Formation of Combustible Hydrocarbons and H2 during Photocatalytic Decomposition of Various Organic Compounds under Aerated and Deaerated Conditions

A possibility of photocatalytic production of useful aliphatic hydrocarbons and H2 from various organic compounds, including acetic acid, methanol, ethanol and glucose, over Fe-modified TiO2 is discussed. In particular, the influence of the reaction atmosphere (N2, air) was investigated. Different gases were identified in the headspace volume of the reactor depending on the substrate. In general, the evolution of the gases was more effective in air compared to a N2 atmosphere. In the presence of air, the gaseous phase contained CO2, CH4 and H2, regardless of the substrate used. Moreover, formation of C2H6 and C3H8 in the case of acetic acid and C2H6 in the case of ethanol was observed. In case of acetic acid and methanol an increase in H2 evolution under aerated conditions was observed. It was concluded that the photocatalytic decomposition of organic compounds with simultaneous generation of combustible hydrocarbons and hydrogen could be a promising method of “green energy” production.


Introduction
Over the past thirty years increased concerns over emissions of greenhouse gases and the depletion of non-renewable resources of fossil fuels has caused the necessity to look for new methods of energy production. From both the ecological and economical point of view conversion of waste and wastewaters into energy is especially desirable. One of the most promising and popular approaches is biogas generation [1,2]. Biogas is a mixture of different gases, mainly methane and carbon dioxide. Its production during anaerobic digestion involves microorganisms, which results in some serious drawbacks of this technology, as the bacteria responsible for methane generation are very sensitive to the environmental conditions, such as oxygen content, pH or presence of certain organic and inorganic compounds [3]. Therefore, wastes or wastewaters containing substances which are toxic or recalcitrant to these microorganisms cannot be used in the traditional biogas production process.
Application of the photocatalytic process instead of the biological one could remove that restriction. Photocatalysis is not selective for any kind of substrates, therefore it might be used for treatment of all contaminants, even those which are toxic to the methanogenic bacteria [4].
Due to its significant activity, stability and low cost TiO2 is widely used as a photocatalyst. Most investigations concerning the photocatalytic treatment of organic compounds in aqueous solutions are focused on their complete mineralization to CO2 and H2O. Usually, during these experiments the composition of the aqueous phase is only monitored. However, determination of the gas phase composition should be also of interest. There are some reports [5][6][7][8][9][10][11] showing that the process of a photocatalytic reduction of CO2 may lead to methane formation.
The first papers concerning the photocatalytic generation of hydrocarbons from organics in liquid phase were published in the 1970s by Kraeutler and Bard [12][13][14]. These authors described a photocatalytic decarboxylation of acetic acid under UV light in the presence of Pt/TiO2 photocatalyst. The reaction in which CH4 and CO2 were evolved as the products was named the "photo-Kolbe" reaction. A few years later Sakata et al. [15] reported methane and ethane formation during photodecomposition of acetic and propionic acids in the presence of bare and Pt modified TiO2.
A possibility of hydrocarbon formation during photodegradation of C1-C3 alcohols in aqueous suspensions of TiO2 was investigated by Dey and Pushpa [16]. They concluded that CH4 and CO2 were the main products of the reaction of methanol, ethanol and 2-propanol. Other hydrocarbons such as ethane, ethene and propene were also detected; however, at relatively low yields. Similar investigations were conducted by Bahruji et al. [17]. The authors used Pt-modified TiO2 in order to increase H2 formation. CH4, CO2, C2H6 and C3H8 were also identified in the gas phase.
Xu et al. [18] reported biomass reforming on Pt/TiO2 (anatase-rutile structure) leading to H2 generation. Methanol, propanetriol, formic acid and glucose were used as the model compounds and sacrificial agents. The possibility of hydrogen production from glucose, sucrose and starch over noble metal-loaded TiO2 photocatalysts was also described by Fu et al. [19]. The results revealed an enhancement of H2 production in case of Pd and Pt modified TiO2 and an inhibition of the efficiency in aerated systems.
Recently, Klauson et al. [20] described the application of TiO2 modified with Pt, Co, W, Cu or Fe for the production of hydrogen, oxygen and low molecular weight hydrocarbons from aqueous solutions of humic substances under anoxic conditions. In the presence of all the above materials the formation of CH4 was observed, although the highest yield was found in case of Pt-TiO2. That photocatalyst was also the most efficient when formation of C2H4, C2H6 and H2 was taken into account.
In the present work an Fe-modified TiO2 photocatalyst was applied for the photocatalytic generation of useful hydrocarbons and hydrogen which could be regarded as the potential source of "green energy". Different organics representing biomass-derived compounds, including an aliphatic acid (acetic acid), aliphatic alcohols (methanol and ethanol) and glucose were used in the experiments. In particular the influence of the reaction atmosphere on the products evolution was investigated. The Fe/TiO2 photocatalyst was chosen on a basis of our previous investigations [21] during which we found that it exhibits high activity in the "photo-Kolbe" reaction using acetic acid as a substrate.

