Surface ligand engineering of Ag₂Te colloidal quantum dots for a self-powered shortwave infrared photodetector

Abstract

Traditional colloidal quantum dots (CQDs) used for shortwave infrared (SWIR) photodetectors, such as mercury (Hg)-based and lead (Pb)-based CQDs, have been developed with high detectivity and fast responses, but they contain toxic heavy metals. As a non-toxic material, emerging Ag₂Te CQDs provide a new path for the development of the next generation of SWIR photodetectors due to their tunable bandgap, solution-processability, low cost, and environmental friendliness. However, the high surface energy and low bond dissociation energy of Ag₂Te CQDs make them prone to decomposition during high-temperature synthesis, resulting in poor monodispersity. Here, we propose a surface modification strategy based on bifunctional thiol ligands, combined with reaction kinetics regulation, to achieve stable synthesis of Ag₂Te CQDs and discuss their application in SWIR photodetectors. The results show that the synthesized Ag₂Te CQDs have high monodispersity, and they help realize a wide response self-powered photodiode from 400 to 1600 nm at 0 V bias. The specific detectivity of the SWIR Ag₂Te CQD photodetector is of the order of 10¹⁰ Jones, the −3 dB bandwidth reaches 211 kHz, the linear dynamic range exceeds 74 dB, and the fast response time is 13.3 μs at room temperature (25 °C). In addition, benefiting from its excellent SWIR detection performance, we further demonstrate the application of the Ag₂Te CQD photodetector in non-contact transmission pulse monitoring.

---

1. Introduction

As an important component of infrared spectroscopy, shortwave infrared (SWIR) radiation exhibits excellent optical and thermal properties, with reduced scattering and absorption effects in atmospheric transmission, allowing it to effectively penetrate smoke, thin clouds, and partial dielectric materials. Most substances have characteristic absorption and reflection spectra within this wavelength range, giving SWIR a unique advantage in the field of spectral analysis.^1–4 With the development of high-performance photodetectors such as HgCdTe and InGaAs, advanced optical systems, intelligent signal processing technologies, and image sensors, SWIR technology has been widely used in military reconnaissance, environmental remote sensing, industrial detection, biomedical imaging, and other fields.^5–7 However, most of these traditional materials adopt an epitaxial growth process, which requires substantial equipment, complex fabrication process, high cost, and refrigeration, hindering the wide application of infrared detectors.^8,9 Colloidal quantum dots (CQDs) have been widely used in the field of SWIR photodetectors due to their easy fabrication and large-scale and low-cost synthesis processes.^10–22 At present, mercury (Hg)-based and lead (Pb)-based CQDs are relatively mature, but due to the limitation of toxicity, it is necessary to strengthen safety monitoring and control during use.^10–19 Therefore, it is of great significance to develop a simple, low-cost, and eco-friendly synthesis process to fabricate non-toxic CQD-based SWIR photodetectors.

Ag₂Te is non-toxic and free of heavy metals, with a narrow direct bandgap and high electron mobility, which can be used to fabricate non-toxic CQD-based SWIR photodetectors.^23–32 However, due to the existence of numerous dangling bonds and defect states on the surface of Ag₂Te CQDs, the surface energy is significantly increased, making them prone to aggregation in solution environments and difficult to maintain a stable monodisperse state and thus affecting device performance.¹ Hence, there is a need to investigate suitable ligands for modifying Ag₂Te CQDs in order to enhance their stability and optical performance. Ahn et al. effectively reduced the surface defect states of Ag₂Te CQDs and improved carrier mobility through a novel solid-phase ligand exchange strategy, successfully constructing a photodetector with a response time of only 72 ns.²⁴ However, the presence of phosphine hindered the effective ligand exchange and surface passivation.^25,26 Therefore, even if the best solid-phase ligand exchange solution was selected to optimize the performance of the Ag₂Te CQDs photodetector, the highest detectivity can only reach the order of 10⁷. Konstantatos's research group successfully fabricated highly monodisperse Ag₂Te CQDs by developing a phosphine-free precursor synthesis method.²⁶ This strategy effectively suppresses the Ostwald ripening phenomenon, enabling the photodetector to achieve a high specific detectivity (D*) of over 3 × 10¹² Jones at room temperature. On this basis, the latest research has used zinc halide control methods to form Ag–X complexes to reduce precursor reactivity, achieving precise control of CQDs in the optical bandgap range of 1.22–0.60 eV.²⁷ The above research progress indicates that the nucleation and growth kinetics of Ag₂Te CQDs can be precisely controlled by optimizing the synthesis process, thereby achieving a CQD system with narrow size distribution and tunable bandgap. Although the phosphine-free method has significant advantages, in the traditional synthesis process, the oleylamine (OLAm) ligand used has weak coordination ability with Ag⁺ ions, resulting in a large number of dangling bonds and defect states on the surface of the CQDs. This not only makes the CQDs prone to aggregation in solution environments, making it difficult to maintain long-term colloidal stability, but also introduces nonradiative recombination centers, thereby affecting device performance.^28,33–35 In response to the above issues, referring to the relatively mature HgTe CQD system, thiols can be used to stabilize the synthesis of CQDs.^13,14 The strong interaction between thiols and metal ions can be used to enhance the dispersion of crystal nuclei, effectively improving the aggregation problem of the CQDs and increasing their solubility in non-polar solvents, thereby helping to improve the quality of subsequent HgTe CQD films and enhance the optoelectronic performance of devices.^13,14

Based on the above analysis, we adopted a phosphine-free synthesis route and further proposed a surface modification strategy based on bifunctional thiol ligands, combined with reaction kinetics regulation, to achieve stable synthesis of Ag₂Te CQDs. The strong binding energy between thiol ligands and metal ions can increase the nucleation threshold and regulate the crystal growth process, laying the foundation for obtaining monodisperse, spherical, and high-quality CQDs. Moreover, stable Ag–S bonds are formed between thiol ligands and Ag, which can directly fill surface defects and construct a dense monolayer on the surface of the CQDs, effectively blocking external quenching factors such as oxygen and moisture. This dense ligand adsorption layer not only passivates surface defects but also effectively blocks direct contact between the CQDs, reducing the tendency for aggregation caused by surface charge interactions or van der Waals forces of the CQDs, thereby achieving bifunctional modification. In addition, octadecene (ODE) is added to both precursor solutions to reduce the reaction activity and further control the growth rate of the CQDs, so as to control accurately the size of the CQDs. The results show that the Ag₂Te CQD-based SWIR photodetector has D* of the order of 10¹⁰ Jones at room temperature, the −3 dB bandwidth reaches 211 kHz, and the fast response time is 13.3 μs. Benefiting from the excellent SWIR detection performance, we further demonstrate the application of the Ag₂Te CQDs in non-contact transmission pulse monitoring, effectively achieving the real-time detection of human pulse signals.

