Effects of REF3 (RE = Y, La, Ce) additives on dehydrogenation properties of LiAlH4

Effects of REF3 (RE = Y, La, Ce) additives on dehydrogenation properties of LiAlH4

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 1 6 4 2 e1 1 6 5 0 Available online at www.sciencedirect.co...

1MB Sizes 3 Downloads 21 Views

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 1 6 4 2 e1 1 6 5 0

Available online at www.sciencedirect.com

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

Effects of REF3 (RE ¼ Y, La, Ce) additives on dehydrogenation properties of LiAlH4 Si Zhou a, Jianxin Zou a,b,*, Xiaoqin Zeng a,b, Wenjiang Ding a,b a

National Engineering Research Center of Light Alloy Net Forming & State Key Laboratory of Metal Matrix Composite, Shanghai Jiao Tong University, Shanghai 200240, China b Shanghai Engineering Research Center of Magnesium Materials and Application & School of Materials Science and Engineering, Shanghai Jiao Tong University, Shanghai 200240, China

article info

abstract

Article history:

The catalytic effects of rare earth fluoride REF3 (RE ¼ Y, La, Ce) additives on the dehydro-

Received 28 March 2014

genation properties of LiAlH4 were carefully investigated in the present work. The results

Received in revised form

showed that the dehydrogenation behaviors of LiAlH4 were significantly altered by the

9 May 2014

addition of 5 mol% REF3 through ball milling. The destabilization ability of these catalysts

Accepted 16 May 2014

on LiAlH4 has the order: CeF3>LaF3>YF3. For instance, the temperature programmed

Available online 14 June 2014

desorption (TPD) analyses showed that the onset dehydrogenation temperature of CeF3 doped LiAlH4 was sharply reduced by 90  C compared to that of pristine LiAlH4. Based on

Keywords:

differential scanning calorimetry (DSC) analyses, the dehydriding activation energies of the

Hydrogen storage materials

CeF3 doped LiAlH4 sample were 40.9 kJ/mol H2 and 77.2 kJ/mol H2 for the first and second

Lithium alanate (LiAlH4)

dehydrogenation stages, respectively, which decreased about 40.0 kJ/mol H2 and 60.3 kJ/

Rare earth fluorides

mol H2 compared with those of pure LiAlH4. In addition, the sample doped with CeF3

Dehydrogenation

showed the fastest dehydrogenation rate among the REF3 doped LiAlH4 samples at both 125  C and 150  C during the isothermal desorption. The phase changes in REF3 doped LiAlH4 samples during ball milling and dehydrogenation were examined using X-ray diffraction and the mechanisms related to the catalytic effects of REF3 were proposed. Copyright © 2014, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.

Introduction Fossil fuel has been used as the major energy resource for both family lives and industries for centuries. However, severe energy and environment crises are now becoming world-wide issues due to the excessive utilization of fossil fuel [1]. Therefore, developing alternative clean energy resources is in great demand. Owing to the advantages such as cleanness,

environmental friendliness and high efficiency, hydrogen energy has been considered as a promising energy resource for future. It provides an ideal solution to meet the global energy requirements while reducing the environmental pollution with improved energy security. However, the bottle neck toward the full implementation of “Hydrogen Economy” [1] is the lack of efficient hydrogen carriers for various industrial demands and particularly for the commercialization of hydrogen-fueled vehicles. Since the serious flaws of

* Corresponding author. National Engineering Research Center of Light Alloy Net Forming & School of Materials Science and Engineering, Shanghai Jiao Tong University, Shanghai 200240, China. Tel.: þ86 21 54742381; fax: þ86 21 34203730. E-mail address: [email protected] (J. Zou). http://dx.doi.org/10.1016/j.ijhydene.2014.05.103 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 1 6 4 2 e1 1 6 5 0

gaseous or liquid state hydrogen storage have already been confirmed [2], solid state hydrogen storage materials are now becoming attractive hydrogen carriers due to their high safety and high gravimetric/volumetric hydrogen capacity [1,3,4]. Among those hydrogen storage materials, Lithium alanate (LiAlH4) has attracted growing interest in recent years [5]. It is generally acknowledged that LiAlH4 theoretically contains 10.5 wt% H2 and decomposes in the following three steps upon heating [6,7].

3LiAlH4 ¼ Li3AlH6 þ 3H2 þ 2Al (150e175  C);

(1)

Li3AlH6 ¼ 3LiH þ Al þ 3/2H2 (180e220  C);

(2)

3LiH þ 3Al ¼ 3LiAl þ 3/2H2 (>400  C).