Photocatalytic Decomposition of Various Organic Compounds: The Influence of a Substrate on the Formation of the Gaseous and Liquid Products
Depending on the substrate, different gases were identified in the headspace volume of the reactor ( Table 1). In case of acetic acid, the main products of its decomposition were CH4 and CO2. Low amounts of C2H6, C3H8 and H2 were also identified. During the photocatalytic degradation of alcohols the following gaseous products were identified: CO2, CH4 and H2 in case of CH3OH and CO2, CH4, C2H6 and H2 in case of C2H5OH. The gaseous products formed during photodegradation of C6H12O6 were CH4, CO2 and H2 ( Table 1). The diversity of the products generated from the applied substrates resulted from their different photocatalytic decomposition pathways. Taking into consideration that some by-products of the organics' degradation must have been generated in the liquid phase, the composition of the reaction solution was also examined. The investigations revealed ( Table 1) the presence of trace amounts of acetaldehyde (CH3CHO) in all cases. Furthermore, methanol (CH3OH) in the case of acetic acid and ethanol decomposition, and ethanol (C2H5OH) and methyl acetate (CH3COOCH3) in the case of acetic acid and glucose degradation were identified. In addition, small quantities of acetone (CO(CH3)2) were detected during the photodecomposition of acetic acid. The amounts of all the products in the liquid phase were very low and no clear dependence of the liquid phase composition on the reaction atmosphere used was found.

Effect of the Reaction Atmosphere on Gas Phase Composition during the Photodegradation of Various Organic Substrates
The concentrations of gaseous reaction products evolved with time of irradiation were continuously monitored during the experiments. Figures 1-4 present changes of the amounts of CO2 and CH4 in the gaseous phase during the processes conducted under either N2 or air atmospheres. In Figures 5 and 6 a comparison of the amounts of C2H6 and H2 evolved after 27 h of the decomposition of the model compounds is shown.

