Deducing transport properties of mobile vacancies from perovskite solar cell characteristics
James M. Cave,Nicola E. Courtier,Isabelle A. Blakborn,Timothy W. Jones,Dibyajyoti Ghosh,Kenrick F. Anderson,Liangyou Lin,Andrew A. Dijkhoff,Gregory J. Wilson,Krishna Feron,M. Saiful Islam,Jamie M. Foster,Giles Richardson,Alison B. Walker
DOI: https://doi.org/10.1063/5.0021849
IF: 2.877
2020-11-14
Journal of Applied Physics
Abstract:The absorber layers in perovskite solar cells possess a high concentration of mobile ion vacancies. These vacancies undertake thermally activated hops between neighboring lattice sites. The mobile vacancy concentration <span class="equationTd inline-formula"><math> N 0</math></span> is much higher and the activation energy <span class="equationTd inline-formula"><math> E A</math></span> for ion hops is much lower than is seen in most other semiconductors due to the inherent softness of perovskite materials. The timescale at which the internal electric field changes due to ion motion is determined by the vacancy diffusion coefficient <span class="equationTd inline-formula"><math> D v</math></span> and is similar to the timescale on which the external bias changes by a significant fraction of the open-circuit voltage at typical scan rates. Therefore, hysteresis is often observed in which the shape of the current–voltage, J–V, characteristic depends on the direction of the voltage sweep. There is also evidence that this defect migration plays a role in degradation. By employing a charge transport model of coupled ion-electron conduction in a perovskite solar cell, we show that <span class="equationTd inline-formula"><math> E A</math></span> for the ion species responsible for hysteresis can be obtained directly from measurements of the temperature variation of the scan-rate dependence of the short-circuit current and of the hysteresis factor <span class="equationTd inline-formula"><math> H</math></span>. This argument is validated by comparing <span class="equationTd inline-formula"><math> E A</math></span> deduced from measured J–V curves for four solar cell structures with density functional theory calculations. In two of these structures, the perovskite is <span class="equationTd inline-formula"><math> MAPbI 3</math></span>, where MA is methylammonium, <span class="equationTd inline-formula"><math> CH 3 NH 3</math></span>; the hole transport layer (HTL) is spiro (spiro-OMeTAD, 2,2<span class="equationTd inline-formula"><math> ′</math></span>,7,7<span class="equationTd inline-formula"><math> ′</math></span>- tetrakis[N,N-di(4-methoxyphenyl) amino]-9,9<span class="equationTd inline-formula"><math> ′</math></span>-spirobifluorene) and the electron transport layer (ETL) is <span class="equationTd inline-formula"><math> TiO 2</math></span> or <span class="equationTd inline-formula"><math> SnO 2</math></span>. For the third and fourth structures, the perovskite layer is <span class="equationTd inline-formula"><math> FAPbI 3</math></span>, where FA is formamidinium, <span class="equationTd inline-formula"><math> HC ( NH 2 ) 2</math></span>, or <span class="equationTd inline-formula"><math> MAPbBr 3</math></span>, and in both cases, the HTL is spiro and the ETL is <span class="equationTd inline-formula"><math> SnO 2</math></span>. For all four structures, the hole and electron extracting electrodes are Au and fluorine doped tin oxide, respectively. We also use our model to predict how the scan rate dependence of the power conversion efficiency varies with <span class="equationTd inline-formula"><math> E A</math></span>, <span class="equationTd inline-formula"><math> N 0</math></span>, and parameters determining free charge recombination.
physics, applied