Transition metal Ti coated porous fullerene C24B24: Potential material for hydrogen storage

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 x x x ( 2 0 1 5 ) 1 e7

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.elsevier.com/locate/he

Transition metal Ti coated porous fullerene C24B24: Potential material for hydrogen storage Chunmei Tang*, Shengwei Chen, Weihua Zhu, Jing Kang, Xiang He, Zhenjun Zhang College of Science, Hohai University, Nanjing, Jiangsu 210098, China

article info

abstract

Article history:

The hydrogen storage capacity of transition metal Ti atoms decorated porous fullerene

Received 17 April 2015

C24B24 is investigated by the pseudopotential density functional method. The C24B24 cage

Received in revised form

contains six B4 rings with the average diameter of 3.88 Å. The Ti atoms are strongly bound

21 May 2015

to six B4 rings. Each Ti atom can adsorb up to six H2 molecules. The calculated average

Accepted 25 May 2015

adsorption energies per H2 for (Ti-nH2)6C24B24(n ¼ 1e6) are in the energy range from 0.24 to

Available online xxx

0.55 eV, which is suitable for hydrogen storage at near-ambient conditions. The Dewar eKubas interaction dominates the adsorption of H2 on the outer surface of Ti6C24B24. The

Keywords:

largest hydrogen gravimetric density of (Tie6H2)6C24B24 is 8.1 wt%, exceeding the 5.5 wt%

Fullerene

by the year 2017 specified by the US department of energy (DOE). Therefore, the stable

C24B24

Ti6C24B24 can be applied as one candidate for hydrogen storage materials at near-ambient

Hydrogen storage

conditions.

Density functional theory

Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.

Introduction Because of the abundance, renewability, high efficiency, and environmentally friendly nature, hydrogen has been attracting much interest as an ideal kind of energy and has the potential to reduce our dependence on fossil fuels, helping to resolve the global warming issue [1]. However, it is a great challenge to find a material suitable for hydrogen storage. Recently, the U.S. Department of Energy (DOE) updated the hydrogen storage targets to a hydrogen gravimetric density of 5.5 wt% by the year 2017 [2]. The adsorption of H2 through chemisorption is held by strong covalent bonds, and their dissociation requires high temperature. On the other hand, the adsorption of hydrogen molecules on materials via physisorption [3] is so weak that storage at near-ambient

conditions is not feasible. In general, materials that bind hydrogen molecularly with average adsorption energy per H2 (Ead) intermediate between physisorbed and chemisorbed states (0.2e0.6 eV) are good candidates to tailor the above requirements [4e6]. During the past decade, the hydrogen storage abilities of Cbased nanomaterials like fullerene, carbon nanotubes, and graphene have been extensively studied. Due to the simple van der walls dominates the binding of hydrogen molecules to the surface of pure C-based nanomaterials, the surfaces cannot efficiently store hydrogen. However, their metal decorated counterparts such as C48B12M12 (M ¼ Fe, Co, Ni) [7], M32B80 (M ¼ Ca and Sc) [8], Rh coated carbon nanotubes [9], metal decorated graphene [10], Li9C60 [11] have exhibited remarkable hydrogen adsorption capacity [12]. Unfortunately,

* Corresponding author. Tel.: þ86 025 83786640. E-mail address: [email protected] (C. Tang). http://dx.doi.org/10.1016/j.ijhydene.2015.05.159 0360-3199/Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Tang C, et al., Transition metal Ti coated porous fullerene C24B24: Potential material for hydrogen storage, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.05.159

