High Performance 0D ZnO Quantum Dot/2D (PEA)2PbI4 Nanosheet Hybrid Photodetectors Fabricated via a Facile Antisolvent Method

Two-dimensional (2D) organic−inorganic perovskites have great potential for the fabrication of next-generation photodetectors owing to their outstanding optoelectronic features, but their utilization has encountered a bottleneck in anisotropic carrier transportation induced by the unfavorable continuity of the thin films. We propose a facile approach for the fabrication of 0D ZnO quantum dot (QD)/2D (PEA)2PbI4 nanosheet hybrid photodetectors under the atmospheric conditions associated with the ZnO QD chloroform antisolvent. Profiting from the antisolvent, the uniform morphology of the perovskite thin films is obtained owing to the significantly accelerated nucleation site formation and grain growth rates, and ZnO QDs homogeneously decorate the surface of (PEA)2PbI4 nanosheets, which spontaneously passivate the defects on perovskites and enhance the carrier separation by the well-matched band structure. By varying the ZnO QD concentration, the Ion/Ioff ratio of the photodetectors radically elevates from 78.3 to 1040, and a 12-fold increase in the normalized detectivity is simultaneously observed. In addition, the agglomeration of perovskite grains is governed by the annealing temperature, and the photodetector fabricated at a relatively low temperature of 120 °C exhibits excellent stability after a 50-cycle test in the air condition without any encapsulation.


Introduction
Photodetection provides a feasible approach for interpreting the unreadable optical information by converting it into electrical signals, which has been sparking burgeoning interests for widespread applications, such as optical telecommunication, environmental monitoring, biomedical imaging, distributing sensor networks, and the internet of things [1][2][3][4]. The rapid expansion of information technology has been fueling the demand for high-performance photodetectors with a high signal-to-noise ratio, sensitivity, response speed, spectrum selectivity, and stability, which were proposed as "5S" figure-of-merits for the evaluation of photodetection capability [5]. Perovskites have emerged as promising photoactive materials with a combination of advantages over conventional semiconductors, including easy solution processability, a large absorption coefficient, a long carrier diffusion length, and high carrier mobility [6,7]. Inheriting from the excellent photoelectric capacities of their three-dimensional (3D) counterparts, 2D perovskites have been recently developed with a considerably improved atmospheric stability owing to the protection of hydrophobic organic chain spacers enclosed in [MX 6 ] 4− metal halide octahedra [8], which is believed to be a savior for long-term stable perovskite-based photodetection [9]. As of now, research on the 2D perovskite-based photodetectors is still in the early embryonic PEAI and 184.0 mg PbI 2 was dissolved in 1 mL of DMF, and the solution was treated with ultrasonication for 20 minutes at room temperature.
Device Fabrication: Prior to spin-coating, each glass substrate with a size of 2.5 × 2.5 cm 2 was ultrasonicated in anhydrous ethanol (99.0%) and acetone (99%) for 20 minutes, respectively. Subsequently, 100 µL of (PEA) 2 PbI 4 precursor solution was initially dropped on the glass substrate, and 100 µL of chloroform antisolvents with various contents of ZnO QDs: 0 mg/mL (PD1), 0.032 mg/mL (PD2), 0.048 mg/mL (PD3), 0.097 mg/mL (PD4), 0.194 mg/mL (PD5), 0.388 mg/mL (PD6), and 0.776 mg/mL (PD7). After incubation for 1 min, the photoactive thin films were spin-coated layer-by-layer at 800 rpm for 20 s and 2000 rpm for 30 s, respectively. After coating 3 layers, the samples were annealed at 120 • C for 1 min between spin-coating for each layer and were finally treated with a 10 min annealing at an identical temperature to improve crystallization. To investigate the annealing temperature effect, the samples fabricated with a 0.097 mg/mL of chloroform antisolvent were annealed on a hot plate in the atmosphere at room temperature (PD8), 80 • C (PD9), 100 • C (PD10), and 140 • C (PD11). Finally, a pair of 80 nm-thick Au electrodes with a spacing distance of 200 µM was deposited via thermal evaporation.