2. Experimental section

2.1. Materials and reagents

The materials used for fabricating the Ag₂Te CQD-based SWIR photodetector include silver acetate (AgAc, Aladdin, AR, ≥99.5%), tellurium dioxide (TeO₂, Aladdin, 99.99%), 1-octanethiol (OTT, Aladdin, ≥98%), ODE (Aladdin, ≥90%), 1-dodecanethiol (DDT, Aladdin, 98%), isopropanol (IPA, Aladdin, >99.7%), toluene (Chron Chemicals, 99.5%), octane (OCT, Aladdin, >99%), and 1,2-ethanedithiol (EDT, Aladdin, 99.5%).

2.2. TeO₂–DDT–ODE precursor (0.2 mol L⁻¹)

0.1596 g (1 mmol) of TeO₂ was added into a 100 mL three-neck flask containing 4 mL of DDT and 1 mL of ODE at 100 °C and under a nitrogen atmosphere to obtain a clear yellow solution (Fig. S1).

2.3. Synthesis of the Ag₂Te CQDs

333.8 mg (2 mmol) of AgAc was added to a three-neck flask containing 2 mL of OTT and 8 mL of ODE and purged with nitrogen for 30 min. When the temperature was raised to 140 °C, the solution turned to a clear orange color and 3.8 mL TeO₂–DDT–ODE was quickly injected at this time. After 5 min of reaction, the mixture was cooled to room temperature by an ice bath. After cooling, an equal volume of IPA was added to the crude solution to precipitate the CQDs and after centrifuging at 10 000 rpm for 5 min, the precipitate was resuspended in 10 mL toluene. Thereafter, the solution was precipitated a second time with an equal volume of IPA. The final collected Ag₂Te CQDs were dried and dissolved in octane for characterization and device fabrication.

2.4. Materials characterization

The X-ray diffraction (XRD) patterns were recorded with a Rigaku Ultima IV using Cu-Kα1 irradiation in the range of 2θ from 5° to 90°, with a step size of 0.02°. The morphology and crystalline structure of the Ag₂Te CQDs were examined by a JEOL 2100 Plus transmission electron microscopy (TEM) instrument at 300 kV. Optical absorption spectra of Ag₂Te CQD solution and film were collected using a Shimadzu UV-3600 Plus spectrophotometer with a range of 300–2500 nm. Surface morphologies of the Ag₂Te CQDs and photodetector cross-sectional microstructure were obtained using a ZEISS Gemini 300 scanning electron microscope (SEM). Ultraviolet photoelectron spectroscopy (UPS) measurements were performed with an electron spectrometer using monochromatic HeI radiation (21.22 eV) of a Thermo Scientific-ESCALAB Xi+ in ultrahigh-vacuum conditions at 5 V bias.

2.5. Fabrication of the Ag₂Te CQD-based SWIR photodetector

The structure of the Ag₂Te CQD-based photodetector is: indium tin oxide (ITO) (cathode)/ZnO (electron transport layer)/Ag₂Te CQDs (active layer)/MoO₃ (hole transport layer)/Au (anode). Its fabrication process is as follows. The patterned ITO conductive glass substrate was wiped with a detergent and sonicated with deionized water, acetone, and IPA for 15 min, followed by drying with a nitrogen stream. The cleaned ITO substrate was treated with UV-ozone for 20 min. The ZnO film (∼40 nm) was prepared by spin coating ZnO precursor on the ITO substrate as in our previous report.³⁶ The Ag₂Te CQD active layer was fabricated by layer-by-layer spin coating on the ZnO substrate, and each iteration consisting of three steps. (1) The solution of Ag₂Te CQDs (80 mg mL⁻¹) was dropped onto the ZnO substrate and spin-coated at 2000 rpm for 20 s. (2) Then the film was covered with EDT/IPA (1 : 100 v/v) solution for 30 s and spin-coated at 2000 rpm for 20 s. (3) Finally, IPA was used to clean the residual EDT on the surface of the Ag₂Te CQD film, the excess IPA was removed by spin coating at 2000 rpm for 20 s, and the film was annealed at 50 °C for 10 min. The above process was repeated 8 times to obtain an Ag₂Te CQD film. Finally, the MoO₃ layer (10 nm) and Au layer (100 nm) were sequentially thermally evaporated under high vacuum (<10⁻⁴ Pa).¹⁷

2.6. Measurements of the photodetector

The current–voltage (I–V) curves, photocurrent broadband response spectrum, D*, response (R), and external quantum efficiency (EQE) were measured with a semiconductor characterization system (PD-QE, Enli Technology) under a xenon light source coupled with a monochromator at room temperature (25 °C), calibrated with a standard Si detector (S1337-1010BQ, Hamamatsu Photonics) and InGaAs detector (Hamamatsu Photonics). The response bandwidth, linear dynamic range (LDR), and response time were measured by a semiconductor parameter analyzer (PD-RS, Enli Technology).

