Ultrasmall SnS 2 quantum dot − based photodetectors with high responsivity and detectivity

: Quantumdots(QDs)oftenexhibituniquebehav-iors because the reduction in lateral size leads to stronger quantum confinement effects and a higher surface-to-volume ratio in comparison with larger two-dimensional nanosheets. However, the preparation of homogeneous QDs remains a longstanding challenge. This work reports the preparation of high-yield and ultrasmall tin disulfide (SnS 2 ) QDs by combining top–down and bottom–up approaches. The as-prepared SnS 2 QDs have a uniform lateral size of 3.17 ± 0.62 nm and a thicknesses 2.39 ± 0.88 nm. A series of self-powered photoelectrochemical-type photodetectors (PDs) utilizing the SnS 2 QDs as pho-toelectrodes are also constructed. Taking advantage of the tunable bandgaps and high carrier mobility of the SnS 2 , our PDs achieve a high photocurrent density of 16.38 μ A cm − 2 and a photoresponsivity of 0.86 mA W − 1 , and good long-term cycling stability. More importantly, the device can display obvious photoresponse, even at zero bias voltage (max), and greater weak-light sensitivity than previously reported SnS 2 -based PDs. Density functional theory calculation and optical absorption were employed to reveal the working mechanism of the SnS 2 QDs-based PDs. This study highlights the prospective applications of ultrasmall SnS 2 QDs and provides a new route towards future design of QDs-based optoelectronic devices.


Introduction
Recently, photodetectors (PDs) based on two-dimensional (2D) materials such as graphene [1,2], black phosphorus (BP) [3,4], and layered metal dichalcogenides (LMDs) [5,6] have been active areas of research because of their unique optoelectronic properties. Among these PDs, graphene is the most studied owing to the high carrier mobility (≈10 5 cm 2 V −1 s −1 ) [7]. However, graphene-based photoelectric devices are greatly limited by their absence of a bandgap. Different from graphene, BP shows great promise for optoelectronic applications benefits from its widely tunable energy gap (from 0.3 eV in the bulk to 2.2 eV for a monolayer) [8]. Nevertheless, BP-based devices are susceptible to degradation under deleterious environmental conditions because of its oxygenated and hydrophilic properties [9]. LMDs encompass a large family of materials (e.g. ReS 2 , MoS 2 , NbSe 2 , CrSe 2 , SnS 2 , and WS 2 ), and many of them exhibit landmark features that subjected to an indirect-direct bandgap transition from the bulk phase to a monolayer [10][11][12]. In addition, the high mobility and unique size, thickness, and flexibility of LMDs result in excellent electronic properties, such as the improved integration levels and containment of short-channel effects [13][14][15]. These merits render LMDs suitable candidates for high-next-generation optoelectronics. Among various semiconducting LMDs, SnS 2 is a promising material for electronic and optoelectronic applications with high charge carrier mobility, regulable bandgap (from 2.41 eV for a monolayer nanosheet to 2.18 eV in the bulk phase), and a high absorption coefficient of 10 4 cm −1 [16]. However, the performance of SnS 2 −based PDs in terms of responsivities and detectivities remains poor, which hinders their practical applications [17][18][19].
Researchers have been made extensive efforts to improve the performance of LMDs-based PDs by constructing heterostructures [20,21], morphology engineering [22][23][24] and chemical doping [25,26]. In particular, tailoring the lateral dimensions of LMDs to quantum dots (QDs) or 0D nanodots is an effective means of adjusting their intrinsic properties or inducing new optical/electrical characteristics [27]. Among these LMDs-based nanostructures, QDs possess many inherent advantages in the field of PDs such as strong quantum confinement effects, large surface area to volume ratios, and additional active edge sites [28,29]. Compared with widely studied MoS 2 and WS 2 [30][31][32], reports on mild preparation methods for SnS 2 QDs remain scarce and incomprehensive. So far, SnS 2 QDs have been synthesized using two classes of methods: top-down method by high-powered and long-time liquid exfoliation [33] and bottom-up route by L-cysteine-assisted hydrothermal method [34]. Nevertheless, the low yield and repeatability, complicated syntheses procedures and lack of redispersed solvents remain bottlenecks, leading to uncontrollable and irreproducible reactions from the various batches. Therefore, a strategy for mass production and reproducible preparations of high-quality of SnS 2 QDs is highly desirable.
Herein, we overcome this performance bottleneck by combining both top-down and bottom-up approaches through liquid-phase exfoliated (LPE) and solvothermal methods for the fabrication of ultrasmall and high-quality SnS 2 QDs with outstanding self-powered PD properties. At zero bias condition, the devices exhibited a high photocurrent density (P ph ) of 0.72 μA cm −2 in 0.1 M KOH.
Furthermore, the photo-response behavior of photoelectrochemical (PEC) SnS 2 QDs-based PDs is studied under different electrolyte, light wavelength, and bias potential and its possible working mechanism as revealed by Density functional theory (DFT) calculations for the first time. In addition, the SnS 2 QDs-based PDs also exhibit good long-term stability over 30 days, which was unimaginable for analogous QD materials. Our work provides a facile method to obtain high quality of SnS 2 QDs, which provides us with new perspectives for designing novel QDs-based PEC PDs.

