Synthesis and catalytic investigation of organophilic Pd/graphite oxide nanocomposites

Article abstract of DOI:10.1016/j.catcom.2011.10.028

Low-loaded, organophilic Pd/graphite oxide (Pd/GO) nanocomposites were synthesized from different Pd complex precursors by applying graphite oxide as a host material and tetradecyltrimethylammonium bromide (C₁₄TAB) as a stabilizer. Structural investigation of the Pd/GO samples was performed by ICP-AES, XRD, N₂ sorption and TEM measurements. It was found that monodispersed Pd nanoparticles were formed, ranging in size between 1 and 6 nm, both on the external surface and in the interlamellar space of GO. The samples proved to be highly active and selective catalysts for liquid-phase alkyne hydrogenations. The variation in the catalytic performances was attributed to the difference in the amount of interlamellar Pd particles, which participated in the reactions as active sites.

Full text of DOI:10.1016/j.catcom.2011.10.028

Catalysis Communications 17 (2012) 104–107
Contents lists available at SciVerse ScienceDirect
Catalysis Communications
journal homepage: www.elsevier.com/locate/catcom
Short Communication
Synthesis and catalytic investigation of organophilic Pd/graphite
oxide nanocomposites
b
b
b
Á. Mastalir a, , T. Szabó , Z. Király , I. Dékány
⁎
a
Department of Organic Chemistry, University of Szeged, H-6720 Szeged, Dóm tér 8, Hungary
Department of Physical Chemistry and Materials Science, University of Szeged, H-6720 Szeged, Aradi vt. 1, Hungary
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 28 July 2011
Received in revised form 21 October 2011
Accepted 26 October 2011
Available online 4 November 2011
Low-loaded, organophilic Pd/graphite oxide (Pd/GO) nanocomposites were synthesized from different Pd
complex precursors by applying graphite oxide as a host material and tetradecyltrimethylammonium bro-
mide (C14TAB) as a stabilizer. Structural investigation of the Pd/GO samples was performed by ICP-AES,
XRD, N2 sorption and TEM measurements. It was found that monodispersed Pd nanoparticles were formed,
ranging in size between 1 and 6 nm, both on the external surface and in the interlamellar space of GO. The
samples proved to be highly active and selective catalysts for liquid-phase alkyne hydrogenations. The
variation in the catalytic performances was attributed to the difference in the amount of interlamellar Pd
particles, which participated in the reactions as active sites.
Keywords:
Graphite oxide
Palladium
© 2011 Elsevier B.V. All rights reserved.
Nanoparticle
Alkyne hydrogenation
Turnover frequency
Selectivity
1. Introduction
to clay minerals [2]. GO also possesses an excellent intercalation ability
and cation exchange capacity and hence a large number of intercalated
Graphite oxide (GO) is a carbonaceous layer-structured material,
which may be produced by the controlled oxidation of graphite
[1–4]. Since the formation of GO has been first reported [5], several
preparation methods have been developed [6,7]. GO is a two-
dimensional solid in bulk form, with strong covalent bonding within
the layers and weaker interlayer contact between intercalated water
molecules [1,3,8]. Several models have been proposed for the
structure of GO [9–12]. One of the most recent models, introduced
by Lerf et al. [13], describes a GO layer as a random distribution of
flat aromatic regions with unoxidized benzene rings and other
regions of alicyclic six-membered rings with tertiary hydroxyl and
epoxide groups. The graphene oxide sheets are terminated with
carboxyl groups, located only at the edges of the single GO sheets
[13,14]. Unlike most lamellar compounds of graphite, GO cannot be
characterized by a definite empirical formula [4], as GO is a nonstoichio-
meric compound and its composition depends on the synthesis condi-
tions [3,15]. Further, GO is strongly hygroscopic and tends to
decompose above 333 K [15]. GO is a hydrophilic material, which may
be readily dispersed in water to form stable colloidal suspensions
[1,3]. Self-assembled films can be prepared from diluted dispersions,
as the GO lamellae are capable of parallel orientation [16]. As a result
of liquid sorption, GO exhibits swelling and disaggregation, similarly
GO materials have been synthesized and investigated [8,17,18]. For the
preparation of metal nanoparticles on graphite oxide or graphene-
derived materials, several methods have been reported in the literature
[19–22], including the photochemical loading of metal nanoparticles on
reduced graphene oxide sheets [21] and a special heating procedure
applied for metal complexes and GO [22]. In the present study, GO has
been utilized as a host material for the formation of Pd nanoparticles
of controlled particle size. The structure of the Pd/GO materials has
been investigated and the samples were tested as catalysts for the
liquid-phase hydrogenations of terminal and internal alkynes under
standard conditions. For the latter reactions, Pd has been recognized
as the most selective metal [23], and Pd nanoparticles supported on
layer-structured materials have been found to be particularly efficient
catalysts [24,25].
