pH-Dependent Photophysical Properties of Metallic Phase MoSe2 Quantum Dots

Fluorescence properties of quantum dots (QDs) are critically affected by their redox states, which is important for practical applications. In this study, we investigated the optical properties of MoSe2-metallic phase quantum-dots (MoSe2-mQDs) depending on the pH variation, in which the MoSe2-mQDs were dispersed in water with two sizes (Φ~3 nm and 12 nm). The larger MoSe2-mQDs exhibited a large red-shift and broadening of photoluminescence (PL) peak with a constant UV absorption spectra as varying the pH, while the smaller ones showed a small red-shift and peak broadening, but discrete absorption bands in the acidic solution. The excitation wavelength-dependent photoluminescence shows that the PL properties of smaller MoSe2-mQDs are more sensitive to the pH change compared to those of larger ones. From the time-resolved PL spectroscopy, the excitons dominantly decaying with an energy of ~3 eV in pH 2 clearly show the shift of PL peak to the lower energy (~2.6 eV) as the pH increases to 7 and 11 in the smaller MoSe2-mQDs. On the other hand, in the larger MoSe2-mQDs, the exciton decay is less sensitive to the redox states compared to those of the smaller ones. This result shows that the pH variation is more critical to the change of photophysical properties than the size effect in MoSe2-mQDs.


Introduction
Two-dimensional (2D) layered transition metal dichalcogenides (TMDs) have been extensively studied [1][2][3][4][5] and their quantum dots (mQDs) have attracted interest because of their potential applications [6][7][8][9]. In the TMDs structures, there are two different phases, semiconducting and metallic (1T/1T'), exhibiting miscellaneous changes in electronic and optical properties [5,[7][8][9][10][11][12]. In general, semiconducting TMDs-QDs show a broadening of excitonic absorption peaks and a higher quantum yield compared to flakes [7][8][9], while the metallic phase TMDs-QDs (TMDs-mQDs) have features of high charge transfer efficiency and photocatalytic and electrocatalytic capabilities [9,13]. Recently, the photophysical properties of two-sized TMDs-mQDs were demonstrated to have different charge excitation and decay pathways, which was ascribed to defect states and valence band splitting [14]. Also, the optical energy band structure of the TMDs-mQDs was reported to be predominantly affected by edge oxidation [15]. Because the defect and the edge oxidation of atomically thin TMDs-mQDs strongly depends on the environmental conditions [16][17][18][19], to study the photophysical properties as changing the pH of solution and the size relating to defects is essential for the practical applications.
Studies of the effects of redox on TMDs flakes have largely been for tuning optical and physical properties, however, the pH dependence of TMDs' PL is still suggestive and speculative based on the results measured from TMDs flakes. So far, the pH-dependent optical properties of TMDs-mQDs have been rarely studied compared to those of semiconducting TMDs-QDs [9,[20][21][22]. This is because the 1T/1T' metallic phase of TMDs is known to be less stable than the 2H semiconducting phase, in addition to having a large band gap and high surface to volume ratios. Semiconducting TMDs-QDs have been reported to exhibit different quantum yields with varying emission wavelengths depending on their size, surrounding conditions, and edge functionalization [4,9,[23][24][25][26][27]. Mishra et al. showed that semiconducting MoS 2 -QDs had a pH-dependent fluorescence switching behavior, which was mainly ascribed to the surface-absorbed functional groups [26]. For tin disulfide quantum dots, the quantum yield (QY) was higher in pH~1 (QY~5.32%) compared to that (QY~1.17%) in pH~12, and it was proposed that this effect was due to the protonation and deprotonation of edge functional groups. Several research groups including ours have developed a synthesis process for TMDs-(m)QDs and studied their optical properties (Figure 1a, see details in [5,10]) [5,14]. In the process, the TMDs were co-intercalated alkali metal-organic compounds and exfoliated to be mQDs showing an exotic exciton