Catalytic effect of mixed Zr-Fe additives on the hydrogen desorption kinetics of Mg H2

Article abstract of DOI:10.1063/1.2840180

Mg films containing Fe-Zr mixed additives were prepared by rf sputtering and their hydrogen sorption kinetic was studied by Sievert technique. We observed that mixed Fe-Zr additives improve the H2 desorption kinetics better than singly Fe or Zr. X-ray diffraction analysis evidences that mixed additives optimize the Fe distribution that forms smaller clusters than in Mg sample with single Fe catalyst. The presence of mixed additives, with Fe having larger diffusivity than Zr, contributes to maintain the atomic Zr dispersion in the Mg matrix, thus, favoring the presence of a greater number of nucleation sites for Mg H2 to Mg transformation.

Full text of DOI:10.1063/1.2840180

Catalytic effect of mixed Zr–Fe additives on the hydrogen desorption kinetics of Mg H 2
Nicola Bazzanella, Riccardo Checchetto, and Antonio Miotello
Citation: Applied Physics Letters 92, 051910 (2008); doi: 10.1063/1.2840180
View online: http://dx.doi.org/10.1063/1.2840180
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/92/5?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Effect of transition-metal additives on hydrogen desorption kinetics of MgH2
Appl. Phys. Lett. 102, 033902 (2013); 10.1063/1.4788746
Study on the physicochemical properties of Fe / CeO 2 catalysts as an effect from different iron loading
AIP Conf. Proc. 1502, 288 (2012); 10.1063/1.4769152
Bimetallic Fe–V catalyzed magnesium films exhibiting rapid and cycleable hydrogenation at 200 ° C
Appl. Phys. Lett. 96, 013108 (2010); 10.1063/1.3275864
Synergy on catalytic effect of Fe–Zr additives mixed in different proportions on the hydrogen desorption from
MgH 2
Appl. Phys. Lett. 94, 204103 (2009); 10.1063/1.3141453
Catalytic effect on hydrogen desorption in Nb-doped microcrystalline Mg H 2
Appl. Phys. Lett. 85, 5212 (2004); 10.1063/1.1829155
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
169.230.243.252 On: Wed, 03 Dec 2014 14:51:38

APPLIED PHYSICS LETTERS 92, 051910 ͑2008͒
Catalytic effect of mixed Zr–Fe additives on the hydrogen desorption
kinetics of MgH2
a͒
Nicola Bazzanella, Riccardo Checchetto, and Antonio Miotello
Dipartimento di Fisica, Università degli Studi di Trento, I-38050 Povo (Trento), Italy
͑
Received 19 December 2007; accepted 14 January 2008; published online 7 February 2008͒
Mg films containing Fe–Zr mixed additives were prepared by rf sputtering and their hydrogen
sorption kinetic was studied by Sievert technique. We observed that mixed Fe–Zr additives improve
the H desorption kinetics better than singly Fe or Zr. X-ray diffraction analysis evidences that
2
mixed additives optimize the Fe distribution that forms smaller clusters than in Mg sample with
single Fe catalyst. The presence of mixed additives, with Fe having larger diffusivity than Zr,
contributes to maintain the atomic Zr dispersion in the Mg matrix, thus, favoring the presence of a
greater number of nucleation sites for MgH to Mg transformation. © 2008 American Institute of
2
Physics. ͓DOI: 10.1063/1.2840180͔
Despite the large number of papers published in the last
years on the hydrogen sorption kinetics in Mg based materi-
als, there is not a clear understanding of the catalysis mecha-
nisms due to the presence of transition metal ͑TM͒ or
transition metal oxides ͑TMOs͒ nanoclusters. Recent experi-
ments seem to support the idea that, in nanocrystalline ball-
milled magnesium, TM and TMO nanoclusters promote the
formation of a destabilized MgH2−␦ phase. Because of the
abundance of H vacancies in this phase, the hydrogen diffu-
sion results accelerated: this phase, thus, acts as gateway
through which hydrogen enters into the Mg grains. This sub-
stoichiometric phase is supposed to grow in presence of the
TMO additive during the ball-milling process because the
TMO additive destabilizes the magnesium hydride in regions
nisms͒ and activation energy as that of the pure MgH2
samples but a shorter nucleation time of the h-Mg phase. We,
thus, suggested that Nb atoms dispersed in the Mg matrix act
as nucleation centers for the h-Mg phase in the parent MgH2
phase.
