ww.sciencedirect.com

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 en e r g y 4 0 ( 2 0 1 5 ) 3 6 3 3e3 6 4 1

Available online at w

ScienceDirect

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

Influence of La on reduction behaviour and Nimetal surface area of NieAl2O3 catalysts for COx

free H2 by catalytic decomposition of methane

Ch. Anjaneyulu a, S. Naveen Kumar a, V. Vijay Kumar a,b, G. Naresh a,b,S.K. Bhargava b, K.V.R. Chary a, A. Venugopal a,*

a Inorganic & Physical Chemistry Division, CSIR-Indian Institute of Chemical Technology, Tarnaka, 500007

Hyderabad, Telangana, Indiab College of Science, Engineering and Health, Advanced Materials & Industrial Chemistry, School of Applied Sciences,

RMIT University, GPO BOX 2476, Melbourne 3001, Australia

a r t i c l e i n f o

Article history:

Received 9 September 2014

Received in revised form

7 January 2015

Accepted 14 January 2015

Available online 9 February 2015

Keywords:

Pure H2

Ni metal surface area

CNT

TPR

Raman spectra

* Corresponding author. Tel.: þ91 40 2719172E-mail address: akula@iict.res.in (A. Venu

http://dx.doi.org/10.1016/j.ijhydene.2015.01.00360-3199/Copyright © 2015, Hydrogen Ener

a b s t r a c t

A series of Ni:Al:La mixed oxides derived from hydrotalcite precursors with varying La/Al

mole ratio were examined for COx free H2 production by catalytic decomposition of CH4.

The physico-chemical characteristics of Ni:Al:La were determined by XRD, UV-DRS, SEM,

TEM, Raman spectroscopy, BET-surface area, TPR and O2 pulse chemisorption measure-

ments. Addition of La to NieAl improved the Ni metal surface area and the reduction

behaviour of NiO particles is dramatically changed. Particle size of Ni was similar to the

size of carbon nano fibre in the deactivated catalyst. A direct correlation between Ni metal

surface area and the CH4 decomposition activity was observed.

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

reserved.

Introduction

Hydrogen has been identified as the fuel of the future because

it has no environmental issues due to absence of harmful

gases such as SOx, COx and NOx formation [1e3]. Due to

negligible natural abundance of hydrogen, conventionally

hydrogen is produced via catalytic steam reforming, partial

oxidation and auto thermal reforming of natural gas or steam

reforming of alcohols [4]. However, a major drawback with

these routes is the formation of large amounts of COx as by-

0, þ91 40 27193510; fax:gopal).72gy Publications, LLC. Publ

products. Hydrogen generated by these methods can be uti-

lized in fuel cells only if CO is completely eliminated. Since its

presence even 10 ppm level in H2 stream strongly deteriorates

the efficiency of fuel cell anodic catalyst [5,6]. Therefore, cat-

alytic decomposition of methane (CDM) has become a prom-

ising approach that produces clean H2 with potentially useful

carbon in the form of nano tubes and nano fibres without COx

contamination [7e11]. Thus, produced hydrogen by CDM

process is COx free that can be directly utilized not only in

PEMFC without further purification also for various applica-

tions [5]. The second product is a graphitic carbon which has

þ91 40 27160921.

ished by Elsevier Ltd. All rights reserved.

Table 1 e Physicochemical characteristics of Ni:Al:La catalysts with constant Ni and varied Al:La mole ratios.

Composition of Ni:Al:La Surfacearea (m2/g)

NiO domainsize (nm)

H2 uptake(mmol/gcat) TPR

O2 uptake(mmol/gcat)

SNi

(m2/gcat)H2 yield

(mols H2/mol Ni)Nominal EDX

(a) 2:1:0 2.06:0.94:0 120.0 8.3 9.6 3.2 17.0 1030

(b) 2:0.8:0.2 1.98:0.81:0.19 104.0 5.8 11.9 6.5 34.6 1550

(c) 2:0.7:0.3 1.97:0.72:0.308 93.9 5.3 13.6 7.2 38.5 1885

(d) 2:0.5:0.5 1.99:0.51:0.49 54.6 2.2 12.4 3.96 21.0 1478

(e) 2:0.1:0.9 1.98:0.09:0.92 30.3 2.0 11.5 0.8 4.3 370

(f) 2:0:1 2.08:0:0.92 20.3 7.3 10.2 0.45 2.6 108

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 4 0 ( 2 0 1 5 ) 3 6 3 3e3 6 4 13634

many unique properties such as high resistant to strong acids,

strong bases, as a H2 storage material, in electronic switches,

good mechanical strength and graphitic carbon with high

surface area makes them useful materials as a catalyst sup-

port in thermo and photocatalytic process [9].

Transition metals such as Ni [7,12e14], Fe [15] and Co [16]

are known active metals for methane decomposition. Partic-

ularly the nickel-based catalysts are active at low tempera-

tures and provide higher yield of H2 per unit mass of the active

component [17e19]. However, these catalysts are easily

deactivated at high temperatures. Hence, modification of Ni

based catalysts opens up an interesting area to increase the H2

yields and catalyst stability at higher temperatures in

methane decomposition [20]. In this regard, introduction of

second metal or metal oxides to a Ni catalyst provides sig-

nificant changes in the activity and stability of the catalysts

[21]. Doping of metals such as Cu, Rh, Pd, Ir and Pt to Ni cat-

alysts were studied by Takenaka et al. and suggested that Pd

doping alone increased the Ni activity through an alloying

effect [20]. However, the other studies revealed that Cu doping

significantly enhanced the performance of Ni catalysts to-

wards methane decomposition for better H2 yields [22e25].

