Yuanhao
Li
a,
Yukun
Wang
*a,
Zuhuan
Lu
a,
Zongming
Yu
a,
Tianyi
Zhang
a and
Wenhong
Sun
*abcd
aResearch Center for Optoelectronic Materials and Devices, Guangxi Key Laboratory for the Relativistic Astrophysics, School of Physical Science and Technology, Guangxi University, Nanning 530004, China. E-mail: ykwang0929@163.com
bState Key Laboratory of Featured Metal Materials and Life-cycle Safety for Composite Structures, Guangxi University, Nanning 530004, China
cMOE Key Laboratory of New Processing Technology for Nonferrous Metals and the Guangxi Key of Processing for Non-ferrous Metals and Featured Materials, Guangxi University, Nanning 530004, China
dThird Generation Semiconductor Industry Research Institute, Guangxi University, Nanning 530004, China
First published on 28th April 2025
Methylammonium lead iodide is widely used in the preparation of photodetectors because of its excellent photovoltaic properties. However, because of the nature of the perovskite polycrystals, the low-temperature solution treatment approach of creating perovskite thin films causes flaws to arise in the material. Undercoordinated lead ions (Pb2+) have been shown to have comparatively low formation energies among all defect species and to be a major contributor to defect creation. Here, we use a straightforward but efficient additive engineering strategy to introduce tetrafluoroterephthalonitrile as an additive into the perovskite precursor solution to passivate perovskite defects and improve the quality of perovskite thin films for the production of high-efficiency perovskite photodetectors. We demonstrated that –CN (cyano) and polyfluorine atoms in the structure of TFTPN may passivate defects caused by lead ions and prevent the releasing of organic cations (MA+), improving the stability of the perovskite structure and the quality of perovskite films. Consequently, we designed perovskite photodetectors with TFTPN that demonstrated exceptional performance in terms of photoresponse, detection and other areas. These devices had a maximum peak external quantum efficiency (EQE) of 91.73%, a lower dark current density of 8.86 × 10−11 A cm−2, a linear dynamic range (LDR) of 105.4 dB, and more. Furthermore, the perovskite photodetector's durability is enhanced by the addition of TFTPN, and after 20 days of storage in an air environment at 25 °C and 20–30% relative humidity, the device retains its original efficiency of 92.3%.
According to our findings, the TFTPN structure's –CN (cyano) and multi-fluorine atoms may passivate lead ion-induced defects and stop organic cations from escaping, boosting the perovskite structure's stability and the perovskite film's quality. In all, we designed a photodetector with a linear dynamic range (LDR) of up to 105.4 dB, a lower dark current density of 8.86 × 10−11 A cm−2, and a maximum peak external quantum efficiency (EQE) of 91.73%. The stability was enhanced, and the device retained 92.3% of its original efficiency after 20 days of storage in an air surrounding with a temperature of 25 °C and a relative humidity of 20–30%.
