Open access peer-reviewed chapter

Glycerol as a Superior Electron Source in Sacrificial H2 Production over TiO2 Photocatalyst

By Masahide Yasuda, Tomoko Matsumoto and Toshiaki Yamashita

Submitted: November 14th 2018Reviewed: March 13th 2019Published: April 11th 2019

DOI: 10.5772/intechopen.85810

Downloaded: 283


Biodiesel fuel (BDF) has gained much attention as a new sustainable energy alternative to petroleum-based fuels. BDF is produced by transesterification of vegetable oil or animal fats with methanol along with the co-production of glycerol. Indeed, transesterification of vegetable oil (136.5 g) with methanol (23.8 g) was performed under heating at 61°C for 2 h in the presence of NaOH (0.485 g) to produce methyl alkanoate (BDF) and glycerol in 83.7 and 73.3% yields, respectively. Although BDF was easily isolated by phase separation from the reaction mixture, glycerol and unreacted methanol remained as waste. In order to construct a clean BDF synthesis, the aqueous solution of glycerol and methanol was subjected to sacrificial H2 production over a Pt-loaded TiO2 catalyst under UV irradiation by high-pressure mercury lamp. H2 was produced in high yield. The combustion energy (ΔH) of the evolved H2 reached 100.7% of the total ΔH of glycerol and methanol. Thus, sacrificial agents such as glycerol and methanol with all of the carbon attached to oxygen atoms can continue to serve as an electron source until their sacrificial ability was exhausted. Sacrificial H2 production will provide a promising approach in the utilization of by-products derived from BDF synthesis.


  • BDF
  • photocatalyst
  • TiO2
  • sacrificial agent
  • glycerol
  • hydrogen

1. Introduction

The major issue in the current world is an urgent need to stop the increase of CO2 levels. A large amount of consumption of fossil resources causes serious environmental problems such as global warming and air pollution. Therefore, biofuels such as bioethanol, bio-hydrogen, and biodiesel (BDF) have gained much attention as renewable and sustainable energy alternative to petroleum-based fuels [1]. However, the problems to be solved for practical uses still remain in each biofuel. In bioethanol, the ethanol concentrations are still too low to isolate pure ethanol by distillation at a low energy cost [2, 3]. Bio-hydrogen is isolated spontaneously from reaction mixtures without operations to separate. However, it is needed to construct newly a supply system to vehicles.

BDF is produced by transesterification of vegetable oil or animal fats with methanol along with the co-production of glycerol [Eq. (1)] [4]. Although methyl alkanoate (BDF) is easily isolated by phase separation, a mixture of glycerol and unreacted methanol remains in aqueous solution as waste. New utilization of these wastes is required. Reforming of glycerol has been extensively investigated through pyrolysis [5, 6], steam gasification [7, 8], and biological reforming [9, 10]. We have focused on photocatalytic reforming over titanium dioxide (TiO2) [11]:


TiO2 has a semiconductor structure with 3.2 eV of bandgap, which corresponds to 385 nm of light wavelength [12]. Therefore, the TiO2 can be excited by 366 nm emitted from a high-pressure mercury lamp. Irradiation of the TiO2 induces charge separation into electrons and holes (Figure 1). Electron excited to the conduction band serves to reduce water to H2. Evolution of H2 is usually accelerated by deposition of noble metals (Pt, Pd, and Au) onto the TiO2. The positive charge (hole) oxidizes hydroxide absorbed on the surface of TiO2 to generate hydroxyl radicals, which is eventually transformed to O2 [13]. However, spontaneous conversion of hydroxyl radical into O2 is inefficient. Moreover, water splitting into O2 and H2 is a large uphill reaction, resulting in rapid reverse reaction.

Figure 1.

Photocatalytic water splitting over TiO2.

On the other hand, the hydroxyl radicals can be effectively consumed by the use of electron-donating sacrificial agents (hole scavengers), thus accelerating the H2 production (Figure 1) [14]. This method is named “sacrificial H2 production.” The sacrificial H2 production is an uphill process, but the energy change is small. Therefore, the sacrificial H2 production proceeds more smoothly compared with water splitting without sacrificial agents, thus providing a convenient method to generate H2 [15]. When one equivalent of hydroxyl radical is consumed, one equivalent of electron is generated to produce 0.5H2.

