In situ-generated PVP-stabilized palladium(0) nanocluster catalyst

in hydrogen generation from the methanolysis of ammonia–borane

Huriye Erdogan, Onder Metin and Saim Ozkar*

Received 11th August 2009, Accepted 11th September 2009

First published as an Advance Article on the web 6th October 2009

DOI: 10.1039/b916459f

Herein, we report the in situ generation of poly(N-vinyl-2-pyrrolidone) (PVP)-stabilized

palladium(0) nanoclusters and their catalytic activity in hydrogen generation from the

methanolysis of ammonia–borane (AB). The PVP-stabilized palladium(0) nanoclusters with an

average particle size of 3.2 � 0.5 nm were formed from the reduction of palladium(II)

acetylacetonate during the methanolysis of AB in the presence of PVP at room temperature.

The palladium(0) nanoclusters are highly stable in solution for extended periods of time,

can be isolated as solid materials, are redispersible in methanol and show catalytic activity after

redispersion. The nanoclusters were characterized by TEM, XPS, FTIR, UV-Vis, XRD, and

SAED techniques. Mercury poisoning experiments indicate that PVP-stabilized palladium(0)

nanoclusters are heterogeneous catalysts in the methanolysis of ammonia–borane. The

PVP-stabilized palladium(0) nanoclusters are highly active and stable catalysts as they provide

23 000 turnovers in hydrogen generation from the methanolysis of AB over 27 h before

deactivation at room temperature. A kinetic study shows that the catalytic methanolysis

of AB is first order with respect to catalyst concentration and zero order with respect to substrate

concentration. The activation energy of the methanolysis of AB catalyzed by PVP-stabilized

palladium(0) nanoclusters was determined to be Ea = 35 � 2 kJ mol�1.

Introduction

Hydrogen is a globally accepted clean fuel which could

overcome the world energy problem and reduce the environ-

mental pollution caused by usage of fossil fuels as an energy

source.1 The main challenge for the widespread application

of hydrogen is its real-time production and its safe and

convenient storage because the low density of H2 makes it

difficult to store in compressed or liquid form.2 Metal

hydrides,3 metal organic frameworks,4 chemical hydrides,5

and hydrocarbons6 have been tested as solid hydrogen storage

materials. Although sodium borohydride has also been

considered as a solid hydrogen storage material,7 its use has

been restricted to portable fuel cell applications because of

its instability in solution without a base8 and inefficacy in

recycling the hydrolysis product.9 Recently, ammonia–borane

(AB) has attracted much attention due to its high hydrogen

capacity (19.6 wt%), high stability, and non-toxicity.10

Hydrogen can be generated from AB via thermal

decomposition,11 acid-catalyzed12 or transition-metal

catalyzed dehydrogenation,13–15 and hydrolysis.16

In contrast to NaBH4, ammonia–borane does not undergo

self hydrolysis in water and only generates hydrogen through

the hydrolysis reaction in the presence of a suitable catalyst.17

Our recent studies have shown that zeolite framework-

stabilized rhodium(0) nanoclusters and poly(N-vinyl-2-

pyrrolidone) (PVP)-stabilized cobalt(0) nanoclusters are

highly active catalysts for hydrogen generation from the

hydrolysis of ammonia–borane.18 However, liberation of

small quantities of ammonia at high AB concentrations and

inefficacy in recycling the hydrolysis product, metaborate, can

pose problems for hydrogen generation from hydrolysis of AB

in fuel cell applications. In this regard, methanolysis appears

to be advantageous for hydrogen generation from AB

(eqn (1)),19,20 as the methanolysis product can be regenerated

and no ammonia is liberated in the methanolysis—two

important requirements for fuel cell applications.10

H3NBH3 þ 4CH3OH�!catalyst NH4BðOCH3Þ4 þ 3H2 ð1Þ

Methanolysis of AB occurs only in the presence of a suitable

catalyst. Herein, we report the employment of in situ-generated,

polymer-stabilized palladium(0) nanoclusters as catalysts in

the methanolysis of AB. The catalyst systems used so far in the

methanolysis of AB are listed in Table 1.

Ramachandran and Gagare have tested various transition-

metal salts such as RuCl3, RhCl3, PdCl2 and CoCl2 in the

methanolysis of AB and shown that RuCl3 is the catalyst of

choice (Table 1).10 Although RuCl3 provides high activity for

the methanolysis of AB, the report lacks information on the

lifetime of the catalyst; as in the other two papers,19,20 it has

not been reported for how long such a high TOF value would

be retained. Also, the issue of whether the catalysis starting

with the transition metal salts is homogeneous or hetero-

geneous has not been addressed.10 It is noteworthy that

the catalytic activities of two Pd catalysts, PdCl2 and

carbon-supported Pd, are very low.10 Similarly, Pd supported

on g-Al2O3 has been reported to have much lower activity in

Department of Chemistry, Middle East Technical University,06531 Ankara, Turkey. E-mail: sozkar@metu.edu.tr;Fax: +90 312 210 3200; Tel: +90 312 210 3212

This journal is �c the Owner Societies 2009 Phys. Chem. Chem. Phys., 2009, 11, 10519–10525 | 10519

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

Publ

ishe

d on

06

Oct

ober

200

9. D

ownl

oade

d by

Uni

vers

ity o

f C

alga

ry o

n 27

/09/

2013

13:

10:0

1.