In the past few years it is found that the rare earth oxides

such as La2O3 can improve the catalytic properties of a variety

of materials. Modification of NieAl by La2O3 is a topic of in-

terest particularly in the steam reforming of alcohols at high

reaction temperatures for the production of hydrogen [26].

The aim of this study is development of catalysts for the

production of COx free hydrogen. In this investigationwe have

examined the influence of La on the nature and distribution of

NieAl for CH4 decomposition. The results indicated that theNi

particle size and its interaction with support played a major

role on the hydrogen yields and longevity of the catalyst.

Based on the obtained data, carbon growth mechanism is

explained and the role of metal support interaction was

discussed.

Experimental

Preparation of NieAleLa hydrotalcite precursors

The Ni:Al:La mixed oxides are obtained from their respective

hydrotalcite (HT) precursors upon calcination in air at 550�C/5 h. The Ni:Al:La HT precursors were prepared by a simple co-

precipitation method. In a typical procedure, solution A con-

taining a mixture of metal nitrates with constant Ni and

varied Al:La mole ratios; solution B containing a base mixture

of 1:1 volume of 2M NaOH and 1M Na2CO3 were added slowly

under vigorous stirring while maintaining a constant pH ~8.5

throughout the addition. Thus produced precipitate is thor-

oughly washed with distilled water until the pH reached to ~7

followed by drying at 100 �C for overnight and subsequently

calcined in static air at 550 �C/5 h. The nominal and estimated

compositions of the Ni:Al:La samples are reported in Table 1.

Characterization of catalysts

The experimental conditions for BET-surface area, powder X-

ray diffraction (XRD), Raman spectroscopy, scanning electron

microscope-energy dispersive X-ray (SEM-EDX) and trans-

mission electron microscopy (TEM) analyses were similar as

reported by us earlier [14,18]. The XRD analysis of Ni:Al:La

samples are carried out on a Rigaku Miniflex X-ray diffrac-

tometer using Ni filtered Cu Ka radiation (l¼ 0.15406 nm) from

2q ¼ 10e80� at a scan rate of 2�min�1. The specific surface

areas of the calcined samples are calculated applying the BET

method by N2 adsorption at liquid N2 temperature in an

Autosorb 3000 physical adsorption apparatus. H2-TPR and O2-

pulse chemisorptions were carried out in a quartz micro-

reactor interfaced to a GC equipped with a thermal conduc-

tivity detector (TCD). The H2 uptakes are measured by a cali-

bration TPR profile of Ag2O. Prior to O2 pulse chemisorption,

the sample (~0.10 g) reduced at 550 �C for 2 h in 5%H2/Ar

stream and flushed in helium at 550 �C/1 h, subsequently, the

sample was cooled to 260 �C in helium, out gassed at this

temperature for 30 min, followed by titration with 5.01% O2

balance helium at 260 �C. The O2 uptakes were calculated

assuming the formation of surface NiO phase [27,28]. The

carbon contents in deactivated samples were measured using

a VARIO EL, CHNS analyser. The SEM-EDX of fresh and deac-

tivated catalysts is recorded using a JEOL-JSM 5600 instru-

ment. Transmission electron microscopy (TEM) is performed

with a JEOL JEM 2010 transmission electron microscope. The

UVeVis spectra measured at room temperature using a UV-

2000, Shimadzu Spectrophotometer equipped with a diffuse

reflectance attachment with an integrating sphere containing

BaSO4 as a reference. The spectra were recorded in the range

between 190 and 800 nm with sampling interval 0.5 nm and

slit width 2 nm and the spectra were converted to Kubel-

kaeMunk function. The Raman spectra for the structural

characterization of deposited carbon in deactivated catalysts

have been acquired with a Horiba Jobin-Yvon lab ram HR

spectrometer using a laser beam excitation of l ¼ 632:81 nm.

Fig. 1 e a: XRD patterns of oven dried NieAleLa samples (a) (2:1:0), (b) (2:0.8:0.2), (c) (2:0.7:0.3), (d) (2:0.5:0.5), (e) (2:0.9:0.1) and

(f) (2:0:1) (ICDD # 15-0087). b: XRD patterns of calcined Ni:Al:La (mole ratio) (a) (2:1:0), (b) (2:0.8:0.2), (c) (2:0.7:0.3), (d) (2:0.5:0.5),

(e) (2:0.9:0.1) and (f) (2:0:1) catalysts [C-NiO (ICDD # 78-0643),B-La2O2CO3 (ICDD # 23-0320),A-Ni2O3 (ICDD # 14-0481]. c: XRD

patterns of reduced Ni:Al:La (mole ratio) (a) (2:1:0), (b) (2:0.8:0.2), (c) (2:0.7:0.3), (d) (2:0.5:0.5), (e) (2:0.9:0.1) and (f) (2:0:1)

samples [C-Ni� (ICDD # 87-0712), A-La(OH)3 (ICDD # 36-1481)].

