Hybrids composed of two-dimensional (2D) and zero-dimensional (0D) materials have demonstrated great application potentials in electronics and optoelectronics. The CdSe@Zn1−XCdXS (CSZCS) quantum dots (QDs) possess unique gradient band structure with a continuously increasing energy level along the radial direction from the center to the surface, which is favorable for light-harvesting, photocarrier transfer and promising for photovoltaic and photodetection applications. Here, the gradient alloyed CSZCS QDs, acting as a photosensitive layer, have been first integrated with 2D InSe as a carrier transport layer. The new 2D–0D hybrids exhibit a 300-fold improvement in responsivity compared with that of pristine InSe, which is much more superior to hybrids composed of core–shell CdSe@ZnS QDs/InSe. Thanks to the low dark current and large photo-gain induced by the photo-gating effect, the responsivity and detectivity of the CSZCS QDs/InSe photodetector can reach up to 30.16 A/W and 1.69 × 1012 Jones, respectively. This work provides a new kind of promising QDs with the gradient alloyed structure that can be explored into 2D–0D hybrids for further development of high-performance photodetectors.

In the last few decades, solution-proceed semiconductor quantum dots (QDs) have been used in electronic and optoelectronic devices due to their size-related bandgap tunability, facile synthesis, high absorption characteristics, and easy large-scale manufacturing.1,2 However, their low carrier mobility severely restricts their performance in optoelectronics. In order to overcome the disadvantage of QDs devices, the hybrids combining the QDs photosensitizer and two-dimensional (2D) materials have recently attracted increasing attention due to their well synergy effect and excellent charge transfer (CT) characteristics.3–5 In the hybrids, the 2D channel materials have been developed from graphene to transition metal dichalcogenides (TMCDs), such as MoS2 and WS2, aiming to combine both high carrier mobility and low dark current.2,6 Meanwhile, the QDs as photosensitive layers have been optimized in terms of their size and composition to achieve beneficial band structure, resulting in high photo-harvest efficiency and broadband response from the visible to infrared range.7 However, the band alignment of QDs and 2D materials, which is crucial for the charge transfer upon light illumination, has not been well investigated.

As typical 2D TMCDs, indium selenide (InSe) possesses a direct bandgap ranging from 1.26 to 2.11 eV by thinning the samples from bulk to single layer.8 The layered InSe exhibits high charge carrier mobility and strong light–matter interaction and thus has been selected as channel materials in this work.9 The varying colloidal QDs, such as CdSe, PbS, and HgTe, have been integrated into the 2D–0D hybrid platform and show great success in photodetector applications in the previous studies.10,11 Recently, we reported the CdSe@Zn1−XCdXS (CSZCS) gradient alloyed QDs with a funnel-shaped band structure and quasi-type-II configuration from the center to the surface. The unique band arrangement is favorable for the separation and transfer of the photoexcited electron–hole pairs, yielding the highly efficient photocatalysis properties.12 More importantly, the gradient alloyed structure of CSZCS QDs can effectively diminish strain from mismatched crystal lattices and reduce the number of interfacial defects in core/shell QDs,13 which could act as the recombination site of excited electrons and holes. The soft confinement potential of the gradient alloyed structure narrows the momentum spectrum of confined carriers and weakens the transition dipole moment with higher band states, which is considered to lower the rate of Auger recombination.14,15 Moreover, the suitable colloidal size and particle composition of gradient alloyed QDs could lead to the excellent ability for electron–hole pair separation, which is already explained by switching between type-I and quasi-type-II configurations via controlling over the electronic structure of alloying in the previous publication.16 Specifically, the continuous increase in composition thickness and particle size may induce ion diffusion, causing the generation of a gradient alloy involving all ions, resulting in an electronic structure converting from type-I to quasi-type-II with a less overlap integral of the electron and hole wave functions.16 Nevertheless, these superior properties of gradient alloyed QDs have been rarely applied in 2D–0D hybrid photodetectors.

In this work, we have fabricated the 2D–0D InSe/CSZCS QDs hybrids for the high-performance photodetectors. For comparison, CdSe@ZnS core/shell QDs with similar elements have been studied under the same condition. We found that both hybrid devices exhibit improved performance in responsivity and detectivity compared to the pristine InSe device (R = 0.101 A/W, D* = 2.05 × 1011 Jones) due to the photo-gating effect, while the improvement by the gradient alloyed QDs (R = 30.16 A/W, D* = 1.69 × 1012 Jones) is more obvious than that by core–shell QDs (R = 5.39 A/W, D* = 6.81 × 1011 Jones), owing to the superior band alignment of gradient alloying for more efficient separation of photocarriers. This work offers a new promising strategy to integrate gradient alloyed QDs owning funnel-shaped energy level into 2D–0D hybrids for photodetector applications.