2

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 x x x ( 2 0 1 5 ) 1 e7

it is difficult to realize these metal decorated materials experimentally, since metal atoms tend to form clusters on the surfaces of nanostructures, and consequently the hydrogen storage capacity drops dramatically. Therefore, the binding energy of a metal atom to the substrate should be larger than the experimental cohesive energy of the bulk metal [13]. It was suggested that alkali metals (AM) and alkaline-earth metals (AEM) can produce a homogeneous coating due to their lightweight and low cohesive energies [8,12]. The hydrogen adsorption on these structures has been demonstrated experimentally, such as Li12C60 [13e15]. Experimentally, substitution the carbon atoms by the boron and nitrogen atoms in the C-based nanomaterials, such as fullerenes, carbon nanotubes, and graphenes, have already been realized [16]. Theoretically, the boron and nitrogen are the neighbors to carbon in the periodic table. It should be noticed that the N atom has one more valence electron than C, thus, the N doped nanomaterials are electron-rich complex and behave like donors, when the surfaces are decorated by the metal atoms, the electrons will transfer from the namomaterials to the metal atoms, if these structures adsorb hydrogen, the metal atoms do not have enough hollow orbitals supplied to the electrons from the H2 molecule, this obviously hander the hydrogen storage ability. However, the boron atom has one less valence electron than C, so the B doped nanomaterials are electronedeficit, when the metal atoms decorate the surfaces, they will lose some electrons, therefore, the hydrogen molecules are strongly polarized by the charged metal positive ion, which is helpful for the adsorption of hydrogen [17]. This has been verified by the calculations on the hydrogen storage abilities of many metal decorated B-based nanostructures. For example, Sc12C48B12 [18] has the better hydrogen storage capacity than that of Sc12C48N12 [19]. Li12C48B12 [20] exhibits more remarkable hydrogen storage ability than that of Li12C48N12 [21]. Lu et al. [22] have found that the porous graphene by N doping obviously weakened the interaction energy between metal atoms and the porous graphene. On the other hand, the B atom is lighter than the N and C atoms, so the B doped hydrogen storage complex should have higher hydrogen gravimetric density than the N atom doped hydrogen storage structures. Therefore, the metal atoms decorated B-based nanomaterials should show the better hydrogen storage capacity and higher hydrogen gravimetric densities than the corresponding Nbased nanostructures. Recently, Guo et al. [23] have reported that the Ti atoms decorated untraditional fullerene cage Ti6C48 could adsorb up to 36 hydrogen molecules. Later, Srinivasu et al. [24] have found that six transition metal atoms (Sc, Ti, V) decorated porous M6C24N24 cages, particularly Ti6C24N24, show remarkable hydrogen storage capacity at near-ambient conditions. More importantly, the transition metal atoms are bound to six N4 rings in the C24N24 cage can effectively avoid the clustering problem. Inspired by this study, we choose the comparable more suitable transition metal atom Ti to decorate the B atom doped porous fullerene C24B24, and then use the density functional theory to study its hydrogen storage capacity. It is calculated that each B4 ring in C24B24 has the average diameter of 3.88 Å and can be expected to effectively bind transition metal atoms, which can avoid the clustering problem. Each Ti

atom can adsorb up to six hydrogen molecules with the Ead ranging from 0.24 to 0.55 eV, which are between physisorbed and chemisorbed states (0.2e0.6 eV). The hydrogen gravimetric density of the (6H2eTi)6C24B24 cage can reach 8.1 wt%, exceeding the 5.5 wt% by the year 2017 specified by US Department of Energy (DOE) [2]. Therefore, the Ti6C24B24 cage can be applied as one candidate for hydrogen storage materials at near-ambient conditions.

Computational details The calculations are performed with the generalized gradient approximation (GGA) [25] based on density functional theory (DFT) [26] as implied in the DMol3 package [27]. The PerdeweBurkeeErnzerhof (PBE) functional [28] and the double numerical basis sets including d-polarization functionals (DNP) are used. For all the hydrogen adsorbed structures, the Grimme method [29] for PBE/GGA is used for dispersion corrections. In addition, calculations are also conducted with another two exchange-correlation functionals, Wang and Perdew functional (PW91) [30] based on GGA and PerdeweWang functional (PWC) [31] based on LDA, with the OBS method [32] considered for the dispersion interactions. The electronic structure is obtained by solving the spin-polarized Kohn-Sham [33] equations self-consistently. The DFT semicore pseudopotential (DSPP) core-treat method [34] for the Ti atom and all-electron method for the C and B atoms are adopted. Self-consistent field procedures are done with a convergence criterion of 105 Hartree on the electron densities. The convergence tolerances are 2.0  105 Hartree, 0.004 ˚ respectively for the energy, the max Hartree, and 0.005 A force, and the max displacement in the optimization. Finally, to obtain the electronic configurations of all the considered structures, the natural bonding orbital (NBO) electrons of all the structures are computed. The ground-state structures are determined by their minimal energies, which are further verified by no imaginary frequency in their harmonic frequency calculations. The accuracy of our computational method is tested by computing the H2 and the porous fullerene C24N24. The calculated bond length and binding energy (Eb) of H2 are 0.75 Å and 4.55 eV, in good agreement with the experimental values of 0.74 Å and 4.53 eV respectively [35]. For C24N24, the calcu˚ and 1.34 A ˚ , in lated bond lengths of CeC and CeN are 1.55 A good agreement with the results gotten by Srinivasu et al. [24] ˚ and 1.34 A ˚ ) separately. Therefore, our computational (1.56 A scheme is suitable to study the structural and electronic properties of nanostructures.

