the dominant mechanism that intrinsically controls exciton behaviors as the size and environmental conditions change remains elusive, although this issue is important for utilizing the photophysical properties in practical applications.

Materials and Methods
Potassium sodium tartrate was selected to synthesize WX 2 -mQDs at low temperature to minimize damage [3,27,28]. To fabricate the WX 2 -mQDs, we followed the previous process for GQDs and TMD-QDs in Refs. [5,22], and the details can be found there. Briefly, the initial steps started with the mixing and grinding of potassium sodium tartrate (200 mg) with WS 2 (WSe 2 ) (20 mg) purchased from Sigma Aldrich in Seoul, Republic of Korea. Then, the ground homogeneous mixtures were reacted in an autoclave vessel at 250 • C for 12 h, and then instantly exfoliated in water with sonication. The sizes of the WX 2 -mQDs were firstly controlled by AAO (20 nm, Whatman, Merch KGaA, Republic of Korea) filtration, and then dialyzed using dialysis tubing (10,000 and 8000 NMWL, Amicon Ultra-15, Merck KGaA, Seoul, Republic of Korea) to separate them into two sizes, which simultaneously removed the remaining salts. Finally, the WX 2 -mQDs were obtained as a dispersed solution in water. After drying, the WX 2 -mQDs were re-dispersed with an appropriately low concentration (<0.1 mg/mL) in 10 mL of water with pH 2 and 11.
The morphologies of the mQDs were analyzed using an AFM (X2-70, Park Systems Corp, Suwon, Republic of Korea) in tapping mode under ambient conditions at the Smart Nature Research Centre. XPS (Sigma Probe, AlKα), UV/Vis spectra (UV-3600, Shimadzu spectrometer, Smart Nature Research Centre, Cheonan, Republic of Korea), fluorescence spectra (Perkin-Elmer LS 55 luminescence spectrometer, PerkinElmer, MA, USA), and transmission electron microscopy (TEM, Titan cubed G2 60-300, FEI, OR, USA) analyses were conducted. PL measurements were carried out using a 325 nm He-Cd continuouswave laser, a monochromatic light from a 300 W Xenon lamp, and UV spectrometers (Maya2000, Ocean Optics, Orlando, FL, USA) as a PL detector at room temperature. Figure 1 schematically shows the experimental process: the sample preparation, size separation, and measurement of the optical properties of WX 2 -mQDs. From the bulk WS 2 (WSe 2 ), the Na, K-tartrate intercalated composites were produced using a thermostatic oven and then radically dropped into the water with mild sonication to be broken into QDs with a certain size distribution. After the size separation via filtration and dialysis, the WX 2 -mQDs were re-dispersed in water with a pH of 2 and 11 for the measurement of their optical properties. More experimental details are provided in the Supporting Information. With the as-prepared WX 2 -mQDs, the structural properties of WX 2 -mQDs were examined at pH 7 and are referred to in the previous work [3]. Transmission electron microscopy (TEM) measurements were performed for the WX2-mQDs suspended in pH 7 using a lacy carbon TEM grid ( Figure 2). The insets of Figure 2a,c are the digital images of dispersed WS 2 -mQDs and WSe 2 -mQDs, respectively. Figure 2b,d are TEM images of WS 2 -mQDs and WSe 2 -mQDs, respectively, with a lattice parameter. From TEM (Figure 2 and Ref. [3]) and AFM ( Figure S1) analysis, the collected WX2-mQDs are revealed to not only have nano-sized structure, but also have two different sizes (WX 2 -mQDs-Ss with diameter of~3 nm similar to the exciton Bohr radius; WX 2 -mQDs-Ls with diameter of 12 nm, sufficiently larger than the exciton Bohr radius). The XPS spectra of the tungsten core level (W4f) ( Figure S2) show four separated peaks, where the former two peaks at 32 and 34.2 eV correspond to the characteristic peaks of oxygen-free W4f and the latter at 35.2 and 37.5 eV originate from oxidized W4f [29]. The binding energies of the oxygen-free W4f peaks are~1 eV lower than those from 2H WS 2 and WSe 2 , indicating that the as-prepared WX 2 -mQDs have metallic phases (1T/1T') even though the semiconducting phase cannot be totally excluded. The edges of the WX 2 -mQDs are modified by the hydroxide (OH − ) and/or trivial tartrate (C 4 H 4 O 6 2− ) groups during the exfoliation and fracturing process. The WX 2 -mQDs are observed to be negatively charged in the neutral pH by zeta potential ( Figure S4). During the protonation or further oxidation, there is little possibility for further edge-functionalization except for H + or OH − because only NaOH or HCl solutions were used. The WX 2 -mQDs were well dispersed in the water with pH of 2 and 11 for hours and they were used in the further experiments without any treatment. However, the aggregated WX 2 -mQDs were also easily re-dispersed by mechanical and ultrasonic agitation even after days.

