Introduction

1-1 Introduction of conjugated polymers

In the late 70’s, Shirakawa et al. found a doped conjugated polymer, polyacetylene, with an unusual feature containing conductivity. The electrical conductivity of polyacetylene was induced to 10³ S-cm^-1 by using electrochemical doping with iodine.

After that, several conjugated polymers, such as poly(p-phenylene) (PPP), polythiophene (PT), polyaniline, polypyrrole, and poly(p-phenylenevinylene) (PPV)

were synthesized to investigate their electro-optical properties, as shown in Figure 1-1.¹ In recent years, π-conjugated rigid-rod polymers with semiconducting properties have

gained great interests since they possess unique optoelectronic properties with potential to be used in flexible electronic devices such as electrochromic devices, transistors, photovoltaics, and light-emitting diodes (LEDs), etc. In all of the conducting polymers, poly(alkylthiophene) (P3AT) especially attracted the much attention as an important material due to its excellent optical and electrical properties as well as exceptional thermal and chemical stability.

The conducting polymer is arranged of a single bond and a double bond by turns.

The arrangement of conducting polymer leads to the electrons and the holes can move along the main molecular chain or across the anther molecular chain. This phenomenon

is called conjugated structure, and it has resonance effect. So conducting polymer, more precisely, called conjugated conducting polymer.

Figure 1-1 The principal conducting polymers.¹

Similar to inorganic semiconductors, conjugated polymers are organic semiconductors composed of electronic energy levels. Electrons are able to occupy in

3

bands rather than in discrete levels and the ground state energy bands either can be

entirely filled or emptied. The band structure of a conjugated polymer originates from the interactions of the π-orbitals of the repeating units throughout the chain. Analogous

to semiconductors, a concept from organic chemistry was introduced that at the lowest level of conduction band is called the lowest unoccupied molecular orbital (LUMO) while at the highest level of valence band is called the highest occupied molecular orbital (HOMO). The difference in energy between HOMO and LUMO is called band gap, Eg. Additionally, a conjugated organic polymer, that is in the metallic conducting regime (~1-10⁴ S/cm), either an insulator or a semiconductor having a small conductivity (10^-10-10⁴ S/cm), can be converted to a conductor via dopping process.

The doping of all conducting polymers can be accomplished by redox doping, where the partial addition (reduction) or removal (oxidation) of electron form the π system of the

polymer backbone. Hence, conjugated organic polymers, having the electronic, magnetic and optical properties of a metal while maintaining the processability, mechanical properties associated with a common polymer, are termed intrinsically conducting polymers (ICP) as “doped” form of polymers. The most critical challenges in developing ideal p-type materials are to design and synthesize a conjugated polymer that simultaneously possesses good film-forming properties, strong absorption ability, high hole mobility, and suitable HOMO-LUMO energy levels.² A fundamental

understanding of molecular design and the benefits of versatile polymer syntheses allow for the effective tailoring of the intrinsic properties of conjugated polymers to serve the desired purpose and address the application needs.

1-2 Configuration and synthesis of poly(alkylthiophene)

The study of polythiophene has intensified in 1980s. Conductivity resulting from electron delocalization is not only interesting property, but also the optical properties of these materials, with dramatic color shifts in response to changes in solvent, temperature, applied potential. However, in that time, the development of polythiophene was restricted due to its poor solubility. Until 1985, Elsenbaumer et al. developed a new polythiophene derivative poly(3-alkylthiophene) (P3AT) which contained alkyl side chain to solve the problem of solubility.³ After that, a lot of relative research had been investigated. There are many ways to synthesize P3AT, such as electrochemical polymerization (Sato1991) and oxidative polymerization by iron(III) chloride (Sugimoto 1986) had been reported in early time, but the configuration of the P3AT synthesized by aforementioned method was regiorandom which will cause the photovoltatic property decreased. Later on, McCullough reported another method for the synthesis of regioregular poly(3-alkylthiophene)s by Grignard metathesis (GRIM) in

5

1992,⁴ who used Grignard reagents and nickel complex catalyst (Ni(dppp)C12) to synthesis a highly head-to-tail (H-T) configuration P3AT. McCullough and Yokozawa independently demonstrated that the Grignard metathesis polymerization of regioregular 3-alkylthiophene proceeds by a living chain growth mechanism instead of the traditionally accepted step growth polycondensation.⁵ As a result, low polydispersities (~1.2-1.3) and well-defined molecular weights can be controlled by the feed ratio of monomer to the Ni catalyst. In 1995, Rieke et al. synthesised a highly head-to-tail (H-T) regioregular P3AT up to 98% using highly reactive "Rieke zinc" to activate the monomer and nickel complex (Ni(dppp)Cl2) as a initiator.⁶ Increasing regioregularity in P3AT through these advanced metal-catalyzed reactions leads to various beneficial outcomes including a red shift in absorption in the solid state with an intensified extinction coefficient and an increase in the mobility of the charge carriers.