3. Results and discussion

3.1. Characterization of the Ag₂Te CQDs

Fig. 1a illustrates the growth process and formation mechanism of the Ag₂Te CQDs. According to the LaMer nucleation growth theory, the CQD synthesis can be divided into rapid nucleation stage (step 1) and slow growth stage (step 2).^37,38 In this work, we used AgAc/OTT/ODE and TeO₂/DDT/ODE complexes as precursors for thermal injection, where TeO₂ was reduced to divalent tellurium by thiols and further reacted with the silver precursor.²⁶ When the tellurium precursor was rapidly injected into the silver precursor, the monomer concentration exceeds the critical supersaturation level and enters the nucleation stage. At this time, due to the strong binding energy between thiol ligands and metal ions, the concentration of free metal ions in the reaction system decreases and the nucleation threshold increases, resulting in the formation of a large number of uniformly sized crystal nuclei at the moment of high-temperature thermal injection.^39–41 When the monomer concentration drops below the critical supersaturation threshold, the CQDs enter the growth stage. Thiol molecules precisely control the ion supply rate through dynamic coordination dissociation equilibrium, and the directional adsorption of thiols promotes the formation of regular spherical CQDs.^13,14 The stable Ag–S bonds formed between thiol ligands and Ag ions can directly fill defect sites and form a dense monomolecular layer on the surface of the CQDs, shielding external quenching factors such as oxygen and water.¹⁴ The dense ligand adsorption layer not only passivates surface defects but also effectively blocks direct contact between the CQDs, reducing the tendency for aggregation caused by surface charge interactions or van der Waals forces of the CQDs, thereby achieving bifunctional modification.^42,43 In addition, ODE was added to the precursor solutions to reduce reaction activity and better control the growth rate of the CQDs.^31,32 Although ODE does not have a coordinating effect, its long-chain alkane structure can significantly increase the viscosity of the reaction system.^23,25,30 The high-viscosity medium effectively suppresses the diffusion rate of silver precursor and tellurium precursor ions in the solution, thereby significantly reducing the probability of effective collision between Ag ions and Te ions.^33,35 The mechanism not only significantly slows down the growth kinetics of the CQDs, but also provides a physical basis for achieving precise control of their growth process. The physical method of controlling ion diffusion and collision frequency by regulating system viscosity may be one of the key factors in obtaining high-quality, monodisperse CQDs.^44–46 After synthesizing the Ag₂Te CQDs, they were purified and dissolved in n-octane to produce CQD ink (Fig. S2). In this non-polar solvent, the coordination-dissociation process between thiol ligands and the surface of the Ag₂Te CQDs is in a dynamic equilibrium state. When external factors (such as temperature changes) may cause CQD aggregation, dissociated thiol molecules can quickly re-coordinate to the active sites on the surface of the CQDs to maintain the integrity of the surface ligand layer, thereby ensuring the dispersion stability of the CQDs.

Fig. 1 (a) A schematic diagram of the growth process and formation mechanism of the Ag₂Te CQDs. (b) The absorption spectrum of the Ag₂Te CQDs. (c) TEM images of the Ag₂Te CQDs. (d) The XRD pattern of the Ag₂Te CQDs. (e) XPS spectra of the Ag₂Te CQDs.

Fig. 1b shows the absorption spectrum of the Ag₂Te CQDs, from which clear exciton peaks can be observed, indicating that the CQDs have good monodispersity. To evaluate the crystal quality of the synthesized CQDs, TEM and XRD were used. Fig. 1c shows TEM images of the Ag₂Te CQDs. It can be seen that the Ag₂Te CQDs are densely arranged without an obvious stacking phenomenon. Meanwhile, they exhibit regular spherical structures and uniform size distributions, further confirming the excellent monodispersity. The appearance of some rod-shaped CQDs in the image can be attributed to the reaction rate being too fast. This also indicates that we use ODE to reduce reaction activity to control the growth rate of CQDs, resulting in most CQDs exhibiting uniform and regular spherical structures. Although there are a few rod-shaped CQD structures, this does not affect the overall properties of the Ag₂Te CQDs. The high-resolution image displays clear lattice stripes with a corresponding interplanar spacing of 0.23 nm, confirming the high crystallinity of the Ag₂Te CQDs.^26,27Fig. 1d shows the XRD patterns of the Ag₂Te CQDs. The characteristic diffraction peaks located at 24.3°, 40.6°, and 44.6° correspond to the (121), (040), and (402) crystal planes of the monoclinic crystal structure of Ag₂Te, respectively.²⁷ In addition, the characteristic peaks observed at 30.2°, 34.3°, and 50.7° are attributed to the ITO substrate. This is because we fabricated thin-film samples for XRD characterization by drop-coating CQD solution on ITO glass and annealing, resulting in some characteristic peaks of the ITO substrate appearing in the pattern. In addition, Fig. 1e presents the X-ray photoelectron spectroscopy (XPS) results of the Ag₂Te CQDs to analyze their composition. In the narrow scans of Ag 3d and Te 3d, peaks were observed at 374.4 eV (3d_3/2) and 368.4 eV (3d_5/2) for Ag atoms, while peaks were observed at 583.0 eV (3d_3/2) and 572.7 eV (3d_5/2) for Te atoms, further confirming the successful synthesis of the Ag₂Te CQDs.²⁴

