Synthesis of solar fuels by a novel photoelectrocatalytic approach

Abstract

The characteristics of nanostructured (a) TiO₂ thin films (based on an ordered array of titania nanotubes) and their performances as photoanode in H₂ production by water splitting or photoreforming of ethanol (in liquid or gas phase) and (b) carbon-nanotube based electrodes for the gas-phase reduction of CO₂ to liquid fuels (mainly isopropanol) are discussed together with their application for the design of a novel photoelectrocatalytic approach for the synthesis of solar fuels.

---

Broader context

The conversion of solar to chemical energy is one of the most attractive routes in the future energy scenario. While in the long term, the use of solar energy will become predominant, in the medium term a better use of renewable resources is needed, including solving the issue of energy storage and transport, and finding a sustainable solution to the CO₂ emissions, because the full transition to non-fossil fuels will require a longer time. In this energy scenario the thermal, photocatalytic, and photoelectrocatalytic reduction of CO₂ under solar irradiation could greatly increase carbon recycling and reduce fossil fuel consumption. Here, we discuss some aspects of the synthesis of solar fuels, using a novel photoelectrocatalytic (PEC) approach. In particular, we present results on the synthesis of solar H₂ by water photoelectrolysis or photoreforming of ethanol, and the conversion of CO₂ to isopropanol.

1 Introduction

The concerns regarding greenhouse gas emissions and fossil fuel shortages together with the increasing fraction of world’s population contributing to a massive use of energy have raised the social pressure to accelerate the transition from fossil to renewable fuels and introduce fundamental changes in the present energy supply system.^1–3

The design and development of future fuels based on a green chemistry approach require the definition of sustainable energy scenarios and effective scientific strategies necessary to address the complex energy and environment issues.⁴ There is the need on a short-term of new technologies for energy saving and efficiency, and use of biomass. On a medium term, a better use of renewable resources is needed, including solving the issue of energy storage and transport, and finding a sustainable solution to CO₂ emissions, because the full transition to non-fossil fuels will require a longer time. Finally, in the long-term, the renewable energy scenario, based in particular on solar energy, will become predominant.^5,6

Therefore, an alternative feedstock for synthesis fuels, from stranded and unconventional fossil fuel reserves to green biomass, represents the short-medium term solution, which integrates with energy saving and an increasing use of renewable energy (wind, geothermal and solar energy, particularly using photovoltaic—PV—cells). However, in a medium-long term perspective, it is necessary to push the role of solar energy in directly producing fuels and recycling CO₂.^7,8 Here, we define as solar fuels the direct production of H₂ from water, and the conversion of CO₂ to fuels using solar energy.^9,10

In fact, PV cells are close to grid-parity, but an issue that is still critical is the storage of electrical energy.^11,12 Therefore, the conversion of solar to chemical energy is an interesting integration to current solar options for storing and transporting solar energy.^13–15 H₂ is a clean energy carrier, but problems still exist for its storage. Its renewable production using solar energy is a target in sustainable energy scenarios and renewable H₂ production would reduce the use of fossil fuels actually necessary for its production.^16–19

On the other hand, the energy density (per unit volume or weight) of gasoline, for example, is larger by far than that possible for H₂ and for electrical energy, even considering future possible developments in storage materials.^20,21 In addition, realistic energy scenarios should consider, besides sustainability, the continuity in usage of existing technologies (and infrastructure) wherever possible. Therefore, the conversion of CO₂ back to fuels using solar energy is also a relevant part of this future energy scenario.^{8–10,14,22–24}

There are two main options in using solar energy to produce fuels. The thermochemical approach uses the high temperatures reached in solar furnaces to produce H₂ (and O₂) from water or CO (and O₂) from CO₂.^25–28 Thermochemical cycles are necessary to lower the temperatures required. An example is the use of a metal oxide that spontaneously reduces at high temperature. The reduced oxide is then oxidized by interaction with H₂O to form H₂ or alternatively by reaction with CO₂ to produce CO. Nevertheless, temperatures above 1200–1400 °C are necessary. This creates a number of issues in terms of materials and stability, cost-effectiveness and productivity. The syn gas (CO/H₂) should then be catalytically upgraded to fuels (methanol, Fischer–Tropsch hydrocarbons). The approach is essentially suitable for solar plants, while it may be difficult to adapt it in a delocalized production of solar fuels.

Delocalization of energy production is an important target for sustainable chemistry and energy⁶ for a better integration with territory and reduction of eco-impact, reduction of fuel transport safety and cost issues, etc.

In the low-temperature approach, solar energy is used by a suitable semiconductor to generate, by charge separation, the electrons and holes, which further react with water and CO₂.^13,19,29,30 The reduction of the latter can be a two-step approach, e.g. generation of electricity in photovoltaic cells, by wind, etc. and then use of the electrons to electrochemically/catalytically reduce carbon dioxide in a physically separate cell. Alternatively, it is possible in a one-step approach by coupling the two processes in a single unit, e.g. photoelectrochemical/catalytic approach. The physical separation of the two reactions of water oxidation and CO₂ reduction in a photoanode and electrocathode, respectively, is necessary to increase the efficiency of the two reactions and limit charge recombination.¹³

The same device can be also used for the production of physically separated flows of H₂ and O₂ during water photoelectrolysis. Moreover, this device could be used to produce renewable H₂ by photocatalytic reforming of chemicals present in waste streams from agro-food or agro-chemical production, such as diluted streams of ethanol, glycerol, etc.³¹ Many diluted waste streams containing ethanol and other organics are too diluted to be used (i) as feed to produce H₂ by catalytic conversion, (ii) for producing methane by anaerobic digestion, or (iii) as feed in fuel cells. The photoreforming of these waste streams to produce H₂ is thus an interesting option.⁵

1.1 The photoelectrocatalytic (PEC) reactor

The separation of a photoinduced process in two physically distinct areas related to water oxidation (to form oxygen, protons and electrons) and proton reduction (to form H₂) or CO₂ reduction (e.g. the photoelectrochemical—PEC—approach) shows many potential advantages as discussed in the previous section.

