Selective hydrogenation of 3-Hexyn-1-ol with Pd nanoparticles synthesized via microemulsions

Article abstract of DOI:10.1016/j.apcata.2017.03.038

In the study at hand we present a design strategy for novel catalysts which can be used for the selective hydrogenation of alkynes to alkenes. The design of the novel catalysts is based on two main ideas, namely (1) the synthesis of Pd nanoparticles via microemulsions and (2) the use of highly-ordered mesoporous silica with a 3-D pore network (FDU-12) serving as support. The nanoparticles are deposited on FDU-12 in two different ways. Firstly, we simply impregnated the support with a dispersion of the nanoparticles. The resulting catalyst was not selective at all; on the contrary, it fully hydrogenated our model alkyne, namely 3-hexyn-1-ol. Secondly, we synthesized the FDU-12 in the presence of the nanoparticles (in-situ synthesis). In this case, we obtained one catalyst which performed as well as the Lindlar catalyst although the metal content was slightly lower and our catalyst contained no Pb. Another catalyst of the same series, prepared in the presence of another stabilizer, performed as well as the NanoSelect catalyst but at a 7 times higher metal content. For the sake of comparison we also impregnated FDU-12 via classical incipient wetness impregnation and again obtained a completely nonselective catalyst. Our results demonstrate that the in-situ synthesis has great potential as regards the development of novel catalysts.

Full text of DOI:10.1016/j.apcata.2017.03.038

Accepted Manuscript
Title: Selective hydrogenation of 3-Hexyn-1-ol with Pd
nanoparticles synthesized via microemulsions
Authors: Thomas Montsch, Moritz Heuchel, Yvonne Traa,
Elias Klemm, Cosima Stubenrauch
PII:
S0926-860X(17)30147-3
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.apcata.2017.03.038
APCATA 16197
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
3-2-2017
24-3-2017
28-3-2017
Please cite this article as: Thomas Montsch, Moritz Heuchel, Yvonne Traa,
Elias Klemm, Cosima Stubenrauch, Selective hydrogenation of 3-Hexyn-1-ol
with Pd nanoparticles synthesized via microemulsions, Applied Catalysis A,
Generalhttp://dx.doi.org/10.1016/j.apcata.2017.03.038
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Selective Hydrogenation of 3-Hexyn-1-ol with Pd Nanoparticles
Synthesized via Microemulsions
Thomas Montsch [a], Moritz Heuchel [a], Dr. Yvonne Traa [a], Prof. Dr.-Ing. Elias Klemm [a]
Prof. Dr. Cosima Stubenrauch
,
*[b]
[a] Dipl.-Chem. T. Montsch, M. Sc. M. Heuchel, PD Dr. Y. Traa, Prof. Dr.-Ing. E. Klemm
Institute of Chemical Technology, University of Stuttgart, Pfaffenwaldring 55, 70569 Germany
E-mail: yvonne.traa@itc.uni-stuttgart.de
[b] Prof. Dr. C. Stubenrauch
Institute of Physical Chemistry, University of Stuttgart, Pfaffenwaldring 55, 70569 Germany
E-mail: cosima.stubenrauch@ipc.uni-stuttgart.de
Graphical abstract
Highlights

Palladium nanoparticles of 4 to 8 nm size with a narrow particle size distribution were
synthesized with a water-in-oil microemulsion technique in an accessible temperature
range.


Mesoporous FDU-12 silica with a three-dimensional pore network was synthesized in the
presence of the nanoparticles, but nanoparticles were also deposited onto the support.
The catalytic activity of the materials was studied in the hydrogenation of 3-hexyn-1-ol
with molecular hydrogen and compared to reference catalysts such as the Lindlar and
TM
NanoSelect catalyst.

