- Email: [email protected]
Fabrication of hollow silica–alumina composite spheres and their activity for hydrolytic dehydrogenation of ammonia borane
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 1 3 6 e1 7 1 4 3
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/he
Fabrication of hollow silicaealumina composite spheres and their activity for hydrolytic dehydrogenation of ammonia borane Naoki Toyama, Tetsuo Umegaki*, Yoshiyuki Kojima Department of Materials & Applied Chemistry, College of Science & Engineering, Nihon University, 1-8-14, Kanda-Surugadai, Chiyoda-Ku, Tokyo 101-8308, Japan
article info
abstract
Article history:
In this study, we investigated the influence of the preparation conditions of hollow silica
Received 20 May 2014
ealumina composite spheres on their activity for the hydrolytic dehydrogenation of
Received in revised form
ammonia borane. Hollow silicaealumina composite spheres were prepared by polystyrene
28 July 2014
template method, and the polystyrene template particles were removed by calcination.
Accepted 10 August 2014
The as-prepared hollow spheres were calcined at 523e873 K for 3 h. From the results of
Available online 10 September 2014
elemental analysis, polystyrene templates were completely removed by calcination at 873 K. small particles around the hollow spheres were observed from the images of
Keywords:
transmission electron microscopy. To obtain homogeneous hollow spheres, the as-
Ammonia borane
prepared hollow spheres were calcined at 873 K for 0e12 h. From the results of trans-
Hollow silicaealumina composite
mission electron microscopy, homogeneous hollow spheres were obtained by calcination
spheres
for 0 h. The activity of the hollow spheres was the 2.6 times higher that of the hollow
Preparation conditions
spheres calcined for 3 h. From the results of activity tests and ammonia temperature-
Hydrolytic dehydrogenation
programming desorption, the activity of the hollow spheres depends on amount of acid
Acid sites
sites. Copyright © 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
Introduction Hydrogen has recently attracted much attention as an alternative energy source to satisfy the increasing demand for effective and clean energy because of its abundance, high energy density, and environmental friendliness [1e3]. However, several challenges need to be addressed before transitioning to a hydrogen economy. Therefore, it is necessary to develop new hydrogen storage materials and efficient catalysts for the hydrogen release reaction [4,5]. Among the
various materials that can be used for hydrogen storage, ammonia borane (NH3BH3) is an attractive candidate because of its high hydrogen capacity (19.6 wt%), and high stability in the solid state or in solution [3,6e20]. Furthermore, this compound can release highly pure hydrogen in the presence of an appropriate catalysts or acids at room temperature, as shown by Eq. (1) [3,7,14,18,20]. NH3 BH3 þ 2H2 O/NHþ 4 þ BO2 þ 3H2
(1)
A number of catalysts or acids have been recently utilized for the hydrolysis of NH3BH3 [3,6e20]. Among these catalysts
* Corresponding author. Tel.: þ81 3 3259 0810; fax: þ81 3 3293 7572. E-mail address: [email protected] (T. Umegaki). http://dx.doi.org/10.1016/j.ijhydene.2014.08.057 0360-3199/Copyright © 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 1 3 6 e1 7 1 4 3
and acids, it has been reported that solid acids such as H-type zeolites (Beta zeolites and Mordenite) exhibit activity for the hydrolysis of NH3BH3 [11]. However, there have been few investigations into solid acids regarding the correlation between their structure and activity. To clarify this correlation, we have focused on hollow spheres. The most frequently used method to prepare hollow spheres is the template-based route. The template method can be classified into two possible methods [21]: the soft template method [22e24] and the hard template method [25e27]. The hard template method in particular has attracted much interest because it can be used to prepare monodisperse particles and easily control the shell and hollow structures. However, the removal of the core particles by selective dissolution in an appropriate solvent or by calcination at optimal conditions in the air is crucial to obtain hollow spheres. Microsized, monodisperse, hollow silica [28,29] and titania [30] spheres have been recently fabricated via a one-step process, with the formation of inorganic shells and the dissolution of the core particles occurring within the same medium. Using this method, microsized, monodisperse, positively charged polystyrene (PS) particles were prepared through dispersion polymerization using the cationic monomer 2-(methacryloyl)-ethyltrimethylammonium chloride (MTC) as the co-monomer [31] or through emulsifier-free emulsion polymerization using a,a0 -azodiisobutyramidine dihydrochloride (AIBA) as the initiator and poly(vinyl pyrrolidone) (PVP) as the stabilizer [32,33]. These PS template particles, which lacked a positively charged co-monomer could also be dissolved in the same medium subsequently after, or even simultaneously, during the coating of the silica shells to directly form the hollow spheres. In previous study, PS template particle were incompletely dissolved in the same medium although hollow silica-nickel composite spheres were prepared through the PS template method. The amount of PS residues is able to be reduced by increasing aging time and amount of aqueous ammonia solution for the preparation, and the catalytic activity of the hollow spheres increases when the amount of PS residues decreases [12]. In this study, we first investigated the influence of preparation condition of hollow silicaealumina composite spheres using the PS template method and subsequently investigated their activity for the hydrolytic dehydrogenation of NH3BH3.
Experimental Hollow composite spheres preparation Hollow silicaealumina spheres were fabricated using the PS template method [31]. Monodisperse PS particles were prepared through emulsifier-free emulsion polymerization as follows: 9.0 mL of styrene (Kanto Chem. Co.), 1.5 g of PVP K30 (Mw z 40,000, Fluka), 0.26 g of the cationic initiator AIBA (Kanto Chem. Co.), and 100.0 mL of deionized water were placed in a 250-mL three-neck flask equipped with a mechanical stirrer, a thermometer with a temperature controller, a N2 inlet, a Graham condenser, and a heating oil bath. The reaction solution was first deoxygenated by bubbling N2 gas
17137
through the solution at room temperature for 1 h. The reaction was then carried out at 343 K for 24 h at a stirring rate of 250 rpm. The obtained PS suspension was centrifuged at 6000 rpm for 5 min and washed with ethanol 3 times; the contents of the PS suspension could be fine-tuned through the addition of appropriate amount of ethanol. 0.0057 g of aluminum isopropoxide (Kanto Chem. Co., >99.0%), 3 mL of aqueous ammonia solution (Kanto Chem. Co., 28 wt%), and 40 mL of ethanol were added to 15 g of the PS suspension, followed by the addition of 155.1 mL of tetraethoxysilane (Kanto Chem. Co., >99.9%). The solegel reaction was then carried out at 323 K for 1.5 h, and the as-prepared hollow spheres were obtained. After drying in a desiccator overnight, hollow silicaealumina spheres were obtained upon calcination under various conditions in air. The obtained white powders were subsequently used in the hydrolysis experiments. After drying in a desiccator overnight, the recycled hollow spheres were obtained. The obtained white powders were also used in the hydrolysis experiments.
Characterization The morphologies of as-prepared hollow silicaealumina composite spheres and the hollow spheres calcined various conditions were observed by mean of transmission electron microscopy (TEM) using a Hitachi FE2000 microscope operating at an acceleration voltage of 200 kV. Fourier transform infrared (FT-IR) spectra of the as-prepared hollow spheres and the hollow spheres calcined various temperature were recorded using a PerkineElmer FT-IR with a resolution of 4 cm1 in the angle attenuated total reflectance (ATR) sampling mode. The hollow spheres powder was placed on the horizon of the internal reflectance crystal where the total internal reflection occurred. A ZnSe crystal with a transmission range of 4000e650 cm1 was used in this experiment. The thermal dissolution temperature of the PS residues within the asprepared hollow spheres was conducted using thermogravimetric e differential thermal analysis (TG-DTA) using Rigaku Thermo plus TG-8120, operated under air with a heating rate of 24 K min1 from room temperature to 1073 K. The amount of PS residues within the as-prepared hollow spheres and the hollow spheres calcined various conditions was estimated by the elemental analysis of C, H, and N using MICRO CORDER JM10. The amount of acid sites of the hollow spheres was measured by neutralization titration with n-butyl amine (Kanto Chem. Co.). About 0.2 g of the hollow spheres was dispersed in 20 mL of ethanol under sonication. The suspension was titrated with a 0.1 M n-butylamine using methyl red (Kanto Chem. Co.) as an indicator. The amount of acid sites of the hollow spheres was calculated as follow: The amount of 0.1 M n-butylamine was divided by amount of composites. Temperature-programmed desorption of ammonia (NH3-TPD) were carried out on BELCAT-B instrument. The analysis was performed by loading 50 mg of the sample into quartz reactor and drying it in a pure He flow at 783 K for 1 h followed by a pure He purge at the same temperature for 1 h. After the sample was cooled to 373 K by a pure He flow, NH3 adsorption was carried out by exposing the sample to NH3eHe gas mixture (95 vol.% He) and then keeping it at 373 K for 1 h. The sample was then purged using a pure He purge allowing the
17138
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 1 3 6 e1 7 1 4 3
accurate detection of the desorbed NH3. The NH3-TPD measurements were conducted using a pure He flow by heating the sample from 373 K to 873 K at a rate of 10 K min1, and the desorbed NH3 molecules were detected by thermal conductivity detector (TCD). The desorption peaks were deconvoluted by using Gaussian function with temperature as variant.