(3)

The amount of hydrogen release is 5.3, 2.6, and 2.6 wt% for reaction (1) (R1), reaction (2) (R2) and reaction (3) (R3), respectively [8]. Though the third step requires temperature as high as 400  C, a hydrogen release of 7.9 wt% from the first two steps is still attracting for potential applications. However, the poor hydrogen sorption properties of LiAlH4, such as the relatively high desorption temperatures and unfavorable hydrogen desorption kinetics of the first two steps, are still far from the industrial requirements. To overcome these obstacles, numerous approaches using ball-milling and catalysts additions have been attempted in recent years [9e17]. Among different catalysts, Ti containing species, such as TiO2, TiC, TiF3, TiN and TiCl3 [9e13], showed great effects in improving the hydrogen desorption properties of LiAlH4. In addition, many carbon materials, including carbon nano fibers (CNFs), graphitic nano fibers (GNFs), single-walled carbon nano tubes (SWCNTs), and multi-walled carbon nano tubes (MWCNTs) [14e17]were also testified to be able to destabilize LiAlH4. Moreover, metal halides, such as VBr3, VCl3, ZrCl4, NbF5, NiCl2 and K2TiF6 [18e23] were doped into LiAlH4 and reduced significantly the decomposition temperatures of LiAlH4. However, as far as we know, limited studies have reported on the desorption behaviors of LiAlH4 doped with metal fluorides and no rare-earth metal fluorides have yet been considered as catalysts for LiAlH4 before. According to Ref. [24], it has been proven that F offers great benefits in improving the dehydrogenation properties of Li3AlH6. Meanwhile, F anions can substitute H anions during dehydrogenation reaction, which will result in favorable thermodynamic modifications [25]. Addition of CeF3 nano particles could successfully improve hydrogen storage properties of NaAlH4 [37]. In terms of borohydrides, it has been validated that RCl3 (R ¼ Y, Dy, Gd) can be used to destabilize LiBH4 [26]. The addition of YF3 to NaBH4 enables the reversibility of NaBH4 by the formations of NaF and YB4 [27]. Chong et al. [28] recently reported that the addition of PrF3 to NaBH4 notably decreased the dehydrogenation temperature and benefited hydrogen sorption kinetics of NaBH4. Considering the experimental and theoretical findings mentioned above, three rare earth metal fluorides, YF3, LaF3

11643

and CeF3, were added into LiAlH4 as catalysts through ball milling with the purpose to explore their effects on the dehydrogenation properties and the corresponding mechanisms were carefully investigated.

Experimental Sample preparation The starting materials, LiAlH4 (97 wt%), YF3 (99.99 wt%) and LaF3 (99.99 wt%), were all purchased from Aladdin Reagent Database, Inc. CeF3 (99.9 wt%) was obtained from Alfa Aesar. No further purification was conducted for all the chemicals. Ball milling of LiAlH4 (typically 1e2 g) with 5 mol% YF3, LaF3 or CeF3 was performed in a QM-QX planetary ball mill with a rotational speed of 300 r/min. The milling time for all samples was set as 8 h with the cycles containing 0.5 h milling time and 0.5 h dwelling time. Stainless steel balls were added into the milling vessel with a ball-to-powder mass ratio of 30:1. All materials handling was performed in a glove-box filled with purified argon (99.9%) before and after sample preparations.

Characterizations Phases in the samples before and after dehydrogenation as well as in the middle of dehydrogenation were examined at room temperature by using a powder X-ray diffraction (XRD) apparatus (D/max 2550VL/PCX) equipped with a Cu-Ka radiation source. Before measurements, samples were flattened uniformly onto a sample holder. The sample holder with arch glass on each side was sealed with airtight tape to avoid direct exposure to air, which shows a broad peak around 2q ¼ 15 in XRD patterns. Differential scanning calorimetry (DSC, Netzsch STA449F3 Jupiter) analyses of the dehydrogenation process were conducted for different samples under 0.1 MPa of argon flow. The heating rate was set at 3, 5 and 10 K/min with the temperature rising from 50 to 300  C. The hydrogen desorption kinetic behaviors were examined by using a Sievert type apparatus manufactured by Shanghai Institute of Microsystem and Information Technology [27,28]. The pressure-composition-temperature (PeCeT) apparatus can be operated at different temperatures with the pressure up to 4.6 MPa. The hydrogen desorption capacity is calculated as the amount of hydrogen released over the total weight of LiAlH4 and REF3 additives. Temperature programmed dehydrogenation (TPD) measurements were performed on the as-received LiAlH4 and REF3 doped samples starting from vacuum. The samples were then heated from room temperature to 250  C at a heating rate of 3 K/min. The hydrogen desorbed was calculated using volumetric methods. For measurements of isothermal dehydrogenation kinetics, the sample vessel was first evacuated to vacuum for 10 min at room temperature and was then rapidly heated with a rate of 15 K/min and kept at the desired temperature.

11644

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 1 6 4 2 e1 1 6 5 0

Results and discussions Dehydrogenation properties Non-isothermal dehydrogenation properties Fig. 1 shows the TPD curves of the as-received, YF3, CeF3 and LaF3 doped LiAlH4 samples. From these curves, it is obvious that REF3 (RE ¼ Y, Ce, La) addition has striking effects on both the first and second stages of decomposition in LiAlH4, especially for the first one. Compared to the as-received LiAlH4, which started to decompose at 170  C, the onset dehydrogenation temperatures are all reduced for REF3 doped samples. For instance, doping with LaF3 and CeF3 can lower down the onset dehydriding temperature of LiAlH4 by 80  C and 90  C, respectively. For the second stage of dehydrogenation, the addition of LaF3 and CeF3 both reduced the onset temperature by around 20  C. Furthermore, from the slops of these curves, it is evident that doping with LaF3 has significant promoting effect on the dehydrogenation rate of the first stage. While for the second stage, doping with CeF3 can accelerate dramatically the dehydrogenation rate. Table 1 summarizes the data obtained from TPD measurements. For the as-received LiAlH4, the hydrogen released from the first stage and the second stage was 4.83 wt% and 2.46 wt%, respectively. The total amount of hydrogen released was 7.29 wt%. The REF3 doped samples showed a reduction in hydrogen desorption capacity, which resulted from the addition of catalysts. Alternatively, YF3 and CeF3 may induce the decomposition of LiAlH4 during ball milling, which will be discussed in the following sections. Fig. 2 gives the DSC curves of the as-received and REF3 doped LiAlH4 samples measured within the temperature range of 50e300  C at the heating rate of 3 K/min. Clearly, there existed two endothermic peaks and one exothermic peak for the un-doped LiAlH4. The first endothermic peak corresponded to the melting of LiAlH4, while the first exothermic peak was related to R1, and the second endothermic peak resulted from R2. Surprisingly, three endothermic peaks and one exothermic peak were observed in the

