View Article Online / Journal Homepage / Table of Contents for this issue

PAPER

www.rsc.org/pccp | Physical Chemistry Chemical Physics

Redox behavior of small metal clusters with respect to hydrogen.

The eect of the cluster charge from density functional resultsw

Published on 29 July 2010. Downloaded by UNIVERSITY OF NEBRASKA on 15/04/2015 02:34:13.

Galina P. Petrova,a Georgi N. Vayssilov*a and Notker Rosch*b

Received 16th March 2010, Accepted 10th May 2010
DOI: 10.1039/c004377j
Tetrahedral model iridium species [Ir4Hn]q+ of dierent charge and hydrogen loading
were described at the density functional level. The energy of dissociative adsorption of hydrogen
was calculated to vary in the small interval from 63 kJ mol1 to 77 kJ mol1 (per H atom).
Adsorption of hydrogen on Ir4 and Ir4+ induces an oxidation of the metal moiety, whereas the
highly charged cluster Ir43+ is reduced upon hydrogen adsorption. The ligand shell acts as charge
buer as the metal moieties of the complexes [Ir4H12]q+ with maximum hydrogen loading carry
very similar eective charges, irrespective of the total charge q. Similar eects were conrmed
to occur on small clusters of other 4d and 5d transition metals.

1. Introduction
Small transition metal clusters in the gas phase or supported
on a metal-oxide surface have been thoroughly studied by
both experimental16 and computational techniques.1,79 Such
systems have been found useful in various catalytic or sorption
processes as supported species13 or directly in the gas phase as
neutral and charged clusters.1,4,10 The interaction of such
clusters with molecular hydrogen is particularly important,
for characterizing the properties of the clusters and their
application as catalysts.5,6 Therefore, we previously explored
computationally the dissociative adsorption of H2 from the
gas phase on isolated and zeolite-supported tetrairidium
clusters.1113 These model studies showed both supported metal
clusters and metal clusters in the gas phase to adsorb up to
12 hydride ligands, i.e. 3 H ligands per Ir atom. Hydrogenated
Ir4 clusters adsorbed on a dehydroxylated zeolite support were
calculated to be the most stable. The formal charge of such
Ir4Hn moieties is 3 e as a result of reverse proton spillover from
bridging hydroxyl groups of the zeolite fragment. Yet, calculated charges of the Ir4Hn moiety were estimated to be notably
smaller, 1.071.64 e.13 From a comparison of the results for
supported clusters with those for neutral model clusters in the
gas phase, one expects the (eective) charge of the cluster to
have a notable inuence on the properties of the hydrogenated
metal clusters.
To explore this eect of the cluster charge on the hydrogen
loading, we extended our studies on tetrairidium species to a
a

Faculty of Chemistry, University of Soa, 1126 Soa,

Bulgaria. E-mail: gnv@chem.uni-soa.bg
Department Chemie and Catalysis Research Center,
Technische Universitat Munchen, 85747 Garching, Germany.
E-mail: roesch@mytum.de
w Electronic supplementary information (ESI) available: Table with
calculated energy characteristics and interatomic distances of the
optimized structures [Ir4Hn]q+ (q = 03; n = 0, 3, 6, 9, 12); table
with various characteristics of the electronic structure of the clusters
modeled; gure showing density of state plots of the Ir 2p core levels as
a function of the cluster charge and the hydrogen loading; gure with
the optimized structures of [M4Hn]q+ species (M = Ru, Rh, Pd, Os,
Pt); gure with variations of the average shifts of the M 2p levels in the
clusters [M4H12]q+ for various charges q. See DOI: 10.1039/c004377j
b

This journal is

the Owner Societies 2010

series of cationic species [Ir4Hn]q+ with total charges q = 13 e.

The highest value of q corresponds to the formal charge of the
zeolite-supported Ir4Hn moiety, while the value at the low end,
1 e, is close to the estimated charges of the supported species, as
just mentioned.13 In the present work, we show that this
variation by only of a few electrons on the whole cluster
changes the direction of the oxidation: the metal moiety is
oxidized for neutral systems or clusters with q = 1 e, but the
hydrogen ligands are oxidized when the cluster carries a charge
q = 3 e. Thus, the charge of the system strongly aects the
direction of the oxidation or reduction during hydrogen loading
of the metal cluster. These eects, rst observed for iridium
clusters, also have been conrmed for hydrogenated clusters of
other 4d (Ru, Rh, Pd) and 5d (Os, Pt) transition metals.