During our investigations on sacrificial H2 production over a Pt-loaded TiO2 (Pt/TiO2) [15], it was found that sacrificial agents with all of the carbon attached oxygen atoms such as saccharides, polyalcohols (e.g., arabitol, glycerol, 1,2-ethandiol), and methanol continued to serve as an electron source until their sacrificial ability was exhausted. Glycerol (1a) and methanol (1b) are by-products from BDF synthesis. The 1a has the potential to produce hydrogen in theoretical yield of seven equivalents, whose combustion energy (ΔH = 1995 kJ mol−1) is larger than ΔH of 1a (1654.3 kJ mol−1) [Eq. (2)]. Also, 1b can produce three equivalents of hydrogen, whose ΔH (855 kJ mol−1) is larger than ΔH of 1b (725.7 kJ mol−1) [Eq. (3)] [16]. Thus, photo-energy can promote uphill process:


2. Outline of conversion of glycerol to hydrogen

Generally, biomass reforming is started by the production of water-soluble materials from biomass through biological treatment as well as chemical reaction [17, 18]. The resulting water-soluble materials (saccharides, amino acids) are converted to biofuels such as ethanol, methane, and hydrogen through various catalytic reactions in aqueous solution. Our biomass reforming is performed in aqueous solution through sacrificial H2 production over Pt/TiO2 using water-soluble materials derived from lignocelluloses [19, 20, 21] and chlorella [22] (Figure 2).

Figure 2.

Outline of conversion of glycerol to hydrogen.

In this chapter, we will show H2 production through sacrificial H2 production over Pt/TiO2 using 1a and 1b from standpoints of construction of renewable energy system and clean synthesis of BDF.

3. Materials and method

3.1 Apparatus

NMR spectra were taken on a Bruker AV 400M spectrometer for CDCl3 solution. LC-MS analysis were performed on a Waters Alliance 2695 under conditions (ESI ionization, capillary voltage 3.5 kV, source temperature 120°C and desolvation temperature 350°C) using column (Waters, SunFire C18, 2.1 mmΦ × 150 mm) and 1% formic acid in MeOH-H2O (6:4) as an eluent solution. GLC analysis of solution was performed on a Shimadzu 14A gas liquid chromatograph with FID detector at a temperature raised from 50 to 250°C using a capillary column (J & W CP-Sil 5CB, 0.32 mmΦ × 50 m).

3.2 Photoreaction apparatus

Reaction vessel was a cylindrical flask with 30 cm of height and 7.5 cm of diameter, which had three necks on the top. A high-pressure mercury lamp (100 W, UVL-100HA, Riko, Japan), which emitted mainly a light at 313 and 366 nm, was inserted into the large central neck of the reaction vessel. The reaction vessel was connected to a measuring cylinder with a gas-impermeable rubber tube to collect the evolved gas. The reaction vessel was set in a water bath to keep it at 20°C. The stirring of the solution was performed by magnetic stirrer.

3.3 Preparation of photocatalyst

Almost all research has used TiO2 in anatase form such as P25 (Degussa Co. Ltd., Germany) and ST01 (Ishihara Sangyo Co. Ltd., Japan) for photocatalytic H2 production. A Pt-loaded TiO2 catalyst (Pt/TiO2) was prepared by photo-deposition method according to the previous literature [23]. An aqueous solution (400 mL) containing TiO2 (4.0 g, ST01), K2PtCl6 (40–400 mg), and 2-propanol (3.06 mL) was introduced reaction vessel, which was large scale of cylindrical flask with 35 cm of height and 9.0 cm of diameter. After the oxygen was purged by N2 gas bubbling for 20 min, the solution was irradiated by stirring. After irradiation for 24 h, the water was entirely removed from the reaction mixture by an evaporator. The resulting black precipitate was washed with water on a filter and then dried under reduced pressure to produce Pt/TiO2 [14]. The Pt content on TiO2 was optimized to be 2.0 wt% by the comparison of the H2 amounts evolved from photocatalytic reaction using 1a (115 mg, 1.25 mmol) over various Pt contents of the Pt-doped TiO2 (100 mg, 1.25 mmol) [15]. The structure of Pt/TiO2 was analyzed by a Shimadzu XRD 7000 diffractometer.

