Boosting Photovoltaic Performance in Organic Solar Cells by Manipulating the Size of MoS2 Quantum Dots as a Hole-Transport Material

The design of photoactive materials and interface engineering between organic/inorganic layers play a critical role in achieving enhanced performance in energy-harvesting devices. Two-dimensional transitional dichalcogenides (TMDs) with excellent optical and electronic properties are promising candidates in this regard. In this study, we demonstrate the fabrication of size-controlled MoS2 quantum dots (QDs) and present fundamental studies of their optical properties and their application as a hole-transport layer (HTL) in organic solar cells (OSCs). Optical and structural analyses reveal that the as-prepared MoS2 QDs show a fluorescence mechanism with respect to the quantum confinement effect and intrinsic/extrinsic states. Moreover, when incorporated into a photovoltaic device, the MoS2 QDs exhibit a significantly enhanced performance (5/10-nanometer QDs: 8.30%/7.80% for PTB7 and 10.40%/10.17% for PTB7-Th, respectively) compared to those of the reference device (7.24% for PTB7 and 9.49% for PTB7-Th). We confirm that the MoS2 QDs clearly offer enhanced transport characteristics ascribed to higher hole-mobility and smoother root mean square (Rq) as a hole-extraction material. This approach can enable significant advances and facilitate a new avenue for realizing high-performance optoelectronic devices.


Introduction
There is enormous interest in two-dimensional van der Waals crystals possessing tunable optical bandgaps and unique electrical properties. As a result, transitional metal dichalcogenides (TMDs, MX 2 ; M = Mo and W; X = S, Se, and Te) have gained considerable attention regarding electronic and optoelectronic devices [1][2][3][4]. Their utilization as electron/hole-transport materials to replace inorganic oxides (e.g., MoO 3 , WO 3 , NiO x , ZnO x , and V 2 O 5 ) and poly (3,4-ethylenedioxythiophene): poly(styrenesulfonate) (PEDOT:PSS) has led to a significant enhancement of the power conversion efficiency (PCE) in bulk-heterojunction (BHJ) solar cells. This enhancement has been achieved because they have clear advantages, such as suitable work function values and a widely tunable band gap; they are also easy to process [5][6][7]. Gu et al. demonstrated the fabrication of a MoS 2 film via deposition of evaporated MoO 3 and assessed its performance as a hole-transport layer (HTL) in inverted organic solar cells (OSCs)-its efficiency reached~8.11% [8]. Yun et al. and Le et al. reported the work function modulation of MoS 2 nanosheets through a p-/n-doping range of 3.5-4.8 eV and application as an HTL in OSC [9][10][11]. Liu et al. studied a solution-processing method ascribed to the exfoliation of MoS 2 via the addition of a hydrophilic surfactant (hexadecyltrimethylammonium chloride), obtaining an efficiency of~7.26% [12]. Despite this significant achievement, intensive studies are still necessary to attain higher efficiency and operational stability through the manipulation of energy level alignment and wettability between layers in OSCs.
Quantum dots (QDs) in OSCs are promising candidates due to their band gap tunability, high absorption coefficient, and multiple exciton generation [13][14][15]. The bandgap can be tuned easily by adjusting the size and composition of QDs. This makes them suitable as electron/hole materials in the OSCs, because the mismatches between energy levels or interfacial defects in planar structures largely cause an undesirable recombination loss of the photogenerated charge carriers, as well as instability regarding the device performance [16][17][18]. Li et al. reported a significant improvement in PCE through CuInS 2 QDs with sequential coatings of a CdS layer, obtaining efficiencies of 4.20% and 5.38%, respectively [19]. Recently, Pan et al. reported a photovoltaic efficiency of~7% by using CuInS 2 /ZnS core-shell QDs [20]. More recently, Lajafi et al. reported the synergistic effect of TMD-based QDs in photovoltaic devices, which evidently suppress interfacial charge recombination between layers, and obtained the enhanced power conversion efficiency (PCE) of~20.11% through low-cost and low-temperature processing [21]. Xing et al. demonstrated UV-ozone (UVO)-treated MoS 2 QDs with a tunable work function as the hole-transport layer in P3HT-and PTB7-Th-based organic solar cells and achieved significantly enhanced performance [22]. Inspired by this contribution [23,24], we argue that tailoring the energy level of TMD-based QDs through size controlling and optimizing band alignment still remain a challenge for realizing high-efficiency OSCs. Our approach can pave the way for further improving the performance of energy-harvesting devices, and the current different-sized MoS 2 QDs can be immediately applicable in energy-harvesting systems suffering from similar problems.
