Palladium nanoparticles encapsulated in a dendrimer networks as catalysts for the hydrogenation of unsaturated hydrocarbons

Article abstract of DOI:10.1016/j.molcata.2014.10.018

A novel method has been proposed for encapsulating palladium nanoparticles up to 5 nm in the matrix of polymeric support networks based on polyamidoamine dendrimers. The shape of the particle size distribution and the catalytic activity of the materials obtained during the hydrogenation of unsaturated compounds depend strongly on the support structure. High activity (TOF up to 86,000 h^-1) has been observed during the hydrogenation of styrene.

Full text of DOI:10.1016/j.molcata.2014.10.018

Journal of Molecular Catalysis A: Chemical 397 (2015) 1–18
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Palladium nanoparticles encapsulated in a dendrimer networks as
catalysts for the hydrogenation of unsaturated hydrocarbons
Edward A. Karakhanova,∗, Anton L. Maksimova,b, Elena M. Zakhariana,
Yulia S. Kardashevaa, Sergey V. Savilova, Nadezhda I. Truhmanovab,
Andrey O. Ivanovb, Vladimir A. Vinokurovc
a Department of Chemistry, Lomonosov Moscow State University (MSU) , 119991 Moscow, Russia
b Institute of Petrochemical Synthesis RAS, 119991 Moscow, Russia
c I.M. Gubkin Russian State University of Oil and Gas, 119991 Moscow, Russia
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 24 July 2014
Received in revised form 14 October 2014
Accepted 15 October 2014
Available online 24 October 2014
A novel method has been proposed for encapsulating palladium nanoparticles up to 5 nm in the matrix
of polymeric support networks based on polyamidoamine dendrimers. The shape of the particle size
distribution and the catalytic activity of the materials obtained during the hydrogenation of unsaturated
compounds depend strongly on the support structure. High activity (TOF up to 86,000 h−1) has been
observed during the hydrogenation of styrene.
© 2014 Elsevier B.V. All rights reserved.
Keyword:
Pd catalysts
Nanoparticles
PAMAM dendrimer
Selective hydrogenation
1. Introduction
nanoparticles have the advantage over other types of metaldendrite
catalysts: the poly(propyleneimine) (PPI) and PAMAM dendrimers
Catalysts utilising metal nanoparticles as active components
deposited on dendrimer matrices have recently found numer-
ous applications in petrochemical and organic syntheses. These
materials are used during the hydrogenation of double bonds [1],
hydroformylation [2,3], amination [4], carbonylation [5] and cross-
coupling reactions (the Suzuki [6], Heck [7] and the Sonogashira
reaction [8]). The main problem with metal nanoparticles as catal-
ysis is their aggregation, which leads to the decrease in the total
specific surface of the catalyst and, as a consequence, to the activ-
ity loss. In a number of cases, the selectivity of the reaction also
depends on the particle size. Metal nanoparticles can be stabi-
lized by the interactions with either the donor atoms of various
organic ligands following the coordination mechanism, or the inter-
nal surface of pores in such systems as zeolite and mesoporous
aluminosilicates.
commonly used as a matrix can be relatively readily synthesized
in laboratory from commercially available precursor [13–18]. We
have proposed a method for synthesising a heterogeneous sup-
port based on polyamidoamine dendrimers, binding them to one
another with special cross-linking agents, such as diisocyanates. It
includes a preliminary cross-linking PPI- dendrimers with various
bifunctional agents (diepoxides, diisocyanates) with subsequent
deposition of a required metal salt and its reduction to the zerova-
lent state. The driving force for the preparation of the catalyst is the
formation of the complex between the ions of the deposited metal
and the amino groups of dendrimers. In this case, the final proper-
ties of the material (nanoparticle size distribution, catalytic activity
and selectivity) are determined by the dendrimer generation, the
size, rigidity, and polarity of the cross linking agent. Thus synthe-
sized system structurally resembles metalorganic frameworks, in
which the dendrimers, rather than metal ions, serve as the matrix
nodes, whose rigidity increases with the increase in the gener-
ation and decrease in the length of the branches. Therefore, we
prepared effective hydrogenation catalysts based on cross-linked
dendrimers and palladium nanoparticles with particle sizes ran-
ging from 1.8 to 2.8 nm based on generation of the PPI dendrimer
[19]. The corresponding material exhibits a high activity when
hydrogenating olefin and phenylacetylene [19–22]. In this paper,
One of the most interesting approaches toward synthesising cat-
alysts based on metal nanoparticles is the use of the dendrimers
to bind metal ions, which are nanoparticle precursors, and to sta-
bilise the resultant particles [9–12]. The dendrimer encapsulated
∗
Corresponding author. Tel.: +7 495 939 53 77; fax: +7 495 939 19 51.
E-mail address: kar@petrol.chem.msu.ru (E.A. Karakhanov).
http://dx.doi.org/10.1016/j.molcata.2014.10.018
1381-1169/© 2014 Elsevier B.V. All rights reserved.

2
E.A. Karakhanov et al. / Journal of Molecular Catalysis A: Chemical 397 (2015) 1–18
we report the design of catalysts based on palladium nanoparticles
and PAMAM dendrimers using the same approach.
diisocyanate (0.65 g, 1.014 mmol) in 50 mL of absolute THF. The
reaction product was obtained as a white powder (2.124 g).
IR (sm−1): 3298 (N-Hst in NH-C( O)); 2954 (C-Hst); 2843 (C-
Hst, CH2-Nst); 1633 (C Ost in NH-C( O)); 1558, 1510 (N-H␦, CH2␦
,
C-N-H␦); 1404, 1300 (CH2␦); 1219, 1126, 1072, 1016 (C-Nst in NH-
C( O)); 823 (aromatic C-H␦); 761, 687, 652 (aromatic C C␦).
XPS (eV): 284.5 (32%), 285.7 (46%), 287.1 (8%), 288.5 (14%) (C 1s,
64.9%); 399.3 (N 1s, 22.1%); 530.8 (O 1s, 12.5%).
2. Experimental
2.1. Analysis
Synthesis of PAMAM-G1-DMPDI. The synthesis was carried out
according to the procedure described above. The reactants included
PAMAM(NH2)8 (1.871 g, 1.309 mmol) and 3,3ꢀ-dimethoxy-4,4ꢀ-
diphenylene diisocyanate (0.518 g, 1.748 mmol) in 50 mL of
absolute THF. The reaction product was obtained as loose white
lumps (2.389 g).
The first-, second- and third-generation polyamidoamine den-
drimers, specifically PAMAM-G1 (PAMAM (NH2)8), PAMAM-G2
(PAMAM (NH2)16), and PAMAM-G3 (PAMAM (NH2)32), were syn-
thesised according to the procedure described in [23]; these
materials were used as starting materials. The cross-linking
agents, which were obtained from Aldrich, included the following:
1,4-butylene diisocyanate (BDI), 1,6-hexamethylene diisocyanate
(HMDI), 1,8-octamethylene diisocyanate (OMDI), p-phenylene
diisocyanate (PDI), and 3,3ꢀ-dimethoxy-4,4ꢀ-biphenylene diiso-
cyanate (DMPDI).
IR (sm−1): 3325 (N-Hst in NH-C( O)); 2929 (C-Hst); 2856 (C-
Hst, CH2-Nst); 1628 (C Ost in NH-C( O)); 1547 (N-H␦, CH2␦, C-N-
H␦); 1462, 1431 (CH2␦); 1261, 1119, 1038 (C-Nst in NH-C( O)); 820
(aromatic C-H␦); 768, 652 (aromatic C C␦).
Synthesis of PAMAM-G2-HMDI. The synthesis was carried out
according to the procedure described above. The reactants included
PAMAM(NH2)16 (2 g, 0.615 mmol) and 1,6-hexamethylene diiso-
cyanate (0.26 mL, 1.64 mmol) in 50 mL of absolute THF. The reaction
product was obtained as a yellowish sticky mass (2.011 g).
Synthesis of PAMAM-G2-PDI. The synthesis was carried out
according to the procedure described above. The reactants included
PAMAM(NH2)16 (2 g, 0.615 mmol) and p-phenylene diisocyanate
(PDI) (0.262 g, 1.64 mmol) in 50 mL of absolute THF. The reaction
product was obtained as a yellowish sticky mass (2.215 g).
Synthesis of PAMAM-G2-DMPDI. The synthesis was carried out
according to the procedure described above. The reactants included
PAMAM(NH2)16 (768 mg, 0.236 mmol) and 3,3ꢀ-dimethoxy-4,4ꢀ-
diphenylene diisocyanate (711 mg, 2.4 mmol) in 50 mL of anhy-
drous methanol. The reaction product was obtained as a white
powder (892 mg).
The IR spectra were obtained with a Nicolet IR2000 (Thermo Sci-
entific) using a method involving multiple distortions in the total
internal reflection with Multi-reflection HATR accessories; these
accessories contained a ZnSe crystal (45◦) for use at different wave-
lengths and a resolution of 4 cm−1
.
The X-ray photoelectron spectroscopy (XPS) studies were per-
formed with a Kratos Axis Ultra DLD electronic device equipped
with a photoelectron analyser and an OPX-150 slowing potential.
The photoelectrons were excited using X-rays from an aluminium
anode (Al K␣ = 1486.6 eV) at 12 kV and 20 mA. The photoelectron
peaks were calibrated using the C 1s carbon line at 284.8 eV.
The palladium in the samples was quantified using induc-
tively coupled plasma atomic emission spectroscopy (ICP-AES)
with an IRIS Interpid II XPL instrument (Thermo Electron Corp.,
USA) through the radial and axial detection at 310 and 95.5 nm.
The transmission electron microscopy (TEM) studies of the sam-
ples were performed on a LEO912 AB OMEGA transmission electron
microscope.
IR (sm−1): 3325 (N-Hst in NH-C( O)); 2929 (C-Hst); 2856 (C-Hst
,
CH2-Nst); 1627, (C Ost in NH-C( O)); 1547 (N-H␦, CH2␦, C-N-H␦);
1462, 1431 CH2␦); 1261, 1157, 1113, 1038 (C-Nst in NH-C( O));
874, 820 (aromatic C-H␦); 768, 729, 710, 658 (aromatic C C␦).
Synthesis of PAMAM-G3-HMDI. The synthesis was carried out
according to the procedure described above. The reactants included
PAMAM(NH2)32 (2 g, 0.29 mmol) and 1,6-hexamethylene diiso-
cyanate (0.25 mL, 1.547 mmol) in 50 mL of absolute THF. The
reaction product was obtained as a white powder (2.124 g).
2.2. Synthesis of dendrimer compounds and characterization
Synthesis of PAMAM-G1-BDI. A 1.1-g portion of PAMAM(NH2)8
(0.773 mmol) and 50 mL of absolute THF were placed in a 100-mL
single-neck flask equipped with a magnetic stirrer and a reflux con-
denser. Under these conditions, the dendrimer swelled but did not
dissolve. Afterwards, 0.13 mL (1.031 mmol) of 1,4-butylene diiso-
cyanate was added with stirring. The reaction proceeded for 12 h at
70 ◦C before the product mixture was evaporated in a rotary evap-
orator at 50 ◦C. The product was obtained as loose yellow lumps
(1.357 mg).
IR (sm−1): 3327 (N-Hst in NH-C( O)); 2929 (C-Hst); 2854 (C-Hst
,
CH2-Nst); 1628 (C Ost in NH-C( O)); 1556 (N-H␦, CH2␦, C-N-H␦);
1469, 1431, 1358 (CH2␦); 1261, 1153, 1037 (C-Nst in NH-C( O)).
Synthesis of PAMAM-G3-PDI material. The synthesis was car-
ried out according to the procedure described above. The reactants
included PAMAM(NH2)32 (2 g, 0.29 mmol) and p-phenylene diiso-
cyanate (PDI) (0.248 g, 1.57 mmol) in 50 mL of absolute THF. The
reaction product was obtained as a white powder (2.549 g).
Synthesis of PAMAM-G1-HMDI. The synthesis was carried out
according to the procedure described above. The reactants included
PAMAM(NH2)8 (1.088 g, 0.761 mmol) and 1,6-hexamethylene
diisocyanate (0.16 mL, 1.014 mmol) in 50 mL of absolute THF. The
reaction product was obtained as loose yellow lumps (1.259 g).
IR (sm−1): 3327 (N-Hst in NH-C( O)); 2916 (C-Hst); 2862 (C-Hst
,
CH2-Nst); 1631 (C Ost in NH-C( O)); 1566 (N-H␦, CH2␦, C-N-H␦);
1477, 1433, 1335 (CH2␦); 1265, 1153, 1047 (C-Nst in NH-C( O));
904, 862, 827 (aromatic C-H␦); 766, 669 (aromatic C C␦).
IR (sm−1): 3302 (N-Hst in NH-C( O)); 2933 (C-Hst); 2858 (C-Hst
,
CH2-Nst); 1658 (C Ost in NH-C( O)); 1552 (N-H␦, CH2␦, C-N-H␦);
1477, 1433, 1387 (CH2␦); 1263, 1157, 1120, 1036 (C-Nst in NH-
C( O)).
2.3. Preparation of Pd-dendrimer compounds and
characterization
Synthesis of PAMAM-G1-OMDI. The synthesis was carried out
according to the procedure described above. The reactants included
PAMAM(NH2)8 (0.568 g, 0.397 mmol) and 1,8-octamethylene
diisocyanate (0.15 mL, 0.77 mmol) in 50 mL of absolute THF. The
reaction product was obtained as yellow loose lumps (0.738 g).
Synthesis of PAMAM-G1-PDI. The synthesis was carried out
according to the procedure described above. The reactants
included PAMAM(NH2)8 (1.088 g, 0.761 mmol) and p-phenylene
Synthesis of G1-BDI-Pd. To a 50-ml flask equipped with a
magnetic stirrer and a reflux condenser was added 1.1 g of PAMAM-
G1-BDI in 40 mL of absolute chloroform. Afterwards, 787 mg
(3.51 mmol) of Pd(OAc)2 was added to the resulting suspension
with stirring. The reaction proceeded for 12 h at 70 ◦C.
After the reaction, the suspension was evaporated to dryness in
a rotary evaporator. The intermediate product was a dark brown

