Monatshefte für Chemie - Chemical Monthly

, Volume 148, Issue 7, pp 1311–1321 | Cite as

Co-production of hydrogen and carbon nanotube-silica fiber composites from ethanol steam reforming over an Ni-silica fiber catalyst

  • Natthawan Prasongthum
  • Chaiyan Chaiya
  • Chanatip Samart
  • Guoqing Guan
  • Paweesuda Natewong
  • Prasert Reubroycharoen
Original Paper


Nickel supported on silica fiber (Ni-SF) catalysts was successfully synthesized by sol–gel-assisted electrospinning of SFs followed by the conventional impregnation of the Ni salt and calcining. Their activity for the co-production of hydrogen (H2) and carbon nanotube-silica fiber (CNT-SF) composites in the ethanol steam reforming (ESR) process was investigated. The effects of the ESR reaction temperature, steam-to-carbon ratio (S/C), space–time (W/F), and Ni loading level on the reaction activity as well as H2 production and CNT-SF characteristics of the Ni-SF catalyst were investigated. The Ni-SF catalyst was highly effective at the simultaneous production of H2 and CNT. The optimized condition of the ESR in terms of the ethanol conversion and H2 yield was achieved at 600 °C, an S/C ratio of 9, and W/F of 18 gcat h mol−1 with a maximum H2 yield of 55%, while the best quality and quantity of the CNT (36%) formed along with a H2 yield of 29% was obtained at an Ni loading of 30 wt%, S/C ratio of 1, and W/F of 9 gcat h mol−1. The novel CNT-SF composite obtained from ESR exhibited a relatively high surface area and easy accessibility, making it a promising catalyst support for various processes.

Graphical abstract


Nickel-based catalyst Steam reforming of ethanol Silica fiber Carbon nanotubes 


Hydrogen (H2) is considered as a clean energy. It can be directly utilized in a fuel cell to generate electricity and is also a potential feedstock in the oil refinery and chemical industry. Currently, the conventional H2 production is by the steam reforming of fossil fuel, but this releases extra undesirable greenhouse gases to the atmosphere [1, 2], as well as depletes the non-renewable fossil fuel stocks. Therefore, new renewable feedstocks for replacing the conventional fossil fuels are receiving more attention. Ethanol is one of the interesting sources for H2 production, since it is a clean source that is derived from renewable biomass [3]. The successful production of H2 via ethanol steam reforming (ESR) over various catalysts has been reported in many research groups. The ESR is an endothermic reaction, which produces H2 and numerous by‐products such as acetaldehyde, acetone, ethylene, CO, CO2, CH4, and carbon deposition. The origin of the deactivation by carbon deposition on the catalysts during the ESR is attributed to the following routes: (a) the polymerization of ethylene formed by ethanol dehydration on the acid sites of the catalyst; (b) the decomposition of methane; and (c) the Boudouard reaction [4]. According to the literature [5], there are two types of carbon deposited occurring during the ESR: amorphous carbon and filamentous carbon (CNT and CNF). The valuable CNT product is receiving considerable attention, since the CNT has unique structural properties such as a high specific surface area, high thermal stability, and excellent electrical conductivity. These advantages allow it to be used in various applications, especially in composite materials for improving the physical and chemical properties of other materials [6, 7]. The CNT can be synthesized by various methods such as arc discharge, pyrolysis, laser ablation, plasma-assisted deposition, and thermal chemical vapor [8]. There are a few reports in relation to the simultaneous production of hydrogen and carbon nanotubes from the ESR; therefore, the ESR is an interesting approach in converting ethanol into the co-production of H2 and CNT.

Previous studies have investigated the ESR for the production of H2 over various active metals including noble metal-based metals (Ru, Rh, Pd, and Pt) and transition metals (Ni, Co, and Fe). The noble metal-based metals exhibited excellent activity and stability in the ESR; however, their application in large scales is limited by their high costs [9]. Among transition metal catalysts, Ni-based catalysts have been widely used in the ESR because of its inexpensive and inherent activity on C–C and C–H bond cleavage. The Ni-based catalysts have been dispersed on the external and internal surfaces of porous materials such as silica (SiO2) and alumina (Al2O3) [10]. However, the active sites deep inside the pores of the Ni catalysts are less accessible by the reactants which limits their applications. Fibrous materials, such as silica fiber (SFs), have been shown to be a good support for Ni-based catalysts due to their extremely open morphology and ease of accessibility. The use of Ni-SFs for the steam reforming of glycerol has been reported [11], where Ni-SF catalysts exhibited a higher catalytic activity and stability than similar porous supported Ni catalysts. However, SFs have a low surface area, since they are non-porous, whereas the high surface area and electron transfer properties of CNT or CNF make them promising composite materials for improving the physical and chemical properties of other materials [12]. For this reason, the combination of the SFs and CNT as a composite is of great interest. Hybrid composites of CNT and conventional porous Al2O3 were developed as an alternative support for a cobalt catalyst in Fischer–Tropsch synthesis [13], and were found to have a remarkable catalytic activity and stability as well as outperforming other supports.