However, for a practical use of PEC solar cells, the design of the cells should be quite different from that used commonly in literature based essentially on slightly-modified conventional electrochemical cells operating in the liquid phase.³¹

The anode and cathode in the PEC device should be in the form of a thin film separated from a proton-conducting membrane (Nafion® for example, but other membranes could be used) and deposited over a porous conductive substrate which allows the efficient collection/transport of the electrons over the entire film as well as the diffusion of protons to/from the membrane. It is also necessary to allow an efficient evolution of the gas (O₂ and H₂ on the anode and cathode sides, respectively, in the case of water photoconversion), or of oxygen on the anode side, and efficient diffusion of CO₂ and evolution of the products of reaction on the cathode side (for CO₂ photoreduction). The reactor geometry and the spatial relation between reactor and light source are also important, as well as the efficient control of the temperature to avoid overheating during operations.³¹

On the cathode side, gas phase operations are preferable particularly for CO₂ reduction to avoid the problem of gas cap over the electrode, limited solubility of CO₂ and changing the type of products formed.

Electrochemical utilization of CO₂ has been studied for many years, as reviewed recently.^32,33 There are two main approaches, depending on whether the conversion of CO₂ is studied in aqueous or non-aqueous solutions. Formic acid is the main reaction product in the electrolysis of aqueous solutions of CO₂, because the carbon dioxide anion radical easily forms by addition of one electron to CO₂ on the electrode surface. Desorption of this product is faster than its consecutive reduction. In addition, a problem in the utilization of CO₂ in aqueous solution derives from its low solubility in water at standard temperature and pressure. Higher pressures are necessary to increase the CO₂ concentration in the liquid phase, but electrode stability in these conditions is limited.

Solvents with high solubility for CO₂ are used in the non-aqueous electrochemical reduction of CO₂. However, high CO₂ solubility requires larger current density, but low electrolytic conductivity leads to high ohmic losses. Another problem is that high current densities are necessary to maximize the formation of hydrocarbons, and also a fast deactivation is present in these conditions.³⁴

Solventless electrocatalytic reduction of CO₂ would overcome these problems and allow the formation of more valuable products (liquid fuels) as will be discussed later. A novel design for the PEC solar cell is necessary for this scope.

2 Experimental

2.1 Synthesis and characterization of the anodic nanostructured materials

Nanostructured titania thin films were prepared by anodizing titanium discs (Alfa Aesar, 35 mm dia.) of 0.025 mm thickness and 99.96% purity. Prior to any electrochemical treatment, the discs were sonicated in distilled water, and then in isopropyl alcohol for 10 min each. Afterwards, they were dried in air. Anodic oxidation was carried out in a stirred electrochemical cell working at room temperature. An Autolab PGSTAT30 potentiostat/galvanostat in a two-electrode set-up was used to apply different anodizing voltages in the range of 15–50 V for times of up to 6 h. Three types of solutions were used as electrolyte: i) an aqueous solution containing 0.5 wt % HF in double distilled water; ii) an ethylene glycol solution and iii) a glycerol solution with 2 vol % H₂O and 0.3 wt % NH₄F. A platinum electrode served as the cathode. The geometry of the cell was described elsewhere.^35–37 The voltage was kept constant at the set-up value for the whole anodization time. Further details about the anodization conditions are reported in Table 1.

Table 1 Synthesis conditions used for the preparation of some titania-based materials

Electrolyte composition	Aqueous solution			Ethylene glycol
	0.5 wt % HF	0.5 wt % HF	0.5 wt % HF	0.3 wt % NH4F	0.3 wt % NH4F
	—	NH4OH	NH4OH	2 vol % H2O	2 vol % H2O
pH	0	6	6	7	7
Anodization potential/V	15	20	20	50	50
Anodization time/min	45	45	360	210	360

---

Atomic absorption (AA) and X-ray fluorescence (XRF) spectroscopy were used to check the purity of the solutions and analyze the amount of Ti dissolved in solution after anodization. Samples prepared in aqueous solution were cleaned with deionized water at the end of the anodization process; samples prepared in organic solvent were briefly sonicated (30 s) instead to remove debris from the top of the nanotubes. To induce crystallization in the anatase phase, the nano-structured substrates, after preparation, were annealed at 450 °C in air for 3 h with heating and cooling rates of 2 °C min⁻¹.³⁷ The structural and morphological characterization of the materials was performed by scanning electron microscopy (Hitachi S4800) with FEG (field emission gun) and by transmission electron microscopy (TEM) (JEOL-JEM 2010).

The oxide layer sizes (nanotubes length and diameter) were directly obtained from SEM cross-section images. The chemical composition of the oxide layers was determined by energy-dispersive X-ray spectroscopy (EDX).

The characteristics of the obtained titania nanotubes, such as size, shape, packing density and length, depend on several synthesis parameters: electrolyte type, applied voltage, pH, anodization procedure, etc.³⁸ In all cases, the 1D nanostructures are aligned perpendicularly to the metallic substrate and have a cylindrical shape and narrow size distribution.

2.2 Synthesis and characterization of the cathodic nanostructured materials

Commercial multiwalled carbon nanotubes (CNT) from Applied Science Inc. (XT24PS) were used as starting materials after post-treatment in inert gas at 700 °C. This treatment removes the amorphous carbon present on the outer surface and reduces the number of structural defects, thus increasing the graphitic character of the CNT. After this heat treatment, the samples were treated in nitric acid under reflux (HNO₃ 65%) at 100 °C for 10 h to create oxygen functional groups on the surface of the CNTs, which are necessary to better anchor the metal particles. Then, the CNTs were impregnated with 10 wt % Fe or Pt using Fe(NO₃)₃·9H₂O or H₂PtCl₆ as precursors. After drying at room temperature for 24 h, the samples were calcined for 2 h at 350 °C and finally reduced under H₂ at 400 °C for 2 h.

The cathode was prepared by depositing a suspension of Fe/ or Pt/CNT in ethanol on carbon cloth (CC). The metal loading was about 0.5 mg cm⁻². The active phase is located between the gas diffusion layer at direct contact with CO₂ in the gas phase and the Nafion membrane layer.