E.A. Karakhanov et al. / Journal of Molecular Catalysis A: Chemical 397 (2015) 1–18
3
IR (sm−1): 3250 (N-Hst in NH-C( O)); 2918 (C-Hst); 2850 (C-
st, CH2-Nst); 1699, 1637 (C Ost in NH-C( O)); 1550, 1510 (N-H␦,
CH2␦, C-N-H␦); 1439, 1390, 1321 (CH2␦); 1244, 1205, 1173, 1142,
1024 (C-Nst in NH-C( O)); 949, 837, 814 (aromatic C-H␦); 733, 654
(aromatic C C␦).
XPS (eV): 284.4 (43.78%), 285.7 (36.26%), 286.8 (8.23%), 288.3
(11.73%) (C 1s, 69.1%); 335.2 (83%) (Pd 3d5/2, 6.5%), 337.9 (17%)
(Pd 3d3/2); 399.5 (N 1s, 14.6%); 531.2 (O 1s, 9.8%). ICP-AES: 13.12%
Pd.
powder (1.787 g, 100%). This material was placed in a 50-mL flask
equipped with a magnetic stirrer and a reflux condenser before
being suspended in a mixture containing 30 mL of chloroform and
10 mL of methanol. Sodium borohydride (1.334 g, 35.1 mmol) was
added to the resulting suspension with stirring. The reaction mix-
ture turned black, and a violent evolution of gas was observed. The
reaction proceeded for 12 h at 60 ◦C. After the reaction, the sus-
pension was evaporated to dryness in a rotary evaporator, and the
resulting precipitate was washed twice with water and methanol
before drying in air with heating. The product was obtained as a
black powder (533 mg).
H
Synthesis of G2-HMDI-Pd. The synthesis was carried out accord-
ing to the procedure described above. The reactants included
PAMAM-G2-HMDI (300 mg) and Pd(OAc)2 (48.6 mg, 0.216 mmol)
in 30 mL of methylene chloride. The intermediate product was sus-
pended in 30 mL of methylene chloride and 10 mL of methanol
before adding sodium borohydride (82.4 mg, 2.16 mmol). The prod-
uct was a black layered powder (130 mg).
IR (sm−1): 3334 (N-Hst in NH-C( O)); 2922 (C-Hst); 2850 (C-Hst
,
CH2-Nst); 1626 (C Ost in NH-C( O)); 1558, 1525 (N-H␦, CH2␦, C-
N-H␦); 1477, 1435, 1356 (CH2␦); 1281, 1219, 1147, 1051 (C-Nst in
NH-C( O)).
XPS (eV): 284.7 (37.96%), 285.8 (40.22%), 286.1 (6.43%), 288.4
(15.4%) (C 1s, 64.5%); 335.0 (87%) (Pd 3d5/2, 4.7%), 337.9 (13%) (Pd
3d3/2); 399.5 (N 1s, 19.9%); 531.2 (O 1s, 10.9%). ICP-AES: 15.83% Pd.
Synthesis of G1-HMDI-Pd. The synthesis was carried out accord-
ing to the procedure described above. The reactants included
PAMAM-G1-HMDI (623 mg) and Pd(OAc)2 (448 mg, 1.985 mmol)
in 40 mL of absolute chloroform. The intermediate product was
suspended in 30 mL of chloroform and 10 mL of methanol before
adding sodium borohydride (757 mg, 19.85 mmol). The product
was a black powder (275 mg).
IR (sm−1): 3315 (N-Hst in NH-C( O)); 2931 (C-Hst); 2860 (C-Hst
,
CH2-Nst); 1624 (C Ost in NH-C( O)); 1550 (N-H␦, CH2␦, C-N-H␦);
1435, 1335 (CH2␦); 1261, 1068 (C-Nst in NH-C( O)).
XPS (eV): 284.4 (64.1%), 285.5 (19.7%), 286.4 (8.7%), 288.1 (7.5%)
(C 1s, 79.2%); 338.5 (Pd 3d3/2, 0.3%); 399.5 (N 1s, 3.3%); 531.5 (O 1s,
4.0%). ICP-AES: 1.61% Pd.
Synthesis of G22-HMDI-Pd. The synthesis was carried out accord-
ing to the procedure described above. The reactants included
PAMAM-G2-HMDI (500 mg), Pd(OAc)2 (211 mg, 0.941 mmol) and
sodium borohydride (357.6 mg, 9.41 mmol). The product was a
black powder (114 mg).
IR (sm−1): 3324 (N-Hst in NH-C( O)); 2924 (C-Hst); 2852 (C-
Hst, CH2-Nst); 1624 (C Ost in NH-C( O)); 1554, 1523 (N-H␦, CH2␦,
IR (sm−1): 3329 (N-Hst in NH-C( O)); 2924 (C-Hst); 2856 (C-Hst
,
C-N-H␦); 1485, 1433, 1386 (CH2␦); 1254, 1149 (C-Nst in NH-C( O)).
XPS (eV): 284.7 (40.1%), 285.7 (37.7%), 287.2 (11.9%), 288.3
(8.6%) (C 1s, 52.5%); 335.7 (62%) (Pd 3d5/2, 8.9%), 338.2 (38%) (Pd
3d3/2); 399.5 (N 1s, 13.8%); 531.1 (O 1s, 19.6%). ICP-AES: 14.8% Pd.
Synthesis of G1-OMDI-Pd. The synthesis was carried out accord-
ing to the procedure described above. The reactants included
PAMAM-G1-OMDI (738 mg) and Pd(OAc)2 (246 mg, 1.097 mmol)
in 40 mL of absolute chloroform. The intermediate product was
suspended in 30 mL of chloroform and 10 mL of methanol before
adding sodium borohydride (417 mg, 10.97 mmol). The product
was a black powder (146 mg).
CH2-Nst); 1620 (C Ost в NH-C( O)); 1549 (N-H␦, CH2␦, C-N-H␦);
1477, 1433 (CH2␦); 1242, 1072 (C-Nst in NH-C( O)).
XPS (eV): 284.7 (82.58%), 285.6 (12.14%), 286.7 (2.7%), 288.3
(2.58%) (C 1s, 90.4%); 334.8 (50%) (Pd 3d5/2, 0.3%), 337.9 (50%)
(Pd 3d3/2); 399.5 (N 1s, 4.8%); 531.5 (O 1s, 4.5%). ICP-AES:
3.33% Pd.
Synthesis of G2-PDI-Pd. The synthesis was carried out according
to the procedure described above. The reactants included PAMAM-
G2-PDI (300 mg) and Pd(OAc)2 (64.8 mg, 0.288 mmol) in 30 mL of
methylene chloride. The intermediate product was suspended in
30 mL of methylene chloride and 10 mL of methanol before adding
sodium borohydride (109.87 mg, 2.88 mmol). The product was a
black layered powder (70 mg).
IR (sm−1): 3307 (N-Hst in NH-C( O)); 2939 (C-Hst); 2854 (C-
Hst, CH2-Nst); 1624 (C Ost in NH-C( O)); 1564, 1523 (N-H␦, CH2␦,
C-N-H␦); 1433, 1360 (CH2␦); 1254, 1163, 1043 (C-Nst in NH-C( O)).
XPS (eV): 284.6 (53.2%), 285.7 (30.34%), 286.6 (4.82%), 288.5
(11.64%) (C 1s, 61.7%); 335.2 (68%) (Pd 3d5/2, 1.2%), 338.0 (32%) (Pd
3d3/2); 399.5 (N 1s, 24.9%); 531.4 (O 1s, 12.2%). ICP-AES: 5.45% Pd.
Synthesis of G1-PDI-Pd. The synthesis was carried out according
to the procedure described above. The reactants included PAMAM-
G1-PDI (531 mg) and Pd(OAc)2 (352 mg, 1.56 mmol) in 40 mL of
absolute chloroform. The intermediate product was suspended in
30 mL of chloroform and 10 mL of methanol before adding sodium
borohydride (595 mg, 15 mmol). The resulting product was a black
powder (533 mg).
IR (sm−1): 3311 (N-Hst in NH-C( O)); 2927 (C-Hst); 2856 (C-
Hst, CH2-Nst); 1631 (C Ost in NH-C( O)); 1549, 1510 (N-H␦,
CH2␦, C-N-H␦); 1450, 1404 (CH2␦); 1298, 1221, 1107, 1016 (C-
Nst in NH-C( O)); 837, 823 (aromatic C-H␦); 760, 648 (aromatic
C
C␦).
XPS (eV): 284.5 (46.28%), 285.5 (33.36%), 286.4 (9.04%), 288.4
(11.31%) (C 1s, 71.4%); 335.2 (50%) (Pd 3d5/2, 1.3%), 338.0 (50%)
(Pd 3d3/2); 399.5 (N 1s, 16.4%); 531.1 (O 1s, 10.9%). ICP-AES:
4.61% Pd.
Synthesis of G22-PDI-Pd. The synthesis was carried out according
to the procedure described above. The reactants included PAMAM-
G2-PDI (500 mg), Pd(OAc)2 (160 mg, 0.714 mmol) and sodium
borohydride (271.2 mg, 7.14 mmol). The product was a black pow-
der (94 mg).
IR (sm−1): 3319 (N-Hst in NH-C( O)); 2925 (C-Hst); 2858 (C-
Hst, CH2-Nst); 1635 (C Ost in NH-C( O)); 1547, 1510 (N-H␦, CH2␦
,
C-N-H␦); 1448, 1404 (CH2␦); 1298, 1221, 1070, 1009 (C-Nst in
NH-C( O)); 849, 827 (C-H␦ in aromatic ring); 760, 644 (C C␦ in
aromatic ring).
IR (sm−1): 3313 (N-Hst in NH-C( O)); 2924 (C-Hst); 2534 (C-
XPS (eV): 284.4 (50.6%), 285.5 (33.2%), 286.8 (8.9%), 288.1 (7.4%)
(C 1s, 65.9%); 335.7 (45%) (Pd 3d5/2, 3.9%), 338.2 (55%) (Pd 3d3/2);
399.5 (N 1s, 13.9%); 531.9 (O 1s, 13.0%). ICP-AES: 6.5% Pd.
Synthesis of G1-DMPDI-Pd. The synthesis was carried out accord-
ing to the procedure described above. The reactants included
PAMAM-G1-DMPDI (2.389 g) and Pd(OAc)2 (0.978 g, 4.362 mmol)
in 40 mL of absolute chloroform. The intermediate product was
suspended in 30 mL of chloroform and 10 mL of methanol before
adding sodium borohydride (1.66 g, 43.62 mmol). The product was
a black powder (275 mg).
H
st, CH2-Nst); 1631 (C Ost in NH-C( O)); 1556, 1510 (N-H␦, CH2␦,
C-N-H␦); 1452, 1404 (CH2␦); 1298, 1221, 1109, 1014 (C-Nst in NH-
C( O)); 833, 821 (aromatic C-H␦); 761, 640 (aromatic C C␦).
XPS (eV): 284.6 (53.94%), 285.6 (29.89%), 286.6 (6.72%), 288.4
(9.45) (C 1s, 75.9%); 335.0 (63%) (Pd 3d5/2, 1.4%), 338.0 (37%) (Pd
3d3/2); 399.5 (95.11%), 401.1 (4.89%) (N 1s, 13.5%); 531.1 (O 1s,
9.2%). ICP-AES: 6.23% Pd.
Synthesis of G32-PDI-Pd. The synthesis was carried out according
to the procedure described above. The reactants included PAMAM-
G2-PDI (604 mg), Pd(OAc)2 (201 mg, 0.898 mmol) and sodium