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

the hydrolysis of AB compared to the other transition metals

such as Ru, Rh and Pt supported on g-Al2O3.21 Despite the

low catalytic activity of palladium in both the hydrolysis and

methanolysis of AB, we have shown that the PVP-stabilized

palladium(0) nanoclusters are highly active and long-lived

heterogeneous catalysts in the methanolysis of AB.

PVP-stabilized palladium nanoparticles have been prepared

via alcohol,22 polyol,23 microwave assisted24 and chemical

reduction methods.25 For catalytic applications, PVP-stabilized

nanoparticles prepared by using one of these methods must be

separated from the synthesis solution and redispersed in the

catalysis solvent . The separation processes generally result in

a significant loss of catalyst. In this regard, in situ generation

of nanoparticles in the medium of the catalytic reaction is

significant. Polymer-stabilized palladium(0) nanoclusters were

generated for the first time in situ by reduction of palladium(II)

acetylacetonate in the presence of PVP during the methano-

lysis of AB. PVP-stabilized palladium(0) nanoclusters are

highly active catalysts in the methanolysis of AB and stable

in solution. They could be isolated from the solution and

characterized by transmission electron microscopy (TEM),

X-ray photoelectron spectroscopy (XPS), Fourier transform

infra-red (FT-IR), X-ray diffraction (XRD), selected area

electron diffraction (SAED) and UV-Visible electronic absorption

spectroscopy. Mercury poisoning was used to determine

whether the in situ generated PVP-stabilized palladium(0)

nanoclusters are homogeneous or heterogeneous catalysts in

the methanolysis of ammonia–borane. The detailed kinetics

of the catalytic methanolysis of AB were also studied by

measuring the volume of hydrogen generated during the

reaction whilst varying the catalyst concentration, substrate

concentration and temperature.

Experimental

Materials

Palladium(II) acetylacetonate (99%), borane–ammonia

complex (497%), PVP-40 (average molecular weight

40 000), were purchased from Aldrich and used as received.

Methanol was purchased from Riedel-De Haen AG

Hannover, and it was distilled over metallic magnesium. All

methanolysis reactions were performed using the distilled

methanol under inert gas atmosphere unless otherwise

specified. Teflon-coated magnetic stir bars and all glassware

were cleaned with acetone and dried in an oven at 150 1C.

In situ generation of PVP-stabilized palladium(0) nanoclusters

and the methanolysis of ammonia–borane

Both the in situ formation of PVP-stabilized palladium(0)

nanoclusters and the catalytic methanolysis of AB were

performed in the same medium. Before starting the

experiment, a jacketed reaction flask (50 mL) containing a

Teflon-coated stir bar was placed on a magnetic stirrer (IKA

RCT Basic) and thermostated to 25.0 � 0.5 1C by circulating

water through its jacket from a constant temperature bath

(Lauda RL6). Then, a graduated, glass cylinder tube filled

with water was connected to the Schlenk tube to measure the

volume of hydrogen gas evolved from the reaction. Next,

1.55 mg palladium(II) acetylacetonate (0.5 mM Pd) and

2.78 mg PVP-40 (2.5 mM) were dissolved in 5 mL methanol

in the Schlenk tube with a stir bar and stirred. Next, 64 mg

ammonia–borane (200 mM AB) was dissolved in 5 mL

methanol elsewhere and then transferred into the Schlenk tube

under argon atmosphere at 700 rpm with constant stirring.

The initial concentrations of AB and palladium were 200 mM

AB, and 0.5 mM Pd, respectively. A molar ratio of AB to

metal precursors greater than 100 was used to ensure complete

reduction of Pd2+ to its zero oxidation state and to observe

the catalytic hydrolysis of AB at the same time. The formation

of nanoclusters took less than 1 min and immediate evolution

of hydrogen gas took place, indicating that the nanoclusters

formed start to catalyze the methanolysis of AB. Catalytic

methanolysis was monitored by measuring the volume of

hydrogen gas evolved every 2 min at constant pressure by

determining the decrease in the water level of the glass tube

which was connected to the Schlenk tube. In a control experi-

ment, the stirring rate was varied in the range 0–800 rpm. It

was found that after 500 rpm, the stirring rate has no

significant effect on the catalytic activity. This indicates that

the system is in a non-MTL (mass transfer limitation) regime

at stirring rates 4500 rpm. In other words, there is no

diffusion problem due to the presence of the polymer in

concentrations of less than 5 mM.

