HIGH PERFORMANCE POLYMER PHOTOVOLTAICS

BACKGROUND

An economic need exists for a practical source of renewable energy that will genuinely reduce dependence upon fossil fuels. Silicon-based solar energy systems have been touted for years as a potential candidate. However, the capital-intensive nature of silicon manufacturing processes contributes to a cost structure that falls significantly short of commercial success. Photovoltaic cells, or solar cells, based on Inherently Conductive Polymers (ICPs) (or conducting polymers or conjugated polymers such as polyacetylene, polythiophene, or poly(phenylene vinylene) offer great potential as significantly lower cost devices because these polymers can be handled like inks in conventional printing processes. A need exists to enable commercialization of conductive polymer solar cells with a versatile technology platform to drive solar cell performance.

Alternative sources of energy, especially renewable energy, are being sought to dramatically change the functional and cost boundaries resulting from current energy sources. This need is heightened by the rapidly increasing cost, environmental impact, and geo-political implications of the world's reliance on fossil fuels. Regulations from the global (e.g. Kyoto) to local level increase the demand for cost-effective renewable energy supply. The use of the sun's rays to create power represents an attractive, zero-emission source of renewable energy.

Many constituencies can benefit from solar technology including, for example, (1) Portable Electronics Manufacturers/Consumers: Solar technology for battery charging will reduce device weight and size and extend usage time; (2) Residential Consumers: Solar technology will reduce overall power costs and may provide electricity to millions of households that currently have no access; and (3) Governments and Industry: Solar technology will reduce reliance on fossil fuels and meet environmental objectives.

Silicon-based solar cells, first demonstrated over 50 years ago (Perlin, John "The Silicon Solar Cell Turns 50" NREL 2004), are the primary technology in the current $5 Billion solar cell market. However, the installed cost of this technology is approximately five to ten times that of traditional power sources. Thus, its cost/performance structure does not facilitate broad market adoption. As a result, solar energy accounts for much less than 1 percent of the nation's current energy supply. In order to expand this reach, and

meet the growing need for renewable energy sources, novel alternative technologies are required.

Conductive polymers are a key component of a new generation of solar cell that promises to significantly reduce the cost/performance barrier of existing solar cells. The primary advantage of a conductive polymer solar cell is that the core materials and the device itself can be manufactured in a low-cost manner. The core materials — similar to plastics - are made in industrial sized reactors under standard thermal conditions. They can be solution processed to form thin films or printed by standard printing techniques. Thus, the cost of a manufacturing plant is orders of magnitudes less expensive than a silicon fabrication facility. This creates a low total-cost solar cell manufacturing platform. Furthermore, conductive polymer solar technology presents flexible, light weight design advantages compared to silicon-based solar cells. While this technology holds great promise, commercialization hurdles remain. A need exists to find compositions which substantially improve performance, hopefully with minimal changes in composition.

Polythiophenes have been used in photovoltaic devices but as high molecular weight materials and/or broad molecular weight distribution generally. For example, one reference describes photovoltaic applications using polythiophenes having reported molecular weight of 87,000. See, Reyes-Reyes et al, Applied Physics Letters, 87, 083506 (2005). Another reference describes photovoltaic applications using polythiophenes having reported molecular weight of 21,100. See, Kim et al, Applied Physics Letters, 86, 063502, (2005). Another reference describes photovoltaic applications using polythiophenes having reported molecular weight of 100,000. See, Yang et al, NanoLetters, 2005, vol. 5, no. 4, 579-583. Molecular weight and polydispersity have often not been recognized as an important parameter to impact performance and in many cases higher molecular weights are taught, suggested, or directly noted as desirable or required. See for example (1) Schilinsky, P et al., Chem. Mater. 2005, 17(8), 2175-2180; (2) Kline, R. et al., Macromolecules (2005), 38(8), 3312-3319; (3) Neher, et al., Adv. Funct. Mater. 2004, 14, 757 — 794. Also, fractionation is not a commercially attractive way to control molecular weight.

A need exists to better understand how to achieve high active layer PV performance in relation to polymer microstructure and morphology.

