MULTI PASS SPRAY DEPOSITION OF P OLYMER THIN FILMS FOR SOLAR CELL APPLICATIONS A n Undergraduate Thesis Presented to the Faculty of the University of Florida By Jessica R. Lederman April 2011
ABSTRACT Spray deposition of polymer thin film s for photovoltaic devices is a fabrication technique that shows great promise as a low cost high throughput deposition method. The thickness profile of a single plane of sprayed poly( 3hexylthiophene) (P3HT) and1(3 methoxycarbo nyl) propyl 1 phenyl (6,6)C61 (PCBM) demonstrated the relative thickness and topographical configuration of multipass spray ed films T hickness profiles of P3HT:PCBM thin films produced under various multipass spray protocol s display distinctive film thicknesses and topography that can be controlled to a reasonable extent to produce continuous thin films over large areas Device efficiencies prove d the viability of fabricating polymer solar cell d evi ces with multipass, multi plane sprayed P3HT:PCBM thin films.
1 TABLE OF CONTENTS P age 1. Introduction .........2 1.1. Polymer Solar Cells.. .....2 1.2. Deposition Methods and Manufacturing 5 2. Experimental Setup and M ethods ...11 2.1. Spray Deposition Setup and Process ...11 2.2. Device Structure and Fabrication P rocedures .13 2.3. Device C haracterization ..18 2.3.1. Optical Transmittance Thickness Mapping ..18 2.3.2. Solar Cell Device Characterization. .....19 3. Spray Deposition Results and Discussion..20 3.1. Single Plane Thickness Profile ....21 3.2. Multiple Plan e Thickness Profile 3.3. Device Characterization Results .....36 4. Conclusion.. REFERENCES .41
2 1. Introduction 1.1. Polymer Solar Cells Developments in solar power generation are revolutionizing the contemporary energy industry. Photovoltaic technology generates electrical power from light energy, namely sunligh t, according to the photovoltaic effect first noted in 1839 by French physicist AlexandreEdmond Becquerel .1 Solar energy is a clean, renewable, saf e, and plentiful energy source with a promising future moving toward common use It has the potential to overcome the financial environmental, and supply issues of traditional energy sources such as coal, oil, and natural gas The grid connected photovoltaic energy capacity in the U nited S tates has strengthened significantly from 476.0 MW in 2007 to 1,256.0 MW in 2009 using traditional solar cells made from inorganic semiconductors .2,3,4 O rganic solar cells ( O SCs ) are a relatively new solar technology that promise to significantly low er manufacturing costs and serve niche markets such as solar powered tents. O SCs are fabricated using thin films of conjugated organic electronic materials including both small molecular materials and polymer s, for solar energy conversion devices The solubility of polymers makes these materials particularly useful for the fabrication of polymer solar cells (PSC) from solution. Of the most accessible PSC systems poly(3 hexylthiophene) (P3HT) and1(3 methoxycarbo nyl) propyl 1phenyl (6,6)C61 (PCBM) have been well studied and produce efficient solar cells with power conversion efficiency of 45% C ommercially available photovoltaic cells are created using inorganic semi conductors, most notably silicon which require s complex fabrication and high costs Interest in PSC s stems from a need to establish an alter na tive to inorganic solar energy systems with lower cost and simplified fabrication; the growth of the organic electronics industry could satisfy this need
3 PSC technology is significant in that semi conducting polymers allow fabrication of light weight and ine xpensi ve solar cells on flexible substrates. Organic materials for PSC devices typically have much lower material cost s than inorganic semiconducting material s and can b e processed via simple and inexpensive methods at room temperature that require low energy input Moreover, organic materials are capable of employing flexible substrates that are advantage ous due to their inexpensive and lightweight nature making them ideal for high throughput processing. F lexible substrates permit integration into existing structures on various surfaces and geometries increas ing the usability of solar energy for both permanent and portable applications Furthermore, flexible solar cells boast lower transportation, maintenance, and installation co sts compared to rigid heavier products like glass substrates .5 Solar cells are an environmentally friendly alternative to conventional power generation methods given that solar energy is a plentiful a nd ever present energy source. PSCs in particular, may be easily recycled due to the low melting point of the polymer materials The ir lightweight nature renders them superior for storage and distribution, reducing the carbon footprint and costs associated with their use Moreover, advances in polymer thin film deposition methods have significantly reduced material waste during PSC production to minimal quantities and require low energy output with room temp e rature processing W ith the global supply of inorganic solar cell material s becoming increasingly rare, semiconducting polymers present an available alternative source PSC devices perform light to electrical power conversion in a different manner than inorganic photovoltaic devices. The photogeneration process in PSCs is a succession of four m ajor steps : absorption, exciton generation, separation of excitons into free electrons and holes and free carrier collection. Absorption of incident photons within the device begins the process
4 of solar energy conversion and is significantly affected by the absorption coefficients of the materials within the device as well as the surface characteristics of the deviceair interface I norganic semiconductor solar cells generate free electron hole pairs upon photon absorption in the material whereas PSC s generate bound electronhole pairs called excitons. In the exciton generation stage, electron hole pairs are produced on a single molecule within a mixed donor acceptor heterojunction. Next, these chargeless excitons can diffuse from molecule to molecule within the bulk material, but can also recombine If the excitons first encounter an interface before recombination, they can be separated into free electrons and holes. This gives rise to holes in the highest occupied molecular orbital (HOMO) of the donor material and electrons in the lowest unoccupied molecular orbita l (LUMO) of the acceptor material illustrated in Figure 1. Collection of charge at the electrodes completes the process of solar cell operation when the electrons and holes are transported through the respective accept or and donor materials toward the anode and cathode, producing a photocurrent. Solar cell efficiency is heavily dependent on the number of independent charge carriers that are generated and thus the exciton generation and exciton separation phases are considered most vital .6,7 Figure 1. Illustration of exciton separation in a donor acceptor heterojunction structure of a PSC.