DSC curves of samples doped with YF3 and LaF3, as seen in Fig. 2b and c. The only possible explanation is that the addition of LaF3 or YF3 rendered partial decomposition of LiAlH4 before melting, which corresponded to the first endothermic peak. The second endothermic peak was clearly due to the melting of the residual LiAlH4. The first exothermic peak and the third endothermic peak should be ascribed to R1 and R2, respectively. According to Fig. 2d, after doping with CeF3, the number of thermal events was reduced to only two exothermic processes with peaks located at 127.4  C and 189.7  C, respectively. The first one was related to the decomposition (R1) of the solid LiAlH4 and the second corresponded to R2. Apparently, no melting peak was observed in this DSC profile. This phenomenon can be attributed to the fact that doping with CeF3 enables LiAlH4 to decompose at a temperature much lower (~40  C) than its melting temperature. Many works have reported the doping of various catalysts in LiAlH4 and have investigated the effects of these catalysts on the thermal decomposition behaviors of LiAlH4. It is worth noting here that, for LiAlH4 doped with MnFe2O4 [29], TiC [10], TiF3 [11], Al3Ti [30], TiO2 [9], Nb2O5 and Cr2O3 [31], Fe [32], NbF5 [21], NiCl2 and TiN [33], the first dehydrogenation step of solid LiAlH4 is exothermic. Unlike these studies, for all the three REF3 doped samples in the present work, the first dehydrogenation process is endothermic, which is in agreement with the results shown in Refs. [12] and [23]. This difference may be explained by the study of Chen et al. [34], which claimed that the first dehydrogenation step of LiAlH4 is intrinsically endothermic. Adding REF3 may help to reveal the endothermic nature of the first decomposition event by uncoupling the first decomposition event from the melting-decompositionsolidification process that appears at 165e180  C in the undoped LiAlH4. Through the above results, both the significant reduction in peak temperatures and the revelation of the endothermic nature for the first dehydrogenation stage indicated that the dehydrogenation properties of LiAlH4 could be significantly improved by the addition of REF3. In particular, CeF3 worked as a more effective catalyst than YF3 and LaF3. Yet it is still worth noting that the decomposition temperatures measured by DSC was higher than those obtained by PCT for the same sample, which was also observed in other works [9,23]. This phenomenon can be mainly attributed to the different atmospheres in DSC and PCT measurements.

Dehydrogenation activation energy (Ea) To further study the desorption behaviors of LiAlH4 with and without the addition of REF3, the dehydrogenation activation energy (Ea) for the first and second stage of the as-received LiAlH4 and REF3 doped LiAlH4 were determined by using Kissinger method [35]:   dln Tb2 P

d Fig. 1 e Temperature programmed desorption (TPD) curves of the as-received LiAlH4 and LiAlH4 doped with 5 mol% YF3, CeF3 and LaF3.

  ¼ 1 TP

Ea R

(4)

where b is the heating rate, Tp the peak temperature, and R the gas constant. Fig. 3 demonstrates the DSC curves of these four samples with a heating rate of 3, 5 and 10 K/min. Just as

11645

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 1 6 4 2 e1 1 6 5 0

Table 1 e Data obtained from TPD measurements for dehydrogenation in as-received and REF3 doped LiAlH4 samples. Samples

LiAlH4 LiAlH4 þ 5 mol% YF3 LiAlH4 þ 5 mol% LaF3 LiAlH4 þ 5 mol% CeF3

Dehydrogenation temperature of R1 ( C)

Dehydrogenation temperature of R2 ( C)

Final temperature of dehydrogenation ( C)

Hydrogen released from R1 (wt%)

Hydrogen released from R2 (wt%)

170 140 90 80

187 180 167 167

255 250 220 215

4.83 3.74 3.74 3.74

2.46 2.12 1.76 1.86

expected, the peaks of dehydrogenation shifted toward higher temperatures as the heating rate increased. The plots based on the DSC data and using Kissinger method were shown in the insets of Fig. 3. Additionally, the temperatures corresponding to those peaks on the DSC curves obtained at different heating rates in Fig. 3 were listed in Table 2. According to the Kissinger method, the Ea values for the two stages of the undoped and REF3 doped samples were given in Table 3. For comparison, results from other published works were also included. Evidently, for samples doped with CeF3 and YF3, Ea values for the first stage were 40 and 19.7 kJ/ mol H2 lower than that of the as-received LiAlH4, respectively. For the second stage, Ea values for REF3 doped samples were all reduced. In particular, from Table 3, Ea value for the second

stage of CeF3 doped sample was only 77.2 kJ/mol H2, which was 60.3 kJ/mol H2 lower than that of the as-received LiAlH4 and was the lowest among all reported data. The outcomes above clearly indicated that adding REF3 (RE ¼ Y, La, Ce), especially CeF3, greatly reduced the energy barriers for both the two dehydrogenation stages of LiAlH4.