3.2. Device performances of the Ag₂Te CQD SWIR photodetector

Fig. 2a shows the device structure of the ITO/ZnO/Ag₂Te CQDs/MoO₃/Au photodetector (detailed in the experimental section). The dense Ag₂Te CQD film was fabricated by multi-layer spin coating using traditional solid-phase ligand exchange methods (Fig. S3). Fig. S4 shows the cross-sectional SEM image of the photodetector, indicating that the thickness of the Ag₂Te CQD layer is about 278 nm. Due to the limited resolution of SEM, it is difficult for the ZnO layer and MoO₃ layer to form clear and distinguishable morphological features and interlayer boundaries, so they cannot be clearly observed in the image. Referring to our previous work, the thicknesses of ZnO layer and MoO₃ layer are about 40 nm and 10 nm, respectively.¹⁷ The band alignment of the ZnO/Ag₂Te CQD heterojunction before contact can be used to understand its optoelectronic properties. Using the Tauc method, the optical bandgap of the Ag₂Te CQD film was calculated to be 0.90 eV based on the absorption spectrum (Fig. S5).⁴⁷ In addition, UPS can be used to reveal the maximum valence band (E_V) and Fermi energy (E_F) of the Ag₂Te CQD film, calculated as 4.93 eV and 4.31 eV, respectively (Fig. S6). As a commonly used wide-bandgap semiconductor material, ZnO has a bandgap of 3.25 eV and a minimum conduction band (E_C) and Ev of 4.0 and 7.25 eV.³⁶ Based on the above data, the energy band alignment of the Ag₂Te CQD heterojunction before contact can be plotted as shown in Fig. 2b. The values of E_V and E_C of the Ag₂Te CQD film are located within the bandgap of ZnO, forming a typical type-I energy band alignment. Fig. 2c shows the current density–voltage (J–V) curves of the Ag₂Te CQD photodetector in the dark and for different wavelengths. It is worth noting that when the reverse bias voltage exceeds −0.2 V, trap-assisted tunneling and Ohmic leakage current become the dominant leakage current mechanism. On the one hand, trap states significantly promote carrier tunneling under strong electric fields, leading to an exponential increase in current.⁴⁸ On the other hand, the Ohmic leakage current caused by pinholes, cracks, or interface defects in the film also significantly increases with the increase of reverse bias voltage. In addition, fabrication of too thin a film can lead to insufficient depletion of the device, resulting in excessively high local electric field strength. This not only further exacerbates the trap-assisted tunneling effect, but also reduces the bias stability of the device, making it more prone to early soft breakdown at higher biases, resulting in a sharp increase in leakage current.⁴⁹ In addition, the photodetector exhibits rectification characteristics under the built-in electric field between ZnO and Ag₂Te CQDs. The built-in electric field between ZnO and Ag₂Te CQDs is beneficial for reducing the recombination of photogenerated carriers, lowering the dark current of the device, and achieving a self-powered photodiode under 0 V bias. The device exhibits significant photocurrent values under radiation of different wavelengths, indicating that the Ag₂Te CQD photodetector can achieve broadband photoresponse. Fig. 2d shows the EQE curve of the photodetector, and Fig. S7 shows a partial magnification image. It can be seen that there is a maximum EQE of 17.48% at 440 nm, 3.41% at 820 nm, and 0.90% at 1000 nm. Generally speaking, the photoresponse spectrum of the photodetector is closely related to the absorption spectrum of the CQD solution. However, it can be seen in Fig. 1b that the absorption peak of the Ag₂Te CQD solution is at 1018 nm, while the EQE curve produces a local peak at 820 nm and is very low in the range of 1000–1400 nm. The possible reasons include the following three aspects. (1) The aggregation and close packing of CQDs caused by spin coating and thermal annealing processes during the transition of CQDs from solution state to solid thin film. The phenomenon can cause dipole coupling, energy transfer, and increased surface defects between adjacent CQDs, ultimately leading to a significant broadening of the thin-film absorption spectrum, thereby endowing the device with broadband optical response characteristics.¹³ (2) We conducted performance tests on the photodetector using a xenon lamp light source, and its output optical power density varied significantly and unevenly with wavelength (Table S1).¹⁶ Therefore, the local peak formed on the EQE curve may be the result of the combined action of the inherent photophysical properties of the CQD film and the spectral distribution of the test light source. (3) We used a thick ITO conductive glass substrate (thickness of 1.1 mm) in the device. Although ITO has high optical transmittance in the visible-light region, it will significantly decrease in the near-infrared region. The incident near-infrared photons are strongly absorbed by the thick ITO substrate, resulting in a significant reduction in the number of effective photons that ultimately reach the Ag₂Te CQD active layer and participate in photoelectric conversion, leading to an extremely low EQE response in this band.

Fig. 2 (a) A schematic diagram of the structure of the Ag₂Te CQD photodetector. (b) A band diagram of the Ag₂Te CQD photodetector. (c) J–V curves of the Ag₂Te CQD photodetector in the dark and at different wavelengths. (d) The EQE curve of the Ag₂Te CQD photodetector.

As an important indicator of photodetectors, the LDR reflects the linear ratio between photocurrent and light intensity, which can be calculated by eqn (1), where P_max and P_min correspond to the maximum and minimum optical power in the linear photocurrent region, respectively.¹⁶ As shown in Fig. 3a, the Ag₂Te CQD photodetector exhibits excellent linear response characteristics, maintaining a good linearity relationship between photocurrent and incident power over a wide dynamic range of over 74 dB. To comprehensively characterize the performance of the device, we further investigate its frequency response characteristics. Fig. 3b shows that the −3 dB bandwidth of the Ag₂Te CQD photodetector reaches 211 kHz, confirming its excellent high-frequency operational stability. In terms of dynamic response characteristics, the response time (τ_rise) of a photodetector is defined as the time required for the output signal of the detector to rise to a stable value of 90% after the light signal is irradiated, reflecting its perception speed of the light signal. The recovery time (τ_fall) is defined as the time it takes for the output signal to drop to a stable value of 10% after the light signal stops illuminating, reflecting the speed at which the detector returns to its initial state.²⁷ In the Ag₂Te CQD photodetector, the strong built-in electric field formed at the interface of the ZnO/Ag₂Te heterojunction significantly improves the carrier migration efficiency by effectively driving the separation and transport of photogenerated carriers. This characteristic is validated in time-domain response testing (Fig. 3c), where the device exhibited a fast τ_rise of 13.3 μs and τ_fall of 12.4 μs, confirming that the heterojunction interface achieved efficient separation and extraction of photogenerated carriers. It is worth noting that the relationship between the measured response time and the 3 dB bandwidth deviates significantly from the empirical formula of f_3dB = 0.35/τ_rise. This difference is mainly due to the fact that Ag₂Te CQD photodetectors are not ideal first-order RC low-pass filter systems, and their frequency response characteristics are influenced by multiple physical mechanisms, including trap state distribution, surface state effects, and carrier recombination dynamics, among other complex factors.⁵⁰Fig. 3d shows the photocurrent response and recovery curves of the Ag₂Te CQD photodetector at different wavelengths at 0 V bias, indicating that the Ag₂Te CQD photodetector can achieve self-powered mode.^51–53 Generally speaking, the photoresponse spectrum of the photodetector is closely related to the absorption spectrum of the CQD solution. However, it can be seen in Fig. 1b that the absorption spectrum of Ag₂Te CQD solution does not completely match the photoresponse spectrum of the photodetector. The possible reasons include the following two aspects. (1) The aggregation and close packing of CQDs caused by spin coating and thermal annealing processes during the transition of CQDs from solution state to solid thin film. The phenomenon can cause dipole coupling, energy transfer, and increased surface defects between adjacent CQDs, ultimately leading to a significant broadening of the thin-film absorption spectrum, thereby endowing the device with broadband optical response characteristics.¹³ (2) We conducted performance tests on the photodetector using a xenon lamp light source, and its output optical power density varied significantly and unevenly with wavelength (Table S1).¹⁶ Therefore, the response spectrum of the photodetector is the result of the combined action of the inherent photophysical properties of the CQD film and the spectral distribution of the test light source. Furthermore, under different periodic illumination, the photocurrent of the Ag₂Te CQD photodetector exhibits good repeatability (Fig. S8). The values of R and D* and partial magnification images of the Ag₂Te CQD photodetector under different wavelengths of illumination are shown in Fig. 3e, 3f, and Fig. S9. At a 0 V bias and 1000 nm, R is 7.22 mA W⁻¹ and D* is 2.81 × 10¹⁰ Jones, indicating that the Ag₂Te CQD photodetector has a good response to incident light. LDR, R, and D* are measured and calculated at 0 V bias, following the standard of reported self-powered photodetectors.^51–53 Therefore, eqn (1)–(5) are used for calculations.