Because 3-alkylthiophene is an asymmetrical molecule, there are three relative

orientations in poly(3-alkylthiophene) configuration when the two thiophene rings are coupled between the 2- and 5-positions. The first of these is the 2-5′ or head-to-tail coupling (HT), the second is 2-2′ or head-to-head coupling (HH), and the third is 5-5′ or

tail-to-tail coupling (TT), as shown in Figure 1-2. The polymer configuration of the P3AT can be analyzed by H-NMR. Chemical shift on the proton was different of the thiophene ring (H1) and α-methylene-H of the side chain, as shown in Table 1-1 and

Figure 1-3. The best configuration is HT-HT for the photovoltaic property, which arrangement of alkyl side is more regular than other configurations. Otherwise, the

HT-HT configuration leads to the arrangement of polymer backbone, which form a coplanar structure by increasing π-π stacking, and it can be proved by X-ray

diffraction.⁷ The morphology of HT-HT P3AT is lamellar,⁸ and the 2-D lamellar structure with interlayer spacing of H-T P3ATs was shown in Figure 1-4.

S

Figure 1-2 Configurations of poly(3-alkylthiophene).⁶

7 S

H CH₂ CH₂

R

n 1

α β

Table 1-1 H-NMR chemistry shift (ppm) of H in different regiostmctures.⁶

HT-HT TT-HT HT-HH TT-HH

H1 6.98 7 7.02 7.05

head-to tail head-to-head

α-methylene-H 2.8 2.58

β-methylene-H 1.72 1.63

Figure 1-3 NMR spectra of (a) regiorandom P3HT and (b) regioregular P3HT.⁶

Figure 1-4 2-D lamellar structure with interlayer spacing of H-T P3ATs. (a) Intermolecular π stacking between thiophene rings. (b) Lamellar stacking.⁸

1-3 Catalyst-transfer polymerization: Grignard metathesis, GRIM

Polycondensation with a catalyst exploits a new mechanism for chain-growth polycondensation: that is a catalyst-transfer mechanism, in which the catalyst activates the polymer end group, followed by reacting with another monomer and transfer of the catalyst to the chain end of elongated polymer, in other words, the polymerization progress in a similar manner to biological polycondensation. The chain growth mechanism had been discussed extensively in recently years. In 2004, a credible mechanism was reported by Mccullough et al. on Macromolecules. The assumption is

(a)

(b)

9

shown in Figure 1-5,⁹ first, a 2,5-dibromo-3-hexlythiophene monomer reacts with Grignard regent (RMgX) to form 2-bromo-5-bromomagnesium-3-hexylthiophene, then it reacts with Ni(dppp)Cl2 (dppp = 1,3-bis(diphenylphosphino)propane) to yield a new organonickel compound. Next, the reductive elimination occurs, and the organonickel compound transfers to an associated pair quickly, which consists of the tail-to-tail aryl halide dimmer and nickel(0). The dimmer undergone oxidative addition to nickel center generates nickel complex fast, in which terminal C-Br bond reacts with a new monomer to form an organonickel compound, following by another reductive elimination quickly.

The next following steps are as the same as aforementioned. The growth of polymer chain is accomplished by insertion of one monomer at a time in the reaction cycle. After many experiments, they defined Grignard method as a living nature polymerization, and also found that the molecular weight of polymer can be predicted by the molar ratio of monomer to Ni(dppp)Cl2, which means that one Ni(dppp)Cl2 compound initiates one polymer chain. Therefore, the polydispersity index (PDI) is quite narrow.⁹ Furthermore, T. Yokozawa used matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometry to prove all polymer have the same end qroup (one bromine atom and one hydrogen atom). Based on the result and another experiments, four important points were clarified: (1) the polymer end groups are uniform among molecules, one end group is Br and the other is H; (2) the propagating end group is a

polymer-Ni-Br complex; (3) one Ni molecule forms one polymer chain; and (4) the chain initiator is a dimer of 2-bromo-5-bromomagnesium-3-hexylthiophene formed in situ. They proposed a mechanism of chain-growth polycondensation as shown in Figure

1-6. First, two Grignard nucleophilic additions to a nickel catalyst generate the intermediate. A reductive elimination involving carbon-carbon bond formation accompanied by Ni migration and insertion into the terminal C-Br bond keeps the living chain capable of further reacting with 2-bromo-5-bromomagnesium-3-hexylthiophene.