dynamics [14,25,28]. The PL shift in TMDs-(m)QDs is thought to be due to the size-dependent quantum confinement effect and the localized (delocalized) states theoretically related to the vacancy and oxidation defects. The as-prepared MoS 2 -mQDs were also examined for bio-imaging applications and showed biocompatibility with bright luminescence [29]. However, TMDs-mQDs have usually been studied in the neutral state, how environmental influences such as acidic or alkaline solution combined with the size effect on the TMDs-mQDs has not been thoroughly probed yet. known to be less stable than the 2H semiconducting phase, in addition to having a large band gap and high surface to volume ratios. Semiconducting TMDs-QDs have been reported to exhibit different quantum yields with varying emission wavelengths depending on their size, surrounding conditions, and edge functionalization [4,9,[23][24][25][26][27]. Mishra et al. showed that semiconducting MoS2-QDs had a pH-dependent fluorescence switching behavior, which was mainly ascribed to the surface-absorbed functional groups [26]. For tin disulfide quantum dots, the quantum yield (QY) was higher in pH~1 (QY~5.32%) compared to that (QY~1.17%) in pH~12, and it was proposed that this effect was due to the protonation and deprotonation of edge functional groups. Several research groups including ours have developed a synthesis process for TMDs-(m)QDs and studied their optical properties (Figure 1a, see details in [5,10]) [5,14]. In the process, the TMDs were co-intercalated alkali metal-organic compounds and exfoliated to be mQDs showing an exotic exciton dynamics [14,25,28]. The PL shift in TMDs-(m)QDs is thought to be due to the sizedependent quantum confinement effect and the localized (delocalized) states theoretically related to the vacancy and oxidation defects. The as-prepared MoS2-mQDs were also examined for bio-imaging applications and showed biocompatibility with bright luminescence [29]. However, TMDs-mQDs have usually been studied in the neutral state, how environmental influences such as acidic or alkaline solution combined with the size effect on the TMDs-mQDs has not been thoroughly probed yet. In the present work, we investigate the effect of redox states on the photophysical properties of two sizes (Φ~3 nm and ~12 nm) of MoSe2-mQDs. The separately collected two sizes MoSe2-mQDs were re-dispersed in water with three different pH of ~2, 7, and 11; and their photoluminescence (PL), UV-absorbance, excitation wavelength-dependent PL (PLE), and time-resolved PL (TRPL) were measured. For the PLE, the excitation wavelength (λex) changed from 250 nm to 550 nm, showing that the emission wavelengths redshifted as the λex increased. From the TRPL spectra of smaller MoSe2-mQDs depending on pH, the exciton radiatively decayed with a weak red-shift of peak position in pH 2, while it increased the red-shift as the pH increased. On the other hand, the larger ones did not clearly show the red-shift of PL peak as varying the pH. In the present work, we investigate the effect of redox states on the photophysical properties of two sizes (Φ~3 nm and~12 nm) of MoSe 2 -mQDs. The separately collected two sizes MoSe 2 -mQDs were re-dispersed in water with three different pH of~2, 7, and 11; and their photoluminescence (PL), UV-absorbance, excitation wavelength-dependent PL (PLE), and time-resolved PL (TRPL) were measured. For the PLE, the excitation wavelength (λ ex ) changed from 250 nm to 550 nm, showing that the emission wavelengths red-shifted as the λ ex increased. From the TRPL spectra of smaller MoSe 2 -mQDs depending on pH, the exciton radiatively decayed with a weak red-shift of peak position in pH 2, while it increased the red-shift as the pH increased. On the other hand, the larger ones did not clearly show the red-shift of PL peak as varying the pH.

Materials
Bulk MoSe 2 was purchased from 4 sicence (iNexus, Inc., Seoul, Korea). Potassium sodium tartrate was purchase from Sigma-Aldrich (St. Louis, MO, USA) and used without further purification. Chemicals including NaOH, HCl, ethanol, methanol, and acetone were purchase from Chemicals Duksan corp. and used without further purification.

