Titanium Dioxide as a Catalyst Support
Titanium dioxide, has, over the past two decades emerged as a very efficient catalyst due to its photocatalytic activity. However, transition metal oxides have long acted as supports to catalysts, and the physical properties of titanium dioxide have made it an excellent candidate in this role.
A critical review of its use as a catalyst support in the hydrodesulfurization process, in polymer electrolyte membrane fuel cells and in the water-gas shift reaction is given.
Studies show that titanium dioxide with high surface area has been prepared and results in the hydrodesulfurization process demonstrate that it is able to promote higher catalytic activities than conventional alumina supports. Its high stability has played an important role in polymer electrolyte fuel cells, where it has been able to withstand the corrosion effects of electrochemical oxidation. The steam reforming process used in fuel cells to generate hydrogen produces poisonous carbon monoxide which can be removed by the water-gas shift reaction. Operating conditions for conventional water-gas shift catalysts are unsuitable in fuel cells, and the results discussed within this review show that using titanium dioxide as a catalyst support may be a viable alternative.
Titanium dioxide, also known as titania, is a remarkable ceramic material with unique properties that have proved useful in a range of applications. It has a very high refractive index, which gives it the ability to disperse light better than most other materials. This property has been exploited to give the white pigment used in paint, and in sunscreen to disperse ultraviolet radiation away from the skin.
Thin layer films of titanium dioxide are used to cover medical implants. These films allow integration between the implant and bone tissue, which greatly reduces the possibility of implant rejection by the body. This use illustrates the inertness and biocompatibility of titanium dioxide.
In the presence of ultra violet light, titanium dioxide behaves as a very efficient photocatalyst. This property, attributed to the materials strong oxidising power, is utilised in chemical cleaning processes. Applications in self-cleaning glass, air purification and treatment of cancerous tumours have all benefited from this property.
Studies in recent years have shown that, in addition to titanium dioxide being an excellent candidate as a catalyst, it is just as effective as a support to a catalyst. The properties exhibited by titanium dioxide, such as large surface areas, have sparked a great deal of interest among researchers to develop novel methods and processes where it may be able to replace conventional supports.
This review will explore and discuss recent developments in research surrounding the use of titanium dioxide as a support for catalysts in a variety of applications.
The term catalysis was first defined in 1835 by J. J. Berzelius to “describe the property of substances that facilitate chemical reactions without being consumed in them” . This definition allows for the possibility that a catalyst may be able to either increase or decrease the rate of a chemical reaction. However, in most occurrences, the addition of a catalyst affects the rate by increasing it. Catalysts are employed in over 90% of chemical processes, and so their importance cannot be underestimated .
Catalysts can be of two types: heterogeneous or homogeneous. The difference between these two forms is the composition of the phases present during reaction. Heterogeneous catalysts are present in a different phase to reactants and products, usually solid, whereas homogeneous catalysts are present in the same phase as reactants and products, usually liquid.
Research carried out in the topic area of this review has involved using titanium dioxide as the support for a catalyst, of which the catalyst has been the solid phase. This introduction will therefore focus on heterogeneous catalysis.
When a catalyst is present in a different phase to the reactants and products, such as in heterogeneous catalysis, the advantage is that it is relatively simple to separate the catalyst from the product e.g. separating a liquid and a solid. In addition, heterogeneous catalysts are able to withstand more extreme conditions (e.g. temperature, pressure) than an equivalent homogeneous catalyst .
1 illustrates the process that occurs when using a heterogeneous catalyst with the example reaction of the hydrogenation of ethene.
Ethene and hydrogen are passed over a metal catalyst (usually nickel) and adsorb onto its surface, shown in the first step. Hydrogen dissociatively chemisorbs on to the surface of the metal catalyst, and then migrates to the adsorbed ethene molecule (second step). This initially yields a surface alkyl, as shown in the third step, and following migration of the second hydrogen atom, the saturated hydrocarbon ethane is formed (fourth step) .
So, in general, the process that occurs in a chemical reaction involving a heterogeneous catalyst follows:
i. Adsorption of reactants onto catalyst surface
ii. Chemical reaction of reactants
iii. Desorption of reaction products.
In a chemical reaction, the catalyst must provide an alternative pathway that reactants can follow to yield products. If the pathway provided by the catalyst has lower energy barriers than those of the non-catalysed form of the same reaction, the presence of the catalyst would give significant improvements in the rate of the reaction .
The following discusses the properties of a catalyst and its support that may affect the energy barriers of a pathway provided during reaction.
Arguably, the most important physical characteristic of a heterogeneous catalyst is its surface area and porosity. The catalyst should be formed as “finely divided substrates or crystallites” with ample internal pores that are readily accessible. The specific surface area is the surface area of the catalyst divided by the mass of the sample. A high surface area results from many fine connected particles, and one gram of a catalyst support may have a surface area equivalent to that of a tennis court .
During a catalysed reaction, it is the metals surface that allows adsorption of reactants, and so the number of surface metal sites must be maximised to achieve a high catalytic activity. A supported metal particle that has a diameter of 2.5 nm has approximately 40 percent of its atoms on the surface . If the metal particles are able to fuse together then this would decrease the surface area provided for adsorption, so it is important that there is a separation preventing the formation of bulk metal.
In heterogeneous catalysis, adsorption is essential if the catalyst is to work effectively. However, if adsorption is too strong, the active sites on the catalyst would become blocked, and the reaction would no longer be catalysed. The ability of a metal in adsorbing and desorbing reactants and products is an important factor that must be taken into consideration when choosing a heterogeneous catalyst. There are a limited number of metals that function as effective catalysts .
Small metal particles are often unstable and susceptible to sintering (a process where the particles adhere together) at the temperatures that catalytic reactions operate. Therefore, the small metal particles that are used in a heterogeneous catalyst must be stabilised to prevent sintering. This is achieved by depositing the particles inside the pores of an inert catalyst support .
Many heterogeneous catalysts show optimum functionality when supported upon another material, often single- or mixed-metal oxides.
Supported catalysts consist of three key components: the active phase, the promoter and the support. The active phase is the only phase that directly takes part in a chemical reaction, and the catalytic effect is dependent on the presence of this phase. The promoter is a phase that adds enhancement or stability to the catalytic activity. The support is the material upon which the active phase is deposited.
In a heterogeneous catalyst, the active phase functions through its surface and so its exposure to reactants is important if it is to have a high catalytic activity. This is most effectively achieved by dispersing the active phase upon another material, namely the support, which allows the active phase to be composed of particles as small as 1 nm. Without the support, this is not possible. The mass of the active phase can range from less than 1 wt. % (e.g. platinum metals in reforming) to 50 wt. % (e.g. metal sulfide catalysts) .
Utilising a support has many advantages in catalysis. Efficient dispersion of the active phase can maximise the surface area covered, and reduce the amount of the phase required for its application. This can contribute greatly to reducing costs, a major factor taken into consideration in industry. The support provides “mechanical properties needed for a long-lasting operation (hardness, resistance to crushing or erosion)”  and also provides greater resistance to sintering.
Initially, the materials used as a support were naturally derived such as silicates, aluminium and magnesium hydroxysilicates and minerals. Since the 1940s, the catalyst manufacturing industry expanded and there was a need to develop catalyst supports that exhibited properties that could be adapted, because natural materials did not always display the ideal properties required .
Synthetic catalyst supports are manufactured from metal oxides which have high surface areas, are relatively cheap and exhibit excellent chemical and thermal stability. For example, an unsupported copper oxide catalyst is thermally unstable at high temperatures, but copper oxide-supported upon high surface area aluminium oxide (also known as alumina) is thermally stable . Theoretically, it is possible to use any material that displays a high surface area as a catalyst support, but ideally it should conform to the other criteria .
The support should have a porous structure that allows reactants and products to diffuse in and out of easily. The ability to form the support into different physical shapes must also be taken into account as this allows its use in a range of physical conditions. The ability of the support to form interactions with the active phase is important because it can provide stabilisation to the active phase if it has an unusual particle structure .
