Peumans, Uchida and Forrest (Nature 425, 158 (2003)) showed that annealing a vacuum-co-deposited copper phthalocyanine (CuPc):PTCBI film while it is mechanically confined under a deposited metal cap yields a pinhole-free, interpenetrating donor–acceptor network and roughly doubles the power conversion efficiency of comparable bilayer cells, and they interpreted the annealing-temperature-dependent morphology with a cellular-automaton model of thermally activated molecular exchange. The mechanistic and model-level basis of this confined-annealing route, however, was never established: the nanoscale pathway by which the rigid cap biases bulk phase segregation is unidentified, the cohesive- and interfacial-energy assumptions of the cellular automaton were adopted for convenience and remain untested for CuPc/PTCBI, the molecular-exchange kinetics lack direct experimental validation, the mechanism by which a second anneal restores the open-circuit voltage is inferred but unproven, and the long-term stability of the confined-annealed morphology is undemonstrated. This record states that unresolved problem: establish the physical mechanism (energetics, kinetics, and confinement effect) of confined annealing in this system and validate or replace the cellular-automaton interpretation, so that confinement-controlled demixing becomes a predictive morphology-engineering tool for vacuum-deposited small-molecule semiconductor blends.

Background

An organic photovoltaic cell converts light to electricity in a thin film of organic semiconductors. Because the photogenerated exciton (the tightly bound electron–hole pair created by photon absorption, with a binding energy of a few tenths of an eV) is neutral, it must diffuse to a donor–acceptor interface to be split into free charges. Exciton diffusion lengths in such materials are typically 5–20 nm, while the optical absorption length is ~100 nm, so a flat planar donor-on-acceptor (bilayer) junction harvests only a small fraction of the absorbed light — the exciton diffusion bottleneck. Polymer solar cells escaped this bottleneck by spin-coating a blend whose constituents demix during drying into an interpenetrating donor–acceptor network, a so-called bulk heterojunction. For vacuum-deposited small-molecular-weight materials (used when layer thickness and purity must be controlled at the ångström level), forming an analogous bulk heterojunction was long elusive: co-deposited mixed films do not demix on a useful length scale, and annealing them to drive demixing produces large crystallites, surface roughening and pinholes that short-circuit the device. Peumans, Uchida and Forrest (2003) solved the fabrication problem with confined ("strained") annealing: they co-deposited the archetypal donor copper phthalocyanine (CuPc) with the acceptor 3,4,9,10-perylenetetracarboxylic bisbenzimidazole (PTCBI), capped the film with a ~1000 Å Ag layer, and annealed at TA1≈T_{\mathrm{A1}}\approx 450–550 K. The rigid cap suppresses the surface stress-relief and pinhole formation, while phase segregation proceeds in the bulk, yielding ~10–20 nm CuPc-rich and PTCBI-rich domains. After cap removal, deposition of a bathocuproine (BCP, a wide-gap exciton-blocking/transparent electron-transport interlayer) plus Ag cathode, and a second anneal at TA2T_{\mathrm{A2}} that restored the open-circuit voltage (VOCV_{\mathrm{OC}}, the cell voltage at zero net current, set by contact and interface energetics), cells reached ηP=(1.5±0.1)%\eta_{\mathrm{P}}=(1.5\pm0.1)\%, about twice the comparable bilayer. To interpret the morphology, the authors used a cellular-automaton simulation — a lattice model in which each cell is occupied by CuPc or PTCBI, the configuration energy is a sum of nearest-neighbor contact energies EM(i),M(j)E_{M(i),M(j)} with cohesive energy per mole Ecoh=3NAEA,AE_{\mathrm{coh}}=3N_{\mathrm{A}}E_{A,A}, and molecular-exchange attempts occur at a thermally activated rate R(T)=R0exp⁡(−ΔEA/kT)R(T)=R_{0}\exp(-\Delta E_{\mathrm{A}}/kT). Because the required pairwise energies were unknown, they assumed EcohPTCBI=EcohCuPcE_{\mathrm{coh}}^{\mathrm{PTCBI}}=E_{\mathrm{coh}}^{\mathrm{CuPc}} (with EcohCuPc=176E_{\mathrm{coh}}^{\mathrm{CuPc}}=176 kJ/mol) and the interfacial-energy relation 2ECuPc−PTCBI=ECuPc−CuPc=EPTCBI−PTCBI2E_{\mathrm{CuPc-PTCBI}}=E_{\mathrm{CuPc-CuPc}}=E_{\mathrm{PTCBI-PTCBI}} for convenience; the model nevertheless reproduced the observed domain coarsening with annealing temperature, and the trade-off it predicts between the exciton-diffusion efficiency ηED\eta_{\mathrm{ED}} and the carrier-collection efficiency ηCC\eta_{\mathrm{CC}} (the probability that a free carrier generated at an interface reaches its electrode) matches the measured device behavior, in which carrier collection — not exciton harvesting — limits performance. From the temperature dependence of the external quantum efficiency (the fraction of incident photons converted into collected electrons), the authors extracted an activation energy ΔEA=(1.1±0.1)\Delta E_{\mathrm{A}}=(1.1\pm0.1) eV for the morphology-changing process and estimated a maximum operating temperature of roughly 53 °C for a ten-year morphological lifetime. What was demonstrated, therefore, is a processing route whose phenomenology is consistent with a convenient lattice model; what was left open is the underlying physics, as summarized in the problem statement below. For context, the field subsequently pursued other routes to small-molecule bulk heterojunctions — CuPc:C60 planar-mixed and tandem architectures (Forrest, 2005), then non-fullerene acceptors — so the confined-annealing demonstration and its open questions are specific to the CuPc:PTCBI platform.

