Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Metal oxides have been of interest in processing, synthesis, characterization, and fabrication in both polymer-inorganic hybrid and dye-sensitized solar cells. TiO2 [1–5], ZnO [6–9], CuO [10], and Nb2O5 [11] have been used as effective charge transport medium in solar cells. Different morphologies of these metal oxides have been synthesized for better charge transport across solar cells. These metal oxide nanostructures are chosen to provide large interfacial area and enhance charge transport across active layer. Metal oxide-based inorganic nanostructures can also improve environmental stability to cells, which is a major cause of degradation in cell performance. In this section, commonly used metal oxides (e.g., TiO2, ZnO, Nb2O5, and CuO) will be discussed for their role in fabrication of polymer solar cells.

Processing and Synthesis of Metal Oxide Nanostructures

In the past decades, metal oxide nanostructures have attracted great interest due to their potential applications in optoelectronics. Controlling the size and shape of nanostructures is crucial in developing novel properties in nanoscience research. Considerable efforts have been devoted in obtaining various nanostructures such as nanoparticles, nanospheres, nanorings, nanopolyhedrons, nanorods, nanowires, nanobelts, and nanosheets [12, 13], which showed that nanostructure properties and device performance greatly depend on their size, morphology, composition, and surface properties that can be tuned by synthetic methods. Numerous techniques have been developed to synthesize metal oxide nanostructures with a wide range of compositions, sizes, sophisticated crystallite shapes, and complex assembly properties. Although vapor-phase processing has been successfully employed for the preparation of metal oxide nanostructures, especially for one-dimensional nanostructures [14–17], solution-based chemical (wet chemistry) synthesis methods have become more versatile with regard to the controlled variation of structural, compositional, and morphological features. In fact, solution-based chemical synthesis such as sol–gel [18–22], chemical (co)precipitation [23], hydrothermal [24–28], combustion [29, 30], spray pyrolysis [31], microemulsion [25, 26, 32, 33–46], and electrospinning [24–28] has received considerable attention since they offer the possibilities to control homogeneity, purity of phase, size distribution, surface area, and microstructure uniformity. In this work, we will discuss general synthesis methods that are used to prepare several important semiconductors, such as TiO2, ZnO, Nb2O5, CuO, and others with different nanostructures.

TiO2

Historically, Fujishima and Honda reported electrochemical water splitting property on a TiO2 electrode in 1972 [45, 47, 48], and Brian O’Regan introduced high surface area TiO2 films in dye-sensitized solar cells in 1991 [49, 50]. TiO2 is one of the most extensively studied materials with more than 80,000 publications over the past 5 years based on the database, web of knowledge. TiO2, a large-band gap semiconductor, is known to be a very useful nontoxic, environmentally friendly, corrosion-resistant material used as pigment, paint, cosmetics, and catalyst. It has been widely studied with applications in photocatalysts, biosensors, and dye-sensitized solar cells (DSSCs) due to its unique optical and chemical properties [13, 34, 36, 47, 49, 51–55]. As a photocatalyst with large surface area, TiO2 facilitates diffusion of photoinduced electrons and holes towards the surface before their recombination. TiO2 having high dielectric constant and refractive index has also been widely used as optical coatings, beam splitters and antireflection coatings. In addition, TiO2 has also been reported in use as humidity, hydrogen, and oxygen sensors [56–60]. TiO2 has three crystalline polymorphs including anatase, rutile, and brookite. Rutile is a thermodynamically stable phase with smaller band gap than anatase. This section focuses on the recent progresses in the synthesis of TiO2 nanostructures such as nanoparticles, nanorods, nanotubes, nanofibers, and various doping in TiO2 nanostructures.

Nanoparticle

The anatase TiO2 nanoparticles have been one of the most widely studied among various nanostructures with high surface area and strong absorption capacity. For industrial-scale production, the pyrolysis of TiCl4 at high temperatures is used to produce Degussa P-25 TiO2 nanoparticles. The laboratory scale synthetic approaches to fabricate TiO2 nanoparticles include sol–gel [61–65], hydrothermal [28, 63, 66], solvothermal [67, 68], chemical vapor deposition [16, 69, 70], and microemulsion [25, 26, 32]. Among them, hydrothermal is widely used for the production of small particles of TiO2 [28, 63, 71–73]. TiO2 nanoparticles can be synthesized by hydrothermal processing of either peptized precipitates of a titanium precursor with water [28] or titanium alkoxide in an acidic ethanol–water solution with particle sizes in the range of 7–25 nm by changing the concentration of Ti precursor and the solvents [63].

The sol–gel method from hydrolysis of a titanium precursor is another easy and widely used process to prepare TiO2 nanoparticles. This process usually proceeds via a hydrolysis step of titanium alkoxide followed by condensation. A series of thorough studies have been conducted by Sugimoto et al. using the sol–gel method to prepare TiO2 nanoparticles with different sizes and shapes by tuning the reaction parameters [74–78]. Under different pH conditions, the selective adsorption of shape controller on specific crystal planes of TiO2 nanoparticles can tune the growth rate of these planes. Finally the size and shape is controlled by the growth rate in different crystal planes [74, 75]. Furthermore, the colloid dispersions of TiO2 nanoparticles prepared by sol–gel method are efficient precursor to make various TiO2 films on different substrates or nanostructures with various soft and hard templates [19, 21, 61, 73, 79–81].

