Jinhua
Li‡
ab,
Jiantao
Wang‡
c,
Yecheng
Zhou‡
d,
Chengzhuo
Yu
c,
Heng
Liu
c,
Xingnan
Qi
c,
Ruxue
Li
e,
Yuhui
Hua
ab,
Yinye
Yu
d,
Rui
Chen
e,
Dafa
Chen
a,
Lingling
Mao
a,
Haiping
Xia
*ab and
Hsing-Lin
Wang
*cf
aDepartment of Chemistry, Shenzhen Grubbs Institute, Southern University of Science and Technology, Xueyuan Avenue 1088, Shenzhen 518055, China. E-mail: xiahp@sustech.edu.cn
bDepartment of Chemistry, College of Chemistry and Chemical Engineering, Xiamen University, Xiamen 361005, China
cDepartment of Materials Science and Engineering, Southern University of Science and Technology, Xueyuan Avenue 1088, Shenzhen 518055, China. E-mail: wangxl3@sustech.edu.cn
dSchool of Materials Science and Engineering, Sun Yat-sen University, Zhongshan West Road 135, Guangzhou 510275, China
eDepartment of Electrical and Electronic Engineering, Southern University of Science and Technology, Xueyuan Avenue 1088, Shenzhen 518055, China
fGuangdong Provincial Key Laboratory of Energy Materials for Electric Power, Southern University of Science and Technology, Xueyuan Avenue 1088, Shenzhen 518055, China
First published on 13th July 2022
Perovskite solar cells (PSCs) require low work function (WF) cathodes to collect electrons, but often chemically reactive metals are used for devices with inverted configuration. Reactive metals (such as Ag, Cu and Al) with low WF encounter easy oxidation and corrosion, which threatens the long-term stability of devices. Herein, we tailor an organometallic carbolong derivative to modulate the cathode interface in inverted PSCs for the enhancement of power conversion efficiency (PCE) and stability. Density functional theory and surface WF characterization reveal that this organometallic compound can reduce the WF of metals by forming interfacial and molecular dipoles, which reduce the energy barrier for electron transport from the electron transport layer to the external metal cathode. By using this carbolong derivative to modulate the cathode interface, inverted PSCs based on the commonly used Ag cathode obtain a PCE of 21.46% with a remarkable FF of 83.14%. By replacing low-WF Ag with high WF Au, the devices achieve more than 20% PCE and improved ambient stability, and can maintain over 85% of the initial PCE for over 500 h in an inert environment under the maximum power point (MPP) tracking. This work provides a significant route for the realization of high-efficiency and stable PSCs by integrating rationally designed cathode interfacial materials and chemically stable metals.
In general, a cathode interlayer (CIL) is introduced between the ETL and the cathode metal to modulate the interfacial energy barrier and facilitate the carrier transport.10,11 In addition, a dense and hydrophobic CIL spacer is efficacious to protect the underneath layers from metal electrode diffusion and moisture penetration.12,13 One alternative approach to achieve a long-lived cathode is to design CIL materials that can both modify the interfaces and allow the use of chemically stable metals.14 For example, Fang et al. reported a benzotriazole CIL to form a protective layer by suppressing the reaction between the degraded perovskite and Cu electrode.9 Li and Yang et al. reported the use of a phenyl phosphine-inlaid semiconducting polymer as the CIL in inverted PSCs to attain a simultaneous enhancement of device efficiency and thermal/lighting stability.15 Although the CILs with ample molecular structures have been reported to improve the device performance and stability, the molecular design and employment of novel CIL materials that allow the use of high-WF and chemically stable metals as cathodes are still under explored. Our recent work showed that the utilization of a series of novel organometallics with π-conjugating frameworks (carbolong complexes16) can drastically reduce the metal WF and hence achieve high device PCE and superior operational stability for both perovskite solar cells17 and organic solar cells.18,19 Interestingly, the coordination sphere modification of the carbolong framework would improve the device performance. As shown in Fig. 1, the introduction of carbolong complex a1 (M-CCC core), of which the iridium center is chelated by a tridentate carbon–carbon–carbon (CCC) ligand, as the CIL material in inverted PSCs obtained champion PCEs of 20.65% (Ag as the metal cathode) and 18.09% (Au as the metal cathode).17 Furthermore, when the modified framework b1 (M-CCO core) was employed as the CIL, the device performance, especially for the high work-function Au cathode, was enhanced significantly.17 The remarkable improvement of efficiencies in PSCs by using distinct organometallic frameworks as interfacial materials prompts us to consider whether the rational design of the carbolong framework is still applicable to modulate the interfacial properties of the cathode in inverted PSCs, thus boosting the device PCE.
