BARIUM OXIDE AS AN INTERMEDIATE LAYER FOR POLYMER TANDEM SOLAR CELL. A Thesis. Presented to. The Graduate Faculty of The University of Akron

BARIUM OXIDE AS AN INTERMEDIATE LAYER FOR POLYMER TANDEM SOLAR CELL A Thesis Presented to The Graduate Faculty of The University of Akron In Partial Fulfillment of the Requirements for the Degree Master of Science Zhehui Li May, 2013

BARIUM OXIDE AS AN INTERMEDIATE LAYER FOR POLYMER TANDEM SOLAR CELL Zhehui Li Thesis Approved Accepted Advisor Dr. Xiong Gong Department Chair Dr. Robert Weiss Committee Member Dr. Alamgir Karim Dean of the College Dr. Stephen Z. D. Cheng Committee Member Dr. Yu Zhu Dean of the Graduate School Dr. George R. Newkome Date ii

ACKNOWLEDGEMENTS I would like to thank first and foremost my research advisor Dr. Xiong Gong for his patience, encouragement, and guidance throughout the course of this research. Also, my gratefulness is given to all the group members: Mr. Tingbin Yang, Mr. Hangxing Wang, Ms. Xilan Liu, Mr. He Ren, Mr. Wei Zhang, Mr. Chao Yi, Mr. Bohao Li, Mr. Kai Wang, Ms. Chang Liu for their warm caring about my life and helpful comments and suggestions on research. Finally, I would like to express my deepest gratitude to my parents, Mr. Jun Li and Ms. Xiaoyan Xu for their love and support. iii

ABSTRACT Polymer solar cells (PSCs), a member of organic solar cell family, have attracted increasing research interest. PSCs possess significant advantages over their inorganic solar cell counter parts: mechanical flexibility, light weight, low expense, and the potential to achieve roll-to-roll large-scale production. Tandem solar cells, in which two solar cells are linked to take more use of the solar energy, were fabricated with each solution processed layer using Bulk Heterojunction (BHJ) materials comprising semiconducting polymers and fullerene derivatives. For years, tremendous efforts have been put in seeking for an efficient intermediate layer to successfully connect two sub-cells together in the tandem structure. Even though many kinds of intermediate layers such as ZnO/MoO 3 etc. have been explored, most of them suffer from the low conductivity or complicated manipulation disadvantages. Barium oxide (BaO) is both high conductive and wide bandgap n-type semiconductor. We successfully fabricated the polymer tandem solar cell using all thermal vacuum deposition fashion with BaO/Ag/MoO 3 as an intermediate layer. V OC of the tandem structure is the sum of the component cells demonstrating our proposed intermediated layer can efficiently connect sub-cells with no potential/energy loss. iv

TABLE OF CONTENTS Page CHAPTER I. INTRODUCTION 1 1.1. General Background 1 1.2. Polymer Solar Cells 4 1.2.1. Working Principle 4 1.2.2. Device Geometry 8 1.2.3. High Performance: Three Essential Parameters 10 1.3. Polymer Tandem Solar Cells 12 1.3.1. General Background 13 1.3.2. Efficient Single Cells 15 1.3.3. Efficient Intermediate Layer 17 1.3.4. Tandem Polymer Solar Cell Characterization 21 1.3.5. Processing Issues of the Tandem Structure. 22 II.. EXPERIMENT 23 2.1. Materials Preparation 23 2.2. Device Fabrication Procedures 24 2.3. Calibration and Characterization 26 v

2.4. UV-Vis Absorption Spectrum 26 2.5. Atomic Force Microscopy (AFM) Observation 26 2.6. Sol-gel ZnO Nanoparticles Preparation 26 III. RESULTS 28 3.1. Energy Levels 28 3.2. Performance Investigation 29 3.2.1. Current Density-Voltage (J-V) Characteristics 29 3.2.2. UV-Vis Absorption 33 3.2.3.Atomic Force Micrometer (AFM) Images Observation 35 3.3. Comparison of the Intermediate Layers 36 IV. DISCUSSION, CONCLUSIONS AND OUTLOOK 40 REFERENCES 44 v

LIST OF FIGURES Figure Page 1.1 Number of scientific publications contributing to the subject of polymer solar cell(s) 3 1.2 (a) The unit cell of silicon; (b) Simplified energy band diagram for a semiconductor 5 1.3 (a) Bulk Heterojunction (BHJ) Structure of the active layer in polymer solar cell, (b) working principle of polymer solar cells 7 1.4 Photoinduced process in the D-A system. (a) photoinduced charge transfer in a forward direction;(b) exciton recombination happens on a time scale of nm; (c) charge transfer in a back direction 8 1.5 (a) The conventional device structure; (b) Bulk heterjunction configuration in organic solar cells 9 1.6 The organic solar cells with (a) conventional geometry and (b) inverted geometry 9 1.7 Current-Voltage Characteristics of a polymer solar cell under illumination (red line) and in the dark (black line) 10 1.8 (a) Typical tandem solar cell device geometry and (b) simplified procedure of the stacking process of two sub-cells 14 1.9 The current-voltage characteristics of two sub-cells under illumination. The front cell delivered more photocurrent than the bottom cell 16 1.10 Basic principle of an organic tandem solar cell using an intermediate layer. The arrows indicate the hole currents and the electron currents. ETL denotes the electron transport layer and HTL indicates the hole transport layer. 18 1.11 Simplified energy level diagram of the metal and n-type semiconductor (a) before contact and (b) band bending in the Ohmic contact 18 vi

