Palladium nanoparticles on dendrimer-containing supports as catalysts for hydrogenation of unsaturated hydrocarbons

Article abstract of DOI:10.1134/S0965544112050052

A new method has been proposed for encapsulation of palladium nanoparticles with a size of up to 2.5 nm in the matrix of special supports, the polymer networks based on poly(propylene imine) dendrimers, synthesized for this purpose. It has been shown that the particle size distribution of the materials obtained and their catalytic activity in the hydrogenation reactions of unsaturated compounds substantially depend on the specific features of the support structure. A high activity (TOF up to 15000 h^-1) has been observed in the hydrogenation of styrene. The catalysts can be repeatedly used without loss of activity. Pleiades Publishing, Ltd., 2012.

Full text of DOI:10.1134/S0965544112050052

ISSN 0965ꢀ5441, Petroleum Chemistry, 2012, Vol. 52, No. 5, pp. 289–298. © Pleiades Publishing, Ltd., 2012.
Original Russian Text © E.A. Karakhanov, A.L. Maksimov, A.V. Zolotukhina, S.V. Kardashev, T.Yu. Filippova, 2012, published in Neftekhimiya, 2012, Vol. 52, No. 5, pp. 323–332.
Palladium Nanoparticles on DendrimerꢀContaining Supports
as Catalysts for Hydrogenation of Unsaturated Hydrocarbons
E. A. Karakhanov, A. L. Maksimov, A. V. Zolotukhina, S. V. Kardashev, and T. Yu. Filippova
Faculty of Chemistry, Moscow State University, Moscow, Russia
eꢀmail: kar@petrol.chem.msu.ru
Received April 11, 2012
Abstract—A new method has been proposed for encapsulation of palladium nanoparticles with a size of up
to 2.5 nm in the matrix of special supports, the polymer networks based on poly(propylene imine) dendrimꢀ
ers, synthesized for this purpose. It has been shown that the particle size distribution of the materials obtained
and their catalytic activity in the hydrogenation reactions of unsaturated compounds substantially depend on
the specific features of the support structure. A high activity (TOF up to 15000 h–1) has been observed in the
hydrogenation of styrene. The catalysts can be repeatedly used without loss of activity.
DOI: 10.1134/S0965544112050052
Catalysts that contain metal nanoparticles as active tion mode with an HATR Multiꢀreflection accessory
components have found wide application in petroꢀ using a ZnSe 45
crystal for different wavelength
°
chemical and organic synthesis. They are used in douꢀ
bleꢀbond hydrogenation, hydrodechlorination, cross
coupling, and other reactions [1–5]. As a rule, the
synthesis of heterogeneous catalysts of this type is
fraught with the necessity of reaching a regular
arrangement of nanoparticles on the support and a
narrow particleꢀsize distribution [5–8].
One of the most interesting approaches to the synꢀ
thesis of catalysts based on metal nanoparticles is the
use of dendrimers as a stabilizing agent [9–12]. We
have proposed a method for synthesizing heterogeꢀ
neous supports from poly(propylene imine) dendrimꢀ
ers by their bonding to one another with special
crosslinking agents, such as diepoxides and diisocyanꢀ
ates. Using this method, we prepared effective hydroꢀ
genation catalysts based on crosslinked dendrimers
and ruthenium nanoparticles [13]. The material
exhibited a high activity in hydrogenation of double
bonds of olefins and aromatic compounds.
ranges with 4ꢀnm resolution. An Xꢀray photoelectron
spectroscopy (XPS) study of the samples was perꢀ
formed with a LASꢀ3000 instrument equipped with an
OPXꢀ150 retardingꢀfield photoelectron analyzer.
Photoelectrons were excited by Xꢀrays from an alumiꢀ
num anode (Al
K = 1486.6 eV) at a tube voltage of
α
12 kV and an emission current of 20 mA. Photoelecꢀ
tron peaks were calibrated by the C 1s carbon line with
a binding energy of 285 eV.
Palladium in the samples was quantified by inducꢀ
tively coupled plasma atomic emission spectroscopy
(ICPꢀAES) using an IRIS Interpid II XPL instrument
(Thermo Electron, the United States) with radial and
axial detection at wavelengths of 310 and 95.5 nm.
Transmission electron microscopy (TEM) studies of
the samples were performed with a LEO912 AB
Omega electron microscope.
In this paper, we report some results on the design
of catalysts based on palladium nanoparticles using the
same approach.
