The Computational Simulation Of Alkali-metal Doping And Interface Characteristics In Cu-based Thin Film Solar Cells

Solar energy is an important renewable energy source in the 21st century.The development of cheap solar cells with high photoelectric conversion efficiency is very important for the utilization of solar energy.The thin film solar cells,i.e.,Cu（In,Ga）Se₂or Cu₂ZnSn（S,Se）₄,are typical representative of such solar cells whose cell conversion efficiencies currently reach 22.6%and 12.6%,respectively.Inspired by the greatly improved performance of Cu（In,Ga）Se₂ and Cu₂ZnSn（S,Se）₄ solar cells by the doping of alkali metal Na,people have also paid great attention to the doping of other alkali metals（such as K,Rb,etc.）in the solar cells.However,Li,as the lightest alkali metal,had received little attention,and its experimental research on doping is very scarce.Moreover,the existing few experimental results gave a complicated picture on Li doping.Therefore,the doping possibility and doping effects of Li in Cu（In,Ga）Se₂ and Cu₂ZnSn（S,Se）₄ solar cells remain to be explored.At the same time,alkali metal doping may introduce alkali-compounds in Cu（In,Ga）Se₂ and Cu₂ZnSn（S,Se）₄ thinfilms,and their influence on the absorber is also a problem worth exploring.Besides,Cu（In,Ga）Se₂ and Cu₂ZnSn（S,Se）₄ solar cells are typical heterojunction solar cells,where the band offset at the interface is one of the most essential characteristics distinct them from the homojunction solar cells.It is generally believed that the band offset is the only factor affecting the current flow,however,a large number of theoretical results indicate that there may be other influential factors besides the band offset.The clarification of this issue is very important for the design of high efficiency heterojunction solar cells.This thesis is divided into six chapters.In the first chapter,the research status about the doping of alkali metals in Cu（In,Ga）Se₂ and Cu₂ZnSn（S,Se）₄ solar cells are introduced.The second chapter mainly introduces the theoretical basis and calculation methods involved in this study.It includes two parts,namely the density functional theory of first-principles（DFT）calculation and the basic principle for the flow of current in solar cells.These theoretical foundations are important to the subsequent understanding of the research in this thesis.In the third chapter,we studied the doping possibility of Li in Cu（In,Ga）Se₂ and Cu₂ZnSn（S,Se）₄,and explained the complex experimental results.In this chapter,we first calculated the defect formation energy of lithium.It was found that formation energy is much lower than that of Na,and it would be even lower than zero under certain process conditions.This result theoretically proved the feasibility of doping Li.We further explored the reason why it is difficult for Li to be doped into the absorber layer in some experiments.We have found that due to the strong electropositive nature of lithium atoms,only Li₂Se and LiInSe₂ compounds can be formed during doping process.Since the melting point of these two compounds and the raw material LiF are higher than the actual process temperature,the Li atoms would be immobilized in the compound,and thus they are difficult to move and diffuse.By summarizing all the lithium doping processes in the experiments,we found that the effect of Li doping strongly depends on the process:if Li is mixed in the morphology of solution/powder in the raw material of the absorber layer during the synthesis process,the doping effect of Li is very good;if Li is doped into a synthesized absorber,the diffusion of Li is faint.The above results showed that an appropriate process route is required to achieve good Li-doping effects.In the fourth chapter,we studied the alkali-compounds that may be introduced in Cu（In,Ga）Se₂,Cu₂ZnSn（S,Se）₄,and their influence on the absorber layer.By calculating the energy band structures and analyzing the energy band components,we found that when an alkali-compound is formed via replacing of Cu atoms by alkali metal atoms,there would be no d orbitals of Cu in the compounds.As a result,there would be no anti-bonding state formed by hybridization of d orbital of Cu and p orbital of anions in the valence bands,and the maximum of valence-band of the compound is lower than the absorption layer,which induces larger band gap.We further calculated the optical properties of alkali-compounds,and found that in the high-density region of solar spectrum,their light absorption coefficients are very low.This feature means that the alkali-compounds do not affect the absorption of photons in the absorption layer,which also explained why the alkali metal doping can achieve such great success in Cu（In,Ga）Se₂ and Cu₂ZnSn（S,Se）₄ solar cells.The fifth chapter mainly focused on the current flow at the interface of heterojunction solar cells.Our calculation results showed that not only the band offset,but also the effective density of states in the conduction band of the buffer layer,temperature,and doping can affect the flow of current.Interestingly,all these parameters affect the concentration of majority carriers in the near-interface region of the buffer layer.Additionally,we found that the concentration of majority carriers in the near-interface region is a more fundamental factor than the conduction band offset affecting the current.This is because the low carrier concentration would reduce the conductivity in its region,which hinders the flow of photo-generated current and results in a decrease of current.Therefore,the band offset,the effective density of states in conduction band of the buffer layer,as well as doping can affect the current by changing the concentration of majority carriers.We also explored the effect of interface defects on current flow.We found that interface defects can also affect current by affecting the concentration of majority carriers at the near-interface region.The above studies make us have a more profound understanding on the flow of current at the interface in the heterojunction solar cells.On this basis,several approaches can be proposed to improve the current collection in the heterojunction.Chapter 6 summarized all the reserach and discusses prospect of this thesis.