1.12 Schematic energy level diagram at open circuit of a double heterojunction solar cell with highly doped layers as recombination contact 19 1.13 Dark Current Density verse Voltage (J-V) characteristics of a tandem cell before and after light illumination. 20 2.1 Molecular structures of PCPDTBT, P3HT and PCBM respectively 24 2.2 Polymer tandem solar cell geometries with (a) PCPDTBT:PCBM as an upper layer and (b) P3HT:PCBM as an upper layer 25 3.1 Energy levels of the single cell composed of bulk heterojunction polymer blends (a) P3HT: PCBM and (b) PCPDTBT: PCBM. 28 3.2 The energy levels of the tandem solar cells composing the upper polymer layer of (a) P3HT : PCBM and (b) PCPDTBT:PCBM. 29 3.3 The current density-voltage (J-V) characteristics of single reference cells using identical P3HT:PCBM polymer systems and tandem cell. (a) J-V curves under illumination and (b) in dark. 31 3.4 The current density-voltage (J-V) characteristics of single reference cells using P3HT:PCBM and PCPDTBT:PCBM and the tandem cell. J-V curves (a)under illumination and (b) in dark. 33 3.5 UV-Vis absorption spectra of a PCPDTBT:PCBM bulk heterojunction composite film, a P3HT:PCBM bulk heterojunction composite film, and a bilayer of the two as relevant to the tandem device structure. a.u. optical density. 34 3.6 AFM images of (a) MoO 3 surface morphology of BaO/Ag/MoO 3 intermediate layer and (b) PEDOT:PSS on ITO coated glass substrate. The islands observed are due to surface roughness. Note that the islands distribution is more intensive for (b), indicating the roughness is higher for PEDOT:PSS. 35 3.7 AFM height profiles of ZnO nanoparticles under the condition of (a) fast annealing and (b) slow annealing 36 3.8 The intermediate layer composed of ZnO/MoO 3. (a) Current Density-Voltage (J-V) Characteristics and (b) Sol-gel preparation of ZnO nanoparticles 37 3.9 The electronic performance of BaO/Ag/MoO 3 intermediate layer. (a) Diode property of BaO/MoO 3 P-N junction; (b) Conductive property of BaO/Ag/MoO 3 intermediate layer 38 viii

3.10 AFM images of two types of intermediate layers. (a) ZnO/MoO 3 and (b) BaO/Ag/MoO 3 on ITO coated glass substrate. The islands observed are due to surface roughness 39 ix

LIST OF TABLES Table Page 1. Photovoltaic performance of the single reference cell P3HT : PCBM and the corresponding tandem cell. 31 2. Photovoltaic performance of the single reference cells P3HT : PCBM/PCPDTBT : PCBM and the corresponding tandem cell. 33 x

CHAPTER I INTRODUCTION 1.1. General Background Nowadays, environmental pollution and resource depletion are the problems that need to be solved urgently. Due to heavily environmental pollution brought by the widely used traditional energy sources such as oil and gasoline, people are seeking for an environmentally-friendly alternative energy source. Harvesting nature energy to generate power is regarded as one of the promising methods. In this spirit, solar energy is one of the best available alternatives, for its embedded nature of both clean and unlimited. The photovoltaic effect in Silicon (Si) was first proposed in 1954 in Bell Laboratory and the power conversion efficiency (PCE) was reported reaching 6%. 1 Since then, the inorganic-based solar cells including but not limited to Si such as GaAs, CdTe, CIFGS have been intensively explored. In the past decade, the technology of PV is mushrooming at an annual rate of 48% and gradually commercialized. 2 Despite the fact that the inorganic photovoltaic has been booming quite fast, it only takes account for less than 0.1% of the energy demand world widely. Unfortunately, there are many embedded disadvantages of inorganic PV responsible for its bottleneck of development. On the one hand, the Silicon processing consumes large quantities of acid and much poisonous waste is disposed into the environment; 1

On the other hand, the installation of Silicon based photovoltaic (PV) expense is as high as 1500usd/m 2, which inevitably hinders its wide application. To circumvent those issues, researchers are looking for a better solar cell in regards of pollution-free and low-cost. No doubt inspired by discovering the ultra-fast photo induced charge transfer in 1992 3, collaborative efforts by interdisciplinary researchers in the fields of synthetic chemistry and optical physics have been put in the study of organic solar cells. Based on organic materials, this new member of PV is regarded as a promising alternative because it is low processing expense, light weight, and could be fabricated in a continuous fashion. In this way, the organic solar cell can be implemented on flexible substrate carrying the possibility of achieving roll-to-roll printing technique. 4 Polymer photovoltaic is not a precise definition but typically considered as a generation of OPV, and it means applying semiconducting conjugated polymers 5 as active materials within the thin film PVs. The first highly conductive polymer was reported in 1977. 6 The highly chemically doped polyacetylene can form a new class of conducting polymers, and the electrical conductivity property can be systematically and continuous varied over a large magnitude. Another notable event in the polymer solar revolution occurs at 2000 when Heeger, MacDiarmid and Shirakawa were nominated Nobel Laureates in recognition of their outstanding contribution in discovery and development of conducting polymers. Figure1.1 displays an overview of the current development tendency of polymer solar cells. 7 2

Figure 1.1 Number of scientific publications contributing to the subject of polymer solar cell(s). Search done through ISI, Web of Science, 2007 The principal working mechanism of polymer solar cells is: First, the conjugated polymer with localized π electrons can absorb sunlight and forms a coulombically bound pair of electron-hole named exciton; Second, this pair of electron-hole is dissociated at bulk hetrojucntion interface into free charge carriers and those carriers transport through active layer and finally reach the electrodes. The detailed working principle and the device geometry will be disclosed later. To date, thanks to the tremendous efforts by interdisciplinary researchers, the power conversion efficiency of a single polymer solar cell has been pushed to a value which is competent with their inorganic counterparts. Furthermore, the theoretic value of OPV is around 20% and pushing the single cell to 10% has become a reality through thoughtful design of electron-donor polymer and careful device fabrication. Besides using better materials to fabricate the solar cell, another logical thinking is to modifying the device structure. In this spirit, the newly created tandem solar cell was proposed. 3