Synthesis of Catalysts
Synthesis of DAB3ꢀHMDI polymer. A 125ꢀmg porꢀ
tion of DAB(NH2)16 (0.074 mmol) and 50 ml of absoꢀ
lute THF were placed into a 100ꢀml oneꢀneck flask
equipped with a magnetic stirrer and a reflux conꢀ
denser. Under these conditions, the dendrimer swelled
EXPERIMENTAL
Firstꢀ and thirdꢀgeneration poly(propylene imine)
dendrimers DABꢀPPIꢀG1 (DAB(NH2)4) and DABꢀ
PPIꢀG3 (DAB(NH2)16) synthesized according to the
procedure described in [14] were used as starting
materials. The crosslinking agents were 1,6ꢀhexameꢀ
thylene diisocyanate (HMDI) and 1,4ꢀbutylene diisoꢀ
cyanate (BDI), both from Aldrich.
but did not dissolve. Then, 63
amethylene diisocyanate was added with stirring. The
reaction was run at 70 over 12 h. The product mixꢀ
µ
L (0.4 mmol) of hexꢀ
°С
ture was evaporated to dryness in a rotary evaporator,
the residue was washed with water and methanol, cenꢀ
trifuged, and dried in a rotary evaporator at 50 С. The
°
IR spectra of the samples were recorded on a Nicoꢀ product was obtained in the form of ocher colored,
let IRꢀ200 spectrometer in the attenuated total reflecꢀ loose, sticky lumps. The product yield was 191 mg.
289

290
KARAKHANOV et al.
Synthesis of DAB3ꢀBDI polymer. The synthesis was
conducted according to the same procedure. The 2.9%), 340.9 (Pd 3d3/2
reactants were DAB(NH2)16 (500 mg, 0.3 mmol) and 10.7%); 531.9 (O 1s, 9.0%).
3d5/2, 2,
butylene diisocyanate (0.11 L, 0.79 mmol) in 50 ml
of absolute THF. The product was ocher colored,
loose, sticky lumps. The yield was 1.25 g.
IR, cm–1: 3300 (N–Hst in NH–C(=O)); 2930 (C–
Hst); 2850 (C–Hst, CH2–Nst); 1750, 1620, 1570
µ
ICPꢀAES: 8.87% Pd.
Synthesis of DAB3ꢀBDIꢀPd. The reactant was
122 mg of DAB3ꢀBDI in 25 mL of absolute chloroꢀ
form. To this slurry, 133 mg (0.59 mmol) of pallaꢀ
dium(II) acetate was added with stirring. The reaction
(C=Ost in NH–C(=O); 1480, 1460 (N–H , CH2
,
δ
was run at 70 С for 12 h. At the end of the reaction, the
°
δ
X⎯N–H ); 1420 (CH2 ); 1380, 1340, 1270, 1240,
slurry was evaporated to dryness in a rotary evaporator.
The intermediate product, which was a dark brown
powder in an amount of 275 mg, was then suspended
in a mixture of 30 mL of chloroform and 10 mL of
methanol. Sodium borohydride (225 mg, 5.9 mmol)
was added to the resulting suspension and allowed to
δ
δ
1160, 1030 (C–Nst in NH–C(=O); 990, 930, 860, 795
(N–H ); 750 (CH2 )
.
δ
γ
Synthesis of DAB3ꢀHMDI2 polymer. The reaction
was conducted according to the same procedures as
described above. The reactants were DAB(NH2)16
(1.75 g, 1.04 mmol) and hexamethylene diisocyanate
(4 mL, 25.5 mmol) in 60 ml of absolute THF. The
product was a light yellow powder. The yield was 5 g
(83%).
react for 12 h at 60 С. The yield of the dry product was
°
75 mg.
XPS, eV: 285.0 (C
.4%); 340.9 (Pd 3 3/2); 398.7, 402.6 (N
530.6 (O 1 , 21.9%); 534.6 (Pd p3/2, 9.0%).
ICPꢀAES: 22.58% Pd.
Synthesis of DAB1ꢀHMDI2ꢀPd
1
s, 63.0%); 336.1 (Pd
3d5/2,
, 7.7%);
7
d
1s
IR, cm–1: 3334 (N–Hst in NH–C(=O)); 2934 (C–
Hst); 2587 (C–Hst, CH2–Nst); 1685, 1619, 1573
CH2 X–N–
,
(C=Ost in NH–C(=O); 1478 (N–H
,
δ
δ
500 mg of DAB1ꢀHMDI2 and 50 mL of absolute chloꢀ
roform. To this suspension, 160 mg (0.7 mmol) of
Pd(OAc)2 was added with stirring. The reaction was
H ; 1420 (CH2 ); 1372, 1252, 1215, 1145, 1039 (C–
δ
Nδst in NH–C(=O); 771 (N–H ); 735 (CH2 )
.