Experimental procedures for the hydrolysis of NH3BH3 Hollow silicaealumina composite spheres (0.8 g) and H-BEA zeolites (JRC-Z-HB25(1), Catalysis Society of Japan, 0.8 g) were placed in a two-neck round-bottomed flask in air at room temperature: one neck was connected to a gas burette, and the other was connected to an addition funnel. The reaction was initiated by stirring the mixture of the hollow spheres; the aqueous NH3BH3 solution (0.14 wt%, 3.5 mL, Aldrich, 90%) was added from the addition funnel, and the evolution of gas was monitored by using the gas burette.
Result and discussion To confirm calcination temperature removing PS residues within the as-prepared composites, TG-DTA curves for the asprepared composites were obtained, as shown in Fig. 1. From the results of the TG analysis, the as-prepared composites exhibited a weight loss of 60.7% in the temperature range of 573e923 K, which corresponds to the decomposition of the PS residues within the as-prepared composites. The DTA curve revealed an exothermic peak at 670 K due to the loss of the decomposition of the PS residues. The result indicates that the PS residues were included within the as-prepared composites. To remove PS residues in the as-prepared composites, the as-prepared composites were calcined at various temperatures. The as-prepared composites calcined at various temperatures were characterized using FT-IR spectroscopy and elemental analysis. Fig. 2 display the FT-IR spectra of the asprepared composite and the composites calcined at 573, 723, and 873 K for 3 h. The as-prepared composites and the composites calcined at 573 K exhibited the characteristic bonds of PS at approximately 1600, 1492, 1449, 754 and, 698 cm1 [29], while the composites calcined at 723 and 873 K did not exhibit the characteristic bonds of PS. In addition, the results of elemental analysis confirmed that the amount of carbon in the as-prepared composites and composites calcined 573, 723, and 873 K were 70, 43, 11, and 0.3 wt%, respectively. From
0
-40
DTA [μV]
TG [wt%]
-20 DTA
-60 -80
-100 273
TG 473
100 50 0 -50 -100 -150 -200 -250 1073
673
873
Temperature [K] Fig. 1 e TG-DTA curve for as-prepared hollow silicaealumina composite spheres.
Fig. 2 e ATR-IR spectra of (a) as-prepared hollow silicaealumina composite spheres, and hollow silicaealumina composite spheres calcined at (b) 573, (c) 723, and (d) 873 K for 3 h.
these results, the PS residues appeared to be completely removed by the calcination above 873 K, which indicates that the amount of PS residues decreased as the calcination temperature increased. The morphologies of composites calcined at various temperatures were evaluated using TEM. Fig. 3(a), (b), (c), and (d) shows the TEM images of the as-prepared composites and the composites calcined at 573, 723, and 873 K for 3 h, respectively. From the results, homogeneous hollow spheres were obtained all the composites. On the other hands, the composites calcined at 873 K were observed small particles around the hollow spheres. These small particles were probably almost silica particles because the chemical composition of the hollow spheres was about Si/Al ¼ 25/1. Activity for hydrolytic dehydrogenation of NH3BH3 of the as-prepared hollow spheres and the hollow spheres calcined at 573, 723, and 873 K for 3 h were compared. Fig. 4 shows time course of hydrogen generation from aqueous NH3BH3 solution in the presence of the as-prepared hollow spheres and the hollow spheres calcined at 573, 723, and 873 K, respectively. The evolution of 2.5, 4, 8, and 4 mL hydrogen was finished in 5, 8, 14, and 6 min in the presence of the as-prepared hollow spheres and the hollow spheres calcined at 573, 723, and 873 K, respectively. The molar ratios of hydrolytically generated hydrogen to the initial NH3BH3 were 0.6, 1.0, 2.1, and 1.0, respectively. On the other hand, amount of acid sites of the asprepared hollow spheres and the hollow spheres calcined at 573, 723, and 873 K were measured by neutralization titration. From the results, amount of acid sites of the as-prepared hollow spheres and the hollow spheres calcined at 573, 723, and 873 K were 0, 0.05, 0.11, and 0.07 mmol g1, respectively. In the present study, same weight of the samples was used for estimation of their activity, and then, the amount of active component in the samples used for the estimation depends on the amount of the PS residues. The results in Fig. 4(a)e(c) suggest that activity of the composites depends on the amount of the PS residues. On the other hand, from the result of Fig. 4(c) and (d), the hollow spheres calcined at 873 K were much lower activity than the hollow spheres calcined at 723 K despite of its lower carbon content than that in the hollow
17139
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 1 3 6 e1 7 1 4 3
Fig. 3 e TEM images of (a) as-prepared hollow silicaealumina composite spheres and the hollow spheres calcined at (b) 573, (c) 723, and (d) 873 K for 3 h.
spheres calcined at 723 K. The result suggests that acid sites of the hollow spheres decrease because of agglomeration of silica and/or alumina particles. Fig. 5 shows the H2/NH3BH3 molar ratios of the hydrogen generated from the aqueous NH3BH3 solution versus the amount of the PS residues within the hollow spheres calcined at 573, 723 and 873 K for 3 h. From the result of the hollow spheres calcined at 573 and 723 K, if the activity depends on amount of PS residues as shown in Fig. 5(a) and (b), hollow
spheres with negligible amount of PS residues show about 2.4 in estimated activity. It is probably that homogeneous hollow spheres with negligible amount of PS residues show higher activity. To obtain homogeneous hollow spheres with negligible amount of PS residues, the as-prepared hollow spheres were calcined at 873 K while varying the holding time. Fig. 6(a) and (b) shows the TEM images of composites calcined at 873 K for 0 and 12 h, respectively. From the results, homogeneous hollow spheres were obtained upon calcination for 0 h while particles with irregular shapes were obtained upon
10 (c)
8 6
(d)
4
(b)
(a)
2 00
5
10
15
20 25 30 Reaction time [min]
35
40
Fig. 4 e Hydrogen generation from the aqueous NH3BH3 solution (0.14 wt%, 3.5 mL) in the presence of hollow silicaealumina composite spheres calcined at (a) 573, (b) 723, (c) 873 K for 3 h, and (d) as-prepared hollow silicaealumina composite spheres. Amount of composites: 0.8 g.
H2 / NH3BH3 mole ratio [ - ]
Hydrogen generation [ mL ]
12
3.0 Estimated activity
2.5
(b)
2.0 1.5
(c)
(a)
1.0 0.5 0
0
10
20 30 40 Carbon content [wt. % ]
50
Fig. 5 e The molar ratio of hydrogen generated from the aqueous NH3BH3 solution (H2/NH3BH3) (0.14 wt%, 3.5 mL) versus the amount of PS residues of the hollow silicaealumina composite spheres calcined at (a) 573, (b) 723, and (c) 873 K for 3 h.
