Increasing the Li-TFSI doping concentration in Spiro-OMeTAD enables efficient and stable perovskite solar cells

Abstract

Li-TFSI/tBP is a classic doping system for Spiro-OMeTAD in typical n–i–p structure perovskite solar cells (PSCs). Unfortunately, the solubility of Li-TFSI in chlorobenzene is limited, and tBP needs to be added to promote its dissolution. However, tBP is a volatile polar solvent that can damage perovskites and Spiro-OMeTAD films. Moreover, even with the addition of tBP, the maximum concentration of Li-TFSI in chlorobenzene is basically 50 mol%, and the free radicals generated by doped Spiro-OMeTAD are limited. If the Li-TFSI concentration is further increased, it will precipitate and hinder the doping effect. Herein, we demonstrated an effective strategy to improve the performances of the PSCs by enhancing the solubility of Li-TFSI and increasing its doping concentration from 50 mol% to 80 mol% through the substitution of tBP with 12-crown-4. The chelation of 12-crown-4 with Li⁺ not only increased the solubility of Li-TFSI in chlorobenzene and enhanced its doping efficiency but also effectively addressed issues such as Li⁺ migration, hygroscopicity, and pinholes caused by the Li-TFSI/tBP system. The PSCs based on this strategy achieved a champion power conversion efficiency (PCE) of 23.99% and maintained 83% of the initial PCE under ISOS-D-3 protocol aging for 30 days.

---

1. Introduction

Photovoltaic technology, which directly converts solar energy into electrical energy, is considered one of the most promising and economically viable renewable technologies. Recently, organic–inorganic hybrid perovskite solar cells (PSCs) have garnered significant attention due to their wide spectral range, excellent absorption coefficient, low exciton binding energy, long carrier diffusion length, and high carrier mobility. Since the first reported version of PSCs with a power conversion efficiency (PCE) of 3.8% in 2009, the efficiency of PSCs has rapidly soared to 26.1% after comprehensive optimization.¹ Currently, state-of-the-art PSCs adopt an n–i–p structure of the transparent conductive substrate/electron transport layer (ETL)/perovskite absorber layer/hole transport layer (HTL)/metal electrode.

The HTL is located at the top of the perovskite absorber layer, preventing direct contact between perovskites and Au or H₂O molecules, as well as extracting and transporting holes. So, the hole transport material (HTM) in the HTL is considered a critical component due to its pivotal role in both efficiency and stability.² To date, 2,2′,7,7′-tetrakis-(N,N-di-p-methoxyphenylamine)-9,9′-spiro-bifluorene (Spiro-OMeTAD, abbreviated as Spiro) has been one of the most commonly used HTMs in n–i–p structure PSCs due to its large bandgap, deep highest occupied molecular orbital (HOMO) level,³ high film quality, and good thermal stability.⁴ However, the Spiro exhibits low hole mobility and poor conductivity,⁵ requiring p-type doping to modulate its electrical properties.⁶

Lithium bis(trifluoromethanesulfonyl)imide (Li-TFSI) and 4-tert-butylpyridine (tBP) are two of the most commonly used dopants,⁷ in which Li-TFSI can promote the photo-oxidation of Spiro by reacting with oxygen radicals,⁸ while tBP can be used to increase the solubility of Li-TFSI in chlorobenzene. Their synergistic effect enables the achievement of a uniform HTL with high hole mobility and conductivity. However, even in an oxygen atmosphere, this traditional doping process still requires a long time, and prolonged oxidation can have a negative impact on the HTL quality and device performance.⁹ Additionally, the solubility of Li-TFSI in chlorobenzene is limited, and tBP needs to be added to promote its dissolution as mentioned above. However, tBP is a polar solvent that can damage perovskites and it is prone to volatilization due to its low boiling point, leaving voids in the HTL.¹⁰ In addition, tBP can reduce the doping effect due to its dedoping mechanism.¹¹

It is worth noting that even with the addition of tBP, the maximum concentration of Li-TFSI in the HTL solution is basically 50 mol%. Due to the limitation of Li-TFSI concentration, the doping efficiency of Spiro is generally lower than 20%.¹² The amount of free radicals generated by oxidizing Spiro is lower than expected, which affects the hole transport performance of the HTL. If the concentration of Li-TFSI is further increased, it will dissolve incompletely, affecting the doping efficiency. Therefore, developing a research strategy that can further increase the concentration of Li-TFSI in the HTL solution to enhance the doping efficiency of Spiro, while avoiding the use of tBP is crucial for efficient and stable PSCs.

Crown ethers possess unique cavity structures and size effects, which enable selective interactions with alkali metal ions. Choosing a crown ether with a cavity size that closely matches the radius of the target ion is crucial for effective chelation. The radius of Li⁺ is approximately 76 pm, while 12-crown-4 offers a cavity size ranging from 120 to 150 pm, making it particularly suitable for strong chelation with Li⁺. Therefore, we selected 12-crown-4 to chelate Li⁺ in Li-TFSI, forming a host–guest complex.^13,14 Herein, 12-crown-4 was introduced into Spiro solution to replace tBP, and the effects of increasing Li-TFSI concentration on the HTL doping and PSC performance were studied. It was found that replacing tBP with 12-crown-4 can indeed further increase the concentration of Li-TFSI in Spiro solution. As the concentration increases from 50 mol% to 80 mol%, the doping efficiency of Spiro is significantly enhanced, generating more Spiro˙⁺. The PSCs based on Spiro:80 mol% Li-TFSI/12-crown-4 achieved a PCE of 23.99%, surpassing the control PSCs based on Spiro:50 mol% Li-TFSI/tBP (22.29%). Additionally, 12-crown-4 with a hydrophobic skeleton, which replaces tBP, can chelate with hygroscopic Li⁺, effectively addressed issues such as Li⁺ migration, hygroscopicity, and pinholes caused by the Li-TFSI/tBP system, ultimately improving the stability of the PSCs.

2. Results and discussion

The doping of Spiro with Li-TFSI is an indirect doping process involving two redox reactions, as shown in chemical reactions (1) and (2).^7a Spiro reacts with O₂ under light to generate weakly bound species. Due to the slow natural oxidation rate, the Spiro exhibits low conductivity and hole mobility, necessitating the introduction of the dopant Li-TFSI to enhance doping efficiency. Li-TFSI stabilizes Spiro˙⁺ by introducing TFSI⁻, indirectly driving the formation of free radicals Spiro˙⁺.