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 en e r g y 4 0 ( 2 0 1 5 ) 3 6 3 3e3 6 4 1 3635

Activity measurements for CH4 decomposition

Catalytic decompositionof CH4 (CDM) overNi:Al:La samples are

carried out at a reaction temperature of 550 �C and atmospheric

pressure in a fixed-bed vertical quartz reactor (i.d ¼ 1.0 cm,

length ¼ 46 cm) operated in a down flow mode. Methane sup-

plied by Bhuruka gases limited (99.99%) was used directly

without further purification. The CDM is carried out at a gas

hourly space velocity (GHSV) of CH4¼ 180 h�1, using ~15e20mg

of the catalyst without any carrier gas. Prior to the reaction, the

catalyst sample is reduced at 550 �C using 5% H2 balance Ar for

2 h. The CDM reaction was continued until the catalysts were

completely deactivated (CH4 conversions below 1%). The

outflowgaswasanalysedbyVarianCP-3800gaschromatograph

equippedwith a carboxen columnandTCDdetector usingN2 as

acarriergas.Theconcentrationofmethaneandhydrogen inthe

product streamwas calculated using a calibrated data. The CH4

conversions and H2 formation was monitored continuously at

regular intervals using the online micro reactor. The out let

streamcontained hydrogen and the unreactedmethane andno

products due to COx (x ¼ 1 and/or 2) is observed. The calcined,

reduced and some of the deactivated catalysts were character-

ized by adsorption and spectroscopic methods.

Results

Powder XRD analysis of Ni:Al:La samples

Fig. 1a shows XRD patterns of oven dried NieAleLa samples.

The diffraction lines corresponding to Ni6Al2(OH)16(CO3.OH)

4H2O phase defined to be hydrotalcite like structure is

observed in all the samples (ICDD # 15-0087) and the main

signal at 2q ¼ 11.5� is very weak in NieLa compared to

NieAleLa samples. The intensities of the diffraction lines are

decreasedwith increase in La loading. This could be explained

by the unusual cationic size of La3þ in order to form a brucite

like structure comparable with Mg2þ. The peak positions and

their relative intensities due to NiO, La(OH)3 and La2O2CO3

phases in the calcined Ni:Al:La samples are consistent with

standard ICDD. The Ni:Al:La with constant Ni and varied Al/La

ratios showed diffraction lines at 2q ¼ 36.8�, 43.3�, 62.7� and

74.8� (Fig. 1b) that are attributed to NiO phase [ICDD # 01-1239].

No peaks corresponding to lanthanum oxides is observed up

to a Al:La mole ratio ¼ 0.7:0.3 indicating that either finely

mixed with NieAl or their crystallite size may be below the X-

ray detection limit. On the other hand, diffraction lines due to

La2O2CO3 along with NiO phases are present at higher La

loadings i.e. Ni:Al:La (2:0.5:0.5), Ni:Al:La (2:0.1:0.9) and Ni:Al:La

(2:0:1) samples. Upon increasing the La loading the NiO peak

intensity is decreased and La2O2CO3 phase (ICDD # 23-0320)

emerged as sharp diffraction lines. Interestingly a well

resolved peak due to Ni2O3 phase at 2q ¼ 44.8�, 31.9�, 51.6� is

also appeared at higher La loadings i.e. Ni:Al:La (2:0.5:0.5),

Ni:Al:La (2:0.1:0.9) and Ni:Al:La (2:0:1) samples. These results

are in good agreement with earlier report on Ni/La2O3/Al2O3

catalysts [29]. The diffraction peaks due to Al2O3 phase is ab-

sent in all the samples. Fig. 1c represents the XRD patterns of

reduced Ni:Al:La catalysts. It shows that the NiO species dis-

appeared in Fig. 1c, showing that all the NiO species are

reduced. The reduced samples showed metallic Ni at

2q ¼ 44.49�, 51.85� and 76.38� [ICDD # 87-0712] along with

La(OH)3 (ICDD # 36-1481) phase. Thus suggesting the La2O2CO3

species are reduced to form La(OH)3 phase.

H2-TPR analysis of fresh calcined Ni:Al:La samples

H2-TPR profiles of the catalysts with varying Al/La ratio are

shown in Fig. 2 and the corresponding H2 uptakes are reported

in Table 1. The TPR analysis indicates the degree of metallic

phase present after activation and heat treatment of the cat-

alysts. The shape of TPR curve assigns the information about

the nature of reduciblemetal oxide species. The NieAl pattern

shows a single broad peak centred at 630 �C, which is due to

the reduction of bulk NiO species [29e31]. The profiles of (b) to

(d) of Fig. 2 show broad signals with shoulders indicating the

Fig. 2 e TPR profiles of calcined Ni:Al:La (mole ratio) (a)

(2:1:0), (b) (2:0.8:0.2), (c) (2:0.7:0.3), (d) (2:0.5:0.5), (e) (2:0.9:0.1)

and (f) (2:0:1) catalysts.

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 4 0 ( 2 0 1 5 ) 3 6 3 3e3 6 4 13636

two stage reduction behaviour. Initially at (Ni:Al:La ¼ 2:0.8:0.2

and/or 2:0.7:0.3), a high temperature reduction peak has Tmax

between 670 and 698 �C, which corresponds to reduction of

Ni2þ interacted with both Al and La. The low temperature re-

gion shoulder peak can be assigned to Ni2þ interaction with La

as the NieLa sample (Fig. 2f) showed strong reduction peak at

very low temperatures and the NieAl demonstrated high

temperature reduction of Ni2þ species. The Ni:Al:La ¼ 2:0.5:0.5

sample showed Ni2þ reduction signals typically in between

NieAl and NieLa samples. A strong reduction centred at a

Tmax of 475 �C with small peak around 670 �C in

Ni:Al:La ¼ 2:0.1:0.9 and reduction maxima at 412 �C with

second peak at 625 �C is observed on Ni:Al:La¼ 2:0:1 sample. It

cannot be ruled out that these low temperature reduction

peaks are pertaining to Ni2O3 species. The shift in Tmax to-

wards high temperatures with decrease in Al/La ratio of upto

Ni:Al:La ¼ 2:0.7:0.3 could be possibly due to strong interaction

between alumina. Further decrease in Al/La ratio led to Ni2þ

species weakly interacted mostly with lanthanum oxide.