17140
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 1 3 6 e1 7 1 4 3
Fig. 6 e TEM images of hollow silicaealumina composite spheres calcined at 873 K for (a) 0 h, silicaealumina composites calcined at 873 K for (b) 12 h, and (c) recycled hollow silicaealumina composite spheres.
composites calcined for 12 h were measured by neutralization titration. From the results, amount of acid sites of the hollow spheres calcined 0 and 3 h, and the composites calcined for 12 h were 0.12, 0.07, and 0.05 mmol g1, respectively. Furthermore, the activity in the presence of hollow spheres calcined for 0 h were same amount of hydrogen, and higher rate of hydrogen generation than that in the presence of HBEA zeolites. Moreover, homogeneous hollow spheres with
12 Hydrogen generation [mL]
calcination for 12 h. From the results, composites calcined for 12 h were not obtained hollow spheres. Additionally, all the hollow spheres calcined at 873 K were completely free of PS residues from the result of elemental analysis. These results indicate that homogeneous hollow spheres with negligible amount of PS residues could be prepared by adjusting the calcination temperature and holding time. Moreover, homogeneous hollow spheres with negligible amount of PS residues were even prepared by calcination at 723 K for 24 h. To determine the effects of the morphologies of the hollow spheres and composites on hydrolytic dehydrogenation of NH3BH3, the activities of the hollow spheres calcined at 873 K for 0, 3, 12 h and H-BEA zeolites were compared. Fig. 7(a), (b), (c), and (d) shows time course of hydrogen generation from aqueous NH3BH3 solution in the presence of the hollow spheres calcined at 873 K for 0 and 3 h, the composites calcined at 873 K for 12 h, and H-BEA zeolites. The evolution of 10, 4, 3.5, and 9.4 mL of hydrogen were finished in 12, 7, 3, and 38 min in the presence of the hollow spheres calcined for 0 and 3 h, the composites calcined for 12 h, and HBEA zeolites, respectively. The molar ratios of hydrolytically generated hydrogen to the initial NH3BH3 were 2.6, 1.0, 0.9, and 2.5, respectively. From these results, the activity increased as the holding time decreased. Additionally, specific surface area of the hollow spheres calcined for 0 and 3 h, and the composites calcined for 12 h were 393, 423, and 308 m2 g1, respectively. The results indicate that the activity does not depend on specific surface area of the hollow spheres and composites, and the activity is likely dependent on acid sites of the hollow spheres and composites. Moreover, amount of acid sites of the hollow spheres calcined for 0 and 3 h, and the
(a)
10
(d)
8 6 (b)
4
(c)
2 (e)
00
5
10
15
20 25 30 Reaction time [min]
35
40
Fig. 7 e Hydrogen generation from aqueous NH3BH3 solution (0.14 wt%, 3.5 mL) in the presence of hollow silicaealumina composite spheres calcined at 873 K for (a) 0, (b) 3, (c) 12 h, (d) H-BEA zeolites, and (e) recycled hollow silicaealumina composite spheres. Amount of the composites: 0.8 g.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 1 3 6 e1 7 1 4 3
negligible amount of PS residues almost show same estimated activity from the result of Fig. 5(a) and (b). On the other hand, our groups reported that hollow silicaezirconia composite spheres could successfully prepare by PS template methods. Hollow silicaealumina composite spheres were higher amount of hydrogen generation than hollow silicaezirconia composite spheres [34]. Additionally, Fig. 6(c) shows the TEM images of the recycled hollow spheres. From the result of Fig. 6(a) and (c), the morphologies of the recycled hollow spheres were not significantly different from each other. To determine the recycle ability of the hollow spheres, the activity of the recycled hollow spheres were compared. The evolution of 1.5 mL of hydrogen was finished in 2 min in the presence of the recycled hollow spheres as shown in Fig. 7(e). From the results, the recycled hollow spheres were much lower activity than the hollow spheres calcined for 0 h. The amount of acid sites of the recycled hollow spheres was 0 mmol g1. The results indicate that almost all the acid sites on the hollow spheres were consumed via the hydrolysis reaction. It has been previously reported that the activation process of NH3BH3 occurs because the acidic protons from the solid acids in the aqueous medium (which promote the dissociation of the BeN bond and the hydrolysis of BH3 species to produce borate ions species along with the release of hydrogen because the acidity of the solid acids is higher than that of boric acid) the results in the borate ions reacting with Hþ to produce boric acid [11]. Based on the results, the effective acid sites for hydrolytic dehydrogenation of NH3BH3 were likely Brønsted acid sites. These results indicate that the amount of Brønsted acid sites of the hollow spheres had a significant impact on their activity. The amount of Brønsted acid sites of the hollow spheres calcined at 873 K for 0 and 3 h, and the composites calcined at 873 K for 12 h was measured using NH3-TPD. Fig. 8 shows NH3TPD profile of the hollow spheres calcined at 873 K for 0 and 3 h, and the composites calcined at 873 K for 12 h. It has been reported that the NH3-TPD profile shows a peak around 400e600 K, indicating the presence of Brønsted acid sites, and a peak over 650 K, indicating the Lewis acid sites [35e37].
Based on the results, NH3-TPD profiles of the composites calcined at 873 K for 0, 3, and 12 h, first peak around at 400e523 K and the broad peak around at 523e600 K was attribute to the presence of Brønsted acid sites. On the other hands, the broad peak around 650e773 K and second peak around 773 K was attribute to the presence of Lewis acid sites. The amount of the Brønsted acid sites calculated from the peak area in the temperature range of about 400e600 K of the hollow spheres calcined for 0 and 3 h, and the composites calcined for 12 h were 0.12, 0.08, and 0.07 mmol g1, respectively. Fig. 9 shows the H2/NH3BH3 molar ratio of hydrogen generated from the aqueous NH3BH3 solution versus the amount of acid sites of the hollow silicaealumina composite spheres calcined at 873 K for 0 and 3 h, and the composites calcined at 873 K for 12 h, respectively. From the result, the activity increased as the amount of Brønsted acid sites in the hollow spheres and composites increased. These results indicated that the activity depends on the amount of the Brønsted acid sites of the hollow spheres and composites.
Conclusion In this study, we investigated the influence of the preparation conditions of hollow silicaealumina composite spheres on their activity for the hydrolytic dehydrogenation of NH3BH3. Hollow silicaealumina composite spheres were prepared by PS template method, and the PS templates particles were removed by calcination. The as-prepared hollow spheres were calcined at 523e873 K for 3 h. From the results of elemental analysis, PS residues in the hollow spheres were completely removed by calcination at 873 K. small particles around hollow spheres were observed from the images of TEM. To obtain homogeneous hollow spheres, as-prepared hollow spheres were calcined at 873 K for 0e12 h. From the results of TEM images, homogeneous hollow spheres were obtained via calcination at 873 K for 0 h. The activities of the hollow spheres calcined at 873 K for 0 and 3 h for hydrolytic
H2 / NH3BH3 mole ratio [ - ]
3.0
TCD signal [μV]
(a)
(b)
(c)
373
473
2.5
673
773
873
Temperature [ K ] Fig. 8 e NH3-TPD profile of hollow silicaealumina composite spheres calcined at 873 K for (a) 0 and (b) 3, and silicaealumina composite calcined at 873 K for (c) 12 h.
(a)
R2
2.0 1.5 1.0
(b) (c)
0.5 0
573
17141
0
0.05 0.10 0.15 Amount of Brønsted acid sites [mmol g-1]
0.20
Fig. 9 e The H2/NH3BH3 molar ratio of the hydrogen generated from the aqueous NH3BH3 solution (0.14 wt%, 3.5 mL) versus the amount of Brønsted acid sites of hollow silicaealumina composite spheres calcined at 873 K for (a) 0 and (b) 3 h, and silicaealumina composites calcined at 873 K for (c) 12 h.
17142
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 1 3 6 e1 7 1 4 3
dehydrogenation of aqueous NH3BH3 solution were compared. The activity of hollow spheres calcined for 0 h was 2.6 times higher that of hollow spheres calcined for 3 h. From the results of NH3-TPD, the activity increased as amount of Brønsted acid sites increased. The result indicated that the activity depends on amount of Brønsted acid sites.