Results and discussions Fig. 1 schematically shows how to get the B-based fullerene cage C24B24 from C60. As shown in Fig. 1, the generation course of B4 rings are in the following: The two carbons indicated by the red ellipse in C60 and connecting two neighboring fivemember rings are removed firstly, then, the four carbon atoms marked by the blue circle in C48 and connecting these two carbons are replaced by four boron atoms. It is known

Please cite this article in press as: Tang C, et al., Transition metal Ti coated porous fullerene C24B24: Potential material for hydrogen storage, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.05.159

3

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 x x x ( 2 0 1 5 ) 1 e7

Fig. 1 e The construct course of C24B24 from the C60.

from Fig. 1 that total 12 carbon atoms are removed and 24 carbons in six rings are replaced by boron atoms in the C60 cage, generating the C24B24 cage. The computed bond lengths of CeC and CeB of C24B24 are respectively 1.42 Å and 1.55 Å, which are in good agreement with those of C48B12 calculated using the same method [36]. The average binding energy per atom (Eb/atom) of C24B24, which is calculated using the formule: Eb/atom ¼ (EC24B24-24EC-24EB)/48. The calculated Eb/ atom of C24B24 is 6.28 eV, which is larger than that of C24N24 (5.78 eV) calculated at the same theory level, therefore, C24B24 has the better thermodynamic stability than that of C24N24. There are six B4 rings in the C24B24 cage. Due to the remarkable hydrogen storage ability of the Ti atoms coated C24N24 cage, we use the transition metal atom Ti to decorate the C24B24 cage. Six Ti atoms are located on the hollow site of six B4 rings, similar to that in the phthalocyanine-TM complexes. [37] As the earlier experimental and theoretical reports have pointed out that the metal-decorated nanomaterials have many advantages and should meet some important requirements, that is, the large distance between metal atoms and the hydrogen storage in the molecular form with a feasible average adsorption energy. Fig. 2 (a) shows the optimized geometric structure of Ti6C24B24. We calculated that the ˚, distance between the neighboring two Ti atoms is 5.23 A which is considerably larger than that of the Ti2 dimers [38], indicating that dimerization is not a highly favorable process for the Ti atom. Moreover, the TieTi bond length of Ti6C24B24 is also longer than that of Ti6C24N24 [24], therefore, the Ti atoms decorated C24B24 is expected to exhibit better hydrogen storage capacity than that of C24N24 coated by the Ti atoms. The Eb of each Ti atom adsorbed on the surface of the C24B24 cage is defined as Eb ¼ ðEC24 B24 þ 6  ETi  ETi6 C24 B24 Þ=6, the positive Eb indicates the exothermic reactivity [39]. Obviously, the calculated average Eb of the Ti atoms to the C24B24 surface is high to 8.58 eV, nearly double of the experimental cohesive energy of bulk Ti (4.85 eV/atom [40]), therefore, the problem of metal aggregative to form cluster is expected to be overcome, and the material will be stable for designing a recyclable hydrogen material. More importantly, the calculated Eb of the Ti atom to the C24B24 cage (8.58 eV) is much larger than that of the Ti atom to the surface of C24N24 (6.89 eV), therefore, C24B24 is more suitable than C24N24 for hydrogen adsorption after the decoration of the Ti atom. To further confirm the thermodynamic stability of Ti6C24B24, we carry out molecular dynamic  algorithm method at the room simulations using the Nose

temperature (T ¼ 300 K) and the high temperature (T ¼ 1000 K) with the 1.0 fs time step. After 1 ps simulation, we find that the Ti atoms are still bound to the B4 rings. Therefore, the Ti6C24B24 structure is rather thermodynamically stable. Fig. 2(b)e(h) shows the optimized structures of (TinH2)6C24B24(n ¼ 1e6). The Ead of (Ti-nH2)6C24B24(n ¼ 1e6) structures are defined as follows. Ead ¼ ETi6 C24 B24 þ 6nEH2  EðTinH2 Þ6 C24 B24