Spiro + O₂ ⇋ Spiro˙⁺O₂˙⁻ (1)

Spiro˙⁺O₂˙⁻ + Li-TFSI → Spiro˙⁺TFSI⁻ + Li_xO_y (2)

From reaction (2), it can be seen that if the concentration of Li-TFSI can be increased, the chemical equilibrium may shift towards the positive reaction direction, generating more Spiro˙⁺ and achieving better doping of the Spiro HTL. However, limited by the solubility of Li-TFSI in chlorobenzene, the concentration of Li-TFSI in the current doping system has reached its upper limit.

An additional 5.5 mg of Li-TFSI was added to 1 mL of chlorobenzene solution containing 9.1 mg of Li-TFSI (equivalent to increasing the concentration of Li-TFSI doped Spiro solution from 50 mol% to 80 mol%) and stirred overnight. As shown in Fig. 1a, it was found that the solution was turbid, showing that the additional Li-TFSI had hardly dissolved, which also confirms the above statement. Subsequently, 3.82 μL of 12-crown-4 was added to the above-mentioned turbid solution and stirred for 1 minute, and the solution became clear, indicating that the addition of 12-crown-4 can indeed improve the solubility of Li-TFSI. This is because 12-crown-4 is a commonly used Li⁺ chelator that can chelate with Li⁺ in Li-TFSI to form a host–guest complex Li⁺(12-crown-4) TFSI⁻, promoting the dissolution of Li-TFSI. Therefore, based on 12-crown-4 replacing tBP, it is possible to dope Spiro with higher concentration of Li-TFSI.

Fig. 1 (a) Schematic diagram of 12-crown-4 improving the solubility of Li-TFSI. (b) Color changes and (c) UV-vis absorption spectra of the Spiro solutions exposed to light under anaerobic conditions. (d) UV-vis absorption spectra of the Spiro films exposed to oxygen and light. (e) The variation of the UV-vis absorption intensity of the Spiro films at 524 nm with time. (f) Calibration curve for measuring Spiro˙⁺ concentrations. (g) ESR spectra of the Spiro solutions. (h) Hole mobility test of devices with the ITO/PEDOT:PSS/perovskite/Spiro/Au structure. (i) Electrical conductivity test of devices with the ITO/Spiro/Au structure.

In order to study the doping effects of higher concentration of Li-TFSI on Spiro, we prepared 1 mL of Spiro solution, in which 3.82 μL of 12-crown-4 was added to replace tBP, doped with different concentrations (50 mol%, 60 mol%, 70 mol%, 80 mol%, and 90 mol%) of Li-TFSI. Meanwhile, the Spiro solution doped with 50 mol% Li-TFSI and tBP was prepared as the control group. Firstly, the color changes of different solutions were observed. The freshly prepared Spiro:Li-TFSI/12-crown-4 solutions and Spiro:50 mol% Li-TFSI/tBP solution were both pale yellow. However, as shown in Fig. S1 (ESI†), after being exposed to light under anaerobic conditions for 2 hours, the Spiro:Li-TFSI/12-crown-4 solutions changed from light yellow to reddish brown, and the color deepened with increasing Li-TFSI doping concentration, which should be related to the generation of Spiro˙⁺. Among them, the Spiro:80 mol% Li-TFSI/12-crown-4 solution, which exhibits the deepest color, presents a striking contrast with the control group solution that is still pale yellow (Fig. 1b). The results indicate that the Li-TFSI/12-crown-4 system can doped Spiro under anaerobic conditions and the color deepening of the Spiro solution can be attributed to the increased production of Spiro˙⁺.

To confirm the above results, the UV-visible absorption and electron spin resonance (ESR) spectra of the Spiro solutions were studied. As shown in the UV-visible absorption spectra of Fig. 1c, after being exposed to light under anaerobic conditions for 2 hours, the Spiro:50 mol% Li-TFSI/tBP solution showed no significant absorption peak, while the Spiro:Li-TFSI/12-crown-4 solutions showed an absorption broad peak at 524 nm, which corresponds to the absorption peak of Spiro˙⁺.¹⁵ The intensity of this absorption peak increases with the increase of the Li-TFSI concentration, and reaches its maximum intensity when the Li-TFSI concentration is 80 mol%. When the concentration of Li-TFSI continued to increase to 90 mol%, it could no longer dissolve completely, resulting in a decrease in the intensity of the absorption peak of Spiro˙⁺. The above results showed that the Spiro solutions containing 12-crown-4 can exhibit the absorption peak of Spiro˙⁺ under anaerobic conditions. In addition, we visualized the variation of the Spiro˙⁺ content with different Li-TFSI concentrations. 80 mol% Li-TFSI/12-crown-4 doping promotes the highest doping efficiency of Spiro. Moreover, from the ESR spectra of Spiro solution oxidized under aerobic and light conditions for 20 days (Fig. 1g), compared to the control group solution, the 80 mol% Li-TFSI/12-crown-4 doped Spiro solution exhibited significantly enhanced paramagnetic peaks in the range of 3500–3520 G, indicating the generation of more Spiro˙⁺.

Furthermore, evolution of Spiro˙⁺ in films was investigated by depositing these Spiro solutions on ITO substrates. These films were exposed to oxygen and light for 20 days to ensure complete oxidation, and the evolution of UV-vis absorption spectra was continuously monitored. The spectral changes of the six doped Spiro films at different time periods are displayed in Fig. S2 (ESI†). Fig. 1d displays the UV-vis absorption spectra of the Spiro films after 20 days and compares the Spiro˙⁺ absorption peak at 524 nm. The peak intensity increased with increasing Li-TFSI doping concentration, and the Spiro film doped with 80 mol% Li-TFSI/12-crown-4 exhibited the strongest absorption peak, which is consistent with the results obtained from the aforementioned solution measurements. The changes in absorption intensity at 524 nm of the Spiro films tracked for 20 days are presented in Fig. 1e, providing the trend of the Spiro˙⁺ content in these films. In the initial stage, the concentration of Spiro˙⁺ in the all films continued to increase, but the films containing 12-crown-4 exhibited a higher concentration of Spiro˙⁺. After 15 days, the absorption peak of the control group almost plateaued, while the Spiro films doped with Li-TFSI/12-crown-4 still showed a continued increasing trend. The results indicate that the Spiro doping in the control group reached saturation after 15 days, while the introduction of 12-crown-4 facilitated more Li-TFSI participation in doping, resulting in a sustained enhancement in the absorption peak of Spiro˙⁺.