Problem Statement

For vacuum-co-deposited CuPc:PTCBI thin films annealed under a rigid metal cap, establish the physical mechanism and the validated energetics of confinement-controlled bulk phase segregation, replacing the source paper's convenience assumptions with measured or computed quantities. Concretely, resolve all of the following linked questions, which serve the single objective of making confined annealing a physically understood, predictive morphology-engineering route for vacuum-deposited small-molecule donor–acceptor blends: (1) Mechanism — characterize the confined-annealed morphology at nanoscale resolution and determine how the mechanically rigid cap biases bulk phase segregation: which strain fields and stress-relief pathways are suppressed at the surface, and what molecular mass-transport pathway (for example vacancy-mediated or grain-boundary/interface-mediated exchange of CuPc and PTCBI molecules) carries the segregation in the film interior; (2) Energetics — determine the actual cohesive energies of CuPc and PTCBI and the CuPc–PTCBI interfacial energy, experimentally or by atomistic calculation, and thereby test the model assumptions EcohPTCBI=EcohCuPcE_{\mathrm{coh}}^{\mathrm{PTCBI}}=E_{\mathrm{coh}}^{\mathrm{CuPc}} (with EcohCuPc=176E_{\mathrm{coh}}^{\mathrm{CuPc}}=176 kJ/mol) and 2ECuPc−PTCBI=ECuPc−CuPc=EPTCBI−PTCBI2E_{\mathrm{CuPc-PTCBI}}=E_{\mathrm{CuPc-CuPc}}=E_{\mathrm{PTCBI-PTCBI}} underlying the cellular-automaton interpretation (the interfacial-energy relation subsumes the equal-cohesive-energy assumption through Ecoh=3NAEA,AE_{\mathrm{coh}}=3N_{\mathrm{A}}E_{A,A}; the cohesive-energy equality alone does not resolve this part), and establish whether the simulated morphologies survive the corrected parameters; (3) Kinetics — provide direct experimental validation of the thermally activated molecular-exchange kinetics, including the inferred activation energy ΔEA≈(1.1±0.1)\Delta E_{\mathrm{A}}\approx(1.1\pm0.1) eV, and hence of the estimated operating-temperature limit near 53 °C for a ten-year morphological lifetime; (4) Second anneal — determine the mechanism by which the post-contact anneal at TA2T_{\mathrm{A2}} restores VOCV_{\mathrm{OC}} (removal of adsorbed contaminants, electrical 'forming' of the donor–acceptor/BCP contact, or another process); and (5) Stability and transport — demonstrate whether the confined-annealed morphology and the interpenetrating network's carrier collection persist under realistic thermal, mechanical, photochemical and illumination stress over device lifetimes, without reintroducing interfacial disorder or surface defects. The cellular-automaton model itself need not be adopted: its cohesive- and interfacial-energy assumptions (part 2) may be refuted in favour of a validated alternative mechanism or energy set, while the nanoscale morphological and mass-transport determination of part (1), and parts (3)–(5), are separate targets that stand independently of it.