5 Prog ress in PSC research has focused on increasing device efficiencies, increasing the semiconducting polymers stability, and increasing the strength of PSCs Although current PSC s have efficiencies much lower than most inorganic solar cells, research is progressing rapidly to close this efficiency gap In 2010, Konarka Technologies demonstrated their world recordbreaking organic based photovoltaic solar cells with efficiencies reaching as high as 8.3%.8 It should be noted that r apid degradation of semiconducting polymers due to their chemical instability is a property of concern for the reliability of PSCs This could become a plus in the future if degradation properties of semiconducting polymers could improve the safety of solar cells in the environment. In addition, many properties of organic materials can be manipulated for specific applications an d, with recent breakthroughs in chemistry improvements in device efficiency and lifetime s are occurring as we speak The PSC field, while relatively young is progress ing toward larger scale accessibility through scientific breakthroughs 1.2. Deposition Methods and Manufacturing In recent years, t he advancement of p rocessing methods for fabricating polymer based thin film devices, and the resulting decreases in production costs, make the realization of mass produced PSCs increasingly within reach. Since research began in the 1980s on the use of polymers for photovoltaic s interest in PSC s has grown rapidly due to significant increases in power conversion efficiencies of devices These developments were aided largely by advances in materials processing.9 ,10 Contemporary d eposition methods of polymer thin films can be grouped into two categories : vapor phase dep osition and solution processed deposition. A variety of methods exist for the fabrication of organic thin films from the vapor phase Prominent vapor phase deposition methods including vacuum thermal evaporation, pulsed laser
6 deposition, sputtering, and plasma polymerization, require that deposition takes place under vacuum conditions .11,12 As a consequence of vacuum conditions as well as the nature of vapor phase deposition, t hese techniques have high startup costs inefficient materials use, high energy expenditures, and low throughput Vacuum deposition processes can be scaled to large scale manufacturing but require a very significant initial investment in equipment The creation of a vacuum during vapor phase deposition guarantees that cells must be made in batch processes. Batch processing techniques are low throughput methods that are not conducive to roll to roll processing. Creation of a vacuum during these processes also requires significant energy expenditure, therefore adding to the e xpense of producing PSC s in this manner Furthermore, organic materials are highly sensitive to process parameters such as temperature whic h can be difficult to control with vapor phase deposition and can reach regions that are unsuitable for processing organic materials. Notable s olut ion processed thin film deposition methods include spray deposition, solution casting spin coating, inkjet printing, and slot di e coating .13 Solution deposition methods deposit polymer thin films dissolved in a solvent such as chlorobenzene, toluene, and chloroform. These methods are ideal for PSC fabricatio n since they are simple processes that can be performed at room temperature. Solutionbased processing require s significantly less energy than vapor phase deposition methods and most methods have much higher throughput, making solution processed organic solar cells less expensive than vapor phase processed solar cells. Processing conditions such as pressure and temperature can be controlled with accuracy and precision for solution phase deposition techniques .13 It should be noted that s olution processed thin films do have drawbacks such as the difficulty of producing multilayer stacks, the reduced control of nanoscale morphology and the potential for solvent residue to remain and
7 influence electronic behavior Still, optimization of solution phase deposition methods show s significant promise in the fabrication of low cost solar cells Some customary solution phase thin film deposition methods, particularly spin coating and solution casting are not conducive to large scale manufacturing. Spin coating, the most commonly used thin film deposition method for PSC device fabrication produces films with highly uniform thickness as determined by spin speed and duration. However, t his method fabricates films in a time consuming batch process similar to those of vacuum deposition methods. Other s olution processed thin film deposition methods particularly spray deposition, slot die coating, and inkjet printing are transformable into r oll to roll processes for large scale manufacturing of PSCs In manufacturing, roll to roll production is the combination of process es for fabricating electronic devices on a roll of flexible pl astic or metal D evices are deposited, patterned and packaged on the roll in an assembly line like process as in Figure 2 Figure 2. Mod el of an organic photovoltaic r oll to roll manufacturing process Conversion of de vice fabrication to largescale, high volume processes is vital to pushing organic electronics into the mainstream of technology Roll to roll production methods have the potential to drive PSC s to the forefront of energy technology benefitting from the combination of high speed production methods While v apor phase thin film deposition methods can be converted to roll to roll processes, they are not conducive to larg e scale and low cost production
8 of PSC s seeing as they require high energy expenditure, high costs and significant periods of time for deposi tion. Application of r oll to roll p rocessing could create a crucial reduc tion in the price of commercially available solar cells to make solar power more prevalent .14 15,16 Spray deposition of solutionprocessable polymers is an especially attractive technique for thin film deposition owing to its suitability for low cost industrial scale manufacturing processes Of the deposition techniques mentioned in this work s pray technology has the unique ability to deposit thin films on geometrically complex surface contours .17 Vak et al were the first to demonstrate nearly comparable power conversion efficiencies of spray coated polymer solar cell devices and spin coated devices when fabricated in air with power conversion efficiencies of 2.83% and 2.93%, respectively .18 Girotto et al. took this a step further demonstrating similar power conversion efficiencies of all spray coated to all spin coated polymer solar cells using a P3HT:PCBM blend, producing a 3.52% power conversion efficiency in spray coated devices and a 3.62% power conversion efficiency in spin coated devices .19 Moving forward, a great deal of work has been done examining fabrication parameters and equipment to improve spray de posited device efficiency. Several different spray nozzle technologies including airbrush, electrodynamic, and ultrasonic spray nozzle systems are commercially available for use in processing polymer thin films for solar cells Many standard spray techniques in the coatings industry utilize either compressed air or high pressure to produce droplets. In contrast, organic electronics spray deposition techniques often utilize ultrasonic spray systems that operate using very high frequency vibrations to produce extremely consistent droplets on the scale of picoliters.21 22,23 Droplets form when capillary waves are generated by a vibration in the nozzle and droplet size is established as a result of this vibration frequency A low vibration frequency produces large and