δ
γ
XPS, eV: 1193.1 (C–N), 11.99.2 (C=O) (
Ekin C 1s,
run at 70 С for 12 h. At the end of the reaction, the
°
65.0%), 404.3 (N 1s, 20.9%), 944.1 (O–N), 949.1
(C=O) ( kin O 1s, 14.1%).
suspension was evaporated in a rotary evaporator. The
resulting dark brown powder obtained in an amount of
745 mg (100% yield) was placed in a flask and susꢀ
pended in a mixture of 35 mL of methanol and 25 mL
of chloroform. Sodium borohydride in an amount of
266 mg (7 mmol) was added by portions with stirring
E
Synthesis of DAB1ꢀHMDI2 polymer. The reaction
was conducted according to the same procedures. The
reactants were DAB(NH2)4 (2 g, 6.33 mmol) and hexꢀ
amethylene diisocyanate (6 mL, 37.5 mmol) in 40 mL
of absolute THF. The yield was 7 g (84%).
to this slurry. The reaction was conducted at 70 С over
°
IR, cm–1: 3335 (N–Hst in NH–C(=O)); 2928 (C–
12 h. The yield of the dry product was 355 mg (62%).
XPS, eV: 285.8 (C 1s, 58.8%); 335.8 (Pd 3d5/2,
4
.
The reactants were
Pd(OAc)2 in 50 mL of chloroform. The yield of the
product DAB3ꢀHMDI2ꢀPdII was 1.31 g (99%).
XPS, eV: 284.8 (C
.4%), 341.6 (Pd 3d3/2
s, 13.9%). ICPꢀAES: Pd, 9.37%.
1
s
, 63.4%); 335.8 (Pd
3d5/2,
5
1
)
, 399.4 (N , 16.5%); 531.5 (O
1s
12 h. At the end of the reaction, the suspension was
cooled to room temperature and separated from chloꢀ
roform by centrifuging. The resulting dry powdered
black precipitate obtained in an amount of 199 mg was
Catalyst Testing Procedure
A catalyst and a substrate taken in the ratio of 1 mg
suspended in a mixture of 9.5 mL of methanol and of the catalyst to 2.2–2.3 mmol of the substrate were
28 mL of chloroform. Sodium borohydride in an placed in a thermostated steel autoclave equipped with
amount of 225 mg (5.9 mmol) was added with stirring an insert test tube and a magnetic stirrer. A solvent was
to this slurry. The reaction mixture immediately added when necessary. The autoclave was tightly
turned black, and violent gas evolution was observed. sealed, filled with hydrogen to have a pressure of 10 or
The reaction was conducted at 60
end of the reaction, the precipitate was separated from tion was run with vigorous stirring at 80
°
С
over 12 h. At the 40 atm, and connected with a thermostat. The reacꢀ
for 1 h or 30
°С
the solvents by centrifuging, washed with water and or 15 min; after that, the autoclave was cooled down
methanol, and dried. The yield of the product was below room temperature. The products were deterꢀ
169 mg.
mined by GLC on a ChromPack CP9001 gas chroꢀ
PETROLEUM CHEMISTRY Vol. 52
No. 5
2012

PALLADIUM NANOPARTICLES ON DENDRIMERꢀCONTAINING SUPPORTS
Table 1. Supports and catalysts synthesized in this work
Dendrimer Crosslinker NCO/NH2 ratio
291
Catalyst notation
Metal content, %
DAB(NH )
OCN(CH ) NCO
NCO/NH = 2/1
DAB ꢀHMDI ꢀPd
12.06
9.97
2 4
2 6
2
1
2
DAB(NH )
OCN(CH ) NCO
NCO/NH = 2/1
DAB ꢀHMDI ꢀPd
3 2
2 16
2 6
2
DAB(NH )
OCN(CH ) NCO
NCO/NH = 1/3
DAB ꢀHMDIꢀPd
8.87
2 16
2 6
2
3
DAB(NH )
OCN(CH ) NCO
NCO/NH = 1/3
DAB ꢀBDIꢀPd
22.88
2 16
2 4
2
3
matograph with a flameꢀionization detector using a the cross linkers would have an effect on the activity
30 m × 0.2 mm column with the bonded SEꢀ30 phase.
Chromatograms were recorded and processed with the
use of the Maestro 1.4 software.
and selectivity of the catalysts, as well as their polarity.
To compare the effect of the size of the reactant denꢀ
drimer on the properties of the catalysts, two supports
based on DAB(NH2)4 and DAB(NH2)16 with HMDI
taken in a ratio of NCO/NH2 = 2 : 1 were synthesized.