The samples were characterized by Zeta potential measurement to analyse the functionalization with acidic O-containing groups after oxidation with nitric acid, TG experiments in inert gas, SEM (scanning electron microscopy) and TEM (transmission electron microscopy).

The cathode operates in contact with the gas phase CO₂ flux (or inert gas in the case of H₂ production), the protons (provided by the opposite side of the solar cell which diffuse to the electrocatalyst through a Nafion® membrane) and the electrons provided by the external circuit and generated on the titania photocatalyst.

2.3 Assembling the PEC solar cell

The PEC solar cell consists of three layers: a nanostructured TiO₂ thin film supported on Ti prepared by anodization, a Nafion® membrane for electrical insulation and proton transport, and an electrocatalyst (Fig. 1). In H₂ photoproduction configuration, the electrocatalyst is a carbon cloth (E-tek®) with a high platinum dispersion (20%). In CO₂ reduction, the electrocatalyst is the electrode prepared by the deposition of Fe/ or Pt/CNT on carbon cloth described in section 2.2. The assembly of the disc was performed by hot pressing at 120 °C under a pressure of 20 kg cm⁻².

Fig. 1 (a) View of the lab-scale PEC device. (b) Image of the photo/electro-catalytic disc. (c) Scheme of the PEC device for CO₂ reduction to fuels and H₂ production.

2.4 PEC reactor

The apparatus for the photo-catalytic experiments consists of a solar illuminator source, a photo-reactor and gas chromatographs for on-line analysis (Fig. 1). The lamp housing is furnished with a Xe-arc lamp (ORIEL, 300 W), a set of lenses for light collection and focusing, and a water filter, to eliminate the infrared radiation.

The PEC reactor, made of Plexiglas and equipped with a quartz window, is homemade. It has a two-electrode configuration with two compartments for separated evolution of H₂ (or CO₂) and O₂. The irradiated area is 5.7 cm². The photoanode is the nanostructured TiO₂ thin film supported over the porous titanium foil. The cathode is the carbon cloth loaded with Pt or Pt(or Fe)/CNT supported on CC. The two electrodes are joint together by a Nafion^® membrane.

1 M NaOH aqueous solution was used as electrolyte in the anodic compartment. The solution circulates continuously between the solar cell and an outer reservoir. In H₂ photoproduction configuration, a 0.5 M H₂SO₄ aqueous solution was used in the cathodic side of the cell. The simplified process is as follows: (1) light crosses the quartz window and reaches the nanostructured film (photoanode) where photo-generated electron and hole pairs are generated and O₂ evolves, (2) protons pass through the Nafion® membrane, while electrons are collected and reach the cathode through an external wire, and (3) protons react with CO₂ in the presence of electrons on the CNT-based electrocatalyst to give liquid fuels or recombine with electrons, over Pt nanoparticles supported on CC, to give H₂. A potentiostat–galvanostat (AMEL 2049) was used to measure the generated photocurrent.

H₂ and O₂ amounts in the gas streams were periodically determined with an on-line gas-chromatograph (GC), equipped with a molecular sieve 5A column and a thermal conductivity detector (TCD).

Moreover in CO₂ reduction, analysis of the products was made by a GC equipped with a flame ionization detector (FID) and a gas-mass GC (MS-GC) apparatus. Before the experiments, the photo-reactor was completely degassed to remove the dissolved oxygen.

Aqueous solutions of hydrolysis products of lignocellulose or waste solutions from biomass fermentation (e.g. diluted solutions containing bioethanol) can also be used to enhance H₂ or CO₂ formation by photoreforming. For this reason, ethanol was used as model reactant in some testing experiments.

3 Results and discussion

3.1 Photoproduction of H₂ and the role of titania thin-film nanostructure

The photoanode in a PEC solar cell, as introduced in section 1.1, should be in the form of a porous thin film allowing a good light harvesting, fast transport of the protons and electrons produced during the water oxidation, and a good contact with both the electron-collector substrate and the proton-conductor membrane. There is, thus, a need to have a specific nanostructure in the photoanode. The use of an array of 1D aligned nanostructures (nanorods, nanotubes, etc.) improves light harvesting and limits charge recombination at the grain boundaries with respect to an assembly of nanoparticles, while maintaining a high geometrical surface area necessary to improve the photoresponse.³⁸ An optimal contact/interface with the H⁺-membrane is necessary.

A further general issue is the need to use preparation methods to produce the photoanode which (i) can be cost-effective, (ii) allows a good control of the nanostructure in terms of 1D-array characteristics, vertical alignment and density, and thickness, and (iii) can be easily scalable to large films (at least 10 × 10 cm). The choice of suitable preparation methods with all these characteristics is quite restricted.³⁸ We have thus focused attention on the anodic oxidation of titanium thin foils to synthesize 1D nanostructured titania nanotube arrays to be used as the photoanode in PEC solar cells.^38–41 Coupled with the use of titania thin films with an oriented nanostructure we have designed a new modular PEC system equipped with a highly compact photoreactor.

These TiO₂ materials show excellent properties in the preparation of solar cells and photoelectrodes.^42–44 We limit discussion here to undoped titania films, but their reactivity properties can be further promoted by doping or creating heterostructures which promote visible light absorption and effective transfer of electrons from the valence to the conduction band of titania.

Table 2 reports the comparison of the performances in H₂ photoreforming of three different titania films: (a) produced by pressing the commercial P25 Degussa TiO₂, (b) prepared by anodic oxidation of Ti foils and (c) prepared by sol–gel dip-coating method. For comparison, the results obtained with P25 Degussa using a slurry-type well-mixed reactor are also reported. This reactor has a quartz window and it is illuminated with the same solar illuminator used for the tests with titania thin film. The results are thus comparable. All these titania samples were loaded with 0.5 wt % Pt by wet impregnation, because without Pt all the samples show very low activity in photoreforming.^45,46

Table 2 Hydrogen evolution rate for different kinds of TiO₂ samples during photoreforming of 10% ethanol in water^d

	TiO2 sample	Hydrogen/mmol h^−1 g^−1
a Degussa P25 TiO2 powder. b ordered array of TiO2 nanotubes prepared by anodization on Ti foil. c TiO2 dense film prepared by sol–gel dip-coating. d In all samples 0.5 wt % Pt was added by wet impregnation.
Slurry	Powder^a	3.5
Film	Pressed powder^a layer on Ti foil	7.0
	TiO2 nanotube array on Ti^b	47.8
	Sol–gel (dip-coating)^c	5.3

---

The data in Table 2 demonstrated well that the specific photoreforming activity (per amount of titania) considerably depends on both the reactor geometry and the nanostructure of the titania thin film.