SUMMARY

Advantages of one or more embodiments present invention include improved photovoltaic performance including efficiency and current densities by controlled use of polymer chemistry. In addition, improved processing and versatility can be achieved.

In one embodiment, the invention provides a composition comprising: at least one regioregular polythiophene having a number average molecular weight of about 9,000 or less as measured by gel permeation chromatography in chloroform, at least one n-type semiconductor.

In another embodiment, the invention provides a photovoltaic device comprising a plurality of electrodes and at least one active layer comprising at least one regioregular polythiophene having a number average molecular weight of about 9,000 or less as measured by gel permeation chromatography in chloroform, and at least one n-type semiconductor.

In another embodiment, the invention provides a composition comprising: at least one regioregular polythiophene having a number average molecular weight of about 9,000 or less as measured by gel permeation chromatography in chloroform, at least one n-type semiconductor, and at least halogenated aromatic solvent.

In another embodiment, the invention provides a method of forming a film comprising: preparing a composition comprising: (i) at least one regioregular polythiophene having a number average molecular weight of about 9,000 or less as measured by gel permeation chromatography in chloroform, (ii) at least one n-type semiconductor, and (iii) at least halogenated aromatic solvent; removing solvent to prepare a film comprising the regioregular polythiophene and the n-type semiconductor, and annealing the film.

In a preferred embodiment to achieve the most desired performance, the number average molecular weight is about 6,000 to about 9,000 as measured by GPC in chloroform. Also, the polythiophene can have a polydispersity index of about 1.1 to about 1.5.

BRIEF DESCRIPTION OF THE FIGURES Figure 1 illustrates a typical conductive polymer photovoltaic (solar) cell.

Figure 2 illustrates ground state of a p/n junction (left) and exciton formation and dissociation (right). The exciton is the bound hole (+) and electron (-). Arrows indicate the dissociation of the electron to the n-type material and conduction out of the device via the aluminum cathode. The hole (+) is conducted to the ITO anode. The resultant energy is given by the charge q times the open circuit voltage V ₀₀ .

Figure 3 illustrates current voltage curves for working example 1.

Figure 4 illustrates current voltage curves for working example 2 (comparative).

Figure 5 illustrates current voltage curves for working example 3.

DETAILED DESCRIPTION

AU references cited herein are hereby incorporated by reference in their entirety.

Electrically conductive polymers are described in The Encyclopedia of Polymer Science and Engineering, Wiley, 1990, pages 298-300, including polyacetylene, poly(p- phenylene), poly(p-phenylene sulfide), polypyrrole, and polythiophene, which is hereby incorporated by reference in its entirety. This reference also describes blending and copolymerization of polymers, including block copolymer formation.

In one embodiment, the invention provides a composition comprising: at least one regioregular polythiophene having a number average molecular weight of about 15,000 or less as measured by gel permeation chromatography, and at least one n-type semiconductor. The compositions can be used as active layers in photovoltaic devices. PHOTOVOLTAIC/SOLAR CELL DEVICES

Solar cells are generally described including several preferred embodiments. A traditional conductive polymer solar cell can comprise five components (see for example Figure 1). A transparent electrode such as indium tin oxide (ITO) coated onto plastic or glass can function as the anode. It can be approximately 100 run thick and allow the light to enter the solar cell. The anode can be coated with up to 100 run of a hole injection layer (HIL). The HIL can planarize the ITO surface and facilitate the collection of positive charge carriers (holes) from the light-harvesting layer to the anode. The opposite electrode, or cathode, can be made of a metal such as calcium or aluminum, and is typically for example 70 run thick or more. It may include a thin conditioning layer (e.g. less than 1 run of lithium fluoride) that can increase lifetime and performance. In some cases, the cathode may be coated onto a supporting surface such as a flexible plastic or glass sheet. This electrode can carry electrons out of the solar cell and complete the electrical circuit.