The three-dimensional schematic diagram of the hybrid 2D–0D photodetector device structure is shown in Fig. 1(a). The hybrid photodetectors were prepared by spin-coating CSZCS QDs over the InSe field-effect transistor (FET). From the optical image, as shown in Fig. 1(b), the well-distribution of the hybrid system indicates a uniform coating of CSZCS QDs over the entire InSe flakes.17 The InSe transport layer was fabricated on a Si/SiO2 (300 nm) substrate via micromechanical exfoliation techniques.18 The few-layer InSe flake was first identified by optical contrast measurements and then confirmed by atomic force microscopy (AFM).19 The height profile of the red line in the inset of Fig. 1(b) indicates that the thickness of the few-layer InSe is about 9.5 nm. The optical image of InSe/CSZCS QDs with a channel width of 10 μm is illustrated in Fig. 1(b), and the source–drain electrodes are Cr/Au (10/50 nm). The detailed procedures for the hybrid device fabrication were described in the Experimental section. Figure 1(c) shows the Raman and photoluminescence (PL) spectra, showing the typical phonon modes and the direct bandgap at 1.26 eV of InSe layers. Three Raman peaks located at 115, 177, and 225 cm−1 can be assigned to the phonon modes of A11g, E12g, and A21g, which is consistent with the previous reports.20 For a clearer view, the scanning electron microscopy (SEM) images of the pristine InSe and hybrid 2D–0D device were provided in the Fig. S1 of the supplementary material.

FIG. 1.

Structure characterizations: (a) three-dimensional (3D) schematic view of the hybrid InSe/CSZCS QDs photodetector. (b) Optical image of the few-layer InSe/CSZCS QDs photodetector, where the QDs cover the entire device surface. The inset in (b) shows the optical images of the pristine InSe FET before adding QDs, and the red line indicates the measurement result of AFM. (c) Raman spectra and PL spectra (inset) of few-layer InSe. (d) HRTEM images of the OA capped CSZCS QDs. (e) Absorption spectra and PL spectra of the CSZCS QDs capped with OA. (f) PL spectra of the few-layer InSe/CSZCS QDs hybrid (black) and the bare QD film (red).

FIG. 1.

Structure characterizations: (a) three-dimensional (3D) schematic view of the hybrid InSe/CSZCS QDs photodetector. (b) Optical image of the few-layer InSe/CSZCS QDs photodetector, where the QDs cover the entire device surface. The inset in (b) shows the optical images of the pristine InSe FET before adding QDs, and the red line indicates the measurement result of AFM. (c) Raman spectra and PL spectra (inset) of few-layer InSe. (d) HRTEM images of the OA capped CSZCS QDs. (e) Absorption spectra and PL spectra of the CSZCS QDs capped with OA. (f) PL spectra of the few-layer InSe/CSZCS QDs hybrid (black) and the bare QD film (red).

Close modal

The CdSe@Zn1−XCdXS QDs were successfully synthesized in our previous report.12 Based on the energy dispersive spectra (EDS) in Fig. S2, the Zn/Cd atomic ratio is ∼4.0. Thus, X is calculated to be 0.2 in this sample. The morphology of CSZCS QDs is characterized by transmission electron microscopy (TEM). Figure S3 and Fig. 1(d) exhibit the typical TEM and high-resolution TEM (HRTEM) images, respectively, of the oleic acid (OA)-capped CSZCS QDs. Narrow size distribution of the QDs with an average size of 8.5 nm is shown in Fig. S3. Although the particle size of the QDs is small, well-resolved lattice fringes are observed in the HRTEM image illustrated in Fig. 1(d), implying good crystallinity of CSZCS QDs. The corresponding QDs revealed the lattice spacing of 0.32 nm, which should be assigned to the (100) plane of ZnS.21 

It is well known that the optical properties of quantum dots can be characterized by absorption and PL spectra.22 According to previous studies,16 the gradient alloyed structure possesses a quasi-type-II energy band, which not only expanded the exciton yield but also greatly increased the electron–hole separation efficiency compared to the core–shell structure with type-I energy band structure. The absorption and PL spectra of OA capped CSZCS QDs suspensions were provided in Fig. 1(e). The multi-excitonic absorption covering the whole visible range up to 650 nm was observed. This broad absorption range without an obvious excitonic peak further revealed the gradient alloyed structure of the obtained CSZCS QDs. Apart from forming a smooth bandgap gradient, alloying also relaxed the lattice strain of unmatched crystal structures.23–25 Meanwhile, the excited electron was delocalized over the whole structure and thereby increased the photoluminescence quantum yield (PLQY) due to the gradient alloyed structure.