Acetic Acid
In general, the main mechanism responsible for a photocatalytic decomposition of CH3COOH is its decarboxylation initiated by the photogenerated holes (h + ). This reaction, known as the "photo-Kolbe" reaction, leads to the production of one mole of CO2 and one mole of CH4 from one mole of CH3COOH: Moreover, recombination of methyl radicals might take place, which results in a formation of C2H6, except from CH4 [12][13][14][15][21][22][23][24]. Formation of C2H6 and H2 can be written as follows: Further, as can be seen in Table 1, formation of C3H8 can also occur. A possible mechanism of propane generation can be as follows [15]: or: Nevertheless, the present results clearly show that the CH4/CO2 ratio after 27 h of irradiation was 0.88 when a N2 atmosphere was applied and 0.78 when the process was conducted in the presence of air. This suggests that reaction (1) was not the only one proceeding in the system. From Table 1 it can be found that aside from methane, ethane was also formed. In this process methyl radicals are consumed. Therefore, the amount of ethane should be also taken into consideration. Assuming that two methyl radicals form one molecule of C2H6 the CH3 • /CO2 ratio can be calculated. After 27 h of irradiation of acetic acid solution the amount of C2H6 evolved in a N2 atmosphere was 0.05 mmol C2H6/mol CH3COOH, whereas under aerated conditions it was 0.09 mmol C2H6/mol CH3COOH. Thus, the CH3 • /CO2 ratio was 0.93 and 0.82 for N2 and air atmosphere, respectively. However, the values are still below 1. Incorporation of C3H8 in the calculations also does not allow one to get a ratio of 1, since the amount of propane was an order of magnitude lower than that of ethane. These results suggest that formation of carbon dioxide might also be due to the mineralization of CH3COOH to H2O and CO2: Reaction (8) is understandable when the aerated conditions are considered; however, the obtained results revealed that it also proceeded in the N2-purged system. In our previous paper [21] we have discussed higher evolution rate of CO2 compared to CH4 by the reaction of CH3COOH with the photogenerated oxygen. This O2 as well as the hydroxyl radicals might be responsible for the mineralization of CH3COOH [21], which leads to higher CO2 evolution.
The results shown in Figure 1 revealed that the amounts of CH4 and CO2 evolved under aerated conditions were more than two times higher compared to a N2 atmosphere (1.72 vs. 3.85 mmolCH4/molCH3COOH and 1.95 vs. 4.93 mmolCO2/molCH3COOH, respectively, after 27 h). Higher efficiency of CH4 evolution under the aerated conditions can be explained by more effective separation of e − /h + pairs in the presence of O2, being an efficient electron scavenger, and acetic acid, which is known as an effective hole scavenger. Therefore, in the presence of both oxygen and CH3COOH the "photo-Kolbe reaction" should occur more easily, what was confirmed by the results presented in Figure 1. Moreover, it was found that the concentration of O2 in the headspace volume of the reactor decreased from 21 to 12 vol.% after 27 h of irradiation, which confirms that oxygen was consumed in the process. The obtained results ( Figure 5) also revealed higher efficiency of C2H6 evolution in the aerated compared to the N2 purged system. Ethane formation (Reaction (2)) is initiated by the photogenerated holes, therefore, can easily proceed under both deaerated and aerated conditions. However, like in case of methane, more efficient separation of e − /h + pairs contributes to the enhancement of ethane formation. Moreover, the presence of O2 can result in the increase of the amount of C2H6 by enabling of its formation according to the following equation [23,24]:  As shown in Table 1, amongst the products of CH3COOH decomposition hydrogen was also present. As in case of other gases, evolution of H2 was significantly higher in an air atmosphere compared to a N2 one ( Figure 6). After 27 h of irradiation the amounts of H2 were 0.04 and 0.81 mmolH2/mol CH3COOH in N2 and air purged system, respectively. The data discussed above show that the photocatalytic conversion of CH3COOH into hydrocarbons and hydrogen was significantly more effective in the presence of air than in the N2 purged system.