[16]
[17]
Acknowledgment Special thanks are given to Dr. T. Yoneda for assistance with the use of NH3-TPD and to Ms. M. Takagi for helping with elemental analysis.
references
[18]
[19]
[20] [1] Sperling D, DeLuchi MA. Transportation energy futures. Annu Rev Energy 1989;14:375e424. [2] Schiapbach L, Zu¨ttel A. Hydrogen-storage materials for mobile applications. Nature 2001;414:353e8. [3] Umegaki T, Yan JM, Zhang XB, Shioyama H, Kuriyama N, Xu Q. Preparation and catalysis of poly(N-vinyl-2pyrrolidone) (PVP) stabilized nickel catalyst for hydrolytic dehydrogenation of ammonia borane. Int J Hydrogen Energy 2009;34:3816e22. [4] Hu¨gle T, Ku¨hnel MF, Lentz D. Hydrazine borane: a promising hydrogen storage material. J Am Chem Soc 2009;131:7444e6. [5] Jena P. Materials for hydrogen storage: past, present, and future. J Phys Chem Lett 2011;2:206e11. [6] Brockman A, Zheng Y, Gore J. A study of catalytic hydrolysis of concentrated ammonia borane solutions. Int J Hydrogen Energy 2010;35:7350e6. [7] Umegaki T, Xu Q, Kojima Y. Effect of L-arginine on the catalytic activity and stability of nickel nanoparticles for hydrolytic dehydrogenation of ammonia borane. J Power Sources 2012;216:363e7. [8] Chandra M, Xu Q. Room temperature hydrogen generation from aqueous ammonia-borane using noble metal nanoclusters. J Power Sources 2007;168:135e42. [9] Kalidindi SB, Sanyal U, Jagirdar BR. Nanostructured Cu and Cu@Cu2O core shell catalysts for hydrogen generation from ammonia-borane. Phys Chem Chem Phys 2010;10:5870e4. [10] Umegaki T, Yan JM, Zhang XB, Shioyama H, Kuriyama N, Xu Q. Hollow Ni-SiO2 nanosphere-catalyzed hydrolytic dehydrogenation of ammonia borane for chemical hydrogen storage. J Power Sources 2009;191:209e16. [11] Chandra M, Xu Q. Dissociation and hydrolysis of ammoniaborane with solid acids and carbon dioxide: an efficient hydrogen generation system. J Power Sources 2006;159:855e60. [12] Umegaki T, Ohashi T, Xu Q, Kojima Y. Influence of preparation conditions of hollow titania-nickel composite spheres on their catalytic activity for hydrolytic dehydrogenation of ammonia borane. Mater Res Bull 2014;52:117e21. € [13] Zahmakıran M, Durap F, Ozkar S. Zeolite confined copper(0) nanoclusters as cost-effective and reusable catalyst in hydrogen generation from the hydrolysis of ammoniaborane. Int J Hydrogen Energy 2010;35:187e97. € [14] Rakap M, Ozkar S. Hydrogen generation from the hydrolysis of ammonia-borane using intrazeolite cobalt(0) nanoclusters catalyst. Int J Hydrogen Energy 2010;35:3341e6. [15] Tong DG, Zeng XL, Chu W, Wang D, Wu P. Magnetically recyclable hollow Co-B nanospindles as catalysts for
[21]
[22] [23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
hydrogen generation from ammonia borane. J Mater Sci 2010;45:2862e7. Jiang HL, Umegaki T, Akita T, Zhang XB, Haruta M, Xu Q. Bimetallic Au-Ni nanoparticles embedded in SiO2 nanospheres: synergetic catalysis in hydrolytic dehydrogenation of ammonia borane. Chem Eur J 2010;16:3132e7. Zahmakıran M, Ayvalı T, Philippot K. In situ formed catalytically active ruthenium nanocatalyst in room temperature dehydrogenation/dehydrocoupling of ammonia-borane from Ru (cod) (cot) precatalyst. Langmuir 2012;28:4908e14. Basu S, Brockman A, Gagare P, Zheng Y, Ramachandran PV, Delgass WN, et al. Chemical kinetics of Ru-catalyzed ammonia borane hydrolysis. J Power Sources 2009;188:238e43. Cheng F, Ma H, Li Y, Chen J. Ni1xPtx (x ¼ 0e0.12) hollow spheres as catalysts for hydrogen generation from ammonia borane. lnorg Chem 2007;46:788e94. € Rakap M, Ozkar S. Zeolite confined palladium(0) nanoclusters as effective and reusable catalyst for hydrogen generation from the hydrolysis of ammonia-borane. Int J Hydrogen Energy 2010;35:1305e12. Zoldesi CI, Imho A. Synthesis of monodisperse colloidal spheres, capsules, and microballoons by emulsion templating. Adv Mater 2005;17(7):924e8. Schmidt HT, Ostafin AE. Liposome directed growth of calcium phospate nanoshells. Adv Mater 2002;14(7):532e5. Nakashima T, Kimizuka N. Interfacial synthesis of hollow TiO2 microspheres in ionic liquids. J Am Chem Soc 2003;125:6386e7. Putlitz BZ, Landfester K, Fischer H, Antonietti M. The generation of “armored latexes” and hollow inorganic shells made of clay sheets by templating cationic miniemulsions and latexes. Adv Mater 2001;13(7):500e3. Arnal PM, Weidenthaler C, Schu¨th F. Highly monodisperse zirconia-coated silica spheres and zirconia/silica hollow spheres with remarkable textural properties. Chem Mater 2006;18:2733e9. Titirici M, Antonietti M, Thomas A. A generalized synthesis of metal oxide hollow spheres using a hydrothermal approach. Chem Mater 2006;18:3808e12. Yang Z, Niu Z, Lu Y, Hu Z, Han C. Templated synthesis of inorganic hollow spheres with a tunable cavity size onto core-shell gel particles. Angew Chem Int Ed 2003;42:1943e5. Deng Z, Chen M, Zhou S, You B, Wu L. A novel method for the fabrication of monodisperse hollow silica spheres. Langmuir 2006;22:6403e7. Leng W, Chen M, Zhou S, Wu L. Capillary force induced formation of monodisperse polystyrene/silica organicinorganic hybrid hollow spheres. Langmuir 2010;26(17):14271e5. Cheng X, Chen M, Wu L, Gu G. Novel and facile method for the preparation of monodispersed titania hollow spheres. Langmuir 2006;22(8):3858e63. Chen M, Zhou S, You B, Wu L. A novel preparation method of raspberry-like PMMA/SiO2 hybrid. Macromolecules 2005;38:6411e7. Song X, Gao L. Fabrication of hollow hybrid microspheres coated with silica/titania via sol-gel process and enhanced photocatalytic activities. J Phys Chem C 2007;111:8180e7. Yokoi T, Sakamoto Y, Terasaki O, Kubota Y, Okubo T, Tatsumi T. Periodic arrangement of silica nanospheres assisted by amino acids. J Am Chem Soc 2006;128:13664e5. Umegaki T, Hosoya T, Toyama N, Xu Q, Kojima Y. Fabrication of hollow silica-zirconia composite spheres and their activity
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 1 3 6 e1 7 1 4 3
for hydrolytic dehydrogenation of ammonia borane. J Alloy Compd 2014;608:261e5. [35] Sakthivel A, Dapurkar SE, Gupta NM, Kulshreshtha SK, Selvam P. The influence of aluminium sources on the acidic behaviour as well as on the catalytic activity of mesoporous H-AlMCM-41 molecular sieves. Microporous Mesoporous Mater 2003;65:177e87.
17143
[36] Wang Y, Lang N, Tuel A. Nature and acidity of aluminum species in AlMCM-41 with a high aluminum content (Si/ Al ¼ 1.25). Microporous Mesoporous Mater 2006;93:46e54. [37] Fang X, Liu Z, Hsieh M-F, Chen M, Liu P, Chen C, et al. Hollow mesoporous aluminosilica spheres with perpendicular pore channels as catalytic nanoreactors. ACS Nano 2012;6(5):4434e44.