These results manifests that certain loading of La to NieAl

dramatically changed the nickel and alumina interactions.

Shift in Tmax towards high temperature is an indication of

strong interaction of nickel particles with alumina; conversely

reduction occurring at low temperatures is attributed to nickel

interaction with La particles.

UV-DRS analysis of calcined and reduced NieAleLa samples

The diffuse reflectance ultravioletevisible (UV-DRS) spectra of

the NieAleLa fresh catalysts are reported in Fig. 3a. The ab-

sorption band position was determined by the first-derivative

of absorption bands. It can be seen that all catalysts exhibited

strong absorption band in the range of 250 to 370 nm. This

band is due to charge transfer transitions from O 2p to metal.

By increasing La content a strong and broad band grows in the

region 200 to 370 nm. Compared with catalyst NieLa (280 nm)

and NieAl (281 nm), the absorption bands of Ni:Al:La catalysts

in the spectra were all blue-shifted with the decrease of La

content, and the minimum absorption band at 255 nm was

found at lower Al:La mole ratios i.e. ¼ 0.8:0.2 and/or ¼ 0.7:0.3

indicating that there was a weak interaction between Ni and

La species [32]. Catalysts with Al:La ratio above 0.5:0.5 also

exhibited the same absorption band position at 280 nm. It is

well known that the position of the absorption edge of semi-

conductor powders is strongly affected by particle size, shift-

ing significantly to lower wavelength by decreasing particle

size. Absorption bands in blue shifted samples at 255 nm are

indication for the formation of small sized nickel particles.

XRD analysis of NieAl and NieLa revealed the grain size of

NiO is 8.3 and 7.5 nm respectively, whereas the NieAleLa

samples have an average crystallite size of 2.5 nm. In contrast

to the fresh, the reduced samples exhibited weak absorption

bands in the region 200 to 370 nm (Fig. 3b). With increase in La

loading the absorption band intensities decreased and the

bands are red shifted in the reduced samples.

BET surface area and O2 pulse chemisorption measurements

The nominal, measured compositions and the BET surface

areas of Ni:Al:La samples are reported in Table 1. The NieAl

showed a high surface area of about 120 m2/g which is sub-

stantially decreased upon addition of La. The decrease in

surface area is quite detrimental at higher La loading

(Al:La ¼ 0.5:0.5 mol ratio). This can be possibly explained by

the formation of large cluster of segregated La2O2CO3 species

as observed from XRD analysis (Fig. 1a) at higher loadings of

La. The other possible reason is that alumina which mainly

contributes for high surface area than the lanthanum. The Ni

metal surface area (SNi) of NieAleLa catalysts estimated by O2

pulse chemisorption studies and the results are reported in

Table 1. The NieAl sample exhibited ~17.0 m2/g whereas the

NieLa (2:1) showed a very low Nimetal surface area ~2.6 m2/g.

Addition of La to NieAl increased the Ni metal surface area up

to a certain loading (i.e. Ni:Al:La ¼ 2:0.7:0.3) and found to be

38.5 m2/g. At higher loading of La, the Ni metal surface area is

drastically decreased (4.3 m2/g).

Catalytic activity measurements

Fig. 4 shows the activity during CH4 decomposition with time

examined at a reaction temperature of 550 �C. The CDM ac-

tivities were measured until the CH4 conversions reached

below 1% assuming the complete deactivation of catalysts.

The CH4 conversion on NieAl is 42% at 1 h, which is fallen to

below 1% after 15 h. The Ni:Al:La (2:0.8:0.2) and Ni:Al:La

(2:0.7:0.3) sustained little longer than the other samples.

Complete deactivation has been observed at a TOS of 50 and

60 h for the Ni:Al:La (2:0.8:0.2) and Ni: Al: La (2:0.7:0.3)

respectively. At higher La/Al ratio i.e. Ni:Al:La ¼ 2:0.1:0.9 and

Ni:Al:La ¼ 2:0:1 samples, a drastic decrease in activity is

observed. The cumulative H2 yields (Table 1) are in consistent

with CH4 conversion data. The increasing order of H2 yields

over Ni:Al:La catalysts are (2:0.7:0.3) > (2:0.8:0.2) > (2:0.5:0.5) >(2:1:0) > (2:0.1:0.9) > (2:0:1). In the comparative analysis NieAl

exhibited higher H2 yields than the NieLa. Table 1 illustrates a

direct relationship between Ni metal surface area and the

CDM activity with respect to H2 yields. The high CDM activity

of Ni:Al:La (2:0.7:0.3) may be explained by high Ni metal sur-

face area of the catalyst [33,34].

Fig. 3 e a: UV-Vis diffuse reflectance spectra of calcined Ni:Al:La (mole ratio) (a) (2:0.8:0.2), (b) (2:0.7:0.3), (c) (2:0.5:0.5), (d)

(2:0.9:0.1), (e) (2:0:1) and (f) (2:1:0) catalysts. b: UV-Vis diffuse reflectance spectra of reduced Ni:Al:La (mole ratio) (a) (2:1:0), (b)

(2:0.8:0.2), (c) (2:0.7:0.3), (d) (2:0.5:0.5), (e) (2:0.9:0.1) and (f) (2:0:1) catalysts.