Results and Discussion
(WSe2), the Na, K-tartrate intercalated composites were produced using a thermostatic oven and then radically dropped into the water with mild sonication to be broken into QDs with a certain size distribution. After the size separation via filtration and dialysis, the WX2-mQDs were re-dispersed in water with a pH of 2 and 11 for the measurement of their optical properties. More experimental details are provided in the Supporting Information. With the as-prepared WX2-mQDs, the structural properties of WX2-mQDs were examined at pH 7 and are referred to in the previous work [3]. Transmission electron microscopy (TEM) measurements were performed for the WX2-mQDs suspended in pH 7 using a lacy carbon TEM grid ( Figure 2). The insets of Figure 2a,c are the digital images of dispersed WS2-mQDs and WSe2-mQDs, respectively. Figure 2b,d are TEM images of WS2-mQDs and WSe2-mQDs, respectively, with a lattice parameter. From TEM (Figure 2 and Ref. [3]) and AFM ( Figure S1) analysis, the collected WX2-mQDs are revealed to not only have nano-sized structure, but also have two different sizes (WX2-mQDs-Ss with diameter of ~3 nm similar to the exciton Bohr radius; WX2-mQDs-Ls with diameter of ~12 nm, sufficiently larger than the exciton Bohr radius). The XPS spectra of the tungsten core level (W4f) ( Figure S2) show four separated peaks, where the former two peaks at 32 and 34.2 eV correspond to the characteristic peaks of oxygen-free W4f and the latter at 35.2 and 37.5 eV originate from oxidized W4f [29]. The binding energies of the oxygen-free W4f peaks are ~1 eV lower than those from 2H WS2 and WSe2, indicating that the as-prepared WX2-mQDs have metallic phases (1T/1T') even though the semiconducting phase cannot be totally excluded. The edges of the WX2-mQDs are modified by the hydroxide (OH − ) and/or trivial tartrate (C4H4O6 2− ) groups during the exfoliation and fracturing process. The WX2-mQDs are observed to be negatively charged in the neutral pH by zeta potential ( Figure  S4). During the protonation or further oxidation, there is little possibility for further edgefunctionalization except for H + or OH − because only NaOH or HCl solutions were used. The WX2-mQDs were well dispersed in the water with pH of 2 and 11 for hours and they were used in the further experiments without any treatment. However, the aggregated WX2-mQDs were also easily re-dispersed by mechanical and ultrasonic agitation even after days.   Figure 3 shows the UV-Vis absorbance spectra of WS2-mQDs-S (a), WS2-mQD WSe2-mQDs-S (c), and WSe2-mQDs-L (d), respectively. The as-prepared TMD-QD show monotonically decreasing UV-Vis absorbance spectra, indicating the metall erties of the QDs [3]. In the protonated WX2-mQDs-Ss (black in Figure 3a,c), a new tion peak clearly appeared at ~4.2 eV (H-transition). This is in contrast to the qu tinuously increasing absorbance spectra of the oxidized WX2-mQDs-Ss (     Figure 3 shows the UV-Vis absorbance spectra of WS2-mQDs-S (a), WS2-mQDs-L (b), WSe2-mQDs-S (c), and WSe2-mQDs-L (d), respectively. The as-prepared TMD-QDs clearly show monotonically decreasing UV-Vis absorbance spectra, indicating the metallic properties of the QDs [3]. In the protonated WX2-mQDs-Ss (black in Figure 3a,c), a new absorption peak clearly appeared at ~4.2 eV (H-transition). This is in contrast to the quasi-continuously increasing absorbance spectra of the oxidized WX2-mQDs-Ss (Figure 3a,c) and WX2-mQDs-Ls (Figure 3b,d). The H-transition band is much higher than the optical transition bands (~1.9 and ~2.3 eV) between the conduction band and the valance band as shown in the insets of Figure  3a,c [30,31]. Interestingly, the optical transition peak at ~2.1 eV in the protonated WX2-mQDs-Ss is not observed in the WX2-mQDs-Ls (inset Figure 3b,d). It should also be noted The H-transition band is much higher than the optical transition bands (~1.9 and 2.3 eV) between the conduction band and the valance band as shown in the insets of Figure 3a,c [30,31]. Interestingly, the optical transition peak at~2.1 eV in the protonated WX2-mQDs-Ss is not observed in the WX 2 -mQDs-Ls (inset Figure 3b,d). It should also be noted that the H-transition is a feature of only protonated WX 2 -mQDs-Ss. The continuously increasing absorbance in the oxidized WX 2 -mQDs-Ss and WX 2 -mQDs-Ls was attributed to the overlap of high-energy excitonic absorptions and strong electron-phonon coupling [22]. However, we postulate in this work that the H-transition originates from the quantum confinement enhanced by the protonation of WX 2 -mQDs-Ss because the protonation can suppress the edge effect of other functional groups and it supports the formation of highly localized excitonic states [32]. On the other hand, the oxidation induces the sub-energy states around the band edges of the WX 2 -mQDs-Ss and WX 2 -mQDs-Ls, resulting in a continuously increasing absorbance [15,33]. The effective band gap (E eff ) in the QDs can be derived by Equation [11]: where E g is the band gap of bulk material, h is Plank's constant, m r is the reduced mass of the exciton, r is the radius of the quantum dot, and ε 0 (ε r ) is the permittivity of free space and the relative permittivity of the material. Based on the size distribution of the WX 2 -mQDs ( Figure S1), the bandgap distribution of WX 2 -mQDs-Ss is 4~6 eV, whereas it is much lower than 2 eV in the WX 2 -mQDs-Ls. The observed H-transition is located within the bandgap of the WX 2 -mQDs-S. The weak band around 2 eV is due to the size distribution of the WX 2 -mQDs-S. The absence of a UV-Vis absorbance peak in the WX 2 -mQDs-Ls can be explained by the oxidation-induced sub-energy states and an E eff much lower than 2 eV. Figures 4, 5 and S4 show the comparison of PL spectroscopic results between the protonation and oxidation of WS 2 -mQDs-S and WSe 2 -mQDs-S. In Figure S4a, the protonated WS 2 -mQDs-S show a harmonic PL (hPL) peak at~2.95 eV and the full width at half maximum (FWHM) of~0.78 eV, while in the oxidized WS 2 -mQDs-S, the hPL spectrum redshifts as much as~0.03 eV (1%) and broadens the FWHM by about 0.17 eV (21.8%). The WSe 2 -mQDs-S show a similar trend in the hPL spectra, with a deviation of less than 5% ( Figure S4b). The excitation wavelength (λ ex )-dependent PL (PLE) spectra were measured to examine the change in the optical energy band structure in WX 2 -mQDs-Ss and WX 2 -mQDs-Ls according to the protonation and oxidation. Figure 4 shows the PL spectra of WS 2 -mQDs-S (a) and WSe 2 -mQDs-S (b), respectively, with a variation in λ ex from 250 nm to 450 nm. In the protonated WX 2 -mQDs-Ss, the PL peak at~2.9 eV radically redshifts to~2.4 eV when the λ ex is larger than 370 nm. The intensity sharply decreases from λ ex 290 nm (~4.28 eV), which is around the H-transition band. On the other hand, the oxidized WX 2 -mQDs-Ss show a gradual shift of the PL peak from~2.9 eV at λ ex~2 50 nm tõ 2.25 eV as the λ ex increases up to 450 nm. Additionally, the spectra have a shoulder on the lower emission energy (~2.5 eV) with a broad FWHM. This indicates that although in the protonated WX 2 -mQDs-Ss most excitons efficiently occur with high excitation energy decay at~2.9 eV as the λ ex is higher than 330 nm, in the oxidized WX 2 -mQDs-Ss the correlation between the emission energy and the λ ex becomes strong with the decreasing efficiency and another energy band for exciton decay. Figure 5 shows the PL intensity of WS 2 -mQDs-S (a) and WSe 2 -mQDs-S (b), respectively, at a specific emission wavelength (λ em ) as a function of the excitation energy. In the PL intensity vs. excitation energy (E ex ) of the WX 2 -mQDs-Ss ( Figure 5), two effective E ex bands (~4.96 eV and~4.2 eV) are observed. Among them, the E ex~4 .2 eV looks to be correlated with the H-transition of the UV-Vis spectra. The other E ex~4 .96 eV being higher, by as much as~0.76 eV, than the E ex~4 .2 eV is most likely due to the balance band splitting [3,4] and/or the conduction band spin splitting [34]. This trend is similarly observed in the oxidized WX 2 -mQDs-S. However, the two E ex bands are more broadened and the boundary between the two bands becomes vague. Moreover, the relative intensity at the lower E ex band increases compared to that in the oxidized WX 2 -mQDs-Ss. This suggests that the oxidation of the WX 2 -mQDs-Ss considerably affects the exciton behaviors even though it does not overwhelm the quantum confinement. The measured PL properties are listed in Table 1.   