9 coarse droplets whereas a high vibration frequency produce s smaller, more refined droplets.21, 24 One of the greatest advantages of this technology is that u ltrasonic ally atomized droplet features are not controlled by the inert gas flow rate allowing flow rate adjustments without altering droplet size. Ultrasonic spray systems also ha ve the benefit of operating with a widerange of solution concentrations in a clog free manner and with excellent operational efficiency at very low flow rates (mL/min) Overspray, defined as the material and solvent that reflect off a substrat e due to momentum, is minimal with ultrasonic spray in comparison with other spray techniques due to the low velocity nature of the spray. The round spray pattern of ultrasonic spray nozzles is also favorable for PSC depos ition in that it produces large area coverage with a very thin film.22 24 ,25 In general, s pray deposition of organic thin films can be divided into two categories: single pass spray and multi pass spray. Single pass spray deposition applies one single pass of the nozzle head over a substr ate to produce a continuous liquid film from aggregate droplets Multipass spray deposits solution using many passes of the spray nozzle over the substrate This process works to form films layer by layer at a slow enough rate where droplets dry without coalescing forming a film characterized by the droplet qualities .25 Films fabricated via multipass spray are distinguishable by their fairly rough topography and variable morphology on a macromolecular scale. Single pass spray films have smoother topography and a more homogeneous morphology than multi pass spray films, making them more analogous to spin coated films. Girotto et al. revealed devices with the highest power conversion efficiency recorded for sprayed polymer solar cells (using P3HT:PCBM); These devices showed power conversion efficiencies as high as 3.75% and were produced by way of single pass spray techniques .19 Nevertheless, a majority of current spray coat research is focused on multi pass
10 spray deposition motivated by the prospects of forming advanced device structures using additive processes .17, 19, 20 Film morphology and surface uniformity are properties of immense importance to polymer solar cell thin films as they affect photon absorption, electronic functionality and subsequent deposition processes Multipass spray form s films of droplets that dr y immediately when they hit the substrate leaving their droplet boundaries intact. These boundaries can be observed in Figure 3 as the walls of the droplets Multipass sprayed films have fairly rough topographies due to the droplet nature of the se fil ms and can display considerable shunting in the devices because of unwanted short circuits between the front and back surface contacts. Therefore, it is important to reduce the roughness of these boundaries to a minimum by using ideal solvents and pass lo cations. Future implications of spray deposited PSC modules include the prospect of large scale integration of solar energy into everyday products and life, further revolutionizing the energy sector and improving environment wellbeing.17 ,26 Figure 3. Opti cal microscope image of a spray deposited P3HT:PCBM film. Droplet boundaries can be observed as the dark colored walls of the droplets. Scale bar signifies 0.5 mm
11 2. Experimental Setup and Methods 2.1. Spray Deposition Setup and Process An original ultrasonic spray deposition system was modified to produce a large area active layer in PSC devices The spray assembly, shown in Figure 4, displays five m ain system components : the ultrasonic nozzle (A), the nitrogen gas f low system and meter (B) the syringe pump (C), the DC motor and switch (D), and the hand crank shaft and lever (E). Figure 4. Photograph of spray deposition setup. A 1 30 kHz ultrasonic spray nozzle atomizes sprayed solution which produces an even, downward spray of droplets on the substrate in c onjunction with the nitrogengas flow system The spray nozzle is positioned at a vertical distance of approximately 2 in from substrates. The spray nozzle flow is aimed parallel to the s pray system base and the mounted threenozzle nitrogen gas flow system induces a downward motion in the spray momentarily after solution leaves the nozzle. The power and frequency provided to the ultrasonic spray nozzle are regulated at 60% power and 130 kHz by an ultrasonic power source The nitrogen gas flow meter allows reasonable control of the nitrogen flow between 0 20 scfh. An inert tube is
12 fastened to the rear of the spray nozzle, providing solution from a syringe. The syringe pump secures the syringe and regulates the flow rate of solution to the nozzle. A hotplate is set under t he system base to heat substrates to aid droplet drying. T he DC motor and switch direct the uniaxial motion of the spray nozzle along the x dimension. The motor shifts the nozzle at a constant speed by supplying a steady 12 V .25 The direction and timing of passes are controlled manually by flipping the switch between its three positions (left, neutral, and right). The hand crank shaft permits spray nozzle motion in the y dimension. A lever at the fore end of the hand crank shaft is manually turned to shift the spray nozzle (and above mounted DC motor ) Each full turn of the handle moves the nozzle a little less than 2 mm Since x and y axis motion s are performed manually, nozz le motion in both directions cannot occur in a controlled fashion at the same time. Therefore, m ultipass spray at multiple planes of substrates is performed in stages for the purpose of these experiments. In order to prepare the spray system for deposition, system components must be cleaned and positioned to maintain experiment al consistency A minimum 1 mL of solvent must be sprayed in order to clean the tube and nozzle of any other previously sprayed solutions Typically, it is ideal to employ whichever solvent will be used in the ensuing deposition process to minimize solution contamination A substrate reference location was created on the sys tem base to maximize the allowable area of deposition and maintain experimental consistency without applying extra stress to the spray system components Multipass deposition using the aforementioned system is performed in stages at different planes in the y direction relat ive to the substrates Using the DC motor switch the nozzle is directed from its home position on the left side of the system to the right side of the system, depositing a single spray pass. Another spray pass is conducted from right to left using the DC
13 motor switch to return the syste m to its home position This process is continued until t he designated amount of passes has been performed. The handcrank shaft handle is then rotated to move the nozzle a desired distance typically 4 10 mm, toward the front of the apparatus. Spray pa sses are continued from this new location until the desired numbers of passes have been completed there. This process is carried out until all passes have been performed at their designated planes. M ultipass spray deposition performed at different planes of a substrate is idyllic for implementation in rollto roll PSC manufacturing procedures. A uniform film is created u sing an even amount of passes at different locations on substrate s In this way, multiple substrates can be processed at onc e. 2.2. D evice Structure and F abrication P rocedures PSCs are typically fabricated by depositing four main layers of material on a transparent substrate : the anode, the hole transport layer, the active layer, and the cathode. The arrangement of these layers, depict ed in Figure 5, generat e an electrical current when light is absorbed through the transparent substrate A transparent conducting electrode is deposited atop the substrate acting as the anode in the system Atop the transparent anode layer lies the hole transport layer. The hole transport layer helps improve the surface quality of the anode by smoothing it out Additionally, this layer facilitates the movement of charge between the anode and the active laye r.27 The active layer of a PSC device serves as the donor acceptor layer in the structure. This is where exciton separation occurs. Excitons find an interface that is within their diffusion length in the polymer blend active layer due to the donor accep tor heterojunction structure of the device. Because excitons can find an interface with relative ease inside this device structure, these devices have high quantum efficiencies relative to other organic
14 photovoltaic devices.28 Atop the active layer lays a cathode serving as an electrical contact that completes the circuit. Figure 5. Polymer solar cell structure. The preparation and fabrication steps for the se large area PSCs are shown in the Figure 6. This protocol will further discuss each step of the production process later in this chapter. To c reate a completely roll to roll processed cell, all film layers must be deposited with compatible processes. In this study however a roll to roll compatible process will only be utilized for the spray deposition of the active layer of these PSCs Figure 6. Fabrication steps for polymer solar cell fabrication stepwise.