There can be a junction of p- and n-type semiconductors (approximately 100 run thick) between the electrodes. The p-type material is often referred to as the light harvesting component. This material can absorb photons (light) which energizes an electron from its "ground" state to an excited energy state, leaving behind a positive charge or "hole." This electron-hole combination can be bound together, forming what is called an "exciton." (See for example Figure 2.)

The exciton can diffuse to a junction between the p-type and n-type materials, where the charge can then be separated. The electron and "hole" charges can be then conducted through the n-type and p-type materials, respectively, to the electrodes resulting in the flow of electric current out of the cell. The n-type component can comprise a material with a strong electron affinity such as for example carbon fullerenes, derivatives thereof, titanium dioxide, or cadmium selenide, as described more fully below.

The morphology of the p/n junction can be important to solar cell performance. Excitons can form within the p-type semiconductor and diffuse. See for example, Chirvase et al., Nanotechnology, 15 (2004) 1317-1323. However, they should reach the interface of the n- and p-type materials where the charges can be separated before they relax back into the ground state, or "quench," through other processes. The diffusion length of the exciton can be on the order of nanometers and, therefore, an ideal junction is a "bulk heteroj unction" that is macroscopically interpenetrated, but microscopically has connected, phase-separated domains.

The following references, in addition to references cited in the background section, can be used in practicing the various embodiments of the claimed inventions: (1) Brabec, etal. Adv. Func. Mater. 2001, 11, 374-380; (2) Sariciftci, N. S. Curr. Opinion in Solid State and Materials Science, 1999, 4, 373-378; (3) Sariciftci, N. Materials Today 2004, 36; (4) Hoppe, H.; Sariciftci, N. S. J. Mater. Res. 2004, 19, 1924, (5) Nakamura et al., Applied Physics Letters, 87, 132105 (2005); (6) Paddinger et al., Advanced Functional Materials,

2003, 13, No. 1, January, 85; (7) Kim et al., Photovoltaic Materials and Phenomena Scell-

2004, 1371, (8) J. Mater. Res., Vol. 20, No. 12, Dec. 2005, 3224; (9) Inoue et al., Mater. Res. Soc. Symp. Proc, vol. 836, L.3.2.1; (10) Li et al., J. Applied Physics, 98, 043704 (2005), .

REGIOREGULAR POLYTHIOPHENE

Regioregular polythiophenes are known in the art. They can be a homopolymer or a copolymer, including a block copolymer. The regioregular polythiophene can comprise a 3-substituted polythiophene including a 3-alkyl or a 3-alkoxy polythiophene. Side groups

can be substituted with heteroatoms including oxygen and sulfur. The regioregular polythiophene can be soluble in organic solvents. The degree of regioregularity can be, for example, at least 85%, or at least 90%, or at least 95%, or at least 98%.

In particular, photovoltaic applications for these polythiophenes are described in, for example, US provisional application 60/661,934 filed March 16, 2005 to Williams et al., and US regular application 11/234,373 filed September 26, 2005 to Williams et al., which are incorporated by reference in their entireties.

In addition, synthetic methods, doping, and polymer characterization, including regioregular polythiophenes with side groups, is provided in, for example, U.S. Patent Nos. 6,602,974 to McCullough et al. (describing block copolymers and nanowire morphology) and 6,166,172 to McCullough et al. (describing GRIM polymerization method), which are hereby incorporated by reference in their entirety. Additional description can be found in the article, "The Chemistry of Conducting Polythiophenes," by Richard D. McCullough, Adv. Mater. 1998, 10, No. 2, pages 93-116, and references cited therein, which is hereby incorporated by reference in its entirety. Another reference which one skilled in the art can use is the Handbook of Conducting Polymers, 2 ^nd Ed. 1998, Chapter 9, by McCullough et al. _} "Regioregular, Head-to-Tail Coupled Poly(3-alkylthiophene) and its Derivatives," pages 225-258, which is hereby incorporated by reference in its entirety. This reference also describes, in chapter 29, "Electroluminescence in Conjugated Polymers" at pages 823- 846, which is hereby incorporated by reference in its entirety. Additional references include: McCullough, R. D.; Lowe, R. S. J. Chem. Soc, Chem. Commun. 1992, 70; McCullough, R. D.; Lowe, R. D.; Jayaraman, M.; Anderson, D. L. J. Org. Chem. 1993, 58, 904; Lowe, R. S.; Khersonsky, S. M.; McCullough, R. D. Adv. Mater. 1999, 11, 250; McCullough, R. D.; Tristram-Nagle, S.; Williams, S. P.; Lowe, R. D.; Jayaraman, M. J. Am. Chem. Soc. 1993, 115, 4910.