However, the CSZCS QDs were synthesized at high temperatures, and the process must be assisted by the use of long surface ligands. The long ligands hampered the charge-carrier transport from CSZCS QDs to the InSe layer; thus, it was necessary to exchange them with short ligands.26 To completely remove the remaining long ligands after the electric dishcharge tube 1,2-ethanedithiol (EDT) treatment, the InSe/CSZCS QDs hybrid photodetectors were annealed at 100 °C for 30 min in N2 atmosphere.27Figure 1(f) illustrated a dramatical PL quenching phenomenon when QDs were deposited on the surface of InSe, indicating that the exciton transfer from QDs to InSe can act as an additional nonradiative relaxation channel. According to the previous work, this pathway was attributed to the nonradiative energy transfer (NRET).28–30 

Figure 2(a) shows the cross-sectional schematic diagram of this hybrid device. As noted in Fig. 2(b), both pristine and hybrid devices exhibit n-type characteristics. The field effect mobility was calculated from the relation μ = (L/W) (dIsd/CίVsd dVg), where L and W are the length and width of the sample, Cί is the capacitance per unit area of the corresponding dielectric layer [which is given by Cί = (ε0εr/d), where ε0 is free space permittivity, εr is the substrate’s dielectric constant, and d is the dielectric layer’s thickness].The calculated mobilities before and after spin-coating the quantum dots are 2.41 × 10−2 and 4.02 cm2/(V s), respectively. After deposition of the QDs sensitizing layer, the threshold voltage becomes lower and the on-state of drain current is much improved by 100 times compared to that of pristine InSe. This can be attributed to the CT effect at the InSe/QDs interface. To explore the influence of QDs on the photoelectrical performances, we measured the IV curves and time-dependent current of both the hybrid [Figs. 2(c) and 2(d)] and pristine InSe device [Figs. S4(a) and S4(b)] under the irradiation of 635 nm laser with varying light intensities. We find that the photocurrent is significantly improved via the sensitization with QDs. When the laser wavelength was changed to 405 nm, the same result was characterized for the hybrid device InSe/CSZCS QDs and the pristine InSe FET (Fig. S5).

FIG. 2.

(a) Cross-sectional view of the device operation under illumination. (b) Typical transfer curves of InSe and InSe/CSZCS QDs transistor devices in the dark. (c) The I–V curves of the InSe/CSZCS QDs hybrid device. (d) Time-dependent photoelectric response of InSe/CSZCS QDs hybrid device irradiated by different optical power densities at VDS = 2 V.

FIG. 2.

(a) Cross-sectional view of the device operation under illumination. (b) Typical transfer curves of InSe and InSe/CSZCS QDs transistor devices in the dark. (c) The I–V curves of the InSe/CSZCS QDs hybrid device. (d) Time-dependent photoelectric response of InSe/CSZCS QDs hybrid device irradiated by different optical power densities at VDS = 2 V.

Close modal

Device response speed to optical switching is a critical indicator. As illustrated in Fig. 3(a), the rise and decay time for the hybrid photodetector under irradiation of 635 nm laser are 33.6 and 17.4 ms, respectively. The response time was calculated by the variation of 90% photo-current. To understand the effect of the QDs photosensitive layer on the temporal response, we compared the response time of the original InSe device measured in the same conditions. As can be seen from Fig. 3(b), the pristine InSe detector exhibits relatively slower speed with the rise and decay time of 65.1 and 79.6 ms, respectively. As illustrated in Figs. 3(c) and 3(d), when the laser wavelength was changed to 405 nm, the rise time decreased from 46.7 to 17.5 ms, and the decay time decreased from 32.2 to 17.8 ms after sensitization with QDs. Overall, both rise and fall times were obviously reduced via deposition of CSZCS QDs due to the interface interaction of efficient CT and NRET effect between the QDs and few-layer InSe.11 

FIG. 3.

One cycle of the photoresponse under 635 nm laser at 123.849 mW/cm2 for the InSe/CdSe@Zn1−XCdXS QDs hybrid device (a) and pristine InSe FET (b) estimating both the rise and fall times. (c) and (d) Rise and fall time of the InSe/CdSe@Zn1−XCdXS QDs hybrid device and the pristine InSe FET under 405 nm laser irradiation at a power density of 17.515 mW/cm2.

FIG. 3.

One cycle of the photoresponse under 635 nm laser at 123.849 mW/cm2 for the InSe/CdSe@Zn1−XCdXS QDs hybrid device (a) and pristine InSe FET (b) estimating both the rise and fall times. (c) and (d) Rise and fall time of the InSe/CdSe@Zn1−XCdXS QDs hybrid device and the pristine InSe FET under 405 nm laser irradiation at a power density of 17.515 mW/cm2.