3.4 Photocatalytic H2 production

Pt/TiO2 (100 mg) and the given amounts of aqueous solution of sacrificial agent were introduced to reaction vessel. The volume of the reaction solution was adjusted to 150 mL with water. Oxygen was purged from reaction vessel by N2 gas for 20 min. TiO2 was suspended in aqueous solution by vigorous stirring during the irradiation. Total volume of the evolved gas was measured by a measuring cylinder. Irradiation was performed until the gas evolution ceased. The evolved gas (0.5 mL) was taken through rubber tube using syringe and was subjected to the quantitative analysis of H2, N2, CH4, and CO2. Gas analysis was performed on a Shimadzu GC-8A equipped with TCD detector at temperature raised from 40 to 180°C using a stainless column (3 mmΦ, 6 m) packed with a SHINCARBON ST (Shimadzu).

In order to determine the quantum yield (Φ) for H2 evolution, the H2 amount per hour was measured for various concentrations of 1 (8–40 mM). The H2 amount per hour was converted to Φ using an actinometer which was H2 amount per hour evolved from the sacrificial H2 production using ethanol (0.434 M) at pH 10.0 over Pt/TiO2 (Pt content 1.0 wt%), whose Φ was reported to be 0.057 [24]. Limiting quantum yields (Φ) at an infinite concentration of 1 was determined from the intercept of the double reciprocal plots of Φ vs. the concentration of 1 [25].

4. Results

4.1 Sacrificial H2 production using glycerol (1a) and methanol (1b)

Sacrificial H2 production was applied to 1a and 1b. The Pt/TiO2 (100 mg, 1.25 mmol, 2.0 wt% of Pt) was suspended in an aqueous solution (150 ml) of 1a and 1b, whose concentration was varied in a range of 0.25–1.25 mmol. After O2 was purged from the reaction vessel using N2 gas, UV irradiation was continued under vigorous stirring for 10–17 h until gas evolution had ceased [15]. The evolved gas volumes were plotted against the amounts of sacrificial agent used. In the absence of sacrificial agents, the evolved H2 from water was small (<2 mL). Figure 3A is a typical example of the plots of volume of H2 and CO2 against the amounts of 1a used. Gas volume increased as an increase of the amounts of 1a used. However, the molar ratio of the evolved H2 to 1a (H2/1a) was dependent on the amount of 1a used. Therefore, the H2/1a values were plotted against the molar ratios of 1a to catalyst (1a/catalyst). This plot gave a good linear relationship, as shown in Figure 3B.

Figure 3.

(A) The gas volume evolved from the sacrificial H2 production using glycerol (1a) over Pt/TiO2. (B) Plots of H2/1a and CO2/1a against 1a/catalyst: H2 (●) and CO2 (▲).

The intercept of the plot equaled the limiting amount of H2 (H2max) obtained from 1 mol of 1a when the amount of the catalyst was extrapolated to infinite. The H2max became 7.2. The limiting amount of CO2 (CO2max) obtained from 1 mol of 1a at an infinite amount of the catalyst was also determined to be 3.1 from the plots of CO2/1a against 1a/catalyst (Figure 3B). The H2max and CO2max are summarized in Table 1. If the sacrificial agent (CnHmOp) is entirely decomposed into CO2 and H2O by hydroxyl radicals, theoretically (2n + 0.5m − p) equivalents (P) of H2 will be evolved in the TiO2 photocatalytic reaction [Eq. (4)]. The P values are listed in Table 1. Therefore, the chemical yield of H2 production was defined to be 100 H2max/P. In the case of 1a, the yield of H2 production was found to be 103%. Also, the CO2max value was close to the theoretical value. Similarly, the H2max and CO2max values of 1b were determined to be 3.0 and 1.0, respectively. This shows that 1a and 1b are superior sacrificial agents, which are completely decomposed into CO2 and water by sacrificial H2 production:

Sacrificial agentsFormulaPaProducts/mol mol−1Yield/% bΦc
Glycerol (1a)C3H8O377.23.11030.078
Methanol (1b)CH4O33.01.01000.057
1-Hydroxy-2-propanone (1c)C3H6O274.92.50.30870.045
1,2-Propanediol (1d)C3H8O284.81.0Trace60
1,3-Propanediol (1e)C3H8O284.20.553
1-Propanol (1f)C3H8O94.11.0460.069
2-Propanol (1g)C3H8O91.30.014
Carboxylic acids
Glycolic acid (2a)C2H4O332.81.893
Oxalic acid (2b)C2H2O411.02.0100
Formic acid (2c)CH2O211.01.0100
Acetic acid (2d)C2H4O242.91.70.27100
Pyruvic acid (2e)C3H4O353.92.70.30102
Lactic acid (2f)C3H6O364.12.30.3088
Malonic acid (2g)C3H4O442.62.70.3196
Propanoic acid (2h)C3H6O272.31.033

Table 1.

Sacrificial H2 production over Pt/TiO2 using alcohols (1) and carboxylic acids (2).

Theoretical amount of hydrogen was calculated using Eq. (4).

Total chemical yield of H2 and CH4 = 100 (H2max + 4CH4max)/P.

Limiting quantum yield (Φ) for H2 evolution with infinite amounts of 1.


4.2 Degradation mechanism of 1a and 1b

Generally, hydroxyl radical can abstract hydrogen atom more efficiently from the hydroxylated carbon rather than the non-hydroxylated carbon. Therefore, degradation of alcoholic sacrificial agents proceeds through hydrogen-atom abstraction by hydroxyl radical from the hydroxylated alkyl group [Eq. (5)] [15]. Hydroxyl radical reacts with the secondary alcohols to produce ketones, which does not undergo further degradation. The primary alcohols reacted with hydroxyl radical to produce aldehyde, which undergoes further oxidation to carboxylic acid [Eq. (6)]. Furthermore, H abstraction from carboxylic acid by hydroxyl radical induces decarboxylation from carboxylic acids through the formation of carboxyl radical (RCO2·) [Eq. (7)]. When hydroxyl group was substituted on α-position of carboxylic acid [X = OH in Eq. (7)], the decarboxylation took place more smoothly. Many researchers proposed that the decomposition of carboxylic acids is initiated by hole transfer to the carboxylic group rather than H abstraction by hydroxyl radicals [26, 27, 28, 29]. Thus, the degradation of alcohols proceeds through the formation of carboxylic acids:


In 2009, Kondarides et al. reported sacrificial H2 production from 1a over Pt/TiO2 (0.1–0.5 wt% Pt) [30]. They proposed that the decomposition of 1a proceeded through the formation of methanol and acetic acid which were eventually decomposed into CO2 and H2 in a ratio of 3:7 [31]. Also, in irradiation of Pt/TiO2 in the absence and in the presence of glycerol, they detected H2O2 which was produced by dimerization of hydroxyl radicals [32]. Also, Ratnawati et al. detected a small amount of 1,2-ethanediol and acetic acid in reaction mixture [33]. They elucidated that Pt catalyzed not only reduction of water to H2 but also dehydration of 1a. Bowker et al. examined the photocatalytic reforming of 1a over M/TiO2 (M = 0.5 wt% Pd, 2.0 wt% Au) [34]. However, chemical yield of H2 was still unclear.