Polythiophenes are described, for example, in Roncali, J., Chem. Rev. 1992, 92, 71 1; Schopf et al., Polythiophenes: Electrically Conductive Polymers, Springer: Berlin, 1997.

Polymeric semiconductors are described in, for example, "Organic Transistor Semiconductors" by Katz et al., Accounts of Chemical Research, vol. 34, no. 5, 2001, page 359 including pages 365-367, which is hereby incorporated by reference in its entirety.

Block copolymers are described in, for example, Block Copolymers, Overview and Critical Survey, by Noshay and McGrath, Academic Press, 1977. For example, this text describes A-B diblock copolymers (chapter 5), A-B-A triblock copolymers (chapter 6), and

-(AB) _n - multiblock copolymers (chapter 7), which can form the basis of block copolymer types in the present invention.

Additional block copolymers including polythiophenes are described in, for example, Francois et al., Synth. Met. 1995, 69, 463-466, which is incorporated by reference in its entirety; Yang et al., Macromolecules 1993, 26, 1188-1190; Widawski et al., Nature (London), vol. 369, June 2, 1994, 387-389; Jenekhe et al., Science, 279, March 20, 1998, 1903-1907; Wang et al., J. Am. Chem. Soc. 2000, 122, 6855-6861 ; Li et al., Macromolecules 1999, 32, 3034-3044; Hempenius et al., J. Am. Chem. Soc. 1998, 120, 2798-2804. MOLECULAR WEIGHT

Polymer molecular weights and gel permeation chromatographic measurements thereof are known in the art: see for example Billmeyer, Textbook of Polymer Science, 3 ^rd Ed., 1984, including chapter one; Allcock & Lampe, Contemporary Polymer Chemistry, 1981 including chapters 14 and 15.

The at least one regioregular polythiophene can have a number average molecular weight of about 9,000 or less. The number average molecular weight can be, for example, about 1,000 to about 9,000, or about 2,500 to about 9,000, or about 6,000 to about 9,000, or about 7,000 to about 9,000. Molecular weight is measured by methods described in or substantially analogous to the working examples below and as known by those skilled in the art using GPC and chloroform solvent. This provides a uniform method to compare impact on PV performance.

If the regioregular polythiophene is a block copolymer comprising the regioregular polythiophene segment and a non-regioregular polythiophene segment, then the weight average molecular weight refers to the weight of the regioregular polythiophene segment.

Molecular weight polydispersity index can be narrowly controlled to be for example about 1 to about 1.5, or about 1.1 to about 1.5. The lower polydispersity index can enhance morphology of the active layer so as to increase charge mobilities. This can increase fill factor and short circuit current. DOPING

The regioregular polythiophene can be used without doping. Compositions can be formulated without express addition of dopant. N-TYPE SEMICONDUCTOR

The n-type component or electron acceptor can comprise a material with a strong electron affinity, good electron accepting character, such as particles, microparticles, and

nanoparticles, inorganic particles, organic particles semiconductor particles, carbon fullerenes, fullerene derivatives, soluble fullerenes, titanium dioxide, cadmium selenide, and perylenes or perylene derivatives. The n-type semiconductor can be a molecular material, or a non-polymeric materials, having a molecular weight less than about 2,000 g/mol or less than about 1,000 g/mol. The n-type component can be any component providing a bulk heterojunction structure with the regioregular polythiophene. PCBM is a preferred example, and the structure of PCBM is known in the art. The n-type semiconductor should provide fast transfer, good stability, good solubility, and good processability.