Randomly Oriented and Highly Aligned Nanorods

TiO2 nanorods can be used in novel photovoltaic devices including hybrid bulk heterojunction and dye-sensitized solar cells [82–84]. The synthesis of TiO2 nanorods, especially arrays on the substrates, has attracted much attention. Weller et al. reported the controlled growth of high-aspect ratio anatase TiO2 nanorods by hydrolysis of titanium tetraisopropoxide in oleic acid as surfactant at 80 °C [85]. Li et al. successfully synthesized the near monodisperse TiO2 nanorods with their size, shape, and dispersibility controlled by adjusting reaction temperature, duration, and concentration of the reactants [68]. As previous mentioned, TiO2 nanorods can be prepared by combining sol–gel method and anodic alumina membrane (AAM) template [34, 57]. TiO2 nanorod arrays can be obtained by template-assisted sol–gel electrophoresis [18]. Direct oxidation of titanium metal plate with hydrogen peroxide can also produce crystalline TiO2 nanorods [86]. For application in photovoltaic devices, developing a simple method to grow TiO2 nanorod arrays on the transparent conductive substrates is important. The seed growth process is a commonly used method to prepare one-dimensional nanostructures on substrates. In order to prepare TiO2 nanorod arrays, a TiO2 polymeric sol prepared by sol–gel process was spin-coated on the fluorine-doped tin dioxide (FTO) substrate and then annealed as a seed layer, followed by the growth of nanorod arrays by hydrothermal method [42]. Aydil et al. invented a hydrothermal method for the first time to grow oriented, single-crystalline rutile TiO2 nanorod films on FTO substrates (Fig. 20.1) by simply mixing hydrochloride with titanium precursors [87]. The acidity, concentration, additives of surfactants, processing temperature, and time were used to optimize the diameter, length, and density of the nanorods.

Fig. 20.1
figure 1

FESEM images of oriented rutile TiO2 nanorod film grown on FTO substrate in 30 mL of deionized water, 30 mL of hydrochloric acid, and 1 mL of titanium butoxide at 150 °C for 20 h. (a) Top view, (b) cross-sectional view, (c) and (d) tilted cross section (Reprinted (adapted) with permission from Ref. [87]. Copyright (2009) American Chemical Society.)

Full size image

Nanotubes

The first titania nanotubes were probably synthesized by Hoyer through a template-assisted electrochemical deposition method in 1996 [88]. Since then, template-assisted methods [20, 89] and others including hydro-/solvothermal methods with or without templates [20, 24, 80, 90–92] and anodic oxidation method [89, 93, 94] have been reported. The hydrothermal method has been widely used to prepare TiO2 nanotubes since Kasuga et al. reported that the hydrothermal treatment of TiO2 particles in NaOH resulted in the formation of anatase TiO2 nanotubes with large surface areas in 1998 [80]. Teng et al. further studied the structural features of nanotubes obtained by NaOH treatment on TiO2 with different posttreatments [90]. Today, this method is widely used for preparing TiO2 nanotubes. Crystallized TiO2 nanotube arrays can be obtained by anodic oxidation of titanium foil and annealed at high temperature [43, 89, 93, 94]. The length and diameter of the TiO2 nanotubes can be controlled over a wide range with an applied potential from 1 to 25 V in optimized phosphate/HF electrolytes. Depositing Ti film on the transparent conductive substrates, followed by an anodic oxidation process, has become a typical method to grow TiO2 nanotube arrays [41, 44, 95, 96].

Nanofibers

TiO2 nanofibers can be fabricated by the template-assisted sol–gel method with TiO2 sol and porous alumina membranes or ‘track-etched’ polycarbonate filters as templates [20]. The template method is usually complex, cost- and time-consuming. Su et al. developed a soft hydrothermal chemical process via the reactions of amorphous TiO2 gel (or commercial TiOSO4 particles) and NaOH solution to synthesize TiO2 nanofibers with high surface area [97]. Today, TiO2 nanofibers can be conveniently prepared by electrospinning an alcohol solution that contains a polymer of high molecular weight and a titanium alkoxide precursor. Calcination in a subsequent step leads to the formation of polycrystalline nanofibers made of anatase TiO2 with controllable diameters and interesting porous structures (Fig. 20.2) [56, 98–102]. Figure 20.2 shows extremely long TiO2/polymer composite nanofibers (a) can be produced by electrospinning method, and the anatase TiO2 nanofibers (b) with a diameter around 50 nm can be easily obtained by the calcination of these composite nanofibers.

Fig. 20.2
figure 2

(a) SEM image of TiO2/PVP nanofibers that were electrospun from an ethanol solution containing Ti(OiPr)4 (0.1 g/mL) and PVP (0.03 g/mL). The electric field strength was 1 kV/cm. (b) SEM image of the same sample after it had been calcined in air at 500 °C for 3 h. (c) Histogram showing the size distribution of nanofibers contained in the calcined sample. The size distribution was obtained from the SEM images of about 100 nanofibers. (d) XRD pattern of the same calcined sample. All diffraction peaks can be indexed to those of the anatase phase of titania (Reprinted (adapted) with permission from Ref. [101]. Copyright (2003) American Chemical Society.)

Full size image

Doping in TiO2 Nanostructures

As a wide band gap semiconductor, TiO2 can only utilize the ultraviolet light in solar spectrum. In order to extend its optical absorption to visible light region, doping metals or nonmetals into TiO2 is well studied. Doping TiO2 will mainly maintain TiO2 crystal structure, but generate some favorable changes in their electronic and optical properties [12]. Typically, substitution of Ti4+ cations in TiO2 with other transition metals is much easier than replacing O2− anions with other anions due to the differences in charge states and ionic radii.

Three types of methods are typically used to dope TiO2 nanostructures: wet chemistry, high-temperature treatment, and ion implantation on TiO2 nanostructures. Wet chemistry usually realizes the doping by adding metal ion dopants to the titanium precursor, and then the mixed solution undergoes similar processes as the synthesis of pure TiO2 nanostructures. Choi et al. successfully prepared TiO2 nanoparticles doped with 21 different metal ions by sol–gel method and found the presence of metal ion dopants significantly influenced the photoreactivity, charge carrier recombination rates, and interfacial electron-transfer rates [103]. Nagaveni et al. performed a systematic study of W, V, Ce, Zr, Fe, and Cu ion-doped anatase TiO2 nanoparticles by a solution combustion method and found that the solid solution formation was limited to a narrow range of concentrations of the dopant ions [30]. Li et al. synthesized highly crystalline and near monodisperse TiO2 nanoparticles and nanorods doped by Sn, Fe, Co, and Ni ions by well-controlled solvothermal reactions [68]. Anpo et al. prepared TiO2 nanoparticles doped with Cr and V ions with an ion implantation method. Plasma-enhanced CVD, ion beam-induced CVD, and radiation-frequency (RF) thermal plasma are also used to prepare metal-doped TiO2 nanoparticles [12].