Catalysts used in processes involving hydrogen (e.g. hydrogenation, hydrodesulfurization, hydrodenitrogenation) use supports that have high surface areas, whereas low surface area supports are used in selective oxidation processes (e.g. olefin epoxidation). In cases where the active phase is too reactive, a nonporous low surface area support can be used to reduce the contact times of the reactants on the catalyst. In cases where the active phase has a moderate reactivity, a porous low surface area support is often utilised .
The nature and size of the pores of a catalyst support affect the catalytic activity greatly. An open pore structure ensures the best use of the available pore volume, but if molecular size and shape selectivity is required in the reaction, a bimodal pore distribution is recommended. This distribution consists of large channels connected by narrower regions, and only molecules of a certain size or shape may pass through. This type of distribution is observed in zeolite catalysts .
Exposure of a catalyst support to the atmosphere results in the adsorption of foreign molecules, which may assist in stabilising the surface. In many cases, these foreign molecules are molecular water, leading to the formation of surface hydroxyls . Heating to a temperature between 100 - 150 °C can cause desorption of these molecules , but some surface OH groups and coordinatively unsaturated oxygen ions may remain, the former acting as Brønsted acid sites and the latter as Lewis base sites. The presence of these sites is important for supported catalysts in that they provide anchoring sites for active phase compounds during the preparation of the catalyst .
Supported catalysts can be prepared by a range of methods. One method may be more efficient in preparing the desired catalyst over another, and this would depend on the properties that are required of the catalyst. The methods and their key steps are summarised below.
Impregnation: In this method the pores of the support are filled with a solution of the active phase, and the solvent is then evaporated. The pores can be filled either by spraying the support with the active species solution or by adding the support to the solution in known amounts. This is beneficial in that it prevents excessive use of the active species, which can often be costly. Drying ensures the solvent evaporates and this is carried out at high temperature. If the active phase consists of more than one metal catalyst it is important that selective adsorption does not occur; this may lead to the formation of an undesired concentration of the metals in the prepared catalyst. The impregnation method is commonly used when preparing small amounts of a catalyst .
Adsorption: Metal salts or metal ion species can be selectively adsorbed onto a support by either physisorption or by the formation of chemical bonds with the active phase on the support. This method is suitable when control is required over the amount of active phase present and dispersion over the surface of the support, and as such is widely used in industry when preparing supported catalysts .
Coprecipitation: Salt solutions of the active phase and the catalyst support are prepared and a precipitating agent (e.g. NaOH or NaHCO3) is added. Hydroxy salts or hydroxides precipitate and these are filtered off. During the drying and calcination stages carbon dioxide and water are removed, and during the reduction stage oxygen is removed. The final product is a porous catalyst that exhibits a high surface area. This procedure is favoured when the aim is to obtain a material with the optimum catalytic activity per unit volume and materials are cheaply available. It is important to keep the initial solution homogeneous so that both the active phase and the support precipitate simultaneously .
Deposition: This is the process of laying down an active phase on the surface of a support. It is achieved by condensing a metal vapour over an agitated support, a process known as sputtering. A drawback of this method is the requirement of a high vacuum, which may limit its use to only preparing ‘model' catalysts .
Chemical Vapour Deposition (CVD): Similar to the Deposition method, a support is coated with a vapour, but in this technique the vapour is either a volatile inorganic or organometallic compound .
These experimental methods have been utilised in the preparation of catalysts, and exact techniques are described when required. In some cases, literature papers concerning essentially the same experiment have prepared catalysts by different methods, which have led to interesting results. These results are discussed in the sections to follow.
Global energy consumption continues to increase every year, with the majority of energy provided by non-renewable sources such as coal, oil and natural gas. Over the past 50 years, oil has been the world's dominant energy resource. In 2008, 35% of the world's total energy consumption was sourced from oil , and although this is a small decrease over the past decade, the importance of this fuel cannot be emphasised enough.
Oil is composed of hydrocarbon chains of different lengths, and along with these very useful products there are other non-hydrocarbon compounds which must be removed prior to use. These are compounds of sulfur, nitrogen and oxygen, and metals and salts.
Sulfur, which is present in the highest proportion of all the non-hydrocarbon compounds, can cause serious problems if it is not removed from oil. During combustion, it reacts with oxygen to form sulfur dioxide, a major contributor to pollution. Sulfur species, even at very low concentrations, can poison the metal catalysts rendering them ineffective. In catalytic converters, sulfur chemisorbs onto the active sites of catalysts and reacts, forming strong sulfur-metal bonds . This reduces the life of the catalyst drastically.
Authorities worldwide have imposed environmental legislation to reduce the harmful emissions from fuels, which has identified the need to use ‘cleaner' fuel technology. This demand has prompted studies into more efficient refining processes that remove these non-essential compounds from oil.
A key step in the refining of oil involves the removal of sulfur and its compounds (e.g. thiols, organic sulfides and thiophenes), a process known as hydrodesulfurization. It is a catalysed reaction and can occur via two different pathways, as illustrated in 2.
DDS: Direct route; HYD: Hydrogenative route
The HYD route shows the hydrogenation of carbon-carbon double bonds followed by removal of sulfur, whereas the DDS route shows the removal of sulfur without hydrogenation of any carbon-carbon double bonds.
In the direct route a displacement occurs, where a hydrogen atom replaces the sulfur atom in the hydrocarbon chain without hydrogenation of any carbon-carbon double bonds. In the hydrogenative route, at least one aromatic ring adjacent to the sulfur-containing ring is hydrogenated, followed by removal and replacement of the sulfur atom by hydrogen. It is also possible for the sulfur atom to be removed before hydrogenation of the aromatic ring .
In the hydrogenative route, hydrogenation of a carbon-carbon double bond destabilises the aromatic ring, which weakens the carbon-sulfur bond. This facilitates the insertion of the metal (catalyst) atom which results in the removal of the sulfur atom.
Insertion of the metal (catalyst) atom in the carbon-sulfur bond in the direct route occurs because the metal-sulfur bond that forms is energetically favourable. The sulfur is removed in the form of hydrogen sulfide .
The most commonly used catalysts in hydrodesulfurization are cobalt-molybdenum or nickel-molybdenum supported upon alumina. The catalytic phase is molybdenum, and cobalt and nickel are promoters. Molybdenum is present in a sulfide form, MoS2 . Tungsten and ruthenium catalysts are also reported to have been used [10, 11].
It is understood that at the edges and corners of MoS2 crystallites the molybdenum centre is able to stabilise a coordinatively unsaturated site. The sulfur atom of the sulfur-containing hydrocarbon binds to this site, and unbinds from the hydrocarbon. The coordinatively unsaturated site is no longer vacant, so hydrogen (H2) reacts with the bound sulfur atom, yielding hydrogen sulfide. This leaves the site vacant and available for further reaction, and the process repeats .
The industrially used alumina support has a high surface area, typically of the order of 230 m2/g , and this results in a very efficient dispersion of the active species. In order to meet the stringent regulations on sulfur content in fuels, research is underway into higher activity catalysts.
An alternative to researching new catalysts is to vary the catalyst support, which plays an important role by modifying the active species and taking part in the hydrodesulfurization reaction by acting as a co-catalyst [9, 14]. Studies have shown that some supports promote excellent catalytic activities, but are limited by their low surface area, which reduces the effect of increased activity.
Over the past twenty years, many studies have been conducted using mixed-oxides, investigating their use as supports in hydrodesulfurization catalysts . Investigations into single oxides have been less common.
In recent years, titanium dioxide-supported catalysts have been investigated. They have shown higher activity than those supported on alumina : initial studies report that molybdenum catalysts supported on titanium dioxide are 4.4 times more active than those supported on aluminium oxide . The following section describes the similarities and differences between these studies, focussing on the experimental aspects. For simplicity, results are summarised within a table and a discussion follows.
Despite the increased activity of titania in comparison to conventional alumina supports, its typical surface area is much smaller: the paper by Dzwigaj et al. reports the surface area to be as low as 50 m2/g , although many studies have used titania exhibiting a much larger surface area, as described in the following sections.