15 One of the most important stages in device fabrication is substrate preparation for the reason that preparation and cleaning methods can significantly affect the impending processes and device performance. For the purposes of this investigation, PSC devices were created on 1in2 transparent glass substrates with thicknesses of 1 mm. The glass substrates were purchased with a pre deposited 100 nm layer of i ndium t i n oxide (ITO). ITO is the current industry favorite anode material for PSC s owing to its low electrical resistivity tremendous environmental stability, and hig h transparency to visible light .27 2 9 ITO slides were cleaned using compressed air to re move any foreign matter from substrate surface s 1.2 cm wide copper tape pieces were fixed firmly across the cent er of the ITO glass substrates. Tape covered substrates were submerged in a n etching solution composed of 25% HNO3 and 75% HCl for 4 minutes and then immediately removed into deionized water to remove excess etchant The copper tape was then removed and excess residue was cleaned from the substrates using acetone. To compl ete substrate preparations before thin film deposition, substrates were wiped with a t exwipe to remove any surface debris. Substrates were further cleaned by way of sonication in four, 15 minute baths using a sequence of beakers filled with deionize d water + Tergitol surfactant deionized water, acetone, and isopropanol. After substrates were removed from the isopropanol solution they were blown dry with an air gun to remove as much excess solution as possible. Substrates underwent a 15 minute UVozone cleaning to increase the work function of the ITO and dissociate contaminant molecules from the surface of the substrates .30 Following these preparations, substrates were ready for deposited processes A hole transport layer composed of Poly( 3,4ethylenedioxythiophene) poly(styrenesulfonate) known as PEDOT:PSS for short was deposited using a spin coating process PEDOT :PSS is a transparent, conducting polymer mixture where PEDOT carries
16 positive charge and PSS carries a negative charge.3 1 Figure 6 shows the chemical structures of PEDOT and PSS. A s ubstrate was blown with compressed air in order to remove any surface matter a nd then positioned in the center of the spin coat ing instrument. PEDOT:PSS was placed on the substrate using a syringe t o cover the entire center area and the substr ate was spun for 60 seconds at a rate of 8000 rpm Using an isopropanol coated swab, PEDOT:PSS was removed from the edges of the substrate S ubstrates were subsequently annealed on a hot plate at 150 C for 10 minute s. During this annealing process, substrates were covered with a tin foil shield to keep particles in the air from settling on the surface. Substrates are now prepared for active layer deposition. Figure 7. Chemical structures PSC materials PEDOT, PSS, P3HT and PCBM.6 ,32 In this study, deposition of the OPV cell active layer was performed using ultrasonic spray deposition of a P3HT:PCBM blend. PSCs utilizing P3HT:PCBM, whose chemical structures are shown in Figure 7, are considered some of the most effic ient organic solar cells at this time P3HT:PCBM photovoltaic cells have excellent stability under outdoor exposure conditions, lasting up to a year.6 I n order to deposit the active layer substrates were passed into a glove box containing the spray system to maintain a controlled environment of 0.1 ppm water
17 Th e base of the spray system was heated to 50 C using the underlying hotplate A 3 mL polypropylene syringe was filled up with a 2 mg/mL solution of 1:0.8 ratio P3HT:PCBM solution in chlorobenzene A pump rate of 0.20 mL/min was utilized for the purpose of these experiment s Substrates were blown with nitrogen to remove any surface matter and placed on the heated base of the spray system for approximately two minutes. An even flow of P3HT:PCBM solution was established using a nitrogengas flow between 1013 sc f h. Multipass sp ray deposition was performed with varying number of passes on diffe rent planes of the substrates to observe the e ffects of various procedures on film thickness and uniformity. Following completion of the spray protocol, samples were immediately removed from the heated system base. To conclude the active layer deposition process the P3HT:PCBM film was wiped from the edges substrates using swabs saturated with chloroform. Blank 1 in2 or 2 in2 gla ss slides of 1 mm thickness underwent simultaneous sonication cleanings and a one minute UVozone cleaning with the ITO covered glass device substrates. Blank glass slides underwent P3HT:PCBM spray deposition in parallel with the PSC devices These slides provide d insight into the P3HT:PCBM film morpholo gy and thickness characteristics of the simultaneously deposited devices It should be noted tha t the P3HT:PCBM active layer of control devices were fabricated using a spin coating process in a glove box. After substrates were blown with nitrogen and placed in the center of the spin coater 200 L of an 18 mg/mL solution of P3H T:PCBM in chlorobenzene was deposited at a rate of 1000 rpm for 60 seconds. In a similar fashion to sprayed devices, P3HT:PCBM at the substrate edges of spin coated devices was wiped way using swabs immersed in chloroform
18 To conclude thin film deposition processes of PSCs a 100200 nm thick aluminum cathode was evaporated atop the P3HT:PCBM layer Because of the rough surface morphology of the spray coated P3HT:PCBM film an effective cathode required a thickness of at least 100 nm to create an even film with good contact The combination of the 1 cm wide large area aluminum cathode crossed over the 1.2 cm wide ITO strip created a 1.2 cm2 PSC device. A final device anneal step was performed to complete fabrication This step functioned to remove solvent and change the nanoscale morphology of the films to enhance photocurrent generation efficiency Substrates wer e blown with nitrogen to remove excess surface matter and subsequently annealed on a hotplate in a glove box at a temperature of 150 C for 30 minutes. Substrates were removed from the hotplate and set aside to cool before characterization. 2.3. Device C haracterization 2.3.1. Optical T ransmittance Thickness Mapping Optical transmittance data was used to determine the thickness and surface roughness of spray coated P3HT:PCBM films. I mages of sprayed glass slide s were taken using the white light source of a digital optical transmission microscope with a bandpass filter centered at a wavelength, of 498 nm. Sample images were taken using a constant integration time to ensure even saturation of all images when multiple images are connected Images were processed using the open source image processing software ImageJ A p lot of the films transmittance T was created by calculating the ratio of transmittance through the sprayed film and the blank glass slide. A film thickness map was computed by applying Equation 1 to the film transmittance plot
19 in ImageJ to determine film thickness, t at all locations of the plot. The a bsorption coefficient for P3HT:PCBM films at = 498 nm is = 0.0086 nm .25 = ( ) (1) Film thickness profile data were calculated in ImageJ using an average sample thickness along the width of the transmission plot A lengthwise thickness plot displays this data to show surface thickness and uniformity. 2.3.2. Solar Cell Device Characterization PSC devices were characterized by measuring the total current density of devices. These measurements were taken using an Agilent 4155C for devices in the dark as well as under illumination by incident power PIn, of 100 MW / c m2 under simulated AM 1.5 solar spectrum conditions produced by a 150 W Xe arc lamp. The maximum power, Pmax, and fill factor, FF, of devices were calculated according to Equation 2 and Equation 3, where Voc represents the open circuit voltage of the device and Jsc represents the short circuit current of the devices. Device power conversion efficiencies, PCE, were determined via Equation 4. Figure 8 demonstrates a theoretical current density voltage plot of a solar cell, showing the locations of these variables .25 = (2) = (3) = ( 4)
20 Figure 8. Th eoretical current density vs. voltage plot f or a solar cell. 3. Spray Deposition Results and Discussion The main focus of this work was to characterize the thickness profiles of spray deposited films and correlate the film quality to device efficiencies of sprayed solar cells. Single line thickness profile analysis of thin films using optical transmittance thickness mapping demonstrated the thickness of passe s completed at a single plane of a substrate. Multiple line thickness profiles created from optical transmittance thickness data aid ed in analysis of surface uniformity across films Device efficiency results based on examination of PSC devices fabricated using varied multipass spray parameters gave perspective on spray deposition requirements fo r PSCs fabricated with P3HT:PC BM active layers Table 1 displays the fabrication parameters for films examined in this wor k. Figure 9 illustrates the spray deposition configuration and corresponding designations that will be utilized throughout this work.