Fig. 1 Modification of the carbolong framework and the device performance results achieved by their usage as CILs combining both low-WF Ag and high-WF Au cathodes in inverted PSCs. |
Herein, we report the irida-carbolong derivative 1 that exhibits a brand-new organometallic framework as the CIL in inverted PSCs to enhance the device performance and stability. This material can extensively reduce the WF of external metals in inverted PSCs by forming interfacial molecular dipoles. Hence, the employment of compound 1 as the CIL in inverted PSCs reduces the energy barrier for electron transport from the ETL to the external metal, which facilitates the fill factor (FF) and current density (Jsc). The inverted PSCs based on the Ag cathode obtain a champion PCE of 21.46% with a remarkable FF of 83.14%. As a proof of concept, we replaced the low-WF Ag with high WF Au as the cathode for the fabrication of high performance and stable inverted PSCs. The devices based on the Au cathode with the CIL achieve more than 20% PCE and can maintain over 85% of the initial PCE for over 500 h in an inert environment under the maximum power point (MPP) tracking. This work establishes a route for the achievement of high performance and stable PSCs by combining the rationally designed CIL material with the chemically stable metal.
To verify the capacity of compound 1 as the CIL in inverted PSCs, we fabricated photovoltaic devices with the structure of ITO/PTAA/perovskite/PC61BM/CIL/Ag (Fig. 3(A)).20 All the layers except for the electrodes are solution processed, which is low-cost and straightforward.21 The perovskite composition is (CsPbI3)0.05[(FAPbI3)0.9(MAPbBr3)0.1], whose related characterization including surface SEM morphology (Fig. S5, ESI†), film XRD (Fig. S6, ESI†), UV-Vis absorption and PL curves (Fig. S7, ESI†) can be found in the ESI.† We recorded the J–V curves with the forward scan and reverse scan of the champion devices without and with the CIL as shown in Fig. 3(B) and (C), respectively. Their detailed photovoltaic parameters are listed in Table S1 (ESI†). The pristine device without the CIL presents a highest PCE of 17.39%, with an open-circuit voltage (Voc) of 1.08 V, a short-circuit current density (Jsc) of 22.79 mA cm−2, and a fill factor (FF) of 70.38%. The optimized concentration of compound 1 is 0.8 mg mL−1, which enables a film with 8 nm thickness as measured using a spectroscopic ellipsometer. Under this optimal condition, the champion device with compound 1 as the CIL delivers a highest PCE of 21.46%, with a Voc of 1.08 V, a Jsc of 23.88 mA cm−2, and a FF of 83.14%. In comparison, the effects of CIL thickness on the device performance were also taken into consideration as shown in Table S2 (ESI†), which clearly shows that the thickness of 8 nm is the best. To show the superiority of the irida-carbolong derivative 1 of this work to the other cathode interlayers, we fabricated the PSCs with the widely-used bathocuproine (BCP) as the CIL. The photovoltaic parameters of the 8 nm BCP based device are also included in Table S1 (ESI†). The PSCs with BCP possess a highest PCE of 19.02%, with a Voc of 1.07 V, a Jsc of 21.57 mA cm−2, and a FF of 82.40%, still lower than our compound 1 based device. In addition, we compared our results with the reported literature as listed in Table S3 (ESI†). The performances of our CIL included devices are among the highest reported values for inverted PSCs without passivation processing. The EQE values of both devices with and without compound 1 are shown in Fig. 3(D). After inserting the CIL, the device presents an enhanced EQE compared with the pristine device at every measured wavelength. The integrated Jsc of the pristine device is 21.09 mA cm−2 while that of the CIL-modified device is 22.12 mA cm−2. The calculated Jsc values of the pristine device and the CIL-modified device show 7.46% and 7.37% difference with those measured from J–V curves. To confirm the PCE under stabilized conditions, we measured the maximum-power-point (MPP) outputs of typical devices with and without the CIL, and the results are plotted in Fig. 3(E). The MPP outputs within 600 s under AM 1.5G one sun illumination give a stabilized device PCE of 19.5% for the CIL-modified device and 15% for the pristine device. To investigate the device reproducibility, the PCE values of pristine and CIL-modified devices with a respective total number of 20 were statistically studied and their distributions are presented in Fig. 3(F). The average PCE values with standard deviation of the pristine and CIL-modified devices are 15.21 ± 1.42% and 19.61 ± 0.9%, respectively, which are highly in line with the J–V and MPP measurements. Therefore, compound 1 as the CIL plays an important role in enhancing the device performance.