where I_Light is the photocurrent, I_Dark is the dark current, P_in is the radiation intensity, A (2.4 mm²) is the effective area of the device, e is the charge of the electron (1.69 × 10⁻¹⁹ C), and a semiconductor characterization system (PD-QE, Enli Technology) was used to analyze the noise frequency of the photodetector and directly measure the noise current (I_N). Then, the noise equivalent power (NEP) and D* are calculated by eqn (4) and (5) respectively.^54–57 To evaluate the repeatability and stability of the Ag₂Te CQD photodetector, we conducted further tests as shown in Fig. 3g. After 200 cycles of illumination under 900 nm at 0 V bias, the photoresponse of the photodetector remains at its initial value without significant degradation and has a high switching ratio, indicating a good repeatability.⁵⁸ The good cyclic stability results obtained in this study provide preliminary experimental data support for future research on long-term operational stability and environmental adaptability, which is of great significance for promoting the practical application of the non-toxic Ag₂Te CQD photodetector.

Fig. 3 (a) LDR, (b) response bandwidth, and (c) transient response of the Ag₂Te CQD photodetector under 520 nm at 0 V bias. (d) Photocurrent response and recovery curves of the Ag₂Te CQD photodetector under different wavelengths at 0 V bias. (e) R and (f) D* of the Ag₂Te CQD photodetector under different wavelengths. (g) Repeated response/recovery curves over 200 cycles for the Ag₂Te CQD photodetector under 900 nm at 0 V bias.

Compared with recently reported Ag₂Te CQD photodetectors (Table 1),^{23,24,26–30} our photodetector exhibits excellent comprehensive performances. For example, compared with ref. 23 and 24, our photodetector has a higher EQE. Compared with ref. 24 and 27, it has a higher specific detectivity. In terms of −3 dB bandwidth, our photodetector performs significantly better than most of those reported^24,26,27 (some references do not even report this parameter). In addition, our photodetector is significantly superior to that of ref. 27 in terms of the two key parameters of specific detectivity and response time. More importantly, our proposed surface modification strategy of the bifunctional thiol ligands based on nucleation mechanism not only effectively solves the problem of easy aggregation of Ag₂Te CQDs in solution, but also significantly simplifies the fabrication process without relying on common doping treatments, core–shell structures, or complex device processing techniques. The strategy provides new ideas for the development of low-cost and scalable photodetectors, and has important scientific significance and application value.

Table 1 Performance comparisons of various Ag₂Te CQD-based photodetectors

Material	Main strategy	EQE (%)	D* (Jones)	Response and recovery time (μs)	−3 dB bandwidth (MHz)	LDR (dB)	Ref.
Ag2Te	Optimization of synthesis parameters	0.14@1400 nm	—	—	—	—	23
Ag2Te	Solid-phase ligand exchange	0.2@1550 nm	10^7	—/0.072	0.1	78.8	24
Ag2Te	Phosphine-free and buffer layer	30@1400 nm	3 × 10^12	1.3/3.3	0.11	118	26
Ag2Te	Zinc halide-controlled synthesis	30@1550 nm	1.4 × 10^8	97/196	0.0185	93	27
Ag2Te	Surface engineering and doping control	30@1450 nm	10^11	0.025/—	>5	150	28
Ag2Te	Oleylamine passivation	16@1700 nm	9 × 10^10	—/2.7	—	—	29
Ag2Te	Core–shell structure	8.4@1550 nm	1.1 × 10^11	—/0.038	—	63.8	30
Ag2Te	Surface modification strategy based on bifunctional thiol ligands	0.54@1400 nm	8.9 × 10^10	13.3/12.4	0.211	74	This work

---

3.3. Application demonstration of non-contact transmission pulse monitoring

The pulse is one of the important physiological signs of the human body, which has significant importance in clinical diagnosis and exercise monitoring.^59–68 However, the traditional pressure sensors often require prolonged compression of the skin when measuring pulse, which may cause tissue congestion or nerve compression damage, especially for infants, burn patients, or people with sensitive skin.⁵⁹ Therefore, non-contact pulse monitoring technology is gradually receiving attention because it does not require physical contact and can avoid the risk of skin damage caused by friction or pressure.^60–63 The non-contact pulse monitoring methods mainly include reflective and transmissive methods. Although reflective sensors are easy to apply, they are susceptible to interference from motion artifacts, such as handshaking that may cause optical path deviation, requiring complex algorithms to compensate for signal distortion.^64,65 In a transmissive architecture, the light source and detector are positioned opposite each other, forming a relatively stable optical path. Combined with the high penetration of near-infrared light, this design can effectively suppress signal attenuation caused by motion. In addition, the transmissive optical signals can also carry multiple vascular-related information such as blood volume waves (BVP), blood oxygen saturation (SpO₂), and vascular-elasticity, providing richer and more accurate data for clinical and exercise monitoring.^60–63

