Perovskite Solar Cells


Published on

This presentation summarizes history and recent development of perovskite solar cells. If you have any questions or comments, you can reach me at

Published in: Science, Technology, Business
  • Be the first to comment

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide
  • a cheap and scalable approach that is also the main strength of alternative technologies such as organic photovoltaics, dye-sensitized solar cells and colloidal quantum dot-based solar cells.
  • By time-resolved transient photoluminescence, 2 papers (published at the same time) claimed extremely high carrier diffusion lengths of > 100 nm. In one of the papers, the mixed halide perovskite has an even longer diffusion length of >1000 nm. charge-carrier extraction model based on diffusion was used to estimate the charge carrier diffusion lengths, obtained minimum estimates of 130 and 90 nm for electrons and holes, respectively. (compared to the only 10 nm for polymers in PV.)
  • Coupled/correlated electron-hole pairs. Conventional solar cells, such as silicon p-n junction, photogeneration of free electron-hole pairs occurs throughout the bulk semiconductor in conventional cells. Excitonic solar cells, such as OPV and DSSC, consists of cells in which light absorption results in the production of a transiently localized excited state that cannot thermally dissociate (binding energy ≫ kT). In other words, the absorbed photon produces a neutral excitation, not free carriers. Therefore, a dissociation interface is required.

  • and colleagues were working on photovoltaics called dye-sensitized solar cells (DSSCs). Unlike conventional silicon solar cells, DSSCs consist of a blend of organic light-absorbing dyes coating tiny inorganic particles such as titanium dioxide (TiO2), which in turn are surrounded by a charge-conducting electrolyte. In standard DSSCs, when a dye molecule absorbs a photon, the light boosts the energy of one of the electrons in the dye, enabling it to jump onto a TiO2 particle. From there, it skips from particle to particle until it reaches an electrode, where it’s collected and sent through a circuit to do work. Meanwhile, another electron jumps from the electrolyte to the dye to restore it to its original state.

    He says that it took one of his students 2 years to find a recipe to make the material sturdy enough for a brief demonstration.
  • DSSC are considered to be ‘excitonic’ photovoltaics that require a large surface area for charge separation between electron-accepting TiO2 and the adsorbed dye layer.
  • Either a solution of the organic and inorganic materials is used to create the film or the two components are thermally evaporated together inside a vacuum chamber.
  • Perovskite Solar Cells