The study by Dzwigaj et al. focuses on a conventional titania support (72 m2/g) and a high surface area support prepared by a novel method developed by the Chiyoda Corporation that exhibits a higher surface area (120 m2/g). The Chiyoda support provides the opportunity to increase the molybdenum loading to 8 - 12 wt. %, which is comparable to that of alumina-supported catalysts. Conventional titania used in previous studies limited the loading to 6 wt. %. Molybdenum catalyst was prepared by the impregnation method using ammonium heptamolybdate as the precursor. Different concentrations of this solution were used to obtain catalysts containing varying amounts of molybdenum; 1.5, 2.5, 4.0 and 5.0 Mo atoms/nm2. Higher molybdenum loadings (7 - 20 Mo atoms/nm2) were achieved by completing two successive impregnations i.e. a sample containing 7 Mo atoms/nm2 was prepared by a second impregnation of a sample already containing 4 Mo atoms/nm2. There is no mention in the article that a promoter has been used. The catalysts were reacted with a H2/H2S mixture, which yielded the molybdenum sulfide species. The catalytic properties were investigated in the hydrodesulfurization of dibenzothiophene, a thiophene derivative that is commonly found in oil.
A study by Gulková et al.  further investigated the effect of a titanium dioxide support on hydrodesulfurization activity. The catalysts were prepared containing molybdenum trioxide, followed by impregnation with ammonium heptamolybdate. Similarly to the study by Dzwigaj et al. the catalysts were then reacted with a H2/H2S mixture. Although the authors make no mention of a molybdenum sulfide species, it is assumed that this is the molecule yielded from the reaction. Three titania supports with surface areas of 140, 230 and 407 m2/g were used and the molybdenum loading was varied from 2 - 32 wt. %. The catalytic properties were investigated in the hydrodesulfurization of thiophene and were compared against a commercial alumina catalyst (15 wt. % molybdenum loading; surface area: 210 m2/g).
Escobar et al.  prepared a molybdenum catalyst supported on nano-structured titanium dioxide in the same method as used by Dzwigaj et al. and Gulková et al., except that there was a mention of the cobalt promoter. Following the impregnation with ammonium heptamolybdate, and prior to reaction with a H2/H2S mixture, an aqueous solution of cobalt acetate was used to impregnate cobalt onto the catalyst. Catalytic activity was measured in the hydrodesulfurization of dibenzothiophene.
The impregnation method was also used in a study by Cortés-Jácome et al. . Differences in this study in comparison to those described above are the use of nanotubular titania, and the investigation of the effect of pH on catalytic activity. The molybdenum loading was 3 Mo atoms/nm2 and a cobalt promoter was used. Impregnation with ammonium heptamolybdate was carried out on three samples under different pH: natural (5.6), basic (10.0) and acidic (~1.8). Reaction with a mixture of H2/H2S yielded molybdenum sulfide. Catalytic activity was investigated in the hydrodesulfurization of dibenzothiophene.
Nanotubular titania was also studied by Toledo-Antonio et al. . Catalysts were prepared by contacting nanotubular titania with ammonium heptamolybdate at pH 5.6. The amount of molybdenum atoms in each sample was varied; 3, 4 and 5 atoms/nm2. Cobalt was impregnated on to the samples using a cobalt acetate solution and reaction with a mixture of H2/H2S yielded the molybdenum sulfide active phase. The catalysts were tested in the hydrodesulfurization of dibenzothiophene.
The studies mentioned thus far have all used the impregnation method when preparing catalysts. A different approach has been taken by Inoue et al.  where a multi-gelation method was employed. In this procedure an inorganic oxide gel is produced by swinging the pH of the solution several times in sequence. Titanium dioxide was prepared by this method using titanium chloride and ammonia solutions. Hydrodesulfurization activity was investigated in the hydrodesulfurization of oil containing 1.3 wt. % sulfur.
Although studies have shown that using a titanium dioxide support provides the opportunity to achieve higher catalytic activities, there may also be other factors which can affect activities. This was the motivation for a study conducted by Roukoss et al.  where the effect of the method used to impregnate cobalt onto titania-supported catalyst was investigated. Ammonium heptamolybdate was used to impregnate molybdenum onto titania to 2.5 atoms/nm2. Cobalt (II) nitrate hexahydrate was used to impregnate cobalt onto samples of supported molybdenum oxide and molybdenum sulfide (prepared by reaction with H2/H2S mixture). Another sample of molybdenum sulfide was impregnated with cobalt using cobalt acetylacetonate. Catalytic activity was determined by the hydrodesulfurization of thiophene.
In these studies, surface areas were determined by nitrogen physisorption at -196 °C using the Brunauer-Emmet-Teller (BET) equation.
The experimental details have shown that most studies have used similar preparation techniques. The results of these studies are summarised in Table 1.
Table 1: Summary of results obtained in articles discussed
TiO2 surface area (m2/g)
Pore volume (cm3/g)
Pore diameter (nm)
Mo loading (wt. %)
Pseudo-rate constant (m3/kgMo s)
Dzwigaj et al. (2001)
Gulková et al. (2008)
Various; ranging from 2 - 32
Escobar et al. (2005)
2.36 × 10-4
19.88 × 10-4
Cortés-Jácome et al. (2007)
15.8 × 105
126.3 × 10-5
20.0 × 105
155.9 × 10-5
5.5 × 105
46.7 × 10-5
Toledo-Antonio et al. (2009)
2.00 × 103
15.6 × 105
3.58 × 103
26.5 × 105
6.44 × 103
43.0 × 105
1.11 × 103
7.3 × 105
2.71 × 103
17.6 × 105
HDS: hydrodesulfurization; DBT: dibenzothiophene; - : no data provided
Table 2: Results obtained from study by Inoue et al. 
No. Of gelation
TiO2 surface area (m2/g)
Pore volume* (cm3)
Pore diameter* (nm)
Inoue et al. (2004)
*Values of pore volume and pore diameter are estimations from graphical representations provided in the paper. Actual s have not been provided.
Table 3: Results obtained from study by Roukoss et al. 
TiO2 surface area (m2/g)
Mo loading (wt. %)
Roukoss et al. (2009)
CoMoS-O: addition of cobalt (II) nitrate hexahydrate on supported molybdenum oxide
CoMoS-S: addition of cobalt (II) nitrate hexahydrate on supported molybdenum sulfide
CoMoS-acac: addition of cobalt acetylacetonate on supported molybdenum sulfide
Comparison of the data obtained in each of the studies, shown in Table 1, shows that there is no relationship between the titanium dioxide surface areas, pore volume, pore diameter or molybdenum loading. Conversely, the multi-gelation method used by Inoue et al. (Table 2) shows that as the number of gelations increased from one to six, the surface area decreased, while both the pore volume and pore diameter increased. It appears from these results that there is better control over the physical properties of the titanium dioxide support when prepared using this method rather than the commonly used impregnation method. The paper explains that the physical properties depend on the size, shape and aggregation of the primary and secondary particles. It is also stated that the multi-gelation method has achieved higher surface areas of titania, however, this is also commonly observed using the impregnation method. The authors comment that the high surface area “may be due to the growth of crystal particles with gelation times” .
The first studies investigating the use of a titania support had surface areas comparatively smaller than those prepared in subsequent studies. Dzwigaj et al. reports the highest achieved surface area of a fully prepared catalyst (ready for hydrodesulfurization) as 110 m2/g, while later studies have surface areas typically above 200 m2/g. However, at the time of the study by Dzwigaj et al. the newly achieved surface area by the Chiyoda Cooperation was considered to be a great achievement because it was nearly twice that of conventional titania (120 m2/g cf. 72 m2/g). It was shown that up to a molybdenum loading of 7 wt. % the catalytic activities (no s provided, only graphical representation) of the conventional support and the Chiyoda support were similar, but for loadings above 7% the Chiyoda support achieved higher activities. The authors attribute this effect to the higher surface area of the Chiyoda support and the two-step impregnation method which allowed for the dispersion of a high molybdenum amount.
The high surface area (407 m2/g) achieved by Gulková et al. is unusually high considering the same preparation method was used as other studies, which have not achieved such high surface areas. The authors expected catalytic activity to increase with higher molybdenum loadings, however this was not observed. The molybdenum loading was varied from 2 - 32 wt. % but the highest activity was observed at a loading of approximately 20 wt. %, regardless of the surface area of the support. It was also expected that the higher the surface area of the support, the higher the catalytic activity. Again, this was not observed; the 407 m2/g support had the highest activity, while the 230 m2/g had the lowest activity.
