钙钛矿太阳能电池的性能和稳定性受到其结构、组成和光物理性质的纳米尺度变化影响。虽然可以通过组成工程、接触工程和表面钝化等策略提高其性能，但这些策略对电池性能和稳定性的影响在不同尺度上的具体效应仍不清晰。为了完整理解卤化物钙钛矿的性能和退化，开发能够在操作条件下测量完整设备的显微技术至关重要。
基于上述问题，Samuel D.Stranks团队通过多模态原位显微镜工具包，测量出和空间相关的关联纳米尺度的电荷传输损失、复合损失和化学组成。研究发现，具有最高宏观性能的设备在初始性能空间异质性方面最低，这是传统显微镜中缺失的关键联系。界面工程对于实现稳健的设备至关重要。一旦界面稳定，均匀化电荷提取和最小化局部功率转换效率变化的成分工程对于提高性能和稳定性至关重要。
研究发现钙钛矿可以容忍化学上的空间无序，但不能容忍电荷提取的空间无序。通过组成工程，可以实现对电荷提取和局部功率转换效率变化的最小化，从而提高性能和稳定性。
Fig. 1 | Device operando microscopy reveals DCDH solar cell performance is tolerant to even dramatic spatial optoelectronic and chemical heterogeneity. a, Schematic of perovskite solar cell under bias being illuminated either by a white light-emitting diode (LED) array for the luminescence measurements or monochromatic hard X-rays for the nXRF measurements. Note that the optical and X-ray measurements are not simultaneously acquired. b, Hyperspectral PL spectra (at VOC and VMPP) of regions marked in e. c, Comparison of electrical JV curve (red, red arrow points to corresponding y axis) and area-averaged optical JV curve (black, black arrow points to corresponding y axis) of DCDH solar cell. Grey shaded areas show the distribution of JV curves across the map. d, Optical JV curves of the marked regions in e. e, PL centre of mass (COM) energy plot of a region of a DCDH solar cell at VOC. f, Br:Pb map from the marked region in b extracted by nXRF. g, Internal VOC (Δµ). h,i, Optical short circuit current extraction efficiency (ΦPL(0 V)) (h) and optical PCE (VMPP✕ΦPL(VMPP)) of the same region as shown in e (i).
Fig. 2 | Local reductions in performance are evident in microscopic JV curves of DCDH perovskite solar cells after extended operation. a,b, Optical PCE maps of the same area of a fresh (a) and operated (b) DCDH solar cell after 100 h at VOC, 65 °C and 1 sun illumination. c,d, show Br:Pb ratio maps extracted from nXRF from the two regions marked in b after the 100 h of operation. e,f, Optical JV curves before (e) and after (f) ageing from the points marked in a and b. Solid lines are reverse scans, dashed lines are forward scans. g,h, PL spectra from the same marked areas before (g) and after (h) ageing.
Fig. 3 | Multimodal microscopy on DCTH perovskite solar cells reveals reduced device stability and increased microscopic phase segregation compared to DCDH analogues.a,b, Optical PCE maps of the same area of a fresh (a) and operated (b) DCTH solar cell after 100 h at VOC, 65 °C and 1 sun illumination. c,d, Br:Pb ratio (c) and PL centre of mass energy maps (d) extracted from the region marked in b after the 100 h of operation. e,f, Optical JV curves (e) before and after (f) operational stress from the points marked in a and b. Solid lines are reverse scans, dashed lines are forward scans. g, Normalized Br:Pb ratio histograms for a pristine sample (orange) and the marked region of the operated (red) sample. h,i, PL spectra extracted from the same marked areas before (h) and after (i) ageing. j, Optical PCE histogram for this DCTH solar cell before (orange) and after (red) operation. Arrows in g and j are guides to the eye from fresh to operated.
Fig. 4 | Interfacial chemistry and spatial PCE disorder predict performance and stability of mixed-cation, mixed-halide perovskite solar cells.a–c, Optical PCE maps of pristine control 2PACz/TCTH (a), Me-4PACz/TCTH (b) and 2PACz/TCTH + PI (c) passivation devices. d, Representative JV curves of the pristine interface-modified devices mapped in panels a–c. e, Optical PCE distributions and corresponding Gaussian fits of the interface-modified devices. The numbers over the distributions represent the FWHM of the distribution. f, Initial optical PCE disorder (FWHM of PCE distribution) versus initial PCE (mean of PCE distribution) for a range of perovskite devices, where each point is an individual device. Linear regression shows a Pearson’s r value of −0.71,Spearman’s r value of −0.77 and a P value of «0.01 (two-sided tests). Shaded regions represent the 95% confidence interval of the linear fit from a Student’s t distribution percent points function (n = 75 devices). g, JV curves after operational stress of the representative devices shown in d. h, Scatter plot of initial PCE disorder versus PCE loss (%) during operation for the perovskite composition series on 2PACz. (Device numbers for comparison are DCDH = 4, DCTH = 4, TCTH = 8). i, Scatter plot of initial PCE disorder versus PCE loss (%)during operation for the interfacial modification series. (Device numbers for comparison are 2PACz/TCTH = 11, MeO-2PACz = 4, Me-4PACz/TCTH = 14,2PACz/TCTH + PI = 8, 2PACz/TCTH+LiF=4). Solid markers are the mean of a given device type, semi-transparent markers are individual devices. Error bars in h and i are standard deviations.
本文来源：DOI: 10.1038/s41560-024-01660-1
https://doi.org/10.1038/s41560-024-01660-1