4
E.A. Karakhanov et al. / Journal of Molecular Catalysis A: Chemical 397 (2015) 1–18
borohydride (341.2 mg, 8.98 mmol) The product was a black lay-
ered powder (96 mg).
3. Results and discussion
XPS (eV): 284.4 (44.95%), 285.5 (36.65%), 286.5 (6.93%), 288.5
(11.47%) (C 1s, 59.2%), 335.2 (51%) (Pd 3d5/2, 1.5%), 338.0 (49%) (Pd
3d3/2); 399.5 (N 1s, 26.7%); 531.0 (O 1s, 12.6%). ICP-AES: 6.8% Pd.
Synthesis of G2-DMPDI-Pd. The synthesis was carried out accord-
ing to the procedure described above. The reactants included
PAMAM-G2-DMPDI (743 mg) and Pd(OAc)2 (0.386 g, 1.722 mmol)
in 40 mL of absolute chloroform. The intermediate product was
suspended in 30 mL of chloroform and 10 mL of methanol before
adding sodium borohydride (0.654 g, 17.22 mmol). The product was
a black powder (926 mg).
3.1. Catalyst characterization
To prepare the networked polymer matrices used as sup-
ports for palladium nanoparticles, we used first-, second-
and third-generation polyamidoamine (PAMAM) dendrimers
(PAMAM(NH2)8), (PAMAM(NH2)16) and (PAMAM(NH2)32, respec-
tively). The cross-linking agents were compounds that form
conformationally labile links between dendrimers including 1,4-
butylene diisocyanate - OCN(CH2)4NCO (BDI), 1,6-hexamethylene
diisocyanate - OCN(CH2)6NCO (HMDI), 1,8-octamethylene diiso-
cyanate - OCN(CH2)8NCO(OMDI), as well as compounds that form
conformationally rigid links between the dendrimers (p-phenylene
IR (sm−1): 3305 (N-Hst in NH-C( O)); 2933 (C-Hst); 2864 (C-
Hst, CH2-Nst); 1635, 1589 (C Ost in NH-C( O)); 1516 (N-H␦, CH2␦
,
C-N-H␦); 1454, 1392, 1321 (CH2␦); 1240, 1211, 1174, 1124, 1072,
1030 (C-Nst in NH-C( O)); 853, 814, 796 (aromatic C-H␦); 756, 648
(aromatic C C␦). XPS (eV): 284.3 (48.96%), 285.7 (42.65%), 288.2
(8.39%) (C 1s, 72.1%); 335.1 (38.88%) (Pd 3d5/2, 1.0%), 337.9 (61.12%)
(Pd 3d3/2); 399.5 (N 1s, 13.5%); 530.9 (42.26%), 532.6 (57.74%) (O
1s, 13.4%); 532.7 (Pd 3p3/2). ICP-AES: 4.09% Pd.
diisocyanate -
OCNC6H4NCO (PDI) and 3,3ꢀ-dimethoxy-4,4ꢀ-
biphenylene diisocyanate–p-(H3COC6H3NCO)2 (DMPDI). The size,
rigidity and polarity of the resulting cross-links were expected
to affect the activity and selectivity of the catalysts. To compare
the effects on the properties of the final hybrid material relative
to the size of the cross-linker, materials based on PAMAM(NH2)8,
PAMAM(NH2)16, PAMAM(NH2)32 with NCO/NH2 ratios of 1:3 were
synthesised (Table 1). The XPS data indicated that the process
occurred through the traditional mechanism for the addition
of an amine to an isocyanate to form–NH-C( O)-NH–fragments
(Scheme 1).
Metal nanoparticles were encapsulated in dendrimers accord-
ing to the procedure described in [24]: complexation with a
transition metal salt followed by a reduction with sodium borohy-
dride (Scheme 2). The polymer matrices were based on dendrimers
cross-linked with diisocyanates; these materials were used to
prepare the palladium catalysts. The metal deposition step was car-
ried out in chloroform due to good solubility of palladium acetate
and the swelling ability exhibited by this solvent with polymers
containing urethane groups. The reduction to Pd(0) was achieved
using sodium borohydride.
Synthesis of G3-HMDI-Pd. The synthesis was carried out accord-
ing to the procedure described above. The reactants included
PAMAM-G3-HMDI (800 mg) and Pd(OAc)2 (61.56 mg, 0.274 mmol)
in 30 mL of methylene chloride. The intermediate product was sus-
pended in 30 mL of methylene chloride and 10 mL of methanol
before adding sodium borohydride (104.37 mg, 2.736 mmol). The
product was a darkly coloured, resin-like polymer (530 mg).
IR (sm−1): 3244 (N-Hst in NH-C( O)); 2918 (C-Hst); 2860 (C-Hst
,
CH2-Nst); 1630 (C Ost in NH-C( O)); 1547 (N-H␦, CH2␦, C-N-H␦);
1454, 1335 (CH2␦); 1254, 1190, 1132, 1070, 1020 (C-Nst in NH-
C( O)).
XPS (eV) 284.8 (52.5%), 285.8 (31.38%), 287.0 (4.87%), 288.1
(9.46%), 292.2 (1.8%) (C 1s, 74.2%); 338.2 (Pd 3d3/2; 0.3%); 399.5
(85.09%), 401.4 (14.91%) (N 1s, 14.7%); 531.2 (O 1s, 10.8%). ICP-AES:
1.66% Pd.
Synthesis of G3-PDI-Pd. The synthesis was carried out according
to the procedure described above. The reactants included PAMAM-
G3-PDI (800 mg) and Pd(OAc)2 (246 mg, 1.096 mmol) in 30 mL of
methylene chloride. The intermediate product was suspended in
30 mL of methylene chloride and 10 mL of methanol before adding
sodium borohydride (416 mg, 10.944 mmol). The product was a
darkly coloured, rubber-like polymer (357 mg).
The amount of metal deposited on the polymer depended on the
dendrimeric structure, as well as the nature and density of the cross
linker. The metal content (in %) for each of the prepared materials
was determined by ICP-AES, as listed in Table 1.
3.2. X-ray photoelectron spectroscopy (XPS) study
IR (sm−1): 3315 (N-Hst in NH-C( O)); 2923 (C-Hst); 2844 (C-Hst
,
The obtained materials were characterised using XPS technique.
The XPS data are presented in Table 1. Palladium atoms in two
distinct states were observed for the G1-BDI-Pd, G1-HMDI-Pd,
G1-OMDI-Pd, G1-PDI-Pd and G1-DMPDI-Pd samples. The binding
energies for Pd 3d5/2 were 335.7 eV and 335.2 eV for the HMDI and
PDI cross-linking agents and for the BDI, OMDI and DMPDI cross-
linking agents, respectively. It can be attributed to the metallic
palladium. The binding energies for the HMDI and PDI cross-linking
agents and the BDI, OMDI, and DMPDI cross-linking agents were
338.2 eV and 337.9 eV, respectively. The line for this binding energy
was closer to that of palladium oxide (according to the energy
position and shift relative to the value of metallic palladium). In
the G2-HMDI-Pd and G3-HMDI-Pd samples, one state of palladium
atoms was observed: Pd+2. However, the Pd content was twice as
high in G22-HMDI-Pd as in G2-HMDI-Pd. In G22-HMDI-Pd, two
shifts that indicate two oxidation states of palladium (Pd0 and
Pd+2) were apparent. In G1-PDI-Pd, G2-PDI-Pd, G22-PDI-Pd, G32-
PDI-Pd and G3-PDI-Pd, as shown in Table 1, two oxidation states
were observed for palladium. Therefore, while using the PDI cross-
linking agent to synthesise the catalyst, the dendrimer generation
and the Pd content do not influence the state of the palladium; the
same is true when using the HMDI cross-linking agent.
CH2-Nst); 1633 (C Ost in NH-C( O)); 1547, 1508 (N-H␦, CH2␦, C-N-
H␦); 1448, 1404 (CH2␦); 1298, 1241, 1211, 1182, 1107, 1043, 1011
(C-Nst in NH-C( O)); 924, 862, 806 (aromatic C-H␦); 756, 694, 671,
654 (aromatic C C␦).
XPS (eV): 284.3 (52.43%), 285.3 (26.50%), 286.2 (10.8%), 288.3
(10.28%) (C 1s, 60.7%) 335.1 (85%) (Pd 3d5/2, 1.6%), 338.0 (15%) (Pd
3d3/2); 399.5 (92.44%), 401.3 (7.56%) (N 1s, 21.9%); 531.6 (O 1s,
15.8%). ICP-AES: 13.73% Pd.
2.4. Catalytic test
A catalyst and a substrate (4 mg of the catalyst per 1 ml of sub-
strate) were placed in a thermostated steel autoclave equipped
with an insert test tube and a magnetic stirrer. If necessary, the sol-
vent was added before sealing the autoclave tightly, filling it with
5–10 atm hydrogen, and connecting it to a thermostat. The reac-
tion was performed with vigorous stirring at 80 ◦C for 15 min or
1 h; afterwards, the autoclave was cooled to below room tempera-
ture. The products were analysed using gas-liquid chromatography
(GLC) on a ChromPack CP9001 gas chromatograph with a flame
ionisation detector. The column was 30 m × 0.2 mm with an SE-30
grafted phase.
The binding energies for Pd 3d5/2 and Pd 3d3/2 are 335.7 eV
and 338.2 eV, respectively, i.e., they are slightly lower than that