The authors assume that these results are due to differences in impurities of the titania supports used in the investigation. The credibility of this article is debatable as there has been very little reasoning for the results obtained.
It is interesting to note that during catalyst preparation, the surface area of the support has reduced from its initial surface area when in a powdered form. For example, in the study by Escobar et al. the dry titania powder exhibited a surface area of 343 m2/g, which decreased to 225 m2/g when fully prepared and impregnated with molybdenum. This reduction was also observed by Cortés-Jácome et al. where the surface area decreased from 369 m2/g to 204 - 229 m2/g (dependent on the pH during preparation). It is understood that the titania supports undergo structural changes during catalyst preparation, resulting in these changes .
Cortés-Jácome et al. prepared a catalyst sample at a natural pH (5.6). It was found to exhibit a slightly larger surface area than the catalysts prepared at basic and acidic pHs. The physical properties of this sample have exactly the same physical properties of a sample prepared by Toledo-Antonio et al. whose experiment was also conducted at pH 5.6.
Roukoss et al. (Table 3) investigated the effect of varying the compound used to impregnate the cobalt promoter onto molybdenum. In addition to the titania supports, an alumina support and silicon dioxide support were studied. Results showed that for the titania support, the use of different cobalt compounds had no effect on the catalytic activity, but differences were observed with the alumina and silica support.
Some, but not all, of the studies have provided rate constants for the catalysts investigated. Although the units that these constants are given in are the same, there are large differences in the actual rate constant values. The studies have used the same method of preparation and rate constants were determined for the hydrodesulfurization of dibenzothiophene. The conditions under which the reactions were carried out are also the same: 320 °C, 1000 rpm mixing speed in a batch reactor and n-hexadecane solvent. Despite this, in some cases the differences in rate constant values are of many magnitudes, which raise concern. With such values, it is not possible to make comparisons between them, given that it is expected that the rate constants should be within close range of each other.
Within a study, the only comparisons of rate constants are between the values obtained for an alumina-supported catalyst and a titania-supported catalyst. There are no references or comparisons to previous studies. Apart from stating that the rate constant for the titania-supported catalyst is higher than that for the alumina-supported catalyst, there are no other comments. It must be remembered that the objective of these studies was to create a high surface area titania support, but just as important is to show that the supports are capable of yielding fast reactions. It is no use having high surface area supports which give largely differing rates of reaction.
The studies reviewed have shown that titanium dioxide is a promising candidate as a catalyst support in hydrodesulfurization. High surface areas have been prepared, and their characterisation has shown that they display ideal physical properties required of a catalyst support. The commonly used impregnation method does not appear to display control over the physical properties of the supports, however Inoue et al. has shown that the multi-gelation method can be used to prepare supports with particular physical properties.
Future studies should encompass the positive outcomes of studies already conducted; e.g. a study that investigates the multi-gelation method at natural pH and varying the method used to impregnate cobalt (or a suitable promoter) may well produce catalysts with activities much higher than those currently achieved.
A key objective should also concentrate on achieving consistent rate constants in hydrodesulfurization reactions employing titania as a support, since those currently achieved do not appear to show much correlation.
Polymer Electrolyte Membrane Fuel Cells
Hydrodesulfurization has been demonstrated as a very useful reaction in the process of removing sulfur compounds from oil. However, the decline in the quantity of oil remaining on Earth has prompted large investments into researching alternative energy production. There is much research within the field of renewable energy, which is a required investment into the long-term future. Many renewable sources already exist, such as wind and solar power. However, the availability of the “fuel” behind these sources is not consistent, and so their reliability cannot be guaranteed.
Therefore, although naturally available renewable energy sources have proved to be a success, they make up only a small percentage of global energy production.
In order to reduce the use of non-renewable sources, renewable technologies must be developed that have greater reliability. One such technology is fuel cells. These are, as defined by William Grove in 1839, “an electrochemical “device” that continuously converts chemical energy into electric energy (and some heat) for as long as fuel and oxidant are supplied” . Fuel cells are an example of a thermodynamically open system; the fuel is provided from an external source, whereas in a thermodynamically closed system, such as a battery, the fuel is stored within the system. Fuel cells are a promising technology considering they have efficiencies higher than conventional energy systems, as show below in Table 4.
Table 4: Efficiencies of energy systems. Adapted from A. Kirubakaran et al. (2009) 
29 - 42%
6 - 19%
40 - 60%
In the reaction that occurs within a fuel cell, the only by-product is water, making them an environmentally friendly and very attractive technology. Common types of fuel cells include polymer electrolyte (PEMFCs), solid oxide (SOFCs), direct methanol (DMFCs) and molten carbonate (MCFCs) fuel cells. Different fuels are used in each type of fuel cell: PEMFCs use hydrogen, SOFCs run on petroleum products such as hydrogen, methanol and butane, DMFCs are fuelled by methanol and MCFCs use hydrogen as the fuel but have a molten carbonate electrolyte.
All fuel cells operate in the same manner. At the anode, an oxidation occurs where the fuel is oxidised into electrons and protons (equation (1)), and at the cathode, a reduction occurs, where oxygen is reduced to oxide species (equation (2)). Depending on the electrolyte present, oxide ions or protons move through the ion-conducting electrolyte and combine with protons or oxide ions, respectively, to produce electricity and water. These two processes are summarised by the electrochemical equations:
At the anode: 2H2 ' 4H+ + 4e- (1)
At the cathode: 4H+ + 4e- + O2 ' 2H2O (2)
In addition to the different fuels used, fuel cells operate under different conditions and yield slightly different efficiencies, as summarised in Table 5.
Table 5: Operating temperatures and efficiencies of different fuel cells. Adapted from A. Kirubakaran et al. (2009) 
Operating temperature (°C)
50 - 100
800 - 1000
60 - 200
40 - 50%
Although efficiencies of fuel cells are only slightly different, the operating temperatures vary greatly. This means that the running cost is very high for a fuel cell that operates at a high temperature, and it would also require a longer start-up time compared to one that operates at a low temperature. For this reason, PEMFCs, which have a short start-up time, are promising candidates for transportation applications .
3 illustrates the components of a PEMFC. The polymer electrolyte membrane (PEM) is a key part and it is made by substituting hydrogen atoms in polyethylene for fluorine atoms, which yields polytetrafluoroethylene, and then adding sulfonic acid. The final product is a sulfonated tetrafluoroethylene membrane that provides acid sites on sulfonic acid groups. Protons are able to pass from one acid site to another and this forms the basis of proton exchange in PEMFCs. The membrane material was developed by the DuPont Company, and is give the name Nafion®.
The polymer membranes are ideal for use in fuel cells because they display a number of ideal characteristics. The membrane resists chemical attack, has the ability to absorb a lot of water, can be manufactured into very thin films, and carbon and fluorine bonds in the membrane make it inert and durable .
The polymer membrane separates the anode and cathode, and this structure is known as the membrane electrode assembly (MEA) and is the central part of the fuel cell. Both the anode and cathode are coated in a catalyst, which increases the rate of the oxidation and reduction reactions that occur.
The flow plates deliver a flow of reactant gases (oxygen and hydrogen) to the MEA via the flow channels. The channels aid in distributing the gases to the electrodes, and they collect any resulting current so they must be made of a material that has good electrical conductivity. Carbon is often a popular choice .
Water and heat are released from vents on an opposing side of the fuel cell, so that they do not interfere with delivery of the reactant gases.
A single PEMFC provides a voltage of approximately 0.7V, so to achieve the required voltage a number of single cells are connected in series. This forms a fuel cell stack.
When constructing the fuel cell, the choice of material for the catalyst and catalyst support must be carefully considered. Commonly used are platinum-based catalysts  supported on high surface area carbon . The presence of a high surface area support has great importance in enhancing the overall rate of reaction, as discussed earlier in this review. Carbon displays a large surface area, has a high electrical conductivity, and its surface has an adequate pore structure . However, there are associated problems with the carbon support, prompting research into other possible catalyst supports.