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/he
Fabrication of hollow silicaealumina composite spheres and their activity for hydrolytic dehydrogenation of ammonia borane Naoki Toyama, Tetsuo Umegaki*, Yoshiyuki Kojima Department of Materials & Applied Chemistry, College of Science & Engineering, Nihon University, 1-8-14, Kanda-Surugadai, Chiyoda-Ku, Tokyo 101-8308, Japan
article info
abstract
Article history:
In this study, we investigated the influence of the preparation conditions of hollow silica
Received 20 May 2014
ealumina composite spheres on their activity for the hydrolytic dehydrogenation of
Received in revised form
ammonia borane. Hollow silicaealumina composite spheres were prepared by polystyrene
28 July 2014
template method, and the polystyrene template particles were removed by calcination.
Accepted 10 August 2014
The as-prepared hollow spheres were calcined at 523e873 K for 3 h. From the results of
Available online 10 September 2014
elemental analysis, polystyrene templates were completely removed by calcination at 873 K. small particles around the hollow spheres were observed from the images of
Keywords:
transmission electron microscopy. To obtain homogeneous hollow spheres, the as-
Ammonia borane
prepared hollow spheres were calcined at 873 K for 0e12 h. From the results of trans-
Hollow silicaealumina composite
mission electron microscopy, homogeneous hollow spheres were obtained by calcination
spheres
for 0 h. The activity of the hollow spheres was the 2.6 times higher that of the hollow
Preparation conditions
spheres calcined for 3 h. From the results of activity tests and ammonia temperature-
Hydrolytic dehydrogenation
programming desorption, the activity of the hollow spheres depends on amount of acid
Acid sites
sites. Copyright © 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
Introduction Hydrogen has recently attracted much attention as an alternative energy source to satisfy the increasing demand for effective and clean energy because of its abundance, high energy density, and environmental friendliness [1e3]. However, several challenges need to be addressed before transitioning to a hydrogen economy. Therefore, it is necessary to develop new hydrogen storage materials and efficient catalysts for the hydrogen release reaction [4,5]. Among the
various materials that can be used for hydrogen storage, ammonia borane (NH3BH3) is an attractive candidate because of its high hydrogen capacity (19.6 wt%), and high stability in the solid state or in solution [3,6e20]. Furthermore, this compound can release highly pure hydrogen in the presence of an appropriate catalysts or acids at room temperature, as shown by Eq. (1) [3,7,14,18,20]. NH3 BH3 þ 2H2 O/NHþ 4 þ BO2 þ 3H2
(1)
A number of catalysts or acids have been recently utilized for the hydrolysis of NH3BH3 [3,6e20]. Among these catalysts
* Corresponding author. Tel.: þ81 3 3259 0810; fax: þ81 3 3293 7572. E-mail address: [email protected] (T. Umegaki). http://dx.doi.org/10.1016/j.ijhydene.2014.08.057 0360-3199/Copyright © 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 1 3 6 e1 7 1 4 3
and acids, it has been reported that solid acids such as H-type zeolites (Beta zeolites and Mordenite) exhibit activity for the hydrolysis of NH3BH3 [11]. However, there have been few investigations into solid acids regarding the correlation between their structure and activity. To clarify this correlation, we have focused on hollow spheres. The most frequently used method to prepare hollow spheres is the template-based route. The template method can be classified into two possible methods [21]: the soft template method [22e24] and the hard template method [25e27]. The hard template method in particular has attracted much interest because it can be used to prepare monodisperse particles and easily control the shell and hollow structures. However, the removal of the core particles by selective dissolution in an appropriate solvent or by calcination at optimal conditions in the air is crucial to obtain hollow spheres. Microsized, monodisperse, hollow silica [28,29] and titania [30] spheres have been recently fabricated via a one-step process, with the formation of inorganic shells and the dissolution of the core particles occurring within the same medium. Using this method, microsized, monodisperse, positively charged polystyrene (PS) particles were prepared through dispersion polymerization using the cationic monomer 2-(methacryloyl)-ethyltrimethylammonium chloride (MTC) as the co-monomer [31] or through emulsifier-free emulsion polymerization using a,a0 -azodiisobutyramidine dihydrochloride (AIBA) as the initiator and poly(vinyl pyrrolidone) (PVP) as the stabilizer [32,33]. These PS template particles, which lacked a positively charged co-monomer could also be dissolved in the same medium subsequently after, or even simultaneously, during the coating of the silica shells to directly form the hollow spheres. In previous study, PS template particle were incompletely dissolved in the same medium although hollow silica-nickel composite spheres were prepared through the PS template method. The amount of PS residues is able to be reduced by increasing aging time and amount of aqueous ammonia solution for the preparation, and the catalytic activity of the hollow spheres increases when the amount of PS residues decreases [12]. In this study, we first investigated the influence of preparation condition of hollow silicaealumina composite spheres using the PS template method and subsequently investigated their activity for the hydrolytic dehydrogenation of NH3BH3.
Experimental Hollow composite spheres preparation Hollow silicaealumina spheres were fabricated using the PS template method [31]. Monodisperse PS particles were prepared through emulsifier-free emulsion polymerization as follows: 9.0 mL of styrene (Kanto Chem. Co.), 1.5 g of PVP K30 (Mw z 40,000, Fluka), 0.26 g of the cationic initiator AIBA (Kanto Chem. Co.), and 100.0 mL of deionized water were placed in a 250-mL three-neck flask equipped with a mechanical stirrer, a thermometer with a temperature controller, a N2 inlet, a Graham condenser, and a heating oil bath. The reaction solution was first deoxygenated by bubbling N2 gas
17137
through the solution at room temperature for 1 h. The reaction was then carried out at 343 K for 24 h at a stirring rate of 250 rpm. The obtained PS suspension was centrifuged at 6000 rpm for 5 min and washed with ethanol 3 times; the contents of the PS suspension could be fine-tuned through the addition of appropriate amount of ethanol. 0.0057 g of aluminum isopropoxide (Kanto Chem. Co., >99.0%), 3 mL of aqueous ammonia solution (Kanto Chem. Co., 28 wt%), and 40 mL of ethanol were added to 15 g of the PS suspension, followed by the addition of 155.1 mL of tetraethoxysilane (Kanto Chem. Co., >99.9%). The solegel reaction was then carried out at 323 K for 1.5 h, and the as-prepared hollow spheres were obtained. After drying in a desiccator overnight, hollow silicaealumina spheres were obtained upon calcination under various conditions in air. The obtained white powders were subsequently used in the hydrolysis experiments. After drying in a desiccator overnight, the recycled hollow spheres were obtained. The obtained white powders were also used in the hydrolysis experiments.
Characterization The morphologies of as-prepared hollow silicaealumina composite spheres and the hollow spheres calcined various conditions were observed by mean of transmission electron microscopy (TEM) using a Hitachi FE2000 microscope operating at an acceleration voltage of 200 kV. Fourier transform infrared (FT-IR) spectra of the as-prepared hollow spheres and the hollow spheres calcined various temperature were recorded using a PerkineElmer FT-IR with a resolution of 4 cm1 in the angle attenuated total reflectance (ATR) sampling mode. The hollow spheres powder was placed on the horizon of the internal reflectance crystal where the total internal reflection occurred. A ZnSe crystal with a transmission range of 4000e650 cm1 was used in this experiment. The thermal dissolution temperature of the PS residues within the asprepared hollow spheres was conducted using thermogravimetric e differential thermal analysis (TG-DTA) using Rigaku Thermo plus TG-8120, operated under air with a heating rate of 24 K min1 from room temperature to 1073 K. The amount of PS residues within the as-prepared hollow spheres and the hollow spheres calcined various conditions was estimated by the elemental analysis of C, H, and N using MICRO CORDER JM10. The amount of acid sites of the hollow spheres was measured by neutralization titration with n-butyl amine (Kanto Chem. Co.). About 0.2 g of the hollow spheres was dispersed in 20 mL of ethanol under sonication. The suspension was titrated with a 0.1 M n-butylamine using methyl red (Kanto Chem. Co.) as an indicator. The amount of acid sites of the hollow spheres was calculated as follow: The amount of 0.1 M n-butylamine was divided by amount of composites. Temperature-programmed desorption of ammonia (NH3-TPD) were carried out on BELCAT-B instrument. The analysis was performed by loading 50 mg of the sample into quartz reactor and drying it in a pure He flow at 783 K for 1 h followed by a pure He purge at the same temperature for 1 h. After the sample was cooled to 373 K by a pure He flow, NH3 adsorption was carried out by exposing the sample to NH3eHe gas mixture (95 vol.% He) and then keeping it at 373 K for 1 h. The sample was then purged using a pure He purge allowing the
17138
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 1 3 6 e1 7 1 4 3
accurate detection of the desorbed NH3. The NH3-TPD measurements were conducted using a pure He flow by heating the sample from 373 K to 873 K at a rate of 10 K min1, and the desorbed NH3 molecules were detected by thermal conductivity detector (TCD). The desorption peaks were deconvoluted by using Gaussian function with temperature as variant.