Figure 5 shows the PL intensity of WS2-mQDs-S (a) and WSe2-mQDs-S (b), respectively, at a specific emission wavelength (λem) as a function of the excitation energy. In the PL intensity vs. excitation energy (Eex) of the WX2-mQDs-Ss ( Figure 5), two effective Eex bands (~4.96 eV and ~4.2 eV) are observed. Among them, the Eex~4.2 eV looks to be correlated with the H-transition of the UV-Vis spectra. The other Eex~4.96 eV being higher, by   Figure 5 shows the PL intensity of WS2-mQDs-S (a) and WSe2-mQDs-S (b), respectively, at a specific emission wavelength (λem) as a function of the excitation energy. In the PL intensity vs. excitation energy (Eex) of the WX2-mQDs-Ss ( Figure 5), two effective Eex bands (~4.96 eV and ~4.2 eV) are observed. Among them, the Eex~4.2 eV looks to be correlated with the H-transition of the UV-Vis spectra. The other Eex~4.96 eV being higher, by   7 and S5 show the PL spectroscopic results from the larger quantum dots, WS 2 -mQDs-L and WSe 2 -mQDs-L. The hPL peak and FWHM of protonated WS 2 -mQDs-L are~2.83 eV and~0.89 eV, respectively, which are~0.12 eV (4%) redshifted and~0.11 eV (14%) broadened compared with those from the protonated WS 2 -mQDs-S ( Figure S5a). This can be attributed to the size effect. On the other hand, the oxidized WS 2 -mQDs-L show a more redshifted peak (~2.63 eV) and a broadened FWHM (~1.06 eV). WSe 2 -mQDs-L shows a similar trend in the hPL spectrum to the WS 2 -mQDs-L ( Figure S5b) (see details in Table 1). Consequently, it is clear that the WX 2 -mQDs-Ls show redshifts (0.12-0.14 eV) in the peak and broadness (0.09-0.11 eV) of the FWHM as the size increases from~3 nm to~12 nm. However, the variance in the hPL peak and the FWHM due to the protonation/oxidation in the WX 2 -mQDs-Ls is 0.16-0.2 eV (5.6-7.1%) and 0.17-0.21 eV (17-25%), respectively.    This indicates that the pH variation causes more change in the hPL spectrum WX2-mQDs-Ls compared to the size increment. From the PLE spectra ( Figure 6), th in the protonated WX2-mQDs-Ls appears at ~2.9 eV until λex~330 nm and then redsh This indicates that the pH variation causes more change in the hPL spectrum of the WX 2 -mQDs-Ls compared to the size increment. From the PLE spectra ( Figure 6), the peak in the protonated WX 2-mQDs-Ls appears at~2.9 eV until λ ex~3 30 nm and then redshifts as the λ ex increases, which is similar to that of the WX 2 -mQDs-Ss. This implies that the intrinsic optical band of the WX 2 -mQDs is located at~2.9 eV, demonstrating a large Stokes shift due to the strong exciton binding energy [35]. However, when the WX 2 -mQDs-Ls are oxidized, the highest emission appears at~2.5 eV and the emission peak at~2.9 eV becomes a shoulder in the spectrum. This is in an inverse shape to that shown in the WX 2 -mQDs-Ss ( Figure 4) and the protonated WX 2 -mQDs-Ls. Moreover, the peak at~2.9 eV is obviously suppressed when λ ex ≥ 370 nm. Interestingly, another dramatic change is observed in the PL intensity vs. E ex (Figure 7). In the protonated WX 2 -mQDs-Ls, the PL intensity varying with a maximum intensity at λ ex~4 .9 eV is similar to that in the WX 2 -mQDs-Ss as the E ex changes from 5.5 eV to 2.5 eV. On the other hand, when oxidized, the PL intensity becomes flat with curves and the maximum intensity appears at a lower E ex than 4 eV. This indicates that in the oxidized WX 2 -mQDs-Ls the excited exciton states are spread and the oxidation more critically affects the optical energy band structure of the WX 2 -mQDs. The origin of the PLE peak shift is generally due to the size