21 Figure 9. Illustration of the multipass spray deposition configuration where d=distance between spray planes. Table 1. Sample fabrication parameters of spray coated and spin coated films on blank glass substrates Name Deposition Method Total Passes Distance Between Planes (mm) Number of Planes Passes per Plane Sample 1 Spray 100 0 1 100 Sample 2 Spray 450 4 9 50 Sample 3 Spray 450 8 5 90 Sample 4 Spray 360 4 9 40 Sample 5 Spray 360 6 6 60 Sample 6 Spray 480 6 6 80 Sample 7 Spray 540 6 6 90 Reference Sample Spin ----3.1. Single Plane Thickness P rofile Multipass spray deposition of 100 passes was performed in the x dimension on sample 1 at a single plane in the y dimension of the 2in2 glass substrate in order to determine the distribution of the thickness profile in the y dimension that is characteristic of the spray system constructed Figure 10 presents optical microscope image s of sample 1 at a variety of locations from the plane of the spray nozzle showing the variation in droplet density along the length of
22 the sample The low droplet density in Figure 10 (a) is due to the distance of this position being greate r than 25 mm from direct spray of the nozzle In Figure 10 (c), the droplet morphology of sprayed polymer thin films is especially evident. Figure 10. Optic al microscope images of sample 1 at (a) low film thickness, (b) intermediate film thickness, and (c) high film thickness Scale bars signify 0.5 mm. Figure 11 displays the full length thickness map and 3D surface plot of sample 1 created using film transmittance data. A gradual variation in film topography of the sample can be noted with the peaks of the film congregating at points that received the most direct spray from the nozzle. As distance from direct spray increased, the peaks of the film decreased In this way, the topography of the sample formed a mountainous shape. Figure 1 1. Full length t hickness map and 3D surface plot of a single line of sample 1. (b ) (a ) (c )
23 Thickness plots presented in Figure 12 reveal the characteristic rough surface composition and highly nonuniform thicknesses of spray coated polymer thin films. In Figure 12 (a) a fairly even average thickness can be noted throughout the ascent and descent of the profile. Outlying peaks and valleys in the profile are the result of noise. Optical microscope thickness measurements do introduce some error into the system due to nonlinear absorption effects of films and the saturation of detector pixels through thin regions. Error is also present as a result of low signal to noise ratios for thick regions of film which must absorb most of the light and transmit small signals to the detectors. B ecause of the particularly wide spray radius created by the spray system, a 1 in x 2 in blank glass slide was utilized to capture as wide a spray radius as possible to ensure that a wide range of spray was captured on the substrate. Two blank slides supported the longe r than normal length of the substrate and, as with typical optical micros cope transmittance mapping, the ratio of film transmittance to glass slide transmittance was taken during image processing. However, since the two supporting blank slides could not m atch up to each other at the edges, a very minute variation in transmittance through the sprayed film arose at this junction This variation is evident in that the noise of the film thickness profile has an extremely outlying value at the location of slid e mismatch In Figure 12 ( b) adjustments were made to account for noise and to display an even distance on both sides of the spray maximum. The linear fall off of spray thickness from the maximum is likely the result of the spray pattern produced by the nozzle. Spray deposited by the nozzle is delivered at a constant rate to the substrate which layers most significantly nearest to the nozzle and decreases linearly as distance from the nozzle increases. Th is plot is more in line with the expectation that film thickness will not be an extreme maximum peak across the profile of a single plane of spray and that the apex of the profile is at the junction of 2 linear regions Although the sample profiles in Fig ure 12
24 appear as the intersection of 2 linear regions, a Gaussian fit curve was applied to the profiles for analysis purposes Figure 12. Thickness plots of sample 1 displaying (a) the f ull length film t hickness of the film and ( b) an adjusted profile to account for optical microscope transmittance nois e. Table 2 displays statistical data on the thickness profile of sample 1, as well as its adjusted thickness profile A nalysis of sample 1 revealed that at its maximum, 100 spray passes produced a thickness of 159 5 nm without accounting for optical microscope transmittance inaccuracies Taking optical microscope tr ansmittance inaccuracies into account, a maximum thickness of 124 5 nm occurred in the film A maximum value closer to the adjusted peak value appears more accurate since this peak exists as a group of outlying data point s Analysis of sample 1s full length thickness pr ofile also concluded that it had a full width at half maximum (FWHM) of 24 mm The FWHM data for the sample s profiles are shown in Table 2. Thickness mapping also ascertained that the spray setup produces a spray radius greater than 1 in on a substrate. 0 15000 30000 45000 6000 0 20 40 60 80 100 120 140 160 180 Thickness (nm)Length Along Sample (mm) 15000 30000 45000 0 20 40 60 80 100 120 140 160 180 Thickness (nm)Length Along Sample (mm) (a ) (b ) 0 10 20 30 40 50 10 20 30 40 50
25 Table 2. Statistical analysis of thickness profiles for sample 1 All thickness measurements are in units of nm with an error of 5 nm. FWHM in units of mm. Name Maximum Minimum Max Min Mean FWHM Sample 1 159 0 159 50 24 Sample 1 Adjusted 124 13 111 58 18 3.2. Multiple Plane Thickness P rofile Multipass spray depos ition was performed across the x dimension on sample s 2 and 3 at various planes in the y dimension, as outlined in Table 3, amounting to 450 total spray passes on each substrate. Table 3 also displays the total distance traveled by the spray nozzle over each substrate calculated using Equation 5. = ( 5) Table 3. Spray f abricat ion protocols for samples 2 and 3. All distances are in units of mm. Name Total Passes Dist ance Between P lanes Number of Planes Passes per Plane Total Nozzle Distance Traveled Sample 2 450 4 9 50 36 Sample 3 450 8 5 90 40 Figure 13 shows single line optical microscope image s thickness maps and 3D surface plot s of samples 2 and 3. The rough topographies observed in the 3D surface plots in Figure 13 are characteristic of ultrasonic multipass spray deposited films.