 6n

(1)

Where,ETi6 C24 B24 ; EH2 ， EðTinH2 Þ6 C24 B24 ， EðTi（n1）H2 Þ6 C24 B24 is the total energy of Ti6C24B24, H2, (Ti-nH2)6C24B24，and (Ti-(n-1) H2)6C24B24 respectively. It can be seen from Table 1 that the Ead of (Ti-nH2)6C24B24(n ¼ 1e6) basing on the PBE/GGA method are from 0.22 to 0.55 eV, which are between physisorbed and chemisorbed states, so Ti6C24B24 should be a good candidate for reversible hydrogen storage at near-ambient conditions. However, some results calculated by the PW91/GGA method and most results calculated by the PWC/LDA method exceed 0.6 eV, here, the OBS method is considered for the dispersion interaction. It is widely known that LDA usually overestimates the Ead [41e44], moreover, too strong binding will result in the dissociation of the hydrogen molecules in (TinH2)6C24B24(n ¼ 1e6). Therefore, the PBE/GGA with the Grimme method considered for dispersion corrections can generate relatively good adsorption energies. Thus, it has been widely used for some other hydrogen storage systems [41e44]. The adsorption energy of each H2 molecule onto the Ti6C24B24 can be defined by the consecutive adsorption energy (Er) values defined as follows: Er ¼ E½Tiðn1ÞH2 6 C24 B24 þ 6EH2  EðTinH2 Þ6 C24 B24


 6

(2)

The consecutive adsorption energy is an important index for testing the adsorption capacity of hydrogen storage materials. The positive Er means spontaneous adsorption can occur between the hydrogen molecule and the Ti6C24B24 structure. If the Er is small or negative, the adsorption of H2 is difficult. It can be found from Table 1 that the Er for the seventh H2 adsorbed on each Ti atom is only 0.01 eV. Therefore, we can conclude that each Ti atom can adsorb up to 6 hydrogen molecules. Moreover, the computed hydrogen gravimetric density of (Tie6H2)6C24B24 is 8.1 wt%, exceeding the 5.5 wt% by the year 2017 specified by the US DOE [2]. It can be known from Table 1 that the energy gaps (Egs) between the highest occupies molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) of (TinH2)6C24B24(n ¼ 2e6) are all around 0.70 eV, much larger than

Please cite this article in press as: Tang C, et al., Transition metal Ti coated porous fullerene C24B24: Potential material for hydrogen storage, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.05.159

4

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 x x x ( 2 0 1 5 ) 1 e7

Fig. 2 e (a) Optimized geometric structure of Ti6C24B24; (b) ~ (h) Optimized geometric structures of (nH2-Ti)6C24B24(n ¼ 1e6). those of Ti6C24B24 (0.21 eV) and (TieH2)6C24B24 (0.25 eV), indicating that the adsorption of H2 molecules can enhance the kinetic stability of the structure. Focusing on the hydrogen storage materials with metal atoms acting as adsorption centers, we consider the 18electron rule [44] that the number of H atoms bound to each metal atom is nH ¼ 18  nev  nes  nem , where the nev is the number of valence electrons in the free metal atom, it should be a constant once the metal is chosen, When the metal atoms are bound to the substrate in a stable configuration, nes is the bond numbers between the metal atom and the substrate, it also should be a constant. If the metal atoms are far away from one another, the nem can be neglected. If the metal atoms are close to one another, the neighboring metal atoms will interact with each other, the nem value should be the number of

metalemetal bond. The 18-electron rule has been used to predict the maximum adsorbed H2 molecules in the hydrogen storage materials, such as metal decorated graphene [45] and so on. As to Ti6C24B24, the core valence electrons of the Ti atom are frozen in our calculation, therefore, the nev for the Ti atom should be 2 when only considering the 3d valence electrons. In addition, the nes and nem for the Ti atom should be 4 and 0 respectively. Thus, the maximum H2 molecules adsorbed by each Ti atom of the Ti6C24B24 structure should be 6, in accordance with our calculation conclusion. The calculated TieH ˚ , which agrees well with bond length of (TieH2)6C24B24 is 1.99 A ˚ calculated by the PW91 method [40]. All the H2 the 1.90 A molecules introduced to Ti6C24B24 are in the form of dihydrogen molecules. The bond length of HeH is elongated to ˚ , slightly longer than that of the free H2 molecule about 0.83 A