Device Characterization: The morphologies of ZnO QD-decorated (PEA) 2 PbI 4 nanosheets were characterized by a transmission electron microscope (TEM, JEM-1400 Plus, JEOL, Tokyo, Japan), and scanning electron microscope (SEM) images of each sample were obtained by a Regulus8100 system (Hitachi, Tokyo, Japan). The elemental analysis of the samples was carried out by an energy dispersive X-ray spectroscopy system (EDS, Regulus8100, Hitachi, Tokyo, Japan), and the surface morphological characterization was realized with atomic force microscopy (AFM, Dimension Icon, Bruker, Karlsruhe, Germany). The crystallinity of the samples was analyzed by X-ray diffractometer (XRD, D8 Advance, Bruker, Karlsruhe, Germany; D/MAX-RB, Rigaku, Tokyo, Japan). The absorption spectra were recorded on a UV/vis spectrophotometer (UV-2600, Shimadzu, Kyoto, Japan), and the steady-state photoluminescence (PL) spectra were required by a spectrofluorophotometer (RF-6000, Shimadzu, Kyoto, Japan) with an excitation light of 350 nm. The time-resolved PL (TRPL) spectra of (PEA) 2 PbI 4 thin films were recorded with a spectrofluorometer (Fluo Time 300, PicoQuant, Berlin, Germany) with a pumping laser of 375 nm and a probing laser of 526 nm. The photoelectric performances of all devices were measured with a semiconductor characterization system (4200, Keithley, Cleveland, OH, USA), and a monochrome adjustable light source (CME-OPS1000, Microenerg, Beijing, China) was engaged as an illumination source. Figure 1 shows the effect of ZnO QD concentrations on the morphological evolution of the zero-dimensional (0D) ZnO QD and 2D (PEA) 2 PbI 4 nanosheet hybrid thin films. The fabrication process of ZnO QD/(PEA) 2 PbI 4 nanosheet photodetectors is depicted in Figure 1a. Briefly, the ZnO QD/(PEA) 2 PbI 4 thin films were fabricated by blending ZnO QD chloroform solution with various concentrations after dropping 100 µL PEAI and PbI 2 mixture on glass substrate. According to previous works, highly crystallized 2D perovskite nanosheet thin films with a continuous morphology could be hardly achieved [25], and an effective approach for improving crystallinity was reported by adding antisolvents during thin film growth due to the accelerated nucleation site formation and grain growth rates [26]. Providing the distinctive polarity (µ C = 1.06 D) of chloroform over that of DMF (µ D = 3.82 D) [11], the nucleation site formation and grain growth rates were significantly boosted, and thus, the packed morphology of the pristine (PEA) 2 PbI 4 nanosheet thin films was obtained, as shown in Figure 1b. As shown in Figure S1a, the ZnO QDs are well-decorated on the surface of the (PEA) 2 PbI 4 nanosheet with an average diameter of 6.22 nm after blending into the (PEA) 2 PbI 4 nanosheet precursor solution. In addition, (PEA) 2 PbI 4 nanosheets generally crystallized in the monoclinic space group C2/m with lattice parameters a = 3.25 nm, b = 0.61 nm, and c = 0.62 nm [27], and wurtzite ZnO with relatively smaller lattice parameters a = b = 0.32 nm and c = 0.52 nm [28], allowing preferential adsorption of ZnO QDs on (PEA) 2 PbI 4 nanosheets. As a result, ZnO QDs were evenly decorated on the (PEA) 2 PbI 4 nanosheets, as clearly witnessed in Figure S1b, and bright diffraction spots of (200) and (020) planes [29] in the selected area electron diffraction (SAED) pattern evidenced the existence of (PEA) 2 PbI 4 in the mixed solution, as shown in Figure S1c. As shown in Figures 1 and S1d, the packed and continuous surface morphology was comparably observed for each ZnO QD/(PEA) 2 PbI 4 nanosheet sample, with a variation of ZnO QD concentrations between 0 and 0.776 mg/mL. Meanwhile, the neglectable effect of ZnO QDs on the surface morphology can also be witnessed with a similar rootmean-squared roughness (RRMS) for the devices blending with various concentrations of 0 mg/mL (14.4 nm) and 0.097 mg/mL (17.1 nm), as shown in Figure S1e-f. nm after blending into the (PEA)2PbI4 nanosheet precursor solution. In