Research highlights
1. By introducing two different PPPc technologies (i.e., CW- and ns-PPPc), the behavior of trapped carriers on the ns to ms time scale in working PeSCs is revealed for the first time;
2. A kinetic model was established to calculate the trap state density ratio between different devices, and it was found that the trap density after surface passivation was reduced by ~50 times. ;
3. Prove that interface issues play a dominant role in carrier recombination and transport in devices, and confirm that charge drift/diffusion to the interface will be affected by surface passivation.
1. The problem of using optics to study trap states in perovskite cells
Currently, the performance of perovskite solar cells (PeSC) is approaching the Shockley-Queisser limit of single cells, and its further development requires fundamentally reducing non-radiative recombination in PeSC devices. Electronic defects, called carrier "traps," are primarily responsible for nonradiative recombination under single solar excitation. Therefore, a deeper understanding of the nature of trap states and the dynamic behavior of carriers trapped in PeSCs is crucial to minimize voltage losses to further improve device performance. Traditional optical spectroscopy techniques make it difficult to selectively observe the dynamics of carrier trapping in solar cells. Thermal energy or infrared (IR) light can be used to activate or detect trapped carriers to generate trap information at picosecond and nanosecond scales. In the authors' previous work, the ultrafast trapping filling process in PbS solar cells was observed using femtosecond optical pump-infrared pulsed photocurrent (PPPc) technology, which is particularly sensitive to trapped carriers. Unfortunately, the limited time window does not facilitate the evaluation of slow trap filling and trap-assisted recombination processes.
2. Introduction to results
Artem A. Bakulin of Imperial College London, Ziming Chen's team and Gao Feng of Linköping University introduced ns-ms level infrared optical activation PPPc spectroscopy to overcome the limitations of existing spectroscopic techniques and reveal the dynamics of trapped carriers in working PeSCs. By utilizing photocurrent detection, PPPc can successfully overcome the limitations of traditional optical methods. The behavior of trapped carriers in untreated and surface-passivated FA0.99Cs0.01PbI3 (where FA = CH(NH2)2) perovskite solar cells was studied, and the corresponding trapped carrier dynamics were established. science and drift-diffusion models. The results show that the trap process mainly comes from traps in the perovskite body and traps at the interface with the charge (hole) extraction layer. Quantitative interface passivation significantly reduces the total number of trap sites in the device but does not change the energy of the traps. This work is the first report to fully and directly reveal the dynamics of trap filling and trap-assisted recombination processes in working perovskite solar cells.
3. Results and Discussion
Point 1: Experimental device and working mechanism of PPPc
In order to control the trap density and evaluate the effectiveness of defect reduction strategies in perovskite solar cells, FA0.99Cs0.01PbI3-based solar cells were prepared. The surface passivation device uses n-octylammonium iodide (OAI) to passivate the interface trap device between perovskite and Spiro-OMeTAD. The structure of the device is shown in Figure 1a. Compared with the untreated device, the surface-passivated device has a larger open circuit voltage (Voc), which indicates that the trap-assisted recombination on the perovskite surface is effectively suppressed, attributed to the interaction of OAI molecules on the FA+ and I− interfaces. Simultaneous passivation of vacancies. As shown in Figure 1c, the low-angle (2θ < 10°) X-ray diffraction signals shown in Supplementary Figure 1e, f show no evidence of the formation of a 2D perovskite phase on the 3D perovskite surface after OAI deposition.
Figure 1 FA0.99Cs0.01PbI3 device characteristics
Figure 2a shows the concept of the experimental setup for quasi-steady-state PPPc spectroscopy. The device has a very simple layout and consists of two diode lasers (pump and infrared pulse), beam overlap optics, intensity modulator, lock-in amplifier, and device holder. The pump and pulse beams overlap in space and are focused on the sample. A light intensity modulator modulates the infrared pulse beam and sends a reference signal to a lock-in amplifier to detect the infrared-induced photocurrent (rJIR). The optical intensity modulator can also be moved to the pump beam position to detect the pump-induced photocurrent (rJPump). The continuous wave-based PPPc setup (CW-PPPc) uses an 808 nm continuous wave laser as the pump beam and a 980 nm continuous wave laser as the pulsed beam, modulated by a light intensity modulator in the range of 37–4000 Hz, so that Enables access to information on trapped carriers on μs to ms time scales. Furthermore, ns-resolution PPPc (ns-PPPc) uses a pulsed laser of 800 nm (pulse duration approximately 40 fs) as the pump beam, and a pulsed laser of 1064 nm (time resolution approximately 0.5 ns) as the pulse beam, Transient information of trapped carriers on the ns to μs time scale has been achieved. The timing between pump and pulse is controlled by an electronic delay generator, triggering the pulse with an accuracy of ~10 ns.