In this work, we proposed a novel clean process for the synthesis of a co-production of hydrogen and CNT on the SF by ESR. The effects of the reaction temperature, steam-to-carbon ratio (S/C), space–time (W/F), and Ni loading level on the gas production (including H2) and CNT deposition were investigated. The CNT hybridized onto the SFs combined the advantages of a hollow carbon nanostructure and the extremely open structure of SFs which could be used as a support for the Ni catalyst in steam reforming processes to solve the mass transfer limitation problem and to enhance the catalytic activity and stability.

Results and discussion


The surface morphologies of the SF support and Ni-SF catalysts with 10, 20, and 30 wt% Ni loadings (10Ni-SF, 20Ni-SF, and 30Ni-SF, respectively) are shown in Fig. 1. The surface of the SF support was smooth with an average diameter of 260 nm, as estimated using the SemAfore program. The NiO particles were decorated on the SF support and had an average size of 14.7–27.0 nm (Table 1). The average size of the NiO particles tended to increases and became more aggregated with increasing Ni loading levels, reaching 27 nm at an Ni loading of 30 wt%.
Fig. 1

SEM micrographs of a SF support and the fresh catalysts: b 10Ni-SF, c 20Ni-SF, and d 30Ni-SF

Table 1

Textural properties of the SF support and Ni-SF catalysts


Surface area/m2 g−1

Pore volume/cm3 g−1

Pore diameter/nm





























aDetermined by EDS analysis

bAverage Ni particle size determined by XRD using the Debye–Scherrer equation

cAverage Ni particle size determined by FESEM analysis using the SemAfore program

The textural properties of the supports and catalysts are also presented in Table 1. The actual contents of Ni as determined by EDS analysis were fairly similar to the theoretical Ni loading levels. The surface area (3 m2 g−1) and pore volume (0.0003 cm3 g−1) of the SF support were quite low, suggesting that the obtained SF support was a non-porous structure. The surface area and pore volume increased with increasing Ni contents, especially at 30 wt% Ni, up to the highest surface area of 19 m2 g−1 and pore volume of 0.0754 cm3 g−1 for the 30Ni-SF, since at higher Ni loading levels, the NiO particles covered all of the SF surfaces (Fig. 1), and so increased the surface area [14].

The XRD patterns of the SF support and catalysts with different Ni contents are displayed in Fig. 2. The broad diffraction peak located at a 2θ value of 21.3° was assigned to the amorphous SiO2 support. The diffraction peaks at 2θ values of 36.8°, 42.8°, 62.4°, 75.4°, and 79° were the characteristic peaks of NiO [15], and the peak intensities tended to increase with increasing loading levels of Ni. The NiO crystal size, as estimated by Scherrer’s formula using the Ni (2 0 0) peak, is presented in Table 1. The crystal sizes of all samples were in the range of 20.2–26.6 nm and became slightly larger as the Ni content increased. The aggregation of NiO was the main reason for the increased NiO crystal sizes at higher Ni loadings [16]. This result was in agreement with the SEM results.
Fig. 2

XRD patterns of the SF and the fresh 10Ni-SF, 20Ni-SF, and 30Ni-SF catalysts

The H2-TPR profiles of the catalysts are shown in Fig. 3, where they all showed two major reduction peaks. The first peak at 310–440 °C was assigned to the reduction of NiO with large particles which had no or weak interaction with the SiO2 support. The second peak at a higher temperature of 450–530 °C corresponded to the reduction of NiO with small particles or those that were well distributed and strongly interacted with the SiO2 support [17]. When the Ni loading level increased from 10 to 30 wt%, the intensity of the reduction peaks became stronger, which was due to the increased proportion of Ni in the catalyst. In addition, the main reduction peaks shifted towards higher temperature regions with increasing Ni loading levels, which was probably due to highly dispersed NiO particles at lower Ni loading levels. On the other hand, NiO tended to aggregate at higher loading levels. This introduced H2 diffusion limit in the NiO phase, resulting in a high reduction temperature [17].
Fig. 3

H2-TPR profiles of the fresh 10Ni-SF, 20Ni-SF, and 30Ni-SF catalysts

Hydrogen obtained from the ESR

The effect of different Ni loading levels, W/F, reaction temperature, and S/C ratio on the ethanol conversion level and product distribution for the ESR over the Ni-SF catalysts were studied by sequential univariate analysis. The results for the different Ni loading levels (10, 20, and 30 wt%) at 600 °C, W/F of 9 gcat h mol−1, and S/C of 1 are shown in Fig. 4a. The ethanol conversion level increased to 100% with increasing Ni loading levels from 10 to 20 wt% and remained at this level with a 30 wt% Ni loading. The dominant products measured in the outlet gas from the ESR were H2, CH4, CO, and CO2. The H2 yield seemed to decrease slightly as the Ni loading level increased from 10 to 20 wt%, while the CO selectivity decreased. In addition, the CH4 concentration increased slightly with increasing Ni loading levels, reaching 36% at an Ni loading of 30 wt%, suggesting the presence of a methanation reaction. The methanation reaction is known to be favored with large NiO particle sizes [17]. This consumes H2 and so results in the decreased H2 yield. Overall, an increased Ni loading level did not effectively improve the H2 yield but led to the production more undesirable CH4.
Fig. 4