Table 1 List of the catalysts used in hydrogen generation from the methanolysis of AB at 25 1C. The life time is given by the number of totalturnovers (TTO). Turnover frequency (TOF) values were estimated from the data given in respective references

Catalyst Amount of AB and catalyst TOF/min�1 TTO Ea/kJ mol�1 Ref.

1 RuCl3 2.9 mmol, 0.03 mol% 173 — — 102 RhCl3 2.9 mmol, 2 mol% 101 — — 103 CoCl2 2.9 mmol, 2 mol% 3.7 — — 104 NiCl2 2.9 mmol, 2 mol% 2.9 — — 105 RANEYs Ni 2.9 mmol, 2 mol% 3.6 — — 106 Pd/C 2.9 mmol, 2 mol% 2.0 — — 107 PdCl2 2.9 mmol, 2 mol% 1.6 — — 108 Cu NPs 1 mmol, 15 mol% 0.12 — — 199 Cu2O 1 mmol, 15 mol% 0.21 — — 1910 Cu–Cu2O 1 mmol, 15 mol% 0.26 — — 1911 Co–Co2B 10 mmol, 20 mol% 7.5 — — 2012 Ni–Ni3B 10 mmol, 20 mol% 5.0 — — 2013 Co–Ni–B 10 mmol, 20 mol% 10.0 — — 2014 Pd NCs 2 mmol, 0.5 mol % 22.3 23000 35 This work

10520 | Phys. Chem. Chem. Phys., 2009, 11, 10519–10525 This journal is �c the Owner Societies 2009

Publ

ishe

d on

06

Oct

ober

200

9. D

ownl

oade

d by

Uni

vers

ity o

f C

alga

ry o

n 27

/09/

2013

13:

10:0

1.

View Article Online

Self-methanolysis of ammonia–borane

64 mg AB were dissolved in 10 mL methanol in a jacketed

reaction flask (50 mL) containing a Teflon-coated stir bar. The

reaction flask was thermostated to a temperature in the range

15–35 1C by circulating water through its jacket from a

constant temperature bath. The experiment was started by

simultaneously closing the reaction flask and turning on the

stirrer at 700 rpm. Then the reaction was followed in the same

way as described above.

Instrumentation

The TEM and SAED images were obtained using a JEM-2010

(JEOL) TEM instrument operating at 200 kV. The samples

used for the TEM experiments were harvested from the

in situ-generated PVP-stabilized palladium(0) nanocluster

solution described above. The nanocluster solution was

centrifuged at 8000 rpm for 8 min. The separated nanoclusters

were washed with acetone to remove the excess PVP and other

residuals. Then, the nanocluster sample was redispersed in

5 mL methanol. One drop of the colloidal solution was

deposited on the silicon oxide coated copper grid and evaporated

under inert atmosphere. Samples were examined at magnifi-

cations between 100 and 400 K. The samples used for XPS

analysis were prepared by evaporating the methanol from the

nanocluster solution, as described above, by rotary evaporation

under vacuum. The resulting solid particles were washed with

acetone to remove the residuals from the solid sample under

N2 atmosphere. The X-ray photoelectron spectrum was taken

by using a SPECS spectrometer equipped with a hemispherical

analyzer and using monochromatic Al Ka radiation (1450 eV,

the X-ray tube working at 15 kV and 350 W) and a pass energy

of 48 eV. To better access the metal core in the sample, the

polymer matrix was scraped from surface of the sample via

bombardment by argon ions at 3000 eV for 3 min. The sample

prepared for the XPS analysis was also used for FT-IR

analysis. FT-IR spectra of neat PVP and PVP-stabilized

palladium(0) nanoclusters were taken from a KBr pellet on a

Bruker-Advance FTIR Spectrophotometer using Opus

software. To examine the final reaction product by11B-NMR, a certain amount of sample was taken directly

from the Schlenk tube into quartz NMR tube after the

stoichiometric hydrogen generation from the catalytic

methanolysis of AB finished. A few drops of CDCl3 solution

were added into the a quartz NMR tube. NMR spectra were

recorded on a Bruker Avance DPX 400 with an operating

frequency of 128.15 MHz for 11B-NMR. UV-Vis electronic

absorption spectra of palladium(II) acetylacetonate and

PVP-stabilized palladium(0) nanoclusters were recorded in

methanol solution on a Varian-Carry100 double beam

instrument. XRD patterns of PVP-stabilized palladium(0)

nanoclusters were recorded on a Rigaku Miniflex diffracto-

meter with Cu Ka (30 kV, 15 mA, l = 1.54051 A), over a 2yrange of 5 to 901 at room temperature.