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 en e r g y 4 0 ( 2 0 1 5 ) 3 6 3 3e3 6 4 1 3637

Characterization of deactivated catalysts

XRD analysis of deactivated Ni:Al:La catalystsFig. 5 shows the XRD patterns of deactivated Ni:Al:La cata-

lysts. The diffraction signals corresponding to graphitic car-

bon at 2q ¼ 26.28, 45.2, 53.9 and 77� and metallic Ni phase are

observed [13,20]. At lower La loadings graphite carbon and

metallic Ni phases are found. In the comparative analysis the

NieAl catalyst exhibited higher carbon yield than the NieLa

sample. High intensity of XRD signal for graphitic carbon (at

2q ¼ 26.28�) on Ni:Al:La (2:0.7:0.3) reveals substantial deposi-

tion of carbon, that is in good agreement with CH4 conversion

activity.

Raman spectroscopic analysis of deactivated Ni:Al:La catalystsDecomposition of CH4 onNi surface leads primarily to carbidic

carbon (Ni2C: surface carbon/a-carbon), which subsequently

transformed into graphite carbon (b-carbon) either by bulk

Fig. 4 e Methane conversions with time on stream until

complete deactivation of the catalysts. CH4 flow

rate ¼ 30 mL min¡1, reaction temperature ¼ 550 �C,catalyst weight ~15 to 20 mg.

driven or surface diffusion mechanism [35]. However, at high

reaction temperatures i.e. above 450 �C the Ni2C will be

decomposed to form Ni and carbon. Thus the formed carbon

diffuses through the Ni particle where it gets precipitated in

the form of graphitic carbon [33]. The characteristic Raman

bands at 1320 cm�1 and 1580 cm�1 are attributed to disorder

(ID) and ordered (IG) graphitic carbons respectively [36]. Fig. 6a

represents Raman spectra of the deactivated catalysts. All the

samples show presence of both ordered and disordered car-

bon. A plot of La loading against ID/IG and carbon yield is

represented in Fig. 6b. This correlation exemplifies that the

catalyst composition with an optimum loading of Al:La

ratio ¼ 0.7:0.3 exhibited higher activity than the other sam-

ples. The CDM activity on Ni surface seems to depend upon

the nature of carbon deposited on its surface. The catalyst that

deposits highly ordered carbon seems to sustain longer than

the sample that is depositedwith higher amount of disordered

Fig. 5 e XRD patterns of deactivated Ni:Al:La (mole ratio) (a)

(2:0.7:0.3), (b) (2:0.8:0.2), (c) (2:0.5:0.5), (d) (2:1:0), (e) (2:0.9:0.1)

and (f) (2:0:1) catalysts. [C-Graphitic carbon, A-metallic

Ni].

Fig. 6 e a: Raman spectra of deactivated Ni:Al:La (mole ratio) (a) (2:1:0), (b) (2:0.8:0.2), (c) (2:0.7:0.3), (d) (2:0.5:0.5), (e) (2:0.9:0.1)

and (f) (2:0:1) catalysts. b: Relationship between carbon yields against ID/IG vs La/La þ Al ratio [D-Disordered carbon, G-

Ordered carbon].

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 4 0 ( 2 0 1 5 ) 3 6 3 3e3 6 4 13638

carbon (Fig. 4). The La in close proximity to NieAl seems to

destabilise the carbidic carbon (a-carbon) as a result encap-

sulation of a-carbon on Ni surface seems to be reduced.

Therefore the rate of conversion of a-carbon into graphitic (b-

carbon) may be enhanced in the presence of lanthanum. At

higher loadings of La, the surface coverage of Ni is decreased

as observed from the Ni metal surface areas of the catalysts

(Table 1). Consequently the CH4 decomposition rates are

lowered when compared to the samples with lower La. It is

estimated that Ni:Al:La ¼ 2:0.7:0.3 showed higher number of

reducible NiO species which approximately 30% enhance-

ment against NieAl sample. According to Hernandez et al.

doping of La to CeO2, creates additional labile oxygen va-

cancies could cause relatively high mobility of bulk oxygen

species within the lattice cell [37]. These results clearly

demonstrated that certain amount of La has dramatically

changed both the number of reducible species and the Ni

metal surface area (Table 1).

SEM and TEM analysis of deactivated Ni:Al:La catalystsFig. 7a shows SEM images of deactivated catalysts. Formation of

whisker type carbon nanofibers is observed. Appearance of

bright spots is an indication of Ni particle at the tip of filamen-

tous carbon nano fibre [13,38]. TEM analysis clearly indicated

that size of Ni particle at the tip of carbon nano fibre ismore are

less similar to the size carbon nano fibre (Fig. 7b). These results

corroborate the mechanism proposed by Baker et al. in which

methane decomposition takes place on surface Ni particle and

the deposited carbon gets precipitated at the interface [33].

Discussion

The studies pertaining to CH4 conversion over Ni based

particularly La2O3 modified NieAl catalysts and their physi-

cochemical characteristics has been a topic of interest from

fundamental as well as industrial point of view. The Ni:Al:La

(2:0.7:0.3) sample exhibited higher H2 (1885 mol H2/mole Ni)

yield than the other compositions. The high activity of the

Ni:Al:La (2:0.7:0.3) is attributed to high Ni metal surface area.