Isothermal dehydrogenation properties Fig. 4 demonstrates the isothermal dehydrogenation curves of as-received and REF3 doped LiAlH4 at 125  C and 150  C. Apparently, at 125  C, for pristine LiAlH4, nearly no hydrogen was released within 2 h and the hydrogen desorption was barely 0.38 wt% in 3 h. The hydrogen released by samples doped with YF3, LaF3, and CeF3 in 3 h were 3.34 wt%, 3.46 wt%

Fig. 2 e DSC profiles of as-received (a) and YF3 (b), LaF3 (c), CeF3 (d) doped LiAlH4 samples measured in the temperature range of 50e300  C at the heating rate of 3 K/min.

11646

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 1 6 4 2 e1 1 6 5 0

Fig. 3 e DSC profiles of as-received (a) and YF3 (b), LaF3 (c), CeF3 (d) doped LiAlH4 samples measured at the heating rate of 3 K/ min (1), 5 K/min (2) and 10 K/min (3). Inset graphs are the Kissinger plots for the first and second dehydrogenation step.

and 3.56 wt%, respectively. According to the amount of hydrogen released from R1 listed in Table 1, the first decomposition step of REF3 doped LiAlH4 was almost completed. When the temperature was raised to 150  C, it was clearly observed that the onset dehydrogenation process of the as-

Table 2 e Peak desorption temperatures of the asreceived and REF3 doped LiAlH4 samples measured by DSC at different heating rates. Samples

LiAlH4

LiAlH4 þ 5 mol% YF3 LiAlH4 þ 5 mol% LaF3 LiAlH4 þ 5 mol% CeF3

Heating rate/ K min1

Peak temperature for R1/ C

Peak temperature for R2/ C

3 5 10 3 5 10 3 5 10 3 5 10

180.9 190.7 206.4 157.2 168.2 185.9 151.5 158.5 172.5 127.4 139.6 163.5

244.7 254.3 263.7 223.0 229.4 243.6 217.2 224.0 236.7 189.7 199.6 216.2

received LiAlH4 was going through an incubation period which lasted for 60 min. The hydrogen release reached about 2.25 wt% in 2 h, indicating that even the first stage of decomposition was far from being finished. In contrast, the period taken by samples doped with YF3, LaF3 and CeF3 to complete the first stage was 45, 35 and 30 min, respectively. Admittedly, the second stage decomposition of LiAlH4 did not happen yet at this temperature for REF3 doped samples. Nevertheless, the dehydrogenation kinetics of LiAlH4 was still considerably improved by the addition of REF3. From the above results, the catalytic effect of REF3 on the dehydrogenation rate of LiAlH4 has an order: CeF3>LaF3>YF3, which is in accordance with the results of DSC and thermal desorption measurements.

Dehydrogenation mechanisms To explore the reactions occurred between LiAlH4 and REF3 (RE ¼ Y, La, Ce) during ball milling and dehydrogenation, XRD measurements were conducted on REF3 doped LiAlH4 samples at different states and the results were presented in Figs. 5e7. From Fig. 5a, it is evident that, after ball milling, there appeared weak peaks from both Al and Li3AlH6 beside those peaks belonging to LiAlH4 and YF3. Since no other products

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 1 6 4 2 e1 1 6 5 0

11647

Table 3 e Comparison of activation energies of asreceived and doped LiAlH4. Parameter Desorption activation energy (kJ/mol H2)

Samples As-received With CeF3 With LaF3 With YF3 As-received

With nano Nb2O5 With K2TiF6 With TiCl3$1/3AlCl3

1st step 2nd step 80.9 40.9 81.5 61.2 122 81 86 64.5 78.2 89

137.5 77.2 118.4 113.1 153 108 101 79 90.8 103

Ref. This work

[2]a [5] [10] [31] [23] [5]

In the present work, the heating rates® (3, 5, 10 K/min) set for the DSC measurements and the purity of the as-received LiAlH4 (~97 wt %) are quite close to those in Ref. [2]. Thus the Ea values given in Ref. [2] are suitable for comparison with the data obtained in this study. a

were observed in the XRD pattern of ball milled sample, the formation of Al and Li3AlH6 was more likely a result of the decomposition of LiAlH4 alone rather than the reaction between LiAlH4 and YF3. After dehydrogenation at 170  C (Fig. 5b), the increase in intensity of peaks from Al and Li3AlH6 suggested that R1 was completed. Meanwhile, new peaks corresponding to LiF and YH3 appeared, indicating that the interaction between LiAlH4 and YF3 occurred upon heating. When the temperature was raised to 250  C (Fig. 5c), peaks from YF3 phase, though still existed, were largely weakened with the enhancement of peaks from YH3 and the appearance of peaks from YH2. According to Ref. [38], the formation of YHx(x ¼ 2, 3) may be related to the interchange between F and H due to their similar ionic radius. In addition, it has been calculated that YH3 can transform to YH2 at 287  C under 1 bar of hydrogen [40], which may explain the existence of YH2 at 250  C. Many studies have proven the catalytic effects of metal hydrides on the decomposition of complex metal hydrides. For instance, TiH2 and KH formed in K2TiF6 doped LiAlH4 played important roles for the first dehydrogenation step of LiAlH4 [23]. Addition of KH notably decreased the