Synthesis of 2D and 0D SnS 2 nanostructures
Briefly, SnS 2 NSs were prepared from SnCl 4 ·5H 2 O and thioacetamide (TAA) by a hydrothermal method in 20 mL of DI water. The autoclave was sealed and maintained at 160 • C for 12 h in an electric oven. After the hydrothermal reaction, the obtained SnS 2 NSs were washed with water and ethanol several times. Subsequently, SnS 2 powder was sonicated in NMP solvent at a concentration of 6 mg mL −1 for 8 h. The few-layer SnS 2 was obtained and centrifuged at a speed of 6000 rpm for 5 min. Thereafter, the obtained suspension was delivered to a three-neck flask under argon and heated in an oil bath at 140 • C for 3 h. Thus, high−quality SnS 2 QDs with uniform size were formed. Before the devices were fabricated, the SnS 2 QDs were stored in the NMP solution.

Characterization
The morphologies of the SnS 2 QDs were investigated by TEM and high-resolution TEM (JEOL-JEM-2100F) and AFM (Shimadzu SPM-9700, Dimension instrument with 512−px resolution). Powder XRD was carried out by a PhilipsX'Pert Pro Super diffractometer with Cu-Ka radiation ( = 0.154 nm) at scanning rate of 5 • min −1 .
Raman spectra were examined on a high-resolution confocal Raman spectroscopy (HR Evolution) with a 633 nm laser excitation. The S 2p and Sn 3d envelope were characterized by XPS (Thermo K-alpha+). UV-vis absorption spectra were collected on a shimadzu-3600plus spectrometer.

Device fabrication
First, ITO substrates were washed ultrasonically with acetone, ethanol, and DI water in order for 20 min total. After drying in nitrogen (N 2 ), 1 mg of SnS 2 QDs was added to 1 mL of Nafion/NMP (0.05 wt%) and sonicated for 30 min (320 W) to obtain a homogeneous slurry. Thereafter, the mixture was dropped onto the conductive side of ITO glass, which was dried in a vacuum oven at 80 • C overnight.

Photoresponse measurement
The photo-response behavior of SnS 2 QDs was evaluated with a photoelectrochemical (PEC) system on an electrochemical workstation (CHI 760E). The SnS 2 QDs-coated ITO glass, platinum plate, and a KCl saturated Ag/AgCl as the working electrode, counter electrode, and reference electrode, respectively. The SnS 2 QDs-based devices were illuminated under different simulated lights (mixed light from 300 to 800 nm; monochromatic light: 365, 380, 400, 475, and 550 nm) and applied in five electrolytes (0.1, 0.5, 1.0 M of KOH; 0.1 M Na 2 SO 4 ; and 0.1 M HCl). The linear sweep voltammetry measurements were collected from 0.0 to 1.0 V at a scanning rate of 10 mV/s. The electrochemical impedance spectroscopy was performed over a frequency range from 1 to 100 KHz with amplitude of 0.005 V in the dark. The photocurrent−time (I−T) curves were collected at bias voltages from 0 to 0.6 V with an interval of 5 s.