Close modal

Responsivity, as a key figure of merit in the photodetector, was defined as the photocurrent at a given optical power, following the equation R = Iph/P, where Iph is the photocurrent and P is the incident optical power.31 In principle, the responsivity was inversely proportional to the optical power density. Figure 4(a) displayed the responsivity of the hybrid and original InSe photodetector with different optical power densities at 635 nm laser irradiation with VDS = 2 V. Under 635 nm irradiation, the responsivity decreased with the increase in incident power because the scattering of light and recombination of charge carrier are more serious at strong light illumination.32 More importantly, due to the presence of the QDs photosensitive layer, the responsivity of the hybrid device was significantly improved under the same conditions, and the maximum responsivity (VDS = 2 V, P = 0.120 mW/cm2) was 30.16 A/W, which is 300 times larger than that of the InSe device (0.101 A/W). The same decay laws under irradiation of 405 nm laser were observed in Fig. 4(b), and the responsivity of the hybrid device is improved by 130 times from 0.282 to 37.48 A/W compared with that of pristine InSe (VDS = 2 V, P = 0.121 mW/cm2).

FIG. 4.

(a) Responsivity and (c) detectivity as a function of excitation power for the InSe/CdSe@Zn1−XCdXS QDs hybrid device under 635 nm. (b) Responsivity and (d) detectivity as a function of excitation power for the InSe/CdSe@Zn1−XCdXS QDs hybrid device and pristine InSe under 405 nm.

FIG. 4.

(a) Responsivity and (c) detectivity as a function of excitation power for the InSe/CdSe@Zn1−XCdXS QDs hybrid device under 635 nm. (b) Responsivity and (d) detectivity as a function of excitation power for the InSe/CdSe@Zn1−XCdXS QDs hybrid device and pristine InSe under 405 nm.

Close modal

As another significant parameter for the performance of photodetectors, the detectivity reflects the sensitivity of photodetection under the noise-level floor of its dark current. It can be calculated by the following equation: D* = RA1/2/(2eIdark)1/2, where R is the responsivity, A is the active area of the device, e is the unit charge, and Idark was the dark current.33 We assume that the shot noise from direct current mainly contributes to the noise. This InSe/CSZCS QDs hybrid photodetector exhibited a high detectivity of 1.69 × 1012 Jones at 0.120 mW/cm2 under 635 nm illumination in Fig. 4(c). As presented in Fig. 4(d), the maximum detectivity of the InSe/CSZCS QDs hybrid photodetector under 405 nm irradiation can reach up to 3.11 × 1012 Jones. The high performance of the hybrid photodetector was attributed to the high absorption of QDs and the high carrier mobility of InSe.

In order to further explore the mechanism of charge separation and interaction between CdSe@Zn1−XCdXS QDs and InSe (∼9.7 nm), we introduced another core/shell CdSe@ZnS QDs. The diameter of CdSe QDs cores is 7 nm, and the thickness of the ZnS shell is 2 nm. The core/shell QDs possess the type I energy band. The ECB and EVB of QD core CdSe are −4.25 and −6.21 eV, respectively. The conduction band of the shell layer ZnS is −2.50 eV, and the valence band is −6.50 eV.34 Similar to the InSe/CSZCS QDs photodetector, a hybrid device InSe/CdSe@ZnS QDs was fabricated. An optical image and IV curves of the device were provided in Fig. S6. We have compared the photoelectric performance of the hybrid InSe/core–shell QDs photodetector with the original InSe device under the 635 nm, as shown in Fig. 5. The calculated R values of hybrid devices give more than 38-fold (from 0.14 to 5.39 A/W) enhancement in responsivity compared to those of the pristine InSe devices. The detectivity was six times higher than that of its pristine InSe counterpart (from 1.15 × 1011 to 6.81 × 1011 Jones). The rise time is reduced from 61.4 to 49.7 ms, and the decay time is from 55.6 to 26.3 ms. The similar optoelectronic performance of the device under 405 nm laser irradiation was displayed in Fig. S7.

FIG. 5.

Photoresponse of the InSe/CdSe@ZnS QDs hybrid device (a) and pristine InSe (d) with different optical power densities at 2 V bias and 635 nm laser. One cycle of the photoresponse under 635 nm laser at 123.849 mW/cm2 for the InSe/CdSe@ZnS QDs hybrid device (b) and pristine InSe device (e) to estimating both the rise and fall time. (c) Detectivity and (f) responsivity as a function of excitation power for the InSe/CdSe@ZnS QDs hybrid device and pristine InSe under 635 nm.

FIG. 5.