Preparation of MoSe 2 -mQDs and Dispersion
Potassium sodium tartrate was selected in order to synthesize metallic phase MoSe 2 -mQDs at low temperature to minimal damage [5,14,29]. The detail for fabrication can be consulted in [14]. Briefly, the initial was the mixing and grinding of potassium sodium tartrate (200 mg) with MoSe 2 (20 mg). Then, the ground homogeneous mixtures reacted with the autoclave vessel at 250 • C for 12 h then instantly exfoliated in water with sonication. The sizes of the MoSe 2 -mQDs were controlled by AAO (20 nm) filtration following the filtration (10,000 and 8000 NMWL, Amicon Ultra-15) methods, and then dialyzed in dialysis tubing, which simultaneously removed the remaining salts. Finally, the MoSe 2 -mQDs were obtained as dispersed solution in water. After being dried, the MoSe 2 -mQDs were redispersed in 10 mL of water with pH 2, 7, and 11.

Characterization
The morphology of MoSe 2 -mQDs was analyzed using an atomic force microscope (AFM, SPA400, SII, Chiba, Japan) in tapping mode under ambient conditions. UV-Vis spectra (Shimadzu UV-3101PC spectrometer, EVISA, Switzerland), fluorescence spectra (Perkin-Elmer LS 55 luminescence spectrometer, Waltham, MA, USA), and transmission electron microscopy (TEM, Tiatan cubed G2 60-300, FEI, Hillsboro, OR, USA) analyses were conducted. TEM samples were prepared by drying a droplet of the MoSe 2 -mQDs suspensions on a carbon grid. The photoluminescence (PL) measurements were carried out using a 325 nm He-Cd continues-wave (CW) laser, a monochromatic light from a 300 W-xenon lamp, and UV spectrometers (Maya2000, Ocean Optics, Dunedin, FL, USA) as a PL detector at room temperature. The PL excitations were measured by monochromatic light from a 300 W Xenon lamp and a highly sensitive photomultiplier tube as a PL detector. In order to elucidate the recombination dynamics, we carried out time-resolved PL experiments. A mode-locked femto-second pulsed Ti: sapphire laser (Coherent, Chameleon Ultra II, Santa Clara, CA, USA) system was used as an excitation source, and the five wavelengths of the pulsed Ti:sapphire laser (266 nm, 300 nm, 350 nm, 400 nm, and 450 nm) were employed. A streak camera (Hamamatsu, Japan, C7700-01) was utilized to measure the decay profile of the PL spectra at room temperature.

Preparation and Structural Characterization
To study the photophysical properties of TMDs-mQDs depending on their redox states, we synthesized and separately collected two different sizes of MoSe 2 -mQDs (Φ~3 nm, MoSe 2 -mQDs-S and 12 nm, MoSe 2 -mQDs-L) and re-dispersed them in the water with pH 2, 7, and 11 ( Figure 1a). Figure 1b shows transmission electron microscope (TEM) images of MoSe 2 -mQDs with average sizes of~3 nm and~12 nm. From the AFM images (Figure 1c), the height of the quantum dots is 1~2 nm, corresponding to a thickness of less than three layers of MoSe 2 . The detailed structural properties of the MoSe 2 -mQDs were reported in [14]. On the other hand, the PL peak of MoSe 2 -mQDs-L appears at 437 nm with a FWHM of 0.84 eV, which further red-shifts to 461 nm (∆λ~24 nm) and broadens to 1.07 and 1.08 eV of FWHM (∆FWHM~0.23, 0.24 eV) in pH 7 and 11 (Figure 2b). By comparing the PL spectra of MoSe 2 -mQDs-S and MoSe 2 -mQDs-L, the pH variation was determined to have more effectively changed the PL spectrum than the size effect. Figure 2c,d shows the UV-vis absorbance spectra of MoSe 2 -mQDs-S and MoSe 2 -mQDs-L, respectively, according to the variation in pH. The absorbance spectrum of MoSe 2 -mQDs-S in pH 2 appears to have discrete absorption energy bands, however, this dimmed in pH 7 and 11. On the other hand, the MoSe 2 -mQDs-L shows