The tandem solar cell where two sub-cells are connected in series through an intermediate layer is one of the most commonly employed tandem structure. The first reported two terminal organic tandem solar cell was proposed by Hiramoto et al. 8 Previously, people are focusing on the small molecular to make sub-cells, because the applied dry coating fashion is an easy way to stack different sub-cells together. Nowadays, thanks to the development of polymer chemistry, tremendous types of polymers with good electrical conductivity have been successfully synthesized. This breakthrough has overcome the choice limitation of available material and more work was emphasized on the polymer based tandem solar cells afterwards. It was Kawano et al. 9 that demonstrated the first polymer based tandem solar cells. The intermediate layer in the tandem solar cells has been very attractive to numerous investigators since it is key point to connect sub-cells successfully. Several types of intermediate layers in either evaporation or solution processed fashion were explored to connect two sub-cells, and its modification is never overstated. I would disclose the deep-in knowledge about the intermediate layer later. 1.2. Polymer Solar Cells 1.2.1. Working Principle Polymer solar cells (PSCs) is an important member of the OPV family, distinguished by utilizing the π-conjugated polymer as an active component. To better understand the working principle in the PSC, we can simply compare it with the inorganic solar cell, Si based photovoltaic, to be specific. Crystal Si possesses a diamond lattice structure, with each silicon covalent bonded to another four silicon 4

atoms. The pure crystal Si is commonly regarded as a semiconductor material, and the Fermi level is located at the middle of the valance band and conduction band. Figure 1.2 shows the one silicon lattice and the simplified band structure of pure silicon. 10 Figure 1.2 (a) The unit cell of silicon; (b) Simplified energy band diagram for a semiconductor. 10 As we know, electric conductivity is proportional to the concentration of mobile charge carriers and therefore the electric conductivity for pure silicon is comparatively low. To overcome this issue, one commonly used method is to add dopants, also called impurities, into the pure Si. The electric conductivity of the semiconductor is considerable increased after being doped. The element chosen to be a dopant usually possesses or lacks an extra electron compared to silicon. Undoped silicon carries the equal number of electrons and holes and it is called intrinsic silicon, and dopant will generate an excess of either electron or hole. Hence, there are two types of doped silicon: n-type silicon, and n refers to negatively charged carriers (electron); p-type silicon, and p refers to positively charged carriers (hole). When the n and p-type silicon is connected together, the p-n junction is formed. This p-n junction is a key point and a platform for energy conversion and exciton generation in the inorganic PV. Same case with organic solar cells, the importance of donor-acceptor interface can never be overstated. Similarly, the organic solar cells 5

also possess a donor-acceptor (D-A) interface like p-n junction in inorganic photovoltaic. The state-of-art active layer structure is called Bulk Heterojunction (BHJ) structure which was first reported in 1995 11. This active layer is commonly consisted of two materials, namely: an electron donor material and an electron acceptor material. Usually, the conjugated polymer serves as an electron donor and a fullerene derivative as an electron acceptor. Blending the donor materials with the acceptor materials together prepared by dissolving them in the common solvent and spin cast to form a BHJ structure is a good way to enhance the interfacial area and to break photoexcited excitons into free charge carriers. Poly (3-hexylthiophene) (P3HT) is one of the commercially available donor materials and 1-(3-methoxycarbonyl) propyl-1-phenyl[6,6]c61 (PCBM) is acceptor material. When shining the light to the active layer, an electron of the donor material will absorb a photon and be excited from the Highest Occupied Molecular Orbital (HOMO) level to the Lowest Unoccupied Molecular Orbital (LUMO) level, leaving a hole in the HOMO level. Because of the small dielectric constant of organic materials, this pair of electron and hole (called exciton) is tightly coulombically bound. At the interface of donor and acceptor, driven by the difference of electron affinity, this pair of electron and hole is dissociated. Afterwards, the free electron and hole transport though the bulk and reach the respective electrodes. To be concluded, the process of conversion of light into electricity by PSC can be described by the following steps 12 : 1. Absorption of photon leads to the formation of an exciton; 2. Excion is dissociated at the D-A interface; 3. Free charge carriers transport through the bulk volume; 4. Free 6

charge carriers accumulate at electrodes, respectively. The bulk heterojunction (BHJ) configuration of the active layer and working principle of OPV are schematically shown in Figure 1.3 Figure 1.3 (a) Bulk Heterojunction (BHJ) Structure of the active layer in polymer solar cell, (b) working principle of polymer solar cells. Copyright 2010 Elsevier Ltd. There are two critical steps determining how efficiently the device can convert solar energy to electrical energy, one is the efficient excition dissociation and the other is the efficient charge transport through bulk active layer. The ultrafast photophysical studies demonstrate that the photoinduced charge transfer in the D-A blends happens on a time scale of 50fs. 13 For efficient photovoltaic devices, the created charges need to be transported to the appropriate electrodes within the exciton life time. Figure 1.4 schematically shows the comparison between photoinduced charge transfer and its competing processes like photoluminescence and back transfer which usually happen on the time scale larger than ns. 7 7

Figure 1.4 Photoinduced process in the D-A system. (a) photoinduced charge transfer in a forward direction;(b) exciton recombination happens on a time scale of nm; (c) charge transfer in a back direction. Copyright Springer-Verlag Berlin Heidelberg. 1.2.2. Device Geometry The bilayer structure of solar cell was first proposed by C. W. Tang in 1986. 14 In the first reported bilayer structure, two layers of small molecules are layer-by-layer vacuum deposited in the vertical direction on the indium tin oxide( ITO) coated glass substrate. The device is finalized by the thermal deposition of back electrode, namely Ag. In such a device, only the exciton created within the distance of 10-20nm from the interface can be efficiently dissociated, and the thicknesses of two active layers are heavily limited. To circumvent those issues, the Bulk Heterojunction (BHJ) structure was proposed. 11 As mentioned before, Bulk Heterojunction is a blend of the donor and acceptor components in a bulk volume. The advantages of BHJ over bilayer structure is twofold: First, in this nano-scale interpenetrating network, each donor-acceptor (D-A) interface is within 10-20nm length scale to guarantee efficient exciton dissociation; Second, BHJ can tremendously increase orders of magnitude of the interfacial area favoring more exciton generation. The idea of the conventional 8

device structure and BHJ configuration of active layer are schematically displayed in Figure 1.5. Figure 1.5 (a) the conventional device structure; (b) Bulk Heterojunction configuration in polymer active layer. Copyright 2001, Kirchberg in Tirol, Österreich In general, there are two geometries existing for a single cell in terms of the functions of electrodes. The conventional structure and the inverted structures are schematically shown in the Figure 1.6. In the normal geometry, the device is built up by stacking the buffer layers and the active layer in sequence on top of ITO ( a high work-function metal ) and finally covered with a vacuum deposited layer of Al ( a typical low work-function metal ). This conventional structure suffers from the poor stability issue because of the easily oxidized Al. Also, Al is hard to achieve roll-to-roll large-scale printing mass production. To overcome those shortcomings, the alternative inverted device was proposed 15,16, where the two electrodes are in the opposite positions. Figure 1.6. The organic solar cells with (a) conventional geometry and (b) inverted geometry. Copyright 2006, American Institute of Physics 9