P25 TiO₂ Degussa, a classical reference material in photocatalysis, shows a specific activity about two times higher when used in the form of a thin film than as suspended powder (slurry reactor), due to the reduced light scattering when used as a compacted thin film. The sample prepared by sol–gel dip-coating, also a conventional method for preparing titania thin films, shows performances quite similar to those of the thin film prepared by compacting the P25 TiO₂ nanoparticles.

The slightly worse performances could be attributed to the presence of an about 15% rutile phase together with anatase in P25 TiO₂, while only anatase is present in the sample prepared by sol–gel. It is known that rutile acts as an antenna to extend the photoactivity into visible wavelengths and the structural arrangement of the similarly sized TiO₂ crystallites creates catalytic “hot spots” at the rutile/anatase interface.⁴⁷ The difference between the thin film produced by sol–gel dip-coating or by pressing TiO₂ preformed nanoparticles (P25) are minimal in comparison to the one-order of magnitude higher activity of the thin film in the form of an ordered array of vertically-aligned TiO₂ nanotubes (produced by anodic oxidation). It is thus evident how the nanostructure of TiO₂ has a marked effect.

The reasons are multiple. A better light harvesting, lower rate of recombination at the grain boundaries, reduced effects related to formation of a H₂ cap, faster electron transport and charge separation, and a nanostructure which increases the absorption of the visible-light component.^35–41 The result of these multiple effects is that the specific activity of this TiO₂ nanotube array thin film has significant higher activity in photoreforming.

The activity of nanostructured TiO₂ films in H₂ production by photoreforming is a factor of about 6–7 higher than that by water splitting, but a similar trend is observed, evidencing that the basic mechanistic factors determining the photoactivity are the same in the two reactions. Table 3 evidences this concept by comparing the behaviour in H₂ production by photoreforming and water splitting of two nanostructured TiO₂ films prepared by anodic oxidation at two different voltages. As reported earlier,^35,37 the increase of the anodization voltage leads to thicker films, but with a larger diameter of the TiO₂ nanotubes and thus less dense packing. The sample prepared at higher anodization voltage results in more active in H₂ photoproduction. There is a good correspondence between activity increase in water splitting due to the different characteristics of the nanostructured film, and photoreforming activity.

Table 3 Hydrogen evolution rate and photo-generated current during water splitting and photoreforming for samples anodized at different voltage

	Water splitting		Photoreforming
Anodization voltage	20 V	40 V	20 V	40 V
H2/μmol h^−1	3.3	5.3	19.3	37.1
Photocurrent density/μA cm^−2	29.0	40.4	210.0	335.0

---

Table 3 also evidences the concept that a parallel trend is observed in terms of photocurrent generated during the reaction of H₂ photoproduction. This is also a general observation we made in accordance with the suggestion that it is preferable to indicate the reaction of H₂ production from water on titania as photoelectrolysis instead of water splitting, as is often used.

The nanostructure of the titania film produced by anodic oxidation is strongly influenced from the specific conditions of anodization and, as a consequence, the photocurrent and photoactivity are also considerably dependent on the preparation. Fig. 2 reports the increase in photocurrent measured in a series of nanostructured TiO₂ films upon irradiation with a low-power solar lamp (60 W tungsten lamp). In general, the increase of the time of anodization leads to an increase in the thickness of the film with an increase of the photocurrent.

Fig. 2 (a) Photocurrent increments upon irradiation with a 60 W tungsten lamp of some titania nanotube arrays prepared in aqueous and non-aqueous electrolytes. (b) Anodization time of the same titania arrays.

However, while minor differences are observed using either water or glycerol as the solvent, a very large difference is observed using ethylene glycol as the solvent. In this case, an about one order of magnitude higher photocurrent density is observed, particularly for the longer anodization times (>6 h).

Fig. 3 reports the SEM images of some of the samples reported in Fig. 2, prepared either in glycerol or water as the solvent. The different characteristics of the samples may be noted—in terms of diameter of the nanotubes, their wall thickness and packing density of the arrays. The thickness of the films varies between 200 and 700–800 nm. Nearly straight channels could be evidenced in cross-section images, although sometimes a bamboo shape could be also noted. These characteristics have an influence on the photocurrent properties, as evidenced in Fig. 2, but they are of minor relevance with respect to the effect of the synthesis in ethylene glycol.

Fig. 3 SEM images (top view) of titania nanotube arrays prepared by anodic oxidation under different conditions: (a) in glycerol containing 0.5 wt % NH₄F at pH 6 applying a 20 V potential for 45 min; (b) in water containing 0.5 wt % HF at pH 0 applying a 15 V potential for 45 min; (c) in water containing 0.5 wt % HF at pH 4 (adjusted with NH₄OH) applying a 20 V potential for 45 min.

By using ethylene glycol as solvent during the anodization, significantly thicker films were obtained, from 10 microns to higher values, depending on the anodization time and voltage. Fig. 4 reports an example of the titania films which can be obtained. It may be noted that in this case very regular straight TiO₂ nanotubes with thick walls are obtained. At longer times of anodization, the presence of an amorphous titania deposit on the surface is observed. After removing this debris from the surface by sonication, the performance in H₂ production improves by about 10%.

Fig. 4 SEM images (cross section) of nanotube arrays obtained by anodic oxidation of a Ti foil in ethylene glycol containing 0.3 wt % NH₄F and 2 vol % H₂O applying a 50 V potential for 6 h. The thickness of the titania nanostructured film is about 14.6 μm; tube internal diameter is 40 ÷ 43 nm while tube external diameter is in the range 100–105 nm.