    1. 1. All rights reserved. ©2014, Wenchun Feng Perovskite Solar Cells Wenchun Feng Garfunkel Group Department of Chemistry and Chemical Biology Rutgers University 3/7/2014 Image credit: B. Zhang, W.C.L. Glenn & M. Liu. Current Address: Kotov Group, Department of Chemical Engineering, University of Michigan, Ann Arbor Email:
    2. 2. All rights reserved. ©2014, Wenchun Feng Schematic 2 Perovskite Precursor Solution CH3NH3I PbI2 fluorine-doped tin oxide substrate spin coat then annealFTO FTO TiO2 FTO TiO2 Perovskite spin coat perovskite then anneal spin coat spiro-OMeTAD FTO TiO2 Perovskite spiro-OMeTAD FTO TiO2 Perovskite spiro-OMeTAD Electrode Deposition leave in air to dope via oxidation Au Au Au -3.9 -5.4 -4.0 -5.2 -5.1 hv - + - + + E (eV) LUMO HOMO CB TiO2 (CH3NH3)PbI3 spiro-OMeTAD Au HOMO Adv. Funct. Mater. 2014, 24, 151. Sci. Rep. 2012, 2. DOI: 10.1038/srep00591. Image Credit: M.B. Johnston (CH3NH3)PbI3
    3. 3. All rights reserved. ©2014, Wenchun Feng Perovskite Solar Cells Made Simply – C&EN Video 3 For a C&EN video, please see
    4. 4. All rights reserved. ©2014, Wenchun Feng Perovskite Solar Cell is the Poster Child for Solar Energy Conversion Source: Web of Science 4 Henry Snaith
    5. 5. All rights reserved. ©2014, Wenchun Feng 5
    6. 6. All rights reserved. ©2014, Wenchun Feng Science. 2013, 342, 794 Chem. Mater. 1999, 11, 3028 Science. 1999, 286, 945 J. Chem. Soc., Dalton Trans., 2001, 1, 1 In the late 1990s, David Mitzi (IBM) made TFT and LED with organometal halide perovskites. In general, Good light emitters = Good light absorbers = Potentially Good PV materials. However, due to its instability, he didn’t pursue them as viable PV material. OPV DSSC CIGS 6
    7. 7. All rights reserved. ©2014, Wenchun Feng What is Perovskite?  A perovskite structure is any material with the same type of crystal structure as calcium titanium oxide (CaTiO3), known as the perovskite structure ABX3.  First discovered by Gustav Rose in 1839 and named after Russian mineralogist L. A. Perovski. 7
    8. 8. All rights reserved. ©2014, Wenchun Feng CH3NH3PbX methylammonium lead halide Nature. 2013, 12, 1087 J. Mater. Chem. A, 2013, 1, 15628 Phase Transition (CH3NH3PbI3): Orthorhombic  Tetragonal  Cubic 162 K 327 K (54 C) The organic ligand is disordered in Tetragonal and Cubic phase. 8
    9. 9. All rights reserved. ©2014, Wenchun Feng Material Properties: Good for Photovoltaics, but with Caution  Cheap Manufacturing:  Lower manufacturing costs expected: directly deposited from solution  Caution: Encapsulation needed, which may increase cost  Material Properties for High Efficiency Photovoltaics:  1. High Optical Absorption Coefficient  2. Excellent Charge Carrier Transport (crystallinity, diffusion length, carrier mobility)  3. Promising Device Parameter: High Voc of >1.1 V is reported  Stability:  Study shows it can maintain more than 80% of its initial efficiency after 500 hours.  Caution: More studies needed. Lifetime of 15 years has not been demonstrated. The ultimate goal of 15-year-lifetime not demonstrated.  Other Real World Concerns (equally important but omitted here):  Toxicity from Pb  Scaling Problem Nature. 2013, 12, 1087 Nature Comm. 2013, DOI: 10.1038/ncomms3885 Science. 2013, 342, 317 Nature. 2013, 499, 316 9
    10. 10. All rights reserved. ©2014, Wenchun Feng High Optical Absorption Coefficient Perovskite Absorption Coefficient: 1.5X104 cm-1 at 550 nm, which is one order of magnitude higher than conventional ruthenium-based organometallic N719 dye Nanoscale, 2011, 3, 4088 10
    11. 11. All rights reserved. ©2014, Wenchun Feng Crystallinity Tetragonal Structure Lattice Parameters: a = 8.825 Å, b = 8.835 Å, c = 11.24 Å. Science. 2012, 338, 643. 11
    12. 12. All rights reserved. ©2014, Wenchun Feng Carrier Diffusion Length quartz CH3NH3PbI3 PCBM quartz CH3NH3PbI3 Spiro-OMeTAD electron-extracting layer hole-extracting layer Science. 2013, 342, 341 Science. 2013, 342, 344 e h PL quenching originates from the charge-carrier extraction across the interface. 12
    13. 13. All rights reserved. ©2014, Wenchun Feng Remaining Question: Is the Solar Cell Excitonic?  Similar electron and hole diffusion lengths (130 nm and 100 nm in average),  either indicate similar mobilities for both electrons and holes,  or indicate that the predominant diffusion species is the weakly bound exciton.  Based on the work so far, we cannot tell whether the photoexcited species are excitons or free charges Conventional Solar Cell (p-n junction Si) Excitonic Solar Cell (DSSC, OPV) + - + - + - + - interface exciton free charges hv hv J. Phys. Chem. B 2003, 107, 4688 Science. 2013, 342, 341 Science. 2013, 342, 344 13