Table 1 e Eg, RC-C, RB-C, RTi-B of Ti6C24B24, together with the Ead for H2, Er, Eg, RC-C, RB-C, and RTi-B of (Ti-nH2)6C24B24(n ¼ 1e7). Ead (eV)

Ti6C24B24 (H2eTi)6C24B24 (2H2eTi)6C24B24 (3H2eTi)6C24B24 (4H2eTi)6C24B24 (5H2eTi)6C24B24 (6H2eTi)6C24B24 (7H2eTi)6C24B24

Er (eV)

PBE/GGA

PW91/GGA

PWC/LDA

PBE/GGA

PW91/GGA

PWC/LDA

e 0.55 0.42 0.33 0.28 0.25 0.24 0.22

e 0.68 0.59 0.49 0.46 0.46 0.44 0.43

e 1.07 0.93 0.77 0.71 0.71 0.67 0.68

e 0.55 0.30 0.14 0.15 0.12 0.19 0.01

e 0.68 0.49 0.31 0.38 0.43 0.33 0.09

e 1.07 0.80 0.45 0.54 0.69 0.48 0.06

Eg (eV)

˚) RC-C(A

0.21 0.25 0.70 0.69 0.70 0.70 0.70 0.62

1.50 1.50 1.47 1.47 1.47 1.47 1.47 1.47

˚) RB-C(A

˚) RTi-B(A

PBE/GGA 1.55 1.55 1.55 1.55 1.55 1.55 1.55 1.55

2.17 2.15 2.18 2.18 2.18 2.18 2.18 2.18

International Journal of Hydrogen Energy, Tang et al.

Please cite this article in press as: Tang C, et al., Transition metal Ti coated porous fullerene C24B24: Potential material for hydrogen storage, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.05.159

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 x x x ( 2 0 1 5 ) 1 e7

5

Fig. 3 e The difference charge densities of (a) Ti6C24B24 and (b) (TieH2)6C24B24. The charge density is given in the unit of ˚ 3. a.u./A

˚ ), similar to the experimental value (0.74 A ˚ ) [35], and (0.75 A that in many hydrogen storage structures, such as Ti doped zigzag graphene nanoribbons [46], TiC58$nH2 [23] and so on. Interestingly, even though the distance between the two H ˚ for the initial configuatoms in our calculations is set to 3 A ration, the two H atoms approach each other and finally form an H2 molecule. We calculate that the two hydrogen atoms in one H2 molecule have the different positive NBO charge. Thus, repulsive interaction appears between two H atoms in the H2 molecule, causing the HeH bond length elongate to about ˚ . However, the two H atoms of the top sixth H2 molecule 0.83 A bound to each Ti have almost the same charge and the HeH

˚ , longer than others. For the (Tie6H2)6C24B24 distance is 0.90 A structure, we can notice that for each Ti atom, the distance ˚ , much shorter than that between the sixth H2 and Ti is 1.79 A between the Ti atom and other five H2 molecules on the side ˚ ), so the adsorption for the top sixth H2 should be (1.90 A stronger than that for the side five H2 molecules, this also increase the repulsive interaction between the two H atoms in the sixth H2 molecule. We present in Fig. 3 the difference charge densities of Ti6C24B24 and (TieH2)6C24B24, which equal to the total densities subtracting the densities of the isolated atoms. The different value for the charge densities is shown in the color from red to

Fig. 4 e The local densities of states (LDOS) analysis of free H, Ti and B in Ti6C24B24 and those of Ti and H in (Ti-nH2)6C24B24 (n ¼ 1 and 6). Please cite this article in press as: Tang C, et al., Transition metal Ti coated porous fullerene C24B24: Potential material for hydrogen storage, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.05.159

