Chapter 1 Introduction
1.2. Motivation
Cu2O is a promising material for applications in photovoltaic devices since its theoretic solar energy conversion efficiency was estimated up to 18%. However, the highest conversion efficiency reported so far is a modest 2% [32]. This reduced value is attributed to the poor crystallinity in conjunction with numerous defects present in the Cu2O structure.
In this research, we electrodeposited Cu2O films under different pH values, potentials, and deposition times. Next, we attempted to im prove the photoelectrochemical performance of for hydrogen generation by a post-annealing process with the aim to reduce the inherent defects. Afterward, we recorded the properties for solar energy conversion using a simple photoelectrochemical measurement in hydrogen evolution.
Figure 1.1 A schematic of crystal structure of cubic Cu2O. The dark atoms are O and the remaining ones are Cu atoms.
Chapter 2
Literature Review
2.1. Fabrication of Cu
2O
2.1.1. Method of Chemical Synthesis
Fabrication of semiconducting nanoparticles has attracted significant attention these years for their unique chemical and physical properties. Various routes have been developed to synthesize narrow-sized, well-distributed, highly crystalline particles with a variety of morphologies. We are interested in the Cu2O due to its self-assembling character, and its potential application in solar energy conversion. A typical chemical synthetic route for the Cu2O listed below;
)
Copper ions often form copper hydroxide first in an alkaline electrolyte, and then reduced by a reducing agent (L-ascorbic acid, hydrazine). There are many factors that might affect the crystal growth of Cu2O. For example, a well-known determinant is the surfactant. Others variables including pH, processing step, and reducing agent also play certain roles. So far, various Cu2O structures including spheres, cubes, octahetras, and wires have been synthesized [2-22]. A brief summary on their synthesis work is provided below.
A. Nanosphere
The Cu2O exhibits a direct band-gap of 2 eV. It is known that the quantum size effects on the optical properties of indirect band-gap semiconductor are significantly different from that of direct band-gap one. Previously, Yin et al. were interested in the size effect of Cu2O on its optoelectronic properties [2]. In 2005, they synthesized
The transition was observed to reveal a blue-shift as a function of decreasing particle size.
Similarly, Li et al. synthesized the Cu2O nanospheres using copper acetate as a precursor, and NaBH4 as a reducing agent in a DMF solvent [3]. The reaction temperature was about 80~90 ℃ and the resulting Cu2O nanospheres were in diameter of 200 nm. In following gas sensor application (Fig 2.2), their results indicated that the Cu2O nanospheres were a better gas sensor for flammable gas than CuO and Cu2O in octahedral shapes at a reaction temperature of 210 ℃. It was concluded that the surface of Cu2O was converted to Cu2O2-x, an active state in high temperature that makes it more active than CuO. Moreover, it is likely that the stacking arrangement of spherical structure engendered a larger surface area that delivers a higher performance.
Figure 2.1 a) TEM images of self-assembled 6 nm diameter Cu2O nanocrystals, b) synthetic procedure for Cu2O nanocrystals [2].
Figure 2.2 SEM images of Cu2O nanocrystals and gas sensor system [3].
B. Nanocubes
In 2003, Murphy et al. synthesized the Cu2O nanocubes using polyethylene glycol (PEG) as a surfactant [4]. They discussed the effect of synthesis steps and surfactant concentration. Their results indicated the procedure which mixed copper sulfate and PEG600 first, followed by addition of NaOH and ascorbic acid was leading to desirable uniformity and monodispersity. In addition, the cubic size was proportional to the surfactant concentration in a range of 25 to 200 nm.
Hung et al. adopted a seed-mediated synthesis approach to control the size of Cu2O nanocubes in 2007 [5]. The surfactant employed was sodium dodecyl sulfate (SDS). The cubic size was controlled by aggregation of Cu2O seed particles, and the size range could be adjusted from 40 to 420 nm with a rather high yield (Fig 2.3).
Chang et al. synthesized nanocubes of 28 nm in edge length by a simple one-pot route in 2008 [6]. They used fructose as both the stabilizer and weak reducing agent, which would first form Cu-fructose complexes. Subsequently, the complex was reduced to CuOH-fructose complexes at a higher pH. Then the complex was reduced further by L-ascorbic acid to form Cu2O nanocubes. They observed that small Cu2O nanocubes exhibited a hollow structure at first, and then the hollow structure was filled when the reaction time progressed longer. The Cu2O nanoparticles filled the holes through typical Ostwald ripening behavior and relevant images shown in Fig 2.4
.