Herein, we demonstrate a simple and effective route for improving PCE by using sizecontrolled MoS 2 QDs as the HTL in OSCs. The MoS 2 QDs were fabricated via the formation of a MoS 2 /potassium-sodium tartrate (KNaC 4 H 4 O 6 ·H 2 O) intercalation compound, the exfoliation of MoS 2 QDs with a random size distribution in the selected solvents, and size sorting into 5-and 10-nanometer MoS 2 QDs. From photoluminescence (PL), time-resolved PL (TRPL), and transient absorption spectroscopy measurements, the as-prepared MoS 2 QDs demonstrated clear characteristic peaks for the MoS 2 QDs (5 and 10 nm), ascribed to the quantum confinement effect and intrinsic/extrinsic states, respectively. Moreover, the MoS 2 QDs were evaluated as a hole extraction layer in the OSCs, and they played a key role in attaining high performance in the OSCs (5-nanometer QD for PTB7 and PTB7-Th; PCE: 8.32% and 10.57%, J SC : 19.2 and 20.1 mA cm −2 , V OC : 0.73 and 0.80 V, and FF: 59% and 65%, respectively). Overall, the results suggest that scientific strategy through TMD-based QDs (5 and 10 nm MoS 2 QDs) can provide noteworthy progress for high performance in optoelectronics.

Preparation of MoS 2 QDs
Transition metal dichalcogenide (MoS 2 , Sigma-Aldrich Chemicals Co, Miamisburg, OH, USA) and potassium sodium tartrate (Sigma-Aldrich Chemicals Co, Miamisburg, OH, USA) were chosen to fabricate MoS 2 -based quantum dots (5 and 10 nm). The MoS 2 powder (20 mg) and potassium sodium tartrate (200 mg) were mixed by grinding under N 2 atmosphere. Then, the grinded homogeneous mixtures were moved into the homemade reactor and kept at 250 • C for 24 h. The MoS 2 interaction compound was deposited into the selected solvents for fabricating the MoS 2 QDs. After the purification process of the randomly dispersed MoS 2 QDs by a 10 KD MWCO dialysis tube, the size sorting of the QDs was conducted by the centrifuge filtration method (10,000 NMWL, Amicon Ultra-15). Then, the 5-and 10-nanometer QDs dispersed in the distilled water were obtained, and the QD solutions were freeze-dried. Finally, the 5-and 10-nanometer QD powders were obtained.

Device Fabrication
A conventional device was selected to evaluate the performance of PTB7:PC 71 BM and PTB7-Th:PC 71 BM-based OSCs with 5 and 10 nm MoS 2 QDs introduced as the hole-transport layer. The device architecture was glass/ITO/PEDOT:PSS/MoS 2 QDs/photoactive materials/Al. First, ITO was coated on the glass and cleaned using deionized water, acetone, and isopropyl alcohol, then the substrates were dried in an oven. The PEDOT:PSS on the ITO was spin-casted at 4000 rpm for 1 min and baked at 150 • C for 30 min. Then, 5-and 10-nanometer MoS 2 QDs were spin-coated on the PEDOT:PSS/ITO substrates at 4000 rpm for 1 min, with varying numbers of spin-castings (1, 3, 5, and 10 times), and were baked at 150 • C for 30 min. The substrate was moved into the glovebox under N 2 environment. The p-type materials (PTB7 and PTB7-Th: 12 mg mL −1 ) and n-type material (PC 71 BM: 40 mg mL −1 ) were dissolved in a mixture solvent of chlorobenzene/1,8-diiodooctane (97/3 vol%). After the mixing of p-type/n-type materials (1/1.7 wt%), the blends were spin-coated at 900 rpm for 2 min onto substrates deposited on different-sized MoS 2 QDs. Finally, Al (100 nm) counter electrode was thermally evaporated under vacuum (~10 −6 Torr), which defined the device area of 13 mm 2 . Device measurements were conducted in the glovebox by a Xenon arc lamp solar simulator. J−V characteristics were measured under AM 1.5 G illumination (100 mW cm −2 ) with a Keithley 2635 A source measurement unit. The vertical carrier mobility was measured by the space-charge-limited current (SCLC) method [25,26], and the device architecture was hole-only (ITO/PEDOT:PSS/holetransport layer (5 nm/10 nm MoS 2 QDs)/photoactive materials/Au). Hole mobility was calculated by the Mott-Gurney relation. EQE measurement was performed under ambient conditions using a PV Measurements QEX7 Solar Cell QE Measurement System (PV Measurements, Inc., Washington, DC, USA).