To compare the dependence of the properties of the
final hybrid material on the size of the crosslinking
entity, DAB(NH2)16ꢀbased materials with an
NCO/NH2 ratio of 1 : 3 were synthesized (Table 1).
RESULTS AND DISCUSSION
To prepare network polymer matrices as supports
for palladium nanoparticles, we used the firstꢀ and
thirdꢀgeneration poly(propylene imine) (DAB) denꢀ
drimers DAB(NH2)4 and DAB(NH2)16, respectively.
The crosslinking agents were 1,4ꢀbutylene diisocyanꢀ
The IR and XPS data suggest that the process folꢀ
lows the conventional mechanism of the reaction of
isocyanates with amines yielding –NH–C(=O)–
ate OCN(CH2)4NCO
diisocyanate OCN(CH2)6NCO
that form conformationally labile links between the
(
BDI
)
and 1,6ꢀhexamethylene
(
HMDI , compounds
)
dendrimers. It was assumed that the size and rigidity of NH– fragments (Scheme 1).
HN
O
O
HN
O
NH
H
N
N
N
O
O
O
H
N
NH
O
N
N
H
O
NH
N
N
O
N
N
HN
O
HN
H
N
N
H
N
NH
N
N
O
O
N
N
O
O
HN
O
O
O
N
N
H
N
N
O
HN
NH
O
O
HN
N
O
N
O
HN
NH
NH
NH2
O
H2N
NH
O
N
O
HN
O
NH
HN
N
HN
O
O
O
NCO
NH
O
OCN
N
HN
N
O
N
N
N
N
N
THF, 70°C, NCO/NH = 2/1
2
NH
HN
HN
NH
N
NH
O
O
N
O
N
N
O
O
O
NH2
N
H2N
HN
O
HN
O
O
O
NH
HN
NH
HN
HN
HN
N
N
O
O
NH
O
N
N
HN
N
N
O
O
O
O
N
HN
O
N
NH
NH
NH
NH
N
O
O
HN
N
O
HN
O
O
H
N
O
N
HN
O
N
N
O
N
N
H
H
H
N
N
N
N
N
H
O
O
N
O
O
O
HN
NH
N
HN
NH
O
Scheme 1. Synthesis of polymer materials based on crosslinked dendrimers.
PETROLEUM CHEMISTRY Vol. 52
No. 5
2012

292
KARAKHANOV et al.
Table 2. XPS data for the materials
Pd 3d
,
At. concn.,
%
At. concn.,
At. concn.,
%
At. concn.,
%
5/2
Material
C 1s, eV
%
N 1s, eV
O 1s, eV
531.9
3d , eV
3/2
DAB ꢀHMDIꢀPd
336.1;
340.9
2.9
286.8
77.5
399.3;
401.7;
405.3
10.7
9.0
3
DAB ꢀBDIꢀPd
336.1;
340.9
7.4
4.0
5.4
285.0
285.1
284.8
63.0
58.8
63.4
398.7
399.6
399.4
7.7
18.4
16.5
530.6
531.5
531.5
21.9
18.8
13.9
3
DAB ꢀHMDI ꢀPd
335.8;
341.4
1
2
DAB ꢀHMD ꢀPd
335.8;
341.6
3
2
The polymer matrices based on diisocyanateꢀ dride (Scheme 2). The metal deposition step was conꢀ
crosslinked dendrimers were used to prepare pallaꢀ ducted in chloroform because of good solubility of
dium catalysts. Metal nanoparticles were encapsulated palladium acetate in this solvent and its swelling power
in the dendrimers according to the procedure reported with respect to polymers containing urethane groups.
in [11], which involves complexation with a transition The reduction to Pd(0) was achieved with the use of
metal salt followed by reduction with sodium borohyꢀ sodium borohydride.