The reducibility of NiO sites on Ni:Al:La (2:0.7:0.3) shifted to-

wards high temperature region (Fig. 2). Shift in Tmax to high

temperature is an indication of strong interaction between Ni

particles with support. It is quite contrast to the hydrocarbon

processing where carbon act as a catalyst poison whereas in

CH4 decomposition, carbon deposition is the determining

factor for longevity of the catalyst [22]. The deposited carbon is

in both ordered and disordered forms as can be seen from

Raman spectra of deactivated catalysts. It is reported that the

growth of the graphitic carbon either carbon fibres or nano-

tubes depends on the nature of Ni particle such as distribution

[14], crystal size and its interaction with support [31,34,39].

The process of carbon formation was not crystal clear, but

it is believed that during the decomposition process, methane

is adsorbed followed by deposition of carbon on the exposed

surface of the Ni0 catalyst. Further the diffusion of deposited

carbon occurs through the opposite face of the metal where

they crystallize in the form of continuous graphite structure.

Such graphitic carbon formation between Ni0 and support led

to detach the Ni0 with support. In such situation the catalysts

that showed continuous activity exerts growth of Ni0 at the tip

of the carbon nano fibre or tube as evidenced from SEM and

TEM analysis [22]. The deactivation of the catalyst could be

due to blocking of active catalytic sites by carbonaceous de-

posits on the catalyst surface [21]. The SEM images showed

formation of whisker type carbon nano fibres (or tubes) on the

highly active Ni:Al:La (2:0.7:0.3) sample. In the case of Ni:Al:La

(2:0.1:0.9) and Ni:Al:La (2:0:1) that showed poor hydrogen

yields, displayed formation of tiny sized grass type carbon

pillars. A drastic decrease in CDM activity of high loaded La

samples may probably be due to substantial decrease in Ni

metal surface area (Table 1). XRD patterns of the deactivated

catalysts clearly showed a sharp and strong diffraction line

corresponding to graphitic carbon over Ni:Al:La (2:0.7:0.3)

sample. The reduction behaviour of NiO is dramatically

modified by the addition of La to NieAl as was seen in TPR

patterns of Ni:Al:La catalysts. The reduction temperature of

NiO is increased upto a La loading of 2:0.7:0.3 from there

Fig. 7 e a: SEM images of deactivated Ni:Al:La (mole ratio) (a) (2:0.7:0.3), (b) (2:0.9:0.1) and (c) (2:0:1) samples. b: TEM image of

deactivated Ni:Al:La (mole ratio) (2:0.7:0.3) catalyst.

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 en e r g y 4 0 ( 2 0 1 5 ) 3 6 3 3e3 6 4 1 3639

onwards drifted to low temperatures. The H2 uptakes were

varied although constant Ni amount being used in all the

NieAleLa which is an indication of influence of La on the

reduction behaviour of NiO (Table 1).

The EDX analysis indicated that there was no loss of Ni

during the preparation of catalysts. Hence the loss in

reducibility is probably due to nickel presence in a surface

phase containing Ni, Al and La in which nickel interaction is

higher. At higher loadings of La, the Tmax shifted to very low

temperatures. This shift is attributed to lowering interaction

between NiO and support due to an increase in the interac-

tion between lanthanum and aluminium oxides, particularly

at a La loading of Ni:Al:La ¼ 2:0.7:0.3 the metal support

interaction was maximum. According to Ruckenstein and Hu

the weakly interacted NiO reduces at lower temperatures

below i.e. 500 �C [40].

The reduction profiles indicated two reduction maxima

one mainly at high temperatures (lower La) and the other one

at extremely low temperatures in presence of higher amount

of La. These results thus emphasize that the reduction of NiO

in Ni-rich Al-phase is corresponding to the former and the NiO

reduction in Ni-rich La-phase to the latter [31,40]. However,

the XRD analysis of both fresh and reduced Ni:Al:La samples

did not show NiAl2O4 and/or La containing phases such as

LaNiO3 and La2NiO4 [40]. The nickel lattice is changed by the

inclusion of La atoms which seems to be higher at

Ni:Al:La ¼ 2:0.7:0.3 composition.

Decrease in Ni metal surface areas observed while the La

loadings are increased (Table 1). This could be explained by

either formation of large nickel particles or covering of nickel

surface by lanthanum oxide species at exceedingly higher

loadings. The first reason seems correct in this case as the XRD

patterns indicated a decrease in crystallite size of NiO even at

higher La loadings. Slagtern et al. found a higher Ni dispersion

in Ni/Al2O3 when La2O3 was added [41]. Hence the increased

surface Ni area (SNi) is attributed to lanthanumoxide in a close

proximity to nickel that could disperse the Ni metal crystal-

lites. Martinez et al. observed the decrease in diffraction angle

in NieLaeAl catalysts calcined at 750 �C due to enlargement of

cell parameters by the inclusion of La atoms in Ni lattice [39].

In the present study we were unable to explain this phe-

nomenon on Ni:Al:La calcined at 550 �C since the XRD results

showed very broad diffractions lines. Nevertheless, the TPR

results clearly suggested that Ni interaction with alumina is

quite high upto Ni:Al:La ¼ 2:0.7:0.3 which gets weakening

above this composition. We believe that excess La addition

leads to blockade of surface Ni.