a monotonically increasing absorbance at excitation energies lower than 4 eV, which is a feature of the metallic state QDs [14]. At excitation energies higher than 4 eV, the slope of absorbance similarly changed in both of MoSe 2 -mQDs. In addition, the absorbance of MoSe 2 -mQDs-L started to increase at a lower excitation energy than that of MoSe 2 -mQDs-S, regardless of pH. This analysis implies that the protonation besides the size effect enforces the quantum confinement in the MoSe 2 -mQDs, which is obviously observed in the MoSe 2 -mQDs-S with a size close to the Bohr exciton diameter [30]. The quantum yields (QY) of MoSe 2 -mQDs-S were 3.7%, 5.2%, and 5% at pH 2, 7, and 11, respectively, as measured using an absolute PLQY measurement system with an excitation wavelength of 325 nm. The PLQY could not be estimated in MoSe 2 -mQDs-L due to the low quantum efficiency. The values of peak position, FWHM, and QYs are listed on Table 1. These results imply that the oxidation of MoSe 2 -mQDs relieves the localization of the excited charges and the quantum confinement effect. On the other hand, the protonation of MoSe 2 -mQDs induces the energy band structure to be more discrete, which might be a cause of the low quantum efficiency, suggesting that the partial oxidation enhances the quantum efficiency of MoSe 2 -mQDs through forming sub-bands, which assists the intraband transition of charges. measured with an excitation wavelength (λex) of 325 nm. For the MoSe2-mQDs-S (Fig  2a), the PL peak positioned at 418 nm with a full width at half maximum (FWHM) of eV in pH 2, which is slightly red-shifted (Δλ~7 nm) and enlarged (ΔFWHM~0.07-8 eV pH 7 and 11. On the other hand, the PL peak of MoSe2-mQDs-L appears at 437 nm w FWHM of 0.84 eV, which further red-shifts to 461 nm (Δλ~24 nm) and broadens to and 1.08 eV of FWHM (ΔFWHM~0.23, 0.24 eV) in pH 7 and 11 ( Figure 2b). By compa the PL spectra of MoSe2-mQDs-S and MoSe2-mQDs-L, the pH variation was determi to have more effectively changed the PL spectrum than the size effect. Figure 2c,d sh the UV-vis absorbance spectra of MoSe2-mQDs-S and MoSe2-mQDs-L, respectively cording to the variation in pH. The absorbance spectrum of MoSe2-mQDs-S in pH 2 pears to have discrete absorption energy bands, however, this dimmed in pH 7 and On the other hand, the MoSe2-mQDs-L shows a monotonically increasing absorbanc excitation energies lower than 4 eV, which is a feature of the metallic state QDs [14] excitation energies higher than 4 eV, the slope of absorbance similarly changed in bot MoSe2-mQDs. In addition, the absorbance of MoSe2-mQDs-L started to increase at a lo excitation energy than that of MoSe2-mQDs-S, regardless of pH. This analysis implies the protonation besides the size effect enforces the quantum confinement in the Mo mQDs, which is obviously observed in the MoSe2-mQDs-S with a size close to the B exciton diameter [30]. The quantum yields (QY) of MoSe2-mQDs-S were 3.7%, 5.2%, 5% at pH 2, 7, and 11, respectively, as measured using an absolute PLQY measurem system with an excitation wavelength of 325 nm. The PLQY could not be estimate MoSe2-mQDs-L due to the low quantum efficiency. The values of peak position, FWH and QYs are listed on Table 1. These results imply that the oxidation of MoSe2-mQD lieves the localization of the excited charges and the quantum confinement effect. On other hand, the protonation of MoSe2-mQDs induces the energy band structure to be m discrete, which might be a cause of the low quantum efficiency, suggesting that the pa oxidation enhances the quantum efficiency of MoSe2-mQDs through forming sub-ba which assists the intraband transition of charges.