The two electrodes can be further modified by introducing buffer layers on the ITO side and the back metal side. The Hole Transportation Layer (HTL) and Electron Transportation Layer (ETL) are two kinds of buffer layers commonly employed to selectively transport charge carriers. HTL plays a role of selectively transporting holes and blocking electrons and ETL selectively transporting electrons and blocking holes. Poly (3,4-ethylene dioxythiophene): (polystyrene sulfonic acid) PEDOT:PSS 17 and MoO 18 3 are two types of HTL that widely used to improve the charge extraction in the anode side. 1.2.3. High Performance: Three Essential Parameters The current-voltage characteristics of a solar cell under illumination (red line) and in the dark (black line) are shown in Figure 1.7. 19 Figure 1.7 Current-voltage characteristics of a polymer solar cell under illumination (red line) and in the dark (black line). Copyright 2007, American Chemical Society There are three critical parameters for solar cell efficiency: Open Circuit Voltage (Voc), Short Circuit Current Density (Jsc), and Fill Factor (FF). The 10

photovoltaic power conversion efficiency of a solar cell is determined in the following formula. 19 Here, P in is the incident light power density standardized at 1000W/m 2, and I mpp and V mpp are the current and voltage at the maximum power point. Open Circuit Voltage: The open circuit voltage is the point where the current-voltage characteristics under illumination intersect with the vertical coordinates. As mentioned before, one of the two important steps towards efficient solar cells is that the created charges need to be efficiently transported to the electrodes. The driving force for this process is the gradient in the chemical potentials of electrons and holes built up in the donor-acceptor junction, namely built-in potential. This gradient is determined by the difference between the HOMO level of the donor and LUMO of the LUMO level of the acceptor. An agreement has been met that the open circuit voltage is given by this built-in potential. 20 However, Voc is not only related to energy levels of the used materials but also their interfaces, and therefore the Voc of the real device would vary case by case. The open-circuit voltage of a conjugated polymer PCBM solar cell is estimated in the following formula. 21 Voc = (1/e)(E Donor HOMO E PCBM LUMO) 0.3 V Short Circuit Current Density: The short circuit current (Isc) is the point where the current-voltage characteristics under illumination intersect with the horizontal coordinates, and Short Circuit Current Density ( Jsc) is Isc divided by 11

electrode area. It has been demonstrated that Jsc is correlated to the absorption of the active layer and exciton dissociation efficiency, which requires that the materials have a better matched absorption with the solar spectrum and high dielectric constant, respectively. In the ideal case, Isc is determined by the product of photoinduced charge carrier density and the charge carrier mobility within the organic semiconductor. I sc =neµe n; density of charge carriers e; elementary charge µ; charge carrier mobility E; electric field Fill Factor: Telling from the current-voltage characteristics in Figure 1.7, Fill Factor (FF) is determined by the ratio between the area of the yellow rectangle and the area of rectangle with grey border. That is to say, the ratio between V MPP *I MPP (or the maximum power) and V oc *I sc is called the fill factor. In the respect of device physics, charge carriers reaching the electrodes can determine the fill factor, when the built-in potential is lowered towards the open circuit voltage. Fill factor can also be calculated in the following formula: P V J FF = = V J V J MAX MPP MPP oc sc oc sc Fill factor is a very important parameter to achieve solar cell high performance, and it is considerably influenced by the series resistances and finite conductivity of the ITO covered substrate of solar cell. 22 12

1.3. Polymer Tandem Solar Cells 1.3.1. General Background In general, the tandem solar cells can be classified into three categories: 1. Small molecular tandem solar cell; 2. Hybrid tandem solar cell; 3. Fully-solution processed tandem solar cell. Since the discovery of ultrafast photoinduced charge transfer, researchers have been intensively explored how to push upwards the efficiency of solar cells. It was predicted in 2009 that it is possible the PCE can be over 10% analyzed from the theoretically standing point, however, the PCE was staying around 5% at that time. 23 Nowadays, the PCE for a single cell has been reported as high as 13%. One bottleneck of further increasing PCE generates from the inherent nature of a single cell: limited by the band-gap, a semiconductor can t make the best use of every energetic photon in the solar spectrum. People have known a while that the way to broaden the solar spectrum absorption is to make a tandem solar cell. In a tandem solar cell, two or more single cells are absorbing in a complementary wavelength range are stacked together. The most commonly employed device structure is a two terminal tandem cell where two sub-cells are connected in series by an interconnecting layer. Figure 1.8 shows the typical tandem solar cell and the simplified procedure of the stacking process. 13