HN
HN
HN
O
O
O
HN
O
O
NH
NH
N
N
H
H
O
N
N
N
N
N
O
O
O
H
O
H
O
O
NH
O
H
N
N
NH
O
N
N
N
N
O
NH
HN
H
H
O
NH
HN
O
N
N
O
N
N
N
H
O
N
HN
N
N
H
HN
N
N
O
N
HN
N
H
N
N
N
NH
N
N
N
NH
N
O
O
N
O
N
N
N
O
O
O
N
N
O
Pd0
O
O
O
HN
O
O
O
H
N
N
H
N
O
O
O
N
O
H
HN
NH
O
O
NH
N
Pd0
HN
O
N
N
HN
O
HN
NH
N
N
NH
O
HN
NH
NH
O
O
NH
O
O
O
N
NH
HN
NH
O
O
HN
O
N
O
N
O
HN
O
HN
O
1) Pd(PAc)2, CHCl3, 70°C
NH
O
O
N
HN
NH
O
HN
O
HN
N
NH
O
N
O
Pd0
N
N
N
O
O
N
HN
N
O
N
N
2) NaBH4, CHCl3/MeOH, 70°C
N
NH
HN
HN
NH
N
O
NH
O
O
N
NH
N
HN
HN
N
NH
N
N
NH
O
O
N
O
O
O
N
O
N
O
O
HN
HN
O
NH
O
O
O
N
NH
HN
Pd0
HN
HN
O
O
NH
O
O
HN
HN
N
NH
N
O
HN
O
O
NH
HN
HN
N
HN
N
O
O
O
N
N
HN
NH
N
N
O
O
HN
O
O
HN
Pd0
N
N
N
HN
O
N
N
NH
NH
O
N
H
NH
O
O
O
N
N
HN
O
N
O
HN
NH
NH
NH
NH
O
N
HN
O
H
N
N
N
O
O
O
N
HN
HN
O
N
O
N
O
HN
H
N
N
O
O
H
N
N
O
N
O
N
HN
H
N
O
N
H
N
N
O
H
N
N
N
N
O
O
O
H
N
N
N
H
N
H
O
O
O
N
N
O
HN
O
NH
N
H
N
HN
O
O
O
HN
NH
NH
N
O
HN
NH
O
Scheme 2. Encapsulation of palladium nanoparticles in crosslinked dendrimer matrices.
It was found that the amount of the metal deposited higher than for zeroꢀvalence metallic palladium (Pd:
on the polymer depends on the dendrimer structure,
the crosslinker nature, the crosslink density, and the
procedure for isolation of the intermediate and final
products. The metal content in % as determined by
ICPꢀAES for each of the materials is given in Table 1.
The materials obtained were characterized using
the XPS and TEM techniques. The XPS data are preꢀ
sented in Table 2.
335.6, 341.1 eV [15]) and lower than for divalent palꢀ
ladium (PdO: 37.0, 342.2 [15]), thereby confirming
the presence of metal nanoparticles coordinated to
nitrogen or oxygen atoms in the materials. It was
found that only a small portion of the metal occurs on
the surface (Cs
/
C
v
≈
0.3–0.5). The carbon and nitroꢀ
gen energy shifts relative to the usual values (284.8 and
398.6 eV, respectively [15]) in the spectra of the samꢀ
ples are due to partial electron transfer to neutral metal
The binding energies for Pd 3d5/2 and Pd 3d3/2 are
336 and 341 eV, respectively; i.e., they are somewhat particles.
PETROLEUM CHEMISTRY Vol. 52
No. 5
2012

PALLADIUM NANOPARTICLES ON DENDRIMERꢀCONTAINING SUPPORTS
(а) (b)
293
80
60
40
20
0
1
2
3
4
, nm
5
6
50 nm
d
Fig. 1. TEM data and particleꢀsize distribution for DAB ꢀHMDIꢀPd.
3
According to the TEM data, the nature of the
In the case of highly crosslinked samples (Figs. 3,
crosslinker in the case of lowꢀcrosslinkꢀdensity comꢀ 4), the density of metal nanoparticles per unit
posites with Pd nanoparticles affects the particle size area/volume of the support noticeably increases owing
distribution pattern, rather than the size per se to their better retention in the matrix and coordination
(Figs. 1, 2). In the samples prepared with the use of to the amino groups of not only dendrimers but also
hexamethylene diisocyanate, the proportion of partiꢀ the crosslinking agents. In addition, both the size and
cles with the diameter above the mean (2.3–5.0 nm) is the distribution pattern of particles are strongly
high. This difference is substantially lower for the samꢀ affected by the crosslinker nature and the matrix prepꢀ
ples prepared with butylene diisocyanate.
aration procedure. Thus, the density distribution of
(а)
(b)
180
160
140
120
100
80
60
40
20
0
0
1
2
3
4
, nm
5
6
7
50 nm
d
Fig. 2. TEM data and particleꢀsize distribution for DAB ꢀBDIꢀPd.
3
PETROLEUM CHEMISTRY Vol. 52
No. 5
2012

294
KARAKHANOV et al.
(а)
(b)
350
300
250
200
150
100
50
0
1
2
3
, nm
4
5
6
50 nm
d
Fig. 3. TEM data and particleꢀsize distribution for DAB ꢀHMDI ꢀPd.
1
2
particles noticeably increases in the DAB1ꢀHMDI2ꢀ between dendrimer branches, being stabilized by terꢀ
tiary nitrogen atoms.
Pd and DAB3ꢀHMDI2ꢀPd catalysts, correlating with
both the increased palladium content of the samples
and the particle size.