Mercury-poisoning the methanolysis of ammonia–borane

catalyzed by PVP-stabilized palladium(0) nanoclusters.A standard

nanocluster formation and methanolysis experiment was

started with 200 mM AB, 1.0 mM palladium(II) acetylacetonate

and 5.0 mM PVP in 10 mL methanol at 25 � 0.1 1C. After

about 40% conversion of AB, elemental Hg (6.0 g, 3.0 mmol,

ca. 300 equiv.) was added to the solution under stirring.

The reaction was followed for a further 3 h.

Kinetics of the methanolysis of ammonia–borane catalyzed

by PVP-stabilized palladium(0) nanoclusters

In order to determine the rate law of the methanolysis of AB

catalyzed by PVP-stabilized palladium(0) nanoclusters, three

different sets of experiments were performed as described in

the ‘In situ generation of PVP-stabilized palladium(0) nano-

clusters in the methanolysis of ammonia–borane’ section.

Firstly, the AB concentration was kept constant at 200 mM

and the palladium concentration was varied (values of 0.5,

0.75, 1.0, 1.25, and 1.5 mM). In the second set of experiments,

the palladium concentration was kept constant at 0.5 mM and

the AB concentration was varied (values of 100, 150, 200, 250,

and 300 mM). In the third set, the methanolysis reaction was

performed at constant catalyst (1 mM Pd) and constant AB

(200 mM) concentration by varying the temperature in

the range of 15–35 1C to obtain the activation parameters;

activation energy (Ea), entalphy (DHa) and entropy (DSa).

Determination of the catalytic lifetime of PVP-stabilized

palladium(0) nanoclusters in the methanolysis of

ammonia–borane

The catalytic lifetime of PVP-stabilized palladium(0) nano-

clusters in the methanolysis of AB was determined by measuring

the total turnover number (TTON). This experiment was

started with a 20 mL solution containing 0.5 mM palladium(II)

acetylacetonate and 500 mM AB (320 mg) at 25 � 0.5 1C.

After conversion of all the added H3NBH3, checked by

stoichiometric H2 gas evolution (3.0 mol H2/mol H3NBH3),

a new batch of 320 mg ammonia–borane was added and the

reaction was kept going in this way until no hydrogen gas

evolution was observed. The ammonia generation during the

TTO experiments was checked before addition of new batch of

AB by using an acid/base indicator.

Results and discussion

Preparation, characterization and heterogeneity

of PVP-stabilized palladium(0) nanoclusters

PVP-stabilized palladium(0) nanoclusters were formed in situ

from the reduction of palladium(II) acetylacetonate in the

presence of PVP during the methanolysis of AB. In a typical

experiment, polymer and metal precursor were first mixed well

in methanol by stirring at 700 rpm and then AB was added to

the reaction solution in order to form palladium(0) nano-

clusters. As with other amine–boranes,13,18 AB acts as

reducing agent and provides two electrons for the reduction

of Pd(II) to Pd(0). Then, the Pd(0) nanoclusters are formed by

the nucleation and autocatalytic growth mechanism.26 The

abrupt colour change from pale yellow to dark brown after

addition of the AB solution into the mixture of metal

precursor and PVP indicates the formation of palladium(0)

nanoclusters. Concomitantly, hydrogen evolution starts after

This journal is �c the Owner Societies 2009 Phys. Chem. Chem. Phys., 2009, 11, 10519–10525 | 10521

Publ

ishe

d on

06

Oct

ober

200

9. D

ownl

oade

d by

Uni

vers

ity o

f C

alga

ry o

n 27

/09/

2013

13:

10:0

1.

View Article Online

the colour change indicating that the nanoclusters are the

active catalyst in the methanolysis of AB.

The conversion of palladium(II) acetylacetonate to

palladium(0) nanoclusters can be followed in the UV-visible

electronic absorption spectra taken during the reaction

(Fig. 1). The adsorption bands of palladium(II) acetyl-

acetonate at 330, 260 and 225 nm gradually disappear upon

addition of AB into the solution with the concurrent growth of

a new absorption band at 295 nm, attributable to Pd(0)

nanoclusters.27 This spectroscopic observation indicates that

the reduction of palladium(II) to palladium(0) is complete

within less than 5 min.

Survey experiments performed by varying the polymer to

metal ratio showed that the highest stability and catalytic

activity of the PVP-stabilized palladium(0) nanoclusters is

obtained when a five-fold molar excess of polymer is used in

the methanolysis of AB. It is most likely that the complete

dissolution of PVP in methanol achieved by prolonged

vigorous stirring is required to obtain stable palladium(0)

nanoclusters. The PVP-stabilized palladium(0) nanoclusters

are highly stable in solution, even for months, and they can

be isolated from the solution as solid materials by removing

the volatiles in vacuum. The solid palladium(0) nanoclusters

isolated are also stable for months under inert atmosphere and

readily redispersible in methanol. Moreover, when redispersed

in methanol, the palladium(0) nanoclusters still show catalytic

activity in the methanolysis of AB.