addition, (PEA)2PbI4 nanosheets generally crystallized in the monoclinic space group C2/m with lattice parameters a = 3.25 nm, b = 0.61 nm, and c = 0.62 nm [27], and wurtzite ZnO with relatively smaller lattice parameters a = b = 0.32 nm and c = 0.52 nm [28], allowing preferential adsorption of ZnO QDs on (PEA)2PbI4 nanosheets. As a result, ZnO QDs were evenly decorated on the (PEA)2PbI4 nanosheets, as clearly witnessed in Figure S1b, and bright diffraction spots of (200) and (020) planes [29] in the selected area electron diffraction (SAED) pattern evidenced the existence of (PEA)2PbI4 in the mixed solution, as shown in Figure S1c. As shown in Figures 1 and S1d, the packed and continuous surface morphology was comparably observed for each ZnO QD/(PEA)2PbI4 nanosheet sample, with a variation of ZnO QD concentrations between 0 and 0.776 mg/mL. Meanwhile, the neglectable effect of ZnO QDs on the surface morphology can also be witnessed with a similar root-mean-squared roughness (RRMS) for the devices blending with various concentrations of 0 mg/mL (14.4 nm) and 0.097 mg/mL (17.1 nm), as shown in Figure S1e-f. To further investigate the effect of ZnO QD, the developments of optical properties and crystal structures were systematically investigated with diverse concentrations To further investigate the effect of ZnO QD, the developments of optical properties and crystal structures were systematically investigated with diverse concentrations between 0 and 0.776 mg/mL, as shown in Figure 2. As shown in Figure 2a, the elements N, I, Pb, O, and Zn were evenly observed throughout the whole surface of the PD4 device fabricated with 0.097 mg/mL ZnO QD antisolvent solution, indicating a uniform distribution of ZnO QD in the hybrid thin films. In terms of the elementary composition of the samples, the existence of (PEA) 2 PbI 4 was verified with the occurrence of the Kα 1 peak at 0.392 keV of N, Mα 1 peak at 2.346 keV of Pb, and Lα 1 peak at 3.94 keV of I regardless of the ZnO QD concentration [30], and the Lβ 1 peak at 1.035 keV of Zn [31] can only be witnessed with the hybrid thin films blending with 0.097 mg/mL ZnO QDs, as shown in Figure 2b,c. For each sample, the characteristic peaks at 5.5 • , 10.9 • , 16.4 • , 21.9 • , 27.4 • , 33.0 • , and 38.7 • can be assigned to the periodical diffraction peaks of (001) series of reflections ((00x), x = 2, 4, 6), (010), (012), and (014) peaks [32], as shown in Figure 2d. Practically, (002) peaks strikingly appeared in comparison with other peaks, manifesting a preferential crystallization of the thin films [33]. As revealed in Figure 2e, the noticeable exciton absorption peak of (PEA) 2 PbI 4 [34] was constantly observed at 517 nm in the absorbance spectrum for each sample, suggesting an identical bandgap of~2.364 eV the hybrid thin films irrespective of the ZnO QD concentration as conformed with the Tauc plots shown in Figure S2. Correspondingly, the near band emission for each sample was similarly witnessed at 526 nm without any shifting as exhibited with room temperature photoluminescence (PL) spectra in Figure 2f, and the peak intensity gradually elevated along with the increased ZnO QD concentration as observed in Figure 2f, which can be possibly induced by intensified irradiative combination with the reduced defect state density due to the surface passivation with ZnO QDs [35]. To verify the hypothesis, time-resolved photoluminescence (TRPL) spectra of the photodetectors fabricated without (PD1) and with ZnO QD (PD4) were characterized, as shown in Figure 2g. The TRPL curves were well fitted by a multiexponential function with three decay components τ 1 (the excitation relaxation), τ 2 (the interaction between excitons and phonons), and τ 3 (the recombination of excitons with defects) [36]:

Results and Discussion