2. Method and models

The electronic structure calculations were carried out with the
linear combination of Gaussian-type orbitals tting-functions
density functional method (LCGTO-FF-DF)14,15 as implemented in the program PARAGAUSS.16,17 We employed the
gradient-corrected exchange-correlation functional suggested
by Becke (exchange) and Perdew (correlation) (BP).18 We
applied a scalar relativistic variant of the LCGTO-FF-DF
method that aords an explicit description of relativistic
eects by treating all electrons with the DouglasKrollHess
approach of second order.15,19,20 The KohnSham (KS)
orbitals were represented by exible Gaussian-type basis
sets, contracted in generalized form: (6s1p) - [4s1p] for
H,21 (18s13p9d) - [7s6p4d] for Ru, Rh and Pd,22,23 and
(21s17p12d7f) - [9s8p6d4f] for Ir, Os, and Pt.22,23 The
auxiliary basis set, used in the LCGTO-FF-DF method to
represent the Hartree part of the electronelectron interaction,
was derived from the orbital basis set in a standard fashion.14
On each center (except H), this set was augmented by ve
p- and ve d-type polarization exponents, constructed as geometric
series with a factor 2.5, starting from 0.1 au (p) or 0.2 au (d). Only
the p-type series was added at hydrogen centers.
The current study comprises four series of hydrogenated
clusters in the gas phase with net charges q = 03 e (e denotes
Phys. Chem. Chem. Phys., 2010, 12, 1101511020

11015

Published on 29 July 2010. Downloaded by UNIVERSITY OF NEBRASKA on 15/04/2015 02:34:13.

View Article Online

the elementary charge) of the complexes: Ir4Hn, [Ir4Hn]+,

[Ir4Hn]2+, and [Ir4Hn]3+. Unlike in our previous study,13 we
did not apply any symmetry constraints during the structure
optimization. We considered the consecutive dissociative
adsorption of (formally) 3/2 H2 to the species Ir4Hn and in
this way generated structures that contain n = 0, 3, 6, 9, or
12 hydrogen ligands.13,24 The initial positions of the ligands
were chosen similar to those reported earlier for the zeolitesupported hydrogenated moieties.13 Each stationary point was
checked with a normal mode analysis that incorporated all
degrees of freedom to ensure that the reported structures
correspond to local minima. In the spirit of this model study,
we did not search for isomeric structures.
We quantied the stability of a cluster [Ir4Hn]q+ via the
energy change DE of the formal reaction [Ir4]q+ + n/2 H2 [Ir4Hn]q+, i.e. via its relative energies with respect to the
corresponding bare cluster [Ir4]q+:
DE([Ir4Hn]q+) = Etot([Ir4Hn]q+)  Etot([Ir4]q+)  n/2 Etot(H2)
(1)
Etot is the total energy of a system. A negative value of
DE reects the favorable formation of [Ir4Hn]q+ from the bare
cluster after dissociative adsorption of the proper amount of
hydrogen. The basis set superposition error was estimated via
the counterpoise method to 36% of DE. The adiabatic ionization potentials of the bare, [Ir4]q+, and the hydrogenated
clusters, [Ir4Hn]q+ (q = 02), were calculated as dierences of
appropriate total energies Etot. The spin contamination of the
KS determinant of open-shell systems never exceeded 2%.
Experimental core level energies cannot directly be compared with energies of KS orbitals, but changes of KS energies
with respect to a reference provide adequately approximate
core level shifts.25 We estimated average energy shifts of the
Ir 4f shell of [Ir4Hn]q+ clusters, using the Ir 4f KS energies of
the bare tetrahedral clusters [Ir4]q+ as reference. A positive
value of the shift corresponds to a stabilization of the core
levels relative to the reference. We determined eective charges
q of metal moiety Ir4 by tting the electrostatic potential.26
The quoted charges should be taken with due caution as
KohnSham methods with common exchange-correlation
functionals tend to overestimate electron delocalization.27
To probe the general validity of our analysis, we also
modeled [M4Hn]q+ clusters, q = 03, of other 4d and 5d late
transition metals in the same way, but only for the loadings
n = 0, 6, and 12.