Some nonmetal elements such as B, C, N, F, S, Cl, and Br have also been successfully doped into TiO2 nanostructures. Annealing the TiO2 nanostructures under their corresponding gas flow is commonly used to synthesize nonmetal-doped TiO2 nanostructures [12]. For example C- and F-doped TiO2 nanomaterials can be synthesized by heating TiO2 under CO and hydrogen fluoride gas flow, respectively [104, 105]. Hydrolysis of titanium precursors in the solvent-containing dopants has also been reported. N-doped TiO2 nanomaterials have been synthesized by the hydrolysis of titanium precursors in a water/amine mixture, followed by posttreatment of TiO2 sol with amines [39, 106]. S-doped TiO2 nanomaterials were synthesized by mixing TTIP with ethanol containing thiourea [107]. Cl− and Br− co-doped nanomaterials were synthesized by adding TiCl4 to ethanol containing HBr [108].

ZnO

ZnO is a technologically important and environmental-friendly semiconductor with many remarkable properties, such as a direct wide band gap of 3.37 eV, large excitonic binding energy, high electron mobility, large piezoelectric constants, high nonlinear optical coefficients, and radiation hardness. ZnO can be used in many applications including thin film transistors [109], sensors [110], solar cells [38, 111–115], UV photodetectors [116], and piezoelectric power generators [117, 118]. Because the ZnO properties determining the performance of ZnO-derived devices strongly depend on the size and shape, precise control of the morphology of ZnO nanostructures is of importance. Recently, wet chemistry attracted a lot of attention for its advantages such as low cost, low energy consumption, ease of large-scale production, and controllability of the morphology. And a wide variety of crystalline ZnO nanostructures and films were prepared using various solutions containing different chemicals, additives, and seeds or substrates by aqueous solution-based processing. Kawano [119] reviewed various morphologies of ZnO nanostructures and their dependence on preparation conditions including the source chemicals, the role of seeds or substrates, and the presence of organic molecules as a shape modifier. Figure 20.3 shows a schematic illustration of the effects of the guiding parameters. Under a high degree of supersaturation, fine grains are produced through a high rate of nucleation. On the other hand, nanorods elongated in the c direction can be obtained through gradual crystal growth with a low rate of the nucleation under a low degree of supersaturation. The presence of seeds could control the size distribution and crystal orientation. The ZnO nanostructures are basically controlled by tuning the growth rate, growth site, and growth direction. In addition, the degree of the supersaturation, the presence of seeds, and the adsorption of additives are other essential parameters. Many approaches including chemical vapor deposition [120], sputtering [121], sol–gel routes [31], and electrodeposition [46] have also been investigated as the size- and morphology-controlled synthesis method to prepare ZnO crystals.

Fig. 20.3
figure 3

Schematic illustration of the effects of the guiding parameters for the morphological design of ZnO crystals in solutions (Reproduced with permission from Ref. [119])

Full size image

Nanoparticles

Anderson et al. performed the synthesis of ZnO quantum dots (3–6 nm) by a sol–gel method with the addition of LiOH into an ethanolic zinc acetate solution [22]. They also studied the factors that influence the rate of particle growth [22]. Searson synthesized ZnO nanoparticles with different size by precipitation from solutions using Zn(CH3CO2)2 and NaOH in a series of n-alkanols from ethanol to 1-hexanol [122]. ZnO nanoparticles were also been synthesized by ultrasonic irradiation of an aqueous-alcoholic/aqueous-alcoholic-ethylenediamine solution of zinc nitrate and sodium hydroxide [123]. Imai et al. prepared ZnO nanoparticles with size of 30–40 nm by dropping base solutions to Zinc salts solution under 60 °C [124]. Nanoparticles and superstructures of aggregated ZnO nanoparticles were also prepared by surfactant-assisted solvothermal method, where the anions and types of surfactants greatly affected the ZnO structures [125, 126]. Zhang et al. synthesized polydisperse ZnO aggregates by the hydrolysis of zinc precursor in polyol medium at 160 °C [127]. And then, a well-stacked ZnO porous film on FTO (Fig. 20.4) is fabricated by using a drop-cast method. The film is assembled by spherical ZnO aggregates with different diameters, which consists of packed nanocrystallites. Spin coating of ZnO sol combined with calcinations is a common process to prepare thin film of ZnO nanoparticles on various substrates. Magnetron sputtering [121], pulsed laser deposition [17], and spray pyrolysis [128, 129] were also utilized to prepare ZnO nanoparticle films.

Fig. 20.4
figure 4

Morphology and structure of the ZnO aggregate film: (a) SEM image of the cross section of the ZnO aggregate film, (b) SEM image of the top view of the ZnO film consisting of polydisperse aggregates, (c) magnified SEM image of an individual ZnO aggregate, and (d) schematic diagram illustrating the microstructure of aggregated ZnO comprising closely packed nanocrystallites (Reproduced with permission from Ref. [127])

Full size image

Randomly Oriented or Highly Aligned Nanorods

Chemical vapor deposition (CVD) is an efficient method to prepare ZnO nanorods on substrates. The partial oxygen pressure and chamber pressure are parameters that influence the growth mechanism and govern the final ZnO structure [130]. According to the epitaxial crystalline growth mechanism, ZnO nanorods can grow on (100) sapphire at a 30° angle to the substrate [131]. Lee et al. fabricated well-aligned, single-crystalline ZnO nanowires on GaAs substrates by metal-organic CVD [132]. Low-temperature growth routes on fused silica or Si substrates have also been reported [133]. Electrochemical deposition is another technique for achieving uniform and large area synthesis of ZnO nanostructures [46]. Growth of ZnO nanostructures can occur on a general substrate, flat or curved, without any seeds, as long as the substrate is conductive [37]. Under an external electric field, higher nanowire alignment and stronger adhesion to the substrate have been observed [37]. Vayssieres invented an aqueous method without template and surfactant to grow ZnO nanowires and oriented nanorod arrays [134]. Seed growth via a solution-based method is a typical process to prepare ZnO nanoarrays with different diameter, length, and density [114, 115, 135, 136]. ZnO nanocrystal seeds are coated onto cleaned fluorine-doped tin oxide (FTO) substrates firstly. And then the substrates that were immersed into a mixed solution of Zn-precursor, surfactant, and controlled-released bases underwent a chemical bath process to grow ZnO nanoarrays. And calcinations are usually used to improve the crystallinity of the products. Figure 20.5 shows the morphology of ZnO nanoarrays synthesized based on the seed growth method by Yu et al. [114] The average diameter and length of the individual nanorods are ∼ 40 nm and 1 μm, respectively.