Ethanol conversion level and gas product distribution in the ESR with Ni-SF catalyst at time on stream of 360 min under different a Ni loading levels (at 600 °C, W/F of 9 gcat h mol−1, and S/C ratio of 1), b W/F (with 10Ni-SF at 600 °C and S/C of 1), c reaction temperatures (with 10Ni-SF at W/F of 18 gcat h mol−1 and S/C ratio of 1), and d S/C ratios (with 10Ni-SF at 600 °C and W/F of 18 gcat h mol−1)

The effect of the W/F on the ESR catalytic performance with the 10Ni-SF catalyst is displayed in Fig. 4b. The W/F was varied from 9 to 36 gcat h mol−1 by changing the ethanol feed flow rate. The ethanol conversion level was almost 85% at a W/F of 9 and increased to 100% as the W/F increased to 18. The H2 yield and CO selectivity also increased slightly with increasing W/F from 9 to 18. The selectivity of CH4 showed the opposite trend, decreasing as the W/F increased. A large W/F gave a high contact time which promoted C–C bond cleavage. Therefore, the decomposition of ethanol, shown in Eq. (1), was favored, resulting in a high H2 yield and CO selectivity [18]. The high contact time could also lead to CH4 reforming, shown in Eq. (5), and so a low CH4 selectivity:
$${\text{C}}_{ 2} {\text{H}}_{ 5} {\text{OH}} \to {\text{H}}_{ 2} + {\text{CO}} + {\text{CH}}_{ 4}$$
$${\text{CH}}_{ 3} {\text{CH}}_{ 2} {\text{OH}} \to {\text{CH}}_{ 3} {\text{CHO}} + {\text{H}}_{ 2}$$
$${\text{CH}}_{ 3} {\text{CHO}} \to {\text{CH}}_{ 4} + {\text{CO}}$$
$$2 {\text{C}}_{ 2} {\text{H}}_{ 5} {\text{OH}} + {\text{H}}_{ 2} {\text{O}} \to {\text{CH}}_{ 3} {\text{COCH}}_{ 3} + {\text{CO}}_{ 2} + 4 {\text{H}}_{ 2}$$
$${\text{CH}}_{ 4} + {\text{H}}_{ 2} {\text{O}} \to {\text{CO}} + 3 {\text{H}}_{ 2}$$
$${\text{CO}} + {\text{H}}_{ 2} {\text{O}} \to {\text{H}}_{ 2} + {\text{CO}}_{ 2} .$$

However, the H2 yield and CO selectivity showed almost no change when the W/F was further increased from 18 to 36. The maximum H2 yield was 27%.

The effect of temperature on the ethanol conversion level and product distribution in the ESR over the 10Ni-SF catalyst was studied at 400, 500, and 600 °C (Fig. 4c). The ethanol conversion level continuously increased from 52% up to 100% by increasing the reaction temperature from 400 to 600 °C. For the ESR product distribution, the H2 yield and CO2 selectivity also clearly increased, with a maximum of H2 yield (28%) being achieved at 600 °C. In contrast, the ethane and CH4 selectivity tended to decrease slightly with increasing reaction temperature. In the ESR, system involves the formation of an ethoxy species through the dissociative adsorption of ethanol. The ethoxy species would be oxidized to form an aldehyde intermediate. This process was facilitated by water or the OH groups. The acetaldehyde intermediate could either transform to ethane and water by the cleavage of the C–O bond of acetaldehyde and addition of hydrogen or further be oxidized to acetate species by hydroxyl groups from the support. The acetaldehyde and acetate species can further decompose through C–C bond cleavage to form CH4 and CO. The formation of acetone can also occur in the ERS via possible two routes: (a) the coupling of two acetate molecules and (b) the disproportionation of two acetyl species [19]. In experiment, at 400 °C, a negligible level of C2 products (such as acetaldehyde, acetone, and ethane) was detected. The presences of acetaldehyde and acetone were only noticeable at 400 °C with relatively low selectivity of 0.4 and 0.002%, respectively; hence, it was not expressed in Fig. 4. The capability of Ni to cleave C–C bonds was weaker at lower temperatures and the ethanol dehydrogenation [Eq. (2)] to acetaldehyde followed by decomposition of the latter to CO and CH4 [Eq. (3)] was favored, giving a low ethanol conversion level. With increasing temperature, the ability of breaking the C–C bond of Ni was improved, which promoted acetaldehyde steam reforming. Furthermore, the decrease in CH4 selectivity and the simultaneous increased in H2 yield and CO2 selectivity observed were attributed to the CH4 steam reforming and water gas shift (WGS) reactions [Eqs. (5), (6)] [19].