3. Results and discussion

3.1

Structures and stability

The optimized structures of the clusters [Ir4Hn]q+ are provided

in Fig. 1 The average values hIrIri and the spread DR of the
nearest-neighbor distances in the metal moiety are given in
Table 1. As in the case of zeolite-supported species,13 the
average IrIr distance increases with the number of H ligands
coordinated at the cluster, from 247  1 pm in the bare clusters
to 268  2 pm in the clusters with 12 H ligands. These averages
include clusters of all charges studied. The nearest-neighbor
IrIr distances are not uniform within a cluster, as illustrated
11016

Phys. Chem. Chem. Phys., 2010, 12, 1101511020

by rather large values DR, up to B40 pm (Table 1). The

clusters [Ir4H9]3+ and [Ir4H12]3+ exhibit a buttery structure with one of the IrIr distances elongated up to 308 pm.
The terminal HIr bonds, 158163 pm, vary by a few
picometres only as the charge of the cluster changes. In
clusters with a large hydrogen loading, n = 9, 12, and a high
charge, q = 2 e or 3 e, activated H2 molecules, with interatomic distances less than 94 pm,28 form at one of the Ir centers.
Such paired hydrogen ligands feature longer HIr bonds,
169171 pm, than common terminal H ligands.
The relative energies DE of the hydrogenated species are
provided in Table 1 and compared in Fig. 2 as a function of
the number n of hydrogen ligands of the clusters. Similarly to
the data for the neutral tetrairidium clusters in the gas phase,13
the relative energies of the charged clusters [Ir4Hn]q+ increase
(by absolute value) almost linearly with the hydrogen loading
n along each series of clusters with a xed charge q (Fig. 2).
The dissociative adsorption of H2 releases, on average,
D2E = 70  20 kJ mol1 per H atom, but the variations
within each series at xed q clusters are notably smaller, at
most 8 kJ mol1. Hydrogen adsorption is calculated to be
most favorable on complexes with q = 3 e and on neutral
clusters, where D2E = 77 kJ mol1 and 73 kJ mol1,
respectively. The value for the neutral clusters is close to the
adsorption energy, 70 kJ mol1, calculated for models with
C3 symmetry restrictions,13 but smaller than the corresponding
energies calculated for the neutral clusters Ir4H and Ir4H2 of
similar structure, 75 kJ mol1 and 90 kJ mol1, respectively.11
The corresponding adsorption energies for the clusters
[Ir4Hn]+ and [Ir4Hn]2+ are B10 kJ mol1 smaller, 63 kJ mol1
and 66 kJ mol1, respectively. This may be related to the
stronger redox interactions in complexes with charges q = 0, 3 e,
compared to those with q = 1, 2 e (section 3.2). In summary, H
adsorption is quite favorable in all systems studied.11,13
3.2 Electron density distribution
Now, we turn to the central part of this work, the analysis of
the electron density distribution and of the (rst) ionization
potentials of hydrogenated metal clusters of dierent charge.
Metal moieties of zeolite-supported tetrairidium clusters and
of neutral clusters in the gas phase are oxidized after adsorption of hydrogen ligands and the ligands carry a partial
negative charge.11,13 In the following we will discuss the
changes in the electron density distribution as the hydrogen
loading and the cluster charge increase. We will diagnose
these changes via potential-derived charges as well as average
shifts of selected Ir core levels relative to the energies of
the corresponding bare cluster (Table 1, Fig. 3, and Fig. S1
of ESIw).
In the hydrogenated clusters Ir4Hn and [Ir4Hn]+ (n 4 0) the
charge of the metal moiety is above the total charge q of the
system (Fig. 3b). Thus the metal moieties are oxidized through
the adsorption of hydrogen, similar to our earlier results for
zeolite-supported Ir4 clusters.13 The eect increases with the
hydrogen loading. At maximum hydrogen loading, n = 12,
the charge of the Ir4 moiety is 2.20 e in the neutral complex
and 2.37 e in the corresponding monocation. In the corresponding dicationic hydrogenated complexes the charge
This journal is

the Owner Societies 2010

Published on 29 July 2010. Downloaded by UNIVERSITY OF NEBRASKA on 15/04/2015 02:34:13.