Fig. 20.5
figure 5

(a) Top view and (b) cross-section SEM image of nd-ZnO (Reprinted (adapted) with permission from Ref. [114]. Copyright (2010) American Chemical Society.)

Full size image

Nanotower

Growth of ZnO nanocolumns on alumina substrates via oxidation of ZnS in a tube furnace at 950 °C has been reported [137]. The substrates were kept apart from the source at a lower temperature. The synthesized nanocolumns show a layered structure and become gradually thinner. This was attributed to be a result of a gradually decreasing supply of ZnO vapor. Qian et al. [138] developed a novel solution-based method to grow ZnO nanotower on the different substrates including glass, quartz, and polyethylene terephthalate (PET). The towerlike ZnO crystal arrays (Fig. 20.6a–c) were obtained in a reaction mixture solution containing zinc salt, ammonia, ammonium salt, and thiourea. The orientation of these tower like crystals could be controlled by the contents of these reactants at 95 °C. Other methods including CVD [139, 140], MOCVD [15], carbon thermal reduction method [141], and seed growth combined with hydrothermal processes have also been utilized to prepare ZnO nanotower.

Fig. 20.6
figure 6figure 6

FESEM images of the as-prepared towerlike, flowerlike, and tubelike ZnO arrays grown on glass substrate: (a–c) towerlike ZnO with general, better, and best orientation, (d) flowerlike ZnO, and (e, f) tubelike ZnO with high density and low density (Reprinted (adapted) with permission from Ref. [138]. Copyright (2004) American Chemical Society.)

Full size image

Nanoflower

Qian et al. [138] also found that the flower-like ZnO arrays (Fig. 20.6d) can be obtained at 85 °C in the towerlike ZnO reaction solution system. Zhu et al. [142] synthesized flowerlike ZnO by a simple microwave-assisted solution phase approach using an ionic liquid (1-n-butyl-3-methyl imidazolium tetrafluoroborate). Gao et al. [143] performed flowerlike ZnO nanostructures on Si substrate from aqueous solution by the hexamethylenetetramine-assisted thermolysis of zinc-ethylenediamine complex at 95 °C for 1 h. Adschiri et al. [144] identified that a biomolecule, peptide, could recognize and generate ZnO nanoparticles to assemble into a highly ordered flower-type structure. Further researches indicate simply tuning the type and concentration of Zn-precursor/bases used in solution routes, ZnO flowers will be obtained [145–147]. In addition, hydrothermal [27, 148–150], eletrodeposition [33], and CVD methods [139, 151] are also used for the synthesis of ZnO nanoflowers.

Doping in ZnO Nanostructures

By controlling the doping level of ZnO, its electrical properties can be changed from insulator through n-type semiconductor to metal while maintaining optical transparency that makes it unique as a low-cost transparent electrode in flat-panel displays and solar cells to replace In-doped SnO2 [31, 152, 153]. Mn-doped ZnO and Fe-, Co-, or Ni-alloyed ZnO were predicted to stabilize high-Curie-temperature ferromagnetism [154]. P-type doping in ZnO is also possible by substituting either group-I elements such as Li, Na, and K for Zn sites or group-V elements including N, P, and As for O sites [155]. Alloying ZnO with MgO (Eg = 7.7 eV) enables widening of band gap of ZnO and have been considered as a suitable material as barrier layers [156, 157]. However, when alloyed with CdO (Eg = 2.3 eV), it will narrow the band gaps to even the visible spectrum [158, 159]. The doping of ZnO films is usually utilized by CVD [158], MOCVD [14], molecular beam epitaxy [156], atomic layer deposition [160], and sputtering method [157]. Sol–gel process can be used to dope both ZnO film [31] and nanopowder [161]. The thermal decomposition of metal acetylacetonate precursors in a nonoxygen and nonpolar solvent is another method to prepare doped ZnO nanopowders, even the nanoink, a stable ZnO nanoparticle dispersion in the solvent [153].

Nb2O5 Nanoparticles

Nb2O5 is widely used in catalysts, gas sensors, electrochromic devices, and optical filters [162]. Up to now, several processes have been reported to synthesize Nb2O5 nanoparticles. Uekawa et al. prepared 4.5 nm sized Nb2O5 nanoparticles by heating the peroxoniobic acid sol obtained from the peptization of the acid precipitate with the H2O2 aqueous solution [163]. Zhou et al. prepared porous Nb2O5 nanoparticles via a solution-based method combined with calcination without the assistance of any surfactant [162]. Buha et al. synthesized Nb2O5 nanocrystals with size of 18–35 nm using a nonaqueous sol–gel route based on the solvothermal reaction of the corresponding metal chlorides with benzyl alcohol [164]. Recently, Qiao group grew Nb2O5 nanoparticles film by spin-coating the Nb2O5 sol–gel solution prepared by mixing the niobium ethoxide (Nb(OC2H5)5) precursor with ethanol and acetic acid at room temperature and used the layer as a new electron transport layer for double junction polymer solar cells [11]. In particular, the preparation of Nb2O5 layer only takes half an hour and can be done at room temperature in an air atmosphere. When deposited by spin coating, the Nb2O5 sol–gel solution exhibited a high attachability on the polymer–fullerene active layer. Then the spin-coated film got oxidized quickly to form the Nb2O5 layer (Fig. 20.7).