These nanostructured titania thin films may be used in either the liquid phase or gas phase, particularly for the case of H₂ production by photoreforming. Fig. 5 reports the comparison of the rates of H₂ production during ethanol–water photoreforming in the liquid phase (room temperature) and in the gas phase (70 °C). It is evident that in the gas phase, due to the reduced scattering of the light by water and the more efficient desorption of the adsorbed species which quench photoinduced processes, a one order of magnitude higher rate of H₂ production is possible. Gas phase photoreactions in H₂ production are a rather unexplored area, but the results reported in Fig. 5 indicate a quite promising area to investigate.

Fig. 5 Rate of H₂ production profiles for nanostructured TiO₂ thin films with or without Pt and under liquid or gas phase conditions, during photoreforming of 10% ethanol in water. Temperature: liquid phase 40 °C; gas phase 70 °C (temperature of the evaporating solution).

Fig. 5 also evidences the role of the presence and nature of Pt nanoparticles on titania in increasing the rate of H₂ formation. The addition of 0.5 wt % Pt to titania results in a significant increase in the rate of the reaction both in the liquid and gas phases. The usual method of deposition by incipient wet impregnations leads to Pt nanoparticles prevalently in the 3–8 nm range. By photoreduction it is possible to obtain smaller Pt particles (mainly below 2–3 nm) with a minor density and more homogeneous dispersion (see TEM images in Fig. 6). This causes an increase of 2–3 times in the rate of H₂ photoproduction, showing the relevant role of this parameter in the preparation of the samples.

Fig. 6 TEM images of nanotubes decorated with Pt nanoparticles deposited by photoreduction (left side) and wet impregnation (right side).

In Fig. 5, it may also be noted that, particularly in the gas phase, the rate of H₂ formation passes through a maximum after about 2 h of time on stream, to decrease later and stabilize to a value about 15–20% lower. The activity then remains constant at least in lab-scale experiments.

The initial increase is related to the accumulation of H₂ in the photoreactor and lines to the analytical apparatus which has initially been purged with an inert gas. A continuous flow of inert gas was also sent to the reactor during the photoactivity tests. Usually, about four–five times the residence (holding) time is necessary to reach a constant value in transient CSTR-type reactor experiments, e.g. about 2 h in our case. The initial increase of H₂ is thus reasonable, but the presence of the maximum evidences that the initial activity of the titania film is probably higher.

There are different possible explanations, but before discussing them, it is preferable to consider the results concerning the oxygen evolution profile and the photo-generated current simultaneously monitored during the water splitting (photoelectrolysis) experiments reported in Fig. 7.

Fig. 7 Rate of H₂ and O₂ production and photo-generated current profiles for a water splitting experiment in the PEC reactor with a nanostructured TiO₂ thin film (anodized at 20 V) as photoanode.

The same maximum previously observed in Fig. 5 is present, and it may be noted that the maximum in O₂ production occurs previous to those for H₂. However, this could be related to the time necessary to establish the equilibrium through the proton-conductive membrane (Nafion®), because the H₂ is produced on the cathode side and O₂ on the anode side of the PEC reactor. Constant production of H₂ and O₂ is reached after about 3–4 h of time-on-stream. The photocurrent instead shows a very sharp maximum after a few minutes, and then decreases, reaching a constant value about one order of magnitude lower than the maximum value.

The explanation of this effect requires detailed studies, but it may be suggested that it is likely that during the initial part of the reaction, the electrons produced in the photoreaction react with the oxygen produced from water to form O₂⁻ species which quench photoanode activity.^48–50

According to this explanation, the potential photoactivity of the system in water splitting is about one order of magnitude higher, but limited by the presence of oxygen adsorbed species which quench the photoprocesses. The same explanation is probably valid for the results reported in Fig. 5.

Another possible explanation of the sharp maximum in the first few minutes of the water splitting experiment could be due to reaction of the photoelectrons with the O₂ adsorbed on the cathode, which causes a rapid photocurrent decrease to a lower level after oxygen consumption.

The mechanisms in photogenerated surface processes are complex and transient effects are difficult to analyze. In addition, results often refer to conditions (high vacuum) still quite far from real ones. It is thus not possible to have unique conclusive indications and thus alternative interpretations of the transient effect are possible.

It should be remarked, however, that the time scale of the transient effect observed is minutes–hours, and thus much longer than the pico–micro second time scale effects often discussed in literature. The time scale of the transient effect observed already evidences that it is related to the accumulation of surface adsorbed species on the photocatalyst. This result is supported from the observation that the effect is fully reversible, while irreversible effects could be expected, for example, in the case of change of the titania bulk characteristics (creation of Ti³⁺, for example).

It may thus be concluded that a better understanding of these aspects and the possibility to limit their negative effects (for example, favouring oxygen desorption operating at higher temperatures) are key elements to progress towards the production of solar fuels.

3.2 Using the PEC reactor for solar fuels from CO₂

Using the PEC reactor to reduce CO₂ will be possible in the future to develop “artificial trees” capable of capturing the CO₂ and convert it to liquid fuels (hydrocarbons, alcohols).²² Therefore, the implementation of this concept will allow the reduction of the levels of CO₂ in the atmosphere and at the same time capture a renewable source of energy (solar radiation), transforming it in a form (liquid fuels) which can be stored, used and transformed, thus preserving the large investments made on fossil fuels. The liquid hydrocarbons and alcohols can be alternatively used as chemical feedstocks, or renewable fuels.

The PEC reactor shows analogies to PEM fuel cells, i.e. can use the R&D on the latter to reduce the costs and improve scale-up. Some aspects of the characteristics and properties of the nanostructured TiO₂-based thin film we used as photoanode were discussed in section 3.1. The following section will instead discuss some aspects of the cathode side, where particularly the conversion of CO₂ (using the protons and electrons generated in water photodissociation) is made using special electrocatalysts, based on the concept of nanoconfinement.^51–53

It should be mentioned, however, that up to now we studied the photoanode and cathode separately in the frame of two distinct EU projects (NATAMA and ELCAT, respectively). The full assembled PEC reactor was tested only for water splitting with separate H₂ and O₂ production (see Fig. 6), while work is in progress for the case of CO₂ reduction.