Photoresponse of the InSe/CdSe@ZnS QDs hybrid device (a) and pristine InSe (d) with different optical power densities at 2 V bias and 635 nm laser. One cycle of the photoresponse under 635 nm laser at 123.849 mW/cm2 for the InSe/CdSe@ZnS QDs hybrid device (b) and pristine InSe device (e) to estimating both the rise and fall time. (c) Detectivity and (f) responsivity as a function of excitation power for the InSe/CdSe@ZnS QDs hybrid device and pristine InSe under 635 nm.

Close modal

To this end, it was clear that the light detection performance of InSe was greatly improved for both core/shell structures of CdSe@ZnS QDs and gradient alloyed structures of CdSe@Zn1−XCdXS QDs at 635 and 405 nm, respectively. However, in conjunction with the previous description, the increase in the photoelectric performance of InSe for the gradient alloyed structure was always better than that for the core/shell structure under the same conditions. For instance, the maximum responsivity for the gradient alloy system at 405 nm is 37.48 A/W, much greater than the 15.74 A/W in the core/shell system at the same measure condition. From previous studies,16,34 it was clear that the only difference between these two types of QDs was the arrangement of the energy bands determined by their intrinsic structure.

The energy level alignment of the InSe/CSZCS QDs hybrid device was shown in Fig. 6(a). The conduction and valence band edges of few-layer InSe are ECB = −4.60 and EVB = −5.86 eV, respectively.19,35 The gradient alloyed CdSe@Zn1−XCdXS QDs possess a center of CdSe similar to core/shell structure, but the gradient alloying strategy from inner to outer provides systematic control over the electronic structure and enables switching from type-I to quasi-type-II configurations,16 which is more favorable for photogenerated electron–hole separation than that of the core/shell structure. As such, when the gradient alloyed CdSe@Zn1−XCdXS QDs approached few-layer InSe, the high density of the separated electrons and holes can be easily injected to the few-layer InSe through the energy band alignment at the interface. Therefore, the funnel-like energy level is beneficial to electron tunneling and the charge carriers flow from the QDs photosensitive layer to the InSe transport layer, realizing the efficient CT effect in Fig. 6(a) (process II).11 At the same time, another energy transfer form of NRET was also working. Excitons were first generated in the photosensitive layer, and then the exciton energy was transferred non-radiometrically to the InSe layer through dipole–dipole coupling, creating a new electron–hole pair on the InSe film, as shown in the Ⅲ process of Fig. 6(a).36 This matches the PL quenching results in Fig. 1(c). In addition, due to the intrinsic semiconductor property of InSe and quantum dots, both absorb photons when illuminated to produce electron–hole pairs (process I).5 

FIG. 6.

(a) Energy diagram of the interface between InSe and CSZCS QDs after the formation of a heterojunction. Three photoelectrical processes are proposed: I, photon excitation in InSe and the QDs; II, electron transfer from the QDs to InSe via tunneling; and III, exciton transfer from the QDs to InSe via NRET processes. (b) Energy band diagram for a InSe/CdSe@ZnS hybrid, showing NRET as a possible pathway from a photoexcited CdSe@ZnS QD to InSe because the insulating ZnS shell acts as a tunneling barrier and inhibits CT from the CdSe core to InSe.

FIG. 6.

(a) Energy diagram of the interface between InSe and CSZCS QDs after the formation of a heterojunction. Three photoelectrical processes are proposed: I, photon excitation in InSe and the QDs; II, electron transfer from the QDs to InSe via tunneling; and III, exciton transfer from the QDs to InSe via NRET processes. (b) Energy band diagram for a InSe/CdSe@ZnS hybrid, showing NRET as a possible pathway from a photoexcited CdSe@ZnS QD to InSe because the insulating ZnS shell acts as a tunneling barrier and inhibits CT from the CdSe core to InSe.

Close modal

Moreover, the schematic of a core/shell heterostructure and the energy band diagram of core/shell QDs type-I is depicted in Fig. 6(b). The ZnS shell acting as a physical barrier against the charge transfer from the QD core to InSe. The core part (CdSe) in QDs was excited by light to produce electrons and holes crossing the physical barrier (ZnS) with great difficulty, leading to the suppressed CT effect between QDs and InSe.34 Thus, process II was blocked, which was manifested via the reduced photocurrent contributed by the CT effect. However, the 2 nm thick shell could not block the NRET effect between the QDs and the 2D materials.37 The energy of the QDs excited by the laser may be partly transferred to InSe through the NRET effect (process III), which was the source of the increased portion of photocurrent in the hybrid system. Therefore, the interfacial interaction of the gradient alloyed structure InSe/CSZCS QDs was stronger than that of the core–shell structure InSe/CdSe@ZnS, demonstrated by the greater photoelectric performance of the former.