E.A. Karakhanov et al. / Journal of Molecular Catalysis A: Chemical 397 (2015) 1–18
5
Table 1
Catalysts and their characterisation data.
Dendrimer
Cross-linking agent
Indication of a catalyst Pd (%) Pd 3d5/2 (эB) Pd 3d3/2 (eV) Surface atomic concentration (%) Particle size, d (nm)
PAMAM(NH2)8
PAMAM(NH2)8
PAMAM(NH2)8
PAMAM(NH2)8
PAMAM(NH2)8
PAMAM(NH2)16
PAMAM(NH2)16
PAMAM(NH2)16
PAMAM(NH2)16
PAMAM(NH2)16
PAMAM(NH2)16
PAMAM(NH2)32
PAMAM(NH2)32
OCN(CH2)4NCO
OCN(CH2)6NCO
OCN(CH2)8NCO
OCNC6H4NCO
p-(H3COC6H3NCO)2
OCN(CH2)6NCO
OCNC6H4NCO
OCN(CH2)6NCO
OCNC6H4NCO
OCNC6H4NCO
G1-BDI-Pd
15.83
14.8
5.45
6.5
13.12
1.61
4.61
3.33
6.23
6.8
335.0 (87%)
335.7 (62%)
335.2 (68%)
335.7 (45%)
335.2 (83%)
337.9 (13%)
338.2 (38%)
338.0 (32%)
338.2 (55%)
337.9 (17%)
338.5 100%)
338.0 (50%)
337.9 (50%)
338.0 (37%)
338.0 (49%)
337.9 (61%)
338.2 100%)
338.0 (15%)
4.7
8.9
1.2
3.9
6.5
0.3
1.3
0.3
1.4
1.5
1.0
0.3
1.6
1.47 0.08; 2.7 0.24
2.1 0.44
1.84 0.13
G1-HMDI-Pd
G1-OMDI-Pd
G1-PDI-Pd
G1-DMPDI-Pd
G2-HMDI-Pd
G2-PDI-Pd
G22-HMDI-Pd
G22-PDI-Pd
G32-PDI-Pd
G2-DMPDI-Pd
G3-HMDI-Pd
G3-PDI-Pd
2.49 0.25
2.29 0.1; 3.83 0.22
1.74 0.05; 3.55 0.24
1.07 0.02; 1.46 0.09
2.01 0.08; 3.13 0.15
1.09 0.12; 1.89 0.12
335.2 (50%)
334.8 (50%)
335.0 (63%)
335.2 (51%)
335.1 (39%)
p-(H3COC6H3NCO)2
OCN(CH2)6NCO
OCNC6H4NCO
4.09
1.66
13.73
2.22 0.11
1.92 0.49
1.55 0.03;2.25 0.03
335.1 (85%)
Scheme 1. Synthesis of polymeric materials based on cross-linked dendrimers.
Scheme 2. Encapsulation of the palladium nanoparticles in cross-linked dendrimers.
of metallic palladium (0) (Pd: 335.8 eV [25]) and a little bit higher
than that of divalent palladium (PdO: 337.9 [25]) The presence of
these species was confirmed by the presence of metal nanopar-
ticles coordinated by nitrogen or oxygen atoms. According to the
component ratios in the Pd 3d electron spectrum (%), the ratios of
the palladium metal (Pd0) to the oxidised palladium (Pd2+) on the
sample surface are as follows:
- 0:100 for G2-HMDI-Pd, 50:50 for G2-PDI-Pd, 50:50 for G22-
HMDI-Pd, 63:37 for G22-PDI-Pd, 51:49 for G32-PDI-Pd;
- 0:100 for G3-HMDI-Pd and 85:15 for the G3-PDI-Pd.
The XPS data confirmed the presence of the C, N and O atoms, as
well as entities such as CH2CH2C(O)NH (the binding energies for O
and N are 531.5 eV and 399.6 eV, respectively [25]), CH2CH2 (the C
binding energy is 284.9 eV [25]), CH2CH2NH2 (the C binding energy
is 285.5 eV [25]), (CH2NH2)Pd (the C binding energy is 286.8 eV [25])
and CH2CH2C(O)NH (the C binding energy is 288.1 eV [25]). The
- 87:13 for G1-BDI-Pd, 62:38 for G1-HMDI-Pd, 68:32 for G1-OMDI-
Pd, 45:55 for G1-PDI-Pd, 83:17 for G1-DMPDI-Pd;