Electrochemical oxidation in a fuel cell can cause corrosion. Carbon does not exhibit a satisfactory resistance to corrosion, resulting in poor electrode stability. As a consequence, platinum catalyst particles that are supported upon carbon are sintered and released from the carbon support , as was shown in a study conducted by Wang et al. . Sintering reduces the effective surface area of the platinum catalyst, which in turn would reduce the catalytic activity. Replacement of the components in a fuel cell is expensive, therefore increasing the running cost.
Research has been conducted investigating whether inorganic metal oxides are able to act as robust catalyst supports with corrosion-resistant properties. Traditional metal oxides have electrical insulating properties at temperatures below 200 °C , making them unfavourable for use since this is double the typical operating temperature of PEMFCs.
One inorganic metal oxide that has received much attention in terms of its catalyst support properties is titanium dioxide. It has an excellent resistance to corrosion in electrolytic solutions  and high stability in oxidative and acidic environments. Huang et al.  reports that the low electrical conductivity of titanium dioxide prevents its use in fuel cells, but then investigates its use as a support material, which is slightly contradictory.
Gustavsson et al.  and von Kraemer et al.  have conducted studies investigating the activity of a platinum catalyst supported upon titanium dioxide. Gustavsson et al. prepared individual thin films of platinum and titanium which were heated to evaporation and deposited onto Nafion® membranes in a process known as thermal evaporation. The titanium films were oxidised by venting to atmospheric pressure. The ordering of the thin films was varied; some samples had a platinum layer facing the gas (oxygen) side while others had the titanium dioxide layer facing the gas side. Film thickness was either 1.5 or 3.0 nm. Electrochemical activity was measured as current density against a hydrogen anode.
Von Kraemer et al. prepared supports containing carbon in addition to titanium dioxide. The carbon source was Vulcan XC-72 (a commercially available carbon black) and the ratio of titanium dioxide to carbon was varied to determine whether electrode performance increased in addition to the improved stability provided by titanium dioxide. The catalysts were prepared by reacting tetrakisdecyl ammonium bromide and chloroplatinic acid, yielding platinum nanoparticles. These were deposited onto the catalyst support material and electrodes were prepared by adding a Nafion® solution. Thermal stability tests were conducted to determine the stability of the prepared catalysts over long periods of time (~5000 hours).
Huang et al.  do not mention the use of a Nafion® membrane in their study. The electrode was prepared by reacting sodium borohydride, chloroplatinic acid and a solution of titanium dioxide and sodium dodecyl sulfate. The resulting solution was stirred for six hours to allow adsorption of platinum particles onto the support. Polarisation curves were measured for the samples, and an accelerated stress test (as suggested for PEMFCs by the U.S. Department of Energy) was conducted to determine the electrochemical stability and performance of the catalysts. The cell potential was held at 1.2 V over a 200 hour period for the titania-supported catalyst, and an 80 hour period for the carbon-supported catalyst.
Rajalakshmi et al.  predicted that the presence of titanium dioxide in the platinum lattice would prevent agglomeration of platinum particles, provide stability against corrosion, control the nanostructure of the catalyst and effectively disperse platinum atoms on the clusters. Titanium dioxide, prepared by the sol-gel process, was impregnated with chloroplatinic acid and then reduced with sodium borohydride and sodium hydroxide. By increasing the concentration of chloroplatinic acid it was possible to achieve different platinum loadings on the support.
Oxygen reduction reaction plots in the study by Gustavsson et al. showed that when the platinum layer was facing the gas side, the performance of the cathode increased in comparison to a platinum layer deposited directly onto the membrane. The authors state that the increase in performance resulted from a better dispersion of platinum on titanium dioxide and an enhanced proton conduction through the titanium dioxide layer.
When titanium dioxide was facing the gas side, the performance reduced, and this showed that the placement of this layer was important.
Thermal stability tests in the study by von Kraemer et al. showed that the commercial carbon support (containing 20 wt. % platinum) began to degrade after 2000 hours at 210 °C, whereas the titania support containing carbon (50 vol. %) did not display degradation and was highly stable throughout. To show that the presence of the platinum catalyst played a role in the degradation of the commercial carbon support, a thermal stability test was conducted on the as-obtained Vulcan XC-72 carbon material. Throughout the test, the carbon remained stable and showed no signs of degradation. The excellent thermal stability of the titania-carbon support was attributed to a reduced direct contact between platinum and carbon. It was also found that this support presented improved cathode performance when the amount of platinum deposited and the carbon fraction were increased.
Similar thermal stability tests by Rajalakshmi et al., whose study used conventional platinum-carbon and recently synthesised platinum-titania catalysts showed thermal degradation of the platinum-carbon electrode starting at 350 °C and at 450 °C for the titania-supported catalyst. It may be determined that under normal operating conditions (50 - 100 °C) a titania support in a PEMFC should not degrade since the temperature should never reach such extremities.
Results obtained by Huang et al. are in agreement with those described above; after the cell potential was held at 1.2 V for a 200 hour period, the titania-supported catalyst showed very little decrease in performance, whereas the carbon-supported catalyst showed a significant decrease after only 50 hours. This decrease was due to carbon corrosion and detachment and agglomeration of platinum particles, as was predicted by Rajalakshmi et al.
The studies discussed so far have investigated the effect of substituting carbon support with a titania support, or the effect of a mixed titania-carbon support.
In studies by Chhina et al. [36, 37] and in another by Park et al. , the effect of doping the titania support with niobium was investigated. The doped supports were prepared by hydrothermal synthesis and then placed in a solution containing deionized water and platinum chloride. The resulting mixture was reduced using sodium borohydride solution. To compare catalytic activity, a carbon-supported catalyst was prepared using the same method.
Chhina et al. measured the stability of a catalyst in a potential hold test. The cell was held at 1.4 V for 20 hours, and the catalytic activity was measured both before and after the test. The doped titania sample performed slightly worse than the platinum-carbon sample before the test, but after the test the performance of the platinum-carbon sample dropped significantly, whereas only a small decrease was observed in the doped titania sample. Platinum surface area did not change for the doped titania sample, but there was a significant decrease with the platinum-carbon sample after the test.
Comparison of catalytic activity in the study by Park et al. was by oxygen reduction reaction plots. The niobium-doped supported catalyst showed a much higher reduction current which meant it had an excellent catalytic activity in comparison to the carbon supported catalyst. The authors have reasoned this due to the “superior dispersion of Pt [platinum] nanoparticles on Nb-TiO2” . In addition, it was suggested that an interaction between the support and the metal catalyst may partly be responsible for the increased activity.
Promising efficiencies obtained using fuel cells have prompted much research into developing this technology further for large-scale energy production. However, the need to replace degraded carbon supports and sintered platinum catalysts has made PEMFCs an expensive technology, which when compared to conventional energy sources, is not economically viable. Titanium dioxide has shown excellent stability in addition to promoting high catalytic activity, making it a suitable alternative to carbon.
Stability tests of titanium dioxide are within good agreement, and catalytic activities have achieved acceptable levels. The studies have shown that titanium dioxide is economically viable; however, considering this particular area of heterogeneous catalysis is in its infancy stage, further research exploring other possibilities that may yield better results is required.
Water-Gas Shift Reaction
The efficiencies obtained using fuel cells have been discussed in detail, and it is well known that they are a very promising technology. The majority of fuel cells are able to operate on hydrogen fuel, which has been identified as the ideal energy carrier. Combustion of hydrogen is an environmentally clean process; it creates neither greenhouse gases nor pollutants. The energy quantity produced by hydrogen is far superior to that of fossil fuels; the energy per unit mass is nearly three times that of petrol, and nearly seven times that of coal .
However, the major barrier in the development of fuel cells has been hydrogen storage. At ambient conditions hydrogen is a gas and therefore has a low volumetric density . It is possible to compress hydrogen under high pressure; however this poses safety issues in portable applications such as vehicles.
This major drawback has prevented the commercialisation of fuel cells. An alternative solution is to convert readily available fuels such as hydrocarbons into hydrogen on-board a vehicle, which would eliminate the problems associated with storage.