A protocol for quantifying doping of Spiro has been proposed in the literature,¹² which uses Spiro˙²⁺ as a bridge to indirectly quantify the amount of Spiro˙⁺ produced after Li-TFSI doping. Spiro˙²⁺ is the dioxo-free radical form of Spiro, which undergoes a single-step dismutation reaction with Spiro, quantitatively forming two equivalent Spiro˙⁺ (Scheme S1, ESI†). This reaction is quantifiable and a calibration curve is established using Spiro˙²⁺. The abscissa and ordinate of this curve represent the Spiro˙⁺ content and the standardized absorption ratio of Spiro˙²⁺ (SAR_Spiro˙²⁺), respectively. According to the definition SAR_Li-TFSI = SAR_Spiro˙²⁺, SAR_Li-TFSI can be calculated by the absorbance at 524 nm in the UV-vis absorption spectrum of the film divided by the absorbance at the pseudo-isosbestic point at 407 nm to quantify Spiro˙⁺. Fig. 1f shows the quantitative analysis of Spiro˙⁺ based on the calibration curve at different doping levels. As the doping concentration of Li-TFSI increases, more Spiro˙⁺ is generated. When the doping concentration reaches 80 mol% Li-TFSI, the introduction of 12-crown-4 results in a Spiro solution with 7.9 mol% Spiro˙⁺ (i.e., every 80 mol Li-TFSI added to Spiro produces 7.9 mol Spiro˙⁺), demonstrating the highest doping efficiency. However, further increasing the doping concentration leads to a decline in the Spiro˙⁺ yield. Notably, the control group without 12-crown-4 exhibits the lowest doping efficiency, producing only 4.0 mol% Spiro˙⁺. The increase of Spiro˙⁺ helps to improve the conductivity and hole mobility of the Spiro film, providing strong support for the preparation of the efficient PSCs.

Based on the above research results, it was found that unlike the Spiro doping process with the Li-TFSI/tBP system that requires oxygen and light, Spiro can be doped with the Li-TFSI/12-crown-4 system under anaerobic conditions, illustrating that the introduction of 12-crown-4 provides a new doping pathway for doping Spiro with Li-TFSI.¹⁶ As shown in Fig. 2, there are two doping pathways after the introduction of 12-crown-4. The first pathway follows a similar conventional doping path triggered by light and oxygen. Initially, Spiro reacts with oxygen, followed by indirect doping by Li-TFSI, forming stable Spiro˙⁺. Different from the conventional doping path, after participating in the doping process, Li-TFSI no longer forms a by-product of lithium oxide, but instead chelates with 12-crown-4, forming the Li⁺(12-crown-4)TFSI⁻ complex. This process not only avoids the generation of the by-product lithium oxide but also suppresses the migration of Li⁺. The second pathway involves the formation of the Li⁺(12-crown-4)TFSI⁻ complex through chelation between 12-crown-4 and Li⁺. Based on the oxidation potentials of the cyclic voltammetry (CV) measurements in Fig. S3 (ESI†), the HOMO energy levels of Spiro and Li⁺(12-crown-4)TFSI⁻ can be calculated to be −4.99 and −5.43 eV, respectively. Li⁺(12-crown-4)TFSI⁻ has a lower HOMO energy level, demonstrating the strong electron-accepting capability, which can directly induce electrons from Spiro to Li⁺(12-crown-4)TFSI⁻, forming Spiro˙⁺ stabilized by TFSI⁻. The simultaneous occurrence of these two pathways allows more Li-TFSI to participate in doping Spiro, thereby generating more Spiro˙⁺ and ultimately enhancing the doping efficiency, which supports the above research results. Meanwhile, the unique chelating effect of 12-crown-4 also suppresses stability issues caused by excessive Li-TFSI introduction.

Fig. 2 Oxidation mechanism of Spiro doped with the Li-TFSI/12-crown-4.

The hole transport characteristics of the doped Spiro films were investigated using the space charge limited current (SCLC) analysis under dark conditions.¹⁷ The current–voltage (I–V) characteristics of devices with the ITO/PEDOT:PSS/perovskite/Spiro/Au structure are displayed in Fig. 1h, and hole mobility values were calculated using the modified Mott–Gurney law (Scheme S2, ESI†). The hole mobility of the pure hole device doped with 80 mol% Li-TFSI/12-crown-4 is 2.355 × 10⁻⁴ cm² V⁻¹ s⁻¹, while that of the control group is only 1.17 × 10⁻⁴ cm² V⁻¹ s⁻¹. Moreover, the electrical conductivities of the doped Spiro films can be examined by measuring the I–V characteristics of devices with the ITO/Spiro/Au structure. As shown in Fig. 1i, as expected, the calculated electrical conductivity of the Spiro film doped with 80 mol% Li-TFSI/12-crown-4 is 1.56 × 10⁻⁵ S cm⁻¹, higher than that of the control group (1.2 × 10⁻⁵ S cm⁻¹).¹⁸ The significant enhancement in hole mobility and conductivity confirms that 12-crown-4 can indeed promote the doping efficiency of Spiro.