distribution of the mQDs or defects [3,4,36]. However, our above hypothesis is that the size increment and oxidation lead to the occurrence of sub-energy states between the conduction band and valance band. The formation of sub-energy states leads to the migration of excited charges into the lower energy states and helps the recombination of excitons. This is indirectly supported by the more redshifted and broadened PL, smoothly decreasing PL intensity, and flat PLE intensity spectra. In addition, the higher quantum yield in the oxidized WX 2 -mQDs-Ss (4.9 for WSe 2 and 4.1 for WSe 2 ) than that (2.8 and 3.3) of the protonated ones directly proves our hypothesis (Table 1). Figure 8 shows the normalized PL intensity and PL center as a function of measurement time in the protonated and oxidized WX 2 -mQDs-Ss (Figure 8a,c) and WX 2 -mQDs-Ls (Figure 8b,d). The total measurement time of the time-resolved PL (TRPL) was 42 ns with λ ex~2 66 nm ( Figure S6). In the protonated WX 2 -mQDs-S, the PL intensity sharply decreases to less than 0.2 within 20 ns and the PL center shifts by as much~0.1 eV from 3.0 eV, whereas the oxidized WX 2 -mQDs-Ss show a relatively smooth decreasing intensity with a larger center shift (0.3 eV). This suggests that there is more exciton transition into the lower energy states following emissive decay in the oxidized WX2-mQDs-Ss. The protonated WX 2 -mQDs-Ls show a similar variance in the TRPL spectra to the protonated WX 2 -mQDs-Ss with a~0.1 eV lower PL center. However, in the oxidized WX 2 -mQDs-Ls the PL center starts to appear at~2.65 eV and then shifts only~0.1 eV within the measurement time. The PL intensity decreases slightly slower than that of the protonated WX 2 -mQDs-Ls. It should be noted that the average PL lifetimes of the oxidized WX 2 -mQDs-Ls are 3.53 ± 0.11 ns for WS 2 -mQDs-L and 3.42 ± 0.11 ns for WSe 2 -mQDs-L, which are longer than those (2.99 ± 0.21 ns and 3.16 ± 0.14 ns) of the protonated ones ( Figure S7). However, chalcogen dependence is not observed. This indicates that the exciton intraband transition quickly occurs within the broad exciton band above the band edge in the oxidized WX 2 -mQDs-Ls. So far, we have analyzed the UV-Vis, PL, PLE, and TRPL results measured on WX2-mQDs-Ss and WX2-mQDs-Ls according to protonation and oxidation. The solution with a pH of 2 can make the edge of WX2-mQDs protonated and even heal the chalcogen vacancy of the crystal structure. This might enhance the exciton localization by quantum confinement and suppress the occurrence of sub-energy states. Moreover, this protonation might lead to energy dissipation through the hydrogen bond interacting with the surrounding solvent molecules, e.g., H2O. On the other hand, oxidation-induced defect states appear around the top valance band and the lowest intrinsic conduction band. Figure 9 is schematically drawn for the optical energy band structures of the protonated WX2-mQDs-Ss and the oxidized WX2-mQDs-Ls. In the protonated WX2-mQDs-Ss, there are discrete energy states due to the enhanced quantum confinement that causes the localization of charges. On the other hand, the oxidized WX2-mQDs-Ls have a broad exciton band above the band edge resulting in exciton delocalization and there are also extra sub-energy states under the conducting band and over the valance band [4]. However, the simple increment in size might not be critical and rather have a negative effect on the quantum yield of WX2- So far, we have analyzed the UV-Vis, PL, PLE, and TRPL results measured on WX 2 -mQDs-Ss and WX 2 -mQDs-Ls according to protonation and oxidation. The solution with a pH of 2 can make the edge of WX 2 -mQDs protonated and even heal the chalcogen vacancy of the crystal structure. This might enhance the exciton localization by quantum confinement and suppress the occurrence of sub-energy states. Moreover, this protonation might lead to energy dissipation through the hydrogen bond interacting with the surrounding solvent molecules, e.g., H 2 