The morphology and particle size of the PVP-stabilized

palladium(0) nanoclusters were studied by using TEM

(Fig. 2a). Palladium(0) nanoclusters with average particle size

of 3.2 � 0.5 nm are obtained as shown in the histogram

(Fig. 2b). The AB reduction of palladium(II) acetylacetonate

in the presence of PVP yields nearly-spherical nanoclusters

with no agglomeration. Compared to the size of PVP-stabilized

palladium nanoparticles prepared by using the well-known

alcohol reduction method (3–8 nm depending on the PVP/Pd

ratio),22 the use of mild reducing agent AB and low

temperature leads to the formation of palladium nanoclusters

in smaller particle size and narrower size distribution. In

particular, it is noteworthy that the PVP-stabilized Pd(0)

nanoclusters prepared by AB reduction are stable in solution

against agglomeration while the nanoparticles previously used

as catalysts in the methanolysis of AB20 have been shown to

undergo rapid agglomeration. This is key in obtaining higher

catalytic activity (see later).

The XPS spectrum of PVP-stabilized palladium(0) nano-

clusters given in Fig. 3 shows two well-resolved peaks at

335.8 and 341.4 eV, which are readily assigned to Pd(0) 3d5/2and Pd(0) 3d3/2, respectively, by comparison with the values of

metallic palladium.28 That no higher oxidation peak for

palladium is observed in the XPS spectrum might be due to

the protection of palladium(0) against oxidation by the surface-

adsorbed PVP during the sample preparation. Also,

comparison of the FT-IR spectra of the PVP-stabilized

palladium(0) nanoclusters and neat PVP, both taken from a

KBr pellet, shows the existence of PVP in the nanoclusters

sample, most probably adsorbed on the surface of the nano-

clusters, since the free PVP molecules should have been

removed while washing with excess acetone.

Fig. 4a shows the powder XRD patterns for the

PVP-stabilized palladium(0) nanoclusters sample. The peak

broadening is characteristic of materials having a nanometer

particle size.29 Four reflections are observed in the XRD

pattern at 2y of 39.5, 45, 59 and 66.51 that could be attributed

to 111, 200, 220 and 311 peaks of elemental palladium,30 but

another peak observed at 21.51 most probably belongs to the

residual ammonium tetramethoxyborate species remaining on

the surface of the nanoclusters. These four reflections are also

Fig. 1 UV-Visible electron absorption spectra taken during the

reaction of palladium(II) acetylacetonate with ammonia–borane in

the presence of PVP in methanol solution.

Fig. 2 (a) TEM image and (b) associated histogram for PVP-stabilized

palladium(0) nanoclusters formed from the reduction of palladium(II)

acetylacetonate (0.5 mM) by ammonia–borane (200 mM) in the

presence of PVP (2.5 mM).

Fig. 3 X-Ray photoelectron spectrum of PVP-stabilized palladium(0)

nanoclusters isolated from the reduction of palladium(II) acetyl

acetonate (0.5 mM) by ammonia–borane (200 mM) in the presence

of PVP (2.5 mM).

10522 | Phys. Chem. Chem. Phys., 2009, 11, 10519–10525 This journal is �c the Owner Societies 2009

Publ

ishe

d on

06

Oct

ober

200

9. D

ownl

oade

d by

Uni

vers

ity o

f C

alga

ry o

n 27

/09/

2013

13:

10:0

1.

View Article Online

seen from the SAED image shown in Fig. 4b indicating the

presence of palladium in cubic structure.11B-NMR spectra taken from the reaction solution after the

complete methanolysis of AB catalyzed by PVP-stabilized

palladium(0) nanoclusters show a single intense peak at

d E 8.6 ppm which is readily assigned to the tetramethoxy-

borate [B(OCH3)4]� anion.10

Kinetics of the methanolysis of ammonia–borane catalyzed

by in situ-generated PVP-stabilized palladium(0) nanoclusters

As the first control test before starting the kinetic studies, the

methanolysis of AB was performed in the absence of catalyst.

In these test experiments, performed at various temperature in

the range of 15–35 1C, no detectable hydrogen gas evolution

was observed from the self-methanolysis of AB over a period

of 24 h. The in situ-generated PVP-stabilized palladium(0)

nanoclusters are found to be highly active catalysts for the

methanolysis of AB at low concentrations and room temperature.

Fig. 5 shows the plots of the volume of hydrogen generated

versus time during the catalytic methanolysis of 200 mM

H3NBH3 solution in the presence of palladium(0) nanoclusters

in different Pd concentrations (0.50, 0.75, 1.00, 1.25, and

1.50 mM) at 25 � 0.5 1C. The hydrogen generation rate was

determined from the linear portion of the plot for each

experiment. The plot of hydrogen generation rate versus

palladium concentration, both in logarithmic scale (the inset

in Fig. 5), gives a straight line with a slope of 1.17 E 1,

indicating that the methanolysis reaction is first order with

respect to the catalyst concentration.