26 Figure 1 3. From left to right: s ingle line optical microscope image, fulllength film thickness map along a single line and 3D surface plot of (a) sample 2 and (b) sample 3 Film thickness profiles of sample s 2 and 3 are displayed in Figure 14. The full length thickness profile s of lines of these samples display visibly lower average film thickness es covering approximately 20% of each sample. This variation in average thickness illustrates a need for further spray passes on at least one more plane of future substrates Adjusted sample thickness profile s in Figur e 14 display thickness data altered to show plot area that does not (b) (a )
27 require further passes. These plots illustrate more even average sample thicknes se s in accordance with the desire for thickness uniformity. Figure 1 4. Full length film thickness profile s and adjusted film thickness profiles of (a) sample 2 and (b) sample 3 Adjusted profiles display thicknesses along the 80% of samples. Thickne ss plot comparisons in Figure 15 show that sample 2 has a more even average thickness than sample 4. This is statistically confirmed in Table 4 by the lower standard deviation of sample 2 than sample 4. The difference in distances between planes stated in Table 3, account for the variation in average thickness between the samples Therefore, a distance of 4 0 4000 8000 12000 16000 20000 60 80 100 120 140 160 180 200 220 240 260 Thickness (nm)Length Along Sample (mm) 4000 8000 12000 16000 20000 60 80 100 120 140 160 180 200 220 240 260 Thickness (nm)Length Along Sample (mm) 0 4000 8000 12000 16000 20000 60 80 100 120 140 160 180 200 220 240 260 Thickness (nm)Length Along Sample (mm) 4000 8000 12000 16000 20000 60 80 100 120 140 160 180 200 220 240 260 Thickness (nm)Length Along Sample (mm) (b) (a) 0 5 10 15 20 25 5 10 15 20 25 0 5 10 15 20 25 5 10 15 20 25
28 mm between spray ed planes used to fabricate sample 2 is more suitable for creating evenly thick films of P3HT:PCBM. Figure 15 Comparison of full length spray film thickness profiles and adjusted film thickness profiles of sample s 2 and 3. Table 4. Statistical analysis of thickness profiles for samples 2 and 3. All thickness data are in units of nm with an error of 5 nm. Name Maximum Minimum Max Min Mean Standard Deviation Sample 2 218 63 154 139 24 Sample 2 Adjusted 218 79 139 144 22 Sample 3 220 64 156 136 27 Sample 3 Adjusted 220 70 149 141 25 Characterization of samples 3 and 4 establish the variation in surface profiles owing to different distances between planes as wel l as the variation of overall film thickness with differing spray pass es at each plane. Table 5 presents fabrication pr ocedures used to produce sample s 4 and 5 with 360 total spray passes 0 4000 8000 12000 16000 20000 60 80 100 120 140 160 180 200 220 240 260 Sample 2 Sample 3Thickness (nm)Length Along Sample (mm) 4000 8000 12000 16000 20000 60 80 100 120 140 160 180 200 220 240 260 Thickness (nm)Length Along Sample (mm) Sample 2 Sample 3 5 10 15 20 25 0 5 10 15 20 25
29 Table 5. Spray fabrication protocols for samples 4 and 5. All units of distance are given in mm. Name Total Passes Distance Between Planes Number of Planes Passes per Plane Total Nozzle Distance Traveled Sample 4 360 4 9 40 36 Sample 5 360 6 6 60 36 Optical microscope images and 3 D surface plots in Figure 16 show signs that sample 4 has a lower surface roughness than sample 5. While these 3D surface plots have differing thickness scale values, the uniformity and magnitude of thickness differences distinctly show that sample 4 is of greater thickness than sa mple 5, even though both films were fabri cated with the same sum total passes. Quantitative confirmation of this thickness variation will be put forth later in this discussion. Distances between sprayed planes in sample 4 and sample 5 play an important role in shaping film topography. Figure 17 illustrates the variations in thickness along single lengthwise lines of these samples. From their thickness profiles, it can be observed that sample 4 has a smoother topography than sample 5. Sample 4 underwent two full turns of the handcrank shaft to create a distance of approximately 4 mm between planes in the y direction. Sample 5 was fabricated using a distance of 6 mm between planes via three turns of the handcrank shaft. Sample 4 shows a significantly lower standard deviation establishing that a distance of 4 mm between planes is more conducive than 6 mm to creating more even sprayed films. Depending upon design requirements, however, a 6 mm distance between planes could prove sufficient since the dif ferences in surface roughness for these treatments are not significantly dissimilar. Further evidence to this effect will be presented later in this work.
30 Figure 16. From left to right: single line optical microscope image, fulllength film thickness map along a single line and 3D surface plot of (a) sample 4 and (b) sample 5. Figure 1 7. Full length film thickness profiles and adjusted film th ickness profiles of (a) sample 4 and (b) sample 5. 0 4000 8000 12000 16000 20000 60 80 100 120 140 160 180 200 220 240 Thickness (nm)Length Along Sample (mm) 0 4000 8000 12000 16000 20000 40 60 80 100 120 140 160 180 200 220 240 Thickness (nm)Length Along Sample (mm) (a ) (b ) (a ) (b ) 0 5 10 1 5 20 25 0 5 10 15 20 25
31 Figure 1 8 displays a th ickness profile comparison plot of sample 4 and sample 5. This profile confirm s observations from the samples 3D surface plots that samples 4 and 5 do have different overall thicknesses despite being composed of the same total spray passes. The cause of this phenomenon can be attributed to the dissimilar multi pass parameters used to fabricate the samples. Sample 4 was sprayed at smaller distances between planes in the y dimension but at a greater quantity of planes in the x dimension compared to sample 5. Each plane sprayed in sample 4 was subsequently sprayed with fewer direct passe s of the nozzle during fabrication which resulted in a thicker, more uniform film. A thicker and more uniform film resulted because the linear fall off of each singular plane of spray increased film thickness at areas not directly sprayed by the nozzle in small, more consistent quantities. Statistical figures in Table 6 quantitatively confirm these thickness and surface roughness observations Sample 4 shows maxi mum and minimum film thickness es that are much closer in range than those of sample 5. Sample 4 also has a higher mean thickness, further verifying visual observations. Figure 18 Comparison of full length spray film thickness profiles of samples 4 and 5. 0 4000 8000 12000 16000 20000 40 60 80 100 120 140 160 180 200 220 240 Thickness (nm)Length Along Sample (mm) Sample 4 Sample 5 0 5 10 15 20 25
32 Table 6. Statistical analysis of thickness profiles for samples 4 and 5. All thickness data are in units of nm with an error of 5 nm. Name Maximum Minimum Max Min Mean Standard Deviation Sample 4 189 59 129 123 21 Sample 5 225 36 189 113 27 Sample 6 and sample 7 are the thickes t films to be discussed in this work with total passes of 480 and 540 nm, respectively Processing of these films occurred according to the procedures laid out in Table 7. Both were sprayed directly from 6 different planes in the y dimension with distances of 6 mm between each plane but sample 7 was spray coated with an additional 10 pass es per plane in the x dimension. Table 7. Spray fabrication protocols for samples 6 and 7. Distances are given in units of mm. Images in Figure 19 show thickness differential between the two samples resulting from the additional passes on sample 7. Visual analysis of the 3D surface plots in Figure 19 suggests that sample topographies are fairly consistent with each other. Quantitative data to this ef fect will be presented later in this work Name Total Passes Distance Between Planes Number of Planes Passes per Plane Total Nozzle Distance Traveled Sample 6 480 6 6 80 36 Sample 7 540 6 6 90 36
