Room-Temperature High-Detectivity Flexible Near-Infrared Photodetectors with Chalcogenide Silver Telluride Nanoparticles.

Herein, flexible near-infrared (NIR) photodetectors were prepared using silver telluride (Ag5Te3) nanoparticles (NPs) for optoelectronic applications. For the main channel materials of the photodetectors, Ag5Te3 NPs were used, which were synthesized in an aqueous solution. Moreover, Ag5Te3 thin films were successfully fabricated on plastic substrates at 150 °C using redistributed Ag5Te3 NPs in aqueous inks. The crystal structure, chemistry, and optoelectronic properties of the synthesized photodetectors were studied. The fabricated flexible Ag5Te3-based photodetectors achieved a detectivity of 6.27 × 109 cm Hz1/2 W-1 (>109) at room temperature under ∼0.35% compressive and tensile strains. The obtained detectivity value exceeds those of two-dimensional inorganic layered material phototransistors-such as MoS2-or commercial thermistor bolometers at room temperature (∼109). Furthermore, the proposed novel method for the synthesis of Ag5Te3 thin films on plastic substrates can be applied to other Ag5Te3-based applications in the future.

Introduction

High-performance near-infrared (NIR) and mid-IR sensors are crucial for image sensing, wireless communication, as well as night vision, security, and bioimaging systems.[1−6] Until now, high-quality low-band gap semiconductors—such as InGaAs, HgTe, and PbS—grown using the epitaxy method have been used as the main channel materials for sensors.[7−9] Although the cutoff wavelength of such sensors can be tuned by controlling the stoichiometry and composition of the grown material, they usually contain heavy metals that are toxic to the environment and human health. Furthermore, these sensors show high detectivity at an extremely low temperature of 4.2 K and high operating voltages. The operation of such sensing systems involves expensive and complex cooling arrangements, hindering their commercialization for applications involving portable and room-temperature operations. Moreover, the epitaxial growth technique is used to fabricate the aforementioned semiconductor films, based on which low-cost large-area applications are challenging. To solve this issue, the solution-processed sol–gel method is used to synthesize high-quality low-band gap semiconductor layers—such as CuO—for NIR sensors.[10] Unfortunately, to convert precursors into high-quality CuO, high-temperature postannealing processes are typically required above 400 °C, higher than the deformation temperature of plastic substrates for flexible NIR sensors.

Bulk silver telluride (Ag2Te) is a technically important candidate for next-generation electronic, optoelectronic, and energy harvesting applications. Ag2Te exhibits high ionic conductivity at room temperature with a very narrow optical band gap (Eg = 0.025 eV), low thermal conductivity, and excellent thermoelectric properties.[11−14]

Notably, the melting temperature of a material, which is related to the annealing temperature, decreases as the material size decreases. For example, the sintering temperatures of nanoparticles (NPs) are significantly lower than those of bulk materials.[15] This trend suggests that high-quality thin films can be successfully formed on plastic substrates for flexible electrical and optical applications. Furthermore, stable NPs show great potential to be used as inks in solutions for solution processes—such as the spin casting or printing process—for large-area applications.[16−21]

Herein, hexagonal-phase Ag5Te3 NPs were synthesized facilely in an aqueous solution without any gas-phase precursor, such as H2Te. High crystalline Ag5Te3 films were fabricated and deposited using redistributed Ag5Te3 NPs via an annealing process at 150 °C in air. The structural, chemical, optical, and optoelectronic properties of the Ag5Te3 NP-based thin films were studied as functions of postannealing temperatures. For the first time, flexible Ag5Te3 NP-based NIR photodetectors were successfully formed on poly (ether sulfone) (PES) substrates. The detectivity of the obtained photodetector was 6.27 × 109 cm Hz1/2 W–1 (>109) under 0.38% compressive and tensile stresses.

Results and Discussion

Figure a shows a typical high-resolution transmission electron microscopy (HRTEM) image of the synthesized Ag5Te3 NPs with 0.055 g of the Te precursor (Na2TeO3). The NPs appeared as spheres with a size of ∼4 nm and showed crystallinity and reasonable aggregation. The selected-area electron diffraction (SAED) pattern of the Ag5Te3 NPs revealed rings (Figure b), indicating polycrystallinity. The final products were affected by the Te2– concentration (Supporting Information Figure S1).

Figure 1

(a) HRTEM image and (b) SAED pattern of the synthesized Ag5Te3 NPs.