DFT calculations
DFT calculations were conducted by using the Vienna Ab-inito Simulation Package (VASP) [35]. The spin polarized Perdew− Burke−Ernzerhof (PBE) exchange-correlation functional and projector augmented wave method was used [36,37]. The cutoff energy of 500 eV and the Monkhorst−Pack k-point mesh of 11 × 11 × 7 for the primitive unit cell of bulk structure and 11 × 11 × 1 for slab model (1, 2, 4, 8 and 12 layers) were used to ensure that atomic positions and lattice constants were fully relaxed until total energy difference and forces were less than 10 −4 eV and 0.01 eV/Å, respectively [38]. The D3 correction with Becke-Johnson damping was included to correct van der Waals interaction in multi-layers structures, close to the experimental value of SnS 2 [39,40]. A higher convergence criterion for total energy, 10 −6 eV, was adopted to calculate electronic structures. The periodic boundary condition was applied in all three directions. A vacuum space of at least 15 Å was applied in the z−direction to avoid the interaction between two periodic images in each slab model.

Morphology and crystal characterization of SnS 2 QDs
Scheme 1 demonstrates a representative procedure for SnS 2 QDs-based PEC-type PDs. SnS 2 nanosheets (NSs) were obtained through LPE of multilayer SnS 2 by using the strongly-polar solvent N-methyl-2-pyrrolidone (NMP) for dispersion. Then, uniform-sized SnS 2 QDs were successfully fabricated by increasing solvothermal method, and this approach could effectively improve the quality of SnS 2 QDs. Subsequently, SnS 2 QDs was deposited on the indium tin oxide a (ITO) glass by drop−coating and used the SnS 2 QDs as the working electrode of a standard triple electrode for PEC-type PDs (refer to the experimental section for details). The transmission electron microscopy (TEM) image in Figure 1a indicates quite uniform SnS 2 QDs with a circular shape. It shows the distribution of SnS 2 QDs particles with a diameter at ca. 3.17 nm (Figure 1b). The typical high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) image of SnS 2 QDs is depicted in Figure 1d. The crystal lattices of SnS 2 QDs were 0.196 and 0.214 nm (Figure 1e), corresponding to the (111) and (102) faces of hexagonal SnS 2 crystals (JCPDS card No. 23-677) [41]. The results of the corresponding selected area electron diffraction (SAED) pattern ( Figure 1c) indexed well to the hexagonal SnS 2 crystal and the marked lattice planes are in good agreement with the previous report [42,43]. Energy-dispersive X-ray spectroscopy (EDS) elemental mapping (Figure 1f) further illustrates the successful synthesis of SnS 2 QDs. Besides, Figure S1 indicates that individual SnS 2 NSs were hexagonal structures with lateral dimensions about 20 nm. The morphology and thickness of the as-prepared SnS 2 QDs was further investigated by atomic force microscopy (AFM) image in Figure 1g. The average thickness of the SnS 2 QDs ranged from 0.62 nm (one layer) to 6.24 nm (10 layers), with an average thickness of 2.39 nm, indicating their predominantly four-layer nature (since the thickness of a monolayer was 0.60 nm) [44,45]. In addition, the suspension state of the as-prepared SnS 2 QDs was very stable, with no significant aggregation and decomposition for at least 1 month, probably due to the protective chemical stability of the NMP solvent for SnS 2 , similar to that of BP ( Figure S2) [46].
In the Raman spectra of 2D and 0D SnS 2 as shown in Figure 2a, the strong characteristic peaks at 314.5 and 313.5 cm −1 correspond to the A 1g vibration mode of SnS 2 NSs and SnS 2 QDs, respectively. This observed in−plane Raman mode of the atomic vibration shift at ca. 1 cm −1 closely correlates with the substantial variation of nanostructured SnS 2 , which is consistent with the literature [41,47]. Furthermore, the elemental composition and binding information of SnS 2 were intensively studied by high-resolution X-ray photoelectron spectroscopy (XPS). Figure 2b and c present the comparison of high-resolution XPS spectra of S 2p and Sn 3d of the as-prepared SnS 2 NSs and SnS 2 QDs samples. Regarding the S 2p spectra (Figure 2b), the peaks located at 161.8 and 162.9 eV are indexed to S 2p 3/2 and 2p 1/2 orbitals of SnS 2 QDs. The peak position of S 2p shows gradually positive shift with the decrease in the lateral size of SnS 2 . The peaks at 495.3 and 486.9 eV are attributed to the Sn 3d 3/2 and Sn 3d 5/2 orbitals of SnS 2 QDs. The similar increase of the binding energy observed in the Sn 3d X-ray photoelectron spectra (Figure 2c).
The morphology-dependent properties of SnS 2 have been investigated by UV−vis. Figure 2d shows the UV-vis spectra of SnS 2 NSs and QDs in NMP. It can also be clearly observed that SnS 2 NSs show absorption ranging from 250 to 800 nm (Figure 2d) and the corresponding bandgap was estimated to be ca. 2.90 eV when taking (Ahν) 0.5 as a function of the photon energy (hν; Figure 2e), which is in agreement with a previous report [48]. After the reaction, the solution changed from brown yellow to a transparent purple, and the final product was collected by centrifugation. The main peak of SnS 2 QDs was blue-shifted compared to SnS 2 NSs. These results originated from the quantum confinement of carriers, similar results were reported in MoS 2 and WS 2 QDs [49]. The crystal structure of the obtained SnS 2 NSs and SnS 2 QDs were systematically investigated by X-ray powder diffraction (XRD) patterns. As shown in Figure 2f, the peak position of SnS 2 QDs are almost the same compared to the XRD pattern of SnS 2 NSs, although the characterized (001) reflection was broadened and most of other diffraction intensities were obviously reduced, indicating the highly exfoliated nature of these QDs [50].