Experimental procedures for the hydrolysis of NH3BH3 Hollow silicaealumina composite spheres (0.8 g) and H-BEA zeolites (JRC-Z-HB25(1), Catalysis Society of Japan, 0.8 g) were placed in a two-neck round-bottomed flask in air at room temperature: one neck was connected to a gas burette, and the other was connected to an addition funnel. The reaction was initiated by stirring the mixture of the hollow spheres; the aqueous NH3BH3 solution (0.14 wt%, 3.5 mL, Aldrich, 90%) was added from the addition funnel, and the evolution of gas was monitored by using the gas burette.
Result and discussion To confirm calcination temperature removing PS residues within the as-prepared composites, TG-DTA curves for the asprepared composites were obtained, as shown in Fig. 1. From the results of the TG analysis, the as-prepared composites exhibited a weight loss of 60.7% in the temperature range of 573e923 K, which corresponds to the decomposition of the PS residues within the as-prepared composites. The DTA curve revealed an exothermic peak at 670 K due to the loss of the decomposition of the PS residues. The result indicates that the PS residues were included within the as-prepared composites. To remove PS residues in the as-prepared composites, the as-prepared composites were calcined at various temperatures. The as-prepared composites calcined at various temperatures were characterized using FT-IR spectroscopy and elemental analysis. Fig. 2 display the FT-IR spectra of the asprepared composite and the composites calcined at 573, 723, and 873 K for 3 h. The as-prepared composites and the composites calcined at 573 K exhibited the characteristic bonds of PS at approximately 1600, 1492, 1449, 754 and, 698 cm1 [29], while the composites calcined at 723 and 873 K did not exhibit the characteristic bonds of PS. In addition, the results of elemental analysis confirmed that the amount of carbon in the as-prepared composites and composites calcined 573, 723, and 873 K were 70, 43, 11, and 0.3 wt%, respectively. From
0
-40
DTA [μV]
TG [wt%]
-20 DTA
-60 -80
-100 273
TG 473
100 50 0 -50 -100 -150 -200 -250 1073
673
873
Temperature [K] Fig. 1 e TG-DTA curve for as-prepared hollow silicaealumina composite spheres.
Fig. 2 e ATR-IR spectra of (a) as-prepared hollow silicaealumina composite spheres, and hollow silicaealumina composite spheres calcined at (b) 573, (c) 723, and (d) 873 K for 3 h.
these results, the PS residues appeared to be completely removed by the calcination above 873 K, which indicates that the amount of PS residues decreased as the calcination temperature increased. The morphologies of composites calcined at various temperatures were evaluated using TEM. Fig. 3(a), (b), (c), and (d) shows the TEM images of the as-prepared composites and the composites calcined at 573, 723, and 873 K for 3 h, respectively. From the results, homogeneous hollow spheres were obtained all the composites. On the other hands, the composites calcined at 873 K were observed small particles around the hollow spheres. These small particles were probably almost silica particles because the chemical composition of the hollow spheres was about Si/Al ¼ 25/1. Activity for hydrolytic dehydrogenation of NH3BH3 of the as-prepared hollow spheres and the hollow spheres calcined at 573, 723, and 873 K for 3 h were compared. Fig. 4 shows time course of hydrogen generation from aqueous NH3BH3 solution in the presence of the as-prepared hollow spheres and the hollow spheres calcined at 573, 723, and 873 K, respectively. The evolution of 2.5, 4, 8, and 4 mL hydrogen was finished in 5, 8, 14, and 6 min in the presence of the as-prepared hollow spheres and the hollow spheres calcined at 573, 723, and 873 K, respectively. The molar ratios of hydrolytically generated hydrogen to the initial NH3BH3 were 0.6, 1.0, 2.1, and 1.0, respectively. On the other hand, amount of acid sites of the asprepared hollow spheres and the hollow spheres calcined at 573, 723, and 873 K were measured by neutralization titration. From the results, amount of acid sites of the as-prepared hollow spheres and the hollow spheres calcined at 573, 723, and 873 K were 0, 0.05, 0.11, and 0.07 mmol g1, respectively. In the present study, same weight of the samples was used for estimation of their activity, and then, the amount of active component in the samples used for the estimation depends on the amount of the PS residues. The results in Fig. 4(a)e(c) suggest that activity of the composites depends on the amount of the PS residues. On the other hand, from the result of Fig. 4(c) and (d), the hollow spheres calcined at 873 K were much lower activity than the hollow spheres calcined at 723 K despite of its lower carbon content than that in the hollow
17139
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 1 3 6 e1 7 1 4 3
Fig. 3 e TEM images of (a) as-prepared hollow silicaealumina composite spheres and the hollow spheres calcined at (b) 573, (c) 723, and (d) 873 K for 3 h.
spheres calcined at 723 K. The result suggests that acid sites of the hollow spheres decrease because of agglomeration of silica and/or alumina particles. Fig. 5 shows the H2/NH3BH3 molar ratios of the hydrogen generated from the aqueous NH3BH3 solution versus the amount of the PS residues within the hollow spheres calcined at 573, 723 and 873 K for 3 h. From the result of the hollow spheres calcined at 573 and 723 K, if the activity depends on amount of PS residues as shown in Fig. 5(a) and (b), hollow
spheres with negligible amount of PS residues show about 2.4 in estimated activity. It is probably that homogeneous hollow spheres with negligible amount of PS residues show higher activity. To obtain homogeneous hollow spheres with negligible amount of PS residues, the as-prepared hollow spheres were calcined at 873 K while varying the holding time. Fig. 6(a) and (b) shows the TEM images of composites calcined at 873 K for 0 and 12 h, respectively. From the results, homogeneous hollow spheres were obtained upon calcination for 0 h while particles with irregular shapes were obtained upon
10 (c)
8 6
(d)
4
(b)
(a)
2 00
5
10
15
20 25 30 Reaction time [min]
35
40
Fig. 4 e Hydrogen generation from the aqueous NH3BH3 solution (0.14 wt%, 3.5 mL) in the presence of hollow silicaealumina composite spheres calcined at (a) 573, (b) 723, (c) 873 K for 3 h, and (d) as-prepared hollow silicaealumina composite spheres. Amount of composites: 0.8 g.
H2 / NH3BH3 mole ratio [ - ]
Hydrogen generation [ mL ]
12
3.0 Estimated activity
2.5
(b)
2.0 1.5
(c)
(a)
1.0 0.5 0
0
10
20 30 40 Carbon content [wt. % ]
50
Fig. 5 e The molar ratio of hydrogen generated from the aqueous NH3BH3 solution (H2/NH3BH3) (0.14 wt%, 3.5 mL) versus the amount of PS residues of the hollow silicaealumina composite spheres calcined at (a) 573, (b) 723, and (c) 873 K for 3 h.