N-type semiconductors of the fullerene and fullerene derivative type can be found in for example US Patent Nos. 5,454,880 and 5,331,183 to Sariciftci et al, which are hereby incorporated by reference in their entireties. Other N-type semiconductors are described in references cited herein. AMOUNT OF COMPONENTS

Weight ratio between the regioregular polythiophene and the n-type semiconductor can be controlled to achieve the desired photovoltaic effects. For example, the weight ratio can be about 4:1 to about 0.5:1, or about 3:1 to about 0.5:1, or about 2.5:1 to about 1:1. The amount can be tailored with one or more other parameters such as for example molecular weight, solvent selection, casting or coating conditions, and annealing temperature and time. FILM FORMATION/PRINTING

Conventional methods can be used to convert compositions to solid form including thin film form and printed forms. For example, the solids can be dissolved or dispersed in one or more solvents, including organic solvents, and roll coated, screen printed, spin cast or ink jet printed, and other known coating and printing methods. Other methods are described in the references cited herein.

Solvent used for blend formation and film casting can be for example an organic solvent, an aromatic solvent such as toluene or xylene, a haloaromatic or halogenated aromatic solvent, a chlorinated aromatic solvent, a chlorinated solvent such as for example chlorobenzene, 1,2-dichlorobenzene, chloroform, or 1,2-dichloroethane. The solid content of the solution or dispersion can be varied to achieve the best PV properties and processing. For example, solid content can be about 20% to about 80%, or about 30% to about 70%.

Film thickness can be for example about 10 nm to about 500 nm, or about 50 nm to about 250 nm, or about 100 nm to about 200 nm.

Films can be thermally annealed as desired. Annealing temperature and time can be adjusted to achieve a desired result. Annealing temperature can be for example about 100 ⁰ C to about 170 ⁰ C, or about 105 ⁰ C to about 155 ⁰ C. The annealing temperature can be below the melting temperature of the regioregular polythiophene. The annealing temperature can be, for example, below, at, or above the glass transition temperature of the regioregular polythiophene. Annealing temperature can be for example about 5°C to about 60 ⁰ C above the glass transition temperature. The glass transition and melting temperature can be affected by the presence of the n-type semiconductor. For example, they can be lowered. DEVICES AND PROPERTIES

Upon making the active layer, active layer properties can be measured by methods known in the art and illustrated below in the working examples. These methods can include measurement of current- voltage curves using solar simulators. Device measurements can be carried out at room temperature or 25°C. Materials can be encapsulated as desired and for example stored under inert gas conditions.

For example, open circuit voltage (V _oc ) can be measured and can be, for example, about 0.55 V or higher, or about 0.58 V or higher, or about 0.6 V or higher.

Short circuit current (I _sc ) can be measured and can be for example about 6 mA/cm ² or higher, or about 7.7 mA/cm ² or higher, or about 9 mA/cm ² or higher, or about 12 mA/cm ² or higher.

Fill factor (FF) can be measured and can be for example about 0.4 or higher, or about 0.56 or higher, or about 0.6 or higher.

Efficiency (E) can be measured and can be for example about 2.5% or higher, or about 3% or higher, or about 4% or higher, or about 5% or higher. Efficiency can be measured at under 100 mW/cm ² .

Known PV device structures can be prepared including devices comprising multiple active layers. The device can comprise a plurality of active layers including stacked active layers.

Solar cell devices are described in for example Dennler and Sariciftci, Proceedings of the IEEE, vol. 93, no. 8, August 2005, pages 1429-1439; US Patent Nos. 6,812,399, and 6,933,436, which are hereby incorporated by reference in their entireties.

METHODS OF MAKING THE COMPOSITION

One aspect is making compositions, including both solvent-based coating compositions and dry coating compositions. For example, the regioregular polythiophene can be synthesized, including purification, with controlled number average molecular weight, e.g, less than 9,000, or 6,000 to 9,000. Then, the regioregular polythiophene can be formulated with the n-type semiconductor such as for example a fullerene derivative and mixed well. Films or coatings can be prepared with solvent removal. Doping can be adjusted as desired.