Methanol
The photocatalytic degradation of methanol under deaerated conditions can be written as [25]: This reaction can also be represented as two half-reactions of oxidation and reduction, respectively: As reported by Chen et al. [25], Reaction (12) cannot occur easily in an aerated system because only few hydrogen atoms are formed in the presence of oxygen. Under such conditions, oxygen is more competitive in capturing the photogenerated electrons, which eventually leads to the formation of H2O2 and OH • .
The obtained results ( Table 1) revealed formation of CO2 and H2 as the only gaseous products of CH3OH decomposition in N2 atmosphere, which confirms the mechanism presented by Equations (10)- (12). Nonetheless, if the only reaction occurring in the investigated system were Reaction (10), the H2/CO2 ratio should be equal to 3, but the experimental data show that the ratio is significantly lower (ca. 0.7-0.8). This suggests that some other reactions proceeded in the system. As in case of CH3COOH, such a reaction can be mineralization of CH3OH yielding CO2 and H2O as products [24]: During the experiments conducted in the air-purged system, the evolution of methane, except from CO2 and H2, was observed ( Figure 2). Its concentration in the gaseous mixture was, however, very low and after 27 h of irradiation it only amounted to 4.26 μmol/molCH3OH. Nonetheless, the observed formation of CH4 might lead to a conclusion that the mechanism of methanol decomposition in the presence of air is not as simple as the one described by Equation (13). For example, a possibility of CO2 photoreduction cannot be excluded here [16]. Dey and Pushpa reported that carbon dioxide, generated during mineralization of methanol, could undergo a methanation reaction by e − and yield CH4. In the case of greater amounts of CO2 (as is the case in this work, when the system was aerated) there is a better chance of it being reduced, which can explain the results shown in Figure 2.
The amount of CO2 evolved in the presence of air was at the end of the experiment about eight times higher compared to the N2 atmosphere (0.627 vs. 0.078 mmolCO2/mol CH3OH, respectively). High CO2 evolution was an effect of methanol mineralization (Equation (13)) and was accompanied by a decrease of O2 concentration in the gaseous phase (from 21 to 16 vol.%). It was also observed that in the presence of N2 no gaseous product evolved from the reaction mixture within the initial 5 h of the experiment. On the contrary, when the reaction was conducted under aerated conditions the evolution of CO2 started after 2 h of irradiation.
Evolution of hydrogen was significantly lower compared to that of CO2 ( Figure 6). After 27 h of irradiation the amount of H2 was 0.06 and 0.13 mmolH2/mol CH3OH in the N2 and air purged systems, respectively.

Ethanol
In the case of ethanol, the main products identified in the gaseous mixture were CH4 and CO2 ( Figure 3). Moreover, some amounts of C2H6 and H2 were also identified ( Figures 5 and 6). The CH4/CO2 ratio was higher in N2 than in an air atmosphere and amounted to 0.83 and 0.05, respectively. This resulted from significantly higher CO2 evolution in the presence of air compared to the N2-purged system (1.45 vs. 0.078 mmolCO2/mol C2H5OH after 27 h). As in case of other substrates a decrease of O2 concentration in the gaseous phase in case of the experiments conducted under aerated conditions was found (from 21 to 15 vol.%). It was also observed that the amount of methane obtained under both conditions was comparable ( Figure 3).
Decomposition of ethanol is more complex compared to methanol due to the presence of the ethyl group in the C2H5OH structure. As a result, the range of intermediate degradation products is very wide [25]. In case of the deaerated conditions the overall reaction of ethanol decomposition can be written as follows [17,24]: The reduction reaction can be represented by Equation (12), like in case of methanol [25]. However, the oxidation reactions are different. Generally, methane can be produced either by the reaction of free methyl radicals with H • or ethanol, or the reaction of acetic radicals with ethanol [24]. In case of the present research, since acetic acid was not identified in the liquid phase (Table 1), the most probable pathway of CH4 formation was the one involving CH3 • and C2H5OH. Moreover, as in case of methanol [16], the reduction of CO2 leading to the methane production cannot be ignored here. Furthermore, methyl radicals can also recombine yielding C2H6 as the product ( Figure 5).
In the presence of oxygen the decomposition of ethanol can be described by the following equation [24]: Equation (14) indicates that H2 should be present amongst the ethanol decomposition products. Indeed, the analysis of the gaseous phase composition revealed evolution of hydrogen under both the aerated and deaerated conditions ( Figure 6). Furthermore, H2 could be produced by a degradation of the intermediate products present in the liquid phase (CH3CHO, CH3OH, Table 1). However, taking into account that their concentrations were very low, this pathway was of minor importance. The amount of H2 formed in the N2 purged system was comparable to that in the aerated system (0.049 vs. 0.051 mmolH2/molC2H5OH, respectively). If we recall the methane evolution under aerated and deaerated conditions (Figure 3) we may find that the reaction atmosphere did not clearly influence the effectiveness of H2 and CH4 formation during ethanol decomposition.