33 Figure 19. From left to right: single line optical microscope image, single line full length film thickness map, and 3D surface plot of (a) sample 6 and (b) sample 7 Film thickness profiles in Figu re 20 also display only modest contrast in the topographies of samples 6 and 7. Adjusted film thickness profiles demonstrate the viabil ity of spray deposition at 6 mm distances between planes to create fairly even sprayed active layers for P SCs. In a similar fashion to samples 2 and 3, samples 6 and 7 required adjustments to account for an additional plane that is lacking adequate spray passes This phenomenon would not occur in roll to roll spray processing since a continuous pattern of planes would be sprayed across the entire substrate. (a ) (b )
34 Figure 20 Full length film thickness profiles and adjusted film th ickness profiles of (a) sample 6 and (b) sample 7 Adjusted profiles display thicknesses along the 80% of samples Overlaid full length and adjusted thickness profiles of samples 6 and 7 in Figure 21 give an exceptional view of the similarities in film range and surface roughness while also s howing the difference s in thickness between the two samples. Statistical data in Table 8 provides additional confirmation of previous visual observations of sample topographies and thicknesses. Standard deviations of adjusted data for samples 6 and 7 show a relatively low disparity in surface topographies of the films. 0 4000 8000 12000 16000 20000 60 80 100 120 140 160 180 200 220 240 260 Thickness (nm)Length Along Sample (mm) 4000 8000 12000 16000 20000 60 80 100 120 140 160 180 200 220 240 260 Thickness (nm)Length Along Sample (mm) 0 4000 8000 12000 16000 20000 80 100 120 140 160 180 200 220 240 260 Thickness (nm)Length Along Sample (mm) 4000 8000 12000 16000 20000 80 100 120 140 160 180 200 220 240 260 Thickness (nm)Length Along Sample (mm) (a ) (b ) 0 5 10 15 20 25 0 5 10 15 20 25 5 10 15 20 25 5 10 15 20 25
35 Figure 21. Comparison of full length spray film thickness profiles and adjusted film thickness profiles of samples 6 and 7. Table 8. Statistical analysis of thickness profiles for samples 6 and 7. Thickness data are in units of nm. Name Maximum Minimum Max Min Mean Standard Deviation Sample 6 234 74 160 158 23 Sample 6 Adjusted 234 89 145 162 22 Sample 7 257 90 166 180 27 Sample 7 Adjusted 257 103 154 188 23 A spin coated reference sample was fabricated wi th a thickness of 150200 nm to compare film morphology and thickness uniformity with spray deposited films This spin coated sample displays the smooth topography and homogenous morphology which are characteristic of this deposition method. A visual comparison of film morphologies between the reference sample and a sprayed sample can be seen in Figure 22 Even though spin coating creates smoother and more uniform films, spray deposited films could display relatively predictable surface roughness properties with the optimization of spray parameters 0 4000 8000 12000 16000 20000 60 80 100 120 140 160 180 200 220 240 260 Thickness (nm)Length Along Sample (mm) Sample 6 Samle 7 4000 8000 12000 16000 20000 60 80 100 120 140 160 180 200 220 240 260 Thickness (nm)Length Along Sample (mm) Sample 6 Samle 7 0 5 10 15 20 25 5 10 15 20 25
36 Figure 2 2. Optical microscope images of P3HT:PCBM films of (a) a spin coated sample and (b) a spray coated sample. Scale bar represents 0.50 mm. 3.3. Device Characterization R esults PSC devices were fabricated using films deposited in the manner outlined in Section 2: E xperimental Setup and Methods. Fabrication parameters and functionality of PSC devices are displayed in Table 9. Devices with mean P3H T :PCBM thicknesses greater than 136 nm or 450 total passes, were operational. Table 9. Device fabrication parameters Mean P3HT:PCBM film thicknesses are based on mean thickness statistics of concurrently sprayed glass substrates. Thickness data are in units of nm. Name P3HT:PCBM Deposition Method Total Spray Passes Mean P3HT:PCBM Film Thickness ( 5 nm) Estimated Aluminum Cathode Thickness Working Device? Device 2 Spray 450 139 175 Yes Device 3 Spray 450 136 175 Yes Device 4 Spray 360 123 130 No Device 5 Spray 360 113 130 No Device 6 Spray 480 158 175 Yes Device 7 Spray 540 180 110 Yes (a ) (b
37 Table 10 reports PSC device characteristics which can be viewed in the current density vs. voltage curves plotted in Figure 23 Device 3 showed a slight increase in PCE from device 2 which may be a result of device 3 having received a larger area of direct spray Sample 2 had a slightly less rough surface than device 3 when measured by optical transmission microscopy. T he site of the most uniform portions of spray may not have been located over the device area. Device 6 had a PCE of 0.18%, nearly double that of device 2. The full length thickness profile of sample 6 showed the second lowest standard deviation of all devices in this work, second only to sample 4 indicating more moderate t opographical variation. The higher PCE of device 6 in comparison to the other devices can be attributed in part to this lower degree of surface roughness and higher degree of film thickness uniformity. Device 7 was sprayed with 540 total passes, the greatest overall number of passes in this work, but did not display the highest PCE. It was expected that device 7 would post the highest PCE due to a greater overall film thickness but it is likely that this was co unteracted by the extremely thin aluminum cathode layer deposited on this device Nevertheless device 7 did possess a FF of 0.285, the highest of all tested devices in this work. Table 10. Properties of characterized PSC devices. Only operational devices are displayed. Name P3HT:PCBM Deposition Method Total Spray Passes Jsc (mA/cm2) Voc (V) Pmax (mW/cm2) FF PCE (%) Device 2 Spray 450 0.854 0.530 0.090 0.201 0.09 Device 3 Spray 450 1.05 0.495 0.110 0.212 0.11 Device 6 Spray 480 1.46 0.540 0.178 0.225 0.18 Device 7 Spray 540 0.612 0.635 0.111 0.285 0.11
38 Figure 23. Current density vs. voltage charts for (a) device 2, (b) device 3, (c) device 6, and (d) device 7. C urrent density versus voltage plots of all working devices are shown in Figure 24. Devices 2, 3, and 6 display overlaying curves, while device 7 displays a form that is more characteristic of high er efficiency solar cells similar to the theoretical curre nt density versus voltage plot in Figure 8. While these devices do not present with high efficiencies, they do prove the viability of fabricating PSC devices with multipass multiplane sprayed P3HT:PCBM thin films. 2.5 1.5 0.5 0.5 1.5 2.5 0.2 0.3 0.8V (V) J (mA/cm2) 2.5 1.5 0.5 0.5 1.5 2.5 0.25 0.25 0.75V (V) J (mA/cm2) 2.5 1.5 0.5 0.5 1.5 2.5 0.25 0.25 0.75V (V) J (mA/cm2) 2.5 1.5 0.5 0.5 1.5 2.5 0.25 0.25 0.75V (V) J (mA/cm2) (a ) (b ) (d ) (c )
39 Figure 24. Current density vs. voltage plot of sprayed PSC devices 2, 3, 6, and 7. 4. Conclusion Polymer solar cells fabricated with P3HT:PCBM active layer s show promise as a future alternative energy technology. The ability to use spray deposition in large scale solar cell manufacturing allows for low cost fabrication of devices on flexible substrates. Using a custom ultrasonic spray deposition system films were created with varying multipass spray parameters on glass only and ITO + PEDOT coated glass substrates co ncurrently. Film thicknesses and uniformity were characterized via optical transmittance thickness mapping of glass only samples Characterization of PSC device power conversion efficiencies was performed by way of current density measurements. Looking forward, knowledge of the single plane thickness profile of P3HT:PCBM films fabricated with this spray system could aid in design of spr ay protocols with more accuracy and 2.5 2 1.5 1 0.5 0 0.5 1 1.5 2 2.5 0.2 0.3 0.8 V (V) J (mA/cm2) Device 2 Device 3 Device 6 Device 7
40 precision. Results of this protocol show ed that 100 passes of spray along a single plane form a thickness profile with a FWHM of approximately 24 mm Sprayed film thickness data put forth statistical evidence supporting the viability of multipass spray at planes with 4 mm and 6 mm distances betw een them. Device characterization figures display ed the effects of P3HT:PCBM spray parameters on device PCEs. I ncreasing the thickness of P3HT:PCBM films in devices using a greater number of total passes and a distance between planes t hat is no greater t han 6 mm should enhance film t hickness control and reduce surface roughness. Using the findings of this work, improvements to multi pass spray deposition procedures can be made in a move toward low cost roll to roll PSC fabrication.