WORKING EXAMPLES

Further description is provided with use of non-limiting working examples. P3HT is poly(3-hexylthiophene) and is regioregular unless otherwise specified. PCBM is the fullerene derivative, l-(3-methoxycarbonyl)-propyl-l-phenyl- (6,6)C ₆ i. PEDOT is poly(3,4-ethylenedioxythiophene). PSS is poly(styrenesulfonate).

A P3HT:PCBM-based organic photovoltaic cell is described with the device architecture πO/PEDOT:PSS/[P3HT:PCBM]/Ca/Al where low MW P3HT (8k) is shown to perform substantially better than higher MW P3HT (19k). Also, the low MW P3HT has a polydispersity index of 1.4.

Active Layer Fabrication, Deposition, and Device Fabrication:

Materials: Two P3HT polymers were used in this invention and were synthesized by the GRIM method. Two different molecular weights of the polymer (M _n = 8k and 19k) were made. Molecular weights were determined by Gel Permeation Chromatography (GPC). The GPC traces for poly(3-hexylthiophenes) were recorded on a Waters 717 Separation Autosampler in combination with a Waters 515 HPLC pump and a Waters 486 Tunable Absorbance Detector where chloroform was the eluent (flow rate 1.0 mL/min, 35 ⁰ C, λ = 254 run) with a series of three Styragel columns (10 ⁵ , 10 ³ , 100 A; Polymer Standard Services) and a guard column. Calibration based on polystyrene standards was applied for determination of molecular weights and toluene was used as an internal standard. Win GPC software package by Polymer Standards Services was used to collect and process the GPC data. The traces were analyzed by manually selecting the UV absorbance vs. elution volume data at both ends of the peak (typically Gaussian) where the baseline absorbance 'leveled off to zero absorbance and using the Win GPC molecular weight algorithms to process the selected data.

Note on MW nomenclature: The two MW versions of P3HT used in this invention were determined by the above method to be 19k, termed 'High MW,' and 8k, termed 'Low MW' and this description is used throughout.

Note about MW determination: The preceding description of MW determination methodology is an example of what is an accurate method of MW determination for poly(3- alkylthiophenes). It is known that polythiophenes aggregate at all dilutions in various solvents, which has implications for molecular weight determination by size-exclusion chromatography (or GPC) as known in the art. The aggregation of PT polymer chains leads to the formation of aggregates which have a size larger than one isolated polymer chain as known in the art. Polythiophenes also are characterized as 'rod' polymers and thus the calibration of GPC-determined molecular weights against polystyrene standards, coil- shaped polymers, leads to shifted experimentally determined MW values, due to the differences in the hydrodynamic volume of rod vs. coil polymer structure as known in the art. These factors have been determined, by calibration to NMR and other MW determination techniques, to result in the GPC determined MWs to be overestimated by a factor of 1.5 to 2.0 of the actual MW of the analyzed polymers.

Since different solvents cause PTs to aggregate differently, it follows that GPC analysis of a PT polymer in different solvents will give different MW values. This has been observed in prior efforts and chloroform, a good solvent for P3HT, was found to give the 'lowest' M _n values while THF was observed to be more variable and give M _n values ca. 1.2 to 2.0 times the molecular weight determined from a chloroform GPC analysis. Hence, chloroform is the solvent to be used.

There is no standardized approach in the literature as to which solvent is best to use for comparative GPC analysis. THF is commonly reported for GPC analyses which means that the reported molecular weights in the literature are, at best, incomparable and potentially overestimated. The molecular weights reported herein are determined with chloroform as the eluent.