    14. 14. All rights reserved. ©2014, Wenchun Feng Carrier Mobility  High carrier mobility is important:  Because, coupled with high charge carrier lifetimes, it allows for the light- generated carriers move large enough distances to be extracted as current, instead of energy loss by recombination. Inorg. Chem. 2013, 52, 9019 Science. 2013, 342, 317 Adv. Funct. Mater. 2005,15, 77 Combining resistivity and Hall effect data, the electron mobility was calculated to be ~ 66 cm2/V·s for CH3NH3PbI3. This is remarkably high (in comparison, PCBM, a common electron acceptor in OPV, has electron mobility of 0.21 cm2/V·s) 14
    15. 15. All rights reserved. ©2014, Wenchun Feng High Voc  Perovskite solar cells are quite effective in generating a high electric voltage, which is represented by a high open circuit voltage (Voc):  CH3NH3PbI2Cl:  optical bandgap of 1.55 eV  Voc of 1.1 V  A voltage drop of only 0.45 eV, competitive with the best thin-film technologies (CIGS: 1.15 eV  0.7 V; Si: 1.1 eV  0.7 V)  Compares favorably to DSSC or OPV, which has a larger 0.7-0.8 V voltage drop, resulting in a small portion of the bandgap being extracted as Voc.  Other optimization predictions: Higher FF is possible (60-70%  over 80%), current can also be higher. Science. 2013, 342, 794 Science. 2012, 338, 643 15
    16. 16. All rights reserved. ©2014, Wenchun Feng Getting the Best of Both Worlds It has been agreed on that the organic-inorganic hybrid nature is what makes it so successful: The organic component renders good solubility to the perovskite and facilitates self-assembly, effectively enabling its precipitation/deposition from solution. The inorganic component produces an extended network by covalent and/or ionic interactions (instead of weaker forces such as Van der Waals or π-π interactions). Such strong interaction allows for precise crystalline structure in the deposited films.” Nature. 2013, 12, 1087 16
    17. 17. All rights reserved. ©2014, Wenchun Feng DSSC: Predecessor to Perovskite Solar Cells  Current Perovskite Solar Cells are built upon the architectural basis for DSSCs  Pioneering work by Grätzel (EPFL, Switzerland) that has garnered ~ 17000 citations Nature. 1991, 353, 737 technical documents 17
    18. 18. All rights reserved. ©2014, Wenchun Feng Replacing Dye with Perovskite  Dyes do not absorb all the incident light, reducing DSSC efficiency.  In 2009, Miyasaka (Toin U. of Yokohama, Japan) turns to perovskite as possible replacement of the dye and achieved 3.8% efficiency.  Problem: Liquid electrolyte dissolved away the perovskite within minutes. Liquid Electrolyte 0.15 M LiI and 0.075 M I2 in methoxyacetonitrile JACS. 2009, 131, 6050. 18
    19. 19. All rights reserved. ©2014, Wenchun Feng Replace Liquid Electrolyte with a Solid Hole Transporting Layer (HTL)  2012, Nam-Gyu Park (Sungkyunkwan U., South Korea) teamed up with Grätzel, over 9% efficiency. HTL layer: spiro-OMeTAD Sci. Rep. 2012, 2, DOI: 10.1038/srep00591 19
    20. 20. All rights reserved. ©2014, Wenchun Feng TiO2 Al2O3 Scaffold and Single Layer 2012, Henry Snaith (Oxford U.): Is TiO2 essential for high efficiency? Switched to an insulating Al2O3 scaffold, expected to see a decrease in efficiency. Surprisingly, the device with Al2O3 has a higher efficiency than that with TiO2 (11% vs. 7.6%). If all that Al2O3 does is scaffolding, what if we get rid of it as well? 2013 15 % Al2O3 solution vacuum deposition Perovskite Precursor Solution CH3NH3I PbI2 2013 11 % Science. 2012, 338, 643. Nature. 2013, 501, 395. Adv. Funct. Mater. 2014, 24, 151 Do not need artificial interfaces for efficient charge separaton! These two results are remarkable, in that they prove that these perovskites work as conventional semiconductors (Si, GaAs, etc). single layer (thin film) device 20
    21. 21. All rights reserved. ©2014, Wenchun Feng Sequential Deposition 2013, Grätzel sticks with the TiO2 structure and tinkered with the deposition step. mesoporous TiO2 PbI2 CH3NH3I Efficiency: 15% Nature. 2013, 499, 316 perovskite 21
    22. 22. All rights reserved. ©2014, Wenchun Feng Vapor-Assisted Deposition  Drawbacks of the above two deposition techniques:  Despite success in labs, both are challenging for large-scale industrial production.  The all-solution process results in decreased film quality, and the vacuum process requires expensive equipment and uses a great deal of energy.  Alternative: Vapor-assisted solution process (Yang Yang, UCLA):  The organic material infiltrates the inorganic matter and forms a compact perovskite film that is significantly more uniform than the films produced by the wet technique. JACS. 2014, 136, 622 surface roughness 23.2 nm 22