Some catalysts showed high activities and selectivities towards the cis-alkenol due to
sintering of the palladium particles which leads to the removal of most of the unselective
edge atoms.
1

Abstract
In the study at hand we present a design strategy for novel catalysts which can be used for the
selective hydrogenation of alkynes to alkenes. The design of the novel catalysts is based on
two main ideas, namely (1) the synthesis of Pd nanoparticles via microemulsions and (2) the
use of highly-ordered mesoporous silica with a 3-D pore network (FDU-12) serving as
support. The nanoparticles are deposited on FDU-12 in two different ways. Firstly, we simply
impregnated the support with a dispersion of the nanoparticles. The resulting catalyst was not
selective at all; on the contrary, it fully hydrogenated our model alkyne, namely 3-hexyn-1-ol.
Secondly, we synthesized the FDU-12 in the presence of the nanoparticles (in-situ synthesis).
In this case, we obtained one catalyst which performed as well as the Lindlar catalyst
although the metal content was slightly lower and our catalyst contained no Pb. Another
catalyst of the same series, prepared in the presence of another stabilizer, performed as well as
the NanoSelectTM catalyst but at a 7 times higher metal content. For the sake of comparison
we also impregnated FDU-12 via classical incipient wetness impregnation and again obtained
a completely nonselective catalyst. Our results demonstrate that the in-situ synthesis has great
potential as regards the development of novel catalysts.
Keywords: Selective Hydrogenation, Nanoparticles, Mesoporous FDU-12, Microemulsions
1
. Introduction
Our modern world is hard to imagine without catalytic processes. For example, over 80% of
all chemical and pharmaceutical production processes depend on catalytic technologies.[1]
Nevertheless, we are far away from thoroughly understanding, let alone designing new
catalysts for specific applications. One of these applications, which will be studied in the
paper at hand, is the selective hydrogenation of alkynes to alkenes. This hydrogenation plays
[
2,3]
a pivotal role for the industrial synthesis of vitamin A and -carotene.
As a model system
we chose the selective hydrogenation of 3-hexyn-1-ol to 3-Z-hexen-1-ol (see Figure 1).
The most prominent solid catalysts, which are used for this reaction, are the Lindlar catalyst
and the NanoSelectTM catalyst sold by BASF. Besides the NanoSelectTM catalyst, there are
[
4]
[5], Pd/SiO
also many other approaches for lead-free Lindlar catalysts , e.g., CeO
2
/TiO
2
2
[
6]
modified with an ionic-liquid layer , isolated Pd atoms in the cavities of mesoporous
polymeric graphitic carbon nitride [7], Pd/C with boron incorporated into the Pd subsurface[8],
[
9]
Pd on a multimetallic support .
2

The selectivity and activity of solid catalysts depend on a number of parameters: the type of
metal, the size and size distribution of the metal crystals (particles), the dispersion and the
total amount of the metal, the character of the catalytic sites, promoters, the type and structure
of the support, process parameters etc. In a nutshell, catalytic processes are very complex and
a quantitative comparison of two different catalysts is very often difficult or even impossible.
Nevertheless, novel catalysts are always benchmarked against existing ones, which is why we
had a closer look at the structure, activity and selectivity of the Lindlar catalyst and the
NanoSelectTM catalyst. For the sake of comparison, we also examined a standard Pd/C
catalyst (Pd on a microporous carbon support) which is unselective for the hydrogenation of
alkynes. We developed a novel catalyst with a view to reaching better activities than those the
commercially available ones can provide. The design of the novel catalysts is based on two
main ideas, namely (1) the synthesis of Pd nanoparticles via microemulsions and (2) the use
of highly-ordered silica with a 3-D pore network (FDU-12) serving as support.
The use of microemulsions for the synthesis of nanoparticles was developed by Boutonnet et
al.[10] This method has the advantages (a) of leading to particles in the nanometer range with a
narrow size distribution, (b) of synthesizing the particles at ambient conditions, i.e. under
energy-efficient conditions, (c) of adjusting the size of the microemulsion droplets and hence
of controlling the particle size[11,12,13], as well as (d) of having easy access to bimetallic
compounds (not addressed in the study at hand).[11,12] The catalytic nanoparticles produced
from microemulsions are metals, alloys, metal oxides and metals coated by metal oxides.[14,15]
The long-term aim is to use these nanoparticles in two different areas, namely in fuel cell
applications and in heterogeneous catalysis. Supporting these nanoparticles with an
appropriate carrier leads to catalysts which have proved to be superior to conventionally
prepared catalysts, i.e. catalysts prepared via impregnation.[14,15,16] The reason for the superior
performance is still unclear. To answer this open question we need more experimental data
for microemulsion-based catalysts and we need to compare these data quantitatively with
those obtained for conventionally prepared catalysts.
The synthesis of the nanoparticles is only one of many steps that need to be taken. The
nanoparticles need to be deposited on a supporting material if one wants to use them in
catalysis, hence finding a proper supporting material is another important issue. As already
mentioned, the type and structure of the support as well as its metal loading are known to
have a huge impact on the activity and selectivity of the catalyst. In the Lindlar catalyst, a
non-porous CaCO
3
support is used, while the support of the NanoSelectTM is either
3