Figure 2.3 A schematic illustration of the procedure used to grow Cu2O nanocubes and related crystal growth process [5].
Figure 2.4 A schematic illustration of the formation of Cu2O nanocubes from hollow to filled structures and corresponding SEM images [6].
C. Octahedrons
Shen et al. prepared the monodispersed Cu2O octahedron nanocrystals with size varying from 45 to 95 nm (Fig 2.5) [7]. They reduced the copper nitrate in a Triton X-100 water-in-oil microemulsion by γ-irradiation. The Triton X-100 provided a stable environment allowing the reaction to take place without any pH adjustment.
Wang et al. synthesized the Cu2O octahedron without adding surfactants or the assistance of organic compounds [8]. They found that adding NH3 solution would affect the ratio for the growth rate along the <111> versus the <100> direction, which influenced the morphology of the product. The morphology was controlled by adjusting the ratio of NH3 to Cu2+, and distinct shapes such as spherical, porous spherical, cubic, and octahedral were obtained.
Figure 2.5 SEM image and TEM image of octahedral Cu2O reduced by γ-irradiation in Triton X-100 [7].
D. Nanowires, Nanotubes
Wang et al. used a simple chemical route to synthesize the Cu2O nanowires with a rather high yield [9]. The PEG20000 was used as a surfactant, and hydrazine was
used as a reducing agent. The reaction proceeded under constant stirring for entire time. The nanowires were 10 to 20 µm long, and their diameters were about 5 to 8 nm (Fig 2.6). Hence, the nanowires exhibited a high aspect ratio.
Hu et al. used cetyltrimethylammonium bromide (CTAB) as a surfactant [10].
The Cu(OH)42- precursor was selectively reduced by various methods to synthesize Cu, Cu2O, and CuO nanotubes. The Cu and Cu2O were reduced by a hydrazine and glucose respectively, and the CuO was reduced by a hydrothermal method. As the concentration of Cu(OH)4
was increased, the nanotubes became nanorodes with a solid structure. The TEM images for nanotubes and nanorodes of Cu, Cu2O, and CuO are shown in Fig 2.7.
Figure 2.6 TEM images of Cu2O nanowires [9].
Figure 2.7 TEM images of a) nanotubes and b) nanorodes [10].
2.1.2. Method of Electrodeposition [11-12]
There are many methods to prepare Cu2O thin films. For example, physical vapor deposition (e.g., thermal evaporation, sputtering), chemical vapor deposition (CVD), and electrochemical deposition have been studied extensively. Among them, the electrodeposition is the most attractive approach. Electrodeposition is a versatile and low-cost technique for preparing thin films of oxide semiconductors. It is possible to grow uniform thin film over large areas in unique shape. Fabrication of Cu2O films by the electrochemical method has been successfully developed.
A. Cathodic Electrodeposition in Cuprous Oxides
Cathodic deposition of Cu2O has been extensively studied before. It is established that electrochemical deposition is able to adjust precisely the driving force for the reaction involved in deposition. This allows the control of structure and phase composition of the resulting films. Many different electrodeposition environments, substrates, and power supply methods have been explored. It is recognized that changing these conditions would alter structure, morphology, and photoelectro- chemical properties of Cu2O.
A commonly used bath was developed in 1987 by Rakhshoni et al, in which copper sulfide, lactic acid, and sodium hydroxide were used [12]. The Cu2O films were electrodeposited successfully under galvanostatic conditions on stainless steel cathodes. It was shown that uniformly oriented Cu2O films could be grown with robust adhesion.
The pH environment is recognized to critically influence the principal reaction route of electrodeposited Cu2O film. The reactions steps involved for the cathodic
deposition of Cu2O are [13]; dissolve to copper(II) or reduce to copper metal in an acid environment.
In 1998, Switzes et al. identified a stable ratio of copper lactic solution, which contained 0.4 M cuprous sulfate and 3 M lactic acid [13]. The pH for the plating bath could be adjusted from 7 to 12. The copper ion would be stabilized by the lactic acid to form Cu(CH3CHOHCOO)2. As a result, it would not precipitate out in high pH environments. Potentiostatic and galvanostatic depositions both confirmed that the preferential orientation of Cu2O film was [100] in a pH 9 bath, and [111] in a bath whose pH value was above 10 (Fig 2.8).