Characterizations
The optical properties of the MoS 2 QDs were measured by UV-Vis spectrophotometer (Agilent Carry 5000, Santa Clara, CA, USA), and the morphology of the QD on the PEDOT:PSS/ITO substrate was analyzed by an Agilent 5500 scanning probe microscope (SPM) running with a NanoScope V controller. Structural properties of the MoS 2 QDs were also analyzed by Raman spectroscopy (LabRAM HR UV/Vis/NIR, excitation at 514 nm). High-resolution transmission electron microscopy (HR-TEM, Tecnai G2 F30, Hillsboro, OR, USA) was measured after the droplet of the QDs suspension on TEM grid. The photoluminescence (PL), time-resolved PL, and excitation-wavelength-dependent PL behavior measurements were conducted at room temperature using a 325-nanometer He-Cd continuous-wave (CW) laser, a mode-locked femtosecond-pulsed Ti:sapphire laser (Coherent, Chameleon Ultra II, Santa Clara, CA, USA), and monochromatic light from a 300 W Xenon lamp, respectively. Work function of the MoS 2 QDs was measured by ultraviolet photoelectron spectroscopy (He I (21.2 eV) discharge lamp, ESCALAB 250Xi, Thermo Fisher Scientific, Kyoto, Japan) under ultrahigh vacuum (<10 −10 Torr). Figure 1a shows the overall synthetic procedure for fabricating the size-controlled MoS 2 QDs. The MoS 2 powder (20 mg) was mixed with potassium sodium tartrate (200 mg) under N 2 atmosphere, placed into the reactor, and heated at 250 • C for 24 h. Then, the compound was cleaved into randomly dispersed MoS 2 QDs in the selected solvent (distilled water, DI, Figure S1) and the aqueous solution was thoroughly dialyzed by a 10 KD MWCO dialysis membrane in DI, until a neutral pH was reached. Finally, the size-sorting of the MoS 2 QDs was conducted via centrifugation in filters (8000 and 10,000 NMWL, Amicon Ultra-15) to 5 and 10 nm lateral-sized MoS 2 QDs at 10,000 rpm for 10 min (bottom (5 nm) and top (10 nm)) in the centrifuge tube. The experimental details are further elaborated in the experimental section. Raman spectroscopy was performed for the bulk MoS 2 , 5 and 10 nm MoS 2 QDs to determine their structural properties (shown in Figure 1b). Two characteristic peaks are observed at~382 and~406 cm −1 , corresponding to E 2g and A 1g . The E 2g and A 1g peaks of the as-prepared MoS 2 QDs exhibit a blueshift of~2 cm −1 and a redshift of~3 cm −1 , respectively, and their distances between the E 2g and A 1g peaks are approximately ∆ 20 and ∆ 21 cm −1 , respectively. These results indicate that the thicknesses of the 5-and 10-nanometer MoS 2 QDs evidently decreased to a mono/fewlayer thickness, which is consistent with the findings of previous studies [27,28]. Figure 1c shows transmission electron microscopy (TEM) images for the overall size distribution of the 5-and 10-nanometer MoS 2 QDs. From the high-resolution TEM (HR-TEM) images, we can observe that there is a clear feature with ordered lattice fringe spacing (0.27 nm) indexed as (100) facet of the MoS 2 crystal. of the MoS2 QDs was conducted via centrifugation in filters (8000 and 10,000 NMWL, Amicon Ultra-15) to 5 and 10 nm lateral-sized MoS2 QDs at 10,000 rpm for 10 min (bottom (5 nm) and top (10 nm)) in the centrifuge tube. The experimental details are further elaborated in the experimental section. Raman spectroscopy was performed for the bulk MoS2, 5 and 10 nm MoS2 QDs to determine their structural properties (shown in Figure 1b). Two characteristic peaks are observed at ~382 and ~406 cm −1 , corresponding to E2g and A1g. The E2g and A1g peaks of the as-prepared MoS2 QDs exhibit a blueshift of ~2 cm −1 and a redshift of ~3 cm −1 , respectively, and their distances between the E2g and A1g peaks are approximately ∆ 20 and ∆ 21 cm −1 , respectively. These results indicate that the thicknesses of the 5-and 10-nanometer MoS2 QDs