41 REFERENCES [ 1] Mallon, K. Renewable Energy P olicy and P olitics: A Handbook for D ecision making, Earthscan, Sterling, Virginia, 2006.  Sherwood, L. U.S. Solar Market Trends 2009, Interstate Renewable Energy Council, 2010. Available at http://irecusa.org/wp content/uploads/2010/07/IREC Solar Market Trends Report 2010_727 10_web1.pdf  Sherwood, L. U.S. Solar Market Trends 2008, Interstate Renewable Energy Council, 2009. Available at http://irecusa.org/fileadmin/user_upload/NationalOutreachDocs/SolarTrendsReports/IRE C_Solar_Market_Trends_Report_2008.pdf  Sherwood, L. U.S. Solar Market Trends 2007, Interstate Renewable Energy Council, 2008. Available at http://irecusa.org/fileadmin/user_upload/NationalOutreachDocs/IREC_Solar_Market_Tre nds_Revision_11_19_081.pdf. Retrieved 2010 07 24  Pagliaro, M.; Ciriminna, R.; Palmisano, G. Flexible solar cells, ChemSusChem 2008, 1, 880890.  Ge, W. An overview on P3HT:PCBM, the most efficient organic solar cell material so far, Solid State Physics II Lecture, University of Tennessee, Knoxville 2009.  Xue, J. Perspectives on o rganic photovoltaics Poly. Rev., 2010, 50, 411419.  NREL Certificate Konarka, 8.3 % under AM1.5 spectrum, November 29, 2010.
42  Hoppe, H.; Sariciftci N. S. Organic solar cells: An overview, J. Mater. Res. 2004, 19, 19241945.  Spangga a rd H.; Krebs, F. A brief history of the development of organic and polymeric photovoltaics, Sol. Energy Mater. Sol. Cells 2004, 83, 125146.  Krebs, H. U. Polymer thin films, Thin Polymer Films Lecture, Georg August Universitt Gttingen, http://www.pcgg.de/people_detail.php?id=9 (March 27, 2011).  Ling, M. M.; Bao, Z. Thin film deposition, patterning, and printing in organic thin film transistors, Chem. Mater., 2004, 16, 48244840.  Piqu, A.; McGill, R. A.; Chrisey, D. B. A new way to deposit organic thin films, The Industrial Physicist American Institute of Physics 2000, 6, 2023.  Krebs, F. C.; Tromholt, T.; Jrgensen, M. Upscaling of polymer solar cell fabrication using full roll to roll processing, Nanoscale 2010, 2, 873886.  Gregg, A.; York, L.; Strnad, M. R oll to r oll m anufacturing of flexible displays, Flexible Flat Panel Displays G.P. Crawford ed., Wiley, Chichester, United Kingdom, 2005, 410445.  Watts M. P. C. Advances in roll to roll processing, Impattern Solutions 2007, http://www.impattern.com/Download/RollToRollProcessing.pdf (April 1, 2011).  Abdellah, A.; Fabel, B.; Lugli P.; Scarpa, G. Spray deposition of organic semiconducting thinfilms: Towards the fabrication of arbitrary shaped organic electronic devices, Org. Electron., 2010, 11, 10311038.
43  Vak D.; Kim, S.; Jo, J.; Oh, S.; Na, S.; Kim, J.; Kim, D. Fabrication of organic bulk heterojunction solar cells by a spray deposition method for low cost power generation, Appl. Phys. Lett., 2007, 91, 8110281103.  Girotto, C.; Moia, D.; Rand, B. P .; Heremans, P. High performance organic solar cells with spray coated hole transport and active layers Adv. Funct. Mater. 2010, 19.  Green, R.; Morfa, A.; Ferguson, A. J.; Kopidakis, N.; Rumbles, G.; Shaheen, S. E. Performance of bulk heterojunction photovoltaic devices prepared by airbrush spray deposition, Appl. Phys. Lett., 2008, 92.  Ultrasonic Atomizer Nozzle for Spraying Liquids 20kHz 130kHz, Sonaer Ultrasonics, 2008, http://www.sonozap.com/Ultrasonic_Atomizer.html (Mar ch 28, 2011).  Steirer, K. X.; Reese, M. O.; Rupert, B. L.; Kopidakis, N., Olson, D. C.; Collins, R. T.; Ginley, D. S. Ultrasonic spray deposition for production of organic solar cells Sol. Energy Mater. Sol. Cells, 2009, 93, 447453.  Steirer K. X.; Berry, J. J.; Reese, M. O.; van Hest, M.; Miedaner, A.; Liberatore, M. W.; Collins, R. T.; Ginley, D. S. Ultrasonically sprayed and inkjet printed thin film electrodes for organic solar cells, Thin Solid Films, 2009, 517, 27812786.  Ultr asonic atomizing nozzle and controller, Spraying Systems Co., http://www. service.spray.com/Literature_PDFs/b660_utrasonic.pdf ( 2009).  Hammond, W. Design and Fabrication of Organic Semiconductor Photodiodes PhD Thesis, University of Florida, 2011.
44  Hoth, C. N.; Steim, R.; Schilinsky, P.; Choulis, S. A.; Tedde, S. F.; Hayden, O.; Brabec, C. J. Topographical and morphological aspects of spray coated organic photovoltaics Org. Electron., 2009, 10, 587593.  Stukova I.; Harris, N. Design of Organic Solar Cells, Old Dominion Univ., Norfolk, Virginia, www.eng.odu.edu/ece/ece486_487/Proposal_Design%2520of%2520Organic%2520Solar %2520Cells.doc.  Peumans, P.; Uchida, S.; Forrest, S. Efficient bulk heterojunction photovoltaic cells using small molecular weight organic thin films, Nature 2003, 425, 158162.  Paine, D. C.; Yeom, H. Y.; Yaglioglou, B. Transparent conducting oxide materials and t echnology, Flexible Flat Panel Displays G.P Crawford ed., Wiley, Chichester, United Kingdom, 2005, 7998.  Kim, S Y.; Lee, J. L.; Kim, K. B.; Tak, Y. H. Effect of ultraviolet ozone treatment of indium tin oxide on electrical properties of organic light emitting diodes, J. Appl. Phys., 2004, 95, 25602563.  Nardes, A. On the conductivity of PEDOT:PSS thin films, Electronic Sys. Eng. Dep. of the Polytech. School of the Univ. of Sao Paulo, Sao Paulo, Brazil, 2007.  Gunes, S.; Neugebauer, H.; Saricifti, N. S. Conjugated polymers based organic solar cells, Chem. Rev. 2007, 107, 13241338.