View Article Online

Fig. 1 Optimized structures of [Ir4Hn]q+ species for q = 0, 1, 2, and 3 and energy changes for the subsequent adsorption of 3/2 H2 (in kJ mol1).
Table 1 Various characteristics of the bare clusters [Ir4]q+ and the hydrogenated clusters [Ir4Hn]q+ (n = 0, 3, 6, 9, 12; q = 03) from density
functional calculations

Ir4
[Ir4]+
[Ir4]2+
[Ir4]3+
Ir4H3
[Ir4H3]+
[Ir4H3]2+
[Ir4H3]3+
Ir4H6
[Ir4H6]+
[Ir4H6]2+
[Ir4H6]3+
Ir4H9
[Ir4H9]+
[Ir4H9]2+
[Ir4H9]3+
Ir4H12
[Ir4H12]+
[Ir4H12]2+
[Ir4H12]3+

Nsa

DE,b kJ mol1

hIrIri,c pm

DR,d pm

q(Ir4),e e

DE(Ir 4f),f eV

0
1
2
3
3
2
1
0
0
1
0
1
1
0
1
0
0
1
0
1

196
179
201
270
487
408
427
475
637
555
544
727
888
761
809
874

248
246
247
247
252
252
251
251
256
256
255
258
262
262
261
257g
267
270
268
266g

1
9
14
17
17
14
12
3
24
28
26
38
32
38
50
60
7
14
28
42

0
1
2
3
1.36
1.89
2.38
2.47
1.47
1.30
1.86
2.17
1.83
1.62
1.93
2.05
2.20
2.37
2.08
2.14

1.07
0.17
0.14
0.56
0.97
0.68
0.32
0.97
1.52
0.89
0.23
1.61
1.67
0.88
0.50
1.72

Number of unpaired electrons in the complex. b Relative stability of the hydrogenated cluster, see eqn (1). c Average nearest-neighbor distance of
the metal moiety. d Dierence between the largest and the smallest nearest-neighbor distances of the optimized structure. e Potential-derived
charge of the metal moiety q(Ir4). f Average shift of the Ir 4f core levels with respect to the value of the corresponding bare cluster; positive values
can be considered to indicate oxidation of the metal moiety. g Distances exceeding 290 pm are not included when calculating the average.

alterations of the Ir4 moiety are quite small, from only 0.14 e
to 0.38 e. On the other hand, the dissociative adsorption of
hydrogen on iridium tetramers of charge q = 3 e results in a
reduction of the metal moiety, as the positive charge of the
iridium moiety in the hydrogenated clusters is always below
the total charge q of the system. In [Ir4H12]3+ the charge of the
metal moiety is by 0.86 e lower than the charge of the bare
cluster, 3 e.
Interestingly, with increasing hydrogen loading, the
charges of the metal moiety approach one and the same value,
2.23  0.15 e (n = 12), independently of the net charge
This journal is

the Owner Societies 2010

q = 03 e of the complex (Fig. 3a). A Mulliken analysis

(see ESIw) reveals the same trend: here, the charge of the metal
moiety gradually decreases to a common value, 0.67  0.23 e,
in the clusters with the highest hydrogen loading. Thus, a shell
of adsorbed hydrogen ligands acts as a charge buer with
respect to the metal cluster.
These charge eects are reected by the shifts of the metal
core levels as a comparison of Fig. 3b and c reveals for the
example of the Ir 4f levels.29 Although core level shifts in
general are the compound result of several factors,30 the
charge of the metal moiety clearly dominates for the systems
Phys. Chem. Chem. Phys., 2010, 12, 1101511020

11017

Published on 29 July 2010. Downloaded by UNIVERSITY OF NEBRASKA on 15/04/2015 02:34:13.