evidently decreased to a mono/few-layer thickness, which is consistent with the findings of previous studies [27,28]. Figure 1c shows transmission electron microscopy (TEM) images for the overall size distribution of the 5-and 10-nanometer MoS2 QDs. From the high-resolution TEM (HR-TEM) images, we can observe that there is a clear feature with ordered lattice fringe spacing (0.27 nm) indexed as (100) facet of the MoS2 crystal. This result is similar to those of previous works [25,29]. Figure 1d exhibits the thickness distributions for the 5-and 10-nanometer MoS2 QDs; they exist within a thickness range of 1-3 nm (~65%). Atomic force microscopy (AFM) images and thickness profiles for the MoS2 QDs are shown in Figure S2. Size distribution for 5-and 10-nanometer MoS2 QDs is presented in Figure S3. The introduction of oxygen or additional elements for MoS2 QDs was evaluated by X-ray photoemission spectroscopy (XPS), as shown in Figure S4. Figure 2a shows digital images of the 5-and 10-nanometer MoS2 QDs dispersed in the DI water before (Top) and after (Bottom) excitation under a 365-nanometer ultraviolet (UV) lamp. Depending on the MoS2 QD size, the emitting colors of the 5-and 10-nanometer are blue and green, respectively. To further verify optical properties of the differentsized MoS2 QDs, we measured ultraviolet-visible (UV-Vis) absorption spectra, PL, and PL excitation (PLE). The bulk MoS2 shows two characteristic peaks at ~620 and ~670 nm (Figure 2b) because of A and B excitonic transitions at the k-point of the Brillouin zone, This result is similar to those of previous works [25,29]. Figure 1d exhibits the thickness distributions for the 5-and 10-nanometer MoS 2 QDs; they exist within a thickness range of 1-3 nm (~65%). Atomic force microscopy (AFM) images and thickness profiles for the MoS 2 QDs are shown in Figure S2. Size distribution for 5-and 10-nanometer MoS 2 QDs is presented in Figure S3. The introduction of oxygen or additional elements for MoS 2 QDs was evaluated by X-ray photoemission spectroscopy (XPS), as shown in Figure S4. Figure 2a shows digital images of the 5-and 10-nanometer MoS 2 QDs dispersed in the DI water before (Top) and after (Bottom) excitation under a 365-nanometer ultraviolet (UV) lamp. Depending on the MoS 2 QD size, the emitting colors of the 5-and 10-nanometer are blue and green, respectively. To further verify optical properties of the different-sized MoS 2 QDs, we measured ultraviolet-visible (UV-Vis) absorption spectra, PL, and PL excitation (PLE). The bulk MoS 2 shows two characteristic peaks at~620 and~670 nm (Figure 2b) because of A and B excitonic transitions at the k-point of the Brillouin zone, respectively [27]. The calculated optical band gap of the MoS 2 flakes is~1.78 eV. In contrast, the as-prepared MoS 2 QDs have a wide absorption peak ranging from 250-400 nm, indicating the excitonic features of the MoS 2 QDs. respectively [27]. The calculated optical band gap of the MoS2 flakes is ~1.78 eV. In contrast, the as-prepared MoS2 QDs have a wide absorption peak ranging from 250-400 nm, indicating the excitonic features of the MoS2 QDs. In addition, the optical absorption peak of the MoS2 QDs is clearly blue-shifted, depending on the size decrease and the presence of oxygen functional groups and defects formed on MoS2 basal plane and at edges. The PL and excitation-wavelength-dependent PL (PLE) were measured to further clarify the mechanism of luminescence in the 5-and 10-nanometer MoS2 QDs (Figure 2c). The PLE spectra of the MoS2 QDs exhibit two peaks at 250 nm, originating from the transition between Mo and S (intrinsic states); additionally, a broad shoulder was observed near 300 nm, related to the transition for the oxygen functional groups and defects (extrinsic states; Figure S4).