Currently, photoplethysmography (PPG) is one of the important methods for non-contact pulse monitoring, and its principle is shown in Fig. 4a. When a light source penetrates the finger, the oxygenated hemoglobin in arterial blood and reduced hemoglobin in venous blood selectively absorb specific wavelengths of light based on their molecular properties.^69,70 The transmitted light modulated by the absorption of human tissue is captured by the photodetector to collect and display pulse signals in real time. Thanks to the fast response speed (13.3 μs), good repeatability, linear dynamic range greater than 74 dB, and 3 dB bandwidth up to 211 kHz of the Ag₂Te CQD-based photodetector, we integrated it into a non-contact transmitted light pulse monitoring system. When the light penetrates human tissue, the strong absorption characteristics of the Ag₂Te CQDs enable them to effectively capture signals of microvascular volume changes caused by pulse pulsation. Fig. 4b shows a real scene using the Ag₂Te CQD photodetector for non-contact transmission pulse monitoring. Red light (620 nm) and near-infrared light (940 nm) were passed through the finger, and the light passing through the finger was received by the Ag₂Te CQD photodetector. Fig. 4c shows the pulse signals using the PPG method at different wavelengths, and the typical systolic and diastolic peaks of the pulse signals can be clearly identified. When the heart is systolic, arterial blood pressure rises, blood vessels dilate, and the amount of blood flowing into capillaries instantly increases, thereby absorbing more light and weakening the light signal received by the detector. On the contrary, when the heart is diastolic, blood pressure decreases, blood vessels retract, blood in capillaries decreases, and the absorbed light decreases, resulting in an enhanced light signal received by the detector. At a wavelength of 620 nm, heart rates are 84 beats min⁻¹ at rest, 107 beats min⁻¹ after 15 squats, and 114 beats min⁻¹ after 20 squats. At a wavelength of 940 nm, the heart rates are 85 beats min⁻¹ at rest, 108 beats min⁻¹ after 15 squats, and 113 beats min⁻¹ after 20 squats. The calculated pulse results are consistent with typical physiological characteristics, indicating the correctness and reliability of our Ag₂Te CQD photodetector.

Fig. 4 (a) A schematic diagram of non-contact transmission pulse monitoring using the Ag₂Te CQD photodetector. (b) An image of a real instance of non-contact transmission pulse monitoring. (c) Pulse signals measured by the Ag₂Te CQD photodetector at 620 and 940 nm under normal conditions, and after 15 squats and 20 squats.

4. Conclusions

In conclusion, we proposed a surface modification strategy based on bifunctional thiol ligands. Thiol was selected as the only ligand on the surface to synthesize Ag₂Te CQDs. The strong binding force between thiol and Ag makes it form a dense adsorption layer on the surface of the CQDs, which can not only passivate the surface defects but also reduce the agglomeration tendency caused by surface charge interactions or van der Waals forces. At the same time, based on reaction kinetics regulation, ODE was used as the synthesis solvent to reduce the activity of the precursor reaction, and then the growth rate of the CQDs was controlled. The photodetector based on the Ag₂Te CQDs synthesized by this strategy can achieve a D* value of 10¹⁰ Jones at room temperature, the −3 dB bandwidth reaches 211 kHz, and the fast response time is 13.3 μs. Benefitting from excellent photoelectric response performance, we further applied the Ag₂Te CQD-based photodetector to non-contact transmission pulse monitoring, effectively achieving the real-time detection of human pulse signals. This work provides a reference for the development of Ag₂Te CQD photodetectors. In future research, we and peer researchers need to establish the correlation between CQD stability and device stability through experiments, further systematically investigate its long-term stability issues, and conduct optimization studies around packaging strategies to significantly improve the performance stability of the Ag₂Te CQD photodetector during long-term operation and under harsh environmental conditions.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information is available. See DOI: https://doi.org/10.1039/d5tc02959g.

Acknowledgements

This work is supported by the National Science Fund for Distinguished Young Scholars (grant no. 62225106), Natural Science Foundation of China (grant no. U24A20229), and Sichuan Innovation Research Group Project (grant no. 2025NSFTD0008). The authors are grateful to Xuan Wei from Analysis and Testing Center, University of Electronic Science and Technology of China, for technical support.