mesoporous titanium silicate or carbon. We decided to use FDU-12 as support since it is a
mesoporous silica material with a 3-D pore system containing large cavities and pore sizes
which are large enough for our nanoparticles to diffuse into. Moreover, the 3-D pore system
allows for good diffusional characteristics in catalysis. In addition, we should be able to avoid
sintering of the metal particles by entrapping them in the cavities. Sintering of the metal
particles happens with all commercially available catalysts (see Section 3.2), and entrapment
of metal nanoparticles has been shown to prevent sintering.[17]
In the study at hand we first describe how we synthesized Pd nanoparticles of different sizes
via microemulsions. These nanoparticles were then deposited on FDU-12 in two different
ways. Firstly, we simply impregnated the support with a dispersion of the nanoparticles.
Secondly, we synthesized the FDU-12 in the presence of the nanoparticles (in-situ synthesis).
For the sake of comparison we also impregnated FDU-12 via classical incipient wetness
impregnation. Finally, the catalytic performance of our novel catalysts is benchmarked
against the Lindlar and the NanoSelectTM catalysts on the one hand, and against the
unselective Pd/C catalyst on the other.
2
2
. Experimental Section
.1. Materials
The chemicals used in this work are listed in Table 1 with their quality and source of origin.
2
2
.2. Methods
.2.1. Chemical Analysis
Chemical Analysis was performed on a Vista-MPX CCD Simultaneous ICP-OES
spectrometer manufactured by Varian. CHN analysis was carried out on a combustion
analyzer Vario El by burning the samples at 1223 K in air and subsequently analyzing the flue
gas with a thermal conductivity detector (TCD).
2
.2.2. N
The pore structure was determined by means of N
Quantachrome. Samples were degassed at 130 °C under vacuum prior to ad- and desorption at
196 °C.
2
Physisorption
2
physisorption on an Autosorb 3-B by
-
4