Results and Discussion
The PLE peak intensity of 5-nanometer MoS2 QDs at 250 nm is higher than that of 10nanometer MoS2 QDs, whereas the PLE peak intensity of 5-nanometer MoS2 QDs at 300 nm is lower than that of that of 10-nanometer MoS2 QDs. These results are consistent with the UV-Vis results. In the PL under excitation at 310 nm, the 5-nanometer MoS2 QDs show a strong peak at ~440 nm (blue emission), whereas the 10-nanometer MoS2 QDs have a maximum peak at ~480 nm (green emission); the PL intensity of 5-nanometer MoS2 QDs is stronger than that of 10-nanometer MoS2 QDs. This result indicates that the as-prepared 5-nanometer MoS2 QDs have a high crystal quality within a size of 1~6 nm and that their luminescence mainly originates from the transition of Mo and S atoms, whereas the luminescence of 10-nanometer MoS2 QDs show a wide absorption peak ranging from 250~600 nm, depending on the diverse sizes from 9~14 nm and the presence of oxygen functional groups and defects formed on MoS2 basal plane as well as at edges.
To elucidate the origin of the PL and the carrier dynamics of the MoS2 QDs, measurements of the PL after adjusting the excitation wavelength (λex) and the TRPL were conducted using a Ti:sapphire laser and a streak camera detector at room temperature, as shown in Figure 3. With increasing the excitation wavelength, the 5-nanometer MoS2 QDs demonstrated maximum PL emissions at 250 nm, and the luminescence emission of the 5nanometer MoS2 QDs notably redshifts from 400 to 520 nm with a steep decay of the PL (Figure 3a). However, the 10-nanometer MoS2 QDs appear to have a similar PL emission strength in the 250-390 nm excitation range, and their PL redshifts from 460 to 550 nm with decreasing PL intensity (Figure 3b). Depending on the excitation wavelength, our fundamental understanding is not currently sufficient to clearly identify the mechanism for the origin of the PL between the quantum size and surface states. Accordingly, we assessed the TRPL of the different-sized MoS2 QDs at a λex value of 266 nm (Figure 3c,d). As time passes, the temporal PL spectra of the 5-nanometer MoS2 QDs reveal a clear redshift of the PL peak position from 400 to 460 nm (Figure 3c), whereas the 10-nanometer MoS2 QDs display a slight shift from 460 to 480 nm (Figure 3d). This indicates that the In addition, the optical absorption peak of the MoS 2 QDs is clearly blue-shifted, depending on the size decrease and the presence of oxygen functional groups and defects formed on MoS 2 basal plane and at edges. The PL and excitation-wavelength-dependent PL (PLE) were measured to further clarify the mechanism of luminescence in the 5-and 10-nanometer MoS 2 QDs (Figure 2c). The PLE spectra of the MoS 2 QDs exhibit two peaks at 250 nm, originating from the transition between Mo and S (intrinsic states); additionally, a broad shoulder was observed near 300 nm, related to the transition for the oxygen functional groups and defects (extrinsic states; Figure S4).
The PLE peak intensity of 5-nanometer MoS 2 QDs at 250 nm is higher than that of 10-nanometer MoS 2 QDs, whereas the PLE peak intensity of 5-nanometer MoS 2 QDs at 300 nm is lower than that of that of 10-nanometer MoS 2 QDs. These results are consistent with the UV-Vis results. In the PL under excitation at 310 nm, the 5-nanometer MoS 2 QDs show a strong peak at~440 nm (blue emission), whereas the 10-nanometer MoS 2 QDs have a maximum peak at~480 nm (green emission); the PL intensity of 5-nanometer MoS 2 QDs is stronger than that of 10-nanometer MoS 2 QDs. This result indicates that the as-prepared 5-nanometer MoS 2 QDs have a high crystal quality within a size of 1~6 nm and that their luminescence mainly originates from the transition of Mo and S atoms, whereas the luminescence of 10-nanometer MoS 2 QDs show a wide absorption peak ranging from 250~600 nm, depending on the diverse sizes from 9~14 nm and the presence of oxygen functional groups and defects formed on MoS 2 basal plane as well as at edges.