Fig. 5 e XRD patterns of 5 mol% YF3 doped LiAlH4 at different states: (a) as milled; (b) dehydrogenated at 170  C; (c) dehydrogenated at 250  C.

decomposition enthalpy of Na3AlH6 [39]. Similarly, PrH2 developed in the 3NaBH4/PrF3 system catalyzed the dehydrogenation processes [28]. Furthermore, YH3 worked to destabilize LiBH4 and resulted in improved hydrogen desorption performance of LiBH4 [40e42]. According to Ref. [43], LiBH4 could also be destabilized by CeH2/LaH2 and a significant improvement in hydrogen storage was observed. In this case, the in situ formed YH3 may catalyze the first dehydrogenation step of LiAlH4. YH3 together with YH2 may act as catalysts to facilitate the decomposition of newly developed Li3AlH6. In addition, LiF formed during dehydrogenation played the role as the nucleation center for Al since they share the same Fm3m space group [11]. Above all, YH3, YH2 and LiF all together provided a synergetic catalytic effect in improving the dehydrogenation kinetics of LiAlH4. The XRD patterns of LaF3 doped LiAlH4 sample at different states were presented in Fig. 6. Fig. 6a showed that all diffraction peaks belonged to LiAlH4 and LaF3 except the broad

Fig. 4 e Isothermal dehydrogenation curves of as-received and REF3 doped LiAlH4 samples measured at 125  C (a) and 150  C (b).

11648

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 1 6 4 2 e1 1 6 5 0

Fig. 6 e XRD patterns of 5 mol% LaF3 doped LiAlH4 at different states: (a) as milled; (b) dehydrogenated at 170  C; (c) dehydrogenated at 250  C. The simulated peak positions of (La, Al)F3 are present by small sticks. peak from the tape, implying that LiAlH4 did not decompose during ball milling and it was merely mechanically mixed with LaF3. After dehydrogenation at 170  C, peaks from Al appeared with the rising temperature, as seen in Fig. 6b. It is noteworthy that peaks located at 2q ¼ 25.1 , 25.8 , 28.7 , 36.3 , 45.5 , 46.4 , 52.6 and 54.7 cannot be assigned to any known phases. However, the distribution of these peaks is fairly similar as that of LaF3, suggesting the same crystal structure shared by them. To our best knowledge, it could only be described as (La, Al) F3, which was caused by the partial substitution of La3þ by Al3þ in the LaF3 lattice. According to the XRD patterns, the (La, Al) F3 peaks could be indexed to a hexagonal unit cell (P-3c1) with unit cell parameters a ¼ 6.940 Å, b ¼ 6.940 Å and c ¼ 7.098 Å, smaller than those of LaF3 (a ¼ 7.187 Å, b ¼ 7.187 Å, c ¼ 7.350 Å)

Fig. 7 e XRD patterns of 5 mol% CeF3 doped LiAlH4 at different states: (a) as milled; (b) dehydrogenated at 170  C; (c) dehydrogenated at 250  C. The simulated peak positions of (Ce, Al)F3 are present by small sticks.