We thought that degradation of 1a was initiated by the oxidation of terminal alcohol by hydroxyl radical. It was thought that glycolic acid (2a) and oxalic acid (2b) were the intermediates intervening in degradation process of 1a. Therefore, we performed sacrificial H2 production over Pt/TiO2 using 2a and 2b. The H2max and CO2max values of 2a and 2b were shown in Table 1. The 2a and 2b were completely decomposed to CO2 and water, since the CO2max values of 2a and 2b were determined to be 1.8 and 2.0, respectively. Although the degradation of 2a could proceed through 2b and/or formic acid (2c), we could not determine which degradation pathway occurred. In the case of 1b, it was thought that 2c was undoubtedly the intermediates intervening in degradation process of 1b. The 2c was completely decomposed to CO2 and water, since the CO2max value of 2c was 1.0. However, 2a, 2b, and 2c were not detected in the reaction mixture of sacrificial H2 production using 1a and 1b due to easy decomposition of these carboxylic acids by hydroxyl radical. Also, Lu et al. have reported the degradation of 2b and 2c, which can adsorb on Pt/TiO2 to give one equivalent H2 under irradiation [35, 36].

According to Eqs. (5)(7), a possible degradation mechanism of 1a and 1b by hydroxyl radical is shown in Figure 4. In the case of 1a, 14 equivalents of hydroxyl radicals were consumed by 1a along with the formation of 3CO2. At the same time, seven equivalents of H2 were evolved. Actually, 7.2 of H2max and 3.1 of CO2max values of 1a were provided from sacrificial H2 production using 1a. In the case of 1b, six equivalents of hydroxyl radicals were consumed along with the formation of one equivalent of CO2 and 3H2, providing actually 3.0 of H2max and 1.0 of CO2max.

Figure 4.

Degradation pathways of glycerol (1a) and methanol (1b) by hydroxyl radical in the sacrificial H2 production over Pt/TiO2.

4.3 Structural dependence on H2 yields in sacrificial H2 production using several alcohols (1c1g)

In order to elucidate the relationship between molecular structure of sacrificial agents and degradation yield, sacrificial H2 production was performed using propane-based alcohols such as 1-hydroxy-2-propanone (1c); 1,2-propanediol (1d); 1,3-propanediol (1e); 1-popanol (1f); and 2-propanol (1g) (Figure 5) as well as the related carboxylic acids (2d2h) [15].

Figure 5.

Propane-based alcohols (1c–1g) as sacrificial agents for the photocatalytic H2 production.

Sacrificial H2 evolution using 1c produced CH4 along with the formation of H2 and CO2. Limiting amount of CH4 (CH4max) obtained from 1 mol of 1c was 0.30 along with 4.9 of H2max and 2.5 of CO2max values. In the case of sacrificial H2 production along with the formation of CH4, the chemical yield was defined by the following equation: Yield = 100 (H2max + 4CH4max)/P. The yield for the sacrificial H2 production using 1c was calculated to be 87%. Moreover, acetic acid (2d) was detected by LC-MS of the reaction solution at low conversion. A peak appeared at 2.24 min of retention time which showed mass peaks at m/z 60 (M+) and 43 (CH3CO+). Therefore, 2d was subjected to sacrificial H2 production. Mozia et al. reported that 2d was decomposed into H2, CO2, and CH4 over TiO2 without Pt [37], although Zheng et al. reported that a trace amount of CH4 was detected from 2d over Pt/TiO2 (Pt = 1.0 wt%) [38]. We determined the chemical yields [39]. The CH4max of 2d was determined to be 0.27 along with 2.9 of H2max and 1.7 of CO2max values. The total yield was calculated to be 100% (=100 (2.9 + 4 × 0.27)/4) in the sacrificial H2 production using 2d. Considering the experimental error, stoichiometric equation for conversion of 2d into H2, CH4, and CO2 was shown in Eq. (8):


It was thought that pyruvic acid (2e) was an intermediate of degradation process from 1c to 2c. The H2max, CO2max, and CH4max values of 2e were found to be 3.9, 2.7, and 0.3, respectively [39]. Degradation scheme of 2e can be expressed by Eq. (9). The yield for the sacrificial H2 production using 2e was 100%. Since the degradation yield of 1c was found to be 87%, the degradation of 1c to H2, CO2, and CH4 proceeded effectively through the formation 2e followed by 2d:


The next sacrificial H2 production was examined using 1d. Oxidation of 1d with hydroxyl radical was initiated by oxidation of primary alcohol part to afford lactic acid (2f). Sacrificial H2 production using 2f produced H2, CH4, and CO2. The H2max, CO2max, and CH4max values of 2f were 4.1, 2.3, and 0.30, respectively. On the other hand, the H2max and CO2max values of 1d were determined to be 4.8 and 1.0, respectively. Trace amount of CH4 was formed. Thus, complete decomposition of 1dinto H2 and CO2 did not take place. Therefore, it is speculated that degradation of 1d proceeds via 2f which was decomposed to acetaldehyde. It is suggested that oxidation of acetaldehyde by hydroxyl radical was slow.