6
E.A. Karakhanov et al. / Journal of Molecular Catalysis A: Chemical 397 (2015) 1–18
Fig. 1. C 1s-spectrum for G1-BDI-Pd.
Fig. 3. C 1s-spectrum for the G1-HMDI-Pd.
Fig. 2. C 1s-spectrum for G1-OMDI-Pd.
Fig. 4. C 1s-spectrum for G1-PDI-Pd.
energy shift for oxygen and nitrogen relative to the normal val-
ues (531.5 eV and 398.6 eV, respectively [25]) in the spectra for the
samples is due to a partial transfer of electrons to the neutral metal
particles. In the G2-PDI-Pd (2) and G3-PDI-Pd samples, the binding
energy of N is 401.3 eV, confirming the presence of C6H5NH and
indicating that the cross-linking agent is p-phenylene diisocyanate.
The carbon deconvolution spectra show that the bond energies
are approximately the same; however, the bond energies of C-C, C-
N, C-Pd and C-O differ significantly. While considering the binding
energies of samples with conformationally labile cross-linkers such
as BDI, HMDI, and OMDI, the intensity of the C-C bonds (from 37.96%
to 53.20%) increases, and the intensity of the C-N bonds decreases
(from 40.22% up to 30.34%) when increasing the chain length. The
highest intensity for the C-Pd bond (11.92%) is observed in the
sample containing the hexamethylene diisocyanate cross-linker,
while the lowest is observed in the sample with octamethylene
diisocyanate cross-linker (4.82%); the intensity of this bond is not
affected by the palladium content.
Fig. 5. C 1s-spectrum for G1-DMPDI-Pd.
The C 1s-spectra for the G1-BDI-Pd (Fig. 1) and G1-OMDI-Pd
(Fig. 2) are almost identical. There are two signals of 288.4 eV and
285.2 eV, which correspond to the energies of the C-O bond and the
C–C and C–N bonds, respectively. The carbonyl bond, C-O, formed
through the interactions of the amino groups on the dendrimer
and the cross-linker, is present in the dendrimer and the periph-
ery. The intensity ratio for the C-O and C-N bonds is the same in
all of the samples (0.38:1). Similarly, the analogous proportion for
NCO/NH2 = 1/2.6.
In addition, the deconvolution graphs for G1-HMDI-Pd (Fig. 3)
and G1-PDI-Pd (Fig. 4) have almost the same shape. Similarly,
we can calculate the NCO/NH2 ratio for these samples: 1/4.3. The
NCO/NH2 ratio is 1/3 in the G1-DMPDI-Pd sample (Fig. 5).
The intensities of the C and Pd bond are almost identical in the
samples with rigid cross-linkers; the percentage and surface atomic
concentration of palladium is approximately twice as high in the
sample of G1-DMPDI-Pd as in G1-PDI-Pd.

E.A. Karakhanov et al. / Journal of Molecular Catalysis A: Chemical 397 (2015) 1–18
7
Using a harder cross-linker (p-phenylene diisocyanate) rela-
tive to hexamethylene diisocyanate increases the cavity size and,
consequently, the particle size: 2.5 and 2.1 nm for G1-PDI-Pd and
G1-HMDI-Pd, respectively (Figs. 10 and 8).
When using
a
more rigid cross-linker (3,3ꢀ-dimethoxy-
4,4ꢀ-diphenylene diisocyanate), larger particles (2.3–2.5 and
3.8–4.0 nm, Fig. 11; 2.0–2.3 nm, Fig. 16) form due to the ␲-
interactions, which retain the palladium particles through the
additional phenyl ring.
3.4. Solid state NMR spectroscopy
The carbon polarisation transfer spectra (13C CPMAS) were
recorded to determine the local structure of the compounds. The
spectra of the G1 series are as follows (Fig. 19).
Some areas of the signals were observed for the G1-BDI-Pd sam-
ple. In the region from 15 to 45 ppm, the carbon atoms on the
methylene units of the initial dendrimer and the cross-linking agent
(BDI) absorb. In addition, the weaker the field of the signal (closer
to 50 ppm), the closer is the carbon atom to the nitrogen atom. The
carbon signals for in the CH2NH2 groups have the highest inten-
sity. The broader profile of the signals reveals the low regularity
of the dendrimer groups in the structure. The signal at approxi-
mately 160 ppm appears to be the carbon in the –NH–(C O)–NH–
group formed during cross-linking. The broad low-intensity signal
at 172 ppm might correspond to an unreacted isocyanate group
Fig. 6. Pd 3d-electron spectrum for G1-BDI-Pd.
The ratio of atomic concentrations of Pd (II)/Pd (0) is ∼1.22 in
the G1-BDI-Pd sample, as shown in Fig. 6, according to the decon-
volution spectrum.
3.3. TEM analysis
Based on the TEM data, the size and shape of the distributions
are affected by the cross-linker and dendrimer generation for com-
posites based on palladium nanoparticles (Figs. 7–18).
(–N
impurities.
Two intense signals at 33 ppm and approximately 40.5 ppm are
C O). The low-intensity signals at 83 and 115 ppm are likely
The monomodal and bimodal particle size distributions are
obtained based on the first generation of the PAMAM dendrimer
while using four different cross-linkers. The distribution is affected
by the chain length of the cross-linker. While using hexamethylene
diisocyanate, a monomodal distribution is observed (Fig. 8), but
while using butylene diisocyanate, a bimodal distribution (1.5 and
2.7 nm, Fig. 7) is observed. This difference might be associated with
the size of the cavities in the dendritic structure. While using this
cross-linker, two adjacent branches of the dendrimer bound due to
the short carbon chain of the cross-linker. Therefore, the cavities
for smaller particles of 1.4–1.6 nm are formed and stabilised by the
tertiary nitrogen atoms.
In addition, the pattern of the dendrimer generation exerts
a visible influence. While using hexamethylene diisocyanate, a
monomodal distribution is observed with maxima of 2.1 and 1.9 nm
for the samples obtained based on the first and third generations,
respectively (Figs. 8 and 17). For a sample obtained based on the
second generation, a bimodal distribution is observed with maxima
at 1.7 and 3.6 nm (Figs. 12 and 14). The distribution of the samples
obtained using another cross-linker, namely p-phenylene diiso-
cyanate, is similar. There is a monomodal distribution in the sample
based on the first generation (Fig. 10), while a bimodal distribution
is observed in the sample based on the second (Figs. 13 and 15)
and the third generations (Fig. 18). Increasing the generation leads
to a bimodal distribution; the 1.5–2-nm particles might be located
between the branches, while the 2.5–3-nm particles are situated in
cavities.
observed in the spectrum for the G1-HMDI-Pd sample. The first
signal corresponds to the carbon in the –CH2 groups ␤ to the amino
groups, while the second corresponds to an ␣ position carbon to the
–NH2 groups. The –N(CH2)3 signal is the low-intensity shoulder at
approximately 55 ppm. The low-intensity signal at approximately
160 ppm corresponds to the –NH–(C O)–NH– group.
The spectrum of the G1-OMDI-Pd sample is similar to that of
the G1-HMDI-Pd; two signals are observed in the aliphatic region
of the dendrimer, and one low-intensity signal corresponds to the
carbide carbon.
An intense signal at approximately 40 ppm (–CH2NH2 group)
spanning from 20 to 50 ppm is observed for the G1-PDI-Pd sample.
The –N(CH2)3 groups are in the shoulder at approximately 55 ppm.
The signals from 120 to 135 ppm correspond to the carbon centre
of the benzene ring; the lowest field signal (at 133.2 ppm) is the
carbon linked to the –NH–(C O)–NH– group. The signal at approx-
imately 160 ppm corresponds to the carbonyl carbon. The spectra
show that the G1-PDI-Pd and G1-PDI samples are identical without
palladium nanoparticles.
For the G1-DMPDI-Pd sample from 20 to 60 ppm, the two
narrow intense signals (40.6 and 55.3 ppm) correspond to the
–CH2NH2 and –N(CH2)3 groups. Notably, the signal at approxi-
mately 55 ppm in other samples was less intense. This effect might
be caused by the insertion of the Pd nanoparticles, which favours
the carbon centre patterning due to their coordination. Therefore,
for cross-linked dendrimers, the type of cross-linking agent might
favour the conditions necessary for Pd nanoparticle coordination in
all dendrimer nitrogen centres due to its bulk and rigidity. In addi-
tion, the –OMe groups bound to the benzene ring absorb at 55 ppm.
The group of signals between 105 and 160 ppm corresponds to the
carbon centres of the benzene ring, with the lowest field signal
(the carbon bound to the OMe group) at approximately 160 ppm
overlapping with the –NH–(C O)–NH– signal.
The palladium content does not affect the particle distribu-
tion in the G2-HMDI-Pd and G22-HMDI-Pd catalysts, as shown in
Figs. 6 and 8. A similar situation is observed for the G2-PDI-Pd and
G22-PDI-Pd samples (Figs. 7 and 9). Increasing the palladium con-
tent increases the particle sizes: 1.7 and 2.3 for 1.66% Pd and 2.1
and 3.2 for 3.33% Pd in the G2-HMDI-Pd and G22-HMDI-Pd samples,
respectively (Figs. 12 and 14). Similarly, while using another cross-
linker (p-phenylene diisocyanate) the particle sizes are as follows:
1.0 and 1.6 nm are for 4.61% Pd, and 1.2 and 1.9 nm are for 6.23% Pd
for the G2-PDI-Pd and G22-PDI-Pd samples (Figs. 13 and 15).
The spectra of the G2 series (the samples based on the second-
generation dendrimer) are shown in Fig. 20.
The interpretation of these spectra is analogous to that of the
G1 series (the samples synthesised based on the second-generation

8
E.A. Karakhanov et al. / Journal of Molecular Catalysis A: Chemical 397 (2015) 1–18
Fig. 7. TEM data and particle size distribution for G1-BDI-Pd.
Fig. 8. TEM data and particle size distribution for G1-HMDI-Pd.
Fig. 9. TEM data and particle size distribution for G1-OMDI-Pd.
dendrimers). Therefore, only the very weak signal attributed to the
–NH–(C O)–NH– group is notable for the samples with the HMDI
cross-linking agent in contrast to the sample with the PDI cross-
linking agent. This weakness in the signal might occur because the
carbide carbonyl is labile and/or the cross-linker content is low.
When comparing the spectra of the G2-PDI-Pd and G22-PDI-
Pd samples, the signals assigned to the dendrimer are different.
G2-PDI-Pd has only one intense signal at approximately 44 ppm
in contrast to G22-PDI-Pd. This difference might be due to the
differing amounts of palladium in the dendrimer. The palladium
nanoparticles in G2-PDI-Pd might coordinate with the –(CH2)3N
centres, leaving the –CH2NH2 centre free and imparting a labil-
ity higher than the original dendrimer; therefore, the polarisation
transfer to these centres is low over a given contact time. The
CH2NH2 sample centres are locked in the G22-PDI-Pd structure,
leaving them visible.
The spectra for the samples with the HMDI cross-linking agent,
including the G3-HMDI-Pd sample, are shown below in Fig. 21. The
spectra are almost identical, except that the signal at approximately
51 ppm for G3-HMDI-Pd is the most intense.