due to the smaller ionic radii of Al3þ (0.51 Å) compared to that of La3þ (1.06 Å). Simulation was carried out using the above unit cell parameters and the simulated peak positions of (La, Al)F3 were also presented in Fig. 5. Clearly, the simulated XRD peak positions agreed well with those obtained experimentally. The newly formed LaH2.3, LiF and (La, Al)F3 all indicated the reaction between LaF3 and LiAlH4 while LaH2.3 and LiF catalyzed LiAlH4 to decompose. When temperature was raised to 250  C, peaks from LaF3 disappeared while those from (La, Al)F3 weakened with the enhancement of peaks corresponding to LaH2.3. Meanwhile, newly-presented peaks from Al11La3/ Al4La3 phases were also observed. Therefore, it can be concluded that the La hydride is derived from the reaction between LaF3 and LiAlH4 upon heating and the La hydride can further react with Al or Li3AlH6 to form LaeAl alloys. It has been proven that the in situ formed metallic alloy phases may exert active effects on the decomposition of LiAlH4. For example, Al3Ti facilitated the catalytic dissociation and recombination of hydrogen molecules on its surface [36], and doping Ti3Al into LiAlH4 also reduced the decomposition temperature of both two steps [44]. Sun et al. [45] reported that TieAl clusters formed in the NaH/Al mixture catalyzed the dehydriding/rehydriding processes. According to Ref. [33], NiAl3 also acted as the dehydrogenation catalyst for the partial decomposition of LiAlH4. It is thus reasonable that Al11La3/ Al4La3 phases formed in the LaF3 doped LiAlH4 exert similar catalytic effects for the hydrogen desorption. From Fig. 7, it is surprising that no diffraction peaks of CeF3 were detected in samples at all states, indicating that a chemical reaction already occurred between CeF3 and LiAlH4 during ball milling. Similar to the XRD patterns of LaF3 doped LiAlH4, unexpected peaks at 2q ¼ 25.4 , 26.2 , 29.2 , 36.8 , 46.3 , 47.2 , 53.6 and 55.6 could only be ascribed to (Ce, Al) F3, in which Ce3þcations in the CeF3 lattice were partially substituted by Al3þ. The (Ce, Al) F3 and CeF3 share the same hexagonal crystal structure with the space group P-3c1. Compared with the unit cell parameters of CeF3 (a ¼ 7.129 Å, b ¼ 7.129 Å, c ¼ 7.287 Å), those of (Ce, Al) F3were calculated to be a ¼ 6.849 Å, b ¼ 6.849 Å and c ¼ 7.003 Å based on the XRD patterns. Simulated peak positions according to the calculated lattice constants were also included in Fig. 7, which agreed well with the experimental ones. Unlike the LaF3 doped sample, Al, LiF and CeH2.51 have already developed during ball milling (Fig. 7a), suggesting a more severe reaction between CeF3 and LiAlH4. As seen from Fig. 7b, when the ball milled sample was heated to 170  C, merely the first decomposition of LiAlH4 happened. However, new peaks of Al4Ce were detected after dehydrogenation at 250  C (Fig. 7c). Meanwhile, peaks from CeH2.51 were still observable. These results suggested that CeF3 might act in a similar way as LaF3 did, and the real catalysts during the decomposition of LiAlH4 would be CeH2.51 and Al4Ce rather than CeF3. From above analyses, the role of REF3 (RE ¼ Y, La, Ce) additives during the ball milling and the dehydrogenation process of LiAlH4 could be clarified. YF3 did not react with LiAlH4 during ball milling but reacted with LiAlH4 upon heating, producing YH2 and YH3. These hydrides together with LiF catalyzed the two decomposition steps of LiAlH4. LaF3 and CeF3played the similar role in catalytic mechanisms by forming hydrides (LaH2.53 and CeH2.51) first and then

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 1 6 4 2 e1 1 6 5 0

producing alloy phases (Al11La3/Al4La3 and Al4Ce) upon heating. The in situ formed rare earth hydrides, Al-rare earth alloys together with LiF provided a synergetic catalytic effect for the remarkably improved dehydrogenation kinetics of LiAlH4. Additionally, La3þ/Ce3þ could be partially substituted by Al3þ in LaF3/CeF3 lattices, leading to the unexpected peaks present in the corresponding XRD pattern. However, the reaction between CeF3 and LiAlH4 has already occurred even during ball milling, representing greater catalytic activity of CeF3 over LaF3. The combined experimental results have proven that the destabilization effect of these REF3 (RE ¼ Y, La, Ce) catalysts on LiAlH4 comes in the order: CeF3>LaF3>YF3.

Conclusions The present work shows that doping REF3 (RE ¼ Y, La, Ce) through ball milling has exhibited a dramatic improvement in dehydrogenation properties of LiAlH4. In particular, CeF3 tends to be a more promising catalyst than the other two. That is, it produces a more significant decrease in dehydrogenation temperatures for both two dehydrogenation steps and notably improves the desorption kinetics of LiAlH4. It is proven from the TPD results that the addition of CeF3 reduces the onset decomposition temperature of LiAlH4 by 90  C compared with that of as-received LiAlH4. Results of DSC measurements reveal that REF3 (RE ¼ Y, La, Ce), especially CeF3, not only reduces the peak temperatures of both two dehydrogenation steps of LiAlH4, but also reveal the endothermic nature of the first dehydrogenation reaction. Furthermore, a significant reduction in Ea values for the two desorption steps has been witnessed in the REF3 (RE ¼ Y, La, Ce) doped LiAlH4 samples. In particular, Ea values of CeF3 doped LiAlH4 are merely 40.9 kJ/ mol H2 and 77.2 kJ/mol H2 for the first and second dehydrogenation steps, which are about 40 kJ/mol H2 and 60.3 kJ/mol H2 lower than those of the as-received LiAlH4, respectively. In contrast to the poor dehydrogenation kinetics of pure LiAlH4, REF3 doped samples enjoyed higher dehydrogenation rates with an order: CeF3>LaF3>YF3. For instance, CeF3 doped LiAlH4 finishes its first dehydrogenation step within 30min upon heating to 150  C. The catalytic mechanisms of REF3 on the dehydrogenation of LiAlH4 were clarified through XRD analyses. For the YF3 doped sample, the in situ developed LiF and YHx(x ¼ 2, 3) worked together to catalyze the dehydrogenation of LiAlH4. While for the LaF3 and CeF3 doped LiAlH4 samples, (La, Al)F3 and (Ce, Al)F3 phases were detected. Meanwhile, besides LiF and rare earth hydrides (LaH2.53 and CeH2.51), intermetallic phases (Al11La3/Al4La3 and Al4Ce) formed upon heating also played an important role for the significantly enhanced dehydrogenation kinetics in LaF3 and CeF3 doped LiAlH4 samples. Therefore, doping REF3 (RE ¼ Y, La, Ce), especially CeF3, is of great value in improving the dehydrogenation performances of LiAlH4 and deserves further investigations.