We have developed a highly photosensitive hybrid photodetector based on gradient alloyed QDs and few-layer InSe. In order to remove the insulating ligands on the QDs surface, the method of using EDT layer by layer is adopted. Under the condition of 635 nm at 0.120 mW/cm2 illumination, the InSe/CdSe@Zn1−XCdXS QDs hybrid photodetector has a responsivity of 30.1 A/W and a detectivity of 1.69 × 1012 Jones. Compared with the core–shell structure InSe/CdSe@ZnS, the excellent performance of the InSe/CdSe@Zn1−XCdXS QDs hybrid photodetector is attributed to the quasi-type-II energy level alignment of the gradient alloyed QDs induced efficient CT effect, and NRet also plays a key role. More importantly, we have demonstrated a lower cost and more efficient research perspective, starting from the quantum dot structure with unique funnel-shaped band alignment to optimize the photoelectric performance of two-dimensional materials.

CdSe@Zn1−XCdXS gradient alloyed QDs were synthesized by slightly modifying the previous methods reported by our group.12 The detailed synthesis procedure is provided in the supplementary material.

Structural analysis and optical characterization of CdSe@Zn1−XCdXS QDs and the InSe sheet were performed using TEM (JEM2010-HR), UV–vis absorption spectroscopy (SHIMADZU UV-3600 Plus spectrophotometer), and AFM (Dimension Fast Scan from Bruker Co., Ltd.). Raman spectra and PL were recorded using a Raman instrument (NOST Technology Co., Ltd. from South Korea; a laser excitation of 532 nm and a spot size of 120 µm). The morphology and element composition were analyzed by scanning electron microscopy (SEM, FEI Quanta FEG 250) with elemental mapping and energy dispersive spectroscopy (EDS).

A few-layer InSe exfoliated from bulk InSe flakes is deposited on a Si/SiO2 substrate. The metal electrodes were prepared via an ultraviolet lithography method. First, a photoresist (AR-P 5350 of ALLRESIST GmbH Company from Germany) was spin-coated with a speed of 3500 rpm and a time of 60 s. After being baked on a hot plate at 100 °C for 4 min, the substrate was exposed to an ultraviolet intensity of ∼30 mW cm−2 for 4 s with no need for alignment via a lithography machine. The chip was then developed using a freshly prepared developing solution (the volume ratio of tetramethyl ammonium hydroxide 25 wt. % water solutions to deionized water = 1:33) for 30 s and dried using a nitrogen gas gun. Moreover, a 10/50 nm Cr/Au layer was deposited onto the substrate with the help of a thermal evaporation instrument and washed by acetone. The backgated electrode was used by degenerated p-type doped Si wafer. To improve the contact resistance, all devices were annealed at 150 °C for an hour in an N2 atmosphere glovebox.

CdSe@Zn1−XCdXS QDs dispersed in octane were spin-coated with 1300 rpm on the fabricated InSe device. To remove the surface ligands of QDs, 2 vol. % EDT diluted in acetonitrile was spin-coated on the hybrid device. Repeat the above operation twice. After spin-coating, the device was subject to thermal treatment at 100 °C for 30 min in a glovebox filled with nitrogen.

CdSe@ZnS QDs were purchased from Suzouxinsuo Co., Ltd. The preparation process is the same as described above.

The FET and photocurrent characteristics were measured using a four-probe station (PSAICPB6A, Precision Systems Industrial Co., Ltd.) equipped with a KEITHLEY B2636B semiconductor source meter under ambient atmosphere. The photo-response behavior of the device was tested by using a 405 and a 635 nm optical-fiber laser, respectively. The diameter of the excitation laser spot is ∼10 mm, and the response time was extracted via a shutter system with the modulator–demodulator function.

See the supplementary material for additional fabrication and electrical data.

This work was supported by the National Natural Science Foundation of China (Grant Nos. 11904108 and 62004071), the China Postdoctoral Science Foundation (Grant No. 2020M672680), and the “The Pearl River Talent Recruitment Program” (Grant No. 2019ZT08X639).

