摘要： Organolead halide perovskites MAPbX3 (MA = CH3NH3+; X = Cl?, Br?, I?) have attracted broad tremendous interest in the past 10 years for applications in solar cells and light-emitting devices. In evaluating the quality of the perovskite materials, spectroscopic characterizations such as static and time-resolved absorption and photoluminescence measurements are essential to examine their photophysical properties. A recent report found that the correct measurement of static absorption spectra of MAPbX3 films is indeed difficult due to the strong light scattering caused by their poor surface coverage or complex microstructures. These morphological complexities seem to be inevitable in thin-film fabrication and should not only affect the steady-state spectroscopic measurements but also can significantly impact the time-resolved spectroscopic characterizations, whose results are crucial for understanding photoinduced carrier dynamics in the examined materials. Photoexcited states in semiconductor materials induce changes in the real and imaginary parts of the dielectric function. This leads to changes in absorption (imaginary part) and reflectivity (real part), which can be substantial for materials with significant values of refractive index such as lead halide perovskites. Transient absorption (TA) spectroscopy is a typical technique that has been broadly used to probe photoexcited state dynamics in perovskites and other semiconductor materials. In TA measurements, a pump laser pulse is used to excite the perovskite films, and the induced absorption changes (ΔA) are recorded as a function of both wavelength and time. With the transmitted light as the probe (Figure 1a), the TA signal (ΔA) is mainly decided by the ratio of the intensity of transmitted probe light with and without pump excitation (see eq S1 in the SI), assuming that the loss of transmitted probe light completely results from the sample absorption. On the basis of the same experimental setup, transient reflection (TR) measurements can also be carried out by using the reflected probe light as detection signal (Figure 1b). The TR signal (ΔR/R) can also be determined by the ratio of the intensity of reflected probe light with and without pump excitation (see eq S4 in the SI). Unlike the TA measurements that mainly probe the bulk property of samples, the TR signal mainly detects the photoinduced reflection variations due to the refractive index change at the sample surface. Therefore, the TR spectrum and kinetics can be significantly different from those of TA even in the same sample. For example, previous TA and TR measurements have found dramatically faster carrier recombination kinetics on the surface than in the bulk of MAPbX3 perovskite films or single crystals because of the presence of more surface defects. There is an abnormal case in the regular TA measurements particularly when performed on the films with large and heterogeneous microstructures (e.g., films with poor coverage, large grains, and pinholes) because the loss of transmitted probe light in their TA measurements likely results not only from the sample absorption but also from the reflection of the film surface or the boundary of microstructures in samples. In this case, the measured transient spectrum, though collected in the transmittance mode as in TA, can contain contributions from both TA and TR signals (see Figure 1c and eq S6 in the SI). This could lead to distorted TA spectra and thus inaccurate analysis of photoinduced kinetics. A solution-processed organic or inorganic halide perovskite thin film is a typical material whose morphological microstructures were found to have significant impact on device performance. Although the photoinduced carrier dynamics in perovskite films has been extensively studied using TA spectroscopy, the possible artifacts in TA results induced by TR signal originating from the photoinduced reflectivity variation of film surfaces and microstructures have been overlooked. Herein, in order to clarify the influence of TR signal in the regular TA measurements, we performed a careful transient spectroscopic analysis on a series of MAPbBr3 perovskite films with different microstructure morphology. Meanwhile, TR measurements on MAPbBr3 single crystals (SCs) were carried out for comparison. We confirmed that the TA spectra measured in MAPbBr3 perovskite films with large and heterogeneous microstructures do comprise non-negligible TR signals from the photoinduced reflection of microstructures, with the weight of contribution increased from ～20 to ～100% as the size of the microstructure increased from <200 nm to 1?2 μm. The presence of TR signal leads to an “artifact” feature in the TA spectra and faster observed kinetics owing to the faster surface carrier recombination, which will thus mislead the analysis of bulk carrier dynamics. We also provided a method to reduce the TR signal in actual TA measurements by adding solvent with its refractive index close to the samples, by which the TR distortion can be suppressed to some extent.