To clarify the regulation of Spiro energy levels, the ultraviolet photoelectron spectroscopies (UPS) of the Spiro films doped with 80 mol% Li-TFSI/12-crown-4 and 50 mol% Li-TFSI/tBP were carried out. Each film was exposed to oxygen and light for 20 days (50 mol% Li-TFSI/tBP doped films contained 4.0 mol% Spiro˙⁺ and 80 mol% Li-TFSI/12-crown-4 doped films contained 7.9 mol% Spiro˙⁺). As shown in Fig. 3a and b, the secondary electron cut-off edge (E_cut-off) of the control group is 17.27 eV, while the E_cut-off of the Spiro film doped with 80 mol% Li-TFSI/12-crown-4 shifts towards lower binding energy by 0.49 eV, to 16.78 eV. According to the formula:

W_F = 21.22 − E_cut-off (3)

the work functions of the control group and Spiro:80 mol% Li-TFSI/12-crown-4 films are calculated to be 3.95 eV and 4.42 eV,¹⁹ respectively. The increase in work functions indicates a better doping effect, possibly due to the increased solubility of Li-TFSI in chlorobenzene. Moreover, Li-TFSI/tBP can only dope Spiro under aerobic conditions, while Li-TFSI/12-crown-4 can dope Spiro under both aerobic and anaerobic conditions. The energy level alignment calculated from the UPS is shown in Fig. 3c. The HOMO levels of the control group and the 80 mol% Li-TFSI/12-crown-4 doped Spiro film are −5.14 eV and −5.19 eV, respectively. The deeper HOMO level of the Spiro film doped with 80 mol% Li-TFSI/12-crown-4 matches better with the valence band of the perovskite, which not only facilitates effective charge extraction, transportation, and collection, thereby enhancing the open-circuit voltage (V_oc) of the PSCs, but also suppresses the accumulation of interface charge carriers, thus reducing hysteresis phenomena.

Fig. 3 UPS spectra of the (a) control group and (b) Spiro:80 mol% Li-TFSI/12-crown-4 films. (c) Energy levels of the TiO₂, perovskite, Spiro:50 mol% Li-TFSI/tBP and Spiro:80 mol% Li-TFSI/12-crown-4 layers. (d) PL and (e) TRPL spectra of the control group and Spiro:80 mol% Li-TFSI/12-crown-4 films. (f) Nyquist plots of the control group and 80 mol% Li-TFSI/12-crown-4 doped PSCs.

To investigate the effect of the increasing Li-TFSI doping concentration on the hole extraction capability of the HTL, steady-state photoluminescence (PL) spectroscopy was conducted on three films: glass/PVK, glass/PVK/Spiro:50 mol% Li-TFSI/tBP, and glass/PVK/Spiro:80 mol% Li-TFSI/12-crown-4. As shown in Fig. 3d, the glass/PVK film without the HTL exhibits the strongest PL emission intensity at 764 nm, while it is significantly quenched upon capping with the HTLs. Compared with the glass/PVK/Spiro:50 mol% Li-TFSI/tBP film, the glass/PVK/Spiro:80 mol% Li-TFSI/12-crown-4 film shows the obvious fluorescence quenching, demonstrating the stronger hole extraction ability, indicating that the Spiro:80 mol% Li-TFSI/12-crown-4 film can extract and transport photogenerated holes from the perovskite layer more efficiently. Furthermore, the carrier lifetimes of these three films were further characterized using time-resolved photoluminescence (TRPL) (Fig. 3e). The test results were fitted according to a double exponential model:

where τ₁ and τ₂ represent the short and long lifetimes, A₁ and A₂ represent the decay amplitudes, and y₀ is a constant related to the baseline offset. The double exponential PL decay process consists of fast and slow decay processes, where the fast decay process is associated with charge carrier quenching from the perovskite to the carrier transport layer, while the slow decay process is related to radiative recombination in the perovskite. Various parameters are presented in Table S1 (ESI†). The average carrier lifetime τ_ave is calculated using τ₁ and τ₂. Compared with the bare perovskite (τ_ave = 84.17 ns), the τ_ave value of the perovskite capped with the Spiro:50 mol% Li-TFSI/tBP film has decreased to 31.32 ns.²⁰ In the perovskite capped with the Spiro:80 mol% Li-TFSI/12-crown-4 film, the τ_ave has been a more pronounced decline, decreasing to 19.09 ns. A shorter fluorescence lifetime signifies that the Spiro:80 mol% Li-TFSI/12-crown-4 film has better hole extraction capability, which is consistent with the PL results. In addition, electrochemical impedance spectroscopy (EIS) was also used to study the charge transfer at the interface of the PSCs.²¹ As shown in Fig. 3f, the equivalent circuit reveals two main interface resistance, that is, series resistance (R_s) and charge transport resistance (R_tr). Compared with the control PSCs based on Spiro:50 mol% Li-TFSI/tBP (R_tr = 98.5 Ω), the PSC based on Spiro:80 mol% Li-TFSI/12-crown-4 exhibits a smaller transport resistance (R_tr = 83.5 Ω) due to its lower HOMO energy level, which better matches the valence band of the perovskite. A smaller transfer impedance represents less resistance to hole extraction at the interface of the PSC, indicating higher hole extraction efficiency and faster carrier transfer at the interface of the perovskite and Spiro:80 mol% Li-TFSI/12-crown-4 film, which agrees with the results of TRPL results.

To understand the effect of the increasing Li-TFSI doping concentration on the photovoltaic performance of PSCs directly, a series of devices with the structure glass/ITO/SnO₂/perovskite/Spiro/Au were prepared. The cross-sectional scanning electron microscopy (SEM) image of the entire PSC based on Spiro:80 mol% Li-TFSI/12-crown-4 is shown in Fig. 4a, demonstrating that the smooth and uniform perovskite with an average thickness of about 500 nm is fully covered by the HTL, providing an efficient and convenient pathway for hole transport from the perovskite layer to the HTL. Fig. 4b displays the J–V characteristic curves measured under AM 1.5G standard illumination for PSCs with different Li-TFSI doping concentrations, and the relevant parameters are listed in Table S2 (ESI†). When the Li-TFSI doping concentration reaches 80 mol%, the PSC achieves the optimal PCE. A further increase in concentration leads to a decreasing trend in PCE. As shown in Fig. 4c, the champion PSC based on Spiro:80 mol% Li-TFSI/12-crown-4 exhibits a best PCE of 23.99% with a V_oc of 1.167 V, a short-circuit current density (J_sc) of 24.49 mA cm⁻² and a fill factor (FF) of 0.839, whereas the control group only achieves a PCE of 22.29% (V_oc = 1.141 V, J_sc = 24.02 mA cm⁻², and FF = 0.813).