DFT calculations of SnS 2
Although monolayers of SnS 2 has been considered by several theoretical [16,44] and experimental groups [39,51,52], the SnS 2 QDs with few-layer have not been systematically studied so far. To gain further insight, this work investigates the evolution of the structural and electronic properties of SnS 2 based on DFT calculations, by using Heyd-Scuseria-Ernzerhof (HSE06) hybrid functional and Perdew−Burke−Ernzerhof (PBE) methods. In contrast to MoS 2 , both bulk SnS 2 and few-layer SnS 2 are indirect band gap semiconductors [53,54]. As indicated by the layer dependence of the band structure of SnS 2 (Figure 3), the conduction-band minimum (CBM) was located at the M high-symmetry point, and the valenceband maxima (VBM) was at a point along the Γ−M line.
Our calculations indicate the thickness-dependence of theoretical indirect band gap (e.g.) of SnS 2 which ranged from 1.32 eV for 12 layers to 1.588 eV for the monolayer, see details in Table S1. As displayed in Figure 3a, the decrease of e.g. with increasing number of layers, which can be credited to the effective reduction of electrostatic interactions by screening the vacuum in the lamellar structure, as well as the quantum confinement of electrons within a quasi-2D material of finite thickness [55]. Notably, using PBE methods alone to describe exchange-correlation interaction of electrons will undervalues the e.g. of the material (≈1 eV), since it ignores the screened Coulomb potential for Hartree-Fock exchange [56,57]. Hence, e.g. of one-layer, four-layer and bulk SnS 2 were also calculated by using hybrid HSE06 functionals ( Figure S3), and obtained values of 2.42, 2.30 and 2.48 eV, respectively, similar to the values of e.g. calculated from the UV-vis spectra (see Figure 2e). As illustrated in Figure S3d, the total and local density of states (DOS) of four−layer SnS 2 indicates that the VBM were dominated by S-3p states, and a large DOS of the VBM was considered as the main contribution to the high photo-conversion efficiency. In contrast, the S-3p and Sn-5s hybridized states played dominant roles in the CBM.
The e.g. has an important influence on the photoresponse behavior of SnS 2 QDs-based PDs, where a decreased e.g. can enhance the absorption of light energy and result in better photo-response performance. To further elaborate on the electrolyte concentration and different electrolytes effects, the band gaps of SnS 2 are calculated under different concentrations and electrolytes, respectively. On the basis of the aforementioned AFM results, the thickness of the prepared SnS 2 QDs was mainly four layers. Therefore, four−layers of SnS 2 QDs was selected as the thickness for further study. As displayed in Figure 3b, the Eg of SnS 2 QDs were substantially modulated with different external electrolyte. For instance, the Eg of low concentration of hydroxyl terminations (1−OH) adsorbed onto four−layer SnS 2 QDs was evaluated to be 0.55 eV, which is remarkably lower than that of pristine four−layer SnS 2 QDs (1.50 eV). Therefore, the photoresponse performance of PDs can be significantly improved by increasing the KOH concentration, which is in good accordance with the experimental results.