where A 1 , A 2 , and A 3 denote the proportionality factor for each decay period. The average decay lifetime (τ avg ) of the device can be given with [37]: As summarized in Table 1, the τ 1 and τ 2 were comparable for the devices, and the significant decrease in τ avg from 5.16 ns (PD1) to 0.94 ns (PD4) was primarily ascribed to the restricted recombination of excitons with defects [38], which can be beneficial for the carrier transportation within the photodetectors. As shown in the inset in Figure 2g, the continuous morphology of the thin films was observed with the PD4 device, and strong green irradiation under UV light illumination verified the high quality of the spin-coated thin films.  As summarized in Table 1, the τ1 and τ2 were comparable for the devices, and the significant decrease in τavg from 5.16 ns (PD1) to 0.94 ns (PD4) was primarily ascribed to the restricted recombination of excitons with defects [38], which can be beneficial for the carrier transportation within the photodetectors. As shown in the inset in Figure 2g, the continuous morphology of the thin films was observed with the PD4 device, and strong   Figure 3a. In order to deeply study the role of ZnO QD, the energy band diagram of the device is shown in Figure 3b. The conduction band minimum (E C ) and valence band maximum (E V ) of ZnO are 4.35 and 7.72 eV [39], and (PEA) 2 PbI 4 possesses relatively higher E C (3.48 eV) and E V (5.71 eV) [40]. Thus, enhanced carrier separation can be correspondingly expected from (PEA) 2 PbI 4 nanosheets to ZnO QDs under light illumination, overcoming the hindrance of charge transfer between isolated nanosheets within the thin films. As revealed in Figure 3c, the current in dark (I off ) of devices linearly elevated depending on the bias voltage without any barrier formation, and I off was comparable among photodetectors, suggesting that incorporation of ZnO QD would not introduce extra noise. As shown in Figure 3d, the elevated photocurrent (I on ) was observed at each voltage with the increased ZnO QD concentrations from 0 to 0.097 mg/mL, resulting in a noticeable increase in I on from 0.46 to 4.55 nA at 10 V under 500 nm illumination owing to improved carrier transportation induced by the efficient defect passivation and photoexcited electron separation. With the further increased concentration, the I on gradually deteriorated to 0.76 nA when the concentration reached 0.776 mg/mL, as shown in Figure 3e, which can because the formation of undesirable ion scattering sites as a result of the aggregation of exceeded ZnO QDs at grain boundaries of (PEA) 2 PbI 4 severely hindered the carrier transportation [41]. Correspondingly, the I on /I off ratio, an important metric to evaluate the antinoise capacity of photodetectors, radically increased by 13-folds from 78.3 to 1040 in comparison with the pristine device, as shown in Table 2. The responsivity (R s ) can be expressed as [42]: where P represents the light intensity (1.97 mW cm −2 ), and S is the active area between the two electrodes (0.016 cm −2 ). As shown in Figure 3f, the R s is developed as a function of the ZnO QD concentrations, leading to the optimized R s of 143.94 µA W −1 for the PD4 device. Given that the shot noise generally determines the total noise in photodetectors, the normalized detectivity (D * ) becomes a crucial metric for the evaluation of the devices, which can be given with [43]: where e denotes the elementary charge (1.6 × 10 −19 C). In contrast to the pristine device (PD1), the D * of the PD4 device obviously elevated from 1.31 × 10 9 to 1.54 × 10 10 Jones, as shown in Figure 3f, and the D * gradually decreased along with the further increased ZnO QD concentration due to the strictly deteriorated I on . The conversion efficiency of incident photons into electrons for photodetectors can be determined by the external quantum efficiency (EQE) [44]: where h, c, and λ represent the Planck's constant, the velocity of light in a vacuum, and the wavelength of irradiated light, respectively. Similar to R s , the EQE markedly boosted from 0.0035% to 0.0357% when the ZnO QD concentration increased from 0 to 0.097 mg/mL, and the deterioration in EQE was observed with further increased concentration as evidenced with Figure S3a. where h, c, and λ represent the Planck's constant, the velocity of light in a vacuum, and the wavelength of irradiated light, respectively. Similar to Rs, the EQE markedly boosted from 0.0035% to 0.0357% when the ZnO QD concentration increased from 0 to 0.097 mg/mL, and the deterioration in EQE was observed with further increased concentration as evidenced with Figure S3a.    