2. Experimental
2.1. Preparation of Pd/GO from Pd(NH3)4(NO3)2
1 g of finely powdered GO was dispersed in 100 cm3 of distilled
water. After adjusting the pH to 10, the suspension was left under
stirring overnight. Pd nanoparticles were generated by using the
precursor Pd(NH3)4(NO3)2 and the cationic surfactant tetradecyl
trimethylammonium bromide (C14TAB). In order to ensure the com-
plete hydrophobization of GO, an excess of C14TAB was applied
⁎
Corresponding author. Tel.: +36 62 544207; fax: +36 62 544200.
E-mail address: mastalir@chem.u-szeged.hu (Á. Mastalir).
1566-7367/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.catcom.2011.10.028

Á. Mastalir et al. / Catalysis Communications 17 (2012) 104–107
105
(nC14TAB =130 nPd). The aqueous solution of C14TAB was added to the
aqueous solution of the precursor and the mixture was left under stir-
ring for 2 h. After reduction of the precursor by an excess of NaBH4,
the solution was added dropwise to the GO suspension. This resulted
in the formation of an organophilic Pd/GO nanocomposite material.
Stirring was maintained for 48 h and then the solid was
centrifuged and purified by washing with a pH=9 NaOH solution 3
times. The sample (Pd/GO1) was dried in an oven at 333 K overnight.
flame ionization detector (FID). The experimental error was 2% for re-
peated runs.
3. Results and discussion
The specific surface areas of the Pd/GO samples were very low
(2.5 m2 g−1 and 3.5 m2 g−1 for Pd/GO1 and Pd/GO2, respectively),
indicating that the interlayers of organophilic GO were highly loaded
with alkyl chains, as related to the intercalation of C14TAB molecules,
and therefore N2 adsorption took place on the external surfaces of the
GO particles.
2.2. Preparation of Pd/GO from K2PdCl4
The GO suspension was prepared in the same way as mentioned
above. Pd nanoparticles were obtained by using the precursor
K2PdCl4 and the cationic surfactant C14TAB, by applying an excess of
surfactant (nC14TAB =130 nPd). The precursor was dissolved in
10 cm3 of distilled water and then the aqueous solution of the
surfactant was added, resulting in the formation of a bright orange
solution, which was left under stirring for 2 h. Reduction of the pre-
cursor was effected by NaBH4, upon which a dark Pd sol was obtained.
This was added dropwise to the GO suspension and then the mixture
was left under stirring for 48 h. The sample (Pd/GO2) was purified
and dried in the same way as described above. For a similar synthesis
procedure, applied for Pd/hydrotalcites [25], it has been found that
the effect of residual surfactant molecules on Pd was negligible and
hence it was unlikely to affect the catalytic performance.
The pore size distribution could not be determined, as GO was
found to be a nonporous material.