To elucidate the origin of the PL and the carrier dynamics of the MoS 2 QDs, measurements of the PL after adjusting the excitation wavelength (λ ex ) and the TRPL were conducted using a Ti:sapphire laser and a streak camera detector at room temperature, as shown in Figure 3. With increasing the excitation wavelength, the 5-nanometer MoS 2 QDs demonstrated maximum PL emissions at 250 nm, and the luminescence emission of the 5-nanometer MoS 2 QDs notably redshifts from 400 to 520 nm with a steep decay of the PL (Figure 3a). However, the 10-nanometer MoS 2 QDs appear to have a similar PL emission strength in the 250-390 nm excitation range, and their PL redshifts from 460 to 550 nm with decreasing PL intensity (Figure 3b). Depending on the excitation wavelength, our fundamental understanding is not currently sufficient to clearly identify the mechanism for the origin of the PL between the quantum size and surface states. Accordingly, we assessed the TRPL of the different-sized MoS 2 QDs at a λ ex value of 266 nm (Figure 3c,d). As time passes, the temporal PL spectra of the 5-nanometer MoS 2 QDs reveal a clear redshift of the PL peak position from 400 to 460 nm (Figure 3c), whereas the 10-nanometer MoS 2 QDs display a slight shift from 460 to 480 nm (Figure 3d). This indicates that the optically excited carriers in the MoS 2 QDs at λ ex value of 266 nm transfer from intrinsic to extrinsic states.   [26,30,31]. Detailed synthesis and device-testing conditions are provided in the methods. Figure 4b shows a band energy diagram of the materials used in this study; the energy level alignment of the MoS2 QDs is suitable [21]. UV photoelectron spectroscopy (UPS) measurements were conducted to determine the Fermi level energy (EF), work function (WF), and valance band (VB). The secondary electron cut-off energies of the He I (21.22 eV) UPS spectra are shown in Figure 4c. The WF of the 5-and 10-nanometer MoS2 QDs are observed at energy positions of ~16.6 and ~17.6 eV, corresponding to WF values of ~4.6 and 3.6 eV, respectively. This result indicates n-type doping of intrinsic MoS2 QDs ascribed to S-vacancies, impurities, and structural defects. The maximum energies of the VB  [26,30,31]. Detailed synthesis and device-testing conditions are provided in the methods. Figure 4b shows a band energy diagram of the materials used in this study; the energy level alignment of the MoS 2 QDs is suitable [21]. UV photoelectron spectroscopy (UPS) measurements were conducted to determine the Fermi level energy (E F ), work function (W F ), and valance band (VB). The secondary electron cut-off energies of the He I (21.22 eV) UPS spectra are shown in Figure 4c. The W F of the 5-and 10-nanometer MoS 2 QDs are observed at energy positions of~16.6 and~17.6 eV, corresponding to W F values of~4.6 and 3.6 eV, respectively. This result indicates n-type doping of intrinsic MoS 2 QDs ascribed to S-vacancies, impurities, and structural defects. The maximum energies of the VB for the 5-and 10-nanometer QDs were estimated at~5.25 eV, as shown in the inset of Figure 4c. These results are consistent with previous works [21,32,33]. The MoS 2 QDs reveal a maximum energy of the VB that is higher than that of the HOMO of the PEDOT:PSS (−5.2 eV). Consequently, MoS 2 QDs effectively act as an HTL in the PSC structures.