2
.2.3. XRD
Crystallographic phases were analyzed with a Bruker D8 Advance Powder X-Ray
Diffractometer with CuK alpha radiation.
2
.2.4. TEM
TEM images were taken with a Philips CM 200 and a FEI Tecnai G2. The particles were
suspended in ethanol with an ultrasonic bath before they were deposited on the sample holder.
Remaining ethanol was evaporated at room temperature. For the palladium nanoparticles, the
average Feret diameter was determined using SigmaScanPro 5.0. The histograms are given in
the Supporting Information. For the catalysts, only rough estimates of the metal particle size
can be given, since averaging over an appropriate amount of particles was not possible.
2
.2.5. Thermogravimetry
The weight loss of PEA, CAB and sodium nitrate during the nitrogen treatment at 550 °C for
6
h was determined using a Setaram Thermogravimetric Analyzer Setsys TG•16/18.
2
.3. Synthesis of the Nanoparticles
The nanoparticles were prepared following the procedure developed by Magno et al.[13]. First,
an aqueous solution of 20 mM Pd(NO was prepared by solubilizing the metal salt in water.
This solution is then used as the aqueous phase in the w/o-microemulsion (H O/metal salt – n-
octane – Brij 30/1-octanol). The composition of the microemulsion was always kept constant
except of the water content: the total surfactant concentration in the oil phase was 
)
3 2
2
B
= (mBrij30
+
m
1-octanol) / (mBrij30 + m1-octanol + mn-octane) = 0.114 and the mass fraction of 1-octanol in the
mixture of Brij30 and 1-octanol was  = m1-octanol / (mBrij30 + m1-octanol) = 0.39. The water
= mwater / mtotal = 0.02 to 0.08. Prior to any nanoparticle synthesis,
content was varied from w
A
the microemulsion was kept at a temperature just below the water emulsification failure
boundary under stirring for about 15 minutes. The nanoparticles (NPs) were then prepared by
adding solid NaBH
4
to the microemulsion containing the metal salt. After having added
NaBH , the sample was kept at constant temperature under stirring for another 15 min. The
4
addition of ethanol resulted in a precipitation of the Pd NPs which were then separated by
centrifugation, washed with ethanol and water for three times and then stored in ethanol. The
absolute masses used for a typical synthesis of the NPs are given in Table 2.
5

2
.4.Synthesis of FDU-12
In a polypropylene flask, 4.5 g Pluronic F-127 and 22.5 g KCl were dissolved in 234 mL
O, 54 mL HCl and 5.4 g Mes. The mixture was stirred for 24 h at 15 °C. Afterwards,
8.75 g TEOS were added, and the mixture was again stirred for 24 h. Finally, the mixture
was transferred in three Teflon-lined autoclaves and hydrothermally treated at 130 °C for
2 h. The gained material was filtered, washed with H O and dried at room temperature over
H
2
1
7
2
-1
night. Calcination was performed at 550 °C for 6 h with a heating rate of 1 °C min under
synthetic air.
2
.5. Synthesis of Pd/FDU-12(0.6)_NP_0.02
The catalyst Pd/FDU-12(0.6)_NP_0.02 was prepared by deposition of pre-synthesized Pd NPs
on the silica support. 15 mg of Pd NPs were suspended with 2 g dry FDU-12 in 100 mL
ethanol for 1 h in an ultrasonic bath. Afterwards, ethanol was evaporated at 40 °C under
reduced pressure in a rotary evaporator, and the material was dried at 100 °C for 5 h.
2
.6. In-Situ Synthesis of Catalyst
The in-situ synthesis approach consists of the incorporation of pre-synthesized Pd NPs during
the synthesis of the FDU-12 silica material. 30 mg of NPs were synthesized as described
above with w
A
= 0.02. After precipitation of the NPs with an eightfold molar excess of
NaBH , a tip of a spatula of PolyC was added as stabilizer. The NPs were added to the
4
reaction mixture prepared for the synthesis of FDU-12 prior to the addition of TEOS. For the
synthesis of the catalysts Pd/FDU-12(4.5)_CAB and Pd/FDU-12(3.7)_PEA, 19 mg CAB and
5
0 µL PEA, respectively, were added together with the NPs to the initial reaction mixture.
Note that the composition of the reaction mixture was the same as described in Section 2.4,
while the amount of all compounds was reduced by 1/9. The removal of the template was
carried out in a stainless steel tube under dry and oxygen-free nitrogen at 550 °C for 6 h to
avoid oxidation of the Pd NPs.
2
.7. Synthesis of Pd/FDU-12(4,6)_Imp
A solution of 0.0635 g tetraaminepalladium(II)chloride in 0.45 mL H
2
O was added dropwise
to 0.5512 g of pre-synthesized FDU-12 under constant slewing. The impregnated material
was dried overnight at 80 °C. Afterwards, it was pressed, crushed and sieved. The fraction of
-1
2
2
00 to 315 µm was reduced in a fixed-bed reactor with hydrogen (50 mL min ) at 350 °C for
h. After cooling, the obtained material was crushed again to a fine powder.
6