The XRD patterns of GO and the Pd/GO nanocomposites are dem-
onstrated in Fig. 1. For the GO host, a well-defined peak was obtained
at 2Θ=12.9°, which corresponds to an interplanar distance (dL) of
0.69 nm. The characteristic d002 reflection of graphite at 26.6° could
not be observed, which confirmed that complete oxidation took
place [27]. As a result, a well-ordered, lamellar structure was formed,
which was more open than that of graphite (dL =0.336 nm) and thus
more susceptible to intercalation [27]. The pattern of Pd/GO1 is com-
prised of three peaks at 2Θ=7.82, 9.08, and 18.16°. The latter peaks
characterize the same basal spacing, 0.97 nm. The first reflection,
appearing with smaller intensity, belongs to another phase, in
which the c-axis repeat distance is 1.13 nm. These distances indicate
that both phases comprise of surfactant-stuffed interlayers with
slightly different amounts or conformations of the intercalated mole-
cules. Pd/GO2 shows a similar diffractogram, but the first two reflec-
tions are markedly shifted toward lower diffraction angles (with d-
values of 1.4 and 1.2 nm). On the other hand, the position of the
2Θ=18° (d=0.5 nm) peak remained unchanged. Since the related
distance is too low to be accounted for a first-order reflection of GO
(d001 =0.66 nm for the air-dry sample, whereas the lowest values
for anhydrous Brodie-GOs are ca. 0.59 nm [15]), we still assign this
peak as a second-order reflection. The fact that the peaks no longer
occur at integral d-values suggests that randomly interstratified
structures were formed, that is, the structure, conformation and
interlayer concentration of surfactant molecules were different. Nev-
ertheless, the relatively low interlamellar distance implies that the
density of accommodated surfactant molecules is not too high and
the molecules lie flat and take a monolayer arrangement [17]. The
XRD patterns do not give conclusive evidence whether the Pd nano-
crystals are inserted between the GO layers, because the Pd loadings
are lower than the detection limit for ultrafine crystallites generally
exhibiting significant broadening and concomitant intensity loss, to
which interstratification may also contribute.
2.3. Structural characterization
The Pd contents of the samples were obtained from inductively
coupled plasma atomic emission spectroscopy (ICP-AES), by using a
Jobin Yvon 24 sequential ICP-AES spectrometer at 229.7 and
324.3 nm. The Pd contents for Pd/GO1 and Pd/GO2 (0.18% and
0.15%, respectively) were only slightly lower than the nominal Pd
loading (0.20%).
The specific surface areas of the Pd/GO samples were determined
by N2 sorption measurements, performed by using a Micromeritics
2375 BET apparatus at 77 K.
X-ray diffraction (XRD) measurements were performed with a
Philips PW 1820 diffractometer (40 kV, 35 mA, CuKα radiation). The
basal spacings of the samples were determined from the first-order
Bragg reflections, by applying a PW 1877 software.
TEM measurements were effected by using a Philips CM10 trans-
mission electron microscope, equipped with a LaB6 cathode and a
Megaview II digital camera. The particle size distributions and the
mean particle diameters were determined as Σnidi/Σni (Σni >250),
by using AnalySIS 3.1. software. The dispersions of the catalysts
were calculated from the mean particle diameters d, as D=0.885/d
[26].
14000
2.4. Catalytic test reaction
GO
12000
Pd/GO1
The Pd/GO materials were tested as catalysts in an automated
hydrogenation apparatus [25], for the liquid-phase hydrogenations of
1-pentyne, 1-hexyne, 3-hexyne, 1-heptyne and 1-phenyl-1-butyne.
For each measurement, 5 mg of sample was used. The reactions were
investigated at 298 K and 105 Pa, with a reactant:Pd (R:Pd) ratio of
2500. Pretreatment of the catalyst was carried out in static hydrogen
at 298 K for 60 min, followed by the addition of solvent (1 cm3 of
toluene), and stirring for 45 min. All reactions were performed under
vigorous stirring (1400 rpm) to eliminate mass transport limitations.
The hydrogen consumption was recorded as a function of reaction
time. The reaction rates were determined from the slopes and the turn-
over frequencies (TOF) were calculated as TOF=7.2538·10−5 RS/D
[s−1], where RS is the initial rate and D is the dispersion of the catalyst.
Quantitative analysis of the products was performed with an HP 5890
gas chromatograph equipped with a HP-I capillary column and a
Pd/GO2
10000
8000
6000
4000
2000
0
5
10
15
20
25
30
2Θ [°]
Fig. 1. XRD patterns of GO, Pd/GO1 and Pd/GO2.

106
Á. Mastalir et al. / Catalysis Communications 17 (2012) 104–107
100 nm
100 nm
Fig. 2. TEM images of the Pd/GO nanocomposites.