The DAB3ꢀHMDI2ꢀPd catalyst is characterized by
a narrower distribution with maximums at 1.8 and 2.3
nm, which can be explained in terms of not only a
smaller size of cavities compared with the material
based on the firstꢀgeneration dendrimer, but also their
rigidity due to the dense crosslinking of the dendrimꢀ
ers by hexamethylene diisocyanate. In this case, the
size of a particle has the upper bound due to its surꢀ
The DAB1ꢀHMDI2ꢀPd material is characterized by
a bimodal particle size distribution with maximums at
1.3 and 2.2 nm, a feature that may be due to the presꢀ
ence of cavities with the corresponding dimensions in
the structure of the dendritic matrix. In addition, small
particles (
d
≤
1.5 nm) can be accommodated in rounding crosslinks and dendrimer branches—a steric
(а)
(b)
800
600
400
200
0
1
2
3
4
5
50 nm
d, nm
Fig. 4. TEM data and particleꢀsize distribution for DAB ꢀHMDI ꢀPd.
3
2
PETROLEUM CHEMISTRY Vol. 52
No. 5
2012

PALLADIUM NANOPARTICLES ON DENDRIMERꢀCONTAINING SUPPORTS
295
(b)
(a)
2000
1500
100
80
DAB ꢀHMDIꢀPd
3
DAB ꢀBDIꢀPd
3
DAB ꢀHMDIꢀPd
3
DAB ꢀBDIꢀPd
3
60
40
20
1000
500
0
0
hexeneꢀ1
octeneꢀ1
deceneꢀ1
hexeneꢀ1
deceneꢀ1
octeneꢀ1
Fig. 5. Dependence of (a) the conversion and (b) the activity of palladium catalysts on the substrate chain length.
limitation—and the lower bound as the minimal disꢀ straints on alkene passing through the crosslinked
tance between the nearest donor atoms (~0.6–0.7 nm dendritic matrix.
for dendrimers and 1 nm for diisocyanate units in the
approximation that the carbon chain envelopes the
particle).
Simultaneously, the proportion of doubleꢀbond
isomerization products in the reaction mixture
increases: the longer the chain, the greater the proporꢀ
The synthesized materials were tested as catalysts tion. Note that the doubleꢀbond isomerization is easy
for hydrogenation of unsaturated compounds. In the to proceed in the presence of palladium (Scheme 3)
case of hydrogenation of terminal linear olefins in the and affords internal alkenes, which are thermodynamꢀ
presence of the materials based on dendrimers and Pd ically more stable. Being sterically more hindered than
nanoparticles, the activity of all the catalysts with a terminal alkenes, the internal isomers cannot take the
low crosslink density decreases with an increase in the conformation required for the reaction in the existent
chain length of the substrate, as shown in Fig. 5, a ligand microenvironment of the catalytic center and,
development that can be attributed to diffusional conꢀ as such, are hydrogenated with a low yield.
Pd
~H
Pd
~H
Pd
~H
Pd
Pd
Pd
Scheme 3. Positional isomerization of the olefin double bond exemplified by octeneꢀ1.
The yield increases in the order DAB3ꢀHMDIꢀ activity further drops by a factor of 2 or 1.6–1.7 for
ꢀ DAB3ꢀBDIꢀPd in agreement with the increase in DAB3ꢀBDIꢀPd or DAB3ꢀHMDIꢀPd, respectively. The
Pd
the Pd content on passing from one catalyst to
another. The activity of the catalysts, calculated as the
turnover frequency (TOF), an amount of substrate
converted per mole of metal per unit time, drops in the
opposite order.
Regardless of the Pd content, the activity of all the
catalysts declines on passing to a bulkier substrate. It
specificity of the process in this case seems to be assoꢀ
ciated with the enhancement of diffusional constraints
for large substrate molecules. Note that despite their
sterically hindered dendrite matrix, these catalysts
turned out to be more active than the conventional
Pd/C catalyst used in the industry, its TOF is 190 or
130 h–1 for hexene and cyclohexene, respectively [12],
decreases by a factor of 2 on passing from hexaneꢀ1 to values that are a few times below those obtained in this
octeneꢀ1. On passing from octeneꢀ1 to deceneꢀ1, the work.
PETROLEUM CHEMISTRY Vol. 52
No. 5
2012

296
KARAKHANOV et al.
Table 3. Hydrogenation of octeneꢀ1 in the presence of denꢀ
drimerꢀbased Pd catalysts
p
ꢀmethylstyrene (Fig. 6, Table 5). It is likely that there
is an optimal substrate size for this material to correꢀ
spond to the distance between branches of the denꢀ
dritic matrix.