Figure a shows the grazing incidence X-ray diffraction (GIXRD) patterns of the Ag5Te3 nanoparticles as functions of postannealing temperatures. Notably, monoclinic β-phase and α-phase Ag2Te were stable below and above 418 K, respectively. However, the GIXRD patterns were consistent with the hexagonal-phase Ag5Te3 structure (JCPDS 18-1186 and 16-0372). The Ag5Te3 phase was a low-temperature stoichiometric phase in the Ag–Te system. Stable Ag5Te3 can be prepared using Ag2Te in the presence of excess Te. The broad GIXRD peaks were attributed to the small size of the crystallographic domain. The crystalline size of the Ag5Te3 films as functions of the postannealing temperatures was calculated using the Scherrer equation:where D denotes the crystalline size, λ denotes the CuKα wavelength, β represents the line broadening at half the maximum intensity, and θ represents the peak position. The crystalline sizes calculated using the (414) plane peaks were 2.31 nm for the synthesized film and 2.31, 3.56, and 8.52 nm for the films annealed at 50, 100, and 150 °C, respectively. These results indicate that the crystalline size of the annealed Ag5Te3 films were larger than those of the as-synthesized Ag5Te3 NPs. The Ag5Te3 film annealed at 150 °C exhibited sharp XRD peaks, representing Ag (JCPDS 04-0783), possibly ascribed to the redox reaction of AgNO3. The final films were a mixture of Ag and Ag5Te3.

Figure 2

(a) GIXRD spectra and (b) calculated crystalline sizes of the as-deposited Ag5Te3 film and films annealed for 1 h at 50, 100, and 150 °C.

Furthermore, the elemental composition of the Ag5Te3 films was confirmed using X-ray photoelectron spectroscopy (XPS) analysis (Supporting Information Figures S2 and S3), where the binding energy values agreed with the previously reported values. The peaks at 368.3 and 374.3 eV corresponded to Ag 3d5/2 and Ag 3d3/2, while those at 572.5 and 582.9 eV corresponded to Te 3d5/2 and Te 3d3/2, respectively. The peaks of Ag 3d at 367.2 and 373.2 eV represented the binding energies of Ag 3d5/2 and Ag 3d3/2, respectively, of metallic Ag0. Furthermore, the peaks at 367.6 and 373.6 eV were ascribed to the binding energies of Ag 3d5/2 and Ag 3d3/2, respectively, of Ag+ in Ag5Te3. This indicates that the final films were a mixture of Ag and Ag5Te3; these results are consistent with the GIXRD data. No peak was detected at 575.2 eV—corresponding to Te(IV)—in this system, regardless of the postannealing temperature.

Figure S4a shows the ultraviolet (UV) absorbance spectra of the Ag5Te3-based film and those annealed at different temperatures. The Eg of the Ag5Te3 films as a function of the postannealing temperatures was evaluated using the Davis–Mott and Tauc models:where hν represents the photon energy, α denotes the absorption coefficient and A is a constant, and Eg denotes the optical band gap of the semiconductor. The exponent n depends on the type of the transition (n = 2 for indirect allowed transition). α is obtained from eq , where A denotes the absorbance of the films and t represents the film thickness. Figure S4b shows a plot of (αhν)1/2 vs hν for the Ag5Te3 thin films as functions of the postannealing temperature. The Eg of the Ag5Te3 thin films was estimated to be 1.48 eV, larger than that of bulk Ag5Te3.[22] The reduced size of Ag5Te3 affects its physical properties because of the energy quantization phenomena.[23]

Figure a shows the optical image of the fabricated flexible Ag5Te3-based photodetectors, and the inset shows the synthesized Ag5Te3 NPs in aqueous solutions. Figure b shows the current–voltage (I–V) curves with/without 808 nm laser illumination. When the Ag5Te3 films were exposed to 808 nm laser, the current increased. Under illumination, the absorbed photons—with energy larger than the Eg of Ag5Te3 (1.48 eV)—generated electron–hole pairs. The generated electron–hole pairs increased the film conductivity. Moreover, the I–V curves show that the photocurrent values were proportional to the biased voltages. Once the laser illumination was turned off, the current decreased. Figure S5 shows the double sweep photocurrent–V curves, indicating electrochemical properties or resistive switching memory properties due to Ag clusters formed during the synthesis.[20,24]

Figure 3

(a) Optical image of the fabricated 150 °C-annealed Ag5Te3 flexible photodetectors; the inset shows the optical image of Ag5Te3-based photodetectors and synthesized Ag5Te3 NP solutions. (b) I–V curves under tensile and compressive strains; the inset shows a schematic of Ag5Te3-based photodetectors.