17140
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 1 3 6 e1 7 1 4 3
Fig. 6 e TEM images of hollow silicaealumina composite spheres calcined at 873 K for (a) 0 h, silicaealumina composites calcined at 873 K for (b) 12 h, and (c) recycled hollow silicaealumina composite spheres.
composites calcined for 12 h were measured by neutralization titration. From the results, amount of acid sites of the hollow spheres calcined 0 and 3 h, and the composites calcined for 12 h were 0.12, 0.07, and 0.05 mmol g1, respectively. Furthermore, the activity in the presence of hollow spheres calcined for 0 h were same amount of hydrogen, and higher rate of hydrogen generation than that in the presence of HBEA zeolites. Moreover, homogeneous hollow spheres with
12 Hydrogen generation [mL]
calcination for 12 h. From the results, composites calcined for 12 h were not obtained hollow spheres. Additionally, all the hollow spheres calcined at 873 K were completely free of PS residues from the result of elemental analysis. These results indicate that homogeneous hollow spheres with negligible amount of PS residues could be prepared by adjusting the calcination temperature and holding time. Moreover, homogeneous hollow spheres with negligible amount of PS residues were even prepared by calcination at 723 K for 24 h. To determine the effects of the morphologies of the hollow spheres and composites on hydrolytic dehydrogenation of NH3BH3, the activities of the hollow spheres calcined at 873 K for 0, 3, 12 h and H-BEA zeolites were compared. Fig. 7(a), (b), (c), and (d) shows time course of hydrogen generation from aqueous NH3BH3 solution in the presence of the hollow spheres calcined at 873 K for 0 and 3 h, the composites calcined at 873 K for 12 h, and H-BEA zeolites. The evolution of 10, 4, 3.5, and 9.4 mL of hydrogen were finished in 12, 7, 3, and 38 min in the presence of the hollow spheres calcined for 0 and 3 h, the composites calcined for 12 h, and HBEA zeolites, respectively. The molar ratios of hydrolytically generated hydrogen to the initial NH3BH3 were 2.6, 1.0, 0.9, and 2.5, respectively. From these results, the activity increased as the holding time decreased. Additionally, specific surface area of the hollow spheres calcined for 0 and 3 h, and the composites calcined for 12 h were 393, 423, and 308 m2 g1, respectively. The results indicate that the activity does not depend on specific surface area of the hollow spheres and composites, and the activity is likely dependent on acid sites of the hollow spheres and composites. Moreover, amount of acid sites of the hollow spheres calcined for 0 and 3 h, and the
(a)
10
(d)
8 6 (b)
4
(c)
2 (e)
00
5
10
15
20 25 30 Reaction time [min]
35
40
Fig. 7 e Hydrogen generation from aqueous NH3BH3 solution (0.14 wt%, 3.5 mL) in the presence of hollow silicaealumina composite spheres calcined at 873 K for (a) 0, (b) 3, (c) 12 h, (d) H-BEA zeolites, and (e) recycled hollow silicaealumina composite spheres. Amount of the composites: 0.8 g.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 1 3 6 e1 7 1 4 3
negligible amount of PS residues almost show same estimated activity from the result of Fig. 5(a) and (b). On the other hand, our groups reported that hollow silicaezirconia composite spheres could successfully prepare by PS template methods. Hollow silicaealumina composite spheres were higher amount of hydrogen generation than hollow silicaezirconia composite spheres [34]. Additionally, Fig. 6(c) shows the TEM images of the recycled hollow spheres. From the result of Fig. 6(a) and (c), the morphologies of the recycled hollow spheres were not significantly different from each other. To determine the recycle ability of the hollow spheres, the activity of the recycled hollow spheres were compared. The evolution of 1.5 mL of hydrogen was finished in 2 min in the presence of the recycled hollow spheres as shown in Fig. 7(e). From the results, the recycled hollow spheres were much lower activity than the hollow spheres calcined for 0 h. The amount of acid sites of the recycled hollow spheres was 0 mmol g1. The results indicate that almost all the acid sites on the hollow spheres were consumed via the hydrolysis reaction. It has been previously reported that the activation process of NH3BH3 occurs because the acidic protons from the solid acids in the aqueous medium (which promote the dissociation of the BeN bond and the hydrolysis of BH3 species to produce borate ions species along with the release of hydrogen because the acidity of the solid acids is higher than that of boric acid) the results in the borate ions reacting with Hþ to produce boric acid [11]. Based on the results, the effective acid sites for hydrolytic dehydrogenation of NH3BH3 were likely Brønsted acid sites. These results indicate that the amount of Brønsted acid sites of the hollow spheres had a significant impact on their activity. The amount of Brønsted acid sites of the hollow spheres calcined at 873 K for 0 and 3 h, and the composites calcined at 873 K for 12 h was measured using NH3-TPD. Fig. 8 shows NH3TPD profile of the hollow spheres calcined at 873 K for 0 and 3 h, and the composites calcined at 873 K for 12 h. It has been reported that the NH3-TPD profile shows a peak around 400e600 K, indicating the presence of Brønsted acid sites, and a peak over 650 K, indicating the Lewis acid sites [35e37].
Based on the results, NH3-TPD profiles of the composites calcined at 873 K for 0, 3, and 12 h, first peak around at 400e523 K and the broad peak around at 523e600 K was attribute to the presence of Brønsted acid sites. On the other hands, the broad peak around 650e773 K and second peak around 773 K was attribute to the presence of Lewis acid sites. The amount of the Brønsted acid sites calculated from the peak area in the temperature range of about 400e600 K of the hollow spheres calcined for 0 and 3 h, and the composites calcined for 12 h were 0.12, 0.08, and 0.07 mmol g1, respectively. Fig. 9 shows the H2/NH3BH3 molar ratio of hydrogen generated from the aqueous NH3BH3 solution versus the amount of acid sites of the hollow silicaealumina composite spheres calcined at 873 K for 0 and 3 h, and the composites calcined at 873 K for 12 h, respectively. From the result, the activity increased as the amount of Brønsted acid sites in the hollow spheres and composites increased. These results indicated that the activity depends on the amount of the Brønsted acid sites of the hollow spheres and composites.
Conclusion In this study, we investigated the influence of the preparation conditions of hollow silicaealumina composite spheres on their activity for the hydrolytic dehydrogenation of NH3BH3. Hollow silicaealumina composite spheres were prepared by PS template method, and the PS templates particles were removed by calcination. The as-prepared hollow spheres were calcined at 523e873 K for 3 h. From the results of elemental analysis, PS residues in the hollow spheres were completely removed by calcination at 873 K. small particles around hollow spheres were observed from the images of TEM. To obtain homogeneous hollow spheres, as-prepared hollow spheres were calcined at 873 K for 0e12 h. From the results of TEM images, homogeneous hollow spheres were obtained via calcination at 873 K for 0 h. The activities of the hollow spheres calcined at 873 K for 0 and 3 h for hydrolytic
H2 / NH3BH3 mole ratio [ - ]
3.0
TCD signal [μV]
(a)
(b)
(c)
373
473
2.5
673
773
873
Temperature [ K ] Fig. 8 e NH3-TPD profile of hollow silicaealumina composite spheres calcined at 873 K for (a) 0 and (b) 3, and silicaealumina composite calcined at 873 K for (c) 12 h.
(a)
R2
2.0 1.5 1.0
(b) (c)
0.5 0
573
17141
0
0.05 0.10 0.15 Amount of Brønsted acid sites [mmol g-1]
0.20
Fig. 9 e The H2/NH3BH3 molar ratio of the hydrogen generated from the aqueous NH3BH3 solution (0.14 wt%, 3.5 mL) versus the amount of Brønsted acid sites of hollow silicaealumina composite spheres calcined at 873 K for (a) 0 and (b) 3 h, and silicaealumina composites calcined at 873 K for (c) 12 h.
17142
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 1 3 6 e1 7 1 4 3
dehydrogenation of aqueous NH3BH3 solution were compared. The activity of hollow spheres calcined for 0 h was 2.6 times higher that of hollow spheres calcined for 3 h. From the results of NH3-TPD, the activity increased as amount of Brønsted acid sites increased. The result indicated that the activity depends on amount of Brønsted acid sites.