The influence of the S/C molar ratio on the ESR catalytic performance over the 10Ni-SF catalyst was studied at 600 °C with S/C ratios of 1, 3, 6, 9, and 12 (Fig. 4d). The ethanol conversion level was 100% at all S/C ratios studied, but the H2 and CO2 yields increased progressively as the S/C ratio increased. The highest H2 yield (55%) was achieved at an S/C ratio of 9. In contrast, the CO and CH4 selectivity decreased with increasing S/C ratios. Thus, the addition of more water into the feed exhibited a positive effect on the H2 yield. The WGS reaction was not flavored at a low S/C ratio, and so a high selectivity to CO and CH4 was obtained. Increasing the S/C ratio to 9 promoted the WGS and CH4 steam reforming reactions, which resulted in an increased H2 and CO2 concentration, associated with a gradual decrease in the CH4 and CO concentrations [20]. However, further increasing the S/C ratio from 9 to 12 did not obviously improve the H2 yield or change the concentration of CO and CO2. The unchanged H2 yield was attributed to that higher water content led to the competition between ethanol and water to be adsorbed on the active sites [21].

Time on stream of the catalysts with different Ni loading levels (10, 20, and 30 wt%) at 600 °C, W/F of 9 gcat h mol−1, and S/C of 1 is shown in Fig. 5. The ethanol conversion level increased with increasing Ni loading levels. As for the 30Ni-SF catalyst, the ethanol conversion level increased dramatically as time on stream increased and reached 100% at 90 min. After that, the ethanol conversion was stable for all time on stream. The similar trend was observed on the 20Ni-SF catalyst, but the ethanol conversion level reached 100% at 150 min. The ethanol conversion level of the 10Ni-SF catalyst increased gradually as time increased and reached a constant level with 85% after 180 min. The results indicated that the catalysts did not deactivate during the reaction. Similar results were also observed for all reaction parameters including S/C ratios, W/F, and reaction temperatures.
Fig. 5

Time on stream of the Ni-SF catalysts with different Ni loading levels at 600 °C, W/F of 9 gcat h mol−1, and S/C of 1

Synthesis of CNTs on the Ni-SF catalysts during the ESR reaction

Representative FESEM images of the CNT deposited over the Ni-SF catalysts produced from the ESR at 600 °C, W/F of 9 gcat h mol−1, and an S/C ratio of 1 are shown in Fig. 6. All the SEM images revealed carbon filaments that were highly entangled with a web-like structure on the Ni-SF catalysts. These carbon filaments increased in density with increasing Ni loading levels in the Ni-SF catalysts and covered all of the surfaces of the 30Ni-SF catalyst. The TEM image of the spent 30Ni-SF catalyst (Fig. 6d) revealed that the carbon filaments were CNT with an average outer diameter of 22–30 nm, which was close to the average NiO particle sizes obtained by the XRD analysis. This result was supported by the specific surface area and pore volume of the spent catalyst (Table 2). Both the surface area and the pore volume increased with increasing Ni loading levels, suggesting that the amount of CNT could play an important role in increasing the specific surface area of the spent catalysts, since CNT is well known for its high surface, which was beneficial for further enhancement of the catalytic activities of the catalyst [22].
Fig. 6

SEM images of CNTs deposited in the ESR over the spent a 10Ni-SF, b 20Ni-SF, c 30Ni-SF catalysts at 600 °C, W/F of 9 gcat h mol−1, and S/C ratio of 1, and d TEM image of CNTs deposited in the ESR over the spent 30Ni-SF catalysts at 600 °C, W/F 9 gcat h mol−1, S/C ratio of 1, and time on stream at 360 min

Table 2

Textural properties of the spent Ni-SF catalysts and formed CNT-SF composites


Surface area/m2 g−1

Pore volume/cm3 g−1

Pore diameter/nm

















aPurified CNT-SF composites produced from ESR over the 30Ni-SF catalysts at 600 °C, W/F of 9 gcat h mol−1, S/C ratio of 1, and time on stream at 360 min

Raman analysis was used to characterize the CNT formed by ESR on the 10Ni-SF catalyst at different temperatures, and the results are shown in Fig. 7. In the spectrum, two Raman peaks were observed at 1210–1360 and 1550–1660 cm−1, which were attributed to the D band and G band, respectively. The D band corresponded to disordered carbon structures in the CNTs, while the G band was attributed to the highly structural order in the CNTs. The intensity ratio (IG/ID) is associated with the structure of CNT, where a high IG/ID ratio indicated a highly organized structure of the produced CNTs [23]. The order of the IG/ID ratio calculated from Fig. 7 was as follows: 400 °C (IG/ID = 0.79) < 500 °C (IG/ID = 0.90) < 600 °C (IG/ID = 0.92). Thus, the IG/ID ratio increased with increasing reaction temperature, implying that high-quality CNT were formed at an ESR reaction temperature of 600 °C.
Fig. 7