The development of photoelectric performance for the photodetectors depending on the ZnO QD concentrations was further investigated at various spectrum wavelengths from Nanomaterials 2022, 12, 4217 9 of 16 300 to 600 nm and testing periods, as shown in Figure 4. As observed in Figure 4a, the effect of ZnO QD concentrations on the R s was consistently witnessed throughout the whole response spectra between 300 and 550 nm, and the R s of the device PD4 was consequently superior to the photodetectors at each wavelength. Particularly, the optimized R s was obtained at 500 nm for most devices, and the R s of the devices PD4 (374.57 µA W −1 ) at 300 nm was evidently higher than that at 500 nm (143.94 µA W −1 ), which can be ascribed to the additional photogenerated carriers from ZnO QDs under UV illumination. As a result, the similar behaviors of D * and EQE were observed depending on the variation of spectrum wavelengths, as shown in Figure S3b,c, indicating that blending with ZnO QDs can be an advantageous strategy for the improved performance for the 2D perovskite photodetectors over a broad wavelength range. As shown in Figure S3d, the I on /I off ratio of the device PD4 was always higher than that of the device PD1, and the decreased ratios below 375 nm were derived from the unfavorable excitation at much lower light intensities. The photodetector PD4 remained at excellent stability with a negligible deterioration after a continuous test of 50 cycles in the air, as evidenced in Figure 4b. The τ rise (the period for the current elevated from 10% to 90% of the I on ) and τ fall (the period for the current declined from 90% to 10% of the I on ) of each photodetector was within 150 ms, as shown in Figure 4c-i. Compared with the pristine device, the photodetectors fabricated with ZnO QDs exhibited faster τ rise and τ fall , suggesting accelerated carrier mobility within the hybrid thin films.
The evolution of the morphological and optical characteristics of ZnO QD/(PEA) 2 PbI 4 nanosheet hybrid thin films fabricated with an identical ZnO QD concentration of 0.097 mg/mL was investigated with a variation of annealing temperatures from room temperature to 140 • C, as shown in Figure 5. As shown in Figure 5a-c, the uniform and dense morphology was evenly observed at each temperature, and agglomeration gradually occurred along with the increased temperature, which can be beneficial for the crystallization of the hybrid perovskite thin films [45][46][47]. Above 120 • C, the severe truncation was observed between grains, as shown in Figure 5d, which can be induced by the partially melt crystallization of (PEA) 2 PbI 4 [48]. As shown in Figure 5e, there was a slight red shift of the exciton absorption peak as a function of annealing temperature, and thus, the bandgap gradually decreased from 2.368 eV to 2.362 eV due to the expansion of grain size [49], as shown in Figure 5f. As observed in Figure 5g, the PL peak initially descended depending on the annealing temperature induced by the intensified surface scattering along with morphological development, and peak intensity slightly ascended above 120 • C as a result of decreased roughness along with the melt crystallization [50]. In addition, the characteristic peaks were similarly observed without any shift depending on annealing temperatures, and the (002) peak increased after annealing in comparison with the thin films fabricated at room temperature, suggesting a preferential crystallization along [002], as evidenced in Figure 5h.