6

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 x x x ( 2 0 1 5 ) 1 e7

blue (indicated in Fig. 3). The blue color densities present the positive charge densities. It is shown in the figure that there are much positive charge densities distribute on the Ti atoms and the H2 molecules, moreover, the calculated average NBO charge of the Ti atoms and the H atoms are respectively 0.52e and 0.07e respectively, therefore, the electron should transfer from the Ti atom to the C24B24 cage, and the hydrogen molecules are strongly polarized by the charged Ti positive ion. We then consider the interaction mechanism between hydrogen molecules and the Ti6C24B24 cage. The interaction mechanism between Ti6C24B24 and H2 can be ascribed to the well known Dewar coordination and Kubas interaction [47]. The Ti atoms are chemically bound to different nanostructures through hybridization between the LUMO of the nanostructures with the d orbitals of the metal atom (Dewar coordination) [47]. The Ti6C24B24 complex binds many H2 through hybridization between the s orbitals of H2 and the d orbitals of the Ti atom (Kubas interaction). Fig. 4 shows the local densities of states (LDOS) of free H, Ti and B atoms in Ti6C24B24 and those of the Ti atoms and the H atoms in (Ti-nH2)6C24B24 (n ¼ 1, 6). The LDOS are obtained by Lorentzian extension of the discrete energy levels, with weights given by the orbital populations of the levels, and a summation over them. The Fermi energy (Ef) is at 0 eV, shown by the black dotted line in the figure. We can see from Fig. 4 (a) and (d) that the overlap of the orbitals of the Ti atoms and the orbitals of the B atoms in the Ti6C24B24 structure is consistent with the above NBO population analysis. For the (Tie6H2)6C24B24 complex, we can see the overlap of the s bonding orbitals of hydrogen with the d orbitals of the Ti atom (-9 ~ 7 eV) and also the s* antibonding orbitals of hydrogen with the d orbitals of the Ti atom near the Ef. Therefore, we can argue that the DewareKubas interaction makes Ti6C24B24 adsorb H2 molecules. In particular, do these materials have thermal stabilities? To answer this question, we carry out the first-principle molecular dynamics simulation with 2 fs time step at finite temperatures. First, we study the thermal stability of (Tie6H2)6C24B24 at T ¼ 77 K. After 4 ps simulation, the structure still remains intact and no H2 molecule escapes out. We then study their thermal stabilities at the room temperature (T ¼ 300 K). After 4 ps simulation, seven H2 molecules escape from (Tie6H2)6C24B24. It can be expected that more H2 molecules will be released at higher temperature. This indicates that the C24B24 structure decorated by the Ti atoms should be appropriate for hydrogen storage at near-ambient conditions, such as the experimentally isolated nH2-Li12C60 [8].

Conclusions The hydrogen adsorption capacity of transition metal Ti atoms decorated porous fullerene C24B24 is investigated using the density functional method. This porous fullerene is generated through truncated doping of 24 carbon atoms by 24 boron atoms, and contains six B4 rings. These Ti atoms are strongly bound to six B4 rings. The transition metal Ti atoms are strongly bound to these B4 rings, and the average binding energy of each Ti on the C24B24 surface is larger than the experimental cohesive energy of bulk Ti. Each Ti atom can adsorb up to six H2 molecules. The calculated average adsorption energies of hydrogen molecules bound to the

Ti6C24B24 surface are in the range of 0.24e0.55 eV, which is suitable for hydrogen storage under near-ambient conditions. The DewareKubas interaction dominates the adsorption of H2 on the surface of Ti6C24B24. The maximum hydrogen gravimetric density of (Tie6H2)6C24B24 is 8.1 wt%, exceeding the 5.5 wt% by the year 2017 specified by US DOE. Therefore, the stable Ti6C24B24 can be applied as one candidate for hydrogen storage materials at near-ambient conditions.