Figure 1. 8. (a) Typical tandem solar cell device geometry and (b) simplified procedure of the stacking process of two sub-cells. Copyright 2009, Royal Society of Chemistry In 1990, the tandem solar cell was first proposed where the small molecules are employed as an active layer in the two sub-cells. 8 In this spirit, the construction of a tandem structure can be easily manipulated by dry-coating method. However, due to the limited choice of small molecular donor materials which carry considerable different absorption profiles, the hybrid tandem solar cells are further explored. 24 In the hybrid tandem solar cell, the bottom sub-cell is processed from polymers from solution process, while the top cell is still through the thermal deposition of small molecules. Driven by the prospect that it is possible to achieve roll-to-roll large scale production through printing technology, the fully solution processed tandem solar cell is further explored. Not until in 2006 did Kawano et al. 9 developed the first fully solution processed tandem cells using two identical polymer Bulk Heterojunction (BHJ) sub-cells. Generally, an ideal tandem structure would utilize a large band-gap cell as the front cell and the low band-gap cell as the bottom cell. It was in the same year, Hadipour et al. first fabricated a tandem solar cell employing a low band-gap polymer and a large band-gap polymer at the same time, and PCE reported 0.57%. 25 A milestone of the PCE improvement in the tandem solar cell occurs at the year of 14

2007 when Kim et al. demonstrated the efficiency of all solution processed polymer tandem solar cell can be over 6%. 26 The further calculation suggests that tandem solar cell with more than 15% power conversion efficiency is feasible. 27 However, due to several issue such as the still inefficient utilization of solar spectrum, large series resistance, and thermal energy loss, the highest efficiency reported so far is slight larger than 7%. 28 To circumvent those issues and achieve high performance, several criteria should be considered, for instance, sub-cells with minimum absorption overlap, an efficient intermediate layer, and compatible fabrication process. Many research efforts so far have been put into the new polymer design for active layer, and one big problem remaining is how to make a good intermediate layer that can successfully connect two sub-cells with the minimum energy loss. In addition, it is more attractive if the intermediate layer can be both efficient and cheap, so that the fabrication expense would be low down. 1.3.2. Efficient Single Cells In the case that two sub-cells are connected, the three critical parameters of the whole device, namely Open Circuit Voltage (V OC ), Short Current Density (J SC ), Fill Factor( FF) would inevitably differ from the single cells. It has been demonstrated that in the ideal case assuming no potential loss in the intermediate layer or device fabrication derivation from the standard procedure, the V OC of the tandem solar cell is the sum of the two sub-cells: V OC =V Front + V Bottom. On the other hand, the total generated photocurrent will be constant throughout the device. it is also believed that 15

J SC of the tandem cell is limited by the smallest J SC going through the component cell on the condition that the fill factor of the two sub-cells are the same. 29 When in a more realistic case where FF of the two component cells are not identical, J SC of the tandem solar cell is more dominated by the cell having higher FF. To be concluded, the general relationship between V OC and J SC of the tandem solar cell and the component sub-cells is displayed in the following formulas. J Tandem = J Bottom + J Top V Tandem = V Bottom + V Top Because of the narrow solar spectrum absorption of a single cell limits PCE, it is critical to employ two sub-cells with complementary solar spectrum absorption in a tandem solar cell. Due to the embedded nature of the intermediate layer and its non-conductive nature in a 2-terminal tandem cell, characterization of two sub-cells independently in a tandem structure is impossible. However, the simulation statistics demonstrates that the bottom cell (Figure 1.9) 29 generates much more photocurrent than the top cell under short circuit condition. Figure 1.9. The current-voltage characteristics of two sub-cells under illumination. The front cell delivered more photocurrent than the bottom cell. Copyright 2008, Elsevier 16

Besides materials selection for active layer, the thickness of the thin film need to be well tuned in a way that more solar photon can be absorbed in the bottom cell as much as possible to balance the J SC in both top and bottom cells. 1.3.3. Efficient Intermediate Layer The importance of employing an intermediate layer is never overstated. Fabrication of the sub-cells in series without a separation layer in between them will cause the formation of an inverse bulk heterojunction (BHJ) between the donor layer of the top cell and the acceptor of the bottom layer. Hence, the critical step in making a good tandem solar cell is to make an efficient intermediate layer. An inefficient intermediate layer brings about potential loss leading to the V OC of the tandem solar cell is not equal to the sum of the V OC of the component cells. There are three main requirements for such a recombination contact: First, it has to ensure that the electrons from the first sub cell and the holes from the second sub cell meet at the same energy level. Therefore a splitting of the electron and hole quasi-fermi levels has to be avoided. Second, the recombination contact has to be highly transparent to avoid absorption and reflection reducing the power conversion efficiency and disturbing the current matching in the tandem solar cell. Third, it should be compatible to future mass production processes. 30 Figure 1.10 shows the basic principle of an organic tandem solar cell using an intermediate layer. 30 17

Figure 1.10 Basic principle of an organic tandem solar cell using an intermediate layer. The arrows indicate the hole currents and the electron currents. ETL denotes the electron transport layer and HTL indicates the hole transport layer. ITO is the conductive transparent indium tin oxide bottom contact. Copyright 2010, American Institute of Physics In solid-state physics, the work function is determined by the minimum energy needed to remove an electron from a solid to a point outside the solid surface (or energy which is needed to move an electron from the Fermi level into vacuum). 31 When the work function of a metal is smaller than the fermi level of a semiconductor, then after these two materials contact, the electron will flow from the metal into the semiconductor leading to the metal positively charged (Figure 1. 11). Under this condition, there is no barrier for the charge transfer, and we call this contact Ohmic contact. 18

Figure 1.11. Simplified energy level diagram of the metal and n-type semiconductor (a) before contact and (b) band bending in the Ohmic contact. A possible approach to the desired intermediate layer is to make use of thick metal layers. If the layers are thick enough, a closed metal layer is formed which acts as an Ohmic contact. This approach has been introduced in polymer solar cells, because the mental can prevent the underlying layer from dissolving during the spin-casting of the second solar cell. 32,33 The disadvantage of this approach is the high absorbance and reflectance of the metal layer resulting in the losses and unbalanced absorption in the sub-cells. An alternative approach is to use the highly doped semiconducting layers (n-type and p-type) as an intermediate layer (Figure 1.12) 30. Figure 1.12 Schematic energy level diagram at open circuit of a double heterojunction solar cell with highly doped layers as recombination contact. Copyright 2010, American Institute of Physics From this energy level diagram, we can observe that the doped semiconductor layers form Ohmic contact with back and front cells. This ohmic contact can efficiently extract electrons and holes from the two sub-cells. Typical materials for n-type semiconductor are solution processed TiOX and ZnO2, and p-type semiconductors are poly (ethylenediox-ythiophene) doped with poly (styrenesulfonate) (PEDOT: PSS), MOO3, and WO3.34 However, these materials still need to be modified to serve as an efficient part of intermediate layer. One big 19

