Solution‐Processed CsPbBr3 Quantum Dots/Organic Semiconductor Planar Heterojunctions for High‐Performance Photodetectors

Abstract Planar heterojunctions (PHJs) are fundamental building blocks for construction of semiconductor devices. However, fabricating PHJs with solution‐processable semiconductors such as organic semiconductors (OSCs) is a challenge. Herein, utilizing the orthogonal solubility and good wettability between CsPbBr3 perovskite quantum dots (PQDs) and OSCs, fabrication of solution‐processed PQD/OSC PHJs are reported. The phototransistors based on bilayer PQD/PDVT‐10 PHJs show responsivity up to 1.64 × 104 A W−1, specific detectivity of 3.17 × 1012 Jones, and photosensitivity of 5.33 × 106 when illuminated by 450 nm light. Such high photodetection performance is attributed to efficient charge dissociation and transport, as well as the photogating effect in the PHJs. Furthermore, the tri‐layer PDVT‐10/PQD/Y6 PHJs are used to construct photodiodes working in self‐powered mode, which exhibit broad range photoresponse from ultraviolet to near‐infrared, with responsivity approaching 10−1 A W−1 and detectivity over 106 Jones. These results present a convenient and scalable production processes for solution‐processed PHJs and show their great potential for optoelectronic applications.

Transfer characteristics for CsPbBr 3 QDs transistors under dark when the device was measured in the p-type regime. (c) Transfer characteristics for CsPbBr 3 QDs transistors under dark when the device was measured in the n-type regime.
We repeat the measurements of CsPbBr 3 QD transistors several times and we confirm that there are no transistor behaviors. To investigate the charge transport properties of CsPbBr 3 QDs, we fabricated a sandwich-structure diode to characterize the charge transport properties of CsPbBr 3 QDs, as shown in Figure S4a. The conductivity of the materials was estimated to be about 9.62×10 -8 S m -1 from Figure S4b, which is a very low value and shows the low charge transport efficiency in this material.
Then, we further investigated the charge transport of CsPbBr 3 QDs by fabricating a lateral device with structure shown in Figure S4c. In such a structure, the channel length and width are 40 μm and 1000 μm, respectively, and the film thickness is about 120 nm. It turns out that the device shows very low current which is beyond the detection limit of our equipment (Keysight 2912A). Indeed, if we use that conductivity obtained above, we could estimate that the channel resistance is about 3.56×10 12 Ω, which should result in a current on the order of pA, consistent with the results shown in Figure S4d. We noted that only a few papers reporting the transistor behavior of CsPbBr 3 QD transistors with very low current (on the order of 10 -8 A). [1,2] But these CsPbBr 3 QDs (or they name as CsPbBr 3 NCs) have short ligand butylamine (BuAm), while our CsPbBr 3 QDs have long ligand oleic acid and oleylamine. The long ligands result in large barrier for charge transport in QDs, which possibly explain why the charge transport efficiency is low and no transistor behavior can be observed in our CsPbBr 3 QDs.
Finally, we evaluated the dielectric properties of CsPbBr 3 QDs by employing the devices shown in Figure S4e, and the dependence of capacitance on bias voltage is shown in Figure   S4f, from which the dielectric constant was calculated to be 8.4.
In summary, although CsPbBr 3 QDs can transport charges, it has very low conductivity, and thus can be deemed as a dielectric material. These facts explain why we cannot observe the transistor behavior in pristine CsPbBr 3 QD transistors, and why they allow the electrostatic gating of the organic layer on top, as shown in Figure 3a in the manuscript. We studied the influence of CsPbBr 3 QDs thickness on CsPbBr 3 QD/PDVT-10 PHJ phototransistors performance. As shown in Figure S5, we found that when the thickness of the CsPbBr 3 QDs is large enough, the performance of the device is high and weakly dependent on the thickness of CsPbBr 3 QDs; when the thickness of the CsPbBr 3 QDs film becomes thinner, the performance of the device decrease as the thickness of CsPbBr 3 QDs decreases.  Table S1. It is notable that the μ value we used here is slightly lower than the field-effect mobility extracted from the transfer curve shown in Figure 4d, because the field-effect mobility was extracted from the high charge-density region. In this case, the calculated G is about 6×10 4 . On the other hand, G can be obtained by using the formula: = (ℎ / ) , where h is the Planc constant, c is velocity of light, and λ is wavelength of incident light. In this way, G is about 2.75R, which yields a value consistent with the one obtained above at P in = 0.01 mW cm -2 . The results shown in Figure S8a indicate that both photoconductive and photo-gating effect exist in the devices, as the transfer curve is seen to shift both up and left.
In addition, information about the gain mechanism of our devices can be obtained by analyzing the dependence of photocurrent on the power density of incident light. In Figure   S8b, we show the photocurrent of the PQD/PDVT-10 phototransistor as a function of power density in the log-log scale. It is seen that the slope varies from a value close to 1 (linear) to less than 1 (sublinear) as the gate voltage increases, which is also an indication of the combined photoconductive and photo-gating effect in the devices. To understand the band-bending at the PQD/PDVT-10 interface, we measured the work function of the two semiconductors using Kelvin probe system (KP technology 020), with the results shown in Figure S9b. According to these work function values, we can draw the energy-band diagram of the PHJs, as illustrated in Figure S9d. It can be seen that there is a built-in electric field with direction pointing from PQDs to PDVT-10. This build-in electric field is favorable for separation of photo-generated excitons in PQDS and the drifting of holes to PDVT-10, which is desired for high-performance phototransistors.     reported in literatures. [3]