E.A. Karakhanov et al. / Journal of Molecular Catalysis A: Chemical 397 (2015) 1–18
9
Fig. 10. TEM data and particle size distribution for G1-PDI-Pd.
Fig. 11. TEM data and particle size distribution for G1-DMPDI-Pd.
Fig. 12. TEM data and particle size distribution for G2-HMDI-Pd.
3.5. Hydrogenation of unsaturated compounds
to the increase in palladium content, and the dendrimer generation
does not affect the conversion.
3.5.1. Hydrogenation of octene-1
When using catalysts based on the first- and second-generation
PAMAM dendrimers with the PDI cross-linker, the yields
increase in the following order: G2-PDI-Pd « G1-PDI-Pd < G22-PDI-
Pd (Fig. 22b). Moreover, the palladium content is nearly equal in the
G1-PDI-Pd and G22-PDI-Pd samples, but the conversion is approx-
imately 10% higher for G22-PDI-Pd, possibly due to the size of the
palladium particles. The particle size in the first sample is twice
that of the second.
The synthesised materials were tested as hydrogenation cata-
lysts for unsaturated compounds. When a terminal linear olefin,
such as octene-1, was hydrogenated in the presence of the synthe-
sised materials (Fig. 22), the maximum conversions were achieved
using G1-HMDI-Pd and G22-PDI-Pd: 83% and 78%, respectively.
When using catalysts based on the first-, second- and third-
generation PAMAM dendrimers with the HMDI cross-linker, the
yields increase in the following order: G3-HMDI-Pd < G2-HMDI-
Pd « G22-HMDI-Pd < G1-HMDI-Pd (Fig. 22a). This trend corresponds
When using catalysts based on the first generation of den-
drimers and various cross-linkers, another relationship is observed

10
E.A. Karakhanov et al. / Journal of Molecular Catalysis A: Chemical 397 (2015) 1–18
Fig. 13. TEM data and particle size distribution for G2-PDI-Pd.
Fig. 14. TEM data and particle size distribution for G22-HMDI-Pd.
(Fig. 22c). The yields increase in the following order: G1-OMDI-
Pd < G1-BDI-Pd < G1-PDI-Pd < G1-DMPDI-Pd < G1-HMDI-Pd. In this
case, the conversion does not depend on the palladium content
because the yields are almost identical when using G1-BDI-Pd
(15.83%) and G1-PDI-Pd (6.5%), namely 63% and 69%, respectively.
Based on this graph, the conversion is affected by the cross-
linker. A low conversion can be explained by steric hindrance
during the substrate coordination when using conformation-
ally rigid cross-linkers such as para-phenylene diisocyanate and
3,3ꢀ-dimethoxy-4,4ꢀ-biphenylene diisocyanate and “short” cross-
linkers such as 1,4-butylene diisocyanate.
Graphs describing the activity toward octene hydrogenation
are shown in Fig. 23. When using catalysts based on the PAMAM
dendrimers of various generations with the HMDI linker, the G22-
HMDI-Pd sample exhibits highest activity (TOF = 13.6 × 103 h−1
)
(Fig. 23a).
For similar catalysts with the PDI cross-linking agent, the max-
imal value of TOF of 8.5 × 103 h−1 is observed in the G22-PDI-Pd
Fig. 15. TEM data and particle size distribution for G22-PDI-Pd.

E.A. Karakhanov et al. / Journal of Molecular Catalysis A: Chemical 397 (2015) 1–18
11
Fig. 16. TEM data and particle size distribution for G2-DMPDI-Pd.
Fig. 17. TEM data and particle size distribution for G3-HMDI-Pd.
Fig. 18. TEM data and particle size distribution for G3-PDI-Pd.
sample (Fig. 23b); this value is less than that while using the HMDI
cross-linker.
in the presence of palladium is quite facile (Scheme 3), generating
internal alkenes that are more thermodynamically stable internal
alkenes. The latter product is more sterically hindered than termi-
nal alkenes, preventing the substrate from taking the conformation
necessary for the reaction with the available ligand microenviron-
ment; therefore, these substrates are hydrogenated with low yields.
The variations in the octene hydrogenation when using the
first-generation PAMAM dendrimer with various cross-linking
agents are shown in Fig. 23 (graph c). The catalytic activity
decreased in the following order due to the increasing steric
hindrance during substrate coordination: G1-PDI-Pd » G1-DMPDI-
Pd > G1-HMDI-Pd > G1-OMDI-Pd > G1-BDI-Pd.
Simultaneously, the proportion of the double bond isomeri-
sation product increases in the reaction mixture; the higher the
proportion, the longer is the chain. The isomerism of double bonds
3.5.2. Hydrogenation of styrene, p-methylstyrene and
p-tert-butylstyrene
Materials based on palladium nanoparticles and cross-linked
dendrimers were highly active during the hydrogenation of

12
E.A. Karakhanov et al. / Journal of Molecular Catalysis A: Chemical 397 (2015) 1–18
Fig. 19. 13C CP MAS spectra for the samples synthesised based on the first generation PAMAM dendrimer with various cross-linking agents.
Fig. 20. 13C CP MAS spectra for the samples synthesised based on the second-generation PAMAM dendrimer.
styrene and its derivatives, forming the corresponding ethylben-
which has the lowest palladium content. The G3-HMDI-Pd cata-
lyst is the most active during the styrene hydrogenation; its TOF is
35.2 × 103 h−1 (Fig. 24b).
zene (Figs. 24–26, Scheme 4), with a TOF value close to 86 × 103 h−1
,
which is significantly higher than that for 1-octene.
The best catalyst for the hydrogenation of styrene and p-tert-
butylstyrene is G1-PDI-Pd; 100% conversion was achieved in all of
the reactions using this catalyst (Fig. 25a).
In this case, the yields depended on the size of the substrate.
Therefore, for G1-HMDI-Pd, regular increases in activity were
observed when increasing the size of the substrate; this cata-
lyst is not affected by steric hindrance. However, the G1-BDI-Pd,
G1-OMDI-Pd, G22-HMDI-Pd, G3-HMDI-Pd and G3-PDI-Pd catalysts
exhibited a parabolic dependence on the size of the substrate for
the activity; a minimum for p-methylstyrene and maximum for
styrene were characteristic (Figs. 26a and 24a).
The effect of the PAMAM dendrimer generation with the
1,6-hexamethyldiisocyanate linker on the conversion and hydro-
genation activity of styrene is shown in Fig. 24. The highest
conversion is observed with the G1-HMDI-Pd catalyst; the con-
version also increases when increasing the size of the substrate
(Fig. 24a). Therefore, the conversion is affected by the palladium
content rather than the dendrimer generation. The worst catalyst
for the hydrogenation of styrene and its derivatives is G3-HMDI-Pd,
For these materials, the substrate has an optimal size
corresponding to the distance between the branches of the den-
dritic matrix. For the G1-DMPDI-Pd, G2-PDI-Pd, G22-PDI-Pd and

E.A. Karakhanov et al. / Journal of Molecular Catalysis A: Chemical 397 (2015) 1–18
13
Fig. 21. 13C CP MAS spectra for the samples synthesised based on the first-, second- and third-generation PAMAM dendrimers with the HMDI cross-linker.
Fig. 22. The relationship between dendrimer generation and the conversion of the cross-linked palladium catalysts (HMDI (a) and PDI (b)) and the effect of the cross-linking
agent (c). Reaction conditions: 4 mg of the catalyst per 1 mL of the substrate, 80 ◦C, 1 h, 10 atm H2.
Fig. 23. The effect of dendrimer generation on the hydrogenation activity for palladium catalysts with the HMDI (a) and PDI cross-linkers (b), as well as the effect of the
cross-linking agent (c). Reaction conditions: 2 mg of the catalyst per 1 mL of the substrate, 80 ◦C, 1 h, 10 atm H2.

14
E.A. Karakhanov et al. / Journal of Molecular Catalysis A: Chemical 397 (2015) 1–18
Scheme 3. Isomerisation of the double bond, as exemplified by octene-1.
Fig. 24. The effect of the dendrimer generation on (a) the conversion and (b) the hydrogenation activity of the substrate over palladium catalysts with the HMDI cross-linker.
Reaction conditions: (1) styrene: 0.67 mg of the cat., 1 mL substrate, 80 ◦C, 15 min, 5 atm H2; (2) p-Me-styrene: 0.83 mg cat., 1 mL substrate, 80 ◦C, 15 min, 5 atm H2; (2)
p-t-Bu-styrene: 2.5 mg cat., 1 mL of the substrate, 80 ◦C, 15 min, 5 atm H2.
Fig. 25. The effect of dendrimer generation on (a) the conversion and (b) the hydrogenation activity of a substrate for palladium catalysts with the PDI cross-linker. Reaction
conditions: (1) styrene: 0.67 mg of the cat., 1 mL of the substrate, 80 ◦C, 15 min, 5 atm H2; (2) p-Me-styrene: 0.83 mg of the cat., 1 mL of the substrate, 80 ◦C, 15 min, 5 atm H2;
(2) p-t-Bu-styrene: 2.5 mg of the cat., 1 mL of the substrate, 80 ◦C, 15 min, 5 atm H2.
Scheme 4. The hydrogenation of styrene to form the corresponding alkyl benzenes.
G32-PDI-Pd catalysts, which contain a phenyl ring, p-phenylene
and 3,3ꢀ-dimethoxy-4,4ꢀ-diphenylene diisocyanate are used as the
cross-linkers, and a parabolic dependence on the value of the alkyl
substituent para to the vinyl group of styrene is also typical. The
maximum activity is achieved during the hydrogenation of p-tert-
butylstyrene (Fig. 25a).
A similar dependence between the PAMAM dendrimer gen-
eration and the PDI cross-linker is shown in Fig. 25. Therefore,
the catalysts containing p-phenylene diisocyanate as a linker
hydrogenate p-methylstyrene poorly, while the conversion for the
hydrogenation of styrene and p-tert-butylstyrene exceeds 60% for
the G1-PDI-Pd, G22-PDI-Pd and G32-PDI-Pd catalysts. As observed
in this graph, the Pd content does not affect the conversion: the
Pd contents in the G1-PDI-Pd and G3-PDI-Pd samples are 15.83%
and 13.73%, respectively, and the conversions of styrene are 100%
and 20%, respectively. This effect can be explained by the surface
atomic concentration of palladium because, in the second sample,
little palladium is located at the surface (Table 1). The lowest activ-
ity is exhibited by all catalysts containing this active cross-linking
agent during the hydrogenation of p-tert-butylstyrene (Fig. 25a).
The substrate-dependence of conversion during hydrogenation
using catalysts based on the first-generation PAMAM dendrimer
with various cross-linking agents is shown in Fig. 25a. The worst
conversion is observed with the G1-OMDI-Pd sample. When