In the introduction it was evidenced how the conversion of CO₂ to liquid fuels is a key part of the vision for future carbon-neutral fossil fuels.⁹ In addition, it was shown how solar H₂ may be considered a part of this vision, which is more than just a suitable energy vector because there is the need to produce easily transportable and stored fuels, which can be integrated into the existing energy infrastructure. In this sense, we strongly suggest that liquid fuels produced from carbon dioxide and water using solar energy is the preferable route, notwithstanding the difficulties in the reaction.

Some earlier aspects of CO₂ conversion to fuels have been reviewed by Centi and Perathoner²³ discussing the status and perspectives on the heterogeneous catalytic reactions with CO₂. A recent review⁵² and advanced report⁵⁴ reveal more closely the opportunities and prospects in the chemical recycling of carbon dioxide to fuels.

The vision of carbon dioxide as the hydrogen-storage material of the future was reported in an essay by Enthaler,²⁴ where the cycle between CO₂ and formic acid was identified as the preferable option for renewable energy systems, notwithstanding the problems in terms of toxicity and instability associated with formic acid. Various aspects of the chemical fixation of CO₂ in constructing a future low-carbon global economy with reference to energy sources, thermodynamic considerations, net carbon emissions and availability of reagents were discussed by Kerry Yu et al.,⁵⁵ while recent advances in CO₂ capture were discussed by Choi et al.⁵⁶

3.3 Electrocatalytic conversion of CO₂ to liquid fuels

As evidenced in section 1.1, the electrochemical utilization of CO₂ has been studied for many years, but using the conventional electrochemical approach (liquid phase), many problems exist related to the solubility of CO₂, type of products formed, stability, etc. which never allowed CO₂ to pass the lab-stage development. To overcome these problems, we thus proposed to form under solventless conditions liquid fuels such as long-chain hydrocarbons and/or alcohols, which can be easily collected without the need to be distilled from liquid solutions (a quite energy intensive process).^23,51,52

We will discuss here some aspects of this investigation, but remarking that experimentation up to now has been limited to continuous hemi-cell (cathode side). In these cells, the photoanode is substituted with a liquid electrolyte to supply the protons (through the Nafion® membrane) necessary for the reduction of carbon dioxide on the cathode side (operating in the gas phase). A constant potential of about −2 V was applied through the cell to supply the electrons necessary for the reaction of CO₂ reduction. However, we discovered that in these conditions there is also a migration (cross-over) of potassium ions through the membrane (potassium hydrogencarbonate was used as electrolyte on the anode side of the cell).⁵¹ These K⁺ ions react with Fe or Pt nanoparticles on the cathode side causing their dissolution/migration (in the case of Fe) or poisoning (in the case of Pt), thus leading to irreversible deactivation.

The electrolyte is necessary to simulate the half-cell of the full photoelectrocatalytic device, while in the latter no electrolyte is needed, as the protons and electrons are produced by water splitting. The elimination of the electrolyte in the PEC full cell is expected to eliminate this main cause of deactivation, but activity in this direction is only in progress, due to the need to optimize the interface between the titania photoanode and proton membrane, as well as the need to develop membranes operating at higher temperature (based, for example, on supported liquid heteropolyacids or supported ionic liquids containing heteropolyacids) for the motivations discussed in section 3.1.

The features of the electrode used in this gas phase electrocatalytic reduction of CO₂ are close to those used in PEM fuel cells,²⁰e.g. carbon black/Pt on a carbon cloth/Nafion® assembled electrode (GDE—gas diffusion electrode). The electrocatalyst is the Pt supported on carbon black, which is then deposited on a conductive carbon cloth to allow the electrical contact and the diffusion of gas phase CO₂ to the electrocatalyst. The Pt particles are at the contact with Nafion® through which protons diffuse. On the Pt nanoparticles, the gas phase CO₂ reacts with the electrons and protons to be reduced to longer chain hydrocarbons and alcohols, whose relative distributions depends on the reaction temperature. However, using this electrode acetone is the major product of conversion at 60 °C, while at rt the productivities are lower, but longer-chain hydrocarbons are formed.⁵³

Using a similar GDE configuration, consisting of carbon nanotubes instead of carbon black as support, for the electrocatalyst nanoparticles, it is possible to form isopropanol as the main product of reaction. Table 4 also evidences that using carbon nanotubes, it is possible to also use iron nanoparticles instead of a noble metal, although the latter shows a better stability. In addition, it is shown that the use of N-doped carbon nanotubes (N/CNT) allows further improvement in the productivity of making isopropanol.

Table 4 Product distribution at 60 °C in the electrocatalytic reduction of carbon dioxide in the gas phase over Nafion® 117/(Pt or Fe_(10%)/CNT)_20%/carbon cloth GDM (gas diffusion membrane) electrode

Main products	μmoles h^−1 cm^−2 (^a)
	Pt10-CNT	Fe10-CNT	Fe10–N/CNT
a Tests in a semi-batch cell, using a 0.5 KHCO3 electrolyte on the anode side and operating the cathode in the gas phase with a continuous feed of 50% CO2 in humidified nitrogen.
Methanol	4.95 × 10^−4	5.17 × 10^−4	8.70 × 10^−4
Acetaldehyde	4.98 × 10^−4	2.55 × 10^−3	3.23 × 10^−4
Ethanol	6.87 × 10^−3	1.02 × 10^−2	6.71 × 10^−4
Acetone	5.28 × 10^−5	2.26 × 10^−4	5.92 × 10^−5
Isopropanol	1.57 × 10^−2	2.28 × 10^−2	5.74 × 10^−2

---

Not only is the nature of carbon relevant, but also the presence of nanocavities, which could favor the consecutive conversion of intermediates with formation of C–C bonds. Table 5 reports the comparison of the behaviour at 25 °C in the electrocatalytic reduction of CO₂ of two 10 wt % Fe/CNT samples. In the first catalyst, the iron particles are located both on the inner and outer surface of the CNT (the usual situation), as shown by transmission electron microscopy data.