Fig. 20.7
figure 7

Atomic force microscopy (AFM) topography image of (a) MDMO-PPV:PCBM active layer, (b) Nb2O5 ETL deposited on top of the MDMO-PPV:PCBM active layer, and (c) the pH neutral PEDOT:PSS HTL on top of the Nb2O5 layer (Reproduced with permission from Ref. [11])

Full size image

CuO

As a p-type semiconductor, CuO exhibits a narrow band gap (1.2 eV) and shows wide applications in the field of emission materials, catalysts [165], gas sensors, magnetic storage media, electronics, and solar cells.

CuO Nanoparticles

Liu et al. prepared highly stable CuO nanoparticles with size of 2–4 nm in diameter (Fig. 20.8) by heating aqueous Cu(CH3COO)2 and urea solution in the presence of poly[(2-ethyldimethylammonioethyl methacrylate ethyl sulfate)-co-(1-vinylpyrrolidone)] (PQ11) [166]. Wang et al. performed CuO nanoparticles with an average size of ca. 4 nm by microwave irradiation, using copper (II) acetate and sodium hydroxide as the starting materials and ethanol as the solvent [167]. Applerot et al. also reported a sonochemical method to synthesize CuO nanoparticles [168]. Solid-state synthesis [169], green synthesis via starch as surfactant [165], simple precipitation method [23, 170], and thermal decomposition [171] have also been used to prepare CuO nanoparticles.

Fig. 20.8
figure 8

(a) Low, (b) HRTEM images, and (c) their corresponding particle size distribution histogram of the products (inset: the photograph of CuO dispersion in water) (Reproduced with permission from Ref. [166])

Full size image

CuO Hollow Spheres

Hollow nanostructures have been prepared to increase their surface area and reduce density for applications in drug delivery, chemical sensors, photonic devices, and lightweight filler. A variety of methods have been developed to prepare nanostructures with hollow interiors [35, 172, 173]. Zeng et al. found that CuO hollow dandelion-like architectures by a one-pot hierarchical organizing scheme rely primarily on geometric constraints of building blocks [172]. Zhu et al. prepared a CuO hierarchical hollow nanostructure (Fig. 20.9), assembled by nanosheets, in n-octanol/aqueous liquid system through a microwave approach [174]. Controlled experiments revealed that both bubble and interface play key roles in determining the self-assembly process of CuO hierarchical hollow nanostructures. The morphology/size of building blocks and final products could be readily tuned by adjusting reaction parameters. Kong et al. synthesized CuO hollow nanospheres with an average diameter of 400 nm and shell thickness of 40 nm via a thermal oxidation strategy with Cu2O solid nanospheres as the precursor [35]. The formation of CuO hollow nanospheres mainly result from the Kirkendall effect in the temperature-dependent experiments. Park et al. [175] synthesized uniform Cu2O nanocubes by a one-pot polyol process, and the Cu2O nanocubes were converted to polycrystalline CuO hollow nanostructures through a sequential dissolution–precipitation process, by adding aqueous ammonia solutions in air.

Fig. 20.9
figure 9

SEM and TEM images of CuO obtained: (a) low-magnification SEM image, (b) enlarged SEM image, (c) FESEM image, (d) low-magnification TEM image, (e) enlarged TEM image, and (f) HRTEM image (Reprinted from Ref. [174], Copyright (2010), with permission from Elsevier.)

Full size image

CuO Nanowires

Among various metal oxides, CuO has been extensively studied as a p-type metal oxide semiconductor [40, 176]. Jiang et al. described a vapor-phase approach to the facial synthesis of CuO nanowires supported on the surfaces of various copper substrates that include grids, foils, and wires [177]. Hansen et al. introduced and provided details on a large-scale, cost-effective pathway to fabricating ultrahigh dense CuO nanowire arrays by thermal oxidation of Cu substrates in oxygen ambient [178, 179]. The CuO nanowires that are produced by heating copper foil at 500 °C in a pure O2 gas flow for 150 min feature an average length and diameter of ∼15 μm and ∼200 nm, respectively. Umar et al. demonstrated the preparation of a large-scale vertical array of single-crystalline CuO nanowires on different material surfaces [180]. The procedure simply involved a room-temperature liquid–solid growth process of attached CuO nanoseeds on ITO surface in the mixed aqueous solution of Cu(CH3COO)2 and NH3. The field emission scanning electron microscopy (FESEM) image analysis indicated that these nanowires feature uniform size with tiny structures that have diameters and lengths in the range of 10 and 100 nm, respectively, and tend to form a bundle-like structure at the top end of the wires as shown in Fig. 20.10.

Fig. 20.10
figure 10

(a) FESEM image of CuO nanoseed particles prepared using an alcohothermal method for 2 h at 250 °C; (b, c) FESEM images of CuO NWs grown at different growth periods in the mixed aqueous solution of 10 mM Cu(CH3COO)2 and 12 mM NH3 for (b) 2 and (c) 15 h at room temperature (25 °C); (d) high-magnification image of (c). Scale bars are 100 nm (Reprinted (adapted) with permission from Ref. [180]. Copyright (2007) American Chemical Society.)

Full size image

Applications of Metal Oxides in Solar Cells

Organic and organic/inorganic hybrid solar cells have attracted a lot of interest due to their solution-based processing and low cost [181–195]. Figure 20.11 shows the calculated 3D contour plots of polymer LUMO, polymer band gap, and cell efficiency with three typical inorganic acceptors of (a) TiO2, (b) ZnO, and (c) CdSe. It can be seen that the polymer band gaps are almost the same at ∼1.5 eV; however, their LUMO and HOMO energy levels are different depending on what acceptor materials will be used [188].