for the 5-and 10-nanometer QDs were estimated at ~5.25 eV, as shown in the inset of Figure 4c. These results are consistent with previous works [21,32,33]. The MoS2 QDs reveal a maximum energy of the VB that is higher than that of the HOMO of the PEDOT:PSS (−5.2 eV). Consequently, MoS2 QDs effectively act as an HTL in the PSC structures.   Table 1. The optimized control devices based on PTB7 and PTB7-Th active materials show a PCE of 7.24%, a shortcircuit current density (JSC) of 17.8 ± 0.3 mA cm −2 , an open-circuit voltage (VOC) of 0.71 ± 0.01 V, and a fill factor (FF) of 56 ± 1% for PTB7, and a PCE of 9.49%, a JSC of 18.9 ± 0.2 mA cm −2 , a VOC of 0.80 ± 0.02 V, and an FF of 60 ± 2% for PTB7-Th. To assess the synergistic effect of the MoS2 QDs as the HTL, the 5-and 10-nanometer-sized MoS2 QDs with mono-/few-layer thicknesses were deposited on the PEDOT:PSS/ITO substrate and the device performance was evaluated using the optimized donors (PTB7 or PTB7-Th):acceptor (PC71BM) active materials. Introducing the MoS2 QDs significantly enhances performance, compared to the control devices. The photovoltaic results of 5-nanometer MoS2 QD in PTB7-and PTB7-Th-based OSCs are as follows: PCE of 8.32%, JSC = 19.2 ± 0.2 mA cm −2 , VOC = 0.73 ± 0.01 V, FF = 59 ± 1% for PTB7 and PCE of 10.57%, JSC = 20.1 ± 0.2 mA cm −2 , VOC = 0.80 ± 0.01 V, FF = 65 ± 1% for PTB7-Th. Moreover, the hole mobilities for PTB7-based devices fabricated by deposition of 5/10-nanometer QDs were assessed and estimated by the space-charge-limited current (SCLC) method [30,34], as shown in Figure S5 and Experimental section. The devices fabricated by the deposition of 5/10-nanometer MoS2 QDs show enhanced hole mobilities (μh = 2.65 × 10 −4 cm 2 V −1 s −1 and μh = 1.32 × 10 −4 cm 2 V −1 s −1 for 5 and 10-nanometer MoS2 QDs, respectively) compared to that of the control device (μh = 4.2 × 10 −5 cm 2 V −1 s −1 for the control device). Especially, the 5-nanometer MoS2 QD leads to the highest level of hole mobility, which is consistent with the enhanced performance ascribed to values of JSC and FF in the OSCs. Figure S6 exhibits the UV-Vis absorption spectra of the PTB7:PC71BM active material/MoS2 QDs/PEDOT:PSS/ITO films containing different-sized MoS2 QDs (5 and 10 nm) and observing the increase of the overall absorption peaks with addition of MoS2 QDs.   Table 1. The optimized control devices based on PTB7 and PTB7-Th active materials show a PCE of 7.24%, a short-circuit current density (J SC ) of 17.8 ± 0.3 mA cm −2 , an open-circuit voltage (V OC ) of 0.71 ± 0.01 V, and a fill factor (FF) of 56 ± 1% for PTB7, and a PCE of 9.49%, a J SC of 18.9 ± 0.2 mA cm −2 , a V OC of 0.80 ± 0.02 V, and an FF of 60 ± 2% for PTB7-Th. To assess the synergistic effect of the MoS 2 QDs as the HTL, the 5-and 10-nanometer-sized MoS 2 QDs with mono-/few-layer thicknesses were deposited on the PEDOT:PSS/ITO substrate and the device performance was evaluated using the optimized donors (PTB7 or PTB7-Th):acceptor (PC 71 BM) active materials. Introducing the MoS 2 QDs significantly enhances performance, compared to the control devices. The photovoltaic results of 5nanometer MoS 2 QD in PTB7-and PTB7-Th-based OSCs are as follows: PCE of 8.32%, J SC = 19.2 ± 0.2 mA cm −2 , V OC = 0.73 ± 0.01 V, FF = 59 ± 1% for PTB7 and PCE of 10.57%, J SC = 20.1 ± 0.2 mA cm −2 , V OC = 0.80 ± 0.01 V, FF = 65 ± 1% for PTB7-Th. Moreover, the hole mobilities for PTB7-based devices fabricated by deposition of 5/10-nanometer QDs were assessed and estimated by the space-charge-limited current (SCLC) method [30,34], as shown in Figure S5 and Experimental section. The devices fabricated by the deposition of 5/10-nanometer MoS 2 QDs show enhanced hole mobilities (µ h = 2.65 × 10 −4 cm 2 V −1 s −1 and µ h = 1.32 × 10 −4 cm 2 V −1 s −1 for 5 and 10-nanometer MoS 2 QDs, respectively) compared to that of the control device (µ h = 4.2 × 10 −5 cm 2 V −1 s −1 for the control device). Especially, the 5-nanometer MoS 2 QD leads to the highest level of hole mobility, which is consistent with the enhanced performance ascribed to values of J SC and FF in the OSCs. Figure S6 exhibits the UV-Vis absorption spectra of the PTB7:PC 71 BM active material/ MoS 2 QDs/PEDOT:PSS/ITO films containing different-sized MoS 2 QDs (5 and 10 nm) and observing the increase of the overall absorption peaks with addition of MoS 2 QDs. External quantum efficiency (EQE) characteristics for the PTB7-Th-based devices fabricated after deposition with 5-and 10-nanometer MoS 2 QDs are provided ( Figure S7). The results clearly suggest the dependence of EQE values ascribed to deposition of the MoS 2 QD and size, which is in good agreement with the data obtained from the J-V characteristics.  