Fig. 4 (a) The cross-sectional SEM of the entire PSC based on Spiro:80 mol% Li-TFSI/12-crown-4. (b) J–V curves of the PSCs with different concentrations of Li-TFSI. (c) J–V curves of the PSCs based on Spiro:80 mol% Li-TFSI/12-crown-4 and the control group. (d) The PCE, J_sc, FF and V_oc distribution data of the 20 PSCs. (e) External quantum efficiency (EQE) spectra and integrated current densities of the PSCs. (f) Steady-state PCE and J_sc outputs of the PSCs. (g) J–V curves of the PSCs based on Spiro:80 mol% Li-TFSI/12-crown-4 under forward and reverse scans.

Fig. 4d also displays histograms of the PCE, J_sc, FF, and V_oc distributions for 20 PSCs based on 80 mol% Li-TFSI/12-crown-4, which provides a visual comparison of the photovoltaic parameters. Obviously, compared with the control PSCs, the PSCs based on Spiro:80 mol% Li-TFSI/12-crown-4 exhibit the higher V_oc, J_sc, FF, PCE and the better reproducibility due to the energy level matching, higher conductivity, higher hole extraction and faster carrier transfer. The external quantum efficiency (EQE) spectra of the PSCs based on Spiro: 80 mol% Li-TFSI/12-crown-4 and Spiro:50 mol% Li-TFSI/tBP were examined to confirm the accuracy of J_sc values of the PSCs. As shown in Fig. 4e, the calculated integrated J_sc value of the PSCs based on Spiro:80 mol% Li-TFSI/12-crown-4 is 23.88 mA cm⁻², which is in good agreement with the J_sc value obtained from the experimental J–V measurement. Moreover, the PSCs based on Spiro:80 mol% Li-TFSI/12-crown-4 exhibit higher EQE than the control PSCs across almost the entire visible light spectrum. Additionally, to precisely analyze the output performance of the PSCs, the steady-state output at the fixed maximum power point (MPP) under 1 sun illumination for 400 s is shown in Fig. 4f. The PSCs based on Spiro:80 mol% Li-TFSI/12-crown-4 exhibited a stable efficiency of 23.20% and a stable J_sc output of 23.70 mA cm⁻², while the control PSCs showed an output efficiency of 21.04% and a stable J_sc output of 22.77 mA cm⁻². Compared with the control PSCs, the PSCs based on Spiro:80 mol% Li-TFSI/12-crown-4 has a faster photoelectric response and more stable output capability. Besides, the existence of hysteresis effects the PCE accuracy of the PSCs. Under the reverse and forward scanning, the hysteresis of the PSCs based on Spiro:80 mol% Li-TFSI/12-crown-4 is smaller and can be ignored, as displayed in Fig. 4g and Fig. S4 (ESI†). Therefore, these results demonstrate that the Li-TFSI/12-crown-4 system has more potential to improve the PSC performance and process controllability.

Stability is a crucial factor in evaluating the performance of the PSCs. Therefore, based on the ISOS protocol,²² the stability of the unencapsulated PSCs was investigated in three ways: (1) the PSCs were aged at room temperature and RH = 25 ± 5% (ISOS-D-1) for 30 days, and the results are shown in Fig. 5a. The PCE of the PSCs based on Spiro:80 mol% Li-TFSI/12-crown-4 maintained 93% of the initial PCE, while the PCE of the control PSCs retained only 84% of the initial PCE. (2) The PSCs were aged at room temperature and RH = 85% for 30 days. As shown in Fig. 5b, the PSCs based on Spiro:80 mol% Li-TFSI/12-crown-4 maintained 88.6% of its initial PCE after aging, whereas the control PSCs only retained 68% of its initial PCE under the same aging conditions. (3) The PSCs were aged in an environment with 85 °C and RH = 85% (ISOS-D-3) to test the adaptability of the PSCs to extreme conditions. In Fig. 5c, the PSCs based on Spiro:80 mol% Li-TFSI/12-crown-4 still maintained 83% of the initial PCE after aging for 30 days, while the control PSCs retained only 57% of the initial PCE under the same aging conditions. The above stability results indicate that the stability of the PSCs based on Spiro:80 mol% Li-TFSI/12-crown-4 is significantly improved.

Fig. 5 Normalized PCE variations after aging at (a) room temperature and RH = 25 ± 5%, (b) room temperature and RH = 85%, and (c) 85 °C and RH = 85%. SEM images of the (d) fresh and (f) aged perovskite films covered with Spiro:50 mol% Li-TFSI/tBP and (e) fresh and (g) aged perovskite films covered with Spiro:80 mol% Li-TFSI/12-crown-4. AFM images of the perovskite films covered with (h) Spiro:50 mol% Li-TFSI/tBP and (i) Spiro:80 mol% Li-TFSI/12-crown-4. XRD results of the perovskite films covered with (j) Spiro:50 mol% Li-TFSI/tBP and (k) Spiro:80 mol% Li-TFSI/12-crown-4 after aging at room temperature and RH = 25 ± 5%.

Next, we further explored the reasons for improving the stability of the PSCs based on Spiro:80 mol% Li-TFSI/12-crown-4. Firstly, we conducted time-dependent UV-vis measurements on PVK/HTL samples under the irradiation and aerobic environment, to demonstrate the stability of the HTL based on the Li-TFSI/12-crown-4 doping system. As shown in Fig. S5 (ESI†), the absorption intensity within the range of 450–550 nm (corresponding to the characteristic peak of Spiro radicals) of the PVK/HTL sample based on Li-TFSI/tBP significantly decreases with irradiation time, which is due to the reduction of Spiro radicals caused by the dedoping of Spiro. In contrast, the absorption intensity of the Spiro radical characteristic peak in the PVK/HTL sample based on Li-TFSI/12-crown-4 remains almost unchanged with irradiation time, indicating that Li-TFSI/12-crown-4 doped HTL has good stability. Next, the optical photographs of perovskite films covered with different HTLs were directly observed after 30 days of aging (ISOS-D-1). As shown in Fig. S6 (ESI†), it was found that the perovskite film covered with Spiro:50 mol% Li-TFSI/tBP exhibited numerous pinholes, while the perovskite film covered with Spiro:80 mol% Li-TFSI/12-crown-4 showed no obvious pinholes, which is basically the same as before aging.