Nature and growth of filamentous carbon depend on the Ni

particle size and the quantity of nickel present at the catalyst

surface as too small nickel particles avoid the formation of

carbon filament [7,18]. It was also reported that the sample

sustains longer period if the rate of diffusion of carbon

through the nickel particle is higher than the rate of deposi-

tion of carbon on the surface of Ni particle. If reverse is true

the sample deactivates rapidly due to carbon deposition on

the catalyst surface [21,35]. The diffusion of carbon through

the Ni particle may be curtailed due to the presence of La in Ni

lattice. As a result of it, growth of filamentous carbon is

i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 4 0 ( 2 0 1 5 ) 3 6 3 3e3 6 4 13640

hindered, consequently encapsulation of carbon at the pore

mouth of surface Ni particle too high that leads to rapid

deactivation at higher La loadings. Contrary to this effect at

lower La loadings Ni particles interactedwith Al-oxide in close

proximity to lanthanum oxide would enhance the rate of

decomposition of a-carbon (Ni2C) to form b-carbon (graphitic)

for further growth eventually life of the catalyst was extended.

Conclusions

The mixed oxides of NieAleLa were derived from hydro-

talcite precursor exhibited methane decomposition activity.

Addition of lanthanum to NieAl has influenced the reduction

behaviour of Ni particles. The O2 chemisorption studies

revealed that certain amount of La has enhanced the Ni

metal surface of the catalyst Ni:Al:La (2:0.7:0.3). Raman

spectra of deactivated catalysts indicated the presence of

both ordered and disordered carbon. The SEM analysis of

deactivated catalysts presented the formation of carbon

nano fibres. TEM image of deactivated sample displayed that

the size Ni particle is similar to the size of carbon nano fibre.

In the comparative analysis the NieAl exhibited higher

hydrogen yields than NieLa. An optimized La with a

composition of Ni:Al:La ¼ 2:0.7:0.3 demonstrated good to

excellent hydrogen yields. Finally it can be summarized that

an optimum amount of La addition to NieAl has enhanced

the production of COx free hydrogen.

Acknowledgement

The authors CA thank CSIR New Delhi, VVK and GN thanks

RMIT Australia for the award of fellowships. AV and SNK

thanks Dr M Lakshmi Kantam and Dr K S Rama Rao for their

constant help and encouragement.

Appendix A. Supplementary data

Supplementary data related to this article can be found at

http://dx.doi.org/10.1016/j.ijhydene.2015.01.072.

r e f e r e n c e s

[1] Muradov NZ, Veziroǧlu TN. From hydrocarbon tohydrogenecarbon to hydrogen economy. Int J HydrogenEnergy 2005;30:225e37.

[2] Muradov NZ. How to produce hydrogen from fossil fuelswithout CO2 emission. Int J Hydrogen Energy 1993;18:211e5.

[3] Das D, Veziroǧlu TN. Hydrogen production by biologicalprocesses: a survey of literature. Int J Hydrogen Energy2001;26:13e28.

[4] Muroyama H, Nakase R, Matsui T, Eguchi K. Ethanol steamreforming over Ni-based spinel oxide. Int J Hydrogen Energy2010;35:1575e81.

[5] Choudhary TV, Goodman DW. CO-free fuel processing forfuel cell applications. Catal Today 2002;77:65e78.

[6] Choudhary TV, Aksoylu E, Wayne Goodman D. Non-oxidative activation of methane. Catal Rev 2003;45:151e203.

[7] Venugopal A, Naveen Kumar S, Ashok J, Hari Prasad D, DurgaKumari V, Prasad KBS, et al. Hydrogen production bycatalytic decomposition of methane over. Int J HydrogenEnergy 2007;32:1782e8.

[8] Takenaka S, Kobayashi S, Ogihara H, Otsuka K. Ni/SiO2

catalyst effective for methane decomposition into hydrogenand carbon nanofiber. J Catal 2003;217:79e87.

[9] De Jong KP, Geus JW. Carbon nanofibers: catalytic synthesisand applications. Catal Rev 2000;42:481e510.

[10] Dai H, Hafner JH, Rinzler AG, Colbert DT, Smalley RE.Nanotubes as nanoprobes in scanning probe microscopy.Nature 1996;384:147e50.

[11] Dai H, Kong J, Zhou C, Franklin N, Tombler T, Fan S.Handbook of nanophase and nanostructured materials. 2003.p. 96.

[12] Ashok J, Naveen Kumar S, Venugopal A, Durga Kumari V,Subrahmanyam M. COX-free H2 production via catalyticdecomposition of CH4 over Ni supported on zeolite catalysts.J Power Sources 2007;164:809e14.

[13] Ashok J, Naveen Kumar S, Subrahmanyam M, Venugopal A.Production of COX free hydrogen by catalytic decompositionof methane over Ni/HY catalysts. Catal Lett 2007;118:139e45.

[14] Ashok J, Kumar SN, SubrahmanyamM, Venugopal A. Pure H2

production by decomposition of methane over Ni supportedon hydroxyapatite catalysts. Catal Lett 2008;121:283e90.

[15] Ermakova MA, Ermakov DY. Ni/SiO2 and Fe/SiO2 catalysts forproduction of hydrogen and filamentous carbon via methanedecomposition. Catal Today 2002;77:225e35.

[16] Li X, Zhang Y, Smith KJ. Metalesupport interaction effects onthe growth of filamentous carbon over Co/SiO2 catalysts.Appl Catal A Gen 2004;264(1):81e91.

[17] Reshetenko TV, Avdeeva LB, Ismagilov ZR, Chuvilin AR,Ushakov VA. Carbon capacious Ni-Cu-Al2O3 catalysts forhigh-temperature methane decomposition. Appl Catal A Gen2003;247(1):51e63.