Figs. 2 and 3 display the TEM images and the particle size distribu-
of terminal and internal alkynes over Pd/GO1 are summarized in
Table 1. It may be observed that Pd/GO1 was a highly active catalyst
for alkyne hydrogenations. For the transformations of terminal alkynes,
the reaction rates and the turnover frequencies displayed no systematic
variation with the reactant size. The highest reaction rate was obtained
for 3-hexyne. On the other hand, the reaction rate for the hydrogenation
of 1-phenyl-1-butyne was considerably lower, and decreased with the
reaction time. This finding may be regarded as an indirect evidence
for the formation of interlamellar Pd crystallites in the GO host [25], as
the latter crystallites could not be identified by XRD. Such interlamellar
clusters may also be regarded as active sites and tend to be less accessi-
ble for 1-phenyl-1-butyne, comprising an aromatic ring and an internal
alkyne group [19,20]. The selectivity toward the formation of the main
products alkene and cis-alkene exceeded 90% for each reactant and
was nearly 100% for the reactions of 1-pentyne and 3-hexyne. This im-
plies that overhydrogenation and the formation of by-products (either
2-alkene or trans-alkene) were negligible. It may therefore be estab-
lished that Pd/GO1 was a particularly efficient catalyst for the hydroge-
nations of aliphatic alkynes, for which high conversions and alkene
selectivities were achieved at short reaction times.
The data obtained for the catalytic investigation of Pd/GO2 for the
same reactions are displayed in Table 2. Although the Pd contents of
the Pd/GO nanocomposites were very similar, the initial rates and
the TOF values for Pd/GO2 were considerably lower than those for
Pd/GO1. This suggested limited access for the reactant molecules to
the active centers in the interlamellar space of GO. The highest con-
versions were obtained for the hydrogenations of lower alkynes,
whereas Pd/GO2 proved to be a less efficient catalyst for the reactions
of 1-heptyne and 1-phenyl-1-butyne. The catalytic performance of
Pd/GO2 decreased with increase of the reactant size. It may therefore
be suggested that a significant proportion of the catalytically active Pd
nanoparticles was situated in the interlamellar space of GO and the
latter particles were less accessible for the reactants than those on
the external surface of the GO host. The selectivities for the formation
of the main products were similar to those obtained for Pd/GO1.
Overhydrogenation was insignificant for each reactant. For the hydro-
genations of terminal alkynes, small amounts of 2-alkenes were
tions of the Pd/GO samples, respectively. The formation of the Pd
nanoparticles had no effect on the lamellar structure of the GO host
material. A large number of quasi-spherical surface Pd particles can
be distinguished, for which aggregation did not occur. For Pd/GO1,
the particle diameters fell in the range 1–6 nm. The highest relative
frequencies were observed for the 2–4 nm Pd particles (77%), where-
as the number of Pd particles smaller than 2 nm and larger than 4 nm
was negligible. This finding confirms the steric stabilizing effect of the
cationic surfactant molecules, which prevented particle aggregation
[24,25]. The mean particle diameter of the Pd particles was 3.6 nm
and the dispersion of the sample was 25%. The Pd particles for Pd/
GO2 proved to be small, quasi-spherical and monodispersed, similarly
as those for Pd/GO1. The majority of the particles (86%) fell in the
range 1–3 nm. The mean particle diameter was 2.5 nm and the
dispersion was 36%. The application of different precursors may result
in the formation of different species, which may affect the reduction
procedure. Previous studies for K2[PdCl4] indicated that the applica-
tion of a large excess of C14TAB resulted in a ligand exchange reaction
with free Br− [28,29], resulting in the formation of (C14TA)2[PdBr4],
attached to free C14TAB micelles. On the other hand, for the precursor
Pd(NH3)4(NO3)2, an electrostatic interaction with Br− anions and a
ligand exchange reaction of NH3 with the C14TA+ may be suggested.
As reported for positively charged Pd(II) complexes bound to anionic
surfactant micelles, both electrostatic and hydrophobic interactions
are to be considered in the binding process [30].
The catalytic performances of the Pd/GO nanocomposites were in-
vestigated in liquid-phase alkyne hydrogenations, typically regarded
as zero-order reactions, for which the reaction rate is independent
of the reactant concentration [31,32].
This implies that the initial rate remains constant until all the reac-
tant has been transformed. The results obtained for the hydrogenations
60
Pd/GO1
50
40
30
20
10
0
Pd/GO2
Table 1
Hydrogenations of alkynes over Pd/GO1.