Excitation Wavelength Dependent PL (PLE)
To examine the variation in optical energy band structure depending on pH, we measured the excitation wavelength (λ ex )-dependent PL (PLE) from MoSe 2 -mQDs in pH 2, 7, and 11. Figure 3 is the PLE spectra of MoSe 2 -mQDs-S (a, b, and c) and MoSe 2 -mQDs-L (d, e, and f) as a function of λ ex . Similar to WS 2 , MoS 2 , and pristine MoSe 2 -GQs [14,25,28], the MoSe 2 -mQDs showed a red-shift in the PLE peak as the λ ex increased for all pH values. The PLE spectra of MoSe 2 -mQDs-S in pH 2 ( Figure 3a) exhibited a peak shift from~410 nm to~490 nm (∆λ~80 nm) as the λ ex varied from 250 nm to 450 nm. When the pH increased to 7 and 11, the peak shift went up to 535 nm (∆λ~125 nm) at λ ex~4 50 nm, which is a 45 nm further red-shift compared to that in pH 2. This phenomena has been similarly observed in WS 2 -QDs due the size effect, and in MoS 2 -QDs due to the pH effect [25,26]. However, it is important to note that the PLE peak intensity of MoSe 2 -mQDs-S sharply decreased in the range of λ ex larger than~310 nm in pH 2, while it slowly decreased with increasing λ ex in pH 7 and 11. This indicates that in MoSe 2 -mQDs-S the pH variation does critically affect the decay pathway of excitons, most likely due to the formation of sub-bands related to the redox state of MoSe 2 -mQDs. The oxidation can induce an extension of band-edge, resulting in some excited charges transiting to the lower energy state. This is in accordance with the analysis of results shown in Figure 2 and Table 1. A peak shift in the PLE spectra was similarly observed in MoSe 2 -mQDs-L (Figure 3d-f). In pH 2, the PLE peak positions of MoSe 2 -mQDs-L depending on the λ ex were similar to those of MoSe 2 -mQDs-S at pH 2, supporting the idea that the protonation of MoSe 2 -mQDs enforces the exciton decay at the intrinsic energy states. However, in pH 7 and 11, the PLE peaks at λ ex~2 50-310 nm appeared at~460 nm, which are red-shifted as much as~50 nm compared to those in pH 2. The PLE peak position at the λ ex~4 50 nm was~550 nm (∆λ~90 nm) compared to that at λ ex~2 50 nm, which is a smaller peak shift compared to that (∆λ~125 nm) shown in MoSe 2 -mQDs-S. This means that the effect of pH variation in MoSe 2 -mQDs-L is weaker than that in MoSe 2 -mQDs-S, although it is stronger than the size effect. Figure 4 is the time resolved PL (TRPL) spectra depending on the pH of MoSe 2 -mQDs-S (Figure 4a-c) and MoSe 2 -mQDs-L (Figure 4d-f) measured at λ ex~2 66 and 310 nm, respectively, from 0 ns to 42 ns with six steps of time interval. The cumulative spectra of MoSe 2 -mQDs-S for the PL measurement time interval showed a similar trend in peak shift at the three pHs as time went on (Figure 4a-c). However, the peak shift was much clearer in pH 7 and 11, from~415 nm to~465 nm, indicating that the charge delocalization and migration to the lower energy states easily occur when the MoSe 2 -mQDs are oxidized. This is in line with the analysis in Figure 3 that the decay pathway of excitons has an intrinsic bandgap in the protonated MoSe 2 -mQDs-S, while the band edge changes as the pH increases. In comparison, the MoSe 2 -mQDs-L does not show a distinct peak shift in all pHs, although the PL peak position in pH 2 is~430 nm and shifts to~465 nm in pH 7 and 11. Because the PL peaks of MoSe 2 -mQDs-L in pH 2 are closer to the values (~415 nm) of MoSe 2 -mQDs-S than the values (~465 nm) in pH 7 and 11, the size effect looks to be weaker than the pH effect. This again confirms that the protonation of MoSe 2 -mQDs enhances the exciton decay at the band states with an intrinsic bandgap.  (Figure 4a-c). However, the peak shift was much clearer in pH 7 and 11, from ~415 nm to ~465 nm, indicating that the charge delocalization and migration to the lower energy states easily occur when the MoSe2-mQDs are oxidized. This is in line with the analysis in Figure 3 that the decay pathway of excitons has an intrinsic bandgap in the protonated MoSe2-mQDs-S, while the band edge changes as the pH increases. In comparison, the MoSe2-mQDs-L does not show a distinct peak shift in all pHs, although the PL peak position in pH 2 is ~430 nm and shifts to ~465 nm in pH 7 and 11. Because the PL peaks of MoSe2-mQDs-L in pH 2 are closer to the values (~415 nm) of MoSe2-mQDs-S than the values (~465 nm) in pH 7 and 11, the size effect looks to be weaker than the pH effect. This again confirms that the protonation of MoSe2-mQDs enhances the exciton decay at the band states with an intrinsic bandgap.