Fig. 20.11
figure 11

Calculated relationship between polymer band gap (Eg), polymer LUMO, and cell efficiency in single junction hybrid solar cells with three representative inorganic acceptors of (a) TiO2, (b) ZnO, and (c) CdSe. The efficiency was calculated with IPCE = 65 %, FF = 60 % under AM 1.5 with an incident light intensity of 100 mW cm−2. (Reproduced from Ref. [188] with permission from The Royal Society of Chemistry)

Full size image

Titanium Dioxide (TiO2)

Different nanostructures (e.g., nanoparticles, nanotubes, and nanofibers) of TiO2 have been employed for fabricating efficient solar cells. Kwong et al. in 2003 reported the use of TiO2 nanoparticle with poly(3-hexylthiophene) (P3HT) to make a nanocomposite film acting as an active layer [196]. Photoluminescence spectrum of active layer films showed quenching of P3HT emission for all different concentrations of TiO2 which was mixed with P3HT. Cells made from 50 % and 60 % TiO2 concentrations showed improved performance than those made with TiO2 concentrations as 40 % (lowest) and 70 % (highest). Poor performance for lower concentration (40 %) of TiO2 was attributed to recombination of dissociated charges, whereas higher concentration (70 %) of TiO2 led to poor film quality. Solvent selection for mixing TiO2 and P3HT also affected the performance of fabricated cell. Cell fabricated with 60 % TiO2, using xylene as solvent gave best performance. Qiao et al. also used TiO2 nanoparticles to blend with a water soluble polymer (PTEBS) and demonstrated potential to develop environmentally friendly solar cells [82, 197–199]. Yang et al. studied the effect of adding TiO2 nanotubes into P3HT:PCBM blend on device performance in 2010 [200]. Conventional device structure was fabricated with device structure as glass/ITO/PEDOT:PSS/P3HT:PCBM:TiO2/LiF/Al (Fig. 20.12a). Active layer films were prepared by adding TiO2 nanotubes in P3HT:PCBM solution, followed by sonication for 30 min, in order to disperse it uniformly. The solution was then stirred and spin-coated on top of PEDOT:PSS, followed by thermal deposition of LiF (1 nm) and Al (100 nm). Finally the device fabricated was annealed at 150 °C for different time period to optimize morphology of active layer.

Fig. 20.12
figure 12

(a) Device structure of TiO2 nanotube-based solar cells, (b) optical image of fabricated solar cell surface taken at × 500, (c) SEM micrograph of TiO2 nanotubes, and (d) TEM micrograph of TiO2 nanotubes (Reproduced from Ref. [200] with permission from The Royal Society of Chemistry)

Full size image

From Fig. 20.12b, it was observed that TiO2 nanotube aggregates are made from many individual nanotube each having diameter of ∼10 nm. The TiO2 aggregates as a whole had diameter of 300–600 nm, which was much larger than the films made from P3HT:PCBM (140 nm). The surface roughness of active layer film therefore increased after adding TiO2 nanotube aggregates. J-V curves of the control sample without TiO2 aggregates annealed for 9 min showed an efficiency of 2.8 %, whereas device with TiO2 aggregates annealed for 5 min showed an efficiency of 3 % with an open circuit voltage (Voc) of 0.68 V, short circuit current (Jsc) of 0.60, and power conversion efficiency (PCE) of 3.2 %.

Liao et al. reported hybrid solar cells made from nanostructured arrays of TiO2/P3HT in 2012 [201]. The hybrid cell was modified by introducing Z907 and D149 dye molecule showing enhancement in device performance. Introduction of dye molecules enhanced light absorption and improved the TiO2 nanostructure/P3HT interface. Dye molecule acted as a bridge to link TiO2 and P3HT, where the dye carboxylic group links with hydrophilic TiO2 and the dye aryl group connects with hydrophobic P3HT. Rutile TiO2 nanorod (NR) nanoarrays and nanodendrite (ND) structure were used with P3HT to fabricate hybrid solar cells. TiO2 NR and ND structures were modified by treating it with TiCl4 to form nanoparticles on NR and ND surface, thus increasing the surface area. Figure 20.13 shows top and cross-sectional SEM micrograph of TiO2 nanorods treated with TiCl4. TiO2 NR and ND structure act as electron acceptor and provide transport pathway in TiO2/P3HT-based hybrid cell. Two types of device were fabricated: one without any dye modification and the other with dye attached onto both TiO2 NR and TiO2 ND.

Fig. 20.13
figure 13

(a) Top view and (b) cross-sectional TiO2 nanorod SEM images of TiCl4 treated TiO2 NR (Reprinted (adapted) with permission from Ref. [201]. Copyright (2012) American Chemical Society.)

Full size image

J-V characteristics of non-modified TiO2 NR/P3HT solar cells showed poor device performance with efficiency of 0.43 %. The Z907 and D149 dye modified cells exhibited higher efficiency at 0.94 % for Z907 and 1.58 % for D149. The devices were made by using 650 nm thick TiO2 nanorod arrays. Figure 20.14 shows device performance without and with dyes attached onto TiO2 as surface modifier.

Fig. 20.14
figure 14

J-V plot for non-modified, Z907 dye modified, D149 dye modified TiO2 NR array-/P3HT-based hybrid cell (Reprinted (adapted) with permission from Ref. [201]. Copyright (2012) American Chemical Society.)

Full size image

Further enhancement in solar cell performance was observed by using 1.5 μm thick TiO2 NR array and 1.5 μm thick ND array with D149 dye modification. J-V curves showed a significant performance improvement with increasing thickness of TiO2 nanostructured array with TiO2 ND (1.5 μm)–P3HT-based cell showing highest power conversion efficiency at 3.12 % (Fig. 20.15).

Fig. 20.15
figure 15

J-V plots of TiO2 NR array (650 nm thick)/P3HT modified with D149 dye molecules, D149 modified TiO2 NR array (1.5 μm thick)/P3HT, and D149 modified TiO2 ND array (1.5 μm thick)/P3HT (Reprinted (adapted) with permission from Ref. [201]. Copyright (2012) American Chemical Society.)