Cu2O had been electrodeposited on several substrates including Cu, InP (001), ITO, Si, and stainless steel [14-15]. The out-of-plane orientation of film deposited on a polycrystalline substrate depended on the solution pH, and was only slightly affected by the single crystal substrate [16].
Wang et al. carefully examined the effect of pH value on the resulting crystal growth [17]. They observed a new preferred orientation, which is the (110) of Cu2O.
They concluded that the film could be obtained in a narrow pH range from 9.4 to 9.9 (Fig 2.9).
Lee et al. electrodeposited the Cu2O film in a weak acidic electrolyte of 10 mM Cu(NO3)2 by a constant current mode [18]. Their results confirmed that the Cu2O and
Cu were deposited simultaneously. Formation mechanism of Cu2O was investigated by electrochemical analysis utilizing an electrochemical quartz crystal microbalance (EQCM). The superimposition of anodic current and mass change data suggested that the Cu exhibited a lower dissolution overpotential. Therefore, they supplied an anodic current to selectively dissolve the Cu, and obtained a pure Cu2O film afterward. Their results provided a different preparation route for pure Cu2O phase.
Figure 2.8 Plot of relative intensity (I(200)/I(111)) and grain size as a function of bath pH, the applied potential is -0.4 V vs. SCE, and the bath temperature is 60 °C on a stainless steel substrate [13].
Figure 2.9 XRD of (110) oriented Cu2O film and corresponding SEM image [17].
B. Morphology Controlled by Electrodeposition
In 2002, Huang et al. synthesized the Cu2O nanowires by electrodeposition from a lyotropic reverse hexagonal liquid crystalline phase, which was used as a soft template [19]. The liquid crystalline phase was polarized, and aligned in an electric field. Its alignment could be further improved by controlling the distance between the electrodes. The deposited Cu2O nanowires had a diameter ranging from 25 to 100 nm with a high aspect ratio (Fig 2.10).
In 2004, Choi et al. pointed out that the shape of a crystal was determined by the crystallographic planes on its surface [20]. They used an additive, sodium dodecyl sulfate (SDS), in order to tailor the crystal habits of electrochemically grown Cu2O crystals. The SDS was preferentially adsorbed onto {111} faces, and impeded the
crystal growth alone the <111> direction. Its growth process is shown in Fig 2.11.
Furthermore, the preferential adsorption of SDS was pH-dependence, which enabled selective tuning for the growth rate of Cu2O crystals along the <111> directions.
In 2005, they further demonstrated a systematic and simultaneous tuning of the habit and degree of branching in the Cu2O crystals by manipulating certain key conditions, such as electrodeposited voltage, current, temperature, and composition of the solution [21]. The deposition potential-current diagram, as shown in Fig 2.12, summarizes the effect of electrochemical conditions on branching and faceted growth of Cu2O crystals. The crystal growth could be precisely controlled by adjusting the preference for branching or faceting growth conditions through suitable electrodeposition methods. In such way, the crystal shape could be designed and reproduced.
Adding different additives engendered a distinct preferential adsorption effect.
The additives include surfactants, polymers, and specific ions. Choi et al. had studied the pre-grown crystals carefully in 2006 [22]. Different ions exhibited various adsorbing strengths. By adjusting the additives, the growth process could be totally reversed (Fig 2.13). Wang et al. studied the additives effect by adding different amounts of CTAB and cations. The structure for the bulk Cu2O film was also affected by the additives (Fig 2.14) [23].
Figure 2.10 a) Structure of surfactant AOT molecule, and b) SEM image of electro- deposited Cu2O wires [19].
Figure 2.11 a) The scheme of crystal-habit control achieved by preferential orientation adsorption of additives during the crystal growth process and b) crystal shape evolution from cubic to octahedral [20].
Figure 2.12 a) A deposition potential–current diagram summarizing the effect of electrochemical conditions on branching (triangle) and faceting (diamond) growth and b) SEM images of branched Cu2O crystals with varying crystal habits [21].