Z.D. and C.W. contributed equally to this work.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

1.
D.-H.
Kwak
,
P.
Ramasamy
,
Y.-S.
Lee
,
M.-H.
Jeong
, and
J.-S.
Lee
,
ACS Appl. Mater. Interfaces
11
(
32
),
29041
(
2019
).
2.
S.
Zhang
,
X.
Wang
,
Y.
Chen
,
G.
Wu
,
Y.
Tang
,
L.
Zhu
,
H.
Wang
,
W.
Jiang
,
L.
Sun
,
T.
Lin
,
H.
Shen
,
W.
Hu
,
J.
Ge
,
J.
Wang
,
X.
Meng
, and
J.
Chu
,
ACS Appl. Mater. Interfaces
11
(
26
),
23667
(
2019
).
3.
A.
Robin
,
E.
Lhuillier
,
X. Z.
Xu
,
S.
Ithurria
,
H.
Aubin
,
A.
Ouerghi
, and
B.
Dubertret
,
Sci. Rep.
6
,
24909
(
2016
).
4.
H.
Wu
,
Z.
Kang
,
Z.
Zhang
,
Z.
Zhang
,
H.
Si
,
Q.
Liao
,
S.
Zhang
,
J.
Wu
,
X.
Zhang
, and
Y.
Zhang
,
Adv. Funct. Mater.
28
(
34
),
1870239
(
2018
).
5.
H.
Zang
,
P. K.
Routh
,
Y.
Huang
,
J.-S.
Chen
,
E.
Sutter
,
P.
Sutter
, and
M.
Cotlet
,
ACS Nano
10
(
4
),
4790
(
2016
).
6.
V.
Selamneni
,
S. K.
Ganeshan
, and
P.
Sahatiya
,
J. Mater. Chem. C
8
(
33
),
11593
(
2020
).
7.
S.
Pradhan
,
M.
Dalmases
, and
G.
Konstantatos
,
Adv. Mater.
32
(
45
),
2003830
(
2020
).
8.
I.
Eren
,
S.
Ozen
,
Y.
Sozen
,
M.
Yagmurcukardes
, and
H.
Sahin
,
J. Phys. Chem. C
123
(
51
),
31232
(
2019
).
9.
M.
Dai
,
H.
Chen
,
F.
Wang
,
M.
Long
,
H.
Shang
,
Y.
Hu
,
W.
Li
,
C.
Ge
,
J.
Zhang
,
T.
Zhai
,
Y.
Fu
, and
P.
Hu
,
ACS Nano
14
(
7
),
9098
(
2020
).
10.
N.
Huo
,
S.
Gupta
, and
G.
Konstantatos
,
Adv. Mater.
29
(
17
),
1606576
(
2017
).
11.
D.
Kufer
,
I.
Nikitskiy
,
T.
Lasanta
,
G.
Navickaite
,
F. H. L.
Koppens
, and
G.
Konstantatos
,
Adv. Mater.
27
(
1
),
176
(
2015
).
12.
X.
Liu
,
D.
Wen
,
Z.
Liu
,
J.
Wei
,
D.
Bu
, and
S.
Huang
,
Chem. Eng. J.
402
,
126178
(
2020
).
13.
K.
Boldt
,
S.
Bartlett
,
N.
Kirkwood
, and
B.
Johannessen
,
Nano Lett.
20
(
2
),
1009
(
2020
).
14.
G. A.
Beane
,
K.
Gong
, and
D. F.
Kelley
,
ACS Nano
10
(
3
),
3755
(
2016
).
15.
Y.-S.
Park
,
J.
Lim
,
N. S.
Makarov
, and
V. I.
Klimov
,
Nano Lett.
17
(
9
),
5607
(
2017
).
16.
K.
Boldt
,
N.
Kirkwood
,
G. A.
Beane
, and
P.
Mulvaney
,
Chem. Mater.
25
(
23
),
4731
(
2013
).
17.
L.
Gao
,
C.
Chen
,
K.
Zeng
,
C.
Ge
,
D.
Yang
,
H.
Song
, and
J.
Tang
,
Light: Sci. Appl.
5
(
7
),
e16126
(
2016
).
18.
W.
Gao
,
Z.
Zheng
,
Y.
Li
,
C.
Xia
,
J.
Du
,
Yu.
Zhao
, and
J.
Li
,
J. Mater. Chem. C
6
(
46
),
12509
(
2018
).
19.
Q.
Lv
,
F.
Yan
,
N.
Mori
,
W.
Zhu
,
C.
Hu
,
Z. R.
Kudrynskyi
,
Z. D.
Kovalyuk
,
A.
Patanè
, and
K.
Wang
,
Adv. Funct. Mater.
30
(
15
),
1910713
(
2020
).
20.
Q.
Zhao
,
W.
Wang
,
F.
Carrascoso-Plana
,
W.
Jie
,
T.
Wang
,
A.
Castellanos-Gomez
, and
R.
Frisenda
,
Mater. Horiz.
7
(
1
),
252
(
2020
).
21.
X.