Raman spectra of carbon deposited over the 10Ni-SF catalysts at different temperatures: 400, 500, and 600 °C (W/F of 18 gcat h mol−1, S/C ratio of 1, and time on stream at 360 min)

The types and yield of carbon deposited on the spent catalysts were further examined by TG/DTG analysis, with the TGA curves of carbon deposited over the 10Ni-SF catalyst at ESR temperatures of 400, 500, and 600 °C (W/F of 18 gcat h mol−1 and S/C of 1), as shown in Fig. 8. Three weight loss regions were obtained on the catalyst used for ESR at 400 °C. The first weight loss region at 460–530 °C was ascribed to the decomposition of amorphous carbon. The second, at 530–577 °C, was assigned to the decomposition of CNTs with disordered structures, while the third at 597–670 °C was the decomposition of stable CNTs [24]. There were only two weight loss regions for the spent catalyst used for ESR at 500 °C, which were assigned to the decomposition of less stable and stable CNTs, respectively, while for the spent catalyst used for ESR at 600 °C, only CNTs were observed. The results here indicated that a low ESR reaction temperature (400 °C) produced a mixture of amorphous carbon, CNTs, while at 600 °C, CNTs were mainly formed. In experiment, there was no any ethylene product measured, and thus, the CNTs cannot be formed via the ethylene polymerization. However, the negligible amounts of acetaldehyde and acetone were detected. Therefore, the possible carbon deposited routes were an accumulation of acetate-like species, the decomposition of CHx species, and Boudouard reaction as follows [19, 25]:
$${\text{CO}} + {\text{H}}_{ 2} \to {\text{H}}_{ 2} {\text{O}} + {\text{C}}$$
$${\text{CH}}_{ 4} \to {\text{C}} + 2 {\text{H}}_{ 2}$$
$$2 {\text{CH}}_{ 3} {\text{CHO}} \to 2 {\text{CH}}_{ 3} {\text{COCH}}_{ 3} \to \left( {{\text{CH}}_{ 3} } \right)_{ 2} {\text{C}}\left( {\text{OH}} \right){\text{CH}}_{ 2} {\text{COCH}}_{ 3} \to \, \left( {{\text{CH}}_{ 3} } \right)_{ 2} {\text{C}} = {\text{CHCOCH}}_{ 3} + {\text{H}}_{ 2} {\text{O}}$$
$${\text{CH}}_{ 3} {\text{CHO}} \to {\text{coke}} .$$
Fig. 8

TGA curves of carbon deposited over the 10Ni-SF catalyst during the ESR at 400, 500, and 600 °C (W/F of 18 gcat h mol−1, S/C ratio of 1, and time on stream at 360 min)

Figure 9 shows the carbon balance of the ESR on the Ni-SF catalyst at various reaction conditions. It can be seen that the products consisted of gases, liquid (acetaldehyde and acetone), and carbon deposited (amorphous carbon and CNTs). The amounts of amorphous carbon and CNTs were determined by TGA analysis. In Fig. 9a, the percentage of carbon in ethanol transform into CNTs decreased as reaction temperature increased. The CNT yields were obtained for 29, 22, and 20% at reaction temperature of 400, 500, and 600 °C, respectively, while the amorphous carbon was only obtained at 400 °C. The CNTs yield on the 10Ni-SF catalyst at S/C of 1 was about 20% and decreased slightly with increasing S/C ratios to 14% at an S/C ratio of 12 (Fig. 9b). In Fig. 9c, the CNT yield on the 10Ni-SF catalyst produced from ESR at an S/C ratio of 1 and 600 °C decreased from 28 to 19% as the W/F increased from 9 to 36 gcat h mol−1. A higher CNTs yield was obtained at a higher Ni loading (Fig. 9d). The maximum CNTs yield of 36% was obtained at Ni loading of 30 wt%. Based on the results, the optimization of feed ratio, W/F, reaction temperature, and Ni loading was believed to generate quality and quantity of CNTs.
Fig. 9

Carbon balance of the ESR on the Ni-SF catalysts with different reaction conditions: a reaction temperatures (10Ni-SF at S/C ratio of 1 and W/F of 18 gcath mol−1), b S/C ratios (10Ni-SF at 600 °C and W/F of 18 gcath mol−1), c W/F values (10Ni-SF at 600 °C and S/C ratio of 1), and d Ni loading levels (S/C ratio of 1 and W/F of 9 gcat h mol−1 at 600 °C)

As mentioned above, the systematically effect of various conditions (S/C molar ratios, W/F, reaction temperatures, and Ni loadings) on the co-production of hydrogen and CNTs from the ESR over the Ni-SF catalysts was discussed. The results can be summarized as follows: the higher reaction temperature (600 °C) and S/C molar ratio (9) favored the production of H2 with a yield of 55%, which was accompanied by 14% yield of CNTs. In terms of the carbon production, the highest yield of carbon (36%) along with an H2 yield of 29% was obtained at 600 °C with a lower S/C (1) and W/F (9 gcat h mol−1) and a higher Ni loading (30 wt%). It is clear to achieve the co-production yield of H2 and CNTs using the same reaction.