Furthermore, water contact angle tests over time were conducted to observe the hydrophobicity of the Spiro films, and the results are shown in Fig. S7 (ESI†). Compared to the Spiro:50 mol% Li-TFSI/tBP film (83.2°), the Spiro:80 mol% Li-TFSI/12-crown-4 film exhibited a greater water contact angle (90.2°). After being exposed to the ambient environment for 30 minutes, the water contact angle of the Spiro:80 mol% Li-TFSI/12-crown-4 film remained at 74.1°, while the Spiro:50 mol% Li-TFSI/tBP film decreased to 60.8°. It demonstrates that the Spiro: 80 mol% Li-TFSI/12-crown-4 film exhibited the better hydrophobicity, effectively protecting the perovskite film from damage caused by water molecules.

Furthermore, the microstructures of the Spiro films were studied using scanning electron microscopy (SEM) and atomic force microscopy (AFM). As shown in the SEM images of Fig. 5d–g, pinholes and white particulates appeared on the surface of the perovskite film covered with Spiro:50 mol% Li-TFSI/tBP after aging for 30 days at room temperature and RH = 25 ± 5%, which may be due to the volatilization of tBP, migration and precipitation of Li⁺, resulting in the decomposition of the Spiro film and ultimately leading to poor long-term stability of PSCs. In contrast, there were almost no white particles or pinholes in the perovskite film covered with Spiro:80 mol% Li TFSI/12-crown-4, which remained almost unchanged from before aging. This result indicates that the 12-crown-4 can chelate with Li⁺, inhibiting the hygroscopicity of Li⁺ and preventing Li⁺ precipitation. On the other hand, replacing tBP with 12-crown-4 eliminates adverse effects such as pinholes caused by tBP volatilization. Moreover, the hydrophobicity of the 12-crown-4 skeleton can further enhance the hydrophobicity of the HTL, preventing water molecules from entering the perovskite film. As shown in the AFM images of Fig. 5h and i, the root mean square roughness (RMS) of the perovskite films covered with Spiro:50 mol% Li-TFSI/tBP and Spiro:80 mol% Li-TFSI/12-crown-4 is 9.09 nm and 7.53 nm, respectively. In Fig. S8 (ESI†), observation of the cross-sections of the Spiro films reveals that the Spiro:80 mol% Li-TFSI/12-crown-4 film is smoother, which can enhance contact with the perovskite, improve charge carrier transport efficiency, and reduce the attachment points for moisture and oxygen in the air, thus ensuring the long-term stability of the PSC.

The stabilities of the perovskite films covered with Spiro:50 mol% Li-TFSI/tBP and Spiro:80 mol% Li-TFSI/12-crown-4 were further investigated using XRD. As shown in Fig. 5j and k, all the films show the typical diffraction peaks of the perovskite and a dominant peak at 14.07° is assigned to the characteristic (110) lattice planes. After aging for 30 days, the perovskite film covered with Spiro:50 mol% Li-TFSI/tBP exhibited a significant PbI₂ characteristic peak at 12.61°, while only a weak PbI₂ characteristic peak was observed in the perovskite film covered with Spiro:80 mol% Li-TFSI/12-crown-4, indicating that the presence of 12-crown-4 significantly inhibits perovskite degradation.

Therefore, the above results showed that due to the ability of 12-crown-4 to suppress the hygroscopicity of Li⁺, eliminate pinholes caused by tBP volatilization, and the hydrophobicity of the skeleton, the Spiro:80 mol% Li-TFSI/12-crown-4 film can effectively block the intrusion of water molecules and protect the perovskite, exhibiting the good stability of the Spiro film doped with 80 mol% Li-TFSI/12-crown-4, so that the PSCs based on Spiro:80 mol% Li-TFSI/12-crown-4 has good long-term stability.

3. Conclusions

In summary, due to its unique cavity structure, crown ethers can chelate with alkali metal ions to form host–guest complexes, promoting the dissolution of alkali metal salts. Therefore, introducing 12-crown-4 into Spiro solution can chelate with Li⁺ in Li-TFSI, improve the solubility of Li-TFSI in chlorobenzene, and thus increase the doping concentration from 50 mol% to 80 mol% of Li-TFSI in the HTL. Moreover, the formed host–guest complex Li⁺(12-crown-4)TFSI⁻ has the strong electron accepting ability and can directly dope Spiro, thereby producing more Spiro˙⁺. The above two aspects led to an increase in the electrical conductivity and charge carrier mobility of the HTL, ultimately achieving an improvement in the doping efficiency of Spiro. Finally, the PSCs based on Spiro:80 mol% Li-TFSI/12-crown-4 achieved a PCE of 23.99%, which is better than the control PSCs (22.29%). The chelation of 12-crown-4 with Li⁺ can also inhibit the hygroscopicity of Li⁺ in Li-TFSI, eliminate pinholes caused by tBP volatilization, and couple with a hydrophobic skeleton, which can effectively prevent the invasion of water molecules and protect the perovskite. Therefore, the long-term stability of the PSCs based on Spiro:80 mol% Li-TFSI/12-crown-4 is significantly improved, retaining 83% of the initial PCE after aging for 30 days under the ISOS-D-3 protocol, while the control PSCs retained only 57% of the initial PCE under the same aging conditions. Our research results indicate that increasing the concentration of Li-TFSI improves the doping efficiency of Spiro, providing an effective strategy for improving the efficiency and stability of the PSCs.

Data availability

The data supporting this article have been included in the manuscript.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The authors are grateful to the National Natural Science Foundation of China (62374029, 22175029 and 62474033), the Sichuan Science and Technology Program (Grant No. 2024NSFSC0250), the Foundation of China Petroleum & Chemical Corporation (Grant No. 30000000-23-ZC0607-012736850000-23-ZC0607-0045), the Natural Science Foundation of Shenzhen Innovation Committee (Grant No. JCYJ20210324135614040) and the Fundamental Research Funds for the Central Universities of China (ZYGX2021J010 and ZYGX2019Z007) for financial support.