issue of the olution-processed ZnO in as an n-type intermediate layer lies in the inherent nature of its solution process that can t withstand the top polymer sub-cell. Besides efficiently extracting electrons and holes from the two component cells, the intermediate layer should also act as an efficient recombination zone for the charge carriers. A thin layer of mental material is a promising candidate to serve this purpose given the good conductivity and suitable working function. The rectification of the current-voltage characteristics in dark has also been studied recently. Rectification concept originally comes from rectifier device. A rectifier can convert alternating current (AC) to direct current (DC), and this process is known as rectification. In the polymer solar cell, rectification represents how many orders of magnitude of the current at the most positive bias voltage higher than the current at the most negative bias voltage for the current-voltage characteristics in the dark. To date, more and more work is focusing on how to improve the rectification of the current-voltage characteristics, but merely no breakthrough has been make in this area. Some reports have demonstrated that the UV illumination can help to increase the rectification of the current-voltage characteristics in dark (Figure 1.13). 35 20

Figure 1.13 Dark Current Density verse Voltage (J V) characteristics of a tandem cell before and after light illumination. Copyright 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 1.3.4. Tandem Polymer Solar Cell Characterization. Due to the embedded nature of the interlayer and its non-conductive nature of a two-terminal tandem cell, characterization of sub-cells independently in a tandem structure is impossible. Therefore, the way to characterize tandem cells is most based on the knowledge of a single cell. For a single cell, poly (3-hexylthiophene) (P3HT) is one of the most studied polymer BHJ solar cell applications. Important qualities of the thin film include the surface morphology and the degree of crystallinity domain are very necessary in the study of power conversion efficiency. The film morphology can be further modified through thermal annealing leading to better order within the P3HT and mixing of the blend. 36 As a critical parameter, the degree of phase separation can be studied by Atomic Force Microscopy (AFM) and Transmission Electron Microscopy (TEM). 37 In addition, the high-resolution cross-section TEM can reveal the layer by layer feature in the tandem solar cell. The Power Conversion Efficiency (PCE), Open Circuit Voltage (V OC ), Short Current Density (J SC ), and Fill Factor (FF) 21

can be detected in a way that the device is test under the solar simulator and analyzed by Keithley 236 source measurement. 1.3.5. Processing Issues of the Tandem Structure One of the major challenges in fabricating polymer tandem solar cells is layer-by-layer solution procedure without washing away the layer underneath. Initial efforts on the hybrid tandem solar cell structure fabrication utilized a thermal deposition method to build up the bottom cell, and the fully solution processed tandem cells are even more challenging. One approach is to use orthogonal solvents so that solution processing of one layer does not affect the underlying layer. Most polymers for solar cell application are soluble in chlorinated organic solvents such as dichlorobenzene (DCB), chloroform (CB) etc. Therefore, finding a solvent for the top polymer layer that does not dissolve the underlying polymer layer is very difficult. One concern raised by Sista et al is that the use of low boiling point solvent (the solvent drying process is fast) for the upper polymer layer can largely prevent damage to the underlying polymer layer. 22

CHAPTER II EXPERIMENT 2.1. Materials Preparation 2.1.1 Substrate: Conductive and transparent indium tin oxide (ITO) coated glass substrates with a sheet resistance of 15 V and surface roughness ~2nm have been purchased. Rectangular pieces of the correct size for the experiments will be cut especially for solar cell device fabrication. 2.1.2. PEDOT:PSS: the doped p-type soluble semiconductor PEDOT:PSS (VP AI 4083 from H. C. Stark, PEDOT4083) has been purchased from Heraeus Precious Metals GmbH & Co.. The material has been stored in the refrigerator. 2.1.3. Barium oxide: The barium oxide (BaO) has been purchased from Sigma-Aldrich. It is a white powder with the density of 5.72 g/ml at 25 C(lit.). 23

2.1.4. Solvent: For achieving optimum performance, we used chlorobenzene (CB) as the solvent for upper polymer layer and dichlorobenzene (DCB) as the solvent for underlying polymer layer. 2.1.5. PCPDTBT: PCBM and P3HT:PCBM ratio and concentration: The best device performance is achieved when the mixed solution has PCPDTBT:PCBM ratio of 1.0:3.0 with a concentration of 0.7wt % PCPDTBT plus PCBM 2.5wt% in Chlorobenzene (CB), and P3HT:PCBM ratio of 1.0 : 0.8 with a concentration of 1wt% P3HT plus PCBM 0.8wt% in a mixed solvent composed of 97% Dichorobenzene (DCB) and 3% 1,8-diiodooctane (DIO). The molecular structures of the active materials: PCPDTBT, P3HT, PCBM are schematically shown in figure 2.1. 38 O N S N OO S S n S n n Figure 2.1 Molecular structures of PCPDTBT, P3HT, PCBM respectively 2.2. Device Fabrication Procedures 2.2.1. Substrate clearance: The ITO-coated glass substrate was cleaned with detergent, then ultrasonicated in acetone and isopropyl, and subsequently dried in an oven overnight. 2.2.2. Device Fabrication Procedure: Polymer tandem cells were prepared according to the following procedure: The initially cleaned ITO-coated glass substrate 24