To verify the above results in more depth, we performed X-ray photoelectron spectroscopy (XPS) analysis of the perovskite films with or without TFTPN molecules based on the structure as ITO/NiOX/MAPbI3. Here, for ease of presentation, we define the ideal doping concentration as 2 mg mL−1, due to the fact that a large number of studies have shown that a doping concentration of 2.0 mg mL−1 gives the highest efficiency of a perovskite photodetector; the ideal concentration is used for the characterization, unless otherwise stated. Analyzing the with TFTPN encapsulated thin films using XPS, the C 1s-core spectra revealed distinct –CN (285.93 eV) and C–F (289.03 eV) bonding signals (Fig. 2a), demonstrating that TFTPN molecules were successfully incorporated into the ref. 23–25. Meanwhile, the results of energy dispersive X-ray spectroscopy (EDS) showed the N and F elements, which confirmed the above results (Fig. 2b, c and Fig. S1a–c, ESI†). These results provide strong support for subsequent performance tests and mechanistic studies. The addition of –CN groups probably increased the electron cloud density of Pb2+, which is why the peaks of Pb2+ were moved toward lower binding energies after the additive was added.20 The primary peaks that were initially situated at 143.16 eV (Pb 4f5/2) and 138.29 eV (Pb 4f7/2) were both moved toward the low binding energy region (i.e., 142.75 eV corresponds to Pb 4f5/2, and 137.89 eV corresponds to Pb 4f7/2) following TFTPN modification, as can be seen in the XPS spectral curves displayed in Fig. 2d. This result suggests that the strong interaction exists between the –CN group in TFTPN and the undercoordinated Pb2+ ions in the perovskite membrane. Both the I 3d and N 1s spectra revealed a notable shift toward lower binding energies in the membranes changed by TFTPN, as illustrated in Fig. 2e and f. By forming coordination links between TFTPN and I− and MA+ sites, this shift event efficiently lowers the number of defects on the perovskite surface.26 This outcome further demonstrates TFTPN's exceptional capacity to prevent the formation of defects and restore the surface structure of perovskite thin films.27
![]() | ||
Fig. 2 (a) XPS spectra of C 1s of MAPbI3 with TFTPN. (b) EDS image of F. (c) EDS image of N. (d) Pb 4f orbital XPS spectra. (e) I 3d orbital XPS spectra. (f) N 1s orbital XPS spectra. |
In addition, we hypothesized that F− in TFTPN could form a hydrogen bond with the charged MA+, and the above N 1s peak was shifted to the direction of low binding energy, which indicated a charge transfer on the N–H unit of the MA+, and since F− has a strong electronegativity, such a shift may imply the formation of a hydrogen bond between F− and MA+.11,28 This was verified in our 1H NMR characterization, as shown in Fig. S3 and S4 (ESI†). In pure deuterated DMSO solution, the resonance signal of protonated ammonium in MAI can be observed to be located at 7.49 ppm.29 However, after TFTPN modification, the resonance signal of ammonium splits at both 7.49 ppm and 7.56 ppm, forming two distinct peaks. This occurrence implies that the chemical environment of ammonium in MA+ is changed by the interaction between the addition of TFTPN and MA+, leading to two distinct states.30 The creation of hydrogen bonds between F− and MA+ efficiently prevents the movement of MA+.31–34
Subsequently, we verified the chemical interaction between TFTPN and lead iodide by Fourier transform infrared spectroscopy (FTIR) analysis. In the FTIR spectrum of TFTPN (Fig. S5, ESI†), peaks located at 2252 cm−1 and 1260 cm−1 were observed, which corresponded to the stretching vibrations of the C–N and C–F bonds, respectively, and the peak at 1500 cm−1 was attributed to the stretching vibration of the C–C aromatic bond.35–37 Notably, these distinctive peaks can also be found in the TFTPN–PbI2 spectrum. The effective binding of TFTPN to PbI2 to create the TFTPN–PbI2 complex is strongly supported by this discovery.
The above findings reveal that additive TFTPN is capable of forming strong ligand and hydrogen bonds with the perovskite material, and we expect that these strong interactions will have a significant optimizing effect on the morphology of the modified perovskite films. To confirm this, we used scanning electron microscopy (SEM) and atomic force microscopy (AFM) on substrates with the structure of ITO/NiOX/MAPbI3 to thoroughly examine the surface morphology of the perovskite films with or without TFTPN. The cross-section of the ITO/NiOX/MAPbI3 structure with TFTPN (Fig. 3a). As shown in Fig. 3b–e, the surface morphology of the TFTPN additive-modified perovskite films is flatter and denser, with fewer pinholes and a significant increase in grain size. From Fig. 3f and g, we counted the grain size. The average grain size of perovskite films before modification was 205 nm, while the average grain size of the modified films increased to 247 nm. This indicates that the TFTPN additive significantly increased the average grain size of perovskite films. Additionally, as illustrated in Fig. 3h and i, the root-mean-square (RMS) of perovskite films surface roughness with TFTPN decreased to 13.5 nm, a 10.4 nm decrease from the unaltered control films' roughness of 23.9 nm. In summary, the introduction of the additive TFTPN not only improves the morphology of perovskite films but also provides an effective way to prepare high-quality perovskite films by increasing the grain size and reducing the surface roughness.