In this paper, we present a study on the hydrogen desorp-
tion from magnesium hydride samples containing mixed me-
tallic additives, Fe and Zr, at atomic percent level. As previ-
ously discussed, the homogeneous distribution of the catalyst
in form of few atoms aggregates results more efficient in
accelerating the hydrogen desorption from MgH than the
2
5
situation in which large clusters are formed. The additive
mixing studied in this paper will represent a route to limit the
aggregation of one or of both additives in form of large clus-
ters, thus, favoring a better distribution of catalysts ͑small
clusters or a few atoms aggregates͒ into the whole volume of
the magnesium hydride. To this purpose, we have used Zr
close to the MgH -oxide interfaces by the formation of
2
1
mixed Mg-TM oxides. When the additive is TM, these
mixed oxide regions result from mechanical alloying of the
TM nanoclusters and the MgO surface layer of the Mg
and Fe because they have catalytic effects when dispersed in
2
6
particles.
form of nanocluster in the MgH matrix and because they
2
7
In previous papers, by studying samples prepared by rf
sputtering, we observed that the presence of TM nanoclusters
do not form binary phases with Mg matrix. Finally Zr is
heavier than Fe and has larger atomic diameter, 4.32 versus
3.44 Å. They, thus, have different mobility values and we,
thus, expect different clustering processes when sharing the
same host matrix ͑Mg͒ in the sense that mainly the faster
atoms ͑Fe in the present case͒ will make use of the dispos-
able ͑defects͒ nucleation centers.
dispersed in the MgH matrix accelerates the H desorption
2
3
kinetics and there was no evidence of any oxide phase, as
studied by x-ray diffraction analysis ͑XRD͒, transmission
electron microscopy, and extended x-rays absorption fine
4
structure ͑EXAFS͒ analysis. We suggested that TM nano-
clusters act as preferential sites for h-Mg ͑hexagonal͒ nucle-
ation and promote the formation of an interconnected net-
work of h-Mg domains, where H diffusion preferentially
Mg films with thickness in the 10–20 ␮m range were
synthesized on polished graphite ͑g-C͒ substrates by rf mag-
netron sputtering using high purity Ar ͑99.999%͒ as working
gas. The addition of the metallic elements to the Mg samples
was obtained by putting over the Mg target small Zr and Fe
fragments. The additive concentration was measured after
deposition by energy dispersion spectroscopy ͑EDS͒ analy-
sis. The Mg film samples were coated with a 15 nm Pd cap-
ping layer using a Pd target ͑99.95% purity͒ without inter-
rupting the vacuum conditions. Details on the deposition
3
,4
occurs. Two additional effects were also evidenced by the
experiments. First, in Mg samples with Nb content in the
at. % level, the hydrogen desorption kinetics of as prepared
samples, where Nb is not clusterized, was faster than that
observed in the fully activated samples where Nb clusters are
5
present. Second, catalytic effects were also observed when
the metallic additive was present at levels lower than
5
0
.1 at. % in the Mg matrix, that is close to the solubility
3,8
procedure are described elsewhere. The Mg films peeled
limit of Nb in Mg. The previous evidences indicate that the
additive can play a catalytic role also when atomically dis-
off from g-C substrate immediately after exposure to air:
self-supporting film samples were then inserted in a Sievert-
5
persed in the hydride matrix. The desorption curves pre-
9
type apparatus to study the hydrogen absorption and desorp-
sented the same kinetics ͑nucleation and growth mecha-
tion kinetics. The activation procedure consisted on repeated
absorption and desorption cycles at 623 K and 1.000 ͑ab-
a͒Electronic mail: bazzanel@science.unitn.it.
