AVS2011 Session EN+NS-WeM: Organic Photovoltaics

Conjugated polymer/metal oxide hybrid photovoltaic devices have received considerable attention in recent years due to their potential for scalable, low-cost manufacturing. To improve device efficiency, most research has focused on increasing the fill factor and short circuit current. In this article, calcium doped zinc oxide (Zn_xCa_1-xO) is used as electron acceptor in the hybrid poly(3-hexylthiophene) /Zn_xCa_1-xO bilayer solar cells to modify the band offsets and increase the open circuit voltage. XRD results show the sol-gel prepared Ca doped ZnO films have a hexagonal wurtzite phase with no indication of calcium separation. Atomic force micrographs indicate the Zn_xCa_1-xO surface becomes decorated with interconnected ridges approximately 100 nm in height. The rms surface roughness increases with increasing Ca. The band gap obtained from UV-vis absorption spectra widens from 3.20 to 3.60 eV as the Ca concentration increases from x=0 to x=0.12. The work function of the Zn_xCa_1-xO films measured by Kelvin probe is seen to decrease with increasing Ca composition for values of x up to 0.12, which indicates a decrease in electron affinity. The open circuit voltage of hybrid P3HT/Zn_xCa_1-xO devices increases from 570 mV at x = 0 up to 800 mV for x = 0.10. Work is under way to incorporate Ca doped ZnO nanoparticles into bulk heterojunction solar cells to optimize device efficiency.

The fabrication process for bulk heterojunction (BHJ) organic photovoltaic (OPV) devices nearly always includes anode modification steps ranging from solvent cleaning to haloacid treatments to the deposition of interfacial layers such as polymer blends or transition metal oxides. The role of these treatments is not yet fully understood, but they are thought to modify the anode work function, contribute to electron-blocking, alter the anode surface energy, and prevent shunts among other functions. In separate work, conductive atomic force microscopy (AFM) and derivative techniques have been used to simultaneously probe both the morphological and electrical properties of BHJ photovoltaic layers. Previous work has demonstrated a correlation between BHJ nanostructure and properties such as photocurrent at the same scale (~ 20 nm) as the phase separation in these films. Photocurrent variability has also been observed at a length scale orders of magnitude larger the domains in the BHJ layer, and this variation has been speculated to have an origin in the anode or interfacial layers in the OPV architecture.

Carbon nanomaterial thin films are being increasingly investigated for use as transparent electrodes in a variety of optoelectronic devices [1-3]. These flexible and solution-processible films are ideal candidates for organic electronics, including organic photovoltaics (OPVs), where low production costs and mechanical robustness are essential. While carbon nanotube networks have begun to be explored in such devices, the inherent polydispersity of the samples—namely the mix of both semiconducting and metallic species—has prevented the realization of their full potential in these devices and convoluted a complete understanding of their implications on device operation. Herein we report the use of single-walled carbon nanotubes (SWNTs) sorted by electronic type via density gradient ultracentrifugation as the transparent anode in poly(3-hexylthiophene) (P3HT) [6,6]-phenyl-C₆₁ butyric acid methyl ester (PCBM) organic photovoltaic devices. Through a combination of dip coating and nitric acid treatment we achieve film roughnesses comparable to typical transparent oxides such as indium tin oxide. While carbon nanotube films are often electronically doped during processing, either intentionally or as a byproduct of roughness-reducing acid treatments, we find that the application of the quintessential electron-blocking interlayer poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) removes sufficient adsorbed dopant groups to return the semiconducting nanotubes to their original state, vastly reducing their contribution toward current collection and transport. This is observed by both sheet resistance increases and UV-vis-NIR spectrophotometry. Further investigation using X-ray photoelectron spectroscopy reveals that the weakly-bound nitric oxide groups on the surface are almost entirely removed by PEDOT:PSS. By varying the semiconducting and metallic content in the electrodes, we find that metallic content greater than 70 percent yields devices with efficiencies 50 times greater than those comprised of almost entirely semiconducting SWNTs. This observation is counterintuitive considering that freshly acid-treated semiconducting SWNT films often possess a lower sheet resistance than their metallic counterparts [4]. This underscores the advantage of using metallic-enriched populations for transparent conductors, where unstable doping of semiconducting carbon nanotubes complicates processability and ultimately reduces device performance.

[1] T. M. Barnes; et al. Appl. Phys. Lett. 2010, 96, 243309.

[2] S. Kim; et al. Adv. Funct. Mater. 2010, 20, 2310.

[3] J. Li; et al. Nano Lett. 2006, 6, 2472.

Numerous reports have established that improving the performance of bulk heterojunction (BHJ) organic photovoltaic (OPV) devices requires not only the development of materials systems with improved spectral response and higher charge carrier mobility but also the ability to understand and tailor the morphology of these systems. Approaches for morphological characterization have included microscopic techniques such as scanning electron microscopy, transmission electron microscopy, and various atomic force microscopy techniques. Other methods have provided indirect information about active layer morphology by enabling the determination of charge carrier lifetimes. These approaches have included transient absorption spectroscopy, transient photovoltage, and time-of-flight techniques among others. While these methods have proven quite informative, the relationships between nanoscale morphology, device performance, and the underlying electrical characteristics of functioning devices are not yet fully understood.