View Article Online

Fig. 2 Relative energy, DE, of hydrogenated [Ir4Hn]q+ clusters as a

function of the total number of hydrogen atoms adsorbed on the metal
cluster, n: neutral Ir4Hn clusters (circles); [Ir4Hn]+ clusters (triangles);
[Ir4Hn]2+ clusters (rhombs); [Ir4Hn]3+ clusters (squares).

under study. This eect comprises two major contributions, the

stabilization or destabilization due to the eective charge of an
atomic center and the electric eld of the other charged centers
(Ir or H) in the immediate vicinity. At each hydrogen loading n,
the Ir 4f core level shifts decrease with increasing charge of the
Ir4 moiety. In all systems studied, shifts are positive for the
neutral and the monocationic clusters, reecting the oxidation
of the metal centers. In contrast, the metal moiety of the
complexes [Ir4Hn]3+ is reduced and, concomitantly, the Ir 4f
core levels are clearly destabilized, up to 1.73 eV in [Ir4H12]3+,
as result of the dissociative adsorption of hydrogen.
To rationalize the dierent redox behavior with respect to
hydrogen, depending on the clusters charge, we compare
calculated adiabatic ionization potentials (IPs) of the clusters
[Ir4]q+, 6.8 eV, 12.2 eV, 18.2 eV, and 24.7 eV, for q = 03,
respectively, to the (calculated) IP of a H atom, 13.6 eV.
The propensity for the Ir4 moiety undergoing oxidation or
reduction obviously correlates with the IP value of the cluster
[Ir4]q+, whether it is smaller or larger, respectively, than the IP
of hydrogen.
We also estimated the electronegativities w of pertinent
systems according to w = (IP+EA)/2, using the calculated
values of the adiabatic IP and the electron anity (EA): 7.0 eV
for H, and 4.2 eV, 9.5 eV, 15.2 eV, and 21.4 eV for the bare
[Ir4]q+ clusters with q = 03 e, respectively. An alternative
evaluation of the electronegativities from the HOMO and
LUMO energies yields very similar values, 3.9 eV, 9.4 eV,
15.2 eV, and 21.3 eV, for the series of bare clusters, while w(H)
is reduced to 4.7 eV. As expected from the observed specic
redox behavior, the electronegativity of H is intermediate
between those of clusters with dierent charges. Yet, w(H)
appears between the w values of the clusters with q = 0 e and
1 e, while the change in the redox behavior is observed ongoing
from clusters with q = 1 e to 2 e. The reason for this dierence
likely is the reference system, an isolated H atom, used for
comparison with the electronegativities of the metal clusters.
3.3

Comparison with clusters of other transition metals

To check whether the eects of the cluster charge, observed for

the series of neutral and positively charged clusters [Ir4]q+,
11018

Phys. Chem. Chem. Phys., 2010, 12, 1101511020

Fig. 3 Potential-derived charges q of Ir4 moiety in [Ir4Hn]q+ clusters

(a), and variations of these charges (b), as well as average shifts of the
Ir 4f levels (c) with respect to the corresponding bare clusters as a
function of the total number n of hydrogen atoms adsorbed on the
metal moiety: Ir4Hn (circles), [Ir4Hn]+ (triangles), [Ir4Hn]2+ (rhombs),
[Ir4Hn]3+ (squares).

also holds for clusters of other similar transition metals, we

modeled in the same way tetragonal model clusters of various
4d (Ru, Rh, Pd) and 5d (Os, Pt) transition metals; the
optimized structures of these clusters are provided in Fig. S2
of ESI.w In these calculations we included only clusters
[M4Hn]q+ of selected hydrogen loading, n = 0, 6, 12. Similarly
to the series of iridium clusters, the average adsorption energy
per H on dierent clusters was evaluated by a least-squares
t of the relative energies DE of the hydrogenated tetramers [M4Hn]q+ on the hydrogen loading n (Table 2). The
values DE for the neutral clusters dier by B50 kJ mol1,
This journal is

the Owner Societies 2010

View Article Online

Table 2 Various characteristics of the bare clusters [M4]q+ and the hydrogenated metal clusters [M4Hn]q+ for late 4d and 5d transition metals
M (n = 0, 6, 12; q = 03) from density functional calculations
DIP,b eV

Published on 29 July 2010. Downloaded by UNIVERSITY OF NEBRASKA on 15/04/2015 02:34:13.