Fig. 3 (A) Device structure. J–V curves of the best device (B) without and (C) with the CIL. (D) EQE and calculated Jsc curves. (E) MPP output. (F) Device statistics. |
From the analysis of device performance, the employment of compound 1 as the CIL mainly boosts the FF and Jsc, which are highly related to the carrier transport efficiency. Generally, a low-WF metal has a lower interfacial energy barrier relative to the ETL and can improve the carrier transport. The introduction of compound 1 as the CIL was expected to reduce the WF of the Ag surface since we have confirmed that the WF of PCBM is almost not impacted by the CIL as shown in Fig. S8 (ESI†), despite the efficient reduction of the PCBM surface roughness (Fig. S9, ESI†). To verify the WF reduction of metal Ag enabled by the introduction of compound 1, we first conducted the density functional theory (DFT) study via building the adsorption model with different molecular configurations and performed molecular dynamics simulation to obtain the final configuration as shown in Fig. S10 (ESI†). In the final molecular adsorption configuration with the lowest energy as shown in Fig. 4(A), [BF4]− anion tends to adsorb on the Ag surface and thus the organometallic cation is sterically distributed on the anion surface, leading to the formation of the anion–cation molecular dipole (μ1). In addition, the cation in compound 1 is large enough to cover the [BF4]− anion, and parts of the cations prefer to contact with the Ag surface directly, contributing to another interfacial dipole (μ2) on the Ag surface. Both the molecular dipole (μ1) and interfacial dipole (μ2) would synergistically help increase the inner surface potential and reduce the WF of Ag.22 The local potential curve derived from the adsorption model is shown in Fig. 4(B). The WF of the Ag (111) surface is obviously reduced from 4.3 eV to 3.95 eV, which is a 0.35 eV reduction.
To experimentally confirm the WF reduction of the Ag surface after using compound 1 as the CIL, we carried out the UPS measurements. The WF can be calculated according to the equation WF = hϑ − Ecut-off, where hϑ is the incident photon energy of 21.2 eV, and Ecut-off is the energy of the secondary electron cut-off. The WF of the Ag surface shows a 0.5 eV reduction after CIL modification according to the energy of the secondary electron cut-off as shown in Fig. 4(C). Moreover, we further characterized the WF change before and after the coverage of compound 1 on the Ag surface by performing scanning Kelvin probe microscopy (SKPM). As shown in Fig. 4(D) and (E), the surface potential mappings of Ag surfaces before and after the deposition of the CIL reveal a noticeable increase. From the comparison of their extracted line-cut profiles as shown in Fig. 4(F), the surface potential is increased from −254.829 mV to 323.847 mV. According to the equation φM = φtip − eVcp, where φM and φtip refer to the WF of the metal and conductive tip, respectively, e is the charge constant, and Vcp is the contact potential, the change of measured surface potentials before and after the introduction of the CIL reflects a WF reduction of approximately 0.59 eV after introducing compound 1.23
To clarify how the dipoles generated by the CIL affect the metal electrode work function, a schematic diagram of the energy level and dipole distribution is plotted in Fig. 5(A). Due to the CIL dipole layer at the interface of PC61BM and the metal cathode, the vacuum energy level (Evac) is steeply shifted, and it therefore reduces the apparent WF of the metal cathode by Δ.24 This electric dipole interlayer can strongly affect the carrier transport when embodied into photovoltaic devices. To study the carrier transport at the cathode interface in the perovskite-included devices, we collected the transient photocurrent (TPC) decay profiles of the PSCs with and without the CIL. As shown in Fig. 5(B), the device without the CIL has a longer photocurrent decay life time of 0.43 μs, longer than the CIL modified device of 0.19 μs. The shorter lifetime after the usage of the CIL signifies the enhanced carrier extracted to the electrodes in the CIL modified PSCs.25 We further performed time-resolved photoluminescence (TRPL) to characterize the carrier transport dynamics in the PSCs with and without the CIL as presented in Fig. 5(C). The fast extraction of carriers to the electrode would result in less carrier recombination and a shorter TRPL lifetime. The decay spectra were fitted with a bi-exponential function I(t) = A1exp(−t/τ1) + A2exp(−t/τ2), where τ1 is a fast component and τ2 is a slow component. The fitting parameters are listed in Table S4 (ESI†). The PSCs with the CIL have a shorter average decay lifetime τav (τav = A1τ1 + A2τ2), which indicates a facilitated carrier transport compared with the pristine device without the CIL. Much higher carrier transport should result in less interfacial carrier recombination induced by carrier accumulation in PSCs, which were further characterized by EIS. The EIS Nyquist plots of PSCs with and without the CIL are shown in Fig. 5(D). According to the inset equivalent circuit, the EIS resistances mainly contain a transport resistance (Rc) and a recombination resistance (Rrec). A larger semicircle of the PSCs with the CIL corresponds to a larger Rrec, which indicates the inhibited carrier recombination in the devices with the CIL.25 Therefore, the reduction of the Ag WF enabled by the CIL leads to improved carrier transport and suppressed carrier recombination at the ETL/cathode interface, contributing to the much-boosted device FF and Jsc.