References

1. H. Yang, Z. Ma and Q. Wang, ACS Nano, 2024, 18, 30123–30131.
2. X. Tang, M. M. Ackerman, M. Chen and P. Guyot-Sionnest, Nat. Photonics, 2019, 13, 277–282.
3. W. Zeng, Z. Duan, Y. Bu, X. Tang, J. Yang, Z. Yuan, Y. Jiang and H. Tai, Laser Photonics Rev., 2025, 19, 2402163.
4. D. Naczynski, M. Tan, M. Zevon, B. Wall, J. Kohl, A. Kulesa, S. Chen, C. Roth, R. Riman and P. Moghe, Nat. Commun., 2013, 4, 2199.
5. J. Chen, N. Chen, Z. Wang, R. Dou, J. Liu, N. Wu, L. Liu, P. Feng and G. Wang, Chips, 2025, 4, 8.
6. W. Wei, Q. Li, Z. Xu, J. Chen, L. Chen, H. Guo, P. Wang and W. Hu, IEEE Trans. Electron Devices, 2025, 72, 1255–1258.
7. V. Letka, M. Martin, N. Massara, C. Leroux, R. Templier, C. Licitra, J. Richy and T. Baron, ACS Photonics, 2023, 10, 3266–3274.
8. K. Xu, W. Zhou and Z. Ning, Small, 2020, 16, 2003397.
9. V. Pejovic, E. Georgitzikis, J. Lee, I. Lieberman, D. Cheyns, P. Heremans and P. Malinowski, IEEE Trans. Electron Devices, 2022, 69, 2840–2850.
10. Z. Wu, Y. Ou, M. Cai, Y. Wang, R. Tang and Y. Xia, Adv. Opt. Mater., 2023, 11, 2201577.
11. V. Pejović, E. Georgitzikis, I. Lieberman, P. Malinowski, P. Heremans and D. Cheyns, Adv. Funct. Mater., 2022, 32, 2201424.
12. H. Wang, J. Pinna, D. Romero, L. Mario, R. Koushki, M. Kot, G. Portale and M. Loi, Adv. Mater., 2024, 36, 2311526.
13. M. Yu, R. Huang, X. Zhang, X. Hu, X. Ran, H. Xia, M. Yuan, J. Zhang, L. Gao, J. Tang and X. Lan, Adv. Funct. Mater., 2025, 35, 2423299.
14. J. Yang, H. Hu, Y. Lv, M. Yuan, B. Wang, Z. He, S. Chen, Y. Wang, Z. Hu, M. Yu, X. Zhang, J. He, J. Zhang, H. Liu, H. Hsu, J. Tang, H. Song and X. Lan, Nano Lett., 2022, 22, 3465–3472.
15. H. Hu, J. Liu, J. Liu, M. Yuan, H. Ma, B. Wang, Y. Wang, H. Xia, J. Yang, L. Gao, J. Zhang, J. Tang and X. Lan, ACS Nano, 2025, 19, 8974–8984.
16. Y. Bu, Y. Jiang, Z. Duan, W. Zeng, X. Tang, J. Yang, Z. Yuan and H. Tai, Adv. Opt. Mater., 2025, 13, 2500295.
17. W. Zeng, Z. Duan, Y. Bu, H. Yu, C. Wang, X. Tang, J. Yang, Z. Yuan, Y. Jiang and H. Tai, Mater. Horiz., 2025, 12, 2179–2186.
18. B. Wang, M. Yuan, J. Liu, X. Zhang, J. Liu, J. Yang, L. Gao, J. Zhang, J. Tang and X. Lan, Nano Lett., 2024, 24, 9583–9590.
19. K. Sergeeva, S. Hu, A. Sokolova, A. Portniagin, D. Chen, S. Kershaw and A. Rogach, Adv. Mater., 2024, 36, 2306518.
20. H. Wu, Z. Liu, B. Wang, L. Zheng, S. Lian, J. Zhang, S. Zhang, G. Zhang, Z. Xue, S. Yang, X. Cheng, G. Ding, Z. Liu, C. Ye and G. Wang, ACS Photonics, 2024, 11, 1342.
21. S. Zhang, G. Zhang, L. Zheng, Z. Liu, B. Wang, H. Wu, Z. He, Z. Jin and G. Wang, IEEE Electron Device Lett., 2023, 44, 1240.
22. Z. Liu, H. Wu, L. Zheng, W. Zhou, S. Zhang, G. Zhang, G. Wang, Z. Ning and Y. Yu, IEEE Trans. Electron Devices, 2024, 71, 4368.
23. J. Ouyang, N. Graddage, J. Lu, Y. Zhong, T. Chu, Y. Zhang, X. Wu, O. Kodra, Z. Li, Y. Tao and J. Ding, ACS Appl. Nano Mater., 2021, 4, 13587–13601.
24. Y. Ahn, S. Eom, G. Kim, J. Lee, B. Kim, D. Kim, M. Si, M. Yang, Y. Jung, B. Kim, Y. Chung, K. Jeong and S. Baek, Adv. Sci., 2024, 11, 2407453.
25. Z. Liu, A. Liu, H. Fu, Q. Cheng, M. Zhang, M. Pan, L. Liu, M. Luo, B. Tang, W. Zhao, J. Kong, X. Shao and D. Pang, J. Am. Chem. Soc., 2021, 143, 12867–12877.
26. Y. Wang, L. Peng, J. Schreier, Y. Bi, A. Black, A. Malla, S. Goossens and G. Konstantatos, Nat. Photonics, 2024, 18, 236–242.
27. D. Yuan, F. Cao, Z. Han, M. Liu, X. Deng, L. Zhang, X. Liu, S. Cao, X. Liu, Z. Yang, Z. Deng, Y. Li, O. Voznyy, Q. Fan, B. Sun and W. Huang, Nano Energy, 2025, 140, 111026.
28. Y. Wang, H. Wu, C. Roda, L. Peng, N. Taghipour, M. Dosil and G. Konstantatos, Adv. Mater., 2025, 2500977.
29. G. Kim, D. Choi, S. Eom, E. Jung, J. Lee, B. Rehl, S. Kim, S. Hoogland, E. Sargent and K. Jeong, ACS Mater. Lett., 2024, 6, 4988–4996.
30. J. Lee, I. Lee, D. Kang, N. Kim, J. Kim, S. Baek and Y. Kim, Small, 2025, 2412420.
31. Z. Liu, W. Zhao, L. Chen, Y. Chen, Z. Wang, A. Liu and D. Pang, Nano Res., 2024, 17, 10585–10606.
32. H. Pi, X. Li, L. Xie, Y. Cui, L. Jing, B. Chen, Y. Wang and A. Rogach, Adv. Opt. Mater., 2025, 13, 2403489.
33. J. Roo, M. Ibanez, P. Geiregat, G. Nedelcu, W. Walravens, J. Maes, J. Martins, I. Driessche, M. Kovalenko and Z. Hens, ACS Nano, 2016, 10, 2071–2081.
34. L. Trizio, I. Infante and L. Manna, Acc. Chem. Res., 2023, 56, 1815–1825.
35. S. Gallagher, J. Kline, F. Jahanbakhshi, J. Sadighian, I. Lyons, G. Shen, B. Hammel, S. Yazdi, G. Dukovic, A. Rappe and D. Ginger, ACS Nano, 2024, 18, 19208–19219.