Tang
,
K.
Yu
,
Q.
Xu
,
E.
Shi Guan Choo
,
G. K. L.
Goh
, and
J.
Xue
,
J. Mater. Chem.
21
(
30
),
11239
(
2011
).
22.
D.
Prasai
,
A. R.
Klots
,
A.
Newaz
,
J. S.
Niezgoda
,
N. J.
Orfield
,
C. A.
Escobar
,
A.
Wynn
,
A.
Efimov
,
G. K.
Jennings
,
S. J.
Rosenthal
, and
K. I.
Bolotin
,
Nano Lett.
15
(
7
),
4374
(
2015
).
23.
S. M.
Fairclough
,
E. J.
Tyrrell
,
D. M.
Graham
,
P. J. B.
Lunt
,
S. J. O.
Hardman
,
A.
Pietzsch
,
F.
Hennies
,
J.
Moghal
,
W. R.
Flavell
,
A. A. R.
Watt
, and
J. M.
Smith
,
J. Phys. Chem. C
116
(
51
),
26898
(
2012
).
24.
B. G.
Jeong
,
Y.-S.
Park
,
J. H.
Chang
,
I.
Cho
,
J. K.
Kim
,
H.
Kim
,
K.
Char
,
J.
Cho
,
V. I.
Klimov
,
P.
Park
,
D. C.
Lee
, and
W. K.
Bae
,
ACS Nano
10
(
10
),
9297
(
2016
).
25.
N.
Tschirner
,
H.
Lange
,
A.
Schliwa
,
A.
Biermann
,
C.
Thomsen
,
K.
Lambert
,
R.
Gomes
, and
Z.
Hens
,
Chem. Mater.
24
(
2
),
311
(
2012
).
26.
G.
Konstantatos
,
M.
Badioli
,
L.
Gaudreau
,
J.
Osmond
,
M.
Bernechea
,
F. P. G.
de Arquer
,
F.
Gatti
, and
F. H. L.
Koppens
,
Nat. Nanotechnol.
7
(
6
),
363
(
2012
).
27.
J.-H.
Choi
,
A. T.
Fafarman
,
S. J.
Oh
,
D.-K.
Ko
,
D. K.
Kim
,
B. T.
Diroll
,
S.
Muramoto
,
J. G.
Gillen
,
C. B.
Murray
, and
C. R.
Kagan
,
Nano Lett.
12
(
5
),
2631
(
2012
).
28.
J.
Gu
,
X.
Liu
,
E.-c.
Lin
,
Y.-H.
Lee
,
S. R.
Forrest
, and
V. M.
Menon
,
ACS Photonics
5
(
1
),
100
(
2017
).
29.
P. L.
Hernández-Martínez
,
A. O.
Govorov
, and
H. V.
Demir
,
J. Phys. Chem. C
117
(
19
),
10203
(
2013
).
30.
A.
Raja
,
A.
Montoya−Castillo
,
J.
Zultak
,
X.-X.
Zhang
,
Z.
Ye
,
C.
Roquelet
,
D. A.
Chenet
,
A. M.
van der Zande
,
P.
Huang
,
S.
Jockusch
,
J.
Hone
,
D. R.
Reichman
,
L. E.
Brus
, and
T. F.
Heinz
,
Nano Lett.
16
(
4
),
2328
(
2016
).
31.
N.
Huo
and
G.
Konstantatos
,
Nat. Commun.
8
(
1
),
572
(
2017
).
32.
Y.
Sun
,
W.
Song
,
F.
Gao
,
X.
Wang
,
X.
Luo
,
J.
Guo
,
B.
Zhang
,
J.
Shi
,
C.
Cheng
,
Q.
Liu
, and
S.
Li
,
ACS Appl. Mater. Interfaces
12
(
11
),
13473
(
2020
).
33.
Z.
Zheng
,
J.
Yao
,
J.
Li
, and
G.
Yang
,
Mater. Horiz.
7
(
9
),
2185
(
2020
).
34.
M.
Li
,
J.‐S.
Chen
,
P. K.
Routh
,
P.
Zahl
,
C.-Y.
Nam
, and
M.
Cotlet
,
Adv. Funct. Mater.
28
(
29
),
1707558
(
2018
).
35.
F.
Qin
,
F.
Gao
,
M.
Dai
,
Y.
Hu
,
M.
Yu
,
L.
Wang
,
W.
Feng
,
B.
Li
, and
P.
Hu
,
ACS Appl. Mater. Interfaces
12
(
33
),
37313
(
2020
).
36.
F.
Prins
,
A. J.
Goodman
, and
W. A.
Tisdale
,
Nano Lett.
14
(
11
),
6087
(
2014
).
37.
K. M.
Goodfellow
,
C.
Chakraborty
,
S.
Kelly
,
P.
Waduge
,
M.
Wanunu
,
T.
Krauss
,
K.
Driscoll
, and
A.
Nick Vamivakas
,
Appl. Phys. Lett.
108
(
2
),
021101
(
2016
).

Supplementary Material