Catalyst
Conversion, %
TOF, h–1
The catalysts based on highly crosslinked dendrimꢀ
ers appeared to be less active than their counterparts
with a low crosslink density at similar Pd/substrate
ratios (Table 4, Fig. 7). This difference is evidently due
to steric hindrances by the dendritic matrix and the
crosslinks, since all dendrimer amino groups are
involved in the formation of amide bonds with the
diisocyanate crosslinks.
DAB ꢀHMDIꢀPd
22
35
27
32
480
240
380
545
3
DAB ꢀBDIꢀPd
3
DAB ꢀHMDI ꢀPd
1
2
DAB ꢀHMDI ꢀPd
3
2
The data in Table 6 show that the activity of the catꢀ
alysts in the case of styrene drops with an increase in
both the palladium content and the average substrate
size. The decrease in the activity on passing from
DAB3ꢀHMDIꢀPd to DAB3ꢀHMDI2ꢀPd is determined
by enhancement of steric hindrances from the densely
crosslinked dendritic matrix. The activity also
dropped as the size of the substituent in the paraꢀposiꢀ
In the presence of the catalysts with a high crosslink
density, the product yield and the activity per Pd atom
significantly increase, a fact that can be explained by
both a high affinity of the substrate for nonpolar
crosslink chains, whose number is substantially greater
in the case of denser crosslinking, and the stability of
the nanoparticles themselves during the reaction. On
passing from the generation one to generation three
dendrimer, the specific activity increases despite a
somewhat greater size of the particles (Table 3).
tion to the vinyl group further increased (
rene) or substituents appeared at the double bond
transꢀstilbene), with the yields being comparable on
pꢀvinylstyꢀ
(
DAB1ꢀHMDI2ꢀPd and DAB3ꢀHMDI2ꢀPd (55 and
50%, respectively) and considerably below those on
DAB3ꢀHMDIꢀPd (100%). However, the stilbene
hydrogenation activity of the catalysts based on
densely crosslinked dendrimers was higher by a factor
of approximately 1.5–1.7 than that of DAB3ꢀHMDIꢀ
Pd (Tables 4–6), possibly, because of enhanced hydroꢀ
phobicity of the dense dendritic matrix and, as a conꢀ
sequence, a higher affinity for stilbene compared with
the rarely crosslinked supports, which are mainly
composed of hydrophilic PPI dendrimers in DAB3ꢀ
The materials based on Pd nanoparticles and
crosslinked dendrimers showed a high activity in the
hydrogenation of styrene and its derivatives to correꢀ
sponding ethylbenzenes (Tables 4–6), with the TOF
value being very high; it approaches 10000–15000,
i.e., is substantially higher than that for 1ꢀalkenes. The
activity of the catalysts considerably increases with
temperature in this case.
Note that there is a relationship between the prodꢀ
uct yield and the substrate size for this class of cataꢀ
lysts. Thus, a regular fall in activity because of steric
hindrances with an increase in the substrate size was
observed for DAB3ꢀHMDIꢀPd, whereas DAB3ꢀBDIꢀ
HMDIꢀPd
.
The catalysts prepared in this work can be repeatꢀ
edly used without loss of activity. For example, the
Pd is characterized by a parabolic dependence conversion of styrene almost does not decrease for six
of the activity on the substrate size with a maximum at catalytic cycles. The turnover number (TON, the
Table 4. Hydrogenation of styrenes in the presence of dendrimerꢀbased Pd catalysts
DAB ꢀHMDIꢀPd
DAB ꢀBDIꢀPd
DAB ꢀHMDI ꢀPd
DAB ꢀHMDI ꢀPd
1 2
3
3
3
2
Substrate
time, converꢀ TOF, time, converꢀ TOF, time, converꢀ TOF, time, converꢀ TOF,
min sion, % h–1 min sion, % h–1 min sion, % h–1 min sion, % h–1
Styrene
ꢀMethylstyrene
30
10
30
10
60
30
60
100
5240
60
–
100
–
1025
–
60
15
60
15
60
15
60
100
2330
60
15
60
15
60
15
60
100
1930
79 12420
100* 18630
100* 15450
p
100
4560
9440
1645
2680
670
60
–
100
–
895
100
48
84
9
2025
3890
1960
840
95
66
18
5.5
55
1600
5300
350
69
100
81
pꢀtertꢀButylstyꢀ
rene
60
–
100
–
645
–
425
transꢀStilbene
100
–
–
–
50
1185
1080
Reaction conditions: 1 mg of catalyst, 250–400
* 0.5 mg of catalyst.
µ
L of substrate, 80°C, 1 h, 5–10 atm H .