To obtain a better understanding of how additive introduction affects perovskite crystal formation, we used X-ray diffraction (XRD) to compare the growth of perovskite crystals with and without TFTPN. Fig. 4a shows the XRD patterns of perovskite films prepared in MAPbI3-based perovskite precursor solution with TFTPN additions of 0, 0.5, 1.0, 2.0, and 3.0 mg mL−1, respectively. No discernible peak displacement was seen, and all films had a dominating (110) peak at 14.6°, confirming that the perovskite lattice structure was not negatively impacted by the addition of TFTPN. Notably, the perovskite film exhibited the strongest diffraction intensity of the (110) peak when the TFTPN addition was 2 mg mL−1. The results showed that the (110) peak at 14.6° of the perovskite film with TFTPN was not only significantly enhanced in intensity, but also its full width at half peak (FWHM) was significantly narrower (Fig. 4b and c). This finding strongly suggests that the introduction of TFTPN effectively enhances the crystalline quality of the perovskite film.38 The findings of this XRD examination are in excellent agreement with our earlier findings from SEM, which both demonstrate that TFTPN positively promotes the development of perovskite. Based on previous studies, we postulated that the interaction of TFTPN with Pb2+ ions in perovskite precursor solution at the surface of the colloidal particles would be a significant factor influencing the formation of calixarene.39 We used dynamic light scattering (DLS) measurements to confirm this, and the outcomes are displayed in Fig. 4d. We observed a change in the colloidal particle radius distribution in perovskite precursor solution with TFTPN compared to the control sample. The radius of colloidal particles was Gaussian distributed, and the average size of colloidal particles with TFTPN was about 904 nm, while the average size of colloidal particles of control samples was about 1174 nm (Fig. 4e and f). This change may be attributed to the interaction of the TFTPN additive with the colloidal particles in the precursor solution. In particular, TFTPN might coordinate with Pb2+ ions as a ligand, changing the colloidal particles' surface characteristics. The bigger colloidal particles may be “trimmed” into smaller ones by this coordination, which could have an impact on the perovskites' growth process.40 The above DLS measurements further confirm that the introduction of TFTPN plays a key role in the growth and film formation of perovskites. Since the quality of perovskite film formation directly affects its light absorption efficiency, by optimizing the addition of TFTPN, we can expect to obtain higher-quality perovskite films, which will improve the performance of related optoelectronic devices.
Thus, we employed ultraviolet-visible spectroscopy (UV-vis) to thoroughly examine the precise impact of TFTPN on the light absorption capabilities of perovskite films on devices with the ITO/NiOX/MAPbI3 structure (Fig. 4g). According to the conclusions, the TFTPN-modified perovskite films have noticeably higher light absorption in the 530–780 nm wavelength range. This is a clear sign that the growth quality of the perovskite crystals has been effectively improved. To more accurately assess the effect of TFTPN modification on the band gap (Eg) of perovskite thin films, we analyzed it by plotting Tauc plots (i.e., Ahν2versus hν) based on UV-vis spectroscopic data (Fig. 4h). Due to the analytical data, the TFTPN-modified perovskite film's band gap is approximately 1.589 eV, which is smaller than that of the control without surface treatment (1.597 eV). In addition, to further assess the quality of the perovskite films, we calculated the Urbach energy (EU) using the formula:41,42
ln![]() ![]() | (1) |