2
.8. Hydrogenation Experiments
The reaction was performed in a stainless-steel autoclave manufactured by Parr Instruments
with a glass inset. 10 mg of catalyst (except for the experiment with Pd/FDU-
1
2(0.6)_NP_0.02, where 89 mg of catalyst were used to compensate for the lower palladium
amount) were dispersed in 1 mL 3-hexyn-1-ol, 1mL 1-octanol (internal standard) and 20 mL
EtOH. The atmosphere above the reaction mixture was purged three times with nitrogen prior
to the reaction to remove air. About 2.2 MPa hydrogen were provided in a gas burette. The
reaction was started with constant 0.3 MPa hydrogen in the autoclave, which was set to a
-1
temperature of 35 °C and stirred at 800 min . Pressure drop and temperature in the burette
were recorded during the reaction, and the molar amount of hydrogen consumed in the
reaction was calculated. In order to remove traces of palladium, the reaction chamber was
cleaned with nitric acid at 60 °C for at least 2 h after the reaction. Product distribution was
determined by means of gas chromatography. The GC was equipped with a DB-WAX column
(30 m), a flame ionization detector and helium as carrier gas.
3
3
. Results and Discussion
.1. Synthesis of Nanoparticles via Microemulsions
Microemulsions are thermodynamically stable mixtures containing at least water, oil, and a
surfactant. Very often a co-surfactant is added. For certain compositions and temperatures
different microstructures can be found, e.g. droplets, elongated and bicontinuous structures, as
well as networks.[18] The structure that is relevant for the study at hand is the water-in-oil
(w/o) microemulsion which consists of water droplets in a continuous oil phase (Figure 2,
left). An indirect way of determining the microstructure is to study the phase behavior of a
microemulsion because 30 years of phase and microstructure studies have enabled us to
assign structures to certain regions of the phase diagram. The important phase diagram for
this work is the so-called “funnel” diagram (Figure 2, middle). In the absence of water, the
mixture of oil and surfactant (plus co-surfactant) forms one single phase. Adding water, one
observes the solubilisation of water in the system up to a certain amount above which excess
phases form. At low temperatures an oil-in-water microemulsion coexists with an excess oil
phase (2), while at higher temperatures a water-in-oil microemulsion coexists with an excess
water phase (2). As one can see, the two 2-phase regions are separated by a 1-phase funnel (1)
extending from the oil-rich side to higher water content up until the three-phase region (3)
forms. The phase boundary at low temperatures is the so-called near critical boundary (ncb),
while the one at higher temperatures is the so-called water emulsification failure boundary
7