E.A. Karakhanov et al. / Journal of Molecular Catalysis A: Chemical 397 (2015) 1–18
15
Fig. 26. The effect of the cross-linking agent on (a) the conversion and (b) the hydrogenation activity of the substrate over the palladium catalysts. Reaction conditions:
(1) styrene: 0.67 mg of the cat., 1 mL of the substrate, 80 ◦C, 15 min, 5 atm H2; (2) p-Me-styrene: 0.83 mg of the cat., 1 mL of the substrate, 80 ◦C, 15 min, 5 atm H2; (2)
p-t-Bu-styrene: 2.5 mg cat., 1 mL of the substrate, 80 ◦C, 15 min, 5 atm H2.
Scheme 5. The hydrogenation of phenylacetylene to form styrene and ethylben-
zene.
comparing catalysts, the cross-linking agents that were conforma-
tionally flexible linkers such as the BDI, HMDI and OMDI were the
Scheme 6. Hydrogenation of 2,5-dimethyl-2,4-hexadiene.
best. A high conversion was achieved when using the G1-HMDI-Pd
sample due to its superior pore structure; the pores facilitate the
approach of the styrene toward the catalytic substrate. In addition,
the palladium content does not affect the conversion (Fig. 25a).
When using rigid linkers such as PDI and DMPDI, the best result
was observed with the G1-PDI-Pd sample, where the Pd content is
twice that in the G1-DMPDI sample.
A similar situation is observed when considering the activity of
the catalyst based on the first-generation PAMAM dendrimer with
various cross-linking agents (Figs. 24b and 25b). These samples are
the least active toward the hydrogenation of p-tert-butylstyrene
due to steric hindrance. Apparently, the optimal substrate size cor-
responds to the distance between the branches of the dendritic
matrix. During the hydrogenation of styrene, the catalyst activity
decreases in the following order: G1-PDI-Pd » G1-OMDI-Pd > G1-
DMPDI-Pd > G1-HMDI-Pd > G1-BDI-Pd (Fig. 26b).
During the hydrogenation of phenylacetylene over 1 h using
catalysts based on the first generation dendrimers with different
cross-linkers, 100% conversion is achieved with a 99% selectivity
for styrene when using G1-PDI-Pd (Table 2). When hydrogenating
styrene and its derivatives, the results depend on the optimal aver-
age palladium particle size (2.5 nm). When hexamethylene is used
as a cross-linker, the selectivity is 0%; specifically, the hydrogena-
tion runs to completion, forming ethylbenzene in a 100% yield. The
hydrogenation is not affected by the palladium content when using
catalysts based on the first-generation dendrimer.
The G1-DMPDI-Pd and G22-PDI-Pd catalysts are the least
selective toward styrene (the selectivity is 11%), similar to the
phenylacetylene hydrogenation that occurred over 15 min.
The G2-HMDI-Pd and G3-HMDI-Pd catalysts had similar pal-
ladium contents: 1.61% and 1.66%, respectively. These materials
hydrogenate phenylacetylene with almost the same yield and
selectivity. Therefore, the dendrimer generation does not affect the
conversion and selectivity.
3.5.3. Hydrogenation of phenylacetylene
During the hydrogenation of phenylacetylene (Scheme 5) over
15 min, 100% conversion is achieved when using two catalysts:
G1-HMDI-Pd and G1-BDI-Pd (Table 2). G1-PDI-Pd achieves a 98%
conversion. These high conversions may be attributed to the
higher flexibility of the cross-linking agents. The palladium con-
tent exerts no effect on the conversion of phenylacetylene; the
conversions of the G1-HMDI-Pd and G1-PDI-Pd samples are 14.8%
and 6.8%, respectively. The G2-HMDI-Pd, G3-HMDI-Pd and G22-
PDI-Pd samples produce the worst results with 17%, 11% and
15% conversion, respectively. Therefore, the conversion decreases
when increasing the dendrimer generation: G1-HMDI-Pd » G2-
HMDI-Pd > G3-HMDI-Pd. The same trend is observed when using
catalysts with the PDI cross-linking agent: G1-PDI-Pd » G22-PDI-Pd.
All of the samples were selective with respect to styrene except
the G1-DMPDI-Pd and G22-PDI-Pd samples. The hydrogenation
of phenylacetylene over these catalysts generates ethylbenzene
quantitatively; this result may be explained by the rigid cross-
linkers.
The best catalyst for phenyl acetylene hydrogenation is G1-
HMDI-Pd, with 100% yield of styrene being formed in the reaction
for 15 min, and 100% yield of ethylbenzene is produced in the reac-
tion for 1 h. The most active catalyst is the G1-PDI-Pd, with a TOF of
14.5 × 103 h−1 in the hydrogenation of phenyl acetylene for 15 min;
using G2-PDI-Pd, the TOF is 13 × 103 h−1 in hydrogenation for 1 h.
3.5.4. Hydrogenation of 2,5-dimethyl-2,4-hexadiene
During the hydrogenation of 2,5-dimethyl-2,4-hexadiene, four
products were obtained: 2,5-dimethyl-hexane, 2,5-dimethyl-2-
hexene, 2,5-dimethyl-3(trans)-hexene, and 2,5-dimethyl-3(cis)-
hexene (Scheme 6). All of the catalysts were selective toward
2,5-dimethyl-3(trans)-hexene with selectivities ranging from 70
to 77% (Table 3). The table shows that the selectivity for 2,5-
dimethyl-hexane ranges from 0.4 to 2.5%. The selectivity for
2,5-dimethyl-3(cis)-hexene ranges from 3 to 6%, while that of 2,5-
dimethyl-2-hexene ranges from 18 to 24%.

16
E.A. Karakhanov et al. / Journal of Molecular Catalysis A: Chemical 397 (2015) 1–18
Table 2
Hydrogenation of phenylacetylene in the presence of Pd catalysts based on dendrimers.
No.
Catalyst
Conversion % (15 min)
Selectivities, % (15 min)
TOF (h−1
)
Conversion % (1 h)
Selectivities, % (15 min)
TOF (h−1
)
1
2
3
4
5
6
7
8
G1-BDI-Pd
100
100
19
98
67
17
64
25
15
74
11
67
78
100
82
95
0
91
82
88
0
6112
6534
6746
14580
4911
9849
26863
7264
2329
17505
6610
9442
100
100
28
100
100
60
85
58
51
98
64
0
1548
1634
2248
3720
1251
3189
12935
4267
2006
5870
8413
3569
G1-HMDI-Pd
G1-OMDI-Pd
G1-PDI-Pd
G1-DMPDI-Pd
G2-HMDI-Pd
G2-PDI-Pd
G22-HMDI-Pd
G22-PDI-Pd
G2-DMPDI-Pd
G3-HMDI-Pd
G3-PDI-Pd
80
99
11
94
94
88
11
0
9
10
11
12
98
95
93
57
100
93
100
Reaction conditions: 4 mg of the catalyst per 1 mL of the substrate, 80 ◦C, 10 atm H2.
Table 3
Hydrogenation of 2,5-dimethyl-2,4-hexadiene in the presence of Pd catalysts based on dendrimers.
No.
Catalyst
Conversion (%)
Selectivities, %
2,5-dimethyl-
hexane
Selectivities, %
2,5-dimethyl-
2-hexene
Selectivities, %
2,5-dimethyl-
3(trans)-hexene
Selectivities, %
2,5-dimethyl-
3(cis)-hexene
TOF (h−1
)
1
2
3
4
5
6
7
8
G1-BDI-Pd
70
54
38
61
95
9
0.6
0.6
0.8
0.4
0.4
2.5
2.3
1.4
0.7
0.7
2.6
2
19
20
18
21
24
19
22
19
22
21
20
20
21
76
74
77
74
70
75
74
77
73
74
75
75
75
4
5
4
5
6
3
1,7
3
4
4
13176
10851
20778
27966
21578
16657
4525
13352
8610
20159
14572
19747
2605
G1-HMDI-Pd
G1-OMDI-Pd
G1-PDI-Pd
G1-DMPDI-Pd
G2-HMDI-Pd
G2-PDI-Pd
G22-HMDI-Pd
G22-PDI-Pd
G32-PDI-Pd
G2-DMPDI-Pd
G3-HMDI-Pd
G3-PDI-Pd
7
17
18
46
20
11
12
9
10
11
12
13
3,4
3
2,7
1.3
Reaction conditions: 1 mg of the catalyst per 1 mL of the substrate, 80 ◦C, 15 min, 10 atm H2.
During the hydrogenation of 2,5-dimethyl-2,4-hexadiene using
the catalysts based on the first-generation PAMAM dendrimer
and various cross-linkers, the conversion decreases in the fol-
lowing order: G1-DMPDI-Pd > G1-BDI-Pd > G1-PDI-Pd > G1-HMDI-
Pd > G1-OMDI-Pd. This trend may be related to the pore sizes of the
cross-linked dendrimer, while the palladium content and particle
size do not affect the conversion (Table 3).
When using the catalysts based on the PAMAM dendrimers with
hexamethylene diisocyanate, the cross-linking agent conversion
decreases in the following sequence: G1-HMDI-Pd » G22-HMDI-
Pd > G3-HMDI-Pd > G2-HMDI-Pd. This trend may also be explained
by the pore sizes. The difference in conversion between the
G2-HMDI-Pd and G3-HMDI-Pd samples is only 2. Therefore, the
conversion is affected by the palladium content.
Fig. 27. DCPD adsorption on the Pd nanoparticles.
When using catalysts based on the first- and second-
generation PAMAM dendrimers with p-phenylene diisocyanate,
the conversion decreases in the following sequence: G1-PDI-
Pd » G32-PDI-Pd > G22-PDI-Pd. The effects of the palladium content
and its surface atomic content on the activity of the catalyst are
observed.
The most active catalyst during the hydrogenation of 2,5-
dimethyl-2,4-hexadiene is the G1-PDI-Pd sample; its TOF was
28 × 103 h−1. The catalytic activity decreases in the following
order: G1-PDI-Pd > G1-DMPDI-Pd > G1-OMDI-Pd » G1-BDI-Pd > G1-
HMDI-Pd. This trend can be explained by the increased steric
hindrance during substrate coordination. When using catalysts
based on the first-, second- and third-generation PAMAM den-
drimers with the HMDI cross-linker, the activity increases in the
following order: G1-HMDI-Pd < G22-HMDI-Pd < G2-HMDI-Pd < G3-
HMDI-Pd. When using the catalysts with the PDI linker, the catalyst
activity decreases when increasing the PAMAM dendrimer gener-
ation.
3.5.5. Hydrogenation of dicyclopentadiene
Materials based on dendrimer-encapsulated palladium
nanoparticles are effective catalysts during the hydrogena-
tion of dicyclopentadiene (Table 4). Five products are formed: the
endo- and exo-isomers with small quantities of cyclopentadiene,
cyclopentane and cyclopentene (Scheme 7).
During the hydrogenation, the configuration of the initial dicy-
clopentadiene (endo) is maintained, and the less sterically hindered
and higher energy double bond is hydrogenated (Fig. 27), match-
ing the published data [26]. The amount of diene decomposition
products did not exceed 13% for G1-HMDI-Pd, G2-HMDI-Pd, G2-
PDI-Pd, G22-HMDI-Pd, G22-PDI-Pd, and G3-HMDI-Pd. The amounts
of decomposition were 21%, 37% and 31% for G1-BDI-Pd, G1-PDI-Pd
and G1-DMPDI-Pd, respectively.
The highest hydrogenation conversions were 92% and 96% for
the G1-HMDI-Pd and G1-PDI-Pd samples, respectively, due to
their nearly identical particle sizes (2.1 and 2.49 nm). When using