Full size image

Xu et al. in 2012 studied charge transfer in P3HT–TiO2 nanorod composite by photoluminescence quenching [83]. Three different samples were prepared: In first sample P3HT was chemically linked with TiO2 nanorod (nr) by a linker triethoxy-2-thienylsilane (P3HT–Si–nr–TiO2), second sample was composite of P3HT and TiO2 (P3HT/LA–nr–TiO2), made by physical mixing of P3HT and linoleic acid (LA) capped TiO2 nanorods, and third sample was pristine P3HT itself. Steady state photoluminescence quenching spectra of all three samples were recorded (Fig. 20.16). Photoluminescence quenching was observed in P3HT chemically linked with TiO2 nanorods, which was attributed to charge transfer occurring between P3HT and TiO2 nanorods, whereas physically mixed P3HT–TiO2 nanorods showed overlap of PL spectrum with pristine P3HT spectrum, indicating no charge transfer occurring in P3HT–TiO2 composite.

Fig. 20.16
figure 16

Photoluminescence spectra of chemically linked P3HT–Si–nr–TiO2, physically mixed P3HT/LA–nr–TiO2, and pristine P3HT. The measurement was taken at excitation wavelength of 420 nm (Reproduced from Ref. [83] with permission from The Royal Society of Chemistry)

Full size image

ZnO

ZnO, an n-type material, is used as electron acceptor along with electron-donating polymer in fabricating organic–inorganic hybrid solar cell. ZnO acts as electron transport layer with good electron mobility and is solution processable and environmentally stable. Beek et al. in 2006 reported hybrid solar cells made from blend of ZnO nanoparticles and P3HT with efficiency ∼ 0.9 % [7]. Hau et al. in 2008 reported inverted polymer solar cell utilizing ZnO as electron transport layer [202]. Charge collection in inverted solar cell is reverse of regular device structure, in which electrons are collected at front electrode, and holes are collected at back metal electrode. Regular device with structure of ITO/PEDOT:PSS/P3HT:PCBM/LiF/Al and inverted cell with structure of ITO/ZnO/P3HT:PCBM/PEDOT:PSS/Ag were fabricated and tested. Inverted P3HT:PCBM blend (1:0.6 by weight) with a film thickness of 200 nm achieved an efficiency of ∼3.5 %. Regular device structure gave an efficiency of ∼2.4 %. Fill factor for both device structures was comparable, but Jsc and Voc were higher in inverted devices, partly attributed to high interfacial area provided by ZnO nanoparticle layer.

Stability studies of conventional and inverted device were carried out by keeping them in ambient condition and measuring their performance for 40 days. In regular device, it was observed that cell efficiency reduced to less than half of its original value in 1 day and got totally degraded after 4 days (Fig. 20.17a). On the other hand, inverted device was found to be highly stable for 40 days (Fig. 20.17b). All the device parameters remained comparable except Jsc, which slightly decreased. In inverted cells, silver electrode can form silver oxide layer in air and thus increase its work function to −5.0 eV. The work function then matches better with HOMO of PEDOT:PSS that is −5.1 eV. Inset of Fig. 20.17a shows the diode behavior in dark condition. The current density at 2 V on the J-V curves of fresh regular solar cells is 2 orders higher that of solar cells left in air for 4 days. However, inverted device shows a much higher stability with Voc and FF remaining almost the same, while Jsc decreases slightly. Their dark condition J-V curves are also comparable as shown in inset of Fig. 20.17b.

Fig. 20.17
figure 17

J-V plots of (a) regular solar cells characterized in ambient condition over a period of 4 days and (b) inverted solar cell tested in ambient condition over a period of 40 days (Reprinted with permission from Ref. [202]. Copyright (2008) AIP Publishing LLC.)

Full size image

In 2013, Ka et al. studied the ZnO annealing effects on performance of inverted solar cells [203]. Devices fabricated with ZnO annealed at a low temperature of 80 °C exhibited an efficiency of 3.6 % that is comparable to devices processed at higher annealing temperatures. Several characterization results concluded that the optimum efficiency at low temperature annealing was attributed to improvement in band energy alignment, crystallinity, and relative ZnO amount. Device structures and their J-V curves under different ZnO annealing temperatures are shown in Fig. 20.18. For devices with annealing temperatures (Ta) at 25 °C and 60 °C, J-V curve shows S-shape in nature, indicating inefficient and/or unbalanced electron and hole transport to electrodes, therefore leading to cell efficiency (η) lower than 0.2 %. When the annealing temperature increased to 70 °C, the J-V characteristics improve a lot, leading to a Jsc of 9.7 mA/cm2, Voc of 0.59 V, FF of 0.49, and η of 2.6 %. When further increased to 80 °C or above, device efficiency increased to 3.61 % with a higher fill factor at 0.61. Further increase in ZnO annealing temperature (Ta) to 90 °C and 150 °C resulted in similar device performance as observed for Ta = 80 °C (Fig. 20.18).

Fig. 20.18
figure 18

J-V plots of inverted polymer solar cells with ZnO as electron-selective layer, treated at different annealing temperatures. Device structure of inverted polymer solar cells is also shown in inset (Reprinted from Ref. [203]. Copyright (2013), with permission from Elsevier.)

Full size image

Xu et al. in 2013 fabricated organic–inorganic hybrid solar cells and studied donor–acceptor interface by Kelvin probe force microscopy [189]. Cell structure consisted of ITO/ZnO/P3HT/Ag, where two different structures of ZnO (nanoridges and nanorods) were used as active layer for fabricating two different cells. P3HT was spin casted on ZnO films, followed by annealing at 150 °C, for 10 min. Topography and interface studies of ZnO/P3HT films were studied by atomic force microscope (AFM) and Kelvin probe force microscopy (KPFM). Topography of ZnO nanoridge films showed wrinkled shaped morphology which was attributed to stress relaxation caused by slow drying process of film (Fig. 20.19). Surface potential (SP) across the interface of P3HT/ZnO was studied by KPFM, which showed higher SP value for P3HT and lower SP value for ZnO nanoridge region (Fig. 20.19). Surface potential of the interface was measured under dark and illuminated condition. A 0.1 V difference in SP was seen under dark condition and 0.25 V difference under illuminated condition and was attributed to light response of P3HT and ZnO (Fig. 20.19). P3HT shows better light response in comparison to ZnO; therefore, P3HT exhibited higher surface potential than ZnO, allowing a favorable path for electron transfer to ZnO.