Fig. 5 (A) Schematic diagram of the energy level and dipole distribution, (B) TPC, (C) TRPL and (D) EIS curves of the solar cells with and without the CIL. |
As low-WF metals encounter air-exposure-caused oxidation and halogen-diffusion-induced corrosion, it is promising to replace Ag with other high-WF and chemically stable metals.26 For concept demonstration, we employed high-WF and chemically stable metal gold (Au) as the cathode in the inverted PSCs. The J–V curves of Au based champion devices with and without the CIL are exhibited in Fig. 6(A). The photovoltaic parameters are listed in Table S5 (ESI†). The champion device without the CIL has a Voc of 1.08 V, a Jsc of 22.12 mA cm−2, a FF of 59.33% and a PCE of 14.11% while the device with the CIL has a Voc of 1.08 V, a Jsc of 23.92 mA cm−2, a FF of 80.36% and a PCE of 20.82% under reverse scan conditions. Similar to the Ag based devices, the Au based devices also show boosted FF and Jsc after the introduction of the CIL. Due to the higher WF of Au than Ag, there is a larger energy barrier for electron transport at the PC61BM/Au interface, which leads to worse FF and Jsc. The employment of the CIL can reduce the WF of Au as demonstrated by the SKPM measurements (Fig. S11, ESI†), which results in the facilitated carrier transport at the ETL/cathode interface and improved device FF and Jsc.
For stability measurements, we first measured the change of device efficiencies relative to the initial point under ambient conditions. Unencapsulated devices based on the Au cathode and Ag cathode with and without the CIL were stored in air with a humidity of 55 ± 10%. As shown in Fig. 6(B), within 300 h tracking time, the devices with the CIL present better ambient stability than the devices without the CIL and the Au/CIL devices present the best stability, which can hold over 80% of the initial PCE for more than 250 h. By contrast, the PCE of the Au based device without the CIL decreases to nearly 80% of the initial PCE at around 150 h. In addition, the PCE decrease speed of the Ag based devices with the CIL is faster than Au with CIL devices and is similar to the Au without CIL devices, which reveals the more stable nature of the Au cathode. Likewise, the Ag devices with the CIL are more stable than devices without the CIL. The fortified environmental stability of devices with the CIL may come from the increased hydrophobic properties of our CIL as proved via the water contact angle measurements shown in Fig. S12 (ESI†).27 The PC61BM has a water contact angle of 71.3°, while that of the CIL covered PC61BM is 85.7°. The increased hydrophobicity after the introduction of the CIL can mitigate the penetration of humidity, thus enhancing the air stability a lot. We further measured the operational stability under MPP tracking in a glove box with white LED light soaking. Normalized PCEs of devices based on the Au cathode with and without the CIL present continuous decrease as shown in Fig. 6(C). However, the devices with the introduction of the CIL show a delayed decrease to the time of 85% of the initial PCE (T85). The T85 of the CIL-modified device is 500 h while that of the pristine device is only 150 h. Therefore, the introduction of a CIL to the Au based inverted PSCs obviously enhances the device operational stability.
This work focuses on the metal cathode stability enabled by the tailored CIL material. Although many reported devices with Ag, Cu, or Al electrodes had very stable performances, even more stable than the values reported in this work, their achievements were based on the complicated treatment/passivation of the perovskite layer.28–32 Hence, our device efficiency and stability can be further improved by the interfacial/bulk passivation of the perovskite layer, which is beyond the scope of this manuscript and will be reported in our future work.33–35
Footnotes |
† Electronic supplementary information (ESI) available. CCDC 2133382 (1). For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d2qm00452f |
‡ These authors contributed equally to this work. |
This journal is © the Partner Organisations 2022 |