was UV-Ozone treated for 20 mins and hold for another 20 mins. Conducting poly(3,4-ethylenedioxylenethiophene)-polystylene sulfonic acid (PEDOT:PSS) was spin-cast (4000 rpm, 30s) with thickness ~40 nm from aqueous solution (after passing a 0.45 µm filter). The substrate was dried for 10 minutes at 150 C in air, and then after cooling down to room temperature, moved into a glove box for spin-casting of the photoactive layer. The dichlorobenzene (DCB) solution comprised of P3HT (1wt%) plus PCBM (0.8wt%) was then spin-cast at 1200 rpm with thickness ~150 nm on top of the PEDOT layer to become the first charge separation layer of the tandem cell. Then the substrate was placed on a hot plate, thermal annealed at 80 C for 15mins to further increase the degree of crystallinity. Afterwards, the substrate was pumped down in vacuum (~10-6 torr), and a thickness of 5nm barium oxide (BaO), 2.5nm Ag, 5nm molybdenum oxide (M O O 3 ) was thermal deposited at the rate of 0.1A/s in sequence. After finishing the deposition of the intermediate layer, the substrate was transferred into the glove-box to spin coating another polymer system of PCPDTBT: PCBM at the spin rate of 1600rpm/s with thickness~100nm. Then the substrate was thermal annealing at 80 C for 5 mins. Finally, the device was pumped down in vacuum (~10-6 torr) again, and a ~100 nm Al electrode was deposited on top. The deposited Al electrode area defines an active area of the devices as 4.5 mm 2. Therefore, the structure of the polymer tandem solar cell is ITO/40 nm PEDOT/150 nm P3HT: PCBM /5 nm BaO / 2.5nm Ag / 5 nm M O O3 / 100nm PCPDTBT:PSS/5nm Ca/100 nm Al. 25

Another tandem solar cell device was fabricated with the upper polymer system using P3HT:PCBM, and the fabrication procedure is as in a similar fashion as the former stated one. The device geometry with P3HT:PCBM and PCPDTBT: PCBM as upper polymer system are schematically shown in Figure 2.2. Figure 2.2. Polymer tandem solar cell geometry with ( a) PCPDTBT:PCBM as an upper layer and (b) P3HT:PCBM as an upper layer. 2.3. Calibration and Characterization For calibration of the solar simulator, the solar spectrum was carefully minimized using an AM 1.5G filter, and then the light intensity was calibrated using calibrated standard silicon solar cells. Current density-voltage characteristics were measured with a Keithley 236 source measurement unit. 2.4. UV-Vis Absorption Spectrum The absorption spectrum of the reference single cells and the corresponding tandem cells were recorded by using a spectrometer (Hitachi U-3900 PC). 2.5. Atomic Force Microscopy (AFM) Observation 26

The quality of the buffer layer and the intermediate layer including the surface roughness were checked by Atomic Force Microscopy (AFM). The AFM was used in tapping model. The AFM images were taken on a surface area of 5.0µm * 5.0µm. The instrument settings were a scan rate of 0.996 Hertz with a set-point of 330mV. At least two images weretaken from separate locations on each sample to ensure that they are representative. The AFM images will be analyzed using the Nanoscope Analysis software. 2.6. Sol-gel ZnO Nanoparticals Preparation The fabrication of sol-gel processed ZnO nanoparticle films with different surface morphologies were made from spin coating the same precursor solution but annealing under different conditions. The precursor solution, consisting of 0.75M zinc acetate dihydrate and 0.75M monoethanolamine in 2-methoxyethanol was first spun-coated onto indium tin oxide (ITO) substrates at 4000rpm for 40s. For fast annealing treatment, the substrate was immediately placed onto a hotplate that was preheated at 250 o C for 5min. For the slow annealing treatment, the spin-coated substrate was first placed onto a hotplate that was initially at room temperature while it was still not completely dry. The temperature was then raised at a ramping rate of 50 o C/min to 250 o C and the substrates were subsequently removed from the hot plate when the final temperature was reached. 27

CHAPTER III RESULTS 3.1. Energy Levels Because of the embedded nature of the tandem solar cells, it s impossible to investigate the two sub-cells in the tandem structure independently. Hence, we fabricated the two reference cells independently with the geometries: (a) ITO/PEDOT: PSS/P3HT: PCBM/Ca/Al. (b) ITO/PEDOT: PSS/PCPDTBT: PCBM /Ca/Al. The 28

energy levels of the single reference cells are schematically shown in Figure 3.1. Figure 3.1 Energy levels of single cells composed of bulk polymer blends (a) P3HT:PCBM and (b) PCPDTBT:PCBM. As for the tandem solar cell, both two types of tandem solar cells use polymer system of P3HT: PCBM as the bottom cell while different materials for upper cell, the proposed tandem solar cell device geometries are: (a) ITO/PEDOT:PSS/P3HT:PCBM/BaO/Ag/M O O 3 /P3HT:PCBM/Ca/Al. (b) ITO/PEDOT:PSS/P3HT:PCBM/BaO/Ag/M O O 3 /PCPDTBT:PCBM/Ca/Al. Figure 3.2 is the energy-level diagram showing the HOMO and LUMO energies of each of the component materials in a tandem structure. 29

Figure 3.2 The energy levels of the tandem solar cells composing the upper polymer layer of (a) P3HT:PCBM and (b) PCPDTBT:PCBM. 3.2. Performance Investigation 3.2.1. Current Density-Voltage (J-V) Characteristics (a) ITO/PEDOT: PSS/P3HT:PCBM/BaO/Ag/MOO3/P3HT:PCBM/Ca/Al. The Current Density- Voltage (J-V) characteristics under Air Mass 1.5 Global (AM1.5G) illumination and in dark of the reference cell P3HT:PCBM and the tandem structure with two identical sub-cells P3HT:PCBM are shown in Figure 3.3. The photovoltaic performances are summarized in Table 2. 30

a 31

b Figure 3.3 The Current Density-Voltage (J-V) characteristics of single reference cells using P3HT:PCBM and tandem cell fabricated using the identical reference cells. (a) J-V curves under illumination and (b) in dark. Table 1. Photovoltaic performance of the single cells and the corresponding tandem cell. Device PCE (%) V OC (V) Jsc ( ma /cm 2 ) FF (%) P3HT : PCBM 2.31 0.60 6.93 55.4 Tandem 1.00 1.20 1.78 47.0 (b) ITO/PEDOT: PSS/P3HT:PCBM/BaO/Ag/M O O3/PCPDTBT : PCBM/Ca/Al. The Current Density Voltage (J-V) Characteristics of single solar cells and the tandem solar cell with P3HT: PCBM and PCPDTBT: PCBM polymer systems are shown in Figure 3.4. The characterization was under Air Mass 1.5 global (AM1.5G) 32