q(M4),a e
Ru
Rh
Pd
Os
Ir
Pt

2.62
1.92
2.05
2.20
2.23
2.08








0.22
0.05
0.49
0.23
0.15
0.17

D2E,d kJ mol1

w,c eV

q=0

q=2

q=1

q=2

q=0

q=1

q=2

q=3

0.81
0.21
1.02
1.21
1.30
1.83

0.96
2.77
2.31
0.82
0.67
1.05

8.93
9.93
10.08
9.24
9.54
9.82

14.82
16.08
16.39
14.39
15.25
15.81

49.6
42.3
24.3
64.9
73.6
62.2

43.0
38.8
18.4
58.6
52.8
48.2

48.3
45.7
38.7
54.4
65.8
60.2

53.7
67.5
58.3
62.1
76.7
68.6

a
Charge q(M4) of the metal moieties of the clusters [M4H12]q+ averaged for q = 0 e and q = 3 e; the variation of these values is characterized by
half of their dierence D. b Dierence between adiabatic ionization potentials of the clusters [M4H12]q+ and [M4]q+; positive values can be
considered to indicate oxidation of the metal moiety. c Electronegativity of the [M4]q+ clusters estimated from the adiabatic ionization potentials
IP and electron anities EA: w = (IP + EA)/2. d Average adsorption energy per H, evaluated by least-squares ts of the dependence of the relative
energies DE of the hydrogenated clusters [M4Hn]q+ (n = 0, 6, 12; q = 03) on the hydrogen coverage n: DE([M4Hn]q+) = D2E  n; see eqn (1). For
the iridium clusters, the whole set of data for n = 0, 3, 6, 9, 12 was used.

from 73.6 kJ mol1 (Ir) to 24.3 kJ mol1 (Pd). The values

for the clusters with q = 3 e spread over a smaller interval of
B20 kJ mol1, 76.7 kJ mol1 (Ir) to 53.7 kJ mol1 (Ru).
The adsorption energies are lowest (by absolute value) for
q = 1 e or 2 e, and highest for q = 3 e, except for Os4 which
features a slightly larger adsorption energy than [Os4]3+. The
adsorption interaction can increase strongly with cluster
charge; for q = 3 e the energies released were calculated
25 kJ mol1 (Rh) and 34 (Pd) kJ mol1 more exothermic than
in the corresponding neutral clusters.
As for Ir, the oxidation or reduction of the M4 moiety upon
the hydrogen loading depends on the total charge q of the
complexes [M4Hn]q+. This can be seen from the calculated
dierence between the adiabatic IPs of the hydrogenated
clusters [M4H12]q+ and the bare clusters [M4]q+ (Table 2).
For all neutral systems, the IP of the hydrogenated clusters is
larger than the IP of the corresponding bare clusters: this can
be considered as indication of an oxidation of the metal moiety
upon hydrogen adsorption. In contrast, the IP values of the
hydrogenated clusters with q = 2 e are lower than those of the
corresponding bare cluster, i.e. the metal moieties are reduced.
These trends are corroborated by the core level shifts of the
metal atoms (Fig. S3 of ESIw).
For all transition metals studied, the charge buer eect of the
shell of hydrogen ligands on the metal clusters exists in the case
of the highest hydrogen coverage, 12 H; for platinum, this eect
can be identied even with only six hydrogen ligands. The spread
D between maximum and minimum calculated charges q(M4) of
the metal moieties for systems of total charges q = 0 e and q =
3 e varies, from very small for Rh (D o 0.1 e) to D E 1.0 e for
Pd. Surprisingly, the charge of the M4 moiety in [M4H12]q+,
averaged over q = 03 e, is quite similar, 2.082.23 e, for the 5d
metals, while these averaged charges vary notably for the 4d
metals, from 1.92 e for Rh4 to 2.62 e for Ru4. The similarity of
the transition metal clusters modeled may be due to the similar
values of the electronegativity of the clusters, reported in Table 2
for the complexes with q = 1 e or 2 e.

4. Conclusions
The results obtained for neutral and charged iridium clusters
show that a small change of the net charge of the complexes
This journal is