The effect of AB concentration on the hydrogen generation

rate was also studied by performing a series of experiments

starting with various initial concentrations of H3NBH3 while

keeping the catalyst concentration constant at 0.5 mM Pd.

Fig. 6 shows plots of the volume of hydrogen generated versus

time during the catalytic methanolysis of AB at various AB

concentrations. The hydrogen generation rate in units of mL

H2 min�1 was determined from the linear portion of the plot

for each ammonia–borane concentration and used for

constructing the plot of hydrogen generation rate versus AB

concentration, both in logarithmic scale (the inset in Fig. 6).

The slope of the line given in the inset of Fig. 6 is 0.09 E 0,

indicating that the methanolysis reaction is zero order with

respect to the ammonia–borane concentration. Consequently,

the rate law for the catalytic methanolysis of AB catalyzed by

in situ generated PVP-stabilized palladium(0) nanoclusters can

be given as in eqn (2).

�3d H3NBH3½ �dt

¼ d H2½ �dt¼ k½Pd� ð2Þ

Fig. 4 (a) Powder XRD pattern. (b) Selected area electron diffraction

image of PVP-stabilized palladium(0) nanoclusters.

Fig. 5 The volume of hydrogen versus time plot for the methanolysis

of ammonia–borane catalyzed by PVP-stabilized palladium(0) nano-

clusters depending on the palladium concentration. The inset shows

the plot of hydrogen generation rate versus the concentration of

palladium (both in logarithmic scale) at 25 � 0.5 1C.

Fig. 6 Volume of hydrogen versus time plots depending on the

substrate concentrations at constant catalyst concentration (0.5mM).

The inset shows the plot of hydrogen generation rate versus the

substrate concentration (both in logarithmic scale) at 25 � 0.5 1C.

This journal is �c the Owner Societies 2009 Phys. Chem. Chem. Phys., 2009, 11, 10519–10525 | 10523

Publ

ishe

d on

06

Oct

ober

200

9. D

ownl

oade

d by

Uni

vers

ity o

f C

alga

ry o

n 27

/09/

2013

13:

10:0

1.

View Article Online

Observation of the first order kinetics for the catalytic

methanolysis of AB starting with a known Pd(0)n nanocluster

source is a compelling evidence that Pd(0) nanoclusters are the

true catalyst.

Methanolysis of AB catalyzed by PVP-stabilized palladium(0)

nanoclusters was carried out at different temperatures in the

range 15–35 1C, starting with an initial substrate concentration

of 200 mM H3NBH3 and a catalyst concentration of 1.0 mM

Pd, in order to determine the activation parameters of the

reaction. The values of rate constants, k (Table 2), for the

methanolysis of AB catalyzed by PVP-stabilized palladium(0)

nanoclusters were calculated from the slope of the linear part

of each plot in Fig. 7 and used to calculate the activation

parameters (Arrhenius plot is shown in the inset): Arrhenius

activation energy, Ea = 35 � 2 kJ mol�1; activation enthalpy,

DHa = 33 � 2 kJ mol�1; activation entropy, DSa = �150 �3 J K�1 mol�1.

A catalyst lifetime experiment was performed starting with a

20 mL solution of PVP-stabilized palladium(0) nanoclusters

containing 0.5 mM Pd and 0.5 M H3NBH3 plus 5 equiv. PVP

at 25.0 � 0.5 1C. The in situ-generated PVP-stabilized

palladium(0) nanoclusters provide 23 000 turnovers in the

methanolysis of ammonia–borane over 27 h before deactivation,

which makes them long-lived nanocatalysts. It is noteworthy

that this is the first total turnover number reported for the

catalytic methanolysis of AB.

Mercury poisoning

The ability of Hg(0) to poison heterogeneous metal(0)

catalysts,31 by amalgamating the metal catalyst or being

adsorbed on its surface, has been known for a long time.

The Hg(0) poisoning experiment is easy to perform, but it is

not definitive by itself, nor universally applicable.32 The

suppression of catalysis by Hg(0) is considered as evidence

for heterogeneous catalysis. After about 40% of conversion in

a typical methanolysis experiment at 25.0 � 0.5 1C,

300 equivalents of mercury per palladium was added into

the reaction solution and the progress of reaction was followed

by monitoring the H2 volume, as shown in Fig. 8. The

complete cessation of the catalytic methanolysis of AB upon

mercury addition is additional evidence that Pd(0) nanoclus-

ters are the real catalyst in the methanolysis of AB.