NMR Measurements: NMR determination of MW:

For comparison and validation, the fiill 500 MHz ¹ H NMR spectra of a regioregular P3HT was obtained. The two small triplets at d = 2.56 - 2.64 ppm arise from of the H/H and H/Br terminated rr-P3OHT and are usually of different intensity due to the lot to lot variation of Br endcapping. They can be assigned to the benzylic methylene on each of two distinct end units and are integrated as a group. It allows a relatively accurate determination of molecular weight from the integration of end-group resonances relative to the bulk polymer. DP _n for the aforementioned polymer can be calculated from the reported integrands and equals to the ratio of the two triplets, eg. The integrand for the large triplet of the polymer backbone is 2.031 arbitrary units (au) and for the two small triplets, taken together, is 0.056 au. The calculation of the ratio of the end groups to the backbone is 2.031 *2 (2 H's - backbone of polymer) to 0.056 (4 H's - two methylenes on the end groups) and results in 72 monomer units (MW = 166 g/mole for each repeat unit) corresponding to M _n = 14,000 (> 98% of HT). The molecular weight of the polymer was also measured by GPC with chloroform as the eluent (flow rate 1.0 mL/min, 35 ⁰ C, 1 = 254 run) and calibration based on polystyrene standards was applied for determination of molecular weights using toluene as an internal standard (Af _n : 26900, PDI: 1.3). GPC is known to overestimate the MW of PAT's by a factor of 1.5-2 so this is consistent with this NMR characterization.

Active layer solution preparation: P3HT was first dissolved in 1,2-dichlorobenzene (DCB) to make a 30 mg/ml solution, followed by blending with PCBM (purchased from ADS, used as received) in 50 wt% dichlorobenzene. The blend was stirred for 14 h at 40 ⁰ C in a dry nitrogen atmosphere. The ratio of P3HT to PCBM was 2:1.

Device fabrication: Before device fabrication, the ITO-coated glass substrates were cleaned by ultrasonic treatment in detergent, deionized water, acetone and isopropyl alcohol sequentially. A thin layer (30 run; of PEDOT/PSS (Baytron P VP AI 4083) was spin-coated to modify the ITO surface. The films were baked at 120 ⁰ C for 1 h in a nitrogen-filled glove box (< 2 p.p.m. O2 and H2O. The active layer was obtained by spin- coating the blend at 600 rpm. for 60 s and the thickness of film was about 150 nm, as measured with a calibrated profilometer. The active device area was 0.09 cm ² . The films were thermally annealed at 110 ⁰ C for 10 minutes.

The polymer PV devices were fabricated by spin-coating a blend of P3HT/PCBM in 1 : 1 wt/wt ratio, sandwiched between a transparent anode and a cathode. The anode consisted of glass substrates pre-coated with indium tin oxide (ITO), modified by spin- coating polyethylenedioxythiophene/polystyrenesulphonate (PEDOT/PSS) layer (30 nm), and the cathode consisted of Ca (5 nmjcapped with Al (150 ran). The base pressure was 2 - 6 10 ^*6 mBar. Encapsulated devices were tested in air under simulated AMI .5G irradiation (100 mW/cm ² ) using a xenon-lamp-based solar simulator (Oriel 300W Solar Simulator).

Note on Efficiency Nomenclature: Examples 1 and 2 were analyzed without a calibrating the solar simulator (Oriel) and thus the efficiencies are normalized with the high MW P3HT cell characterized as '100' and the low MW P3HT cell as '120,' a 20% improvement in efficiency. Example 3 was analyzed with a calibrated solar simulator (100 mW/cm ² ) and the efficiency value for this cell is absolute.

Example 1 : Low MW P3HT:PCBM active layer (MW = 5k), unoptimized, data are provided in Figure 3. Note 20% greater efficiency with this material over the higher MW analog. Note: uncalibrated solar simulator (AM 1.5) was used as the light source and normalized efficiencies are used for comparison purposes.

Example 2 (COMPARATIVE): High MW P3HT:PCBM active layer (MW = 15k), unoptimized, data are provided in Figure 4. Decreased fill factor and current lend to a decreased efficiency as compared to the low MW version. Note: uncalibrated solar simulator (AM 1.5) was used as the light source and normalized efficiencies are used for comparison purposes.

Example 3: Low MW P3HT taken and optimized for increased efficiency. 2.5% efficiency is reported with calibrated solar simulator with AM 1.5 solar spectrum. Data are provided in Figure 5.