[16]
[17]
Acknowledgment Special thanks are given to Dr. T. Yoneda for assistance with the use of NH3-TPD and to Ms. M. Takagi for helping with elemental analysis.
references
[18]
[19]
[20] [1] Sperling D, DeLuchi MA. Transportation energy futures. Annu Rev Energy 1989;14:375e424. [2] Schiapbach L, Zu¨ttel A. Hydrogen-storage materials for mobile applications. Nature 2001;414:353e8. [3] Umegaki T, Yan JM, Zhang XB, Shioyama H, Kuriyama N, Xu Q. Preparation and catalysis of poly(N-vinyl-2pyrrolidone) (PVP) stabilized nickel catalyst for hydrolytic dehydrogenation of ammonia borane. Int J Hydrogen Energy 2009;34:3816e22. [4] Hu¨gle T, Ku¨hnel MF, Lentz D. Hydrazine borane: a promising hydrogen storage material. J Am Chem Soc 2009;131:7444e6. [5] Jena P. Materials for hydrogen storage: past, present, and future. J Phys Chem Lett 2011;2:206e11. [6] Brockman A, Zheng Y, Gore J. A study of catalytic hydrolysis of concentrated ammonia borane solutions. Int J Hydrogen Energy 2010;35:7350e6. [7] Umegaki T, Xu Q, Kojima Y. Effect of L-arginine on the catalytic activity and stability of nickel nanoparticles for hydrolytic dehydrogenation of ammonia borane. J Power Sources 2012;216:363e7. [8] Chandra M, Xu Q. Room temperature hydrogen generation from aqueous ammonia-borane using noble metal nanoclusters. J Power Sources 2007;168:135e42. [9] Kalidindi SB, Sanyal U, Jagirdar BR. Nanostructured Cu and Cu@Cu2O core shell catalysts for hydrogen generation from ammonia-borane. Phys Chem Chem Phys 2010;10:5870e4. [10] Umegaki T, Yan JM, Zhang XB, Shioyama H, Kuriyama N, Xu Q. Hollow Ni-SiO2 nanosphere-catalyzed hydrolytic dehydrogenation of ammonia borane for chemical hydrogen storage. J Power Sources 2009;191:209e16. [11] Chandra M, Xu Q. Dissociation and hydrolysis of ammoniaborane with solid acids and carbon dioxide: an efficient hydrogen generation system. J Power Sources 2006;159:855e60. [12] Umegaki T, Ohashi T, Xu Q, Kojima Y. Influence of preparation conditions of hollow titania-nickel composite spheres on their catalytic activity for hydrolytic dehydrogenation of ammonia borane. Mater Res Bull 2014;52:117e21. € [13] Zahmakıran M, Durap F, Ozkar S. Zeolite confined copper(0) nanoclusters as cost-effective and reusable catalyst in hydrogen generation from the hydrolysis of ammoniaborane. Int J Hydrogen Energy 2010;35:187e97. € [14] Rakap M, Ozkar S. Hydrogen generation from the hydrolysis of ammonia-borane using intrazeolite cobalt(0) nanoclusters catalyst. Int J Hydrogen Energy 2010;35:3341e6. [15] Tong DG, Zeng XL, Chu W, Wang D, Wu P. Magnetically recyclable hollow Co-B nanospindles as catalysts for
[21]
[22] [23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
hydrogen generation from ammonia borane. J Mater Sci 2010;45:2862e7. Jiang HL, Umegaki T, Akita T, Zhang XB, Haruta M, Xu Q. Bimetallic Au-Ni nanoparticles embedded in SiO2 nanospheres: synergetic catalysis in hydrolytic dehydrogenation of ammonia borane. Chem Eur J 2010;16:3132e7. Zahmakıran M, Ayvalı T, Philippot K. In situ formed catalytically active ruthenium nanocatalyst in room temperature dehydrogenation/dehydrocoupling of ammonia-borane from Ru (cod) (cot) precatalyst. Langmuir 2012;28:4908e14. Basu S, Brockman A, Gagare P, Zheng Y, Ramachandran PV, Delgass WN, et al. Chemical kinetics of Ru-catalyzed ammonia borane hydrolysis. J Power Sources 2009;188:238e43. Cheng F, Ma H, Li Y, Chen J. Ni1xPtx (x ¼ 0e0.12) hollow spheres as catalysts for hydrogen generation from ammonia borane. lnorg Chem 2007;46:788e94. € Rakap M, Ozkar S. Zeolite confined palladium(0) nanoclusters as effective and reusable catalyst for hydrogen generation from the hydrolysis of ammonia-borane. Int J Hydrogen Energy 2010;35:1305e12. Zoldesi CI, Imho A. Synthesis of monodisperse colloidal spheres, capsules, and microballoons by emulsion templating. Adv Mater 2005;17(7):924e8. Schmidt HT, Ostafin AE. Liposome directed growth of calcium phospate nanoshells. Adv Mater 2002;14(7):532e5. Nakashima T, Kimizuka N. Interfacial synthesis of hollow TiO2 microspheres in ionic liquids. J Am Chem Soc 2003;125:6386e7. Putlitz BZ, Landfester K, Fischer H, Antonietti M. The generation of “armored latexes” and hollow inorganic shells made of clay sheets by templating cationic miniemulsions and latexes. Adv Mater 2001;13(7):500e3. Arnal PM, Weidenthaler C, Schu¨th F. Highly monodisperse zirconia-coated silica spheres and zirconia/silica hollow spheres with remarkable textural properties. Chem Mater 2006;18:2733e9. Titirici M, Antonietti M, Thomas A. A generalized synthesis of metal oxide hollow spheres using a hydrothermal approach. Chem Mater 2006;18:3808e12. Yang Z, Niu Z, Lu Y, Hu Z, Han C. Templated synthesis of inorganic hollow spheres with a tunable cavity size onto core-shell gel particles. Angew Chem Int Ed 2003;42:1943e5. Deng Z, Chen M, Zhou S, You B, Wu L. A novel method for the fabrication of monodisperse hollow silica spheres. Langmuir 2006;22:6403e7. Leng W, Chen M, Zhou S, Wu L. Capillary force induced formation of monodisperse polystyrene/silica organicinorganic hybrid hollow spheres. Langmuir 2010;26(17):14271e5. Cheng X, Chen M, Wu L, Gu G. Novel and facile method for the preparation of monodispersed titania hollow spheres. Langmuir 2006;22(8):3858e63. Chen M, Zhou S, You B, Wu L. A novel preparation method of raspberry-like PMMA/SiO2 hybrid. Macromolecules 2005;38:6411e7. Song X, Gao L. Fabrication of hollow hybrid microspheres coated with silica/titania via sol-gel process and enhanced photocatalytic activities. J Phys Chem C 2007;111:8180e7. Yokoi T, Sakamoto Y, Terasaki O, Kubota Y, Okubo T, Tatsumi T. Periodic arrangement of silica nanospheres assisted by amino acids. J Am Chem Soc 2006;128:13664e5. Umegaki T, Hosoya T, Toyama N, Xu Q, Kojima Y. Fabrication of hollow silica-zirconia composite spheres and their activity
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 7 1 3 6 e1 7 1 4 3
for hydrolytic dehydrogenation of ammonia borane. J Alloy Compd 2014;608:261e5. [35] Sakthivel A, Dapurkar SE, Gupta NM, Kulshreshtha SK, Selvam P. The influence of aluminium sources on the acidic behaviour as well as on the catalytic activity of mesoporous H-AlMCM-41 molecular sieves. Microporous Mesoporous Mater 2003;65:177e87.
17143
[36] Wang Y, Lang N, Tuel A. Nature and acidity of aluminum species in AlMCM-41 with a high aluminum content (Si/ Al ¼ 1.25). Microporous Mesoporous Mater 2006;93:46e54. [37] Fang X, Liu Z, Hsieh M-F, Chen M, Liu P, Chen C, et al. Hollow mesoporous aluminosilica spheres with perpendicular pore channels as catalytic nanoreactors. ACS Nano 2012;6(5):4434e44.