Conclusions

PVP-stabilized palladium(0) nanoclusters were easily generated

in situ during the methanolysis of ammonia–borane from

the reduction of a commercially available precursor. The

reduction of palladium(II) acetylacetonate by ammonia–

borane in the presence of PVP yields nearly spherical

palladium(0) nanoclusters with no agglomeration. The PVP

stabilizer can provide enough stabilization for the

palladium(0) nanoclusters so that they are very stable in

solution under inert atmosphere and yet highly active catalysts

for hydrogen generation from the methanolysis of ammonia–

borane at room temperature. PVP-stabilized palladium(0)

nanoclusters could be isolated as stable solid materials and

well characterized. PVP appears to be a strongly stabilizing

ligand for the palladium(0) nanoclusters, providing relatively

small size and narrow size distribution (3.2 � 0.5 nm).

Although a strong stabilizer such as PVP would sturdily

control the nanocluster formation, the particle size is also

expected to be affected by the experimental conditions such as

temperature and concentration (of substrate, precursor, and

stabilizer). Such a study was considered beyond the scope this

paper. PVP-stabilized palladium(0) nanoclusters are isolable

and re-dispersible. More importantly, when re-dispersed in

methanol, they retain their catalytic activity in the methano-

lysis of ammonia–borane. The kinetics of methanolysis of AB

were reported for the first time on the reaction catalyzed

by PVP-stabilized palladium(0) nanoclusters. Their activity

Fig. 7 The volume of hydrogen versus time plots at different

temperatures for the methanolysis of ammonia–borane catalyzed by

PVP-stabilized palladium(0) nanoclusters in the temperature range

15–35 1C. The inset shows an Arrhenius plot (ln k versus the reciprocal

absolute temperature, 1/T (K�1)).

Table 2 The values of the rate constant, k, for the catalytic methano-lysis of AB starting with a solution of 200 mM H3NBH3 and in situ-generated PVP-stabilized palladium(0) nanoclusters (1.0 mM Pd) atdifferent temperatures calculated from hydrogen volume versus time

T/1C Rate constant, k/mol H2 (mol Pd)�1 s�1

15 0.19520 0.26125 0.32630 0.41935 0.516

Fig. 8 The volume of hydrogen versus time plots for catalytic

methanolysis of 200 mM AB with and without addition of 300 equiv.

Hg(0) at 25 � 0.1 1C.

10524 | Phys. Chem. Chem. Phys., 2009, 11, 10519–10525 This journal is �c the Owner Societies 2009

Publ

ishe

d on

06

Oct

ober

200

9. D

ownl

oade

d by

Uni

vers

ity o

f C

alga

ry o

n 27

/09/

2013

13:

10:0

1.

View Article Online

measured in TOF compares well with those of the catalysts

used previously for the methanolysis of AB (Table 1). With the

exception of expensive ruthenium and rhodium catalysts, they

provide the fastest reaction. Furthermore, PVP-stabilized

palladium(0) nanoclusters are long-lived catalysts in the

methanolysis of AB, providing 23 000 turnovers over 27 h

before deactivation at room temperature. Easy preparation,

high stability, and the high catalytic performance make the

in situ-generated PVP-stabilized palladium(0) nanoclusters

promising catalytic candidates to be employed in developing

highly efficient portable hydrogen generation systems using

AB as a solid hydrogen storage material. Although palladium

metal is more expensive than the first row metals, the

in situ-generated PVP-stabilized palladium(0) nanoclusters

are highly active catalysts in hydrogen generation from the

methanolysis of AB, even in low catalyst concentrations at

room temperature. This makes palladium(0) nanoclusters

attractive as catalysts.

Acknowledgements

Partial support of this work by Turkish Academy of Sciences,

TUBITAK (Project No: 108T840). O.M. thanks the

METU-DPT-OYP Program on behalf of Ataturk University.

References

1 R. Siblerud, Our Future Is Hydrogen: Energy, Environment, andEconomy, New Science Publications, Wellington, USA, 2001.

2 Basic Research Needs for the Hydrogen Economy, Report of theBasic Energy Sciences Workshop on Hydrogen Production,Storage and Use, May 13–15, 2003, Office of Science, US Departmentof Energy, http://www.sc.doe.gov/bes/hydrogen.pdf.

3 L. Schlapbach and A. Zuttel, Nature, 2001, 414, 353.4 N. L. Rosi, J. Eckert, M. Eddaoudi, D. T. Vodak, J. Kim andM. O’Keeffe, Science, 2003, 300, 1127.

5 Annual Energy Outlook 2005 With Projections to 2025, EnergyInformation Administration, February 2005, www.eia.doe.gov/oiaf/aeo/pdf/0383(2005).pdf.

6 G. A. Deluga, J. R. Salge, L. D. Schmidt and X. E. Verykios,Science, 2004, 303, 993.

7 B.-H. Liu and Z.-P. Li, J. Power Sources, 2009, 187, 527.8 H. I. Schlesinger, H. C. Brown, A. B. Finholt, J. R. Gilbreath,H. R. Hockstra and E. K. Hydo, J. Am. Chem. Soc., 1953, 75, 215.