(
wefb). Along the wefb discrete, spherical water droplets are formed which grow in size with
increasing water content w
A
.[19] Their sizes can be varied between 10 and 40 nm.[13]
The best strategy for the synthesis of nanoparticles via w/o microemulsions is to combine two
microemulsions which have the same structure but different aqueous solutions. One of the
microemulsions contains the metal salt and the other one contains the reducing agent in the
aqueous phase (Figure 2, left). The synthesis of the nanoparticles must take place along the
water emulsification failure boundary to ensure spherical water droplets of a given size. Once
the two microemulsions are combined, the droplet content is exchanged and nanoparticles are
formed (Figure 2, right). A much cheaper procedure – which is used by many groups – is to
reduce the metal salt in the w/o microemulsion by adding solid NaBH
our previous studies that the resulting nanoparticles are a bit larger than those prepared with a
NaBH
-containing microemulsion, while the monodispersity does not suffer significantly.[11]
We thus decided to use the second procedure for the synthesis of our nanoparticles.
4
. We showed in one of
4
For the preparation of our Pd nanoparticles we chose the quaternary system H O – n-octane –
2
Brij30 – 1-octanol because we had used this system in our previous studies for the synthesis
of Pt NPs and bimetallic NPs. The aqueous phase contained 20 mM of palladium nitrate. The
total surfactant concentration was  = 0.114 and the mass fraction of 1-octanol in the mixture
B
of Brij30 and 1-octanol was  = 0.39. These parameters were chosen because we knew from
our previous work that the water emulsification failure boundary lies in a temperature range
that is easily accessible.[20] The resulting wefb is illustrated in Figure 3. For the preparation of
the NPs the temperature of the sample was adjusted such that it was slightly below the wefb to
ensure that we have one phase and that the droplets are spherical. Subsequently, solid NaBH
4
was added. Nanoparticles were prepared at four different water contents, namely w = 0.02,
A
0
1
.04, 0.06, and 0.08 leading to nanoparticles of 4.3 ± 0.8, 4.6 ± 0.4, 4.7 ± 0.6, and 8.2 ±
.4 nm size as illustrated in Figure 4. (TEM images show the nanoparticles in the
agglomerated state, which is the thermodynamically stable form under the dry vacuum
conditions in a TEM.) These numbers reveal that the NPs resulting from the first three
samples differed only slightly in size, while those obtained at w
Although the overall trend is as expected – larger particles for larger water contents – there is
no clear relation between w and NP size, which is most likely due to the rather uncontrolled
reduction when adding solid NaBH instead of a second NaBH -containing microemulsion.
A
= 0.08 were a bit larger.
A
4
4
Nevertheless, we synthesized nanoparticles in the range of 4 – 8 nm with a very good
monodispersity. In order to work at low metal content we decided to take the nanoparticles
8

resulting from the sample with w = 0.02 for the synthesis of the catalysts, which is described
A
in Section 3.2.
3
.2. Design Strategy of Novel Catalysts
Before describing how we synthesized our catalysts and what their structures look like, we
will outline the main features of three commercially available catalysts, namely the Lindlar
catalyst, the NanoSelectTM and the Pd/C catalyst, since we need these details for a discussion
of our results.
Lindlar Catalyst: In the Lindlar catalyst, the traditional reference catalyst for selective
hydrogenation of alkynes to alkenes, typically 5 wt.% of Pd are deposited on a non-porous
CaCO support with a low dispersion of under 10%. Two main types of palladium particles
3
can be distinguished, namely flat platelets of several nm which are attached flat on the support
surface as well as microdomains of elemental Pd forming three-dimensional aggregates. The
modifier Pb seems to block sites which are responsible for unselective, fast
hydrogenation.[21,22] Additional modifiers are organic bases such as quinoline or piperidine
which also enhance the alkene selectivity.[23] The Lindlar catalyst is used world-wide due to
its high selectivity in semi-hydrogenation of alkynes despite the high metal content and the
need for Pb. It is especially this need for Pb that is controversial when it comes to catalyst
recycling.
TM
TM
NanoSelect : The BASF NanoSelect catalyst was invented as a Pb-free replacement for
the Lindlar catalysts. NanoSelectTM either consists of 0.5 wt.% colloidal Pd on mesoporous
titanium silicate, TiSi
Commercial hexadecyl(2-hydroxyethyl)dimethyl
HHDMA) is used as a water-soluble stabilizer/reducing agent for the preparation of the metal
O
2 6
(LF200) or of 0.6 wt.% colloidal Pd on carbon (LF100).
ammonium dihydrogen-phosphate
(
nanoparticles and remains on the Pd particles as surface modifier. The pores of the support are
too small (~ 3 nm) to allow for the diffusion of the colloids (~ 4-8 nm) into the pores, which
leads to an eggshell distribution of the Pd. Both effects permit high selectivities.[24] Indeed,
NanoSelectTM is as selective as the Lindlar catalyst but has a slightly higher activity at a much
TM
lower metal content and without the need for Pb! NanoSelect , however, has two
drawbacks: (1) At high metal loading, Pd leaching was observed, which is due to the weak
bonding of the metal agglomerates.[25] (2) The long-term stability of the metal-modifier
system is not clear. On the one hand, HHDMA could only partially be removed even after
prolonged washing with organic solvents or hot water. Removal of the stabilizer by
9