Figure 5a,b shows the effect of the 5-nanometer MoS 2 with varying spin-casting/drying cycles on device performance. We further evaluated these photovoltaic devices with different spin-casting/drying cycles (1, 3, 5, and 10 times). Each deposition cycle of the 5-nanometer MoS 2 QD on the PEDOT:PSS surface investigated in this study shows improvement in device performance in different cycles (Figure 5a,b), and the corresponding photovoltaic parameters are plotted in Figures S8 and S9. In particular, the best device performances of PCE (8.32% for PTB7 and 10.57% for PTB7-Th) are obtained after three casting/drying cycles. To further understand enhanced device performance, the surface morphologies of the MoS 2 QDs deposited on the PEDOT:PSS layer were evaluated by AFM, as shown in Figure 5c. The surface of the PEDOT:PSS/ITO without the MoS 2 QD has a rough surface with a high root-mean-square (R q ) value of 2.1 nm ( Figure S10). The surface morphologies and R q values exhibit significant differences with increasing spin-casting/drying cycles. With fewer than three spin-casting/drying cycles, the surface morphology exhibits a finer and smoother surface ranging from 0.7-1.4 nm, while after more than five cycles the film morphology becomes rougher, with R q values in the range of 0.9-1.1 nm. Accordingly, MoS 2 QDs with finer, smoother surfaces are considered to be one of the major factors contributing to the improvement of the corresponding devices.
rough surface with a high root-mean-square (Rq) value of 2.1 nm ( Figure S10). The surface morphologies and Rq values exhibit significant differences with increasing spin-casting/drying cycles. With fewer than three spin-casting/drying cycles, the surface morphology exhibits a finer and smoother surface ranging from 0.7-1.4 nm, while after more than five cycles the film morphology becomes rougher, with Rq values in the range of 0.9-1.1 nm. Accordingly, MoS2 QDs with finer, smoother surfaces are considered to be one of the major factors contributing to the improvement of the corresponding devices.

Conclusions
We introduce a promising solution-processed method for producing 0D MoS 2 QDs with different sizes (5 and 10 nm) via an MoS 2 -potassium sodium tartrate compound. We demonstrated these MoS 2 QDs as a hole-transport layer in the OSCs. The optical results reveal that 5-and 10-nanometer MoS 2 QDs have clear and differing characteristics regarding absorption and PL; their origins correspond to quantum confinement effect and intrinsic/extrinsic states. Furthermore, MoS 2 QDs, when used as a hole extraction layer in OSCs, show synergetic effects in achieving high performance, because of their additional photocurrent generation and favorable hole extraction, as well as their modification of the PEDOT:PSS surface. From these results, we confirm that MoS 2 QDs, when used as an HTL, have great potential for improving the performance of OSCs, and can facilitate new avenues for the fundamental study of organic photovoltaics.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10.3 390/nano11061464/s1, Figure S1: Digital images of the MoS 2 quantum dots, Figure S2: AFM images of 5 and 10 MoS 2 quantum dots, Figure S3: Size distribution of the MoS 2 QDs, Figure S4: XPS spectra of Mo 3d and S 2p for 5 and 10 MoS 2 quantum dots, Figure S5: Hole mobility of hole-only device of PTB7:PC 71 BM with/without MoS 2 QDs by a space charge limited current (SCLC) method, Figure S6: UV-vis absorption spectra of ITO/PEDOT:PSS/hole-transport material/active material films without and with 5 and 10 MoS 2 quantum dots, Figure S7: EQE spectra of PTB7-Th:PCB 71 M based OSCs with 5 and 10 MoS 2 quantum dot as hole-transport layer, Figure S8: Device characteristics of PTB7:PC 71 BM based OSCs with 5 MoS 2 quantum dot as hole-transport layer, Figure S9: Device characteristics of PTB7-Th:PC 71 BM based OSCs with 5 MoS 2 quantum dot as hole-transport layer, Figure S10: AFM image of ITO/PEDOT:PSS film.

Author Contributions:
The manuscript was written through the contributions of all authors. K.H.P. and S.H.S. conceptualized the experiments and wrote the manuscript. All authors contributed to the data analysis and discussion of the results. All authors have read and agreed to the published version of the manuscript.