9 C.-L. Hsueh, C.-H. Liu, B.-H. Chen, C.-Y. Chen, Y.-C. Kuo,K.-J. Hwang and J.-R. Ku, Int. J. Hydrogen Energy, 2009, 34,1717.

10 P. V. Ramachandran and P. D. Gagare, Inorg. Chem., 2007, 46, 7810.11 J.-S. Wang and R. A. Geanangel, Inorg. Chim. Acta, 1988, 148, 185.12 F. H. Stephens, R. T. Baker, M. H. Matus, D. J. Grant and

D. A. Dixon, Angew. Chem., Int. Ed., 2007, 46, 746.13 C. A. Jaska, K. Temple, A. J. Lough and I. Manners, J. Am. Chem.

Soc., 2003, 125, 9424.14 M. C. Denney, V. Pons, T. J. Hebden, D. M. Heinekey and

K. I. Goldberg, J. Am. Chem. Soc., 2006, 128, 12048.15 R. J. Keaton, J. M. Blacquiere and R. T. Baker, J. Am. Chem. Soc.,

2007, 129, 1844.16 T. Umegaki, J.-M. Yan, X.-B. Zhang, H. Shioyama, N. Kuriyama

and Q. Xu, Int. J. Hydrogen Energy, 2009, 34, 2303.17 P. A. Storozhenko, R. A. Svitsyn, V. A. Ketsko, A. K. Buryak and

A. V. Ulyanov, Russ. J. Inorg. Chem., 2005, 50, 980.18 (a) M. Zahmakıran and S. Ozkar, Appl. Catal., B, 2009, 89, 104;

(b) O. Metin and S. Ozkar, Energy Fuels, 2009, 23, 3517.19 S. B. Kalidindi, U. Sanyal and B. R. Jagirdar, Phys. Chem. Chem.

Phys., 2008, 10, 5870.20 S. B. Kalidindi, A. A. Vernekar and B. R. Jagirdar, Phys. Chem.

Chem. Phys., 2009, 11, 770.21 M. Chandra and Q. Xu, J. Power Sources, 2007, 168, 135.22 (a) T. Teranishi and M. Miyake, Chem. Mater., 1998, 10, 594;

(b) Y. Li, E. Bone and M. A. El-Sayed, Langmuir, 2002, 18, 4921.23 R. Xiong, J. M. Maclellan, J. Chen, Y. Yin, Z.-Y. Li and Y. Xia,

J. Am. Chem. Soc., 2005, 127, 17118.24 (a) D. Li and S. Komarneni, J. Nanosci. Nanotechnol., 2008, 8,

3930; (b) D. Xu, S. Barbosa and S. G. Yeates, Polym. Preprints,2008, 49, 915.

25 (a) W. Hou, N. A. Dehm and R. W. J. Scott, J. Catal., 2008, 253,22; (b) Y. Wang, M. Du, J. Xu, P. Yang and Y. Du, J. DispersionSci. Technol., 2008, 29, 891; (c) S. Papp and I. Dekany, ColloidPolym. Sci., 2006, 284, 1049.

26 J. A. Widegren, J. D. Aiken, S. Ozkar and R. G. Finke, Chem.Mater., 2001, 13, 312.

27 (a) A. V. Gaikwad and G. Rothenberg, Phys. Chem. Chem. Phys.,2006, 8, 3669; (b) J. A. Creighton and D. G. Eadon, J. Chem. Soc.,Faraday Trans., 1991, 87, 3881.

28 A. Tressaud, S. Khairoun, H. Touhara and N. Watanabe,Z. Anorg. Allg. Chem., 1986, 540, 291.

29 R. M. Rioux, H. Song, P. Yang and G. A. Somorjai, in MetalNanoclusters in Catalysis and Materials Science: The Issue of SizeControl, ed. B. Corain, G. Schmid and N. Toshima, Elsevier,Netherlands, 2008, ch. 7.

30 Palladium, 1969, JCPDS 46-1043.31 G. M. Whitesides, M. Hackett, R. L. Brainard, J. P. Lavalleye,

A. F. Sowinski, A. N. Izumi, S. S. Moore, D. W. Brown andE. M. Staudt, Organometallics, 1985, 4, 1819.

32 J. A Widegren and R. G. Finke, J. Mol. Catal. A: Chem., 2003,198, 317.

This journal is �c the Owner Societies 2009 Phys. Chem. Chem. Phys., 2009, 11, 10519–10525 | 10525

Publ

ishe

d on

06

Oct

ober

200

9. D

ownl

oade

d by

Uni

vers

ity o

f C

alga

ry o

n 27

/09/

2013

13:

10:0

1.

View Article Online