Acknowledgments This work is financially sponsored by the National Natural Science Foundation of China (Grant Nos. 11104062, 10947132) and Qing Lan Project of Jiangsu Province.

references

[1] Cortright RD, Davda RR, Dumesic JA. Hydrogen from catalytic reforming of biomass-derived hydrocarbons in liquid water. Nature 2002;418:964e7. [2] DOE plan: http://wwwl.eere.energy.gov/ hydrogenandfuelcells/storage/. [3] Dillon AC, Jones KM, Bekkedahl TA, Kiang CH, Bethune DS, Heben MJ. Storage of hydrogen in single-walled carbon nanotubes. Nature(London) 1997;386:377e9. [4] Chen P, Wu X, Tan KL. High H2 uptake by alkali-doped carbon nanotubes under ambient pressure and moderate temperatures. Science 1999;285:91e3. [5] Rosi NL, Eckert J, Eddaoudi M, Vodak DT, Kim J, Keeffe MO, et al. Hydrogen storage in microporous metal-organic frameworks. Science 2003;300:1127e9. [6] Han SS, Goddard WA. Lithium-doped metal-organic frameworks for reversible h2 storage at ambient temperature. J Am Chem Soc 2012;129:8422e3. [7] Tang CM, Chen SW, Zhu WH, Zhang AM, Zhang KX, Zou H. Doping the transition metal atom Fe, Co, Ni into C48B12 fullerene for enhancing H2 capture: a theoretical study. Int J Hydrogen Energy 2014;3(9):12741e8. [8] Wu GF, Wang JL, Zhang XY, Zhu LY. Hydrogen storage on metal-coated B80 buckyballs with density functional theory. J Phys Chem C 2009;113:7052e7. [9] Yao QL, Lu ZH, Jia YS, Chen XS, Liu X. In situ facile synthesis of Rh nanoparticles supported on carbon nanotubes as highly active catalysts for H2 generation from NH3BH3 hydrolysis. Int J Hydrogen Energy 2015;40:2207e15. [10] Guo YH, Lan XX, Cao JX, Xu B, Xia YD, Yin J, et al. A comparative study of the reversible hydrogen storage behavior in several metal decorated graphyne. Int J Hydrogen Energy 2013;38:3987. [11] Yoshida A, Okuyama T, Terada T, Naito S. Reversible hydrogen storage/release phenomena on lithium fulleride (LinC60) and their mechanistic investigation by solid-state NMR spectroscopy. J Mater Chem 2011;21:9480e2. [12] Chen M, Zhao YJ, Liao JH, Yang XB. Transition-metal dispersion on carbon-doped boron nitride nanostructures: applications for high-capacity hydrogen storage. Phys Rev B 2012;86:045459. [13] Sun Q, Wang Q, Jena P, Kawazoe Y. Clustering of Ti on a C60 surface and its effect on hydrogen storage. J Am Chem Soc 2005;127:14582e3. [14] Mauron P, Gaboardi M, Remhof A, Bliersbach A, Sheptyakov D, Aramini M, et al. Clustering of Ti on a C60

Please cite this article in press as: Tang C, et al., Transition metal Ti coated porous fullerene C24B24: Potential material for hydrogen storage, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.05.159

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 x x x ( 2 0 1 5 ) 1 e7

[15]

[16]

[17]

[18]

[19]

[20] [21] [22] [23]

[24]

[25] [26] [27] [28] [29]

[30] [31]

[32]

surface and its effect on hydrogen storage. J Phys Chem C 2013;117:22598e602. Paolone A, Vico F, Teocoli F, Sanna S, Palumbo O, Cantelli R, et al. Hydrogen sorption in Li12C60. J Phys Chem C 2012;116:16365e70. Panchakarla LS, Subrahmanyam KS, Saha SK, Govindaraj A, Krishnamurthy HR, Waghmare UV, et al. Adv Mater 2009;21:4726e30. Karfunkel HR, Dressler T, Hirsch A. Heterofullerenes: structure and property predictions, possible uses and synthesis proposals. J Comput Aided Mol Des 1992;6:521e35. Zhao YF, Kim YH, Dillon AC, Heben MJ, Zhang SB. Hydrogen storage in novel organometallic buckyballs. Phys Rev Lett 2005;94:155504. Zhou Z, Gao X, Yan J, Song DY. Doping effects of B and N on hydrogen adsorption in single-walled carbon nanotubes through density functional calculations. Carbon 2006;44:939e47. Sun Q, Wang Q, Jena P. Functionalized heterofullerenes for hydrogen storage. Appl Phys Lett 2009;94:013111. Brena B, Luo Y. Electronic structures of azafullerene C48N12. 7139 J Chem Phys 2003;119. Lu Ruifeng, Rao Dewei, Lu Zelin, Qian Jinchao, Li Feng, Wu Haiping, et al. J Phys Chem C 2012;116:21291e6. Guo J, Liu ZG, Liu SQ, Zhao XH, Huang KL. High-capacity hydrogen storage medium: Ti doped fullerene. Appl Phys Lett 2011;98:023107. Srinivasu K, Ghosh SK. Transition metal decorated porphyrin-like porous fullerene: promising materials for molecular hydrogen adsorption. J Phys Chem C 2012;116:25184e9. Becke AD. Phys Rev A 1988;38:3098. Perdew JP, Chevry JA, Vosko SH, Jackson KA, Pederson MR, Sing DJ, et al. Phys Rev B 1992;46:6671. Delley B. DMOL is a density functional theory program distributed by Accelrys, Inc. J Chem Phys 1990;92:508. Perdew JP, Buke K, Ernzerhof M. Phys Rev Lett 1996;77:3865. Grimme S. Semiempirical GGA-type density functional constructed with a long-range dispersion correction. J Comput Chem 2006;27:1787. Wang Y, Perdew JP. Spinscalingoftheelectron-gascorrelationenergyinthehigh-densitylimit. Phys Rev B 1991;43:8911. Perdew JP, Wang Y. Accurate and simple analytic representation of the electron-gas correlation energy. Phys Rev B 1992;45:13244. Ortmann F, Bechstedt F, Schmidt WG. Semiempirical van der Waals correction to the density functional description