Table 5 Comparison of the product distribution in CO₂ electrocatalytic reduction at 25 °C on Fe/CNT (10 wt % Fe): the iron particles are located both on the inner and outer surface of CNT (A) or only on the inner surface (B).⁵¹

Main products	mmoles min^−1 ^a
	(A) Fe on inner & outer surface of CNT	(B) Fe only inside CNT
a Tests in a semi-batch cell, using a 0.5 KHCO3 electrolyte on the anode side and operating the cathode in the gas phase with a continuous feed of 20 ml min^−1 CO2 in humidified nitrogen.
Methanol	3.08 × 10^−8	5.15 × 10^−8
C3 hydrocarbons	1.48 × 10^−7	3.41 × 10^−7
Ethanol	0	5.02 × 10^−8
C5 hydrocarbons	2.65 × 10^−8	1.06 × 10^−6
Propanol	6.67 × 10^−8	5.33 × 10^−7
Butanol	1.91 × 10^−7	3.55 × 10^−7
C6 hydrocarbons	4.07 × 10^−7	4.72 × 10^−7
C7 hydrocarbons + aromatics	4.06 × 10^−7	6.49 × 10^−7
C8–C9 hydrocarbons	7.53 × 10^−7	1.19 × 10^−6

---

In the second sample, Fe is localized only inside the CNT.⁵¹ The inside localization leads to an enhanced productivity (total sum of products is about 5 mmol min⁻¹vs. about 2 mmol min⁻¹) and influences the product distribution (in particular, favouring hydrocarbon formation). Larger iron particles are present inside the CNT.

Note that these tests were performed at room temperature, while those reported in Table 4 were obtained at 60 °C. The increase of temperature allows the improvement of productivity by about one order of magnitude, also forming shorter C-chain products (<C6) and with enhanced formation of alcohols with respect to hydrocarbons. A further increase of temperature is probably necessary to optimize the performances, but this would require the use of the full PEC reactor and the change of the type of proton membrane.

Conclusions

Producing solar fuels is a topic of current large scientific and industrial interest which also attracts great public attention.¹⁰

We have analyzed here some aspects related to the characteristics of nanostructured (a) TiO₂ thin films (based on ordered array of titania nanotubes) and their performances as photoanodes in H₂ production by water splitting or photoreforming of ethanol (in liquid or gas phase) and (b) carbon-nanotube based electrodes for the gas phase reduction of CO₂ to liquid fuels (mainly isopropanol) together with their application for the design of a novel photoelectrocatalytic (PEC) approach for the synthesis of solar fuels.

Some of the possibilities for the improvement of the current limits, related to the design of the photoanode and electrocathode, were evidenced. This is an area under fast current development worldwide and thus significant progresses may be obtained in the near future. Nevertheless, the preferable routes and technology options are still to be identified.

We remarked, in particular, the need to produce fuels that are easy to transport and store, which can be integrated into the existing energy infrastructure. In this sense, we suggest that liquid fuels produced from carbon dioxide and water by using solar energy are preferable, notwithstanding the difficulties in the reaction.

We feel that the PEC reactor design discussed in this work is a good attempt in this direction towards the practical implementation of solar fuels, even if its applicability is still a long way off.

Acknowledgements

This work was realized in the frame of the Italian PRIN-07 Project “Renewable H₂”. The authors like to thank Dr D. S. Su, Prof. R. Schlögl (Frizt Haber Institute, Berlin, Germany) for helpful discussions and Mr M. Laganà (CNR-ITAE, Messina, Italy) for SEM measurements. The technical assistance of Mr D. Cosio (Univ. Messina, Italy) in the construction of the PEC reactor is also acknowledged.