The TEM image of the purified CNTs (purification with acids) on the SF support produced from the ESR reaction with the 30Ni-SF catalyst at 600 °C, W/F of 9 gcat h mol−1, and S/C of 1 is shown in Fig. 10. The SF support was fully covered by CNTs with a highly entangled and web-like formation. The average diameter of CNTs on the SF-supported was approximately 20–30 nm.
Fig. 10

TEM image of the purified CNTs-SF composite from the spent 30Ni-SF catalyst (W/F of 9 gcat h mol−1, S/C of 1, reaction temperature of 600 °C, and time on stream at 360 min)

Thermogravimetric analysis of the spent 30Ni-SF catalyst and the purified CNTs-SF composite is presented in Fig. 11. Only one weight loss region at around 650 °C was observed, which was assigned to the decomposition of stable CNTs. The amount of carbon before the treatment was 36 wt% but after treatment was approximately 10 wt%. The decrease in the carbon content after treatment was attributed to the elimination of carbon impurities [26].
Fig. 11

TGA profiles of a spent 30Ni-SF catalyst with time on stream at 360 min (W/F of 9 gcat h mol−1, S/C of 1, and reaction temperature of 600 °C), and b purified CNTs-SF composite

The specific surface area of the spent 30Ni-SF catalyst was 116 m2 g−1, while that of the CNT-SF after removing Ni was 157 m2 g−1 (Table 2). The higher specific surface area after treatment was because the treatment to remove Ni opened the carbon tube tip, which allowed the N2 adsorption into the intertubular pores of the CNT resulting in an increased derived surface area [27].


The successful synthesis of SFs was attained by sol–gel-assisted electrospinning, and they were then used as a support for the impregnated and calcined NiO catalyst in the ESR reaction to produce H2 and CNT. The effect of the reaction temperature, S/C ratio, W/F, and Ni loading level was investigated, with an optimized ESR condition, in terms of the ethanol conversion level and H2 yield, being established as an Ni loading of 10 wt%, 600 °C, S/C of 9, and W/F of 18 gcat h mol−1. At this condition, H2 and CNT yields of 55 and 14%, respectively, were simultaneously observed, while the highest yield of CNT (36%) along with an H2 yield of 29% was obtained at 600 °C with an Ni loading of 30 wt%, S/C of 1, and W/F 9 gcat h mol−1. The novel CNT-SF composites produced in ESR were uniformly distributed with an outer diameter of 20–30 nm. The CNT-SF composites offered the advantages of a large surface area and an entirely open structure, making them appealing for use as a catalyst support in various reactions.


Catalyst preparation

The Ni-SF catalyst was synthesized through the combination of sol–gel and electrospinning techniques followed by the conventional impregnation [28]. In the synthesis of the SF support, tetraethyl orthosilicate (TEOS) was used as the Si source. First, 18 cm3 of TEOS was dissolved in 3 cm3 of distilled water (DW) and stirred for 5 min, followed by the addition of 1 g of concentrated HCl. Then, 9.4 cm3 of ethanol was added into the mixed solution with continuous stirring for an additional 5 min. After that the solution was heated to 55 °C with continuous stirring for approximately 30 min. Finally, the mixed solution was transferred into a disposable syringe equipped with a 0.4 mm diameter needle. The electrospinning condition used for the synthesis of the SF support was a voltage of 15 kV, a feed rate of 5 cm3 h−1, and a tip-to-collector distance of 15 cm. The obtained SFs were dried overnight at 110 °C in an oven and then calcined at 500 °C for 2 h.

The Ni-SF catalysts were prepared by impregnation using Ni(NO3)2·6H2O as the Ni precursor. Ni(NO3)2·6H2O was dissolved in DW and then impregnated onto the SF support, dried at 110 °C overnight, and calcined at 500 °C for 2 h.