In fuel cells, hydrogen can be produced via the steam reforming method. In a reformer, steam is reacted with natural gas to produce hydrogen. A by-product of the reaction is carbon monoxide, which can poison the fuel cell, reducing its performance drastically. Carbon monoxide preferentially adsorbs to the platinum catalyst at the anode, blocking active sites. Baschuk et al.  found that the poisoning effect can be mitigated by operating the fuel cell at a higher temperature, but for PEMFCs this is not favoured if they are to be used in transportation. Therefore, another method to remove carbon monoxide must be utilised.
A process known as the water-gas shift reaction can be used in fuel cells to remove carbon monoxide. The reaction yields hydrogen which is advantageous since it can be used as an additional fuel supply. Carbon monoxide and water are converted to carbon dioxide and hydrogen in the following reaction :
CO + H2O ' CO2 + H2 ΔH = -41.1 kJ/mol (3)
The reaction is catalysed by a noble metal catalyst supported on a metal oxide support  and it is understood that the catalyst and support operate in a bifunctional manner in which both phases participate.
For the reaction pathway there are two proposed mechanistic schemes:
(a) A redox or “regenerative” mechanism, where carbon monoxide adsorbed onto the metal surface is oxidised to carbon dioxide by labile oxygen of the support. Oxygen is then re-oxidised by water in a reaction that forms hydrogen [44, 45].
(b) An associative mechanism, where the adsorbed carbon monoxide interacts with terminal hydroxyl groups of the oxide support to form a formate intermediate that is decomposed to carbon dioxide and hydrogen [46, 47].
The water-gas shift reaction operates with two stages, each using a different catalyst. The first stage, the high temperature shift (HTS), operates at 350 °C with an iron oxide catalyst promoted on chromium oxide. The second stage, the low temperature shift (LTS), operates at approximately 200 °C with a copper catalyst supported on zinc oxide-alumina mixed support. The low temperature shift catalyst has low durability under the oxidising atmosphere in the presence of steam .
The complexity of the reaction, pyrophoric nature of the catalysts, and the high temperatures required (opposed to low temperatures in PEM fuel cells) have currently made this technology unfavourable for use in fuel cells [49, 50].
In the water-gas shit reaction, both reactants, carbon monoxide and water, must be activated. It is difficult to activate water due to its thermodynamic stability . Metals such as copper and iron undergo oxidation by water and therefore activate it, and this property is essential for any metal that may be considered as a potential catalyst in the reaction.
Many recent studies have focussed on new catalysts based on platinum, palladium, gold and rhodium, supported on oxide and mixed-oxide supports such as cerium dioxide , zirconium dioxide , cerium-zirconium mixed oxides , and titanium dioxide.
Platinum catalysts have achieved high activity at low temperatures, making them potential candidates. Studies have shown that for platinum-based catalysts the presence of the oxide support is essential to adsorb and activate water due to thermodynamic instability of platinum oxides under water-gas shift reaction conditions . Of the supports studied, cerium dioxide and titanium dioxide have shown the highest catalytic activities.
Gonazález et al.  report that the platinum catalyst supported on cerium dioxide deactivates over time due to: loss of the active metal surface, irreversible “over” reduction of cerium dioxide by hydrogen, and the formation of carbonate on the surface of the support. Similar effects have been observed with platinum-titania catalyst. To overcome these problems, the authors state that the platinum metal dispersion and the ability of the support to reversibly exchange oxygen must be stabilised. Some studies have investigated using cerium-modified titania supports to determine whether better stabilities are achieved.
Results from a study by Iida et al.  show that the rutile form of titanium dioxide had a relatively higher catalytic activity than the anatase form for the water-gas shift reaction under low temperature conditions. Catalysts were prepared by adding a solution of platinum chloride hexahydrate to the support powder, followed by evaporation. All catalysts had a platinum loading of 3 wt. %. Performance was measured using a conventional fixed bed flow reactor. The catalyst was reduced in a hydrogen stream at 500 °C then cooled in a nitrogen stream at 175 °C. Steam was supplied followed by carbon monoxide and gas composition was determined using a gas chromatograph. The results of the study showed that the reaction followed mechanistic scheme (b). It was suggested that the catalytic activity was largely affected by the interaction between the platinum and the titania support, and the difference in catalytic activity between the rutile and anatase form was explained by the degree of platinum dispersion.
Azzam et al.  prepared a platinum catalyst supported on titania by impregnation of titania with chloroplatinic acid to achieve a platinum loading of 0.5 wt. %. Results showed that the reaction pathway followed by the catalyst was a combination of schemes (a) and (b); that is an associative formate route with redox regeneration.
Liang et al.  prepared three-dimensionally ordered macroporous (3DOM) platinum-titania catalyst. A 3DOM material is characterised with pore size in several hundredths to several tenths of a nanometre, with ball-shaped pores that are closely packed and interconnected through small windows. A titanium dioxide sol was added to a template prepared from polystyrene beads. Platinum was loaded by the impregnation method using chloroplatinic acid. The authors report that the catalyst had exhibited very good catalytic performance for the water-gas shift reaction which was due to the macroporous structure.
Boccuzzi et al.  investigated catalysts consisting of gold supported on iron oxide and titanium dioxide. The iron sample was prepared by the deposition-precipitation method on iron hydroxide with gold tetrachloric acid and the titania sample was prepared by hydrolysis of titanium chloride. Both samples contained a gold loading of 3 wt. %. The authors report that although the activity of the two supports differ significantly, once gold is introduced both supports exhibit similar activities to those of the copper catalyst used in the water-gas shift reaction. The paper focuses on an FTIR study to understand why the two samples have the same activity in the water-gas shift reaction, and there is little mention of the suitability of titanium dioxide as a support. However, considering the activity is high, titania may well be a promising candidate. The authors state that the experimental results can be explained by mechanistic scheme (a).
Park et al.  have studied a mixed-oxide support consisting of cerium dioxide and titanium dioxide with gold, copper or platinum catalyst. A rutile titania sample was used, and cerium was deposited on its surface using metal evaporators. Further details regarding catalyst preparation are not provided. It was found that the catalytic activity was highest for the sample containing platinum catalyst, followed by that containing copper, and gold had the lowest activity, although all activities were very high compared to those used industrially. At low coverage of copper and cerium the catalyst was 8 - 12 times more active than the currently used copper-zinc oxide catalysts. The authors have attributed the very high activity to the unique surface properties of the mixed-metal oxide support.
Although the water-gas shift reaction has been known since 1780 and is successfully operating in industry, only recently since the developments in fuel cells has it become a focal point for research. The drawbacks of the catalysts used in the industrial process outweigh any benefits if they were used in the water-gas shift reaction in fuel cells, which is the reason behind the interest in their research.
It is widely accepted that the catalyst support plays a major role in the catalytic activity in the water-gas shift reaction. Platinum catalysts supported on titania have shown better performance than those currently used, however quantitative comparisons are not possible since studies have not reported catalytic activity values.
Although it is clear that the reduction of the catalyst support is essential, knowledge of the exact mechanism followed remains controversial. Results from studies by [44, 45, 53, 57, 59] are explained by a redox mechanism (scheme (a)) while results from [48, 52, 60] are explained by an associative mechanism (scheme (b)). Azzam et al.  has stated that the mechanism is a combination of both the redox and associative mechanisms.
Since the mechanism has been studied extensively, it shows that it has significant importance and must be considered when developing new techniques and methods to prepare catalysts for the water-gas shift reaction.
Therefore, future research should aim to clarify the exact scheme followed, which may aid in preparing an effective catalyst supported on titanium dioxide.
This review has discussed the use of titanium dioxide as a catalyst support with promising results. In the hydrodesulfurization process, titanium dioxide supports have promoted catalytic activities superior to those of conventional alumina-supported catalysts. This has shown that titanium dioxide is a promising candidate. However, the differences in catalytic activity values between different studies have raised concern and future studies must concentrate on achieving consistent values.