Notes and references

1. V. Balzani, A. Credi and M. Venturi, ChemSusChem, 2008, 1, 26–58.
2. V. G. Dovì, F. Friedler, D. Huisingh and J. J. Klemeš, J. Clean. Prod., 2009, 17, 889–895.
3. N. Z. Muradov and T. N. Veziroglu, Int. J. Hydrogen Energy, 2008, 33, 6804–6839.
4. N. Nakćenović, in World Energy Assessment: Energy and the Challenge of Sustainability, World Energy Council, London, 2000, ch. 9, pp. 333–366.
5. G. Centi and R. A. van Santen, Catalysis for Renewables, Wiley VCH, Weinheim, 2007.
6. G. Centi, F. Trifiró, S. Perathoner and F. Cavani, Sustainable Industrial Chemistry, Wiley VCH, Weinheim, 2009.
7. N. S. Lewis, G. Crabtree, A. Nozik, M. Wasielewski and P. Alivisatos, Basic Research Needs for Solar Energy Utilization, US Department of Energy, Washington DC, 2005.
8. A. T. Bell, B. C. Gates and D. Ray, Basic Research Needs: Catalysis for Energy, US Department of Energy, Washington DC, 2007.
9. G. Centi, S. Perathoner, R. Passalacqua, C. Ampelli, in Carbon-neutral Fuels and Energy Carriers: Science and Technology, ed. N. Veziroglu and N. Muradov, Taylor & Francis, London, 2010, submitted.
10. H. B. Gray, Nat. Chem., 2009, 1, 7.
11. R. Bhandari and I. Stadler, Sol. Energy, 2009, 83, 1634–1644.
12. A. E. Curtright, M. G. Morgan and D. W. Keith, Environ. Sci. Technol., 2008, 42, 9031–9038.
13. A. Currao, Chimia, 2007, 61, 815–819.
14. C. Minero and V. Maurino, in Catalysis for Renewables, ed. G. Centi and R. van Santen, Wiley-VCH, Weinheim, 2007,ch. 16, pp. 351–385.
15. H. Tributsch, Int. J. Hydrogen Energy, 2008, 33, 5911–5930.
16. J. O'M. Bockris, Int. J. Hydrogen Energy, 2008, 33, 2129–2131.
17. J. Nowotny and L. R. Sheppard, Int. J. Hydrogen Energy, 2007, 32, 2607–2608.
18. K. E. N'Tsoukpoe, H. Liu, N. Le Pierrès and L. Luo, Renewable Sustainable Energy Rev., 2009, 13, 2385–2396.
19. A. Kudo and Y. Miseki, Chem. Soc. Rev., 2009, 38, 253–278.
20. G. Centi and S. Perathoner, Eur. J. Inorg. Chem., 2009, 3851–3878.
21. H. Chen, T. N. Cong, W. Yang, C. Tan, Y. Li and Y. Ding, Prog. Nat. Sci., 2009, 19, 291–312.
22. G. Centi and S. Perathoner, Top. Catal., 2009, 52(8), 948–961.
23. G. Centi and S. Perathoner, Stud. Surf. Sci. Catal., 2004, 153, 1–8.
24. S. Enthaler, ChemSusChem, 2008, 1, 801–804.
25. T. Kodama and N. Gokon, Chem. Rev., 2007, 107, 4048–4077.
26. A. J. Traynor and R. J. Jensen, Ind. Eng. Chem. Res., 2002, 41, 1935–1939.
27. C. Agrafiotis, M. Roeb, A. G. Konstandopoulos, L. Nalbandian, V. T. Zaspalis, C. Sattler, P. Stobbe and A. M. Steeled, Sol. Energy, 2005, 79, 409–421.
28. M. Roeb, M. Neises, J.-P. Säck, P. Rietbrock, N. Monnerie, J. Dersch, M. Schmitz and C. Sattler, Int. J. Hydrogen Energy, 2009, 34, 4537–4545.
29. E. E. Barton, D. M. Rampulla and A. B. Bocarsly, J. Am. Chem. Soc., 2008, 130, 6342–6344.
30. R. M. Navarro Yerga, M. C. Alvarez Galván, F. del Valle, J. A. Villoria de la Mano and J. L. Fierro, ChemSusChem, 2009, 2, 471–485.
31. C. Ampelli, R. Passalacqua, S. Perathoner and G. Centi, Chem. Eng. Trans., 2009, 17, 1011–1016.
32. D. L. DuBois, in Encyclopedia of Electrochemistry, Wiley-VCH,Weinheim, 2006, vol. 7A, ch. 6.2, pp. 202–225.
33. M. Gattrell, N. Gupta and A. Co, J. Electroanal. Chem., 2006, 594, 1–19.
34. Y. Hori, H. Konishi, T. Futamura, A. Murata, O. Koga, H. Sakurai and K. Oguma, Electrochim. Acta, 2005, 50, 5354–5369.
35. G. Centi, R. Passalacqua, S. Perathoner, D. S. Su, G. Weinberg and R. Schlögl, Phys. Chem. Chem. Phys., 2007, 9, 4930–4938.
36. C. Ampelli, R. Passalacqua, S. Perathoner, G. Centi, D. S. Su and G. Weinberg, Top. Catal., 2008, 50, 133–144.
37. S. Perathoner, R. Passalacqua, G. Centi, D. S. Su and G. Weinberg, Catal. Today, 2007, 122, 3–13.
38. G. Centi and S. Perathoner, in Catalysis, ed. J. J. Spivey, Royal Society of Chemistry, Cambridge, UK, 2009, vol. 21, pp. 82–119.
39. C. A. Grimes and G. K. Mor, TiO₂ Nanotube Arrays: Synthesis, Properties, and Applications,Springer, Heidelberg, Germany, 2009.
40. P. Schmuki, in Electrochemistry at the Nanoscale, Springer Science, New York, 2009, pp. 435–466.
41. G. Centi and S. Perathoner, in Catalysis, ed. J. J. Spivey, Royal Society of Chemistry, Cambridge, UK, 2007, vol. 20, pp. 367–394.
42. G. K. Mor, K. Shankar, M. Paulose, O. K. Varghese and C. A. Grimes, Nano Lett., 2006, 6, 215–218.
43. D. Kuang, J. Brillet, P. Chen, M. Takata, S. Uchida, H. Miura, K. Sumioka, S. M. Zakeeruddin and M. Grätzel, ACS Nano, 2008, 2, 1113–1116.
44. K. Shankar, J. I. Basham, N. K. Allam, O. K. Varghese, G. K. Mor, X. Feng, M. Paulose, J. A. Seabold, K.-S. Choi and C. A. Grimes, J. Phys. Chem. C, 2009, 113, 6327–6359.
45. M. Antoniadou, D. I. Kondarides and P. Lianos, Catal. Lett., 2009, 129, 344–349.
46. V. M. Daskalaki and D. I. Kondarides, Catal. Today, 2009, 144, 75–80.
47. D. C. Hurum, A. G. Agrios, K. A. Gray, T. Rajh and M. C. Thurnauer, J. Phys. Chem. B, 2003, 107, 4545–4549.
48. A. Yamakata, T.-a. Ishibashi and H. Onishi, J. Mol. Catal. A: Chem., 2003, 199, 85–94.
49. J. Tang, J. R. Durrant and D. R. Klug, J. Am. Chem. Soc., 2008, 130, 13885–13891.
50. O. Dulub, M. Batzill, S. Solovev, E. Loginova, A. Alchagirov, T. E. Madey and U. Diebold, Science, 2007, 317, 1052–1056.
51. M. Gangeri, S. Perathoner, S. Caudo, G. Centi, J. Amadou, D. Bégin, C. Pham-Huu, M. J. Ledoux, J.-P. Tessonnier, D. S. Su and R. Schlögl, Catal. Today, 2009, 143, 57–63.
52. G. Centi and S. Perathoner, Catal. Today, 2009, 148, 191–205.
53. G. Centi, S. Perathoner, G. Winè and M. Gangeri, Green Chem., 2007, 9, 671–678.
54. G. Centi and G. Cum, in Direct Conversion of Methane, Ethane and Carbon Dioxide to Fuels and Chemicals, CAP Report, The Catalyst Group Resources, Spring House, PA, 2008, ch. 4, pp. 85–187.
55. K. M. Kerry Yu, I. Curcic, J. Gabriel and S. C. E. Tsang, ChemSusChem, 2008, 1, 893–899.
56. S. Choi, J. H. Drese and C. W. Jones, ChemSusChem, 2009, 2, 796–854.

---