Optoelectronic performance of SnS 2 QDs
Based on the theoretical investigations of the layerdependent electronic structure of SnS 2 and UV−vis absorp-tion spectra, four−layer SnS 2 QDs in KOH electrolyte with a small bandgap by UV can be easily excited by UV light to generate electrons and enhance their electrochemical properties. Consequently, SnS 2 QDs were utilized as candidate material for PEC-type PDs. Figure 4a shows a schematic diagram of the SnS 2 on the ITO and employed as the working electrode of the (PEC-type) PDs. Generally, the pertinent parameters of photoresponse characteristics including photocurrent density (I ph ), photoresponsivity (R ph ), and specific detectivity (D * ) by the following equations [58]. (3) where I light and I dark denote the current with and without light, respectively; P λ is the light power density; S is the effective area of SnS 2 sample on ITO glass (2 cm 2 ) under illumination; and q is the electron charge constant with a value of 1.60 × 10 −19 C.
The linear sweep voltammetry (LSV) measurements of SnS 2 QDs were conducted in the range of 0-1.0 V under conditions of darkness and simulated light irradiation with different intensities (Figure 4b). There was no obvious redox peak under a bias voltage of 0-0.6 V, which indicates SnS 2 remain stable within bias voltage range. The photoresponse characteristics of SnS 2 QDs-based phtodetectors were systematically studied at increased light power density (P λ ) and different electrolyte solutions. Figure 4c the photocurrent density (P ph ) as a function of both the bias voltages and light power density (i.e., from level I to level VI, see details in Table S2), under simulated light (a mixed light from 300 to 800 nm), P ph increased from 0.72 to 1.5, 2.94, and 7.1 μAcm −2 at Level VI, and the applied bias potential was increased from 0 to 0.2, 0.4 and 0.6 V, respectively. The P ph values were substantially enhanced with ∼10 times, because a higher bias will promote photogenerated carrier efficiency. Notably, a stable "on-off" switching behavior of current density is clearly observed even at unbiased potentials, confirming its promising applications in the self-powered PDs. In addition, the on/off signal of uncoated ITO glass in 0.1 M KOH was negligible, which demonstrates the photoresponse signal under simulated light originated from the SnS 2 QDs.
On the basis of the UV−vis near-infrared absorption spectra of SnS 2 QDs, the absorption peak was ranged from 250 to 550 nm. Therefore, five quasi-monochromatic light wavelengths ( = 365, 380, 400, 475 and 550 nm) were used for evaluating the photoresponse performance of SnS 2 QDs-based PDs. Figure 3d and e display the P ph the relationships with P λ under the light irradiation at various wavelengths and a bias potential of 0.6 V in 0.1 M KOH. It can be observed that P ph shows a decreasing trend from 365 to 550 nm. At level VI, P ph reached 3.55 μAcm −2 (365 nm) and decreased to 0.11 μAcm −2 (550 nm).
This result is in good accordance with the UV-vis spectra, and the same phenomena can be observed in 0.5 M KOH electrolyte (Figure 4g). In contrast to the relationship of P ph − P λ , R ph exhibited an opposite trend as shown in  Table 1. The results indicating the excellent PEC photoresponse behavior of the 2D SnS 2 QDs-based PDs in the this work. Figure 5a displays the typical on/off behavior of SnS 2 QDs in 0.1 M Na 2 SO 4 at level VI and 0.6 V under the irradiation of a 400 nm laser. The photocurrent response as a function of wavelength is similar to the results observed in the KOH electrolyte. Based on the DFT results in Figure 3b, once SnS 2 QDs adsorbs with the OH and SO 4 2− groups, e.g. can be