Figure 2.13 a) SEM images showing the transformation of pre-grown cubic Cu2O crystals over time in a 0.02 M Cu(NO3)2 solution containing 0.17 M (NH4)2SO4 and b) in a 0.02 M copper nitrate solution containing 0.17 M SDS and 0.004 M NaCl [22].
Figure 2.14 SEM images of Cu2O micro-nanostructures deposited on ITO substrates from electrolyte containing 0.02 M Cu(Ac)2, 0.1M NaAc and CTAB with different concentration; a) 0, b) 0.4, c) 0.8, and d) 2.8 mM [23].
2.2. Photoelectrochemistry
2.2.1. Fundamentals of Semiconductor Electrochemistry and Photoelectrochemistry [24-25]
Photoelectrolysis is a vast field that covers a variety of compound formation.
Among them, the photoelectrolysis of water is one of the most attractive subjects.
Because global consumption of energy is rapidly increasing over the years, finding the supply to meet this rising demand is a critical task in the future. Hydrogen offers a great potential for the environmental and energy supply infrastructure. Typically, it can be produced from hydrocarbons and water splitting. In order to drive this
thermodynamically uphill reaction with minimized fossil energy input, the solar energy is considered to be an ideal source.
There are several requirements for any system intended for converting and storing solar energy. First, sunlight must be efficiently absorbed to produce electrons in the excited states within the light-absorbing material. Second, to obtain desirable work either in chemical or electrical form, the photoexcited electrons and their associated vacancies must be separated in space to prevent their recombinations, which produces heat and wastes energy. Third, the photoexcited charges must be energetically and kinetically sufficient to perform a specific chemical reaction, for instance, splitting the water into H2 and O2. Furthermore, these charges must not produce unintended end-products. Lastly, the stability for a photocatalyst needs to be reasonably acceptable. Satisfying all of these requirements simultaneously is a tall order.
A photoelectrochemical cell is used to produce H2 from water electrolysis under sunlight. The possibility of producing H2 using UV light in a photoelectrochemical cell was first demonstrated by Fujishima and Honda in 1972 using a semiconductor material such as TiO2 (Fig 2.15) [24]. Since then, a series of studies have been initiated for the photoelectrolysis of water. A popular approach is to use the semiconductor as the light absorber. Semiconducting solids generally have broad, strong optical absorption characteristics, and effective charge separation is facilitated by electric fields at the interface between the semiconductor and liquid electrolyte.
Presented in Fig 2.16 are several semiconductor-based systems proposed for solar water splitting [26]. The simplest one is a semiconductor-electrolyte cell, which is also named the photoelectrochemical solar cell (PEC).
A semiconductor typically exhibits a band gap (Eg) in the 1-4 eV, which has an important bearing on its optical response. Fig 2.17 presents the band gap for several
semiconductors in contact with an aqueous electrolyte at pH 1. The band gap reflects the solar energy that a semiconductor could absorb and is the indicator for the driving force for subsequent reactions. The other important factor is the relative positions for the participating energy levels in the semiconductor and solution.
The study of semiconductor-electrolyte interfaces has both fundamental and practical motivations. The important point that distinguishes the semiconductor- electrolyte interface from the metal-electrolyte and metal-semiconductor interface is apparent. For a metal, the charge, and thus the associated potential drop, is concentrated at the surface penetrating at most a few Å into the interior. The metal can not support an electric field within itself. Thus, when a metal electrode contacts with an electrolyte, almost all the potential drop at the interface occurs within the Helmholtz region in the electrolyte phase. On the other hand, when a semiconductor is immersed in the same electrolyte, equilibration at the interface requires the flow of charge from one phase to the other and a “band bending” occurs within the semiconductor phase. The net result of equilibrium is that EF =EF,redox and “built-in”
voltage, VSC, develop within the semiconductor phase, as illustrated in Fig 2.18. After equilibration, the Fermi levels are the same on either side of the interface.
The band bending phenomenon is by no means unique to the semiconductor -electrolyte interface. This layer is the space charge region or the depletion layer, which affects the direction of charge flow. We can see the distinct bending direction between a n-type or p-type semiconductor. Their photoelectrochemical currents flow is opposite to each other. The energy scheme for a cell with one n-type semiconductor electrode for photoelectrolysis of water is shown in Fig 2.19. It illustrates the flow of charge in a photoelectrolysis reaction.
