Self-Assembly and Its Impact on Interfacial Charge Transfer in Carbon Nanotube/P3HT Solar Cells
Source abstract
Abstract
Charge transfer at the interface of conjugated polymer and nanoscale inorganic acceptors is pivotal in determining the efficiency of excitonic solar cells. Despite intense efforts, carbon nanotube/polymer solar cells have resulted in disappointing efficiencies (<2%) due in large part to poor charge transfer at the interface. While the interfacial energy level alignment is clearly important, the self-assembly and the interface structure also play a major role in facilitating this charge transfer. To understand and control this effect to our advantage, we study the interface of commonly used conductive polymer poly-3-hexylthiophene (P3HT) and single-walled carbon nanotubes (SWNTs) with a combination of molecular dynamics simulations, absorption spectra experiments, and an analysis of charge transfer effects. Classical molecular dynamics simulations show that the P3HT wraps around the SWNTs in a number of different conformations, including helices, bundles, and more elongated conformations that maximize planar π-π stacking, in agreement with recent experimental observations. Snapshots from the MD simulations reveal that the carbon nanotubes play an important templating role of increasing the π-conjugation in the system, an effect deriving from the π-π stacking interaction at the interface and the 1-dimensional (1D) nature of the SWNTs, and independent of the SWNT chirality. We show how this increase in the system conjugation could largely improve the charge transfer in P3HT-SWNT type II heterojunctions and support our results with absorption spectra measurements of mixtures of carbon nanotubes and P3HT. These findings open possibilities for improved preparation of polymeric solar cells based on carbon nanotubes and on 1D nanomaterials in general.
Source: openalex_inverted_index · Quality: baseline
Research context
OpenAlex topics: Conducting polymers and applications · Carbon Nanotubes in Composites · Organic Electronics and Photovoltaics
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Department of Materials Science and Engineering, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, Massachusetts 02139-4307, United States
School of Engineering Systems, Queensland University of Technology, 2 George Street, Brisbane, Queensland 4001, Australia
References (44)
Show source-provided references
1. Zhu H. · Sol. Energy Mater. Sol. Cells · 2009 · 93 · 146110.1016/j.solmat.2009.04.006 ↗
2. Mayer A. C. · Mater. Today · 2007 · 10 · 2810.1016/S1369-7021(07)70276-6 ↗
3. Kymakis E. · Appl. Phys. Lett. · 2002 · 80 · 11210.1063/1.1428416 ↗
4. Kanai Y. · Nano Lett. · 2008 · 8 · 90810.1021/nl0732777 ↗
5. Kymakis E. · 2005
6. Kymakis E. · Nanotechnol. Law Bus. · 2006 · 3 · 405
7. Li C. · J. Mater. Chem. · 2007 · 17 · 240610.1039/B618518E ↗
8. Ma W. · J. Adv. Funct. Mater. · 2005 · 15 · 161710.1002/adfm.200500211 ↗
9. Kymakis E. · Synth. Met. · 2004 · 142 · 16110.1016/j.synthmet.2003.08.011 ↗
10. Kymakis E. · Synth. Met. · 2002 · 127 · 5910.1016/S0379-6779(01)00592-6 ↗
11. Chen L.-M. · Adv. Mater. · 2009 · 21 · 143410.1002/adma.200802854 ↗
12. Yang C. · J. Am. Chem. Soc. · 2006 · 128 · 1200710.1021/ja063707f ↗
13. Hersam M. C. · Nat. Nanotechnol. · 2008 · 3 · 38710.1038/nnano.2008.135 ↗
14. Liu J. · MRS Bull. · 2010 · 35 · 31510.1557/mrs2010.554 ↗
15. Giulianini M. · Appl. Phys. Lett. · 2009 · 95 · 01330410.1063/1.3173825 ↗
16. Giulianini, M.; Waclawik, E.; Bell, J. M.; Scarselli, M.; Castrucci, P.; De Crescenzi, M.; Motta, N.Evidence of Multiwall Carbon Nanotube Compression Caused by Poly(3-hexylthiophene) Adhesion. Submitted for publication.
17. Gu H. · Adv. Mater. · 2008 · 20 · 443310.1002/adma.200801062 ↗
18. Jensen F. · 2007
19. Frenkel D. · 2002
20. Johnson R. R. · Nano Lett. · 2008 · 8 · 6910.1021/nl071909j ↗
21. Vukmirovic N. · J. Phys. Chem. B · 2009 · 113 · 40910.1021/jp808360y ↗
22. LAMMPS Molecular Dynamics Simulator.http://lammps.sandia.gov. (Accessed March 1, 2010).
23. Plimpton S. J. · J. Comput. Phys. · 1995 · 117 · 110.1006/jcph.1995.1039 ↗
24. Marcon V. · J. Phys. Chem. B · 2004 · 108 · 1805310.1021/jp047128d ↗
25. Allinger N. L. · J. Am. Chem. Soc. · 1989 · 111 · 855110.1021/ja00205a001 ↗
26. Cornell W. D. · J. Am. Chem. Soc. · 1995 · 117 · 517910.1021/ja00124a002 ↗
27. MM3 intramolecular interactions complete the description of the inter-ring torsional barriers and therefore must be used to model the P3HT intramolecular and P3HT-P3HT intermolecular non-bonded terms. On the other hand, the AMBER force field has been tested more extensively to account for the interaction of molecules with SWNTs (see ref.20) and thus this parametrization was chosen to describe the P3HT−SWNT intermolecular interaction.