O. On the other hand, oxidation-induced defect states appear around the top valance band and the lowest intrinsic conduction band. Figure 9 is schematically drawn for the optical energy band structures of the protonated WX 2 -mQDs-Ss and the oxidized WX 2 -mQDs-Ls. In the protonated WX 2 -mQDs-Ss, there are discrete energy states due to the enhanced quantum confinement that causes the localization of charges. On the other hand, the oxidized WX 2 -mQDs-Ls have a broad exciton band above the band edge resulting in exciton delocalization and there are also extra sub-energy states under the conducting band and over the valance band [4]. However, the simple increment in size might not be critical and rather have a negative effect on the quantum yield of WX 2 -mQDs. The decisive factor for quantum confinement is the exciton Bohr radius (R), which can be deduced by the following equation: where ε is the dielectric constant of the material, m 0 is the free electron mass, m r is the reduced mass of the exciton, and a 0 ≈ 0.5 Å. Based on the parameters in previous reports [17,18], the R of WX 2 -mQDs is about 3~4 nm. This suggests that when the size of WX 2 -mQDs is comparable to the R, the quantum confinement more dominantly affects the PL properties than the environmental effect, whereas when the size >>R (for example, the WX 2 -mQDs-Ls) the protonation/oxidation plays an important role in the photophysical properties. However, there must be a transition point of the dominant mechanism between the quantum confinement and the environmental effect as the size increases, and this question still remains elusive.

Conclusions
We investigated the effect of protonation and oxidation on the photophysic erties of two different size WX2-mQDs. The protonated WX2-mQDs showed r sharp harmonic PL spectra with a high emission energy peak, two obvious PLE i peaks, short lifetimes, and low quantum yield. On the other hand, the oxidiz showed relatively broad and lower harmonic PL energy peaks with longer lifeti higher QYs. In the WX2-mQDs-Ls, the variation in PL properties was dominantly by protonation/oxidation rather than the size effect. This demonstrates that in t mQDs with a larger size than the exciton Bohr radius, the photophysical prope more effectively affected by the environmental conditions, whereas in the WX with a size comparable to the exciton Bohr radius, the quantum confinement is a d mechanism that explains the optical properties. This work illuminates the transiti dominant mechanism modulating the exciton dynamics in WX2-mQDs and it is e to be an important guide for the practical application of TMDs-mQDs. However need to investigate the exact size at which the transition from quantum confineme environmental effect occurs.

Conclusions
We investigated the effect of protonation and oxidation on the photophysical properties of two different size WX 2 -mQDs. The protonated WX 2 -mQDs showed relatively sharp harmonic PL spectra with a high emission energy peak, two obvious PLE intensity peaks, short lifetimes, and low quantum yield. On the other hand, the oxidized ones showed relatively broad and lower harmonic PL energy peaks with longer lifetimes and higher QYs. In the WX 2 -mQDs-Ls, the variation in PL properties was dominantly affected by protonation/oxidation rather than the size effect. This demonstrates that in the WX 2 -mQDs with a larger size than the exciton Bohr radius, the photophysical properties are more effectively affected by the environmental conditions, whereas in the WX 2 -mQDs with a size comparable to the exciton Bohr radius, the quantum confinement is a dominant mechanism that explains the optical properties. This work illuminates the transition of the dominant mechanism modulating the exciton dynamics in WX 2 -mQDs and it is expected to be an important guide for the practical application of TMDs-mQDs. However, we still need to investigate the exact size at which the transition from quantum confinement to the environmental effect occurs.