2
PETROLEUM CHEMISTRY Vol. 52
No. 5
2012

PALLADIUM NANOPARTICLES ON DENDRIMERꢀCONTAINING SUPPORTS
297
Table 5. Hydrogenation of styrenes in the presence of Table 6. Comparison of activities of catalysts with the hexꢀ
DAB3ꢀBDIꢀPd at lower temperatures amethylene diisocyanate crosslinker in hydrogenation of
styrenes
Substrate
Conversion, % TOF, h–1
T, °C
p
ꢀMethylꢀ pꢀtertꢀButylstyꢀ
Styrene
Styrene
60
50
60
50
60
74
48
95
90
26
9110
5910
10180
9645
2010
Catalyst
Styrene
styrene
rene
p
p
ꢀMethylstyrene
ꢀMethylstyrene
DAB ꢀHMDIꢀPd
12420
9315
7490
9840
3890
5300
2680
1960
3
DAB ꢀHMDI ꢀPd
pꢀtertꢀButylstyꢀ
rene
3
2
DAB ꢀHMDI ꢀPd
425
1
2
pꢀtertꢀButylstyꢀ
50
20
1540
rene
Reaction conditions: 1 mg of catalyst, 250–400
µ
L of substrate,
Reaction conditions: 0.5 mg of catalyst, 250
5 atm H .
µL of substrate, 10 min,
10 atm H , 80
°
C, 10–15 min.
2
2
(a)
(b)
60°C
50°C
60°C
50°C
100
80
60
40
20
0
12000
10000
8000
6000
4000
2000
0
Fig. 6. Hydrogenation of styrenes in the presence of DAB ꢀBDIꢀPd: (a) conversions and (b) activities.
3
16000
14000
12000
10000
8000
DAB3ꢀHMDIꢀPd
DAB3ꢀHMDI2ꢀPd
DAB1ꢀHMDI2ꢀPd
6000
4000
2000
0
Fig. 7. Effect of the crosslink density and dendrimer generation on the catalyst activity.
PETROLEUM CHEMISTRY Vol. 52 No. 5 2012

298
KARAKHANOV et al.
5. L. Piccolo, A. Valcarcel, M. Bausach, et al., Phys.
amount of product per mole of Pd) is 13590 in this
case.
Chem. Chem. Phys. 10, 5504 (2008).
6. K. Pelzer, Ruthenium Nanoparticles: Synthesis, Characꢀ
terisation and Organisation in Alumina Membranes and
Mesoporous Materials; Applications in Catalysis, PhD
Theses. Toulouse (2003).
7. G. Schmid, M. Baumle, M. Geerkens, et al., Chem.
Soc. Rev. 28, 179 (1999).
8. T. Sawitowski, Y. Miguel, A. Heilmann, and
G. Schmid, Adv. Funct. Mater. 11, 235 (2001).
9. Y. Niu and R. M. Crooks, C. R. Chim. 6, 1049 (2003).
10. E. A. Karakhanov, A. L. Maximov, V. A. Skorkin, et al.,
In summary, the use of nanostructured assemblies
based on crosslinked dendrimers and palladium nanoꢀ
particles makes it possible to synthesize highꢀperforꢀ
mance selective catalysts for hydrogenation of unsatꢀ
urated compounds. The activity and selectivity of such
systems is directly related to the nature and structure
of the dendritic matrix and its affinity for the substrate.
The maximal yields and activity are attained in the
hydrogenation of styrene to corresponding ethylbenꢀ
zenes.
Pure Appl. Chem. 81, 2013 (2009).
11. R. M. Crooks, M. Zhao, L. Sun, et al., Acc. Chem.
Res. 34, 181 (2001).
12. M. Ooe, M. Murata, T. Mizugaki, et al. Nanoletters
999 (2002).
REFERENCES
2,
1. G. Budroni, A. Corma, H. Garcia, and A. Primo, J.
Catal. 251, 345 (2007).
13. E. A. Karakhanov and A. L. Maksimov, A. V. Zoloꢀ
tukhina, and S. V. Kardashev, Pet. Chem. 50, 290
(2010).
2. J. M. Thomas, B. F. G. Johnson, R. Raja, et al., Acc.
Chem. Res. 26, 20 (2003).
14. E. M. M. Berg and E. W. Meijer, Angew. Chem., Int.
3. D. D. Das and Ab. Sayari, J. Catal. 246, 60 (2007).
Ed. Engl. 32 (9), 1308 (1993).
4. N. Ren, Y.ꢀH. Yang, Y.ꢀH. Zhang, et al., J. Catal. 246
,
15. NIST
XPS
Electronic
Database,
215 (2007).
http://srdata.nist.gov/xps.
PETROLEUM CHEMISTRY Vol. 52
No. 5
2012