Figure 1 FA0.99Cs0.01PbI3 device characteristics
Figure 2a shows the concept of the experimental setup for quasi-steady-state PPPc spectroscopy. The device has a very simple layout and consists of two diode lasers (pump and infrared pulse), beam overlap optics, intensity modulator, lock-in amplifier, and device holder. The pump and pulse beams overlap in space and are focused on the sample. A light intensity modulator modulates the infrared pulse beam and sends a reference signal to a lock-in amplifier to detect the infrared-induced photocurrent (rJIR). The optical intensity modulator can also be moved to the pump beam position to detect the pump-induced photocurrent (rJPump). The continuous wave-based PPPc setup (CW-PPPc) uses an 808 nm continuous wave laser as the pump beam and a 980 nm continuous wave laser as the pulsed beam, modulated by a light intensity modulator in the range of 37–4000 Hz, so that Enables access to information on trapped carriers on μs to ms time scales. Furthermore, ns-resolution PPPc (ns-PPPc) uses a pulsed laser of 800 nm (pulse duration approximately 40 fs) as the pump beam, and a pulsed laser of 1064 nm (time resolution approximately 0.5 ns) as the pulse beam, Transient information of trapped carriers on the ns to μs time scale has been achieved. The timing between pump and pulse is controlled by an electronic delay generator, triggering the pulse with an accuracy of ~10 ns.
Figure 2 Experimental device and working mechanism of PPPc
Figure 2b shows that when the sample is not illuminated or is illuminated only by the pump or pulsed beam, the R signal of the lock-in amplifier (i.e., the total infrared induced current) is negligible because the steady-state current from the pump (rJPump) is not are retrieved by a lock-in amplifier, and individual infrared pulse photons do not directly produce detectable photocarriers. When the device is irradiated by pump light and pulse light at the same time, the rJIR signal rises, indicating that rJIR comes from the intra-gap state filled by the moving carrier generated by the pump light. By comparing the signals from the lock-in amplifier, it was concluded that there was no significant change in phase at various chopper modulation frequencies. The amplitude of rJIR varies linearly with infrared light intensity, indicating that rJIR can be used as a measure of trapped carrier concentration. This linear relationship also indicates that only a small fraction of the trapped carriers are reduced by infrared photons. An infrared photoelectron beam can untrap carriers by two possible mechanisms: (i) by a direct optical transition from trap to band or (ii) by thermal activation, after the energy of the absorbed infrared photons increases the temperature of the film. In order to study the possible detrapment process caused by the infrared heating effect, the sample response to temperature changes was evaluated: Figure 2c shows that after heating, the photocurrent decreases, which indicates that the increase in r JIR signal does not originate from the infrared heating effect. In this case, the rJIR signal should originate entirely from the optical transition between the filled trap state and the band state, as shown in Figure 2d, which describes the hole process. After the pump beam is generated, band-edge free holes drift/diffuse to the anode and are occasionally trapped in trap states at the material bulk and interfaces. The trapped holes are immobile and many will recombine from these states, thus producing no photocurrent. Infrared pulse photons have the opportunity to reactivate these immobilized holes from the trap state to VB before recombination, thus generating additional photocurrent detected as rJIR. PPPc spectroscopy is therefore a highly sensitive and selective technique for monitoring carriers trapped in states that serve as recombination centers in devices.
Point 2: Trap carrier concentration and trap density
In the CW-PPPc measurements, the authors varied the intensity of the pump and pulse beams to study the differences in trapped carrier concentration and trap density in pristine and surface-passivated devices. Figure 3a and Supplementary Figure 5 show the intensity-dependent experimental results of surface passivation and pristine FA0.99CS0.01Pbl3. Since rJIR is not only related to the density of traps but also involves factors of charge extraction, the data are presented in the form of rJIR/rJPump to reflect the ratio between trapped carriers and free carriers. rJIR and rJIR/rJPump vary linearly with infrared intensity, indicating that pulsed photons only reduce the number of minority trapped carriers (nTC). Figure 3b shows that at the same pump intensity, the nTC of the pristine device is much higher than that of the surface-passivated device, indicating that more band-edge carriers are trapped in the pristine device. In addition, in both cases, nTC increases with the increase of pump intensity, and nTC and pump intensity follow a power law with the coefficient set to b.
Figure 3 Captured carrier concentration in pristine and surface passivated FA0.99Cs0.01PbI3 devices
Figure 3b shows that the extracted b-values for pristine and surface-passivated devices are 0.94 and 0.15, respectively. These results show that the well state of the surface passivated device has tended to be saturated when nTC is relatively small, while the well state of the original device is still far from saturated when nTC is large. This phenomenon can be attributed to the significant reduction in trap density in surface-passivated devices. Furthermore, considering that OAl passivates the top surface of the perovskite where holes accumulate, and the significant change in nTC before and after surface passivation, the authors believe that hole traps dominate the total trap state of the perovskite. The dependence of rJIR/rJPump on the infrared chopper modulation frequency is shown in Figure 3c. In order to determine the characteristic time scale of the process that generates the frequency-dependent signal, the data were fitted to the Cole-Cole equation commonly used to analyze chopper frequency-dependent measurements. combine.