Catalyst characterization

Nitrogen (N2) physisorption via BET method (Brunauer–Emmett–Teller) on Micromeristics ASAP 2020 system was used to measure the surface area, pore diameter, and pore volume of the material, while X-ray diffraction (XRD) patterns were measured with D8 ADVANCE X-ray diffractometer (Bruker) equipped with Cu Kα radiation (λ = 1.5406 Å) with the radiation source CuKα, operating in the 2θ angle range of 5°–80° with a resolution of 0.02°. The patterns were identified using the Joint Committee on Powder Diffraction Standards (JCPDS cards). Scherrer’s equation was used to determine the average particle size of Ni metal. Temperature programmed reduction of hydrogen (H2-TPR) was conducted using a Micromeristics TPD/TPR 2920 system to determine the reducibility of the catalysts using a quartz reactor under 5% (v/v) H2 (N2 balance). The morphology of each fiber material was observed in field emission scanning electron microscopy (FESEM) using a JEOL JSM-7610F system. The average diameter and particle size of the Ni metal on the SF were estimated from the FESEM images using the SemAfore program. The actual metal content in the catalysts was determined by energy dispersive X-ray spectrometry (EDS; Oxford X-MaxN). Transmission electron microscopy (TEM; JEOL 2100F) was used to observe the surface morphology appearance of the carbon deposited on the Ni-SF catalyst from the ESR reaction. Thermal gravimetric analysis (TG/DTG) was performed on a METTLER TOLEDO STARe system under oxygen atmospheres at a heating rate of 5 °C min−1 from room temperature to 900 °C to study the weight changes and evaluate the yield of carbon deposited.

Catalyst testing

The co-production of H2 and CNTs was synthesized by ESR over the Ni-SF catalyst. The catalytic performance of the Ni-SF catalyst was conducted in a fixed-bed tube reactor. Before ESR, 0.2 g of the catalyst sample was reduced in situ with 10% (v/v) H2 (N2 balance) at 400 °C for 1 h. The ESR was performed at atmospheric pressure using N2 as the carrier gas for 360 min. A water–ethanol mixture was fed into the reactor via an HPLC pump. The effects of the reaction temperature (400, 500, and 600 °C), S/C molar ratio (1, 3, 6, 9, and 12), W/F (9, 12, 18, and 36 gcat h mol−1), and Ni loading level (10, 20, and 30 wt. %) on the gas production and CNTs deposition were investigated. Analysis of the products was performed by gas chromatography (GC) using a flame ionization detector (FID) to analyze the reaction organic products (methane, ethane, acetaldehyde, ethanol, and acetone) and a thermal conductivity detector (TCD) to analyze the gaseous products including H2, CO2, CH4, and CO.

The ethanol conversion level (XEtOH), H2 yield (YH2), and product selectivity were calculated using Eqs. (11)–(13), respectively [5]:
$${\text{X}}_{\text{EtOH}} /\% \;= \left( {F_{{{\text{EtOH}},{\text{ in}}}} {-}F_{{{\text{EtOH}},{\text{ out}}}} } \right)/F_{{{\text{EtOH}},{\text{ out}}}} \times 100$$
$${\text{Y}}_{\text{H2}} / \% \;= {\text{moles \; H}}_{ 2} \; {\text{produced}}/\left( { 6 \times {\text{moles \; EtOH \; converted}}} \right) \times 100$$
$${\text{S}}_{i} /\% \;= {\text{mole\; of gaseous product}}\;i/{\text{mole\; all gaseous products}} .$$

After the reaction, the CNTs obtained from the optimum condition were washed with a mixture of 3 mol dm−3 HCl and 1 mol dm−3 HNO3 solution under sonication at 40–50 °C for 4 h, then washed with DW to pH 7, and dried overnight at 120 °C to obtain the CNTs-SF composites. The morphology and physical properties of the CNTs-SF composites were investigated by TEM and N2 physisorption.



The authors are very grateful for financial support to Science Achievement Scholarship of Thailand (SAST), Overseas Academic Presentation Scholarship for Graduate Students, Graduate School, Chulalongkorn University, the research fund from the Thailand Research Fund (IRG5780001), The National Research Council of Thailand with the National Natural Science Foundation of China (NRCT-NSFC2558-104), and Science and Technology Research Partnership for Sustainable Development (SATREPS), Japan Science and Technology Agency (JST)/Japan International Cooperation Agency (JICA). The authors also acknowledge the support of HORIBA (Thailand) Limited for catalysts characterization by Raman model XploRA PLUS.