The temperature effect on the photoelectric performance of the ZnO QD/(PEA) 2 PbI 4 nanosheet hybrid photoreactors is shown in Figure 6. As shown in Figure 6a, the I off was slightly higher at relatively lower temperatures within the whole bias range due to the denser morphology, and superior I on was constantly obtained with the PD4 device at each bias as observed in Figure 6b, manifesting a sufficient temperature for the high crystallinity (PEA)2PbI4 thin films at 120 • C. Therefore, the Ion at 10 V was elevated by 3.6-folds from room temperature to 120 • C, and the Ion/Ioff ratio correspondingly increased by 21.4-folds as shown in Figure 6c. Depending on the annealing temperature, a comparable trend was witnessed with the Rs, D * , and EQE, as shown in Figure 6d-e. The increase was likewise observed for each metric as a function of temperature, and slight decreases occurred above 120 • C. As evidenced in Figures 6f and S4, the response time was similar for the devices fabricated at various temperatures, suggesting that the ZnO concentration played a more critical role in the response speed than the annealing temperature. As shown in Figure 7, the photodetector fabricated at 120 • C exhibited an optimized performance and a slowly declining response for each device depending on the wavelength derived from the effect of ZnO QDs. Correspondingly, the R s , D *, and EQE of the PD4 device were 374 µA W −1 , 4.01 × 10 10 Jones, and 0.1548% at 300 nm, respectively. As summarized in Table 3, the ZnO QD/(PEA) 2 PbI 4 nanosheet hybrid photodetector PD4 possessed an outstanding performance among perovskite photodetectors [51][52][53][54][55][56][57][58][59], offering an approach for the fabrication of high-performance perovskite photodetectors with excellent stability. The evolution of the morphological and optical characteristics of ZnO QD/(PEA)2PbI4 nanosheet hybrid thin films fabricated with an identical ZnO QD concentration of 0.097 mg/mL was investigated with a variation of annealing temperatures from room temperature to 140 °C, as shown in Figure 5. As shown in Figure 5a-c, the uniform and dense morphology was evenly observed at each temperature, and agglomeration gradually along with morphological development, and peak intensity slightly ascended above 120 °C as a result of decreased roughness along with the melt crystallization [50]. In addition the characteristic peaks were similarly observed without any shift depending on annealing temperatures, and the (002) peak increased after annealing in comparison with the thin films fabricated at room temperature, suggesting a preferential crystallization along [002], as evidenced in Figure 5h. The temperature effect on the photoelectric performance of the ZnO QD/(PEA)2PbI nanosheet hybrid photoreactors is shown in Figure 6. As shown in Figure 6a, the Ioff was slightly higher at relatively lower temperatures within the whole bias range due to the denser morphology, and superior Ion was constantly obtained with the PD4 device at each bias as observed in Figure 6b, manifesting a sufficient temperature for the high crystallin ity (PEA)2PbI4 thin films at 120 °C . Therefore, the Ion at 10 V was elevated by 3.6-folds from room temperature to 120 °C , and the Ion/Ioff ratio correspondingly increased by   Figure 7, the photodetector fabricated at 120 °C exhibited an optimized performance and a slowly declining response for each device depending on the wavelength derived from the effect of ZnO QDs. Correspondingly, the Rs, D *, and EQE of the PD4 device were 374 μA W −1 , 4.01 × 10 10 Jones, and 0.1548% at 300 nm, respectively. As summarized in Table 3, the ZnO QD/(PEA)2PbI4 nanosheet hybrid photodetector PD4 possessed an outstanding performance among perovskite photodetectors [51][52][53][54][55][56][57][58][59], offering an approach for the fabrication of high-performance perovskite photodetectors with excellent stability.

Conclusions
In summary, the 0D ZnO quantum dot (QD)/2D (PEA) 2 PbI 4 nanosheet hybrid photodetectors were fabricated via a facile ZnO QD chloroform antisolvent method under atmospheric conditions. Benefiting from the distinctive difference in polarity between the antisolvent and precursor solutions, the crystallization of the (PEA) 2 PbI 4 thin films was effectively improved as a result of the accelerated nucleation site formation and grain growth rates. The uniform ornamentation of ZnO QDs on (PEA) 2 PbI 4 nanosheets provided an in situ surface passivation for the surface defects as evidenced by the TRPL decay fitting, and the carrier transportation within the perovskite thin films was concurrently enhanced with the intensified charge separation induced by the well-balanced energy band levels. As a consequence, the photoelectric performance of ZnO QD/(PEA) 2 PbI 4 nanosheet hybrid photodetectors sensitively developed with a variation of ZnO QD concentrations, resulting in excellent D * of 1.54 × 10 10 Jones and I on /I off ratio of 1040 for the device fabricated with the 0.097 mg/mL ZnO QD antisolvent.
The annealing temperature-dependent behavior of photoconductive devices originated from the surface morphological evolution of the perovskite thin films, and the photodetector fabricated at 120 • C maintained a fast response speed within 100 ms and outstanding atmospheric stability after a 50-cycle test, offering a facile yet effective strategy for the ever-increasing demand for highly sensitive and long-term stable photodetection based on low-dimensional perovskites.