Propagation via consecutive coupling between the polymer with a Ni complex at the chain end and compound 2-bromo-5-bromomagnesium-3-hexylthiophene elongates the conjugated backbone. Growth continues in such a way that the Ni catalyst moves to the polymer end group. Finally, the hydrogen end group is generated from elimination of nickel complex upon quenching by hydrogen chloride.

11

Figure 1-5 Proposed Mechanism for the Nickel-Initiated Cross-Coupling.⁹

Figure 1-6 Proposed Mechanism of Chain-Growth Polymerization.⁵

In a conventional chain-growth polycondensation mechanism, substituent effect is usually the critical issue that makes it difficult to control the molecular weight and to obtain polymers of very high molecular weights. By comparison, the beneficial features of GRIM are easy to synthesize monodispersed P3ATs without relying on time-consuming polymer fractionation techniques and eliminating the problems arising from non-uniform molecular weight distribution. Moreover, the catalyst-transfer polycondensation can be accomplished both at room temperature and on a large scale, thus, the Grignard metathesis coupling has become the most widely used method for producing P3ATs with predetermined high molecular weights.

Successive application of GRIM polymerization is also expected to be a universal and convenient method to synthesize all-conjugated block copolymers composed of various aromatic ring monomers. Recently, practical synthetic method for all-conjugated block copolymers with phenyl and thiophene rings has been investigated.

It is an important issue that which monomer should be polymerized first for successful synthesis of block copolymers. Yokozawa et al. demonstrated that the uncontrolled postpolymerization occurred while monomer of P3HT was polymerized first followed by monomer of poly(p-phenylene) (PPP). However, successful and well-controlled block copolymerization was carried out in the reverse sequence.¹⁰ The scheme is shown in Figure 1-7. It is associated with the π-donor ability of P3HT and PPP since the

13

π-electrons of the polymers are considered to assist the transfer of the Ni catalyst in GRIM polymerization. The thiophene ring has strong π-donor ability than the phenylene

ring. Therefore, it is difficult for catalyst to move to the terminal C-Br bond of phenylene ring for polymerizing the phenylene monomers while P3HT is polymerized first followed by PPP. The similar result was observed for synthesis of polyfluorene-b-poly(3-hexylthiophene) (PF-b-P3HT) using turbo GRIM reported by McCullough's group.¹¹ More recently, novel heterocyclic block copolymers, poly(3-hexylseleophene)-b-poly(3-hexylthiophene) (P3HS-b-P3HT), was synthesized by Seferos's group.¹² The morphology of P3HS-b-P3HT revealed clear phase separation with distinct domains according to their different heterocyclic constituents.

Figure 1-7 The synthesized sequence of PPP-b-P3HT by Grignardmetathesis.¹⁰

1-4 Self-assembly of conjugated polymer-containing block copolymers

Because of their special optical and electrical properties, conjugated polymers with rigid π-conjugated backbones have gained great interests in the region of polymer

science during the past few years. Recently, block copolymers containing conducting polymer segments, such as poly(p-phenylene) (PPP), polyfluorene (PF), poly(p-phenylenevinylene) (PPV) and polythiophene (PT), are technologically important since the incorporation of conjugated polymer into block copolymers with other functional coil-like polymers can generate distinct mechanical properties and provide novel approach for organic optoelectronic device fabrication. In order to optimize device performance from these structure-directing block copolymers, the definite self-assembly behavior of these conjugated block copolymers to precisely control active layer morphology and interfacial structure is greatly needed. Although the self-assembly behavior of conjugated block copolymers has attracted increasing attention for material science and nanofabrication purpose, the experimental studies for these novel materials are largely unknown.

Recently, there are some approaches which have performed to setup model system for conjugated block copolymers and illustrate their equilibrium thermodynamic phases.