[33] [34] [35] [36]

[37]

[38]

[39]

[40] [41]

[42]

[43]

[44]

[45]

[46]

[47]

7

of solids and molecular structures. Phys Rev B 2006;73:205101. Kohn W, Sham LJ. Self-consistent equations including exchange and correlation effects. Phys Rev 1965;140:A1133. Delley B. Hardness conserving semilocal pseudopotentials. Phys Rev B 2002;66:155125. Lide DR. CRC handbook of chemistry and physics (CRC). Cleveland. 2000. Xie RH, Jensen L, Bryant GW, Zhao JJ, Smith VH. Structural, electronic, and magnetic properties of heterofullerene C48B12. Chem Phys Lett 2003;375:445e51. Zhou J, Sun Q. Magnetism of phthalocyanine-based organometallic single porous sheet. J Am Chem Soc 2011;133:15113e9. Barden CJ, Rienstra-Kiracofe JC, Schaefer HF. Homonuclear 3d transition-metal diatomics: a systematic density functional theory study. J Chem Phys 2000;113:690e700. Lu LH, Sun KC, Chen C. Theoretical study of fullerene derivatives: C28H4 and C28X4 cluster molecules. Int J Quantum Chem 1998;67:187e97. Kittel C. Introduction to solid state physics. 8th ed. New York: John Wiely &Sons, INC.; 2005. Kim YH, Zhao YF, Williamson A, Heben MJ, Zhang SB. Nondissociative adsorption of H2 molecules in lightelement-doped fullerenes. Phys Rev Lett 2006;96:016102. Gao Y, Zeng XC. Ab initio study of hydrogen adsorption on benzenoid linkers in metaleorganic framework materials. J Phys Condens Matter 2007;19:386220. Krasnov PO, Ding F, Singh AK, Yakobson BI. Clustering of Sc on SWNT and reduction of hydrogen uptake: Ab-initio allelectron calculations. J Phys Chem C 2007;111:17977e80. Zhao Yufeng, Lusk Mark T, Dillon Anne C, Heben Michael J, Zhang Shengbai B. Boron-based organometallic nanostructures: hydrogen storage properties and structure stability. Nano Lett 2008;8:157e61. Fair KM, Cui XY, Li L, Shieh CC, Zheng RK, Liu ZW, et al. Hydrogen adsorption capacity of adatoms on double carbon vacancies of graphene: a trend study from first principles. Phys Rev B 2013;87:014102. Lebon A, Carrete J, Longo RC, Vega A, Gallego LJ. Molecular hydrogen uptake by zigzag graphene nanoribbons doped with early 3d transition-metal atoms. Int J Hydrogen Energy 2013;38:8872e80. Kubas GJ. Fundamentals of H2 binding and reactivity on transition metals underlying hydrogenase function and H2 production and storage. Chem Rev 2007;107(:4152.

Please cite this article in press as: Tang C, et al., Transition metal Ti coated porous fullerene C24B24: Potential material for hydrogen storage, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.05.159