E.A. Karakhanov et al. / Journal of Molecular Catalysis A: Chemical 397 (2015) 1–18
17
Table 4
Hydrogenation of dicyclopentadiene in the presence of Pd catalysts based on dendrimers.
No.
Catalyst
Conversion, % (15 min)
% Percentage of decomposition
Internal conversion (%)
Endo/exo
TOF (h−1
)
1
2
3
4
5
6
7
8
G1-BDI-Pd
78
92
17
96
61
21
17
26
79
73
34
9
21
5
9
37
31
6
79
97
18
98
89
23
18
28
91
74
35
10
7:1
23:1
6:1
16:1
10:1
29:1
7:1
28:1
61:1
39:1
15:1
8:1
7908
9982
5006
24206
7462
20934
5918
12531
20352
28647
32873
1052
G1-HMDI-Pd
G1-OMDI-Pd
G1-PDI-Pd
G1-DMPDI-Pd
G2-HMDI-Pd
G2-PDI-Pd
G22-HMDI-Pd
G22-PDI-Pd
G2-DMPDI-Pd
G3-HMDI-Pd
G3-PDI-Pd
4
7
9
13
0.06
2
10
11
12
8
Reaction conditions: 2 mg of the catalyst per 1 mg of the substrate, 80 ◦C, 15 min, 10 atm H2.
Scheme 7. The hydrogenation of dicyclopentadiene.
the first-generation catalysts with various cross-linkers, a rela-
tively high conversion was achieved. The conversion increases
in the following order based on the optimal particle sizes
and palladium contents: G1-OMDI-Pd « G1-DMPDI-Pd < G1-BDI-
Pd « G1-HMDI-Pd < G1-PDI-Pd. The major products underwent the
hydrogenation of only one bond through the endo-isomer. The ratio
of endo- to exo-products ranged from 7:1 to 23:1. Therefore, the
specific activity of the catalysts increases.
highly active and selective catalysts for the hydrogenation of unsat-
urated compounds. The activity and selectivity of these systems
depended on the nature and the structure of the dendritic matrix,
as well as its affinity to the substrate. The maximal yields and activ-
ities are achieved during the hydrogenation of styrene to form the
corresponding ethyl benzene.
Acknowledgments
When using the catalysts based on the first-, second- and
third-generation dendrimers with the HMDI cross-linker, the
yields increase in the following order due to the particle size
and surface concentration of palladium: G2-HMDI-Pd < G22-HMDI-
Pd < G3-HMDI-Pd « G1-HMDI-Pd. In addition, the conversion is
affected by the dendrimer generation. The conversion while using
the G3-HMDI-Pd sample exceeds that obtained while using G2-
HMDI-Pd; the palladium contents of these samples are almost
identical (1.61 and 1.66%, respectively), but the particle sizes are
1 92 nm and 1.74, respectively. The best selectivity is achieved by
using G1-HMDI-Pd and G22-HMDI-Pd: 96% and 97%, respectively.
When using the catalysts based on the first-, second- and third-
generation dendrimers with the PDI linker, the yields increase in
the following order due to the palladium contents and surface palla-
dium content: G2-PDI-Pd « G22-PDI-Pd < G1-PDI-Pd. This trend can
be explained by the average particle sizes: 1.07 and 1.46 nm for
G2-PDI-Pd, 1.09 and 1.89 nm for G22-PDI-Pd and 2.49 nm for G1-
PDI-Pd. The last sample is optimal in that the highest conversion is
achieved while using this sample. G22-PDI-Pd exhibits the highest
selectivity (98%). Similarly, the specific activity and the conversion
of the catalysts varies. The worst catalyst during the hydrogenation
of dicyclopentadiene is G2-PDI-Pd because the average particle size
(1.07 and 1.46 nm) is far from the optimal range (2.1–2.5 nm).
This work was supported by RFBR Grant no. 14-03-3153-A.
References
[1] R.A. Findeis, L.H. Gade, Eur. J. Inorg. Chem. (2003) 99–110.
[2] V. Marval, A.M. Caminade, J.P. Мɑjoral, Organometallics 19 (2000) 4015.
[3] L. Ropartz, R.E. Morris, D.F. Foster, D.J. Cole-Hamilton, Chem. Commun. (2001)
361–362.
[4] Y. Ribourdouille, G.D. Engel, M. Richard-Plouet, L.H. Gade, Chem. Commun.
(2003) 1228–1229.
[5] S. Antebi, P. Arya, L.E. Manzer, H. Alper, J. Org. Chem. 67 (2002) 6623–6631.
[6] M. Pittelkow, K. Moth-Poulsen, U. Boas, J.B. Christensen, Langmuir 19 (2003)
7682–7684.
[7] E.H. Rahim, F.S. Kamounah, J. Frederiksen, J.B. Christensen, Nano Lett. 1 (9)
(2001) 499–501.
[8] K. Heuze, D. Mery, D. Gauss, J.-C. Blais, D. Astruc, Chem. Eur. J. 10 (2004)
3936–3944.
[9] T. Borkowski, P. Subik, A.M. Trzeciak, S. Wolowiec, Molecules 16 (2011)
427–441.
[10] E.A. Karakhanov, A.L. Maksimov, A.V. Zolotukhina, Y.S. Kardasheva, Russian
Chem. Bull. 62 (7) (2013) 1465–1492.
[11] E.A. Karakhanov, A.L. Maximov, Y.S. Kardasheva, V. Semernina, A.V. Zolo-
tukhina, A.O. Ivanov, G. Abbott, E. Rosenberg, V. Vinokurov, ACS Appl. Mater.
Interf. 6 (10) (2014) 8807–8816.
[12] K. Olofsson, C.J. Andren, M. Malkoch, J. Appl. Polym. Sci. 131 (3) (2014) 39876.
[13] Y. Niu, R.M. Crooks, C.R. Chimie 6 (2003) 1049–1059.
[14] M. Bernechea, E. de Jesus, C. Lypez-Mardomingo, P. Terreros, Inorg. Chem. 48
(2009) 4491–4496.
[15] D. Bharathi, A.H. Ronald, M. de V Melgardt, Int. J. Pharm. 284 (2004) 133–140.
[16] Y. Niu, L. Sun, R.M. Crooks, Macromol. 36 (2003) 5725–5731.
[17] S.-K. Oh, Y. Niu, R.M. Crooks, Langmuir 21 (2005) 10209–10213.
[18] R. Andres, E. de Jesus, J.C. Flores, New J. Chem. 31 (2007) 1161–1191.
[19] E.A. Karakhanov, A.L. Maximov, V.A. Skorkin, A.V. Zolotukhina, A.S. Smerdov,
A.Y. Tereshchenko, Pure Appl. Chem. 81 (11) (2009) 2013–2023.
4. Conclusions
Therefore, using nanostructured entities based on cross-linked
dendrimers and palladium nanoparticles allows the synthesis of

18
E.A. Karakhanov et al. / Journal of Molecular Catalysis A: Chemical 397 (2015) 1–18
[20] E.A. Karakhanov, A.L. Maksimov, A.V. Zolotukhina, S.V. Kardashev, Petroleum
Chem. 50 (4) (2010) 290–297.
[21] E.A. Karakhanov, A.L. Maximov, S.V. Kardashev, Y.S. Kardasheva, A.V. Zolo-
tukhina, E. Rosenberg, J. Allen, Macromol. Symp. 304 (1) (2011) 55–64.
[22] E.A. Karakhanov, A.L. Maksimov, A.V. Zolotukhina, S.V. Kardashev, T. Yu Filip-
pova, Petroleum Chem. 52 (5) (2012) 289–298.
[23] W. Soichiro, I. Michiko, J. Photochem. Photobiol. A: Chem. 155 (2003) 57–62.
[24] R.M. Crooks, M. Zhao, L. Sun, V. Chechik, L.K. Yeung, Acc. Chem. Res. 34 (2001)
181–190.
[25] NIST. Electronic data base of the XPS. http://srdata.nist.gov/xps
[26] M. Hao, B. Yang, H. Wang, G. Liu, S. Qi, J. Yang, Ch. Li, J. Lu, J. Phys. Chem. A 114
(11) (2010) 3811–3817.