Acknowledgments Prof. Zou would like to thank the support from the Science and Technology Committee of Shanghai under Nos. 10JC1407700

11649

and 11ZR1417600, and 'Pujiang project' (No. 11PJ1406000). This work is partly supported by Research Fund for the Doctoral Program of Higher Education of China (No. 20100073120007) and from the Shanghai Education Commission (No. 12ZZ017).

references

[1] Schlapbach L, Zu¨ttel A. Hydrogen-storage materials for mobile applications. Nature 2001;414:353e8. [2] Varin RA, Czujko T, Wronski ZS. Nanomaterials for solid state hydrogen storage. New York, NY: Springer Science &Business Media; 2009. [3] Varin RA, Czujko T, Chiu C, Pulz R, Wronski ZS. Synthesis of nanocomposite hydrides for solid-state hydrogen storage by controlled mechanical milling techniques. J Alloys Compd 2009;483:252e5. [4] Grochala W, Edwards PP. Thermal decomposition of the noninterstitial hydrides for the storage and production of hydrogen. Chem Rev 2004;104:1283e315. [5] Andreasen A. Effect of Ti-doping on the dehydrogenation kinetic parameters of lithium aluminum hydride. J Alloys Compd 2006;419:40e4. [6] DilitJ A, Ashby EC. A study of the thermal decomposition of complex metal hydrides. Inorg Chem 1972;11:1230e6. [7] Resan M, Hampton MD, Lomness JK, Slattery DK. Effects of various catalysts on hydrogen release and uptake characteristics of LiAlH4. Int J Hydrogen Energy 2005;30:1413e36. [8] Ares JR, Aguey-Zinsou KF, Porcu M, Sykes JM, Dornheim M, Klassen T, et al. Thermal and mechanically activated decomposition of LiAlH4. Mater Res Bull 2008;43:1263e75. [9] Ismail M, Zhao Y, Yu XB, Nevirkovets IP, Dou SX. Significantly improved dehydrogenation of LiAlH4catalysed with TiO2nanopowder. Int J Hydrogen Energy 2011;36:8327e34. [10] Rafiud D, Lin Z, Li P, Qu XH. Catalytic effects of nano-sized TiC additions on the hydrogen storage properties of LiAlH4. J Alloys Compd 2010;508:119e28. [11] Liu SS, Sun LX, Zhang Y, Xu F, Zhang J, Chu HL, et al. Effect of ball milling time on the hydrogen storage properties of TiF3doped LiAlH4. Int J Hydrogen Energy 2009;34:8079e85. [12] Li L, Qiu FY, Wang YJ, Xu YN, An CH, Liu G, et al. Enhanced hydrogen storage properties of TiN-LiAlH4 composite. Int J Hydrogen Energy 2013;38:3695e701. [13] Chen J, Kuriyama N, Xu Q, Takeshita HT, Sakai T. Reversible hydrogen storage via titanium-catalyzed LiAlH4 and Li3AlH6. J Phys Chem B 2001;105:11214e20. [14] Hima KL, Viswanathan B, Srinivasa MS. Dehydriding behaviour of LiAlH4-the catalytic role of carbon nanofibres. Int J Hydrogen Energy 2008;33:366e73. [15] Hudson MSL, Raghubanshi H, Pukazhselvan D, Srivastava ON. Effects of helical GNF on improving the dehydrogenation behavior of LiMg(AlH4)3 and LiAlH4. Int J Hydrogen Energy 2010;35:2083e90. [16] Ismail M, Zhao Y, Yu XB, Ranjbar A, Dou SX. Improved hydrogen desorption in lithium alanate by addition of SWCNT-metallic catalyst composite. Int J Hydrogen Energy 2011;36:3593e9. [17] Hsu WC, Yang CH, Tsai WT. Catalytic effect of MWCNTs on the dehydrogenation behavior of LiAlH4. Int J Hydrogen Energy 2014;39:927e33. [18] Ares JR, Aguey-Zinsou KF, Elsaesser M, Ma XZ, Dornheim M, Klassen T, et al. Mechanical and thermal decomposition of LiAlH4 with metal halides. Int J Hydrogen Energy 2007;32:1033e40.