Fig. 20.19
figure 19figure 19

(a) AFM image showing ZnO nanoridge topography; surface potential image of P3HT/ZnO film in (b) dark condition and (c) illuminated condition; 3D SP image of P3HT/ZnO film in (d) dark condition and (e) illuminated condition; cross-section profile of SP in the film (Reproduced with permission from Ref. [189])

Full size image

Nb2O5

Nb2O5 is new in the list of oxides, towards which researchers have been showing interest to use as electron transport layer for solar cells. In 2011, Wiranwetchayan et al. reported the use of Nb2O5 in fabricating inverted solar cells [204]. A thin film of Nb2O5 was introduced in between fluorine-doped tin oxide (FTO) and active layer (P3HT:PCBM) with device structure as glass/FTO/Nb2O5/P3HT:PCBM/PEDOT/Ag. The sol of Nb2O5 was used for spin coating a thin film of Nb2O5 on FTO, followed by annealing at 450 °C for 1 h to make it crystalline.

Photovoltaic response of the devices without and with Nb2O5 was studied. The devices without Nb2O5 showed no photovoltaic response; however, devices having a thin film of Nb2O5 exhibited an efficiency of 2.7 %. Dependence of film thickness on device efficiency was also studied. The efficiency decreased to zero with increasing thickness. This was attributed to the Nb2O5 film thickness increasing more than tunneling path length. As conduction band of Nb2O5 is higher than the LUMO of PCBM, therefore it shows little probability of electron transfer from PCBM to Nb2O5. However, Nb2O5-based device showed an efficiency of 2.7 %, suggesting that electron transport to electrode takes place through tunneling. Therefore, as film thickness increased, tunneling distance also increased, thus inhibiting electron transport to the electrode.

Later in 2012, Siddiki et al. reported use of Nb2O5 as electron transport layer in double junction polymer solar cell [11]. Double junction polymer solar cell was fabricated with structure shown in Fig. 20.20a. Nb2O5 was used as an interfacial layer for electron transport between front and back cell. 25 nm thick Nb2O5 film showed >90 % transmittance in the visible region. J-V characteristics of double junction cell and individual cells are shown in Fig. 20.20b. Device efficiency of double junction cell was found to be 0.8 % with a Voc of 1.3 V, which was higher than individual front and back cells. The Jsc of double junction solar cell is less than the individual cells. The series resistance was also found to be increased in the double junction cell. The decrease in Jsc and increase of series resistance were possibly caused by two factors: (1) mismatch in current between front and back cells and (2) barrier formed at the interfacial layer between the two cells. The higher Voc indicated Nb2O5 can be used as electron transport layer in double junction polymer solar cells. Simulation results showed that device performance using Nb2O5 could be improved by current matching of individual cells and reducing series resistance.

Fig. 20.20
figure 20

(a) Device structure of double junction polymer solar cells and (b) J-V plots of front cell, back cell, double junction solar cell (Reproduced from Ref. [11], with permission from the PCCP Owner Societies.)

Full size image

CuO

Copper oxide has been recently used as interfacial layer to improve the stability in polymer solar cells. Wang et al. in 2011 fabricated polymer bulk heterojunction solar cells by introducing CuOx layer in between cathode and P3HT:PCBM [10]. Two device structures were fabricated: one with a thin layer (∼5 nm) of CuOx which was deposited by thermal evaporation in a structure of ITO/PEDOT:PSS/P3HT:PCBM/CuOx/Al and the other was a bilayer of CuOx and LiF in a device structure of ITO/PEDOT:PSS/P3HT:PCBM/LiF/CuOx/Al and ITO/PEDOT:PSS/P3HT:PCBM/CuOx/LiF/Al.

The thickness of LiF was 0.6 nm. The devices without CuOx interfacial layer showed an efficiency of 1.5 %, while that using CuOx as interfacial layer exhibited a similar efficiency but a higher Jsc and lower Voc. Decrease in Voc was attributed to the high work function of CuOx in comparison to aluminum. On the other hand, devices fabricated using a bilayer of CuOx/LiF showed improvement in Voc. Introduction of CuOx improved device stability, but did not show improvement in device performance. Device stability was monitored by keeping the devices made without and with CuOx in air. Devices without CuOx layer showed ∼50 % efficiency decrease after 3 h in air, which was mainly due to decrease in Jsc. The devices made using CuOx/LiF exhibited a 40 % efficiency increase, which was probably due to increase in Voc and FF when kept in air. The increase in Voc and FF was attributed to the oxidation of CuOx in air, leading to increase in shunt resistance and reduction of quenching of excitons and charge carriers occurring at the metal cathode. Device with CuOx layers retained ∼80 % of the initial efficiency after 240 h and ∼50 % after 500 h. Therefore, it was concluded that use of LiF and CuOx improved charge collection and increased Voc and FF of devices.

Conclusion

General synthesis methods of common nanomaterials including TiO2, ZnO, Nb2O5, and CuO with 0D, 1D, 2D, and 3D hierarchical nanostructures and their applications in solar cells have been reviewed in this chapter. Although laboratory-based synthesis methods have been extensively developed, there are still few methods that can realize large-scale, low-cost, and environmental-friendly synthesis processing, especially for the synthesis of some specific structures, such as TiO2 and ZnO nanoarrays. However, applications usually require metal oxides have special structures with aimed properties. For TiO2 and ZnO films used in polymer-inorganic hybrid solar cells as efficient electron transport layer, high crystallinity and surface area are required to facilitate charge separation and transport. Thus, design of ideal structures with large-scale synthesis capability is one of the largest missions of inorganic synthetic chemistry. Nanostructured metal oxides have played an important role in improving polymer-inorganic hybrid solar cell performance. These inorganic metal oxides possess high charge mobility, solution processability, and are environmentally stable. Therefore introduction of metal oxides into solar cells has led to improvement in performance of polymer-inorganic hybrid solar cells. Nanostructured metal oxides have brought new design and properties in fabricating solar cells.