Reactant
Time
[min]
R
TOF
Conversion
[%]
S
a
[cm3 min−1 g−1
]
[s−1
]
[%]b
1-pentyne
1-hexyne
3-hexyne
1-heptyne
21
33
43
21
69
24,475
17,529
31,407
26,726
3287
7.1
5.1
9.1
7.8
1.0
97.0
97.6
91.1
96.6
27.8
99.5
97.6
99.9
90.6
94.6
1-phenyl-1-butyne
1-2
2-3
3-4
4-5
5-6
a
Reaction rate.
particle diameter [nm]
b
Selectivity of the main reaction product (1-alkene and cis-alkene for terminal and
Fig. 3. Particle size distributions of the Pd/GO nanocomposites.
internal alkynes, respectively).

Á. Mastalir et al. / Catalysis Communications 17 (2012) 104–107
107
Table 2
particularly efficient catalyst, which provided high conversions and
alkene selectivities at short reaction times.
Hydrogenations of alkynes over Pd/GO2.
Reactanta
R [cm3 min−1 g−1
]
TOF [s−1
]
Conversion [%] S [%]c
b
1-pentyne
1-hexyne
3-hexyne
1-heptyne
13,653
14,034
8310
6148
1045
2.8
2.8
1.7
1.2
0.2
78.6
88.1
63.5
14.7
6.4
98.2
95.1
98.3
100
100
Acknowledgment
Financial support provided by the TÁMOP-4.2.1/B-09/1/KONV-
2010-0005 project is gratefully acknowledged.
1-phenyl-1-butyne
a
t=45 min.
Reaction rate.
b
References
c
Selectivity of the main product (1-alkene or cis-alkene).
[1] S. Stankovich, R.D. Piner, S.T. Nguyen, R.S. Ruoff, Carbon 44 (2006) 3342–3347.
[2] I. Dékány, R. Krüger-Grasser, A. Weiss, Colloid & Polymer Science 276 (1998)
570–576.
[3] T. Szabó, E. Tombácz, E. Illés, I. Dékány, Carbon 44 (2006) 537–545.
[4] R.J. Beckett, R.C. Croft, The Journal of Physical Chemistry 56 (1952) 929–934.
[5] B. Brodie, Annals Chimie Physical 59 (1860) 466–472.
[6] L. Staudenmaier, Berichte der Deutschen Chemischen Gesellschaft 31 (1898)
1481–1487.
[7] W.S. Hummers, R.E. Offemann, Journal of the American Chemical Society 80
(1958) 1339-1339.
[8] Z. Liu, Z.M. Wang, X. Yang, K. Ooi, Langmuir 18 (2002) 4926–4932.
[9] A. Clauss, R. Plass, H.P. Boehm, U. Hofmann, Zeitschrift fur Anorganische und All-
gemeine Chemie 291 (1957) 205–220.
[10] G. Ruess, Monatshefte fur Chemie 76 (1946) 381–417.
[11] M. Mermoux, Y. Chabre, A. Rousseau, Carbon 29 (1991) 469–474.
[12] T. Nakajima, Y. Matsuo, Carbon 32 (1994) 469–475.
[13] H. He, T. Riedl, A. Lerf, J. Klinowski, The Journal of Physical Chemistry 100 (1996)
19954–19958.
[14] S. Stankovich, R.D. Piner, X. Chen, N. Wu, S.T. Nguyen, R.S. Ruoff, Journal of Mate-
rials Chemistry 16 (2006) 155–158.
[15] T. Szabó, O. Berkesi, P. Forgó, K. Josepovits, Y. Sanakis, D. Petridis, I. Dékány,
Chemistry of Materials 18 (2006) 2740–2749.