    23. 23. All rights reserved. ©2014, Wenchun Feng New Structures  Current Perovskite Structure: CH3NH3PbX  Possible modifications:  Cation (some success)  Metal (challenge)  Halogen (some success, with lingering questions) 23
    24. 24. All rights reserved. ©2014, Wenchun Feng Cation: CH3NH3 vs. HC(NH2)2  Formamidinium cation slightly larger  Bandgap shrunk from 1.55 eV to 1.48 eV  Retained favorable transport property  Best solution-based thin film perovskite solar cell (efficiency: 14.2%). 24 N C N N C Energy Environ. Sci. 2014, 7, 982
    25. 25. All rights reserved. ©2014, Wenchun Feng Metal  In principle, perovskite is a flexible structure type:  Many elements in the periodic table (such as Co2+ , Fe2+ , Mn2+ , Pd2+ , and Ge2+) can be incorporated  Sn:  Earlier work by Mitzi showed that perovskites with Sn show metallic character with small bandgap (not the semiconductors ideal for PV)  Recent work confirmed this notion.  It was also found that the facile oxidation of Sn2+ to Sn4+ gives a metal-like behavior in the semiconductor which lowers the photovoltaic performance.  Cu:  At 2013 Fall MRS meeting, a group from Hokkaido U., Japan, is studying Cu-based materials with a general formula of R2CuBr4. Field effect transistor by inkjet deposition showed weak n-type properties.  In general, replacing Pb with other less or non toxic metals remains a challenge. 25 Science. 1995, 267, 1473 Materials Today. 2014, 17, 16 Dalton Trans., 2011, 40, 5563 Inorg. Chem. 2013, 52, 9019
    26. 26. All rights reserved. ©2014, Wenchun Feng Halogen: MAPbI3, MAPbBr3, MAPbCl3, MAPbI3-xClx  CH3NH3PbI3 bandgap 1.5 eV, CH3NH3PbBr3 2.3 eV, bandgap tailoring can be simply achieved by alloying of both. MAPb(I1-xBrx)3 26 Nano Lett. 2013, 13, 1764 J. Mater. Sci., 2002, 37, 3585 Chem. Mater. 2013, 25, 4613 Adv. Mater. 2013, DOI: 10.1002/adma.201304803 Science. 2012, 338, 643  Why don’t researchers use CH3NH3PbCl3?  Bandgap (3.1 eV) too large for good absorption, appearance is colorless.  Mixed Halide CH3NH3PbI3-xClx: Currently the best perovskite material.  The extent of incorporation of Cl into the perovskite is in debate. Recent results indicate that the Cl incorporation is very low (3-4%), due to the large difference in the ionic radii of Cl- and I- anions. Another XPS study shows Cl:I exactly at 1:2.  This mixed halide have similar bandgaps as CH3NH3PbI3 (another indication that the Cl incorporation is low)  This mixed halide has superior charge transport property than iodide one (diffusion length >1000 nm)  Better stability in air (reason unclear)
    27. 27. All rights reserved. ©2014, Wenchun Feng Snaith Cell Stability Delivers Stable Photocurrent for 1000 hrs under continuous illumination of full spectrum simulated sunlight. However, efficiency suffers from a 50% loss at 200 hr period. Nature Comm. 2013, DOI: 10.1038/ncomms3885 27
    28. 28. All rights reserved. ©2014, Wenchun Feng Grätzel Cell Stability This structure has TiO2 mesoporous structure. less than 20% loss of its initial efficiency after 500 hours. Nature. 2013, 499, 316 Note: Both Snaith and Gratzel cells require encapsulation. Takeaway message: Stability testing shows different results for different devices. Carefully calibrated testing on the same system is needed. 28
    29. 29. All rights reserved. ©2014, Wenchun Feng Summary and Outlook  Perovskite solar cells made strides in efficiency gains over the past 5 years. Further efficiency improvement is predicted to lie in FF and current (Voc is probably saturated at a mere 0.4 V drop).  Photophysical studies show extremely long electron and hole diffusion length (>100 nm; for mixed halides >1000 nm). However, it is unclear whether excitons or free charges are the predominant diffusion species.  Although electron mobility has been studied by one paper, mobility studies in general are lacking.  Preliminary stability testing shows promise. Long term stability testing (which should guarantee a lifetime of 15 years or more) is essential to successful commercialization.  New structures emerging with varying degree of success:  Organic Cation  Toxicity associated with Pb should be mitigated by exploring other metal cations (e.g. Sn, Cu. Unfortunately, they currently do not make good devices).  Halogen (probably the most mature understanding of all three)  Scaling remains a challenge. Best performing devices generally have a low surface area of < 0.1 cm2 .  Efficiency dropped ~ 30% from ~ 0.1 cm2 to 1 cm2 cell area. Mostly by FF decrease. Reason unclear. 29 Nature Photon. 2014, 8, 128