Glucose
The photocatalytic reforming of C6H12O6 is a very complex process which proceeds through numerous steps, in which intermediates such as carboxylic acids, aldehydes and hydrocarbons are formed [19,26]. A detailed probable mechanism of glucose degradation under anaerobic conditions leading to the formation of H2 and CO2 was recently discussed by Fu et al. [19]. The overall reaction can be written as: The present research confirmed the formation of CO2 and H2 (Figures 4 and 6). In addition, small amounts of CH4 were identified as well. From Figure 4 it can be found that the amount of CO2 was higher in the presence of air compared to a N2 atmosphere, which is consistent with the results observed for other substrates. After 27 h of irradiation the amount of CO2 in the gaseous mixture was 0.65 mmolCO2/molC6H12O6 and 0.11 mmolCO2/molC6H12O6 for air and N2, respectively. In the experiment conducted under aerated conditions the concentration of O2 in the gaseous phase decreased from 21 to 15 vol.% which confirms its consumption during glucose decomposition.
No significant difference between the efficiency of hydrogen evolution in the two systems was observed. In case of the N2-purged system the amount of H2 was 0.048 mmolH2/molC6H12O6, whereas in case of the aerated system, it was 0.054 mmolH2/molC6H12O6 ( Figure 6). Similarly, no difference in the amount of methane evolved in the presence and in the absence of oxygen was found. After 27 h of the reaction in both N2 and air atmospheres, the amount of CH4 reached 0.033 mmolCH4/molC6H12O6 ( Figure 4). The observed evolution of methane can be explained by decomposition of by-products formed in the liquid phase (Table 1) as well as CO2 photoreduction, as discussed earlier.

Hydrogen Evolution in the Presence of Oxygen: a Point of Discussion
The results discussed above revealed that the presence of oxygen at a concentration of 21 vol.% or less (i.e., oxygen in air) did not suppress hydrogen evolution during the photodegradation of organic compounds in the performed experiments. What is more, in the cases of acetic acid and methanol a significant enhancement of H2 formation was even observed ( Figure 6). This is somewhat unusual in view of the electron acceptability of O2 and the competitiveness with H + for electron scavenging [16,19,24,25,27].
In order to investigate if the observed phenomenon resulted from the presence of Fe in the photocatalyst structure, an additional experiment was performed. A TiO2 photocatalyst prepared in a similar way to the Fe/TiO2, but without impregnation with Fe(NO3)3, was applied in a process of photocatalytic CH3COOH degradation under N2 and air atmosphere. After 27 h of irradiation it was found that the effectiveness of evolution of CH4 and H2 in the air purged system was higher by 65 and 45%, respectively, compared to the N2 atmosphere. Therefore, it was concluded that the addition of iron was not responsible for the phenomenon described above.
There are very few papers reporting that O2 does not affect negatively or could have a positive influence on hydrogen photogeneration [28][29][30]. Korzhak et al. [28] found that when a small amount of air was introduced to a photocatalytic system containing ethanol, the yield of H2 formation increased. However, in case of mixtures saturated with oxygen or air, hydrogen formation was almost completely suppressed. The authors contributed the observed increase in hydrogen production to the fact that under such conditions the reactions of O2 with active free organic radicals take place with high rate constants. Therefore oxygen is consumed mainly in the process leading to the evolution of additional amounts of hydrogen. Moreover, dissolved oxygen might be involved in stabilization of the radical intermediates thus could enhance the reaction efficiency [31]. Furthermore, organic substrates such as acids, alcohols or glucose, contribute to the improvement of charge separation by scavenging of photogenerated holes and consuming O2 in diverse direct oxidation reactions, which leads to a decrease of the oxygen concentration [32][33][34].
Anyhow, the majority of the work on hydrogen generation with semiconductors dispersed in a solution is carried out in an oxygen-free atmosphere to avoid the back recombination processes, oxygen interferences with the photocatalyst which occurs while forming of superoxides and/or peroxides and the competition of O2 and H + for the reduction sites [15,[17][18][19]26,32,[35][36][37]. Most of the papers which describe the photocatalytic degradation of organics in the presence of O2 are focused on its total mineralization, thus the evolution of H2 is not discussed. We have proved that the negative O2 influence on the H2 generation from different organic substrates is not so evident. In some cases (e.g., decomposition of acetic acid) an increase in H2 evolution yield can even be obtained. Therefore, a broad and detailed discussion is needed in order to explain the discussed phenomenon.