The global need for energy and the declining quantities of fossil fuels remaining have prompted much research into renewable energy sources. Fuel cells are a viable solution; however their current limitations prevent global commercialisation. Electrochemical oxidation in fuel cells causes corrosion, and currently used carbon supports have not displayed adequate resistance. The high stability of titanium dioxide in oxidative and acidic conditions allows it to effectively resist corrosion, and many studies have investigated its use as a potential alternative to carbon. This research has brought the development of fuel cells once step closer to commercialisation, but another limitation poses a significant problem.
Providing hydrogen, the fuel in fuel cells, is currently an issue. Storing the gas at ambient conditions has not been as simple as it appears which has meant alternative solutions should be found. The steam reforming method produces hydrogen but also carbon monoxide which poisons catalysts, so it is essential that it is removed. The water-gas shift reaction is used to remove carbon monoxide, but its complexity and undesired operating conditions has led to research with titanium dioxide acting as the catalyst support, in which initial studies have displayed pleasing results.
This review has reported on the research conducted using titanium dioxide as a catalyst support. Relevant studies have been discussed and it is clear that this remarkable ceramic material is more than just an effective catalyst. Further research may well show new directions in heterogeneous catalysis where the properties of titanium dioxide would be of beneficial use.
1. M. E. Davis and R. J. Davis, "Heterogeneous catalysis," Fundamentals of Chemical Reaction Engineering, McGraw-Hill Higher Education, New York, NY, 2003, pp. 133 - 183.
2. A. Contescu and C. Contescu, "Oxides and related surfaces as catalyst supports," Encyclopedia of Surface and Colloid Science, P. Somasundaran (Editor), vol. 7, CRC Press, New York, 2006, pp. 4405 - 4414.
3. P. Atkins, D. Shriver, T. Overton, J. Rourke, M. Weller and F. Armstrong, "Heterogeneous catalysis," Inorganic Chemistry, Oxford University Press, Oxford, 2006, pp. 694 - 707.
4. I. Chorkendorff and J. W. Niemantsverdriet, "Catalyst supports and preparation of supported catalysts," Concepts of Modern Catalysis and Kinetics, Wiley-VCH, Weinheim, 2003, pp. 189 - 198.
5. G. J. K. Acres, A. J. Bird, J. W. Jenkins and F. King, "The design and preparation of supported catalysts," Catalysis, C. Kemball and D. A. Dowden (Editors), vol. 4, Royal Society of Chemistry, London, 1981, pp. 1 - 30.
6. BP, "Statistical Review of World Energy June 2009," Beacon Press, London, 2009.
7. J. K. Dunleavy, Sulfur as a catalyst poison, Platinum Metals Review 50 (2006), no. 2, 110.
8. D. Zuo, D. Li, H. Nie, Y. Shi, M. Lacroix and M. Vrinat, Acid-base properties of NiW/Al2O3 sulfided catalysts: Relationship with hydrogenation, isomerization and hydrodesulfurization reactions, Journal of Molecular Catalysis a-Chemical 211 (2004), no. 1-2, 179-189.
9. I. Mochida and K. H. Choi, An overview of hydrodesulfurization and hydrodenitrogenation, Journal of the Japan Petroleum Institute 47 (2004), no. 3, 145-163.
10. A. Spojakina, E. Kraleva, K. Jiratova and L. Petrov, TiO2 supported iron-molybdenum hydrodesulfurization catalysts, Applied Catalysis a-General 288 (2005), no. 1-2, 10-17.
11. R. R. Chianelli, G. Berhault, P. Raybaud, S. Kasztelan, J. Hafner and H. Toulhoat, Periodic trends in hydrodesulfurization: In support of the Sabatier principle, Applied Catalysis a-General 227 (2002), no. 1-2, 83-96.
12. M. Daage and R. R. Chianelli, Structure-function relations in molybdenum sulfide catalysts - the rim-edge model, Journal of Catalysis 149 (1994), no. 2, 414-427.
13. S. Dzwigaj, C. Louis, A. Breysse, M. Cattenot, V. Belliere, C. Geantet, M. Vrinat, P. Blanchard, E. Payen, S. Inoue, H. Kudo and Y. Yoshimura, New generation of titanium dioxide support for hydrodesulfurization, 3rd International Conference on Environmental Catalysis, 2001, pp. 181-191.
14. S. K. Maity, M. S. Rana, S. K. Bej, J. Ancheyta, G. M. Dhar and T. Rao, Studies on physico-chemical characterization and catalysis on high surface area titania supported molybdenum hydrotreating catalysts, Applied Catalysis a-General 205 (2001), no. 1-2, 215-225.
15. G. M. Dhar, B. N. Srinivas, M. S. Rana, M. Kumar and S. K. Maity, Mixed oxide supported hydrodesulfurization catalysts - a review, Catalysis Today 86 (2003), no. 1-4, 45-60.
16. J. Ramirez, S. Fuentes, G. Diaz, M. Vrinat, M. Breysse and M. Lacroix, Hydrodesulfurization activity and characterization of sulfided molybdenum and cobalt molybdenum catalysts - comparison of alumina-supported, silica alumina-supported and titania-supported catalysts, Applied Catalysis 52 (1989), no. 3, 211-224.
17. D. Gulkova, L. Kaluza, Z. Vit, J. Horacek, E. Machackova and M. Zdrazil, High surface area MoO3/TiO2 hydrodesulfurization catalysts, Reaction Kinetics and Catalysis Letters 94 (2008), no. 2, 219-226.
18. J. Escobar, J. A. Toledo, M. A. Cortes, M. L. Mosqueira, V. Perez, G. Ferrat, E. Lopez-Salinas and E. Torres-Garcia, Highly active sulfided CoMo catalyst on nano-structured TiO2, International Conference on Gas-Fuel 05, 2005, pp. 222-226.
19. M. A. Cortes-Jacome, J. Escobar, C. A. Chavez, E. Lopez-Salinas, E. Romero, G. Ferrat and J. A. Toledo-Antonio, Highly dispersed CoMoS phase on titania nanotubes as efficient HDS catalysts, 4th International Symposium on Molecular Aspects of Catalysis by Sulfides, 2007, pp. 56-62.
20. J. A. Toledo-Antonio, M. A. Cortes-Jacome, C. Angeles-Chavez and J. Escobar, Highly active CoMoS phase on titania nanotubes as new hydrodesulfurization catalysts, Applied Catalysis B-Environmental 90 (2009), no. 1-2, 213-223.
21. S. Inoue, A. Muto, H. Kudou and T. Ono, Preparation of novel titania support by applying the multi-gelation method for ultra-deep HDS of diesel oil, Applied Catalysis a-General 269 (2004), no. 1-2, 7-12.
22. C. Roukoss, D. Laurenti, E. Devers, K. Marchand, L. Massin and M. Vrinat, Hydrodesulfurization catalysts: Promoters, promoting methods and support effect on catalytic activities, Comptes Rendus Chimie 12 (2009), no. 6-7, 683-691.
23. M. A. Cortes-Jacome, G. Ferrat-Torres, L. F. F. Ortiz, C. Angeles-Chavez, E. Lopez-Salinas, J. Escobar, M. L. Mosqueira and J. A. Toledo-Antonio, In situ thermo-raman study of titanium oxide nanotubes, Catalysis Today 126 (2007), no. 1-2, 248-255.
24. G. Hoogers, "Introduction," Fuel Cell Technology Handbook, G. Hoogers (Editor), CRC Press, Boca Raton, 2002, pp. 1.1 - 1.5.
25. A. Kirubakaran, S. Jain and R. K. Nema, A review on fuel cell technologies and power electronic interface, Renewable & Sustainable Energy Reviews 13 (2009), no. 9, 2430-2440.
26. B. Gou, W. K. Na and B. Diong, "Fundamentals of fuel cells," Fuel cells: Modeling, control, and applications, CRC Press, 2010, pp. 5 - 10.
27. E. Antolini, Formation of carbon-supported PtM alloys for low temperature fuel cells: A review, Materials Chemistry and Physics 78 (2003), no. 3, 563-573.
28. E. Antolini and E. R. Gonzalez, Ceramic materials as supports for low-temperature fuel cell catalysts, Solid State Ionics 180 (2009), no. 9-10, 746-763.
29. E. Antolini, Carbon supports for low-temperature fuel cell catalysts, Applied Catalysis B-Environmental 88 (2009), no. 1-2, 1-24.