the Owner Societies 2010

studied, by only 2 e or less than one electron per metal center,

reverses the character of the redox interaction between the
metal moiety and the hydrogen ligands. While the dissociative
adsorption of hydrogen on the species Ir4 and [Ir4]+ leads to
an oxidation of the metal moiety, the highly charged metal
cluster [Ir4]3+ is reduced upon hydrogen adsorption. This
contrasting behavior with a variation of the charge on the
tetrairidium clusters can be rationalized by the increasing
propensity of the metal moiety to accept electron density with
increasing the charge, as reected in the increasing ionization
potentials of the clusters [Ir4]q+ with the charge q = 03 e.
Such an increase of the ionization potential with increasing
positive charge is expected to occur also for other metal
clusters and indeed has been corroborated for various late
4d and 5d transition metals. At a critical charge value, the
IP of the metal moiety will be close to the IP of the ligand
system and thus the IP will be a convenient parameter for
evaluating the redox behavior of the metal clusters with
respect not only to hydrogen, but also to other ligands.
Therefore, all metal clusters should exhibit a characteristic
(positive) charge value at which the adsorption of hydrogen
changes from being an oxidative process of the metal moiety
to a reduction.
A comparison of the charge distributions in the complexes
[Ir4H12]q+, q = 03 e, with maximum hydrogen loading
showed that the charges of the metal moieties are very similar,
irrespective of the charge q of the complex. This observation
suggests that in these complexes an equilibration of the
electron density takes place such that the metal fragment
carries essentially the same charge, while the charge dierences
are accommodated by the hydrogen ligands. In other words,
the dissociatively adsorbed hydrogen ligands act as a charge
buer with respect to the metal cluster.

Acknowledgements
This work was supported by the Bulgarian National Science
Fund (Contract VUH-303/07), the Bulgarian National Center
of Advanced Materials UNION (Contract DO02-82/2008),
Deutsche Forschungsgemeinschaft, and Fonds der Chemischen
Industrie (Germany).
Phys. Chem. Chem. Phys., 2010, 12, 1101511020

11019

View Article Online

Published on 29 July 2010. Downloaded by UNIVERSITY OF NEBRASKA on 15/04/2015 02:34:13.

References
1 Handbook of Heterogeneous Catalysis, ed. G. Ertl, H. Knozinger
and J. Weitkamp, Wiley-VCH, Weinheim, 1997; A. W. Castleman,
Jr. and P. Jena, Proc. Natl. Acad. Sci. U. S. A., 2006, 103,
1056010569.
2 A. M. Argo, J. F. Odzak, J. F. Goellner, F. S. Lai, F.-S. Xiao and
B. C. Gates, J. Phys. Chem. B, 2006, 110, 17751786; F. Li and
B. C. Gates, J. Phys. Chem. C, 2007, 111, 262267.
3 J. Wei and E. Iglesia, Angew. Chem., Int. Ed., 2004, 43, 36853688.
4 S. Liu, H. J. Zhai and L. S. Wang, J. Chem. Phys., 2002, 117,
97589765; C. Adlhart and E. Uggerud, Chem.Eur. J., 2007, 13,
68836890.
5 S. T. Homeyer, Z. Karpinski and W. M. H. Sachtler, J. Catal.,
1990, 123, 6073; W. M. H. Sachtler and Z. Zhang, Adv. Catal.,
1993, 39, 129220; T. J. McCarthy, G.-D. Lei and W. M. H.
Sachtler, J. Catal., 1996, 159, 9098.
6 B. J. Kip, F. B. M. Duivenvoorden, D. C. Koningsberger and
R. Prins, J. Catal., 1987, 105, 2638; F. W. H. Kampers and
D. C. Koningsberger, Faraday Discuss. Chem. Soc., 1990, 89, 137141.
7 A. Genest, S. Kruger and N. Rosch, J. Phys. Chem. A, 2008, 112,
77397744; K. M. Neyman, C. Inntam, L. V. Moskaleva and
N. Rosch, Chem.Eur. J., 2007, 13, 277286.
8 S. Sicolo, C. Di Valentin and G. Pacchioni, J. Phys. Chem. C, 2007,
111, 51545161.
9 Y. Dong and M. Springborg, Eur. Phys. J. D, 2007, 43, 1518;
R. Grybos, L. Benco, T. Bucko and J. Hafner, J. Chem. Phys.,
2009, 130, 104503.
10 I. Swart, F. M. F. de Groot, B. M. Weckhuysen, P. Grune,
G. Meijer and A. Fielicke, J. Phys. Chem. A, 2008, 112,
11391149; D. J. Harding, T. R. Walsh, S. M. Hamilton,
W. S. Hopkins, S. R. Mackenzie, P. Grune, M. Hartelt, G. Meijer
and A. Fielicke, J. Chem. Phys., 2010, 132, 011101; M. Citir, F. Liu
and P. B. Armentrout, J. Chem. Phys., 2009, 130, 054309.
11 C. Bussai, S. Kruger, G. N. Vayssilov and N. Rosch, Phys. Chem.
Chem. Phys., 2005, 7, 26562663; S. Kruger, C. Bussai, A. Genest
and N. Rosch, Phys. Chem. Chem. Phys., 2006, 8, 33913398.
12 G. P. Petrova, G. N. Vayssilov and N. Rosch, Chem. Phys. Lett.,
2007, 444, 215219.
13 G. P. Petrova, G. N. Vayssilov and N. Rosch, J. Phys. Chem. C,
2007, 111, 1448414492.
14 B. I. Dunlap and N. Rosch, Adv. Quantum Chem., 1990, 21,
317339.
15 N. Rosch, S. Kruger, M. Mayer and V. A. Nasluzov, in: Recent
Development and Applications of Modern Density Functional Theory.