The performance of the photodetectors was investigated under mechanical stress. The I–V curves with/without illumination were obtained using the flexible Ag5Te3-based photodetectors when the substrates were bent until the bending radius of the substrate reached 1.6 cm, corresponding to 0.38% compressive and tensile stresses. The strain values were calculated as follows:where Rc denotes the bending curvature radius of the PES substrates. The performance of the device was retained after 100 cycles in the bending test. The bending direction was parallel to the channel. When the substrate was subjected to 0.38% tensile strain, the photocurrent of the film decreased compared with the case of a flat substrate. Alternatively, under 0.38% compressive strain, the photocurrent of the substrate increased. The change in the photocurrent was attributed to the change in the distance between the neighboring Ag5Te3 NPs in the film owing to the applied strain.[17]

Four representative parameters—responsivity (R), gain (G), detectivity (D*), and photosensitivity (S)—were used to evaluate the performance of the Ag5Te3-based flexible photodetectors. They are calculated using the following equations:where P0 denotes the incident power of the light source, Il denotes the photocurrent, Id represents the dark current, η denotes the quantum efficiency, h is the Planck’s constant, c denotes the light speed, q represents the electron charge, λ represents the wavelength of the light source, Nph denotes the number of absorbed photons per unit time, Nel denotes the number of electrons collected per unit time, τr denotes the carrier transit time, and τ represents the carrier lifetime. The absorbed incident light power obtained using Ra = Pin(1 – e–d), where α denotes the absorption coefficient and d is the distance from the surfaces.[25] If the αd value was considerably smaller than 1, then Ra = Pinαd. The film thickness, absorption coefficient, and actual absorbed incident power were ∼10 nm, 1.96 × 107 cm–1, and 19.57%, respectively. Figure indicates the calculated four representative parameters based on each bias voltage. Using eqs ,6,7,8, the following values were obtained: (i) for voltage = +0.2 V, the R, G, D*, and S values were 2.68 × 10–4 A/W, 4.11 × 10–4, 6.27 × 109 cm Hz1/2 W–1, and 2.75 × 103%, respectively; (ii) for voltage = +2.0 V, the R, G, D*, and S values were 8.41 × 10–4 A/W, 1.29 × 10–3, 5.19 × 109 cm Hz1/2 W–1, and 6.02 × 102%, respectively; (iii) for voltage = +5.0 V, the R, G, D*, and S values were 1.33 × 10–3 A/W, 2.07 × 10–3, 4.04 × 109 cm Hz1/2 W–1, and 2.28 × 102%, respectively; and (iv) for voltage = +10.0 V, the R, G, D*, and S values were 2.55 × 10–3 A/W, 3.91 × 10–3, 5.89 × 109 cm Hz1/2 W–1, and 2.56 × 102%, respectively. All the calculated parameter values of the Ag5Te3 NIR photodetectors are summarized in Table .

Figure 4

Representative parameters calculated based on each bias voltage (a) R and G and (b) D* and S.

Table 1

Calculated Parameter Values of Ag5Te3 NIR Photodetectors

biased voltage (V)	R (A/W)	G	D* (cm Hz1/2 W–1)	S (%)
+0.2	2.68 × 10–4	4.11 × 10–4	6.27 × 109	2.75 × 103
+2.0	8.41 × 10–4	1.29 × 10–3	5.19 × 109	6.02 × 102
+5.0	1.33 × 10–3	2.07 × 10–3	4.04 × 109	2.28 × 102
+10.0	2.55 × 10–3	3.91 × 10–3	5.89 × 109	2.56 × 102

Table shows a comparison of the performance of the NIR photodetectors prepared using other promising materials. In particular, the D* values of the Ag5Te3 NP-based photodetectors were lower than those of photodetectors prepared using Si (D* ≈ 1013 Jones) but higher than those of phototransistors prepared using MoS2, black arsenic phosphorus, SnTe nanoplates, transparent ferroelectric ceramics, or commercial thermistor bolometers at room temperature (D* ≈ 108 Jones).[25−28]

Table 2

Device Performance of NIR Photodetectors Prepared Using Other Promising Materials

materials	R (mA/W)	G	D* (cm Hz1/2 W–1)	S (%)	illumination laser wavelength (nm)	ref.
MoS2	9 × 10–2		7 × 107		850	(25)
black arsenic phosphorus	30		1.06 × 108		2400	(26)
SnTe nanoplates	698		3.89 × 108		980	(27)
lead lanthanum zirconate titanate	2.78 × 10–4		6.96 × 107		1300	(28)
Ag5Te3 NPs	2.68 × 10–1	4.11 × 10–4	6.27 × 109	2.75 × 103	808	this work