References

1. (a) J. Burschka, N. Pellet, S. J. Moon, R. Humphry-Baker, P. Gao, M. K. Nazeeruddin and M. Grätzel, Nature, 2013, 499, 316; (b) A. Kojima, K. Teshima, Y. Shirai and T. Miyasaka, J. Am. Chem. Soc., 2009, 131, 6050; (c) P. J. Shi, Y. Ding, B. Ding, Q. Y. Xing, T. Kodalle, C. M. Sutter-Fella, I. Yavuz, C. L. Yao, W. Fan, J. Z. Xu, Y. Tian, D. Y. Gu, K. Zhao, S. Tan, X. Zhang, L. B. Yao, P. J. Dyson, J. L. Slack, D. R. Yang, J. J. Xue, M. K. Nazeeruddin, Y. Yang and R. Wang, Nature, 2023, 620, 323; (d) W. S. Yang, B. W. Park, E. H. Jung, N. J. Jeon, Y. C. Kim, D. U. Lee, S. S. Shin, J. Seo, E. K. Kim, J. H. Noh and S. I. Seok, Science, 2017, 356, 1376; (e) J. J. Yoo, G. Seo, M. R. Chua, T. G. Park, Y. L. Lu, F. Rotermund, Y. K. Kim, C. S. Moon, N. J. Jeon, J. P. Correa-Baena, V. Bulovic, S. S. Shin, M. G. Bawendi and J. Seo, Nature, 2021, 590; (f) W. Zhang, G. E. Eperon and H. J. Snaith, Nat. Energy, 2016, 1, 16048.
2. (a) C. Cetin, P. Chen, M. M. Hao, D. X. He, Y. Bai, M. Q. Lyu, J. H. Yun and L. Z. Wang, Adv. Sustainable Syst., 2018, 2, 1800032; (b) M. He, D. J. Zheng, M. Y. Wang, C. J. Lin and Z. Q. Lin, J. Mater. Chem. A, 2014, 2, 5994; (c) M. C. Jung, S. R. Raga, L. K. Ono and Y. B. Qi, Sci. Rep., 2015, 5, 9863; (d) R. Li, M. N. Liu, S. K. Matta, A. Hiltunen, Z. F. Deng, C. Wang, Z. C. Dai, S. P. Russo, P. Vivo and H. C. Zhang, Adv. Sustainable Syst., 2021, 5, 2100244; (e) G. S. You, L. Liu, J. Wang, M. Zhao, C. J. Zhao, X. Cai, J. Tang, F. S. Lu and T. G. Jiu, Adv. Sustainable Syst., 2021, 5, 2100107.
3. (a) Z. Hawash, L. K. Ono, S. R. Raga, M. V. Lee and Y. B. Qi, Chem. Mater., 2015, 27, 562; (b) L. K. Ono, P. Schulz, J. J. Endres, G. O. Nikiforov, Y. Kato, A. Kahn and Y. B. Qi, J. Phys. Chem. Lett., 2014, 5, 1374.
4. J. Salbeck, F. Weissortel and J. Bauer, Macromol. Symp., 1998, 125, 121.
5. (a) L. Caliò, M. Salado, S. Kazim and S. Ahmad, Joule, 2018, 2, 1800; (b) Y. S. Liu, Z. R. Hong, Q. Chen, H. J. Chen, W. H. Chang, Y. M. Yang, T. B. Song and Y. Yang, Adv. Mater., 2016, 28, 440; (c) D. Poplavskyy and J. Nelson, J. Appl. Phys., 2003, 93, 341.
6. (a) I. E. Jacobs and A. J. Moulé, Adv. Mater., 2017, 29, 1703063; (b) W. R. Zhao, J. M. Ding, Y. Zou, C. A. Di and D. B. Zhu, Chem. Soc. Rev., 2020, 49, 7210.
7. (a) A. Abate, T. Leijtens, S. Pathak, J. Teuscher, R. Avolio, M. E. Errico, J. Kirkpatrik, J. M. Ball, P. Docampo, I. McPherson and H. J. Snaith, Phys. Chem. Chem. Phys., 2013, 15, 2572; (b) H. J. Snaith and M. Grätzel, Appl. Phys. Lett., 2006, 89, 262114.
8. (a) U. B. Cappel, T. Daeneke and U. Bach, Nano Lett., 2012, 12, 4925; (b) Y. Saygili, H. S. Kim, B. W. Yang, J. J. Suo, A. B. Muñoz-Garcia, M. Pavone and A. Hagfeldt, ACS Energy Lett., 2020, 5, 1271; (c) F. Lamberti, T. Gatti, E. Cescon, R. Sorrentino, A. Rizzo, E. Menna, G. Meneghesso, M. Meneghetti, A. Petrozza and L. Franco, Chem, 2019, 5, 1806; (d) J. Y. Seo, S. Akin, M. Zalibera, M. A. R. Preciado, H. S. Kim, S. M. Zakeeruddin, J. V. Millc and M. Grätzel, Adv. Funct. Mater., 2021, 31, 2102124.
9. (a) Y. Cho, H. D. Kim, J. H. Zheng, J. M. Bing, Y. Li, M. Zhang, M. A. Green, A. Wakamiya, S. J. Huang, H. Ohkita and A. W. Y. Ho-Baillie, ACS Energy Lett., 2021, 6, 925; (b) Z. Li, J. Tinkham, P. Schulz, M. J. Yang, D. H. Kim, J. Berry, A. Sellinger and K. Zhu, Adv. Energy Mater., 2017, 7, 1601451.
10. (a) E. Kasparavicius, A. Magomedov, T. Malinauskas and V. Getautis, Chem. – Eur. J., 2018, 24, 9910; (b) Z. Li, C. X. Xiao, Y. Yang, S. P. Harvey, D. H. Kim, J. A. Christians, M. J. Yang, P. Schulz, S. U. Nanayakkara, C. S. Jiang, J. M. Luther, J. J. Berry, M. C. Beard, M. M. Al-Jassim and K. Zhu, Energy Environ. Sci., 2017, 10, 1234; (c) B. Xu, Z. L. Zhu, J. B. Zhang, H. B. Liu, C. C. Chueh, X. S. Li and A. K. Y. Jen, Adv. Energy Mater., 2017, 7, 1700683.