Segalman et al. studied the self-assembling behavior of poly(2,5-di(2'-ethylhexyloxy)-1,4-phenylenevinylene)-b-polyisoprene (DEHPPV-PI)

15

and discuss their results with previous theoretical predictions.^{13, 14} On the basis of their

experimental results, lamellar structures are observed when the segregation strength between compositional blocks is weak or when the geometric asymmetry parameter (ν)

is low. On the other hand, non-lamellar phases such as hexagonal and spherical structures have been observed in the conditions of high asymmetry or increased segregation strength. Figure 1-8 shows the TEM images and Fourier Transforms of PPV-b-PI block copolymers. Furthermore, the resulting phase behavior derived from these observations is in a qualitative agreement with the theoretical calculations for rod-coil block copolymers since the PPV chains are perfect rigid rods, free of the molecular folding and conformational changes. Additionally, the recent work

contributed by Ho et al.¹⁵ has further discussed the phase diagrams in terms of the ratio of the rod-rod interaction (μ) to the rod-coil interaction (χ) in a

poly(diethylhexyloxy-p-phenylenevinylene-b-methylmethacrylate) (DEH-PPV-b- PMMA) system and showed the phase behavior of these copolymers (Figure 1-9). In 2005, McCullough and coworkers have reported that poly(3-hexylthiophene)-b-poly(3-dodecylthiophene) (P3HT-P3DDT) block copolymers were synthesized via a chain growth mechanism.¹⁶ Also, Hashimoto et al. reported the synthesis of poly(3-hexylthiophene)-b-poly[3-(2-ethylhexyl)thiophene]

(P3HT-b-P3EHT) block copolymer and demonstrated that these block copolymers can

self-assemble into a lamellar morphology in solid state.¹⁷ Meanwhile, Ueda et al. studies the poly(3-hexylthiophene)-b-poly(3-phenoxymethylthiophene) block copolymers system and observed the formation of nanophase-separated lamellar or sheet-like solid state morphologies.¹⁸

Among the various side chains modified conjugated polymers, the highly regioregular P3HT have received more attention since they show high crystallinity from main chain packing and excellent chemical stability and have been applied for photovoltaic and thin film transistors.¹⁹ P3ATs have been traditionally synthesized by electrochemical and oxidative polymerizations. However, these resultant polymers tend to be ill-defined with large polydispersities. Thus, Yokozawa et al. discovered that well-defined all-conjugated block copolymers with different main chains can be synthesized with low polydispersity via a chain growth mechanism. They reported the detailed synthesis on poly(2.5-dialkoxy-1.4-phenylene)-b-poly(N-hexyl-2.5-pyrrole) (PPy-b-PPP)²⁰ and poly(2,5-dihexyloxy-p-phenylene)-b-poly(3-hexylthiophene) (PPP-b-P3HT)¹⁰ block copolymers. McCullough et al. discovered that well-defined end-functionalized P3HTs can be synthesized with low polydispersity via a chain growth mechanism by using the similiar catalyst-transfer polycondensation method.²¹ McCullough et al. further demonstrated that P3HT block copolymers can be synthesized from a linker molecule attached to an end-functionalized P3HT as a macroinitiator via

17

atom transfer radical polymerization (ATRP). They have found that thin and ultrathin films these block copolymers containing polythiophene that were prepared by casting from toluene followed by free evaporation of a solvent reproducibly self-assembled into well-defined nanowires, as shown in Figure 1-10.²²

Figure 1-8 TEM images and Fourier Transforms of PPV-b-PI block copolymers. (a) PPVbPI-31, (b) PPVbPI-41, (c) PPVbPI-57, (d) PPVbPI-71, (e) PPVbPI-81, (f)

PPVbPI-87 and (g) PPVbPI-91. Fourier transforms of these polymers show that both have 6-fold symmetries characteristic of a hexagonal structure. (h) PPVbPI-87 and (i) PPVbPI-91.¹⁴

Figure 1-9 Phase behavior of PPV-b-PMMA block copolymers.¹⁵

Figure 1-10 Nanowire morphology in PS-P3HT block copolymers solvent-cast from toluene and visualized with tapping-mode AFM. Left: height image; right: phase image.²²

19

1-5 Conjugated polymer/inorganic nanoparticles hybrid system

Polymer-inorganic hybrid materials include a broad variety of systems. For instance, a polymer can act as a matrix for dispersed inorganic nanoparticles thus constituting what is known by the name of nanocomposites. Besides other preparation methods, processes based on in situ particle synthesis including sol-gel processes have frequently been applied. Such methods can prevent agglomeration of inorganic species in the final products, which is often a problem when preformed nanoparticles and polymers are mixed, unless the particles are modified with an organic surface layer.