[18] Ashok J, Subrahmanyam M, Venugopal A. Hydrotalcitestructure derived NieCueAl catalysts for the production ofH2 by CH4 decomposition. Int J Hydrogen Energy2008;33(11):2704e13.

[19] Ashok J, Reddy PS, Raju G, Subrahmanyam M, Venugopal A.Catalytic decomposition of methane to hydrogen and carbonnanofibers over Ni�Cu�SiO2 catalysts. Energy Fuels2009;23:5e13.

[20] Takenaka S, Shigeta Y, Tanabe E, Otsuka K. Methanedecomposition into hydrogen and carbon nanofibers oversupported PdeNi catalysts. J Catal 2003;220(2):468e77.

[21] Ashok J, Raju G, Reddy PS, Subrahmanyam M, Venugopal A.Catalytic decomposition of CH4 over NiOeAl2O3eSiO2

catalysts: Influence of catalyst preparation conditions on theproduction of H2. Int J Hydrogen Energy 2008;33(18):4809e18.

[22] Suelves I, L�azaro MJ, Moliner R, Echegoyen Y, Palacios JM.Characterization of NiAl and NiCuAl catalysts prepared bydifferent methods for hydrogen production by thermocatalytic decomposition of methane. Catal Today2006;116(3):271e80.

[23] Echegoyen Y, Suelves I, L�azaro MJ, Moliner R, Palacios JM.Hydrogen production by thermocatalytic decomposition ofmethane over Ni-Al and Ni-Cu-Al catalysts: effect ofcalcination temperature. J Power Sources 2007;169:150e7.

[24] L�azaro MJ, Echegoyen Y, Suelves I, Palacios JM, Moliner R.Decomposition of methane over Ni-SiO2 and Ni-Cu-SiO2

catalysts: effect of catalyst preparation method. Appl Catal AGen 2007;329:22e9.

[25] Pinilla JL, Moliner R, Suelves I, L�azaro MJ, Echegoyen Y,Palacios JM. Production of hydrogen and carbon nanofibersby thermal decomposition of methane using metal catalysts

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 en e r g y 4 0 ( 2 0 1 5 ) 3 6 3 3e3 6 4 1 3641

in a fluidized bed reactor. Int J Hydrogen Energy2007;32(18):4821e9.

[26] Sania M de L, Adriana M da S, Lı́dia OO da C, Jos�e MA,Lisiane VM, Reema S, et al. Hydrogen production throughoxidative steam reforming of ethanol over Ni-based catalystsderived from La1-xCexNiO3 perovskite-type oxides. Appl CatalB Env 2012;121:1e9.

[27] Stytsenko VD. Surface modified bimetallic catalysts:preparation, characterization, and application. Appl Catal AGen 1995;126(1):1e26.

[28] Rynkowski JM, Paryjczak T, Lenik M. On the nature of oxidicnickel phases in NiO/g-Al2O3 catalysts. Appl Catal A Gen1993;106:73e82.

[29] Sanchez-Sanchez MC, Navarro RM, Fierro JLG. Ethanol steamreforming over Ni/LaeAl2O3 catalysts: Influence oflanthanum loading. Catal Today 2007;129:336e45.

[30] Yuehua C, Huidong Z, Hengyong X, Wenzhao L. On thenature of oxidic nickel phases in NiO/g-Al2O3 catalysts. ApplCatal A Gen 2007;331:60e9.

[31] Zhang J, Xu H, Jin X, Ge Q, Li W. Characterizations andactivities of the nano-sized Ni/Al2O3 and Ni/LaeAl2O3

catalysts for NH3 decomposition. Appl Catal A Gen2005;290:87e96.

[32] Garbarino G, Campodonico S, Perez AR, Carnasciali MM,Riani P, Finocchio E, et al. Spectroscopic characterization ofNi/Al2O3 catalytic materials for the steam reforming ofrenewable. Appl Catal A Gen 2013;452:163e73.

[33] Baker RTK. Catalytic growth of carbon filaments. Carbon1989;27:315e23.

[34] Baker RTK, Barber MA, Harris PS, Feates FS, Waite RJ.Nucleation and growth of carbon deposits from the nickelcatalyzed decomposition of acetylene. J Catal 1972;26:51e62.

[35] Dost AA, Dhanak VR, Buckingham S. XPS study of b-carbonhydrogenation on nickel (l11) surface. J Catal1984;89:159e63.

[36] Anajenyulu Chatla, Vijay Kumar Velisoju, Bhargava SureshK, Venugopal Akula. Characteristics of La-modified Ni-Al2O3

and NieSiO2 catalysts for COx-free hydrogen production bycatalytic decomposition of methane. J Energy Chem2013;22:853e60.

[37] Hernandez WY, Laguna OH, Centeno MA, Odriozola JA.Structural and catalytic properties of lanthanide (La, Eu, Gd)doped ceria. J Solid State Chem 2011;184:3014e20.

[38] Rostrup-Nielsen JR. Teknisk Forlag A/S. 1975 [Copenhagen].[39] Martinez R, Romero E, Guimon C, Bilbao R. CO2 reforming of

methane over co-precipitated NieAl catalysts modified withlanthanum. Appl Catal A Gen 2004;274:139e49.

[40] Ruckentsein E, Hu YH. Interactions between Ni and La2O3 inNi/La2O3 catalysts prepared using different Ni precursors. JCatal 1996;161:55e61.

[41] Slagtern A, Olsbye U, Blom R, Dahl IM, Fjellvag H.Characterization of Ni on La modified Al2O3 catalysts duringCO2 reforming of methane. Appl Catal A Gen1997;165:379e90.