36. Q. Liu, L. Li, J. Wu, Y. Wang, L. Yuan, Z. Jiang, J. Xiao, D. Gu, W. Li, H. Tai and Y. Jiang, Nat. Commun., 2023, 14, 6935.
37. C. Murray, D. Norris and M. Bawendi, J. Am. Chem. Soc., 1993, 115, 8706–8715.
38. X. Peng, L. Manna, W. Yang, J. Wickham, E. Scher, A. Kadavanich and A. Alivisatos, Nature, 2000, 404, 59–61.
39. E. Shirzadi, Q. Jin, A. Zeraati, R. Dorakhan, T. Goncalves, J. Abed, B. Lee, A. Rasouli, J. Wicks, J. Zhang, P. Ou, V. Boureau, S. Park, W. Ni, G. Lee, C. Tian, D. Meira, D. Sinton, S. Siahrostami and E. Sargent, Nat. Commun., 2024, 15, 2995.
40. J. Chen, X. Liu, S. Xi, T. Zhang, Z. Liu, J. Chen, L. Shen, S. Kawi and L. Wang, ACS Nano, 2022, 16, 13982–13991.
41. W. Xu, E. Wang and X. Zeng, Acc. Mater. Res., 2024, 5, 1134–1145.
42. S. Jia, M. Hu, M. Gu, J. Ma, D. Li, G. Xiang, P. Liu, K. Wang, P. Servati, W. Ge and X. Sun, Small, 2024, 20, 2307298.
43. X. Deng, S. Zheng, Y. Zhong, J. Hu, L. Chung and J. He, Coord. Chem. Rev., 2022, 450, 214235.
44. J. Li, C. Jin, R. Jiang, J. Su, T. Tian, C. Yin, J. Meng, Z. Kou, S. Bai, P. Müller-Buschbaum, F. Huang, L. Mai, Y. Cheng and T. Bu, Nat. Energy, 2024, 9, 1540–1550.
45. X. Wang, S. Liu, H. Wang and J. Luo, Adv. Funct. Mater., 2024, 34, 2403104.
46. J. Zhang, J. Wu, S. Langner, B. Zhao, Z. Xie, J. Hauch, H. Afify, A. Barabash, J. Luo, M. Sytnyk, W. Meng, K. Zhang, C. Liu, A. Osvet, N. Li, M. Halik, W. Heiss, Y. Zhao and C. Brabec, Adv. Funct. Mater., 2022, 32, 2207101.
47. L. Wu, Y. Ji, H. Dan, C. Bowen and Y. Yang, InfoMat, 2023, 5, 12414.
48. P. Liu, S. Lu, J. Liu, B. Xia, G. Yang, M. Ke, X. Zhao, J. Yang, Y. Liu, C. Ge, G. Liang, W. Chen, X. Lan, J. Zhang, L. Gao and J. Tang, InfoMat, 2024, 6, 12497.
49. Y. Yan, H. Liu, L. Bian, Y. Dai, B. Zhang, S. Xue, Y. Zhou, J. Xu and S. Wang, Mater. Horiz., 2024, 11, 6192–6221.
50. C. Schedel, F. Strauß and M. Scheele, J. Phys. Chem. C, 2022, 126, 14011–14016.
51. M. Sulaman, S. Yang, A. Bukhtiar, P. Tang, Z. Zhang, Y. Song, A. Imran, Y. Jiang, Y. Cui, L. Tang and B. Zou, Adv. Funct. Mater., 2022, 32, 2201527.
52. S. Kim, D. Lee, S. Moon, J. Choi, D. Kim, J. Kim and S. Baek, Adv. Funct. Mater., 2023, 33, 2303778.
53. T. Yan, Z. Li, F. Cao, J. Chen, L. Wu and X. Fang, Adv. Mater., 2022, 34, 2201303.
54. P. Guyot-Sionnest, M. Ackerman and X. Tang, J. Chem. Phys., 2019, 151, 060901.
55. S. Jee, M. Si, J. Choi, D. Kim, C. Kim, D. Yang and S. Baek, Adv. Opt. Mater., 2024, 12, 2303097.
56. M. Si, S. Jee, M. Yang, D. Kim, Y. Ahn, S. Lee, C. Kim, I. Bae and S. Baek, Adv. Sci., 2024, 11, 2306798.
57. M. Si, D. Kim, S. Jeong, M. Yang, J. Kim, S. Lee, I. Lee, J. Jeong, B. Kim, T. Han, B. Kim, Y. Ahn, S. Jee, Y. Jung and S. Baek, Korean J. Chem. Eng., 2024, 41, 3561–3572.
58. Y. Zhang, J. Jiang, Z. Zhang, H. Yu, Y. Lian, C. Han, X. Liu, J. Han, H. Zhou, X. Dong, J. Gou, Z. Wu and J. Wang, J. Mater. Chem. C, 2024, 12, 16714–16721.
59. T. Yokota, P. Zalar, M. Kaltenbrunner, H. Jinno, N. Matsuhisa, H. Kitanosako, Y. Tachibana, W. Yukita, M. Koizumi and T. Someya, Sci. Adv., 2016, 2, 1501856.
60. M. Villarroel, S. Chaichulee, J. Jorge, S. Davis, G. Green, C. Arteta, A. Zisserman, K. McCormick, P. Watkinson and L. Tarassenko, npj Digit. Med., 2019, 2, 128.
61. J. Jorge, M. Villarroel, H. Tomlinson, O. Gibson, J. Darbyshire, J. Ede, M. Harford, J. Young, L. Tarassenko and P. Watkinson, npj Digit. Med., 2022, 5, 4.
62. Z. He, J. Han, X. Du, L. Cao, J. Wang, C. Zheng, H. Lin and S. Tao, Adv. Funct. Mater., 2021, 31, 2103988.
63. J. Huang, J. Lee, J. Vollbrecht, V. Brus, A. Dixon, D. Cao, Z. Zhu, Z. Du, H. Wang, K. Cho, G. Bazan and T. Nguyen, Adv. Mater., 2020, 32, 1906027.
64. H. Jia, Y. Gao, J. Zhou, J. Li, C. Yiu, W. Park, Z. Yang and X. Yu, Nano Energy, 2024, 127, 109796.
65. J. Kim, G. Salvatore, H. Araki, A. Chiarelli, Z. Xie, A. Banks, X. Sheng, Y. Liu, J. Lee, K. Jang, S. Heo, K. Cho, H. Luo, B. Zimmerman, J. Kim, L. Yan, X. Feng, S. Xu, M. Fabiani, G. Gratton, Y. Huang, U. Paik and J. Rogers, Sci. Adv., 2016, 2, 1600418.
66. X. Feng, Y. Ran, X. Li, H. Xu, Q. Huang, Z. Duan, Z. Yuan, Y. Jiang and H. Tai, Mater. Chem. Phys., 2024, 321, 129489.
67. C. Vlachopoulos, K. Aznaouridis and C. Stefanadis, J. Am. Coll. Cardiol., 2010, 55, 1318–1327.
68. Z. Duan, Y. Jiang, Q. Huang, Z. Yuan, Q. Zhao, S. Wang, Y. Zhang and H. Tai, J. Mater. Chem. C, 2021, 9, 13659–13667.
69. H. Xu, J. Liu, J. Zhang, G. Zhou, N. Luo and N. Zhao, Adv. Mater., 2017, 29, 1700975.
70. J. Allen, Physiol. Meas., 2007, 28, 1–39.

---