The improvement in the quality of perovskite films has a positive impact on the extraction and transport of photogenerated carriers. To this end, we performed steady-state (PL) and time-resolved photoluminescence (TRPL) spectroscopic tests for ITO/NiOX/(0, 0.5, 1.0, 2.0, and 3.0 mg mL−1 TFTPN: MAPbI3) structures to probe the carrier extraction properties of perovskite samples at the NiOX/MAPbI3 interface. In comparison to the unmodified MAPbI3 films, the TFTPN-modified films gradually decreased in PL intensity under 532 nm laser illumination, as seen in Fig. 5a. The perovskite film doped with 2 mg mL−1 TFTPN had the lowest PL intensity among all. The reduction in PL intensity suggests that holes are extracted more quickly at the NiOX/MAPbI3 interface, further demonstrating the better PL quenching efficiency and more efficient carrier transport capacity of TFTPN-modified treated perovskite films at the NiOX interface.44 Next, we observed the TRPL image, which was fitted by a double-exponential decay function:45
y = A1e(−t/τ1) + A2e(−t/τ2) + y0 | (2) |
![]() | (3) |
![]() | ||
Fig. 5 (a) Steady-state PL spectra of perovskite films doped with different concentrations of TFTPN. (b) Time-resolved PL decay plots of perovskite films doped with different concentrations of TFTPN. |
The TRPL data revealed that the device at the optimal concentration (39.40 ns) had the shortest decay lifetime when compared to the control device (119.83 ns), based on the specific data values displayed in Table 1. This finding implies that photoinduced carriers have the ability to migrate swiftly from the cluster to NiOX.46 This outcome is in line with previous PL test finding. The creation of stable and high-performing perovskite photodetectors is strongly supported by these observations, which collectively show the superior photovoltaic qualities and stability of premium perovskite films.
Concentration (mg mL−1) | τ 1 (ns) | A 1 (%) | τ 2 (ns) | A 2 (%) | τ avg (ns) |
---|---|---|---|---|---|
0 | 5.36 | 2.20 | 141.94 | 0.43 | 119.83 |
0.5 | 6.25 | 1.64 | 101.74 | 0.49 | 85.45 |
1.0 | 2.93 | 7.85 | 79.56 | 0.41 | 47.87 |
2.0 | 3.47 | 5.62 | 71.33 | 0.29 | 39.40 |
3.0 | 3.96 | 3.74 | 104.89 | 0.35 | 75.88 |
A schematic representation of the planar inverted perovskite photodetectors (PDs) with ITO/NiOX/MAPbI3 (with and without TFTPN)/C60/BCP/Cu structure is provided in Fig. 6a. This model was built to thoroughly examine the precise impact of TFTPN additions on the device performance. External quantum efficiency (EQE) is an important indicator of photoelectric conversion efficiency, which is expressed as the ratio of the number of electrons collected by the photodetector to the number of incident photons and can be expressed as:47
![]() | (4) |
Here, we introduce the responsivity and specific detectivity parameters to further evaluate effective capture of light by photodetectors with or without TFTPN. The responsivity (R) reflects the efficiency of the PDs to respond to different intensities of incident light and can be expressed as:48
![]() | (5) |
Specific detectivity (D*) is a measure of the photodetector's ability to detect weak light signals. When the noise of the detector is mainly due to grain noise, D* can be expressed as:49
![]() | (6) |
Linear dynamic range (LDR) is a core performance metric of photodetectors, which reflects the wide interval over which the photodetector can accurately detect and quantify the intensity of the optical signal, defined as:50
![]() | (7) |
The excellent performance of the devices in terms of EQE, responsivity, and specific detectivity indicated that the incorporation of TFTPN effectively passivated the defects in the perovskite structure. To further investigate the effect of doping with TFTPN doped in the perovskite on the passivation of the defects in the perovskite photodetectors, the J–V characteristics of the photodetectors doped with and without TFTPN were measured in dark conditions. J–V characteristics of perovskite photodetectors (Fig. 6g), the dark current densities of photodetectors with TFTPN were all lower than those of photodetectors without TFTPN under 0 V bias (Table S2, ESI†), and PDs with TFTPN decreased from 5.24 × 10−9 A cm−2 to 8.86 × 10−11 A cm−2 compared to PDs without TFTPN. The results show that the introduction of TFTPN can effectively promote the formation of high-quality perovskite films with low defect density.
