AmericanChemicalSociety.com

Organic polymer blends are being actively explored as materials for solar photovoltaic devices because of their potential to cover large, irregularly shaped areas as low cost.  The long-term stability of such devices remains an issue, and their efficiency does not yet approach that of silicon.  Among the strategies for increasing the efficiencies of organic polymer-based photovoltaic cells, a particularly intriguing one is the incorporation of metallic nanoparticles.  While a number of papers have reported enhancement, negative results have also been reported and other negative results have likely gone unpublished based on discussions with colleagues in the field.  The dominant mechanism for enhancement of solar conversion efficiency is usually assumed to be increased absorption of the incident solar radiation through coupling to surface plasmons of the metal.  However, there may be other interactions between the organic materials and the metal that can be either favorable or unfavorable for overall performance.  We are probing these interactions by exploiting the surface-enhanced Raman effect, which selectively enhances the vibrational Raman scattering from molecules on or very near the surface of plasmonically active nanostructures.

Solar conversion efficiency enhancement has been reported not only in configurations where the metal nanostructures are in direct contact with the light-absorbing polymer but also when they are spaced several tens of nm away.  This is somewhat surprising given the very strong distance dependence of the electromagnetic field enhancement due to surface plasmons.  Our initial studies focused on the polyelectrolyte poly(3,4-ethylenedioxythiophene) polystyrenesulfonate (PEDOT:PSS), which is often incorporated between the active organic layer and the ITO electrode to improve hole collection and is in direct contact with the metal in some of the configurations reported to produce solar conversion enhancement.  PEDOT in commercially available blends with PSS is almost entirely in a positively charged, oxidized form that absorbs negligibly in the visible region of the spectrum.  We have performed surface-enhanced Raman experiments in configurations where colloidal Ag or Au nanoparticles are deposited on glass either prior to (“bottom”) or after (“top”) spin-coating PEDOT:PSS.  In each case a layer of polystyrene is spin-coated on top of the structure for protection.  Spectra have also been obtained from samples in which the PEDOT has been chemically reduced by treatment with hydrazine.  The reduced form absorbs strongly in the visible region of the spectrum.

The plasmonic Raman enhancement averaged over the highly heterogeneous samples is a factor of 5-10, varying somewhat with excitation wavelength and somewhat higher with Ag than with Au.  Many features in the spectrum are sensitive to oxidation state, the presence of metals, and/or excitation wavelength.  However, spectra of the initially reduced material with Ag in the “bottom” configuration show features indicative of oxidized PEDOT.  While these may represent changes in the morphology of the conjugated chains rather than changes in oxidation state, they are intriguing in view of the report that non-emissive "black spots" in organic polymer LEDs arise from dedoping (reduction) of PEDOT.   Nanoscale Ag may be protective against this source of damage.  However, spectra obtained in the presence of Ag at higher light intensities also exhibit new lines in the 950-1100 cm^-1 region that likely result from addition of oxygen to the thiophene ring sulfur.  Although the light intensities in these experiments greatly exceed those in functioning solar cells, accumulated photodamage over long periods of use may be a significant factor limiting the use of silver in these devices.

More recently we have extended these studies to poly(3-hexylthiophene):[6,6]-phenyl C₆₁-butyric acid methyl ester (P3HT:PCBM), one of the most common polymer blends used as the active (light absorbing and charge transporting) material in bulk heterojunction organic polymer solar cells.  We also observe some Raman enhancement with both Ag and Au nanoparticles in various geometries, but there is little evidence for metal-induced changes in morphology or oxidation state or for metal-induced photochemistry.  Interestingly, in some geometries we do observe enhancement of the Raman scattering from P3HT even when it is spaced by several tens of nm from the metal by a layer of PEDOT:PSS.

Our work on PEDOT:PSS was published about the same time as the ACS national meeting in San Francisco in March.  At that meeting I met several of the key players in this field, including a couple who had tried unsuccessfully to observe Raman enhancement of these materials with nanoscale metals.  This meeting led to at least one additional speaking invitation.  More importantly, it led to a collaboration with the group of David Ginger at the University of Washington, who had just shown that triangular silver nanoprisms increase the yield of photoinduced charge carriers in P3HT:PCBM.  They are able to control the size of their nanoprisms such that the plasmon resonances can be tuned across the visible spectrum.  We have received several samples from them and have studied their Raman spectra and Raman enhancements as a function of excitation wavelength.  The Raman enhancements roughly follow what is theoretically expected based on the extinction spectra of the nanoprisms, consistent with a pure electromagnetic enhancement mechanism.  We are continuing to study the Ginger group’s samples and are also working on the methods for synthesizing triangular nanoprisms ourselves so that we can more easily carry out these experiments.