11. S. Wang, M. Sina, P. Parikh, T. Uekert, B. Shahbazian, A. Devaraj and Y. S. Meng, Nano Lett., 2016, 16, 5594.
12. R. L. Forward, K. Y. Chen, D. M. Weekes, D. J. Dvorak, Y. Cao and C. P. Berlinguette, ACS Energy Lett., 2019, 4, 2547.
13. (a) K. S. Diao, L. J. Bai and H. J. Wang, Comput. Theor. Chem., 2011, 964, 18; (b) F. Kaneko, N. Kashihara, T. Tsuchida and K. Okuyama, Macromol. Rapid Commun., 2010, 31, 554; (c) S. Oshima, N. Hirayama, K. Kubono, H. Kokusen and T. Honjo, Anal. Chim. Acta, 2001, 441, 157; (d) K. Reuter, M. R. Buchner, G. Thiele and C. von Hänisch, Inorg. Chem., 2016, 55, 4441; (e) Y. Shen, K. M. Deng, Q. H. Chen, G. Gao and L. Li, Adv. Mater., 2022, 34, 2200978; (f) A. Abate, T. Leijtens, S. Pathak, J. Teuscher, R. Avolio, M. E. Errico, J. Kirkpatrik, J. M. Ball, P. Docampo, I. McPherson and H. J. Snaith, Phys. Chem. Chem. Phys., 2013, 15, 2572; (g) R. V. Hangarge, D. D. La, M. Boguslavsky, L. A. Jones, Y. S. Kim and S. V. Bhosale, ChemistrySelect, 2017, 2, 11487; (h) J. Lu, Y. Y. Qin, Q. Zhang, Y. L. Wu, J. Y. Cui, C. X. Li, L. Wang and Y. S. Yan, Appl. Surf. Sci., 2018, 427, 931; (i) M. Shamsipur and T. Madrakian, J. Coord. Chem., 2000, 52, 139; (j) Z. Y. Lu, Y. Guo, S. W. Zhang, S. C. Wu, R. W. Meng, S. Hong, J. X. Li, H. Y. Xue, B. Y. Zhang, D. H. Fan, Y. F. Zhang, C. Zhang, W. Lv and Q. H. Yang, Adv. Mater., 2021, 33, 2101745.
14. (a) Z. Q. Wan, H. Lu, J. Y. Yang, Y. P. Zhang, F. Y. Lin, J. X. Xia, X. J. Yao, J. S. Luo and C. Y. Jia, J. Energy Chem., 2022, 74, 489; (b) N. Alizadeh, Spectrochim. Acta, Part A, 2011, 78, 488.
15. T. M. Koh, S. Dharani, H. R. Li, R. R. Prabhakar, N. Mathews, A. C. Grimsdale and S. G. Mhaisalkar, ChemSusChem, 2014, 7, 1909.
16. Y. X. Zhang, B. Huang, M. Hu, B. Tan, F. Z. Huang, Y. B. Cheng, A. N. Simonov and J. F. Lu, J. Mater. Chem. A, 2022, 10, 10604.
17. T. Kirchartz, Beilstein J. Nanotechnol., 2013, 4, 180.
18. (a) X. F. Lin, A. N. Jumabekov, N. N. Lal, A. R. Pascoe, D. E. Gomez, N. W. Duffy, A. S. R. Chesman, K. Sears, M. Fournier, Y. P. Zhang, Q. L. Bao, Y. B. Cheng, L. Spiccia and U. Bach, Nat. Commun., 2017, 8, 613; (b) B. Tan, S. R. Raga, A. S. R. Chesman, S. O. Fürer, F. Zheng, D. P. McMeekin, L. C. Jiang, W. X. Mao, X. F. Lin, X. M. Wen, J. F. Lu, Y. B. Cheng and U. Bach, Adv. Energy Mater., 2019, 9, 1901519.
19. (a) F. Ali, C. Roldán-Carmona, M. Sohail and M. K. Nazeeruddin, Adv. Energy Mater., 2020, 10, 2002989; (b) B. de Boer, A. Hadipour, R. Foekema, T. van Woudenbergh, M. M. Mandoc, V. D. Mihailetchi and P. W. M. Blom, Presented at Conference on Organic Optoelectronics and Photonics, Strasbourg, FRANCE, 2004.
20. F. Ansari, E. Shirzadi, M. Salavati-Niasari, T. LaGrange, K. Nonomura, J. H. Yum, K. Sivula, S. M. Zakeeruddin, M. K. Nazeeruddin, M. Grätzel, P. J. Dyson and A. Hagfeldt, J. Am. Chem. Soc., 2020, 142, 11428.
21. D. Wei, F. S. Ma, R. Wang, S. Y. Dou, P. Cui, H. Huang, J. Ji, E. D. Jia, X. J. Jia, S. Sajid, A. M. Eiseman, L. H. Chu, Y. F. Li, B. Jiang, J. Qiao, Y. B. Yuan and M. C. Li, Adv. Mater., 2018, 30, 1707583.
22. M. V. Khenkin, E. A. Katz, A. Abate, G. Bardizza, J. J. Berry, C. Brabec, F. Brunetti, V. Bulovic, Q. Burlingame, A. Di Carlo, R. Cheacharoen, Y. B. Cheng, A. Colsmann, S. Cros, K. Domanski, M. Dusza, C. J. Fell, S. R. Forrest, Y. Galagan, D. Di Girolamo, M. Graetzel, A. Hagfeldt, E. von Hauff, H. Hoppe, J. Kettle, H. Koebler, M. S. Leite, S. Liu, Y. L. Loo, J. M. Luther, C. Q. Ma, M. Madsen, M. Manceau, M. Matheron, M. McGehee, R. Meitzner, M. K. Nazeeruddin, A. F. Nogueira, C. Odabasi, A. Osherov, N. G. Park, M. O. Reese, F. De Rossi, M. Saliba, U. S. Schubert, H. J. Snaith, S. D. Stranks, W. Tress, P. A. Troshin, V. Turkovic, S. Veenstra, I. Visoly-Fisher, A. Walsh, T. Watson, H. B. Xie, R. Yildirim, S. M. Zakeeruddin, K. Zhu and M. Lira-Cantu, Nat. Energy, 2020, 5, 35.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4tc05482b

---