The band gap of Cu2O is about 2 eV. It has attracted substantial attention these years for its capability to splitting water. We summarize some studies on the
photoelectrochemical properties of Cu2O in the next section.
Figure 2.15 a) Fujichima-Honda cell with n-TiO2 photoanode and Pt-cathode, as well as b) a schematic energy level diagram of the cell [24].
Figure 2.16 Schematic diagrams of different types of semiconductor-based systems proposed for solar water splitting; a) solid state photovoltaic cell driving a water electrolyzer, b) cell with immersed semiconductor p/n junction as one electrode, c) liquid junction semiconductor electrode cell, and d) cell with dye -sensitized semiconductor electrode [26].
Figure 2.17 Relative dispositions of various semiconductor band edge positions shown both on the vacuum scale and with respect to the SHE reference. These band edge positions are for an aqueous medium of pH 1 [24].
Figure 2.18 Energy level of a semiconductor-electrolyte interface before equilibrium
(left-hand side) and after equilibrium (right-hand side); a) n-type and b) p-type
semiconductor [24].
Figure 2.19 Energy scheme of a cell with one n-type semiconductor electrode for photoelectrolysis of water. V is stored energy for electrolysis. pEF is Fermi level of photogenerated holes known as quasi-Fermi level, nEF is Fermi level of electron [24].
2.2.2. Photoelectronchemical Properties of Cuprous Oxide
A. Hydrogen Evolution of Cuprous Oxide
In 1998, Domen et al. were first to claim that the Cu2O powder was capable of splitting water under visible light irradiation and its activity was recorded longer than 1900 h (Fig 2.20) [27]. However, the ability of Cu2O was questioned next year by Joung et al. [28]. They used a polycrystalline Cu2O electrode for photoelectrolysis of water but did not measure any cathodic photocurrent under illumination. Nevertheless, it seemed that the Cu2O was able to reduce methylviologen efficiently (Fig 2.21).
Their results implied that the reduction of water was highly unlikely to Cu2O.
Although, the Cu2O is thermodynamically possible to reduce water but it is inefficient
to oxidize water. The conduction band edge of Cu2O is +0.6 V and the oxidation potential of water is +0.57 V at pH 7. As a result, the Cu2O could not supply adequate overpotential for water oxidation. Therefore, Joung et al. concluded that the Cu2O could be a promising material, in conjunction with a suitable redox system as used a p-type photoelectrode in an electrochemical photovoltaic cell, but not for direct photoelectrochemical water splitting.
Rajeshwar et al. used different sacrificial electron donors to promote of Cu2O for water oxidation [29]. They used a two-compartment (Fig 2.22), two-electrode electrochemical cell, to measure the photocurrent. The electrodeposited Cu2O films with or without Ni modification were used as the working electrode. They explored different electrolytes in two compartments which were 0.5 M NaSO4 with 40 mM methyl viologen in cathode compartment and 0.5 M NaSO4 with different sacrificial electron donors in anode. Their results indicated that the hydroquinone at pH 10 delivered the highest performance as the sacrificial electron donor in their system. Ni modification exhibited a better performance, and the mass transfer of methyl viologen species was strongly affecting the resulting reaction. These electron mediators were effective in capturing the photogenerated electrons from the Cu2O before they underwent recombination. Their results implied that after adding suitable electron mediators, the Cu2O could be used as a catalyst for water splitting.
Rajeshwar et al. used different sacrificial electron donors to promote of Cu2O for water oxidation [29]. They used a two-compartment (Fig 2.22), two-electrode electrochemical cell, to measure the photocurrent. The electrodeposited Cu2O films with or without Ni modification were used as the working electrode. They explored different electrolytes in two compartments which were 0.5 M NaSO4 with 40 mM methyl viologen in cathode compartment and 0.5 M NaSO4 with different sacrificial electron donors in anode. Their results indicated that the hydroquinone at pH 10 delivered the highest performance as the sacrificial electron donor in their system. Ni modification exhibited a better performance, and the mass transfer of methyl viologen species was strongly affecting the resulting reaction. These electron mediators were effective in capturing the photogenerated electrons from the Cu2O before they underwent recombination. Their results implied that after adding suitable electron mediators, the Cu2O could be used as a catalyst for water splitting.