Figure a shows the photoswitching characteristics of the Ag5Te3 NP-based flexible photodetector under 50 s illumination at different bias voltages. Without illumination, the conductivity did not decrease abruptly, showing a persistent conductivity effect. Figure b shows the rise and decay times (τr and τd, respectively) as functions of the biased voltages. The τr and τd as functions of the biased voltages were evaluated using the following exponential equations: I = Io – Io(exp(−t/τr)) and I = Io + A(exp(−t/td)), respectively, and using exponential curve fitting. Based on these equations, the following values were obtained: (i) for voltage = 0.2 V, τr = 12.6 s and τd = 105.2 s; (ii) for voltage = 2.0 V, τr = 6.25 s and τd = 66.7 s; (iii) for voltage = 5.0 V, τr = 5.26 s and τd = 12.5 s; and (iv) for voltage = 10.0 V, τr = 2.8 s and τd = 4.0 s. The obtained τr and τd values decreased as the biased voltage increased. Defects and impurities inside the Ag5Te3 channel could induce an extra energy level, affecting the relaxation behavior. However, the increased biased voltage promoted the easy release of confined carriers from the extra energy level, decreasing the decay constants successfully.

Figure 5

(a) Photoswitching characteristics of the 150 °C-annealed Ag5Te3 photodetector exposed to 808 nm laser corresponding to each bias voltage and (b) extracted rise and decay times versus different bias voltages.

Conclusions

In summary, flexible NIR photodetectors were prepared using Ag5Te3 NPs for optoelectronic applications. The Ag5Te3 NPs used for the active channel layer of the photodetectors were synthesized in an aqueous solution, and Ag5Te3 thin films were successfully prepared on PES substrates at 150 °C using redistributed Ag5Te3 NPs in aqueous inks. The structural, chemical, optical, and optoelectronic properties of the Ag5Te3 NP-based thin films were studied as functions of the postannealing temperatures for the first time. The detectivity of the photodetectors was higher than those of two-dimensional material-based phototransistors, such as MoS2, or commercial thermistor bolometers (∼109). Moreover, this method can be used for other Ag5Te3-based applications in the future.

Materials and Methods

Preparation of Materials

All materials used in the experiment were purchased from Sigma-Aldrich. First, 0.4 g of AgNO3 and 1 mL of 1-thioglycerol (capping material) were injected into 100 mL of deionized water. Then, to adjust the pH of the solution to 11.4, 14 mL of 1 M NaOH was added. Na2TeO3 was used as a Te precursor. Next, 1.0, 2.0, 2.5, 12.5, and 25.0 mM Na2TeO3 aqueous solutions were prepared. Thereafter, 10 mL of the prepared solution was added to the aforementioned mixture and 80 mL of a condense solution was prepared via rotary evaporation at 30 °C and 100 rpm. The final phase was affected by the Te2– concentration. Here, hexagonal-phase Ag5Te3 NPs, which were synthesized using 25 mL of the Na2TeO3 precursor solution, were used for the active channel layer of the flexible NIR photodetectors. After adding acetone, the Ag5Te3 NP powders were isolated using a centrifugal separator and then washed with acetone again. The Ag5Te3 NPs extracted via this process were dried in a vacuum environment.

Fabrication of Flexible Photodetectors

To fabricate two-terminal flexible photodetectors, the PES substrates were employed. Gold (Au) electrodes were deposited on the PES substrates via e-beam evaporation with a metal shadow mask. The thickness, length, and width of the Au electrodes were 50 nm, 1000 μm, and 100 μm, respectively. Then, the PES substrate was subjected to a UV/ozone treatment for 15 min to control its surface energy. The redistributed Ag5Te3 NPs in deionized water was spin-coated on PES substrates at 3000 rpm for 30 s. The Ag5Te3 thin films were annealed at 50, 100, and 150 °C for 1 h.

Analysis of Thin Films and Photodetectors

The size, shape, and crystallinity of the Ag5Te3 NPs were examined using TEM and SAED. The phase composition of the Ag5Te3 thin films was analyzed using GIXRD with a small incident angle (0.3°) and Cu Kα radiation (λ = 0.154 nm). The elemental composition of the films was investigated using XPS (Quantera SXM (Physical Electronics, Chanhassen, MN, USA). The optical band gap was determined using UV/visible (UV/vis) spectroscopy. The thickness of the films was examined and measured using a scanning probe microscope (Park NX20, (Park Systems, Suwon, Korea), tapping mode). The optoelectronic properties of the Ag5Te3 photodetectors were examined using a semiconductor parameter analyzer (Keithley 2636B) and a probe station with an 808 nm laser.