In sacrificial H2 production using 1e, H2max and CO2max values of 1e were 4.2 and 0.50, respectively. Moreover, malonic acid (2g, m/z 104 (M+)) was detected in LC-MS of the photolysate. The sacrificial H2 production using 2g showed that the H2max, CO2max, and CH4max values were determined to be 2.6, 2.7, and 0.31, respectively. Degradation scheme of 2g can be expressed by Eq. (10). Although the degradation yield of 2g was relatively high yield (96%), 1e was not completely decomposed, resulting in 0.5 of the CO2max and no CH4 emission. This suggests that the degradation process of 2g is slow:


Moreover, sacrificial H2 production was applied to 1f. The H2max and CO2max values of 1f were determined to be 4.1 and 1.0, respectively. CH4 was not formed. It is suggested that the degradation of 1f proceeded via the formation of propanoic acid (2h). The H2max and CO2max values of 2h were determined to be 2.3 and 1.0, respectively. The decarboxylation of 2h and the subsequent oxidation gave acetaldehyde, which was subjected to the further degradation, but it was slow process [39]. In the case of sacrificial H2 production using 1g, acetone was detected by GLC analysis of the reaction mixture. The H2max value was determined to be 1.3 and CO2 was not evolved. Further degradation of acetone did not proceed.

Based on these results, the degradation pathways of 1c, 1d, 1e, 1f, and 1g by hydroxyl radical are summarized in Figure 6. Though considerable amounts of CO2were evolved from 1c to 1f, the CO2max (0.5–2.5) did not reach the theoretical values. In the case of 1g, CO2 was not formed at all. Thus, in the case of these polyols which have one or two non-hydroxy-substituted carbons, the H2max and CO2max values did not reach the theoretical values. Therefore, we conclude that sacrificial agents with all of the carbon attached to oxygen atoms such as 1a and 1b continued to serve as an electron source until their sacrificial abilities were exhausted.

Figure 6.

Degradation pathway of sacrificial agents by hydroxyl radical in the sacrificial H2 production over Pt/TiO2 using propane-based alcohols: 1-hydroxyl-2-propanone (1c); 1,2-propanediol (1d); 1,3-propanediol (1e); 1-propanol (1f); and 2-propanol (1g).

4.4 Separation of residual glycerol and methanol in BDF synthesis

Vegetable oil was mainly composed of the oleic acid (C17H33CO2H) triglyceride whose average molecular weight was thought to be 884 g/mol. At first, since carboxylic acid was included in used oil as impurity, the amounts of NaOH (a g/kg-lipid) which was required to achieve pH of 8–9 were determined. Lipid (ca. 1 mL, 0.884 g) was solved in 2-propanol (10 mL) and neutralized by an aqueous NaOH solution. In this case, a was determined to be 0 g since fresh vegetable oil was used.

Vegetable oil (150 mL, 136.5 g, 0.154 mol) was set in a reaction vessel. Since usual optimal amount for transesterification of neutral lipid is known to be 3.55 g/kg [11], the amount of NaOH necessary to the transesterification was determined to be 0.485 g (=0 + 0.485 g) by the sum of a g/kg and 3.55 g/kg. Usually, 20% of weight of 1b to vegetable oil is used for BDF synthesis. 1b (30 mL, 23.8 g, 0.743 mol) was mixed with NaOH (0.485 g, 0.012 mol). About half of the mixture of 1b and NaOH was poured in a reaction vessel and then kept at 61°C for 1 h under stirring. Moreover, the remaining mixture of 1b and NaOH was added into the reaction vessel, and the reaction mixture was kept at 61°C for another 1 h.

