We present a comprehensive study of the optical and electrical properties of transparent conductive films made from precisely tuned ratios of metallic and semiconducting single-wall carbon nanotubes. The conductivity and transparency of the SWNT films are controlled by an interplay between localized and delocalized carriers, as determined by the SWNT electronic structure, tube-tube junctions, and intentional and unintentional redox dopants. The results suggest that the main resistance in the SWNT thin films is the resistance associated with tube-tube junctions. Redox dopants are found to increase the delocalized carrier density and transmission probability through intertube junctions more effectively for semiconductor-enriched films than for metal-enriched films. As a result, redox-doped semiconductor-enriched films are more conductive than either intrinsic or redox-doped metal-enriched films.

Transparent conducting (TC) films of single-walled carbon nanotubes (SWNTs) have the potential to replace conventional TC oxides in a wide variety of optoelectronic devices. [1][2][3][4][5] TC-SWNT films are particularly attractive for photovoltaics (PV) due to their high transparency over much of the solar spectrum, excellent electrical conductivity, and the potential for inexpensive roll-to-roll processing. SWNT films have been used, by us and others, in cadmium telluride, [6] copper indium gallium diselenide, [7] and organic PV (OPV) devices. [8][9][10] In several reports, SWNTelectrodes for OPV have been prepared by filtering a sodium dodecyl sulfate (SDS)-stabilized dispersion of SWNTs to form a thin film. [9,10] The film can be released by dissolution of the filter, and then transferred to a transparent substrate.[11] This so-called ''transfer method'' produces highly transparent films with excellent conductivity, but the films possess irregular morphologies and significant roughness, which can lead to short-circuits and overall poor reproducibility during device fabrication.[12] Moreover, the process is not scalable. TC-SWNT films have also been produced for optical and electrical studies by air-brush spraying using surfactant-stabilized SWNT inks. Such films are inhomogeneous because SWNTs sprayed from surfactant solutions agglomerate on heated substrates. To move TC-SWNT electrodes beyond the proof-of-concept stage for PV and other optoelectronic applications, methods are required for producing large-area, transparent, conducting SWNT films that are smooth and homogeneous over large areas.Here, we report methods to prepare SWNT films with high transparency, electrical conductivity, and uniformity, with exceptionally low surface roughness, on arbitrarily large (6 inch  6 inch) substrates by ultrasonic spraying. A side-by-side comparison of OPV devices fabricated on SWNT and indiumdoped tin oxide (ITO) electrodes showed very good performance with energy-conversion efficiencies of $3.1 and 3.6%, respectively, under AM 1.5 illumination. Several factors are critical to the success of the approach. First, we prepared aqueous SWNT dispersions using a high-molecular-weight (MW $90 000) polymeric derivative of cellulose (sodium carboxymethyl cellulose (CMC)). CMC has been previously reported as an excellent agent for dispersing SWNTs in water, [13] and transparent films have been drop-cast, [14] but this is the first report of CMC-based dispersions for spraying SWNT films. A second advance is the use of ultrasonic spraying, which, when combined with the CMC-based dispersions, permits precise amounts of SWNTs to be reproducibly and uniformly dispensed over arbitrarily large areas. In fact, by measuring the weight and optical properties of films as a function of the number of deposited layers, the SWNT absorption coefficient could be accurately determined. Finally, we used SWNTs produced by laser vaporization (LV), which have lower defect densities [15,16] than tubes produced by chemical vapor deposition (CVD). Th...

Improving costs and scale reflect looming opportunities

No abstract

We report two viable organic excitonic solar cell structures where the conventional In2O3:Sn (ITO) hole-collecting electrode was replaced by a thin single-walled carbon nanotube layer. The first structure includes poly(3,4-ethylenedioxythiophene) (PEDOT) and gave a nonoptimized device efficiency of 1.5%. The second did not use PEDOT as a hole selective contact and had an efficiency of 0.47%. The strong rectifying behavior of the device shows that nanotubes are selective for holes and are not efficient recombination sites. The reported excitonic solar cell, produced without ITO and PEDOT, is an important step towards a fully printable solar cell.