Discussion
This result we analysed above implies that in low pH the protonation of MoSe2-mQDs enhances the quantum confinement via the localization of the energy band, while the high pH induces the extension of the band edge, resulting in the large PL peak shift. The high pH of the solution can be attributed to the oxidation of MoSe2-mQDs, which makes a difference in electron density on MoSe2-mQDs in pH 2 and pH 11. Moreover, the higher electro negativity can block the interaction of MoSe2-mQDs with water molecules, causing the PL quenching. The analysis of PLE and TRPL results indicates that the variation of pH is more effective on the change of PL properties in MoSe2-mQDs than the size effect. On the other hand, as the size of MoSe2-mQDs increases to be larger than the exciton Bohr diam-  (a-c). TRPL spectra measured from MoSe 2 -mQD-S at pH 2 (a), pH 7 (b), and pH 11 (c) with λ ex~2 66 nm. (d-f). TRPL spectra measured from MoSe 2 -mQD-L at pH 2 (d), pH 7 (e), and pH 11 (f) with λ ex~3 10 nm. Each PL spectrum has the intensity accumulated for 0-2 (black), 2-6 (red), 6-12 (blue), 12-20 (green), 20-30 (magenta), and 30-42 (dark yellow) ns.

Discussion
This result we analysed above implies that in low pH the protonation of MoSe 2 -mQDs enhances the quantum confinement via the localization of the energy band, while the high pH induces the extension of the band edge, resulting in the large PL peak shift. The high pH of the solution can be attributed to the oxidation of MoSe 2 -mQDs, which makes a difference in electron density on MoSe 2 -mQDs in pH 2 and pH 11. Moreover, the higher electro negativity can block the interaction of MoSe 2 -mQDs with water molecules, causing the PL quenching. The analysis of PLE and TRPL results indicates that the variation of pH is more effective on the change of PL properties in MoSe 2 -mQDs than the size effect. On the other hand, as the size of MoSe 2 -mQDs increases to be larger than the exciton Bohr diameter, the extrinsic band structures related to the structural and oxidation defects become the dominant decay pathway of excitons.

Conclusions
We synthesized two different sized metallic phase MoSe 2 -QDs and investigated their photophysical properties depending on pH. The MoSe 2 -mQDs-S showed narrow PL spectra in all pH compared to MoSe 2 -mQDs-L and a higher quantum yield in pH 7 and 11 than that in pH 2 and the MoSe 2 -mQDs-L. The PLE spectra showed that as the λ ex increased, the PL peaks of MoSe 2 -mQDs-S more red-shifted in pH 7 and 11 than in pH 2. However, in MoSe 2 -mQDs-L, the red-shift of the PL peak was less than those in MoSe 2 -mQDs-S. From the TRPL results, the excitons in MoSe 2 -mQD dominantly decayed with an intrinsic energy bandgap in pH 2, while in pH 7and 11 they clearly migrated into the lower energy states. The PL peaks of MoSe 2 -mQD-showed a small red-shift compared to those shown in MoSe 2 -mQD-S. Consequently, the variation of pH is more effective on the photophysical properties of MoSe 2 -mQD than the size effect. We expect these findings to expand the understanding of photophysical properties of MoSe 2 -mQD, which will be helpful for potential applications.