To elucidate the energetics of the trap states and evaluate whether the passivation process has an impact on the characteristic trap depth, temperature-dependent CW-PPPc measurements were performed. The devices were measured in a nitrogen-filled environment to prevent potential degradation of the perovskite in vacuum conditions, and 3D showed that the total concentration of trapped carriers increases as the temperature decreases from 300 to 250 K. This phenomenon is due to the suppression of the thermal decapture process at low temperatures, resulting in more carriers staying in the gap. Fitting the Arrhenius model to the data, the main activation energy of the trap is ~280 meV. The trap activation energy values of the original device and the surface-passivated device are similar, indicating that the surface trap passivation process mainly reduces the density of traps and has little impact on their feature depth.
Key Point 3: ns-PPPc Reveals Defect Filling Dynamics
In order to reveal changes in the total number of trapped carriers, PPPc experiments were conducted using a synchronized pulse laser source. According to Figure 4a, the maximum amplitude of the nTC signal in the original device is one order higher than that in the surface passivated device, which is consistent with the results of CW-PPPc, indicating that the trap suppression by the OAl interface modulator is very large. The ns-PPPc dynamics show that nTC grows rapidly in both pristine and surface-passivated devices, indicating that considerable trap filling processes occur within the set time resolution (10 ns). Such a fast process may originate from the filling of traps by photocarriers generated near the trap state space, thus attributing this process to the filling of traps in both devices. An instantaneous initial increase in nTC is followed by a slow rise on the -100 ns time scale. The relative contribution of the slow rise (-60%) is much higher (-25%) in the original device compared to the surface-passivated device. Furthermore, the estimated rise time of nTC in the original device (142 ns) is significantly longer than that in the surface-passivated device (33 ns). This suggests that filling of trap states occurs more slowly in the original device, which may be related to a higher density of trap states, slower carrier diffusion, or the presence of different trap species, e.g., "bulk" and "surface" traps . Therefore, the instantaneous component is attributed to carrier trapping in the bulk material, while the delayed rise is attributed to carrier trapping at the charge extraction layer interface. Passivating the hole extraction side of the perovskite surface significantly reduces delayed trapping, again indicating that the majority of trapped carriers are holes. In this case, the change in the time scale of delayed PPPc growth may be related to the accumulation of trapped holes on the perovskite surface to form a positively charged interface charge layer. These interfaces in the original sample may shield the internal fields and slow down the drift/diffusion of carriers to the interface.
Figure 4 ns-PPPc and ns-TAS results for pristine and surface passivated devices
To further understand the charge diffusion/drift characteristics, ns-based transient absorption spectroscopy TAS (ns-TAS) was performed to monitor the average carrier concentration in the perovskite layer on the ns-μs time scale (Figure 4b). ns-TAS shows that the carrier concentration decay rate of surface-passivated perovskite (56.2 ns) is faster than that of pristine perovskite (156.2 ns). Attribute these time scales to carrier extraction because these time scales agree well with previously reported carrier extraction times based on time-resolved PL and photocurrent measurements. The observed time scale is also consistent with that of the delayed PPPc component, indicating that both describe the arrival of photogenerated carriers at the interface between the perovskite and the hole extraction layer. Furthermore, the recombination kinetics of the capture vector are similar in the original and surface-passivated samples, except that they are more pronounced in the longest component (-25 us) in the original sample.
Extended drift-diffusion simulations of the charge concentration dynamics at the hole extraction interface are performed. The extraction rate at the hole transport site in the pristine device is set smaller compared to the surface-passivated device, simulating the presence of a stronger positive interface charge layer in the pristine device. Figure 4c shows the change of nTC concentration at the simulated hole extraction interface with time, qualitatively and quantitatively reproducing the experimental ns-PPPc data. The average hole concentration of the active layer shown in Figure 4d is also in good agreement with the results of ns-TAS. The drift-diffusion model successfully reproduced the ns-PPPc and Ins-TAS data, confirming that interface trapping plays a dominant role in carrier recombination and transport in the device. It also confirms that charge drift/diffusion towards the interface can be affected by surface passivation.
4. Summary
The research results successfully demonstrate that PPPc technology is powerful and highly sensitive in revealing the dynamics, concentration, and activation energy of trapped carriers, which helps to comprehensively understand the role of trap states in optoelectronic devices. It provides the field with a new way to examine trap status. It is also thought that it could be combined with other experiments to provide more information on the operation of traps in working optoelectronic devices, such as device degradation under different circumstances.
5. References
Jiaxin Pan et al. Operando dynamics of trapped carriers in perovskite solar cells observed via infrared optical activation spectroscopy , Nature Commun.
Doi: 10.1038/s41467-023-43852-5 (2023).
E-mail: info@chemborun.com
Tel: +86-574-87178138 
No. 1558, Jiangnan Road,, Ningbo, Zhejiang, China (Mainland)/31