Importantly, polymer-inorganic hybrids can exhibit materials properties which are more pronounced or even differ from those of comparable polymer composites with larger inorganic particles, such as optical properties (e.g., transparency and color, including dichroism), magnetic properties (superparamagnetism), mechanical properties, chemical properties (catalytic or sensory activity), and gas barrier properties. Thus, polymer-inorganic hybrid materials are considered to find application in various areas, for example in photovoltaic cells, optics, sensor technology and electronic devices.

The conjugated polymer–inorganic semiconductor hybrid systems combine the advantages from both organic and inorganic materials. Conjugated polymers (e.g., P3HT), when self-organized into crystal structure, can own a high hole mobility, and can also be easily processed onto the surfaces of both rigid and flexible substrates.

Nanoscale inorganic materials exhibit different optical absorption and photocurrent generation properties from bulk materials due to their quantum size confinement. They have advantages including relatively high electron mobility, high electron affinity and good thermal stability. Solution-processible nanostructured inorganic semiconductors also provide the possibility to have a large interfacial area for efficient exciton dissociation when blending with soluble polymers.²³ One-dimensional (1-D) ordered nanostructure inorganic semiconductors aligned on a substrate can provide an ideally straight pathway for carrier transport. Generally when organic and inorganic components are combined into a heterojunction device, the polymers are used as donors to absorb sunlight and transport holes, while the inorganic semiconductors CdSe function as acceptors to transport electrons. In such devices, an energy conversion efficiency exceeding 3% has been in reach.²⁴ More recently, solid state dye-sensitized solar cell (DSSC) structure has been adopted to fabricate organic (polymer)-inorganic hybrid solar cells.^{25, 26} In these devices, the inorganic semiconductors (e.g., porous TiO2) are sensitized by a traditional dye or a light absorbing inorganic semiconductor (e.g., Sb2S3), and the polymers function as a hole transporter to reduce the dyes and/or also work as an additional donor to absorb light.²⁷ A power conversion efficiency of 5.13%

has recently been achieved in this type of hybrid solar cells. The different functions of organic and inorganic materials provide additional opportunities to improve solar cell

21

performance by taking advantages of organic solar cells, inorganic semiconductor solar cells, and DSSCs. In general, because the interfacial charge separation is the critical step in the whole photovoltaic process, the larger the interfaces, the more the opportunity for the excitons to reach the interfaces, and probably the higher the conversion efficiency.

Therefore, most of the hybrid photovoltaic materials that have been studied are nano-structured composites. There are many types of nanostructures, including particles, rods, tubes, tetrapods, sheets, needles, and porous network, etc., and the way to mix the components together can be either disordered or ordered as shown in Figure 1-11.²⁸

Figure 1-11 Various nano architectures of solar cell materials. (a) Blend of semiconductor nanoparticles and conducting polymer films; (b) blend of semiconductor nanorods and conducting polymer films; (c) blend of semiconductor nano-tetrapods and

conducting polymer films; (d) conducting polymer immersed in porous semiconductor nano-network; (e) blend of semiconductor nanorods arrays and conducting polymer films; and (f) blend of semiconductor nanotube arrays and conducting polymer films.²⁸

1-6 Configuration and synthesis of TiO

nanostructures

Titanium oxide, a wide band gap and n-type semiconductor, exists in three crystalline forms- anatase, rutile and brookite. The more important crystalline phases are anatase and rutile, which occur in atmosphere and are relatively easier to be prepared.^{29, 30} The phase transformation temperature between anatase and rutile is around 600 ^oC, as shown in the phase diagram of TiO2 (Figure 1-12).³¹

Both anatase and rutile phase of TiO2 belong to the tetragonal crystal system. The structures of anatase and rutile phase can be described in terms of chains of TiO6

octahedra. The main differences of these two structures are the distortion of each octahedron and the assembly pattern of the octahedra chains, as shown in Figure 1-13, which reveals the unit cell structures of the anatase and rutile crystals respectively.²⁹

octahedra. The main differences of these two structures are the distortion of each octahedron and the assembly pattern of the octahedra chains, as shown in Figure 1-13, which reveals the unit cell structures of the anatase and rutile crystals respectively.²⁹