sorption͒ or 0.015 ͑desorption͒ MPa H gas pressure. The
2
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
003-6951/2008/92͑5͒/051910/3/$23.00 92, 051910-1 © 2008 American Institute of Physics
69.230.243.252 On: Wed, 03 Dec 2014 14:51:38
0
1

051910-2
Bazzanella, Checchetto, and Miotello
Appl. Phys. Lett. 92, 051910 ͑2008͒
FIG. 1. Hydrogen desorbed fraction vs time at various temperatures for the
Mg sample with mixed Fe–Zr additives at ͑ᮀ͒ 623 K, ͑᭺͒ 613 K, ͑᭝͒
FIG. 2. Hydrogen desorbed fraction vs time at 623 K of MgH sample with
2
Zr ͑᭝͒, Fe ͑᭺͒, and Nb ͑छ͒ single additives and mixed Fe–Zr additives
6
03 K, ͑छ͒ 593 K, ͑૽͒ 583 K, and ͑᭟͒ 573 K. In the inset, we present the
͑
ᮀ͒. The lines are only a guide for the eyes.
Arrhenius plot of the desorption rate constant k. The straight line is the least
square fit.
than 2 at. %͒ the phase transformation has diffusive charac-
ter, as shown by the different values of the reaction order n
structural characterization of the samples was carried out by
x-ray diffraction analysis ͑XRD͒ in conventional Bragg-
Brentano configuration ͑␪-2␪͒ using the Cu K␣ radiation
and of activation energy E . The most important result is the
a
fact that the mixed additives induce faster desorption kinetic,
as compared to pure MgH2 and to MgH2 with Fe or Nb as
single additive, as shown by the ␶1/2 parameter, defined as
the interval time required to transform 50% of MgH2 to Mg.
This is also shown by Fig. 2. Here, we present the H2 de-
sorption curves at 623 K of MgH2 with single catalyst, curve
͑᭺͒ for Fe, ͑᭝͒ for Zr, ͑छ͒ for Nb, and curve ͑ᮀ͒ for MgH2
with mixed additive. Table I also indicates that the hydride
decomposition kinetic with mixed additives is very similar to
that of the sample with 1 at. % Nb described in Ref. 5: the
order of reaction is the same and the activation energy values
are very similar. The much better catalytic effect of the
mixed additives is related to the impressive increase of pre-
exponential factor ͑A͒ of the desorption rate constant ͑k͒
͑
␭=1.5414 Å͒.
SEM and XRD analysis of the as-deposited film ͑not
reported here͒ showed a columnar growth of magnesium film
with no evidence of metallic Fe and Zr. This means that the
presence of the additive is in atomic form or in form of few
atoms aggregates, as previously observed with Mg samples
4
containing the Nb additive. In the present experiment, we
used samples with ͑i͒ mixed additive content of ϳ5 and
ϳ2 at. % for Fe and Zr, respectively, ͑ii͒ Fe single additive
5
at. %, and ͑iii͒ Zr single additive 3 at. %. We have used
additives with at. % concentration because such additive
content gives rise to very fast hydrogen desorption kinetics
on fully activated MgH samples and a Fe content larger than
2
6
4
8
−1
that of Zr because Fe works better as catalyst.
growing from ϳ10 to ϳ10 s . We remark that the im-
proved kinetics compared to that of the sample with Fe or Zr
as single additive cannot be attributed to the larger additive
content because there is a saturation effect in the desorption
In Fig. 1, we report the H desorption curves pertinent to
2
fully activated MgH samples with Fe–Zr mixed additives.
2
The inset shows the Arrhenius plot of the desorption
rate constant ͑k͒ as evaluated by curve fitting procedure
using Johnson-Mehl-Avrami theory for solid state phase
5
catalysis at additive contents of ϳ2 at. %.
In Fig. 3, we present the XRD spectrum of a fully acti-
1
0
transitions. In Table I, we report the kinetic parameters per-
tinent to film samples containing the mixed additives and, as
vated sample after H desorption that provides information
2
on the physical state of the Fe and Zr additives. In the spec-
trum, we observe reflection peaks pertinent to Mg and the
͑110͒, ͑200͒, and ͑211͒ Bragg reflections of the ␣-Fe phase
but no XRD peak attributable to Zr nor to any Fe–Zr phase.