illumination. The photovoltaic performance of the single cells and tandem cells are summarized in Table 2. a 33

a b 1 Figure 3.4 The current density-voltage (J-V) characteristics of single reference cells using P3HT:PCBM and PCPDTBT:PCBM and tandem cell fabricated using the same polymer system. (a) J-V curves under illumination and (b) in dark. Table 2. Photovoltaic performance of the single reference cells and the corresponding tandem cell. Device PCE (%) V OC (V) J SC (ma FF (%) /cm 2 ) P3HT : PCBM 2.31 0.60 6.93 55.4 PCPDTBT :PCBM 2.38 0.65 7.61 48.2 Tandem 1.06 0.95 2.70 41.6 34

3.2.2. UV-Vis Absorption The absorption spectrum of a film of the bulk heterojunction composite of each sub cells containing P3HT:PCBM and PCPDTBT:PCBM polymer systems, respectively, and the bilayer tandem cell composed of P3HT:PCBM/PCPDTBT:PCBM is shown in Figure 3.5. 1.5 P3HT:PCBM PCPDTBT:PCBM P3HT:PCBM/PCPDTBT:PCBM Absorbance (a.u.) 1.0 0.5 0.0 300 350 400 450 500 550 600 650 700 750 800 850 Wavelength (nm) Figure 3.5 Absorption spectra of a PCPDTBT:PCBM bulk heterojunction composite film, a P3HT:PCBM bulk heterojunction composite film, and a bilayer of the two as relevant to the tandem device structure. a.u., optical density. The PCPDTBT:PCBM polymer system has weak absorption in the visible spectral range but has two strong bands: one in the near-infrared (near-ir) between 700 and 850 nm resulted from the interband p-p* transition of the PCPDTBT and one in the ultraviolet (UV) arising primarily from the HOMO-LUMO transition of the PCBM. The absorption of the P3HT:PCBM film falls in the complementary apart of the PCPDTBT:PCBM spectrum and covers the visible spectral range. The electronic absorption spectrum of the tandem structure can be just described as a superposition of the two complementary composites absorption spectra. In addition, the 35

PEDOT:PSS and intermediate layers have negligible absorption in the tandem device structure. 3.2.3. Atomic Force Micrometer (AFM) Images Observation Figure 3.6 illustrates the Atomic Force Micrometer (AFM) tapping mode height images of the intermediate layer BaO/Ag/M O O3 and Hole Transportation Layer PEDOT: PSS coated on the ITO substrate. The islands observed are due to surface roughness and it is clear to see that the distribution of the up-and-down trend is more intense on the PEDOT:PSS surface. The calculated room-mean-square roughness of the intermediate layer and PEDOT:PSS are 0.99nm and 1.72nm, respectively. Figure 3.6 AFM images of (a) M O O3 surface morphology of BaO/Ag/M O O3 intermediate layer and (b) PEDOT:PSS on ITO coated glass substrate. The islands observed are due to surface roughness. Note that the islands distribution is more intensive for (b), indicating the roughness is higher for PEDOT:PSS. 36

3.3. Comparison of the Intermediate layers Figure 3.7 shows AFM height images of sol-gel ZnO nanoparticles fabricated on a ITO coated substrate. It can be seen that the size of ZnO is significantly greater under the fast annealing condition. The calculated Root-Mean-Square roughnesses of the ZnO nanoparticles under fast-annealing and slow-annealing are 3.82nm and 1.39nm, respectively. Figure 3.7 AFM height profiles of ZnO nanoparticles under the condition of (a) fast annealing and (b) slow annealing We fabricated the intermediate layer containing an electron transportation layer sol-gel ZnO nanoparticles and a hole transportation layer thermal deposited MoO 3. The ZnO thin films are prepared in two annealing fashions namely, fast annealing and slow annealing. The electronic performance of the intermediate layer and the procedures are illustrated in Figure 3.8. 37

Figure 3.8 The intermediate layer composed of ZnO/MoO 3 (a) Current Density-Voltage (J-V) Characteristics and (b) Sol-gel preparation of ZnO nanoparticles. As we can tell from the Figure 3.8, the current under the slow annealing condition shows a better diode rectification within the sweep voltage from -1V to 1V. However, when the applied bias was extended as large as 2V, the current under negative bias is just slightly smaller than the current in the forward bias. The obvious diode rectification performance different in the low sweep voltage may tentatively due to the different surface roughness of the ZnO thin film. As the surface get rougher, there are more defects on the surface leading to the electron traps. In that case, under a low bias voltage, the current density versus voltage performance is quite different. When comes to a larger bias, the ZnO and MoO 3 intrinsic property would dominate the diode performance and annealing method does not have much influence. Figure 3.9 illustrates the BaO/MoO 3 P-N junction current density versus voltage performance and the conductivity property of the intermediate layer BaO/Ag/MoO 3. The diode performance of BaO/MoO 3 is comparable to the ZnO/MoO 3 with a slightly decrease in the negative bias voltage. After inserting an 38

electron hole recombination layer Ag in between BaO and MoO 3, the intermediate layer shows a very good electronic conductivity as illustrated in Figure 3.9 (b). Figure 3.9 The electronic performance of BaO/Ag/MoO 3 intermediate layer. (a) Diode property of BaO/MoO 3 P-N junction; (b) Conductive property of BaO/Ag/MoO 3 intermediate layer. 39

To further investigate the quality of these two intermediate layers, we checked the surface morphology of them and make comparison (Figure 3.10).The calculated root-mean-square roughness ZnO/MoO 3 is 1.16nm which is slightly rougher than the surface of BaO/Ag/MoO 3. Figure 3.10 AFM images of two types of intermediate layers (a) ZnO/MoO3 and (b) BaO/Ag/MoO3 on ITO coated glass substrate. The islands observed are due to surface roughness. 40