28. Hummer G. · Nature · 2001 · 414 · 18810.1038/35102535 ↗
29. Tersoff J. · Phys. Rev. B · 1988 · 37 · 699110.1103/PhysRevB.37.6991 ↗
30. Tersoff J. · Phys. Rev. B · 1989 · 39 · 556610.1103/PhysRevB.39.5566 ↗
31. Note that our calculations do not contain explicit solvent molecules; however, the most significant effect from anon-polarsolvent may be a slight screening of the charges within a same P3HT chain. Anon-polarsolvent such as chloroform or toluene, commonly used in P3HT−SWNT solutions for solar cell preparation, would interact in a comparable manner with P3HT and SWNTs, with a net effect (to a first approximation) of slowing down the dynamics of the system without modifying the self-assembly significantly. Introduction of apolarsolvent (like water) would instead change the self-assembly dramatically and lead to significant aggregation within the hydrophobic SWNT and P3HT phases, but for this reason this is not used in practical circumstances for preparation of bulk-heterojunction solar cells. Thus our results refer to the self-assembly in anon-polarsolvent.
32. In Figure 3, PBC have been unwrapped for ease of visualization. This means that all-P3HT domains exist but cannot be seen due to the unwrapping, that causes molecules to move to adjacent cells instead of re-entering the same simulation cell. Clearly all-P3HT domains far from the interface do not crystallize during the simulation as occurs in reality, as this is an intrinsic limit of MD simulations. This does not limit the possibility of studying the very first few layers adjacent to the SWNT−P3HT interface, where crystalline order is in any case disrupted by the SWNTs presence.
33. Bredas J. L. · J. Chem. Phys. · 1985 · 83 · 132310.1063/1.449450 ↗
34. To allow for a 5% tolerance, we actually chose 42° as the threshold value for an angle to be defined as “conjugation-breaking”. We calculatedfevery ps in a 1 ns run, discarded the first 100 ps (equilibration), and averaged to obtain the finalfvalue.
35. Note that while the standard deviation can be significant for a single P3HT chain with 40 rings where the number of sampled ring−ring torsion angles is only 38, when a large number of chains is present in the simulation run the standard deviation decreases significantly (and would go to zero in the limit of infinitely long chains).
36. Wei J. Q. · Nano Lett. · 2007 · 7 · 231710.1021/nl070961c ↗
37. Endo M. · Nature · 2005 · 433 · 47610.1038/433476a ↗
38. Brown P. J. · Phys. Rev. B · 2003 · 67 · 06420310.1103/PhysRevB.67.064203 ↗
39. The relatively slight increase is consistent with the fact that this is mostly an interface effect. Since even at the highest DWNT concentrations the volume fraction of the P3HT is around 80−90%, a relatively small effect has to be expected. The points at small DWNT concentrations below 5% should therefore be considered as a single value despite fluctuations of about 10% in the intensity ratio. This sets the accuracy for these estimates with our method. On the other hand, the increase in the intensity ratio at higher DWNT concentrations is significant and consistent with the other evidences and with the spectra for MWNT−P3HT in theSupporting Information.
40. Bardeen J. · Phys. Rev. Lett. · 1961 · 6 · 5710.1103/PhysRevLett.6.57 ↗
41. Salleo A. · Mater. Today · 2007 · 10 · 3810.1016/S1369-7021(07)70018-4 ↗
42. Hamadani B. H. · Nano Lett. · 2010 · 10 · 161110.1021/nl9040516 ↗
43. For a more complete assessment of charge transfer at this interface, several other mechanisms and factors should be considered: the presence of Forster resonant energy transfer, the existence of a charge transfer state, the possibility of non-adiabatic (vsadiabatic) charge transfer as assumed here, and a correct level alignment (in an independent electron picture of the interface) at the P3HT−SWNT interface, which is partially explored in ref4but is still largely an open problem. The application of more accurate approaches than DFT, which is beyond the scope of this investigation, would also be of great value in an attempt to gain insight into the charge transfer mechanism (see,e.g., ref44).
44. Kanai Y. · J. Mater. Chem. · 2010 · 20 · 105310.1039/B913277P ↗