LAUSR

photovoltaic Effect? Haiyang Zou,† Chunli Zhang,† Hao Xue,† Zhiyi Wu,† and Zhong Lin Wang*,†,‡ †School of Materials Science and Engineering, Georgia Institute of Technology, Atlanta, Georgia 30332-0245, United States ‡Beijing Institute of Nanoenergy and Nanosystems, Chinese Academy of Sciences, Beijing 100083, P.R. China The photovoltaic effect in p−n junctions has been widely studied and used to work as renewable energy sources for portable devices and various appliances. Due to the band gap energy of semiconductors in p−n junctions, carriers are excited by light and then are separated by an internal electric field built in the interface of the junctions, so that the light can be efficiently absorbed and converted into electricity. Nevertheless, the open-circuit voltage is limited by the band gap energy of the semiconductors. The bulk photovoltaic (BPV) effect, caused by the asymmetric distribution of photoexcited carriers in momentum space, can generate a large photovoltage under uniform illumination, which is far beyond its band gap of the corresponding semiconductor. However, the power generation efficiency is very low due to the unusually low short-circuit current, and the materials that exhibit the BPV effect are generally wide-band-gap noncentrosymmetric materials. Yang et al. reported the flexo-photovoltaic (FPV) effect in which the BPV effect can be induced by strain gradients in any semiconductor through the flexoelectric effect. The authors claimed that the strain gradients create very large photovoltaic currents from centrosymmetric single crystals. A silicon-made atomic force microscopy (AFM) tip with a typical radius size of only 8 nm was pressed on a single crystal with a size of more than 0.5 cm× 0.5 cmwith a thickness of more than 0.5 mm to induce strain gradients on the bulk materials. They claimed “the current increases by more than a factor of 100 when the loading force increased from 1 to 18 μN” (actually, from Figure S12A, the current increased by only 3.39 times when the force increased from 1 to 15 μN). The nanosized AFM tip was treated as an ideal rigid spherical indenter, and the Hertzian model was applied to calculate the strain and strain gradient in order to confirm this AFM tip could generate the flexoelectric effect in bulk materials. The author claimed that the flexo-photovoltaic effect was discovered as the photocurrent has a dependence on crystallographic orientation and the light polarization with an amplitude of less than 2 pA. Above all, they believed this FPV effect played a dominating role in the increased photocurrent under strains. Here, we focus on four aspects to discuss the data, mechanisms, and conclusions presented in Yang’s paper and clarify some facts. First, the experimental design is unreasonable as many other issues involved are not properly addressed when the authors drew the conclusions. Second, the physical models adopted for the calculation analysis are irrational, and the results are based on unrealistic assumptions that far exceed the properties of any materials known. Third, the proposed physical model is based on speculation, and the experimental results do not fully support the physical model. Fourth, the statements made in the discussions and conclusions sections are misleading. Therefore, there are many questions and doubts to be clarified, and many statements need to be corrected. We provide some advice on both experimental methods and theoretical models toward the exact quantification of the FPV effect, which are fundamental and critical, to avoid misleading the scientific community and the public readers by the work from Yang and co-authors. More details are elaborated as follows. The Unreasonable Experimental Measurement and Design. Flexoelectricity effect is a coupling effect between electric polarization and strain gradient in centrosymmetric material, which was discovered several decades ago. Studies of flexoelectricity in solids have been scarce due to the extremely small magnitude of this effect in bulk samples. The current studies are mainly focused on materials/structures at the nanoscale as the large strain gradients in nanomaterials can lead to a strong flexoelectric effect. Strain gradients are inversely proportional to size, so thin films are the obvious place to look for flexoelectric effects. Because whenever the thickness exceeds some critical value, the materials may relieve the stress by a formation of misfit dislocations or by twinning, whether the materials are single crystals or not. Yang et al. applied a point force exerted by the tip of a custom-made photoelectric atomic force microscope on the bulk materials with a size of more than 0.5 cm× 0.5 cm with a thickness of more than 0.5 mm to induce strain gradients on the bulkmaterials. Yang et al. did not try to do any experiment to verify whether the flexoelectric effect has been induced or not or measure the strength of the flexoelectric effect induced by the strains. Such important data are missing, and we cannot conclude that the AFM tip can generate the flexoelectric effect on these materials. Every material and every sample is different, including various Young’s modules, surface roughness, defects, and it is inappropriate to guess that the effect can be obtained by comparing with other papers unless it has been verified experimentally. However, the polarization induced by the applied force has never been measured and verified. A conductive AFM probe was employed for both applying pressure and acting as a conductive electrode. Such a design involves many complicated problems and errors. Smooth surfaces, even those polished, are never perfectly flat on a microscopic scale due to asperities. They are rough, with sharp or rugged projections (Figure 1a,b). Single crystals are never atomically flat after the processes of manufacturing and cutting.