  1. 1.
    Ni M, Leung DYC, Leung MKH (2007) Int J Hydrogen Energy 32:3238CrossRefGoogle Scholar
  2. 2.
    Abreu AJD, Lucrédio AF, Assaf EM (2012) Fuel Process Technol 102:140CrossRefGoogle Scholar
  3. 3.
    Contreras JL, Salmones J, Colın-Luna JA, Nuno L, Quintana B, Cordova I, Zeifert B, Tapia C, Fuentes GA (2014) Int J Hydrogen 39:18835Google Scholar
  4. 4.
    Trane R, Dahl S, Skjøth-Rasmussen MS, Jensen AD (2012) Int J Hydrogen 37:6447CrossRefGoogle Scholar
  5. 5.
    Vicente J, Ereña J, Montero C, Azkoiti MJ, Bilbao J, Gayubo AG (2014) Int J Hydrogen 39:18820CrossRefGoogle Scholar
  6. 6.
    Popov VN (2004) Mater Sci Eng, R 43:61CrossRefGoogle Scholar
  7. 7.
    Paradise M, Goswami T (2007) Mater Des 28:1477CrossRefGoogle Scholar
  8. 8.
    Mubarak NM, Abdullah EC, Jayakumar NS, Sahu JN (2014) J Ind Eng Chem 20:1186CrossRefGoogle Scholar
  9. 9.
    Li D, Zeng L, Li X, Wang X, Ma H, Assabumrungrat S, Gong J (2015) Appl Catal B 176:532CrossRefGoogle Scholar
  10. 10.
    Lindo M, Vizcaíno AJ, Calles JA, Carrero A (2010) Int J Hydrogen 35:5895CrossRefGoogle Scholar
  11. 11.
    Reubroycharoen P, Tangkanaporn N, Chaiya C (2010) Stud Sur Sci Catal 175:689CrossRefGoogle Scholar
  12. 12.
    Sun C, Guo Y, Xu X, Du Q, Duan H, Chen Y, Li H, Liu H (2017) Compos A 92:33CrossRefGoogle Scholar
  13. 13.
    Chernyak SA, Suslova EV, Egorov AV, Lu L, Savilov SV, Lunin VV (2015) Fuel Process Technol 140:267CrossRefGoogle Scholar
  14. 14.
    Klaigaew K, Samart C, Chaiya C, Yoneyama Y, Tsubaki N, Reubroycharoen P (2015) Chem Eng J 278:166CrossRefGoogle Scholar
  15. 15.
    Jo SW, Kwak BS, Kim KM, Do JY, Park NK, Lee TJ, Lee ST, Kang M (2016) Chem Eng J 288:858CrossRefGoogle Scholar
  16. 16.
    Gil AG, Wu Z, Chadwick D, Li K (2015) Appl Catal A 506:188CrossRefGoogle Scholar
  17. 17.
    Li Z, Hu X, Zhang L, Liu S, Lu G (2012) Appl Catal A 417:281CrossRefGoogle Scholar
  18. 18.
    Xie H, Yu Q, Wei M, Duan W, Yao X, Qin Q, Zuo Z (2015) Int J Hydrogen 40:1420CrossRefGoogle Scholar
  19. 19.
    Mattos LV, Jacobs G, Davis BH, Noronha FB (2012) Chem Rev 112:4094CrossRefGoogle Scholar
  20. 20.
    Bej B, Pradhan NC, Neogi S (2014) Catal Today 237:80CrossRefGoogle Scholar
  21. 21.
    Compagnoni M, Tripodi A, Rossetti I (2017) Appl Catal B 203:899CrossRefGoogle Scholar
  22. 22.
    Pinilla JL, Utrilla R, Karn RK, Suelves I, La´zaro MJ, Moliner R, Garcı´a AB, Rouzaud JN (2011) Int J Hydrogen 36:7832Google Scholar
  23. 23.
    Wang G, Wang H, Li W, Ren Z, Bai J, Bai J (2011) Fuel Process Technol 92:531CrossRefGoogle Scholar
  24. 24.
    Palacio R, Gallego J, Gabelica Z, Batiot-Dupeyrat C, Barrault J, Valange S (2015) Appl Catal B 504:642CrossRefGoogle Scholar
  25. 25.
    Wang H, Liu Y, Wang L, Qin YN (2008) Chem Eng J 145:25CrossRefGoogle Scholar
  26. 26.
    Hou P, Liu C, Cheng HM (2008) Carbon 46:2003CrossRefGoogle Scholar
  27. 27.
    Li F, Wang Y, Wang D, Wei F (2004) Carbon 42:2375CrossRefGoogle Scholar
  28. 28.
    Ratanathavorn W, Samart C, Reubroycharoen P (2015) Mater Lett 159:135CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Wien 2017

Authors and Affiliations

  • Natthawan Prasongthum
    • 1
  • Chaiyan Chaiya
    • 2
  • Chanatip Samart
    • 3
  • Guoqing Guan
    • 4
  • Paweesuda Natewong
    • 5
    • 6
  • Prasert Reubroycharoen
    • 5
    • 6
  1. 1.Program in Petrochemistry, Faculty of ScienceChulalongkorn UniversityBangkokThailand
  2. 2.Chemical Engineering Division, Faculty of EngineeringRajamangala University of Technology KrungthepBangkokThailand
  3. 3.Department of Chemistry, Faculty of Science and TechnologyThammasat UniversityPathumthaniThailand
  4. 4.North Japan Research Institute of Sustainable EnergyHirosaki UniversityAomoriJapan
  5. 5.Center of Excellence on Petrochemical and Materials Technology, Chulalongkorn University Research BuildingBangkokThailand
  6. 6.Department of Chemical Technology, Faculty of Science, and Center of Excellence in Catalysis for Bioenergy and Renewable ChemicalsChulalongkorn UniversityBangkokThailand

Personalised recommendations