Because Zr and Fe as single additive cannot produce the fast
desorption kinetics shown in Fig. 1, we conclude that the
improved catalytic effect of the mixed additives has two
causes: ͑i͒ aggregation of Fe into nanoclusters and ͑ii͒ atomic
dispersion of Zr ͑that could also form aggregates but consist-
ing of few atoms͒. We remark that the second effect has
permanent character because, contrary to the samples with
single additive, no evidence of Zr clustering is observed also
reference, those pertinent to samples with a single additive
3
and to pure MgH . We note that, in presence of mixed ad-
2
ditives, the hydride to metal phase transition obeys to a
nucleation-and-growth mechanism while in presence of a
single catalyst ͑Fe or Nb, with atomic concentration greater
TABLE I. Evaluated values of the half reaction time ͑␶1/2͒ at T=623 K,
order of reaction ͑n͒, activation energy ͑E ͒, and preexponential factor ͑A͒
a
of pure MgH and samples with different catalysts ͑Nb, Fe, and ZrFe͒.
2
␶1
Ea
A
͑s−1͒
/2
1
Sample
͑s͒
n
͑kJ mol− H͒
after repeated H sorption cycles.
2
The size of the ␣-Fe crystallites is determined to be
8
Pure Mg
3250Ϯ50
110Ϯ10
300Ϯ10
100Ϯ10
4
141Ϯ5
51Ϯ5
78Ϯ5
51Ϯ5
ϳ10
ϳ10
ϳ10
ϳ10
lower than 20 nm by employing the Williamson–Hall
2
4
2
8
Mg+5 at. %Nb
Mg+1 at. %Nb
Mg+5 at. %Fe
1
11
method according to the broadening of the ͑110͒, ͑200͒,
1.5
1
and ͑211͒ Bragg reflection peaks in Fig. 3. We can also
−
4
evaluate a low microstrain value of ␧ϳ6ϫ10 not observed
Mg+͑Fe–Zr͒
70Ϯ10
1.5
94Ϯ5
ϳ10
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
in the Mg sample with single Fe additive. The intensity ratio
1
69.230.243.252 On: Wed, 03 Dec 2014 14:51:38

051910-3
Bazzanella, Checchetto, and Miotello
Appl. Phys. Lett. 92, 051910 ͑2008͒
terms of an increased Fe nanocluster density. Because Fe
nanoclusters act as nucleation centers for the h-Mg phase in
the hydride to metal phase transition, the hydrogen desorp-
3
–6
tion kinetics results, thus, accelerated. In addition, we may
also suggest that in the present samples, the Zr atoms also
can act as stable and distributed active centers for h-Mg
nucleation in the ␤-MgH matrix during desorption, as ob-
2
served in MgH samples with Nb additive atomically
2
5
dispersed. Finally, the very high value of the preexponential
factor ͑A͒ which contributes to increase the desorption kinet-
ics should be explained.
Preliminary results showed that by lowering the Fe con-
centration while increasing the Zr one, then the kinetics of
the hydrogen desorption tends to decrease. This proves that
there is an optimum Fe and Zr atomic concentration which
establishes the best kinetics conditions for cluster-assisted
hydrogen desorption.
To gain a better understanding of the chemical state of
the Zr additive and, thus, of the catalytic process in the
mixed additive samples, we plan to carry out EXAFS analy-
sis to analyze the atomic environment of the Zr atoms.
In conclusion, we proved that mixed additive may con-
stitute a new route to improve hydrogen kinetics in magne-
sium hydrides by increasing the uniform distribution of
nucleation sites for the growing of the h-Mg phase during
hydrogen desorption.
FIG. 3. XRD spectra pertinent to Mg samples with mixed Fe–Zr additives
completely activated͒ after H desorption. In the figure, only the Fe reflec-
tion peaks are labeled and all others peaks are pertinent to h-Mg reflections.