To quantitatively evaluate the passivation effect of TFTPN on perovskite films, the defect density of TFTPN on perovskite films was calculated using the space charge limiting current (SCLC) method. A schematic of the hole-only device, which consists of these functional layers including ITO/NiOX/MAPbI3/Spiro-OMeOxide/MoO3/Cu, is illustrated in Fig. 6h. The defect density can be calculated by the following equation:20,51
![]() | (8) |
One of the main elements influencing the performance of photodetectors is the transit efficiency of photogenerated carriers. Accelerated transient photocurrent rise and fall periods typically indicate faster and more effective carrier transportation within the device. The photodetector's transient photocurrent was measured using both the photodetectors with or without TFTPN under 520 nm light irradiation. As shown in Fig. 7a and b, PDs outstanding and steady dynamic response is demonstrated by the photocurrent curves, which show no discernible attenuation of the photocurrent after multiple cycles of testing under continuous 1 KHz light irradiation. The rise time is the time required for the photocurrent to rise from 10% to 90% of the maximum value, while the fall time is the time required for the photocurrent to fall from 90% to 10% of the maximum value.52 The rise and fall times of the control device are 166 μs and 20 μs, respectively, while the rise and fall times of the wTFTPN device are 122 μs and 20 μs, respectively, and the response time of the optimal device becomes faster after the TFTPN modification treatment.
For photodetectors, having excellent photoelectric performance is not enough to be called an excellent photodetector, and stability is another key aspect of photodetectors. Subsequently, the unencapsulated devices were subjected to a 20-day stability test in air (20–30% relative humidity, 25 °C), as shown in Fig. 7c. After 20 days, we found that the EQE values of photodetectors with or without TFTPN were maintained at 92.3 and 64.2% of their initial EQEs, respectively, and that under the same conditions was mainly attributed to the TFTPN molecules introduced into the perovskites with effective defect passivation and increased perovskite crystallinity. The stability test results show that the introduction of TFTPN molecules can improve the stability of perovskite photodetectors. Given that perovskites are extremely sensitive to moisture, which is regarded as a critical external damage factor, the water contact angle of perovskite thin films was thoroughly analyzed. The water contact angle of the TFTPN-modified perovskite film was considerably raised to 60.85° from 44.49° in the control group, as seen in Fig. 7d. This change indicates that the addition of TFTPN can successfully prevent water intrusion and thereby safeguard the perovskite film.
MAPbI3 precursor solution was prepared by combining MAI and PbI2 (1:
1 molar ratio) in a solvent mixture of DMF and DMSO (9
:
1 volume ratio) at the concentration of 1.2 M. TFTPN powder was added directly to the perovskite precursor solution at 0, 0.5, 1.0, 2.0 and 3.0 mg. All solutions were stirred overnight at a heating temperature of 70 °C and the solutions were filtered using 0.22 μm PTFE (polytetrafluoroethylene) prior to handling. MAPbI3 was spin-coated on the NiOX substrate, first at 3000 rpm for 13 s and then at 5000 rpm for 50 s. In the surplus 45–49 s of the second process, chlorobenzene is dropped onto the spinning perovskite film. Then, the perovskite film was annealed on the hot plate at 100 °C for 20 min to prepare a perovskite layer.
Finally, C60, BCP and Cu were deposited in the vapor deposition system to form the final device with thicknesses of 30, 8 and 100 nm respectively. The effective area of the device is 0.16 cm2.
Footnote |
† Electronic supplementary information (ESI) available: The EDS image of perovskite film with TFTPN. The EDS image of Pb and I. The full XPS spectrum of perovskite film. The 1s orbital XPS spectra of F. The 1H NMR spectra of MAI with or without TFTPN in d-DMSO. The FTIR spectra of PbI2 power, TFTPN power, and mixed power of PbI2 and TFTPN. Measurements of EQE, dark current density, R, and D* curves of PDs doped with different concentrations of TFTPN. Comparison of parameters with previously reported perovskite detectors. PDs performance parameters. See DOI: https://doi.org/10.1039/d5tc01086a |
This journal is © The Royal Society of Chemistry 2025 |