detected as by-products, formed by isomerization [23]. The results of
the catalytic test reactions confirmed that Pd/GO1, obtained from the
precursor Pd(NH3)4(NO3)2, was a more efficient catalyst than Pd/
GO2, prepared from K2PdCl4. As the structures of the Pd/GO nano-
composites and the size distributions of their surface Pd particles
were very similar, the difference in their catalytic performance may
be related to their interlamellar Pd content. The evidence that the
particle diameters for Pd/GO2 were smaller than those for Pd/GO1
(Fig. 3) indicated that Pd/GO2 was more likely to accommodate an
enhanced amount of interlamellar Pd particles. It may be established
that both the surface and interlamellar Pd particles participated in the
reactions as active sites. The catalytic behavior of Pd/GO1 was basical-
ly determined by its Pd content located on the external surface of GO,
and hence this sample may be regarded as a GO-supported Pd cata-
lyst. On the other hand, an enhanced amount of interlamellar Pd
nanocrystals, as suggested for Pd/GO2, tends to decrease the catalytic
performance. For the transformation of 3-hexyne, the Pd/GO samples
proved to be more efficient than the conventional supported Pd cata-
lysts [20], as the former samples provided better alkene selectivities
at high conversions.
[16] N.A. Kotov, I. Dékány, J.H. Fendler, The Journal of Physical Chemistry 99 (1996)
13065–13069.
[17] Y. Matsuo, T. Niwa, Y. Sugie, Carbon 37 (1999) 897–901.
[18] Y. Matsuo, T. Miyabe, T. Fukutsuka, Y. Sugie, Carbon 45 (2007) 1005–1012.
[19] Á. Mastalir, Z. Király, M. Benkő, I. Dékány, Catalysis Letters 124 (2008) 34–38.
[20] Á. Mastalir, Z. Király, Á. Patzkó, I. Dékány, P. L'Argentiere, Carbon 46 (2008)
1631–1637.
[21] G. Moon, Y. Park, W. Kim, W. Choi, Carbon 49 (2011) 3454–3462.
[22] K. Gotoh, T. Kinumoto, E. Fujii, A. Yamamoto, H. Hashimoto, T. Ohkubo, A. Itadani,
Y. Kuroda, H. Ishida, Carbon 49 (2011) 1118–1125.
[23] Á. Molnár, A. Sárkány, M. Varga, Journal of Molecular Catalysis A: Chemical 173
(2001) 185–221.
[24] Á. Mastalir, Z. Király, Gy Szöllősi, M. Bartók, Journal of Catalysis 194 (2000)
146–152.
[25] Á. Mastalir, Z. Király, Journal of Catalysis 220 (2003) 372–381.
[26] P.C. Aben, Journal of Catalysis 10 (1968) 224–229.
4. Conclusions
Graphite oxide was utilized as a host material for the generation of
Pd nanoparticles with controlled particle size. Low-loaded, organo-
philic Pd/GO nanocomposites were prepared by using different com-
plex precursors and C14TAB as a stabilizer. Structural investigation
confirmed that an efficient particle size control was achieved and
the formation of Pd nanoparticles had no effect on the structure of
the GO host. It was found that Pd nanoparticles were formed both
on the external surface and in the interlamellar space of GO. The Pd/
GO samples were tested as catalysts in the liquid-phase hydrogena-
tions of terminal and internal alkynes. Both samples proved to be ef-
ficient catalysts for the reactions of lower alkynes. As compared with
Pd/GO1, Pd/GO2 was suggested to comprise an enhanced amount of
interlamellar Pd particles, which were less accessible for reactants
than those on the external surface of GO. Pd/GO1 proved to be a
[27] C. Hontoria-Lucas, A.J. López-Peinado, J. de López-González, M.L. Rojas-Cervantes,
R.M. Martín-Aranda, Carbon 33 (1995) 1585–1592.
[28] B. Veisz, Z. Király, L. Tóth, B. Pécz, Chemistry of Materials 14 (2002) 2882–2888.
[29] L. Bisson, C. Boissiere, L. Nicole, D. Grosso, J.P. Jolivet, C. Thomazeau, D. Uzio, G.
Berhault, C. Sanchez, Chemistry of Materials 21 (2009) 2668–2678.
[30] F.P. Cavasino, C. Sbriziolo, M. Cusumano, A. Gianetto, Journal of the Chemical So-
ciety, Faraday Transactions 92 (1996) 2263–2268.
[31] G. Carturan, G. Facchin, G. Cocco, S. Enzo, G. Navazio, Journal of Catalysis 76
(1982) 405–417.
[32] C.A. Hamilton, S.D. Jackson, G.J. Kelly, R. Spence, D. de Bruin, Applied Catalysis A:
General 237 (2002) 201–209.