Singlet exciton fission is a process by which a high energy singlet spontaneously decomposes into two low energy triplets. It is an example of a multi-exciton generation process that could allow the efficiency of solar cells to reach beyond the Shockley-Quiesser limit. In this talk we examine singlet exciton fission in solar cells based on tetracene and diphenyltetracene (DPT). Notably we find that the photocurrents in each cell have opposite dependencies on the application of a magnetic field, suggesting that fission increases the photocurrent in tetracene-based solar cells, but decreases the photocurrent in DPT-based solar cells.

We report an organic semiconductor solar cell using tetracene and CuPC (Copper Phthalocyanine) as donors, and C60 as the acceptor. Tetracene absorbs photons in the 450-550-nm region, generating high energy singlets which split into two lower energy triplets, potentially doubling the photocurrent in this part of the spectrum. CuPC extends the absorption of the solar cell to the red part of the spectrum. We also demonstrate bulk heterostructure tetracene-C60 devices and tested them for singlet exciton fission.

We observe that: (i) The addition of the CuPC layer does not affect the flow of excitons from tetracene to the CuPC – C60 junction significantly because of similarities in triplet energies of tetracene and CuPC, (ii) The application of a .5T magnetic field shows ~-1.5% change in photocurrent, confirming singlet fission, (iii) Low temperature quantum efficiency measurements show a drop in the tetracene IQE (internal quantum efficiency), and a singlet fission yield of 72% at room temperature. The drop in performance at low temperature is expected since singlet exciton fission in tetracene is a thermally activated process.(iv) Singlet exciton fission has a very high yield even in bulk heterostructure devices and can be potentially used to improve the performance of polymer solar cells.

The properties of semiconductor nanocrystal-organic films are of interest for applications in photodetectors, light-emitting diodes and photovoltaics. The versatility of these films arises from the ability to tune the bandgaps by adjusting the PbS nanocrystal size. Nanocrystalline photoelectric properties are also strongly affected by the chemical environment presented to them and by their bonding to the surrounding organic films. PbS nanocrystals are simultaneously deposited under vacuum into an organic matrix of evaporated α-sexithiophene while concomitantly modulating the film interface with 50 eV acetylene ions. Cluster beam deposition has been shown to allow the preparation of PbS nanocrystals that are surface terminated with either Pb or S [1]. Surface polymerization on ion assisted deposition (SPIAD) has shown an ability to increase the conjugation of oligothiophenes and otherwise lead to their chemical modification [2-4]. Cluster beam deposition and SPIAD are combined here to modulate the interface between the PbS and the organic matrix by introducing acetylene ions during the formation of the film. X-ray photoelectron spectroscopy analysis shows chemical shifts indicative of acetylene ion modulated interaction between the nanocrystals and sexithiophene. The deposition process was also simulated by a linearly scaled density functional theory method. These computer simulations indicate chemical bonding between the nanocrystals and sexithiophene, which is consistent with the chemical shifts observed by XPS.

1. A. M. Zachary, I. L. Bolotin, D. J. Asunskis, A. T. Wroble, and L. Hanley ACS Appl. Mater. Interf. 1 (2009) 1770

2. S. Tepavcevic, A. M. Zachary, A. T. Wroble, Y. Choi, and L. Hanley, J. Phys. Chem. A 110 (2006) 1618

3. S. tepavcevic, Y. Choi, and L. Hanley, Lang. 20 (2004) 8754

4. W. -D. Hsu, S. Tepavcevic, L. Hanley, and S. B. Sinnott, J. Phys. Chem. C 111 (2007) 4199

Organic materials present a promising direction for potentially cheaper solar cells. One way to use them and increase our understanding (e.g., define basic physical cell performance limits), is hybrid, organic / inorganic photovoltaics (PV). We explore 2 main directions to hybrid PV - with molecules as dipolar films, i.e., use electrostatics

- with molecular monolayers as electronic transport medium.

While we find that incomplete partial dipolar monomolecular films can control solar cell behavior for /single, poly- and nano-crystalline cells, for current to pass through the molecules, we need dense monolayers. Alkyl chain monolayers help form near-ideal Metal-Insulator-Semiconductor (MIS) diodes, with significant PV activity. Surprisingly, though we can actually make *MIS* cells without a separate *I*(nsulator) layer, suggesting that 'MIS' effects are at least partly more “chemical than is often thought.

In this way we demonstrate a near-ambient, simple, potentially low-cost approach to make and modify semiconductor solar cells, using a monolayer of molecules, as short as two carbons, that self-assembles onto the semiconductor (absorber) surface, passivating and buffering it. Good passivation is necessary to express the molecule-induced interface dipole, which can change the semiconductor electron affinity by up to 1 eV.

Good, stable interface passivation along with strong inversion allows minority carriers, generated by absorbed light, to move laterally within the semiconductor top layer, for collection by a minimal-area grid, deposited on the conducting polymer and also minimizes photo-current losses, due to sheet resistance. Thus, ≤ 1 nm thick organic molecules appear to convey a unique advantage over inorganic passivation or buffer layers.