In the inset we present the Fe ͑110͒ reflection peak pertinent to a sample
with only the Fe additive, line ͑a͒, and to a sample with mixed Fe–Zr
additives, line ͑b͒. The thick line is the reference position of the Fe ͑110͒
peak ͑Ref. 12͒.
͑
2
of these diffraction peaks indicates that the ␣-Fe nanoclus-
ters have a random distribution in the Mg matrix. The ␣-Fe
clusters size is lower than in Mg samples with single Fe
additive, ϳ30 nm, as evaluated by the Bragg–Brentano
analysis of the ͑110͒ reflection peak presented in the inset of
Fig. 3, line ͑a͒ ͑this is the only ␣-Fe reflection in the XRD
spectrum͒. Both samples have the same Fe content, ϳ5 at. %
and, assuming that all Fe atoms form precipitates, we con-
clude that the sample with mixed additives presents a higher
Fe cluster density.
We thank Cristina Armellini for XRD analysis. We also
acknowledge Romina Belli for EDS analysis. The research
activity is financially supported by the Hydrogen-FISR Ital-
ian project.
1
The difference in the Fe clustering can be explained by
the lower mobility of Zr atoms, as compared to that of the Fe
atoms. Because the maximum solubility of Zr in Mg is lower
than the present concentration, then Zr atoms segregate in
extended defects such as grain boundaries or dislocations. In
the Mg sample with Zr as single additive, after segregation
A. Borgschulte, U. Boesenberg, G. Barkhordarian, M. Dornheim, and R.
Bormann, Catal. Today 120, 262 ͑2007͒.
H. Gijs Schimmel, J. Huot, L. C. Chapon, F. D. Tichelaar, and F. M.
2
Mulder, J. Am. Chem. Soc. 127, 14348 ͑2005͒.
N. Bazzanella, R. Checchetto, and A. Miotello, Appl. Phys. Lett. 85, 5212
3
4͑
2004͒.
R. Checchetto, N. Bazzanella, A. Miotello, C. Maurizio, F. D’Acapito, P.
Mengucci, G. Barucca, and G. Majni, Appl. Phys. Lett. 87, 061904
͑2005͒.
6
upon diffusion, Zr atoms aggregate forming clusters. In the
sample with mixed additives, these Zr clusters are not ob-
served. As shown by Fig. 3, the larger Fe mobility suggests
that a possible mechanism impeding the Zr aggregation is the
Fe clustering in the disposable nucleation centers which an-
ticipates the Zr precipitation. The lower size of the Fe clus-
ters in the Mg sample with mixed additives is, thus, conse-
quence of a larger density of available nucleation sites as
compared to that in the Mg sample with single Fe additive.
Indeed, Zr atoms also may constitute precipitation centers for
Fe atoms.
5
6
7
8
9
N. Bazzanella, R. Checchetto, A. Miotello, C. Sada, P. Mazzoldi, and P.
Mengucci, Appl. Phys. Lett. 89, 014101 ͑2006͒.
R. Checchetto, N. Bazzanella, A. Miotello, and P. Mengucci, J. Alloys
Compd. 446-447, 58 ͑2007͒.
Phase Diagrams of Binary Magnesium Alloys, edited by A. A. Nayeb–
Hashemi and J. B. Clark ͑ASM International, Materials Park, OH, 1998͒.
R. Checchetto, N. Bazzanella, A. Miotello, and P. Mengucci, J. Alloys
Compd. 404-406, 461 ͑2005͒.
R. Checchetto, G. Trettel, and A. Miotello, Meas. Sci. Technol. 15, 127
10͑
2004͒.
J. W. Christian, The Theory of Transformations in Metals and Alloys, Part
On the basis of the previously discussed Fe clustering
process, we suggest that the presence of mixed additives
gives rise to the optimization of the Fe cluster distribution in
I, 3rd ed. ͑Pergamon, London, 2002͒.
1
1
G. K. Williamson and W. H. Hall, Acta Metall. 1, 22 ͑1953͒.
ICDD Card No. 06-0696.
1
2
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
69.230.243.252 On: Wed, 03 Dec 2014 14:51:38
1