30. J. J. Wang, G. P. Yin, Y. Y. Shao, S. Zhang, Z. B. Wang and Y. Z. Gao, Effect of carbon black support corrosion on the durability of Pt/C catalyst, Journal of Power Sources 171 (2007), no. 2, 331-339.
31. U. Diebold, The surface science of titanium dioxide, Surface Science Reports 48 (2003), no. 5-8, 53-229.
32. S. Y. Huang, P. Ganesan, S. Park and B. N. Popov, Development of a titanium dioxide-supported platinum catalyst with ultrahigh stability for polymer electrolyte membrane fuel cell applications, Journal of the American Chemical Society 131 (2009), no. 39, 13898.
33. M. Gustavsson, H. Ekstrom, R. Hanarp, L. Eurenius, G. Lindbergh, E. Olsson and B. Kasemo, Thin film Pt/TiO2 catalysts for the polymer electrolyte fuel cell, Journal of Power Sources 163 (2007), no. 2, 671-678.
34. S. von Kraemer, J. Wikander, G. Lindbergh, A. Lundblad and A. E. C. Palmqvist, Evaluation of TiO2 as catalyst support in Pt-TiO2/C composite cathodes for the proton exchange membrane fuel cell, Journal of Power Sources 180 (2008), no. 1, 185-190.
35. N. Rajalakshmi, N. Lakshmi and K. S. Dhathathreyan, Nano titanium oxide catalyst support for proton exchange membrane fuel cells, International Journal of Hydrogen Energy 33 (2008), no. 24, 7521-7526.
36. H. Chhina, S. Campbell and O. Kesler, Ex situ and in situ stability of platinum supported on niobium-doped titania for PEMFCs, Journal of the Electrochemical Society 156 (2009), no. 10, B1232-B1237.
37. H. Chhina, S. Campbell and O. Kesler, Characterization of Nb and W doped titania as catalyst supports for proton exchange membrane fuel cells, Journal of New Materials for Electrochemical Systems 12 (2009), no. 4, 177-185.
38. K. W. Park and K. S. Seol, Nb-TiO2 supported Pt cathode catalyst for polymer electrolyte membrane fuel cells, Electrochemistry Communications 9 (2007), no. 9, 2256-2260.
39. M. S. Dresselhaus and I. L. Thomas, Alternative energy technologies, Nature 414 (2001), no. 6861, 332-337.
40. J. Graetz, New approaches to hydrogen storage, Chemical Society Reviews 38 (2009), no. 1, 73-82.
41. J. J. Baschuk and X. G. Li, Carbon monoxide poisoning of proton exchange membrane fuel cells, International Journal of Energy Research 25 (2001), no. 8, 695-713.
42. I. D. Gonzalez, R. M. Navarro, W. Wen, N. Marinkovic, J. A. Rodriguez, F. Rosa and J. L. G. Fierro, A comparative study of the water gas shift reaction over platinum catalysts supported on CeO2, TiO2 and Ce-modified TiO2, Catalysis Today 149 (2010), no. 3-4, 372-379.
43. P. Panagiotopoulou, A. Christodoulakis, D. I. Kondarides and S. Boghosian, Particle size effects on the reducibility of titanium dioxide and its relation to the water-gas shift activity of Pt/TiO2 catalysts, Journal of Catalysis 240 (2006), no. 2, 114-125.
44. R. J. Gorte and S. Zhao, Studies of the water-gas-shift reaction with ceria-supported precious metals, Catalysis Today 104 (2005), no. 1, 18-24.
45. S. Hilaire, X. Wang, T. Luo, R. J. Gorte and J. Wagner, A comparative study of water-gas-shift reaction over ceria supported metallic catalysts, Applied Catalysis a-General 215 (2001), no. 1-2, 271-278.
46. G. Jacobs, U. M. Graham, E. Chenu, P. M. Patterson, A. Dozier and B. H. Davis, Low-temperature water-gas shift: Impact of Pt promoter loading on the partial reduction of ceria and consequences for catalyst design, Journal of Catalysis 229 (2005), no. 2, 499-512.
47. A. Goguet, S. O. Shekhtman, R. Burch, C. Hardacre, E. Meunier and G. S. Yablonsky, Pulse-response tap studies of the reverse water-gas shift reaction over a Pt/CeO2 catalyst, Journal of Catalysis 237 (2006), no. 1, 102-110.
48. H. Iida and A. Igarashi, Characterization of a Pt/TiO2 (rutile) catalyst for water gas shift reaction at low-temperature, Applied Catalysis a-General 298 (2006), 152-160.
49. K. G. Azzam, I. V. Babich, K. Seshan and L. Lefferts, Role of Re in Pt-Re/TiO2 catalyst for water gas shift reaction: A mechanistic and kinetic study, Applied Catalysis B-Environmental 80 (2008), no. 1-2, 129-140.
50. J. A. Rodriguez, J. Evans, J. Graciani, J. B. Park, P. Liu, J. Hrbek and J. F. Sanz, High water-gas shift activity in TiO2(110) supported Cu and Au nanoparticles: Role of the oxide and metal particle size, Journal of Physical Chemistry C 113 (2009), no. 17, 7364-7370.
51. M. A. Henderson, The interaction of water with solid surfaces: Fundamental aspects revisited, Surface Science Reports 46 (2002), no. 1-8, 5-308.
52. G. Jacobs, L. Williams, U. Graham, G. A. Thomas, D. E. Sparks and B. H. Davis, Low temperature water-gas shift: In situ drifts-reaction study of ceria surface area on the evolution of formates on Pt/CeO2 fuel processing catalysts for fuel cell applications, Applied Catalysis a-General 252 (2003), no. 1, 107-118.
53. D. Tibiletti, F. C. Meunier, A. Goguet, D. Reid, R. Burch, M. Boaro, M. Vicario and A. Trovarelli, An investigation of possible mechanisms of the water-gas shift reaction over a ZrO2-supported Pt catalyst, Journal of Catalysis 244 (2006), no. 2, 183-191.
54. W. Ruettinger, X. S. Liu and R. J. Farrauto, Mechanism of aging for a Pt/CeO2-ZrO2 water gas shift catalyst, Applied Catalysis B-Environmental 65 (2006), no. 1-2, 135-141.
55. K. G. Azzam, I. V. Babich, K. Seshan and L. Lefferts, Bifunctional catalysts for single-stage water-gas shift reaction in fuel cell applications. Part 1. Effect of the support on the reaction sequence, Journal of Catalysis 251 (2007), no. 1, 153-162.
56. H. Liang, Y. Zhang and Y. Liu, Three-dimensionally ordered macro-porous Pt/TiO2 catalyst used for water-gas shift reaction, Journal of Natural Gas Chemistry 17 (2008), no. 4, 403-408.
57. F. Boccuzzi, A. Chiorino, M. Manzoli, D. Andreeva and T. Tabakova, Ftir study of the low-temperature water-gas shift reaction on Au/Fe2O3 and Au/TiO2 catalysts, Journal of Catalysis 188 (1999), no. 1, 176-185.
58. J. B. Park, J. Graciani, J. Evans, D. Stacchiola, S. D. Senanayake, L. Barrio, P. Liu, J. F. Sanz, J. Hrbek and J. A. Rodriguez, Gold, copper, and platinum nanoparticles dispersed on CeOx/TiO2(110) surfaces: High water-gas shift activity and the nature of the mixed-metal oxide at the nanometer level, Journal of the American Chemical Society 132 (2010), no. 1, 356-363.
59. C. M. Kalamaras, P. Panagiotopoulou, D. I. Kondarides and A. M. Efstathiou, Kinetic and mechanistic studies of the water-gas shift reaction on Pt/TiO2 catalyst, Journal of Catalysis 264 (2009), no. 2, 117-129.
60. P. Panagiotopoulou and D. I. Kondarides, Effect of the nature of the support on the catalytic performance of noble metal catalysts for the water-gas shift reaction, 1st Conference of the European-Union-Coordination-Action -Coordination of Nanostructured Catalytic Oxides Research and Development in Europe, 2005, pp. 49-52.