Follow-up operation is shown in Figure 7. After cooling, the reaction mixture was separated into a lower layer and an upper layer. The lower layer (solution A) contained 1a and 1b. The upper layer was washed with water (300 mL) and separated to the BDF upper layer. Aqueous solution (solution B) was obtained from the lower layer. In order to check the contamination of lipid to BDF layer, the purity of BDF was determined by the peak-area ratio of methyl and methoxy groups in NMR spectra. The BDF layer contained C17H33CO2Me (114.5 g, 0.387 mol) and unreacted vegetable oil (2.2 g). The yield of C17H33CO2Me (BDF) was 83.7% based on the theoretical amounts of 137 g (0.463 mol).

Figure 7.

Mass balance for BDF preparation and the sacrificial H2 production using residual 1a and 1b.

GLC analysis showed that solution A contained 1a (10.4 g, 0.113 mol) and 1b (6.85 g, 0.214 mol) where molar ratio (b) of 1a to 1b was 1.89. The yield of 1a was 73.3% based on the theoretical amounts of 14.2 g (0.154 mol). NMR analysis of solution A showed that RCO2Na (2.2 g) was contained in solution A. Solution B contained 1b (4.38 g, 0.137 mol) and a small amount of C17H33CO2Na. Thus, 1b was found in both solutions A and B.

4.5 Hydrogen production from residual methanol and glycerol in BDF synthesis

The photocatalytic reforming of 1a and 1b was examined using solution A. Irradiation was performed by a high-pressure mercury lamp under vigorous stirring with a magnetic stirrer. Figure 8 shows the plots of the H2/1a against the molar ratio of 1a to the catalyst (1a/catalyst), which was adjusted to 0.2, 0.4, 0.6, 0.8, and 1.0. From the intercept of the plots, H2max obtained from 1 mol of 1a at an infinite amount of the catalyst was determined to be 12.52. The yields of H2 production of solution A were determined as follows. According to Eq. (11), the H2 amount (P) was theoretically calculated to be 12.67 using P = 7 + 3b and b = 1.89. Since actual H2max was determined to be 12.52, the yield was calculated to be 98.8% (=100H2max/P). The results are summarized in Table 2:

Figure 8.

Determination of H2max values by the plots of H2/1 against 1/catalyst using solution A (◯, b = 1.89) and solution B (◇) obtained from the BDF synthesis.

Residues of BDF synthesisaPhotocatalytic reformingb
1a/mol1b/molP  cH2max  dH2/mol (yield/%)e
Solution Af0.1130.21412.6712.521.41 (98.8)
Solution B0.1373.001.080.15 (36.0)
[ΔH/kJ]g[187][255][445] (100.7)h

Table 2.

Photocatalytic reforming of residues of BDF synthesis.

Transesterification was performed by the reaction of lipid (136.5 g, 0.154 mol) with 1b (23.8 g, 0.743 mol) in the presence of NaOH (0.485 g, 0.012 mol) at 61°C for 2 h. BDF (114.5 g) was isolated.

Photocatalytic reforming was performed by irradiation of Pt/TiO2 in aqueous solution of 1a and 1b obtained from solutions A and B.

The values were the theoretical amounts (P) obtained from Eqs. (11) and (12).

The limiting amount of H2 (H2max) was obtained from Figure 8.

The values in parenthesis were the yield of H2 = 100H2max/P.

The molar ratio (b) of 1b to 1a was 1.89.

The combustion energy (ΔH) of 1a, 1b, and H2 were 1654.3, 725.7, and 285.0 kJ mol−1, respectively [16].

The energy recovery yield was calculated to be 100.7% by Eq. (14).