References

  1. Peter Peumans, Soichi Uchida, Stephen R. Forrest, 'Efficient bulk heterojunction photovoltaic cells using small-molecular-weight organic thin films', Nature 425, 158–162 (2003), DOI 10.1038/nature01949, https://doi.org/10.1038/nature01949
  2. Stephen R. Forrest, 'The Limits to Organic Photovoltaic Cell Efficiency', MRS Bulletin 30, 28–32 (2005), DOI 10.1557/MRS2005.5, https://doi.org/10.1557/MRS2005.5
  3. M. Riede, T. Mueller, W. Tress, R. Schueppel, K. Leo, 'Small-molecule solar cells—status and perspectives', Nanotechnology 19, 424001 (2008), DOI 10.1088/0957-4484/19/42/424001, https://doi.org/10.1088/0957-4484/19/42/424001
  4. Xiaoxia Jiang, Jiguang Dai, Haibo Wang, Donghang Yan, 'The effect of annealing treatment on performance of interdiffused organic photovoltaic devices', Thin Solid Films 516, 6487–6491 (2008), DOI 10.1016/j.tsf.2008.02.026, https://doi.org/10.1016/j.tsf.2008.02.026
  5. Ying Zheng, Sharon K. Pregler, Jason D. Myers, Jiaomin Ouyang, Susan B. Sinnott, Jiangeng Xue, 'Computational and experimental studies of phase separation in pentacene:C60 mixtures', Journal of Vacuum Science & Technology B 27, 169–179 (2009), DOI 10.1116/1.3072516, https://doi.org/10.1116/1.3072516
  6. Han Huang, Yuli Huang, Jens Pflaum, Andrew Thye Shen Wee, Wei Chen, 'Nanoscale phase separation of a binary molecular system of copper phthalocyanine and di-indenoperylene on Ag(111)', Applied Physics Letters 95, 263309 (2009), DOI 10.1063/1.3280858, https://doi.org/10.1063/1.3280858
  7. Jörn-Oliver Vogel, Ingo Salzmann, Steffen Duhm, Martin Oehzelt, Jürgen P. Rabe, Norbert Koch, 'Phase-separation and mixing in thin films of co-deposited rod-like conjugated molecules', Journal of Materials Chemistry 20, 4055 (2010), DOI 10.1039/b927594k, https://doi.org/10.1039/b927594k
  8. Chunfu Zhang, Yue Hao, Shi-Wun Tong, Zhenhua Lin, Qian Feng, En-Tang Kang, Chunxiang Zhu, 'Effects of Cathode Confinement on the Performance of Polymer/Fullerene Photovoltaic Cells in the Thermal Treatment', IEEE Transactions on Electron Devices 58, 835–842 (2011), DOI 10.1109/TED.2010.2100820, https://doi.org/10.1109/TED.2010.2100820
  9. Alexander Hinderhofer, Frank Schreiber, 'Organic–Organic Heterostructures: Concepts and Applications', ChemPhysChem 13, 628–643 (2012), DOI 10.1002/cphc.201100737, https://doi.org/10.1002/cphc.201100737
  10. Xin Li, Yan Chen, Jie Sang, Bao-Xiu Mi, Dan-Hua Mu, Zhi-Gang Li, Hui Zhang, Zhi-Qiang Gao, Wei Huang, 'CuPc/C60 bulk heterojunction photovoltaic cells with evidence of phase segregation', Organic Electronics 14, 250–254 (2013), DOI 10.1016/j.orgel.2012.10.041, https://doi.org/10.1016/j.orgel.2012.10.041
  11. R. Banerjee, J. Novák, C. Frank, C. Lorch, A. Hinderhofer, A. Gerlach, F. Schreiber, 'Evidence for Kinetically Limited Thickness Dependent Phase Separation in Organic Thin Film Blends', Physical Review Letters 110, 185506 (2013), DOI 10.1103/PhysRevLett.110.185506, https://doi.org/10.1103/PhysRevLett.110.185506
  12. Xiaohua Zhang, Pu Fan, Zijun Wang, Ding Zheng, Junsheng Yu, 'Morphology Regulation by Upside-Down Thermal Annealing Based on Blend Thick-Film Organic Solar Cells', Journal of Physical Chemistry C 125, 15194–15199 (2021), DOI 10.1021/acs.jpcc.1c03994, https://doi.org/10.1021/acs.jpcc.1c03994