Photocatalyst
The photocatalyst used in this study was described in details in our previous paper [21]. In brief, the Fe/TiO2 was prepared by an impregnation method using crude TiO2 obtained from the Chemical Factory "Police" (Police, Poland) and (Fe(NO3)3) as the Fe precursor. The sample was calcined at 500 °C. The amount of Fe introduced to the sample was 20 wt.%. The Fe/TiO2 contained anatase, rutile and Fe2O3 phases. The crystallite size of anatase and the anatase over rutile ratio were equal to 9 nm and 87:13, respectively. The specific surface area SBET was 82 m 2 /g.

Photocatalytic Reaction
The photocatalytic reaction was conducted in a cylindrical quartz reactor (type UV-RS-2, Heraeus, Hanau, Germany) equipped with a medium pressure mercury vapour lamp (TQ-150, λmax = 365 nm). The total volume of the reactor was 765 cm 3 (350 cm 3 of a liquid phase and 415 cm 3 of headspace). In the upper part of the reactor a gas sampling port was mounted. At the beginning of the experiment 0.35 dm 3 of CH3COOH, CH3OH, C2H5OH or C6H12O6 solution and 1 g/dm 3 of the photocatalyst were introduced into the reactor. The concentration of the organic substrates was 1 mol/dm 3 in all the experiments.
Before the photocatalytic reaction N2 (in order to eliminate the dissolved oxygen) or air were bubbled through the reactor for 1 h. Then, the gas flow was stopped and UV lamp, positioned in the centre of the reactor, was turned on to start the photoreaction. The process was conducted for 27 h. The reaction mixture containing the photocatalyst in suspension was continuously stirred during the experiment by means of a magnetic stirrer. All the experiments were repeated at least twice in order to confirm the reproducibility of the results. Gaseous products of the reaction were analyzed using a SRI 8610C GC (SRI Instruments, Torrance, CA, USA) equipped with TCD and HID detectors, and Shincarbon (carbon molecular sieve; 2 m, 1 mm, 100-120 mesh), molecular sieve 5 Å (3 m, 2 mm, 80-100 mesh) and 13× (1.8 m, 2 mm, 80-100 mesh) columns. Helium was used as the carrier gas. The composition of the liquid phase was determined using a SRI 8610C GC equipped with a FID detector and a MXT ® -1301 (60 m) column. Hydrogen was used as the carrier gas.

Conclusions
The possibility of photocatalytic generation of combustible hydrocarbons and hydrogen from various organic substrates, including an aliphatic acid (CH3COOH), alcohols (CH3OH, C2H5OH) and sugar (C6H12O6) was demonstrated. The composition of the gaseous phase was influenced by both the applied substrate and the reaction atmosphere. In general, higher efficiency of hydrocarbon and hydrogen generation was obtained under aerated conditions, which is very advantageous from the point of view of possible future applications. In the presence of air, the gaseous phase contained CO2, CH4 and H2, regardless of the substrate used. Moreover, formation of C2H6 and C3H8 in the case of acetic acid and C2H6 in the case of ethanol was observed.
The obtained results revealed that the presence of oxygen did not suppress hydrogen evolution during the photodegradation of organic compounds. In the cases of acetic acid and methanol a significant enhancement of H2 formation was even observed. Further investigations concerning this issue as well as the improvement of the efficiency of the presented system are in progress.