rapidly modulated from 1.50 (four−layer SnS 2 ) to 0.552 (OH group) and 0.520 eV (SO 4 2− group). The experimental results in Figure 5a demonstrated that the photo-response performance in 0.1 M Na 2 SO 4 (3.15 μA/cm 2 ) was slightly lower than that in the 0.1 M KOH (3.55 μA/cm 2 ) electrolyte under a 365 nm laser at level VI, which is consistent with the above theoretical calculations. In addition, the resistance (R) at the interface between the electrolyte and electrode, Figure S4 shows the implies that the type of electrolyte has a strong influence on the photoresponsive behavior, which can be attributed to the synergy between the atomic structure of SnS 2 and the adsorbed functional groups. In addition, the P ph in the HCl electrolyte was negligible compared to that in the 0.1 M KOH and Na 2 SO 4 electrolytes. Therefore, we mainly investigated the photoresponse behavior of SnS 2 QDs-based PDs in KOH and Na 2 SO 4 electrolytes. The response (t res ) and recovery (t rec ) time are other key parameters for PDs in practical applications. Regarding SnS 2 QDs-based PDs, the t res /t rec are measured under the irradiation of 365, 400, 475 and 550 nm at level VI This work in 0.1 M KOH and 0.6 V as demonstrated in Figure 5b. There was no much difference in t res /t rec under different test conditions, indicating a fast photoresponse behavior of SnS 2 QDs-based PDs under various wavelengths. The estimated t res /t rec under the irradiation of 475 nm are 0.11/0.13 s, which are lower than most currently reported PEC-type PDs, including G 2 O 3 films [69], BP NSs [70] and GaSe nanoflakes [71]. As shown in Figure 5c, the electrochemical impedance spectroscopy (EIS) of SnS 2 QDs electrodes in different concentrations of KOH electrolytes revealed the recombination dynamics at the PD-electrolyte interface under the darkness. The interfacial resistance between the PDs and electrolytes decreased with the increase of KOH concentrations. Smaller interfacial resistance values facilitate electron transfer and thus improve the photovoltaic performance [72]. The long-term stability of SnS 2 QDs-based PDs was evaluated both initially and after 1 months in 0.1 M KOH under mixed light. Figure 5d indicates no significant degradation for virgin SnS2 QDs samples (Figure 4d), which decreased by approximately 32.21% after 1 month and remained at an acceptable level. As the test time continuously increasesd, P ph exhibited a gradual decrease, mainly caused by the detachment of the SnS 2 QDs sample from the ITO glass substrate. The long-term stability of SnS 2 QDs can be effectively utilized for practical PEC-type devices, including new nucleic acid tests [73], biosensors [74] and bioanalysis [75].

Conclusions
In summary, ultrasmall and high-quality SnS 2 QDs with a uniform horizontal dimension of 3.17 nm and thickness of 2.39 nm (four layers) were successfully prepared by combining top-down and bottom-up approaches. The as-prepared SnS 2 QDs was employed to make a series of PEC-type photodetectors (PDs), which were systematically characterized under varying monochromatic lights, electrolytes and bias voltages. The SnS 2 QDs-based PDs exhibited a high responsivity of 0.86 mA/W, a high detectivity of 5.83 × 10 9 Jones, and good long-term stability, which are better than most previously reported PEC-type PDs. The excellent and stable periodic optoelectronic performance of SnS 2 QD-based PDs in different electrolytes was explained by the corresponding band structures using density functional theory. The facile approach of preparing the SnS 2 QDs can be employed for the preparation of other low-dimensional metal dichalcogenides QDs with high optoelectronic performances.