Next, photocatalytic reforming was performed with solution B containing 1b. Solution B was neutralized with dilute H2SO4 in order to reduce the effect of excess NaOH on TiO2. After that, an aqueous solution (150 mL) containing 1b (0.25–1.25 mmol) was irradiated in the presence of Pt/TiO2 (100 mg) in a similar manner as solution A. The plots of H2/1b against the molar ratio of 1b to catalyst (1b/catalyst) are overlaid on Figure 8. The H2max values were determined to be 1.08. The H2 yields were calculated to be 36.0% based on the theoretical P (3.00) [Eq. (12)]. In solution B, C17H33CO2Na was converted to C17H33CO2H by neutralization. It is well known that the carboxylic acid can strongly be adsorbed on TiO2. Therefore, it is suggested that the adsorption of C17H33CO2H on TiO2 lowered the photocatalytic activity of TiO2. The presence of C17H33CO2H retarded the H2 production of solution B remarkably.

Total amount of H2 from solutions A and B was calculated to be 1.56 mol by Eq. (13) using 0.113 mol of 1a in solution A and 0.137 mol of 1b in solution B: 0.113 × 12.52 + 0.137 × 1.08 (Table 2). H2 (1.56 mol) whose combustion energy (ΔH) was 445 kJ was evolved from solutions A and B. The ΔH of H2 was compared with ΔHof 1a and 1b. As shown in Table 2, 0.113 mol of 1a and 0.351 mol of 1b were isolated from BDF synthesis which had 442 kJ of ΔH. The energy recovery yield was calculated to be 100.7% by using Eq. (14):


5. Conclusion and perspective

Sacrificial H2 production can produce H2 in aqueous solutions. Gaseous H2 can be spontaneously isolated from reaction mixture without being separated. Therefore, sacrificial H2 production will provide a promising approach in the utilization of 1a and 1b derived from BDF synthesis.

Recent trends are shifting to the development of solar light-responsive photocatalysts. For example, nanotube-type Pt-N/TiO2 (1 wt% Pt) was applied to sacrificial H2 production with 1a where quantum yield for H2 evolution reached 0.37–0.36 [27]. CuO/TiO2 (1.3 wt% of CuO) was used for sacrificial H2 production using 1a [40]. Heteroatom (B, N)-doped Pt/TiO2 catalyst produced H2 in 88.7–90.9% yields from 1a under xenon lamp irradiation [41]. The B, N-doped Pt/TiO2 had absorption in visible light region (400–500 nm). Photo-reforming of 1a over CuOx/TiO2 (Cu = 0.01–2.8 wt%) gave H2 under visible light irradiation [42]. H2 production was performed over a CuO-TiO2 composite using 1a and 1b under sunlight irradiation [43]. Sacrificial H2 production over Ag2/TiO2 from 1a was performed by irradiation with a xenon lamp [44].

BDF market has significantly increased to adhere to energy and climate policies [45]. If H2 is produced by a photocatalytic process using solar energy and biomass-derived sacrificial agents, it will be the most promising process to construct clean BDF synthesis.

Conflict of interest

The authors declare that they have no competing interests.

© 2019 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution 3.0 License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

How to cite and reference

Link to this chapter Copy to clipboard

Cite this chapter Copy to clipboard

Masahide Yasuda, Tomoko Matsumoto and Toshiaki Yamashita (April 11th 2019). Glycerol as a Superior Electron Source in Sacrificial H<sub>2</sub> Production over TiO<sub>2</sub> Photocatalyst, Glycerine Production and Transformation - An Innovative Platform for Sustainable Biorefinery and Energy, Marco Frediani, Mattia Bartoli and Luca Rosi, IntechOpen, DOI: 10.5772/intechopen.85810. Available from:

chapter statistics

283total chapter downloads

More statistics for editors and authors

Login to your personal dashboard for more detailed statistics on your publications.

Access personal reporting

Related Content

This Book

Next chapter

Inclusion of Crude Glycerin in Diets for Sheep

By Marco Túlio Costa Almeida and Josimari Regina Paschoaloto

Related Book

First chapter

Asphaltenes – Problems and Solutions in E&P of Brazilian Crude Oils

By Erika Chrisman, Viviane Lima and Príscila Menechini

We are IntechOpen, the world's leading publisher of Open Access books. Built by scientists, for scientists. Our readership spans scientists, professors, researchers, librarians, and students, as well as business professionals. We share our knowledge and peer-reveiwed research papers with libraries, scientific and engineering societies, and also work with corporate R&D departments and government entities.

More About Us