11650

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 1 6 4 2 e1 1 6 5 0

[19] Blanchard D, Brinks HW, Hauback BC, Norby P. Desorption of LiAlH4 with Ti- and V-based additives. Mater Sci Eng B 2004;108:54e9. [20] Suttisawat Y, Rangsunvigit P, Kitiyanan B, Muangsin N, Kulprathipanja S. Catalytic effect of Zr and Hf on hydrogen desorption/absorption of NaAlH4 and LiAlH4. Int J Hydrogen Energy 2007;32:1277e85. [21] Ismail M, Zhao Y, Yu XB, Dou SX. Effects of NbF5addition on the hydrogen storage properties of LiAlH4. Int J Hydrogen Energy 2010;35:2361e7. [22] Suna T, Huanga CK, Wang H, Sun LX, Zhu M. The effect of doping NiCl2 on the dehydrogenation properties of LiAlH4. Int J Hydrogen Energy 2008;33:6216e21. [23] Li ZB, Liu SS, Si XL, Zhang J, Jiao CL, Wang S, et al. Significantly improved dehydrogenation of LiAlH4 destabilized by K2TiF6. Int J Hydrogen Energy 2012;37:3261e7. [24] Liu SS, Zhang Y, Sun LX, Zhang J, Zhao JN, Xu F, et al. The dehydrogenation performance and reaction mechanisms of Li3AlH6 with TiF3additive. Int J Hydrogen Energy 2010;35:4554e61. [25] Yin LC, Wang P, Kang XD, Sun CH, Cheng HM. Functional anion concept: effect of fluorine anion on hydrogen storage of sodium alanate. Phys Chem Chem Phys 2007;9:1499e502. [26] Toyoto S, Kazutoshi M, Yuko N, Kenji O, Li HW, Tatsuo N, et al. Experimental and computational studies on solventfree rare-earth metal borohydrides R(BH4)3 (R¼Y, Dy, and Gd). Phy Rev B 2008;77:1e8. [27] Zou JX, Li LJ, Zeng XQ, Ding WJ. Reversible hydrogen storage in a 3NaBH4/YF3composite. Int J Hydrogen Energy 2012;37:17118e25. [28] Chong LN, Zou JX, Zeng XQ, Ding WJ. Reversible hydrogen sorption in NaBH4 at lower temperatures. J Mater Chem A 2013;1:13510e23. [29] Zhai FQ, Li P, Sun AZ, Wu S, Wan Q, Zhang WN, et al. Significantly improved dehydrogenation of LiAlH4destabilized by MnFe2O4 nanoparticles. J Phys Chem C 2012;116:11939e45. [30] Balema VP, Wiench JW, Dennis KW, Pruski M, Pecharsky VK. Titanium catalyzed solid-state transformations in LiAlH4 during high-energy ball-milling. J Alloys Compd 2001;329:109e14. [31] Rafiud D, Qu XH, Li P, Lin Z, Mashkoor A. Hydrogen sorption improvement of LiAlH4 catalyzed by Nb2O5 and Cr2O3 nanoparticles. J Phys Chem C 2011;115:13088e99. [32] Varin RA, Parviz R. The effects of the micrometric and nanometriciron (Fe) additives on the mechanical and

[33]

[34]

[35] [36]

[37]

[38]

[39]

[40]

[41]

[42]

[43]

[44]

[45]

thermal dehydrogenation of lithium alanate (LiAlH4), its selfdischarge at low temperatures and rehydrogenation. Int J Hydrogen Energy 2012;37:9088e102. Rafiud D, Qu XH, Li P, Mashkoor A, Lin Z. Comparative catalytic effects of NiCl2, TiC and TiN on hydrogen storage properties of LiAlH4. Rare Met 2011;30:27e34. Langmi HW, McGrady GS, Liu XF, Jensen CM. Modification of the H2 desorption properties of LiAlH4 through doping with Ti. J Phys Chem C 2010;114:10666e9. Kissinger HE. Reaction kinetics in differential thermal Analysis. Anal Chem 1957;29:1702e6. Hendrik WB, Hauback BC, Sesha SS, Craig MJ. Synchrotron Xray studies of Al1-yTiy formation and re-hydriding inhibition in Ti-enhanced NaAlH4. J Phys Chem B 2005;109:15780e5. Fan XL, Xiao XZ, Chen LX, Zhang LT, Shao J, Ge HW, et al. Significantly improved hydrogen storage properties of NaAlH4catalyzed by Ce-based nanoparticles. J Mater Chem A 2013;1:9752e9. Gosalawit-Utke R, Bellosta von Colbe JM, Dornheim M, Jensen TR, Cerenius Y, Minella CB, et al. LiF-MgB2system for reversible hydrogen storage. J Phys Chem C 2010;114:10291e6. Liu YF, Liang C, Zhou H, Gao MX, Pan HG, Wang QD. A novel catalyst precursor K2TiF6 with remarkable synergetic effects of K, Ti and F together on reversible hydrogen storage of NaAlH4. Chem Commun 2011;47:1740e2. Shim JH, Lim JH, Rather SU, Lee YS, Reed D, Kim YY, et al. Effect of hydrogen back pressure on dehydrogenation behavior of LiBH4-based reactive hydride composites. J Phys Chem Lett 2010;1:59e63. Gennari FC. Improved hydrogen storage reversibility of LiBH4destabilized by Y(BH4)3and YH3. Int J Hydrogen Energy 2012;37:18895e903. Shim JH, Lee YS, Suh JY, Cho W, Han SS, Cho YW. Thermodynamics of the dehydrogenation of the LiBH4YH3composite: experimental and theoretical studies. J Alloys Compd 2012;510:L9e12. Gennari FC. Destabilization of LiBH4 by MH2 (M¼Ce, La) for hydrogen storage: nano structural effects on the hydrogen sorption kinetics. Int J Hydrogen Energy 2011;36:15231e8. Micheal MR, Hampton MD, Lomness JK, Slattery DK. Effects of TixAly catalysts on hydrogen storage properties of LiAlH4 and NaAlH4. Int J Hydrogen Energy 2005;30:1417e21. Fang F, Zhang J, Zhu J, Chen GR, Sun DL, He B, et al. Nature and role of Ti species in the hydrogenation of a NaH/Al mixture. J Phys Chem C 2007;111:3476e9.