11020

Phys. Chem. Chem. Phys., 2010, 12, 1101511020

16

17

18

19
20

21
22
23

24
25
26
27
28
29

30

Theoretical and Computational Chemistry, ed. J. M. Seminario,

Elsevier, Amsterdam, 1996, vol. 4, pp. 497566.
T. Belling, T. Grauschopf, S. Kruger, M. Mayer, F. Nortemann,
M. Staufer, C. Zenger and N. Rosch, in High Performance Scientic
and Engineering Computing, Lecture Notes in Computational Science
and Engineering, ed. H.-J. Bungartz, F. Durst and C. Zenger,
Springer, Heidelberg, 1999, vol. 8, pp. 439453.
T. Belling, T. Grauschopf, S. Kruger, F. Nortemann, M. Staufer,
M. Mayer, V. A. Nasluzov, U. Birkenheuer, A. Hu,
A. V. Matveev, A. M. Shor, M. S. K. Fuchs-Rohr,
K. M. Neyman, D. I. Ganyushin, T. Kerdcharoen, A. Woiterski,
A. B. Gordienko, S. Majumder and N. Rosch, PARAGAUSS
version 3.0, Technische Universitat Munchen, Munchen,
2004.
A. D. Becke, Phys. Rev. A: At., Mol., Opt. Phys., 1988, 38,
30983100; J. P. Perdew, Phys. Rev. B: Condens. Matter, 1986,
33, 88228824; J. P. Perdew, Phys. Rev. B: Condens. Matter, 1986,
34, 7406.
O. D. Haberlen and N. Rosch, Chem. Phys. Lett., 1992, 199,
491496.
N. Rosch, A. Matveev, V. A. Nasluzov, K. M. Neyman,
L. Moskaleva and S. Kruger, in Relativistic Electronic Structure
Theory. Part II: Applications, Theoretical and Computational
Chemistry Series, ed. P. Schwerdtfeger, Elsevier, Amsterdam,
2004, vol. 14, pp. 656722.
F. B. van Duijneveldt, IBM Res. Report RJ 945, 1971.
O. Gropen, J. Comput. Chem., 1987, 8, 9821003.
K. M. Neyman, C. Inntam, V. A. Nasluzov, R. Kosarev and
N. Rosch, Appl. Phys. A: Mater. Sci. Process., 2004, 78, 823828;
K. M. Neyman, C. Inntam, A. V. Matveev, V. A. Nasluzov and
N. Rosch, J. Am. Chem. Soc., 2005, 127, 1165211660.
G. P. Petrova, G. N. Vayssilov and N. Rosch, J. Phys. Chem. C,
2008, 112, 1857218577.
G. N. Vayssilov and N. Rosch, J. Catal., 1999, 186, 423432.
B. H. Besler, K. M. Merz and P. A. Kollman, J. Comput. Chem.,
1990, 11, 431439.
A. J. Cohen, P. Mori-Sanchez and W. Yang, Science, 2008, 321,
792794.
G. J. Kubas, Chem. Rev., 2007, 107, 41524205.
Other core levels shift by similar amounts. For instance, the
average shifts of the Ir 2p and Ir 5s levels change synchronously
with the energies of the Ir 4f levels, with deviations of at most 0.03
eV; see Table S2 of ESI.
P. S. Bagus, F. Illas, G. Pacchioni and F. Parmigiani, J. Electron
Spectrosc. Relat. Phenom., 1999, 100, 215236.

This journal is

the Owner Societies 2010