calcination led to sintering of the Pd particles (from 15 to 15-50 nm after 1 h at 700°C), which
ultimately results in a lower activity but unchanged selectivity. Thus, the stabilizer seems to
be irrelevant for obtaining a selective catalyst unless an undetectable amount of residual
stabilizer can cause the high selectivity in the semi-hydrogenation.[24] However, recent studies
indicate that removal of HHDMA by UV-ozone treatment leads to increased conversion but
decreased selectivity.[22]
Pd/C: Palladium on carbon is the most common hydrogenation catalyst in laboratories and
industry.[26] Generally, most Pd particles are smaller than 10 nm, but usually some larger Pd
clusters are also present and the Pd dispersion is rather high.[25,27,28] Palladium on carbon is
not selective for the semi-hydrogenation of alkynes to alkenes.
Aiming to improve the performance of the existing catalysts, we designed a novel catalyst.
Our design strategy was based on the fact that nanoparticles synthesized via microemulsions
often have better catalytical properties than NPs synthesized via other methods.[14,15] Reasons
for this phenomenon have not been found to date. However, taking advantage of this
phenomenon, we first synthesized nanoparticles with a narrow size distribution via
microemulsions before we deposited them on the silica support FDU-12, which has a 3-D
pore network. FDU-12 consists of a face-centered cubic packing of spherical cavities, each of
which is connected to 12 nearest neighboring cavities. Unit cell sizes, diameters of the
cavities, and entrance size to the cavities can be varied[29] (Figure 5, top left). The TEM
picture in Figure 5 (top right) shows that the spherical cavities and the ordered domains
extend over a large area. The nitrogen adsorption isotherm (Figure 5, bottom) has a broad
hysteresis loop. The DFT calculation for porous materials with spherical cavities reveals a
cavity diameter of 20.6 nm. Unfortunately, determining the pore entrance sizes is very
difficult. However, as the desorption branch of the isotherm shows a decline over a broad
pressure range, the distribution of the pore entrance sizes is probably quite broad. Since the
capillary evaporation steps are centered at relative pressures of 0.6 to 0.8, the largest entrance
sizes should have an average diameter of 7 to 14 nm.[30] In any case, the pore entrance size
must be above 5 nm to allow the nanoparticles with w = 0.02 and a size of 4.3 ± 0.8 nm to
A
diffuse into the spherical cavities. The deposition of the NPs onto the support was carried out
via two different routes (Figure 6). For the sake of comparison, we also impregnated FDU-12
via classical incipient wetness impregnation.
1
0

3
.3. Impregnation of Support with Pd Nanoparticles
The nanoparticles were synthesized via microemulsions (see Section 2.3) and stored in
ethanol. The support for the nanoparticles, i.e. FDU-12, was synthesized independently.
Afterwards, the support was impregnated with a suspension of the Pd nanoparticles in
ethanol, which was then evaporated (Figure 6a). In the sample name, X.X denotes the
palladium content in wt.-% related to the dry catalyst, and in NP_(0.0Y), 0.0Y equals the
water content of the microemulsion used for the synthesis of the nanoparticles. Figure 7a
shows a TEM picture of the catalyst produced in this way. Obviously, the nanoparticles
mainly agglomerate around the support without really diffusing into the support pore system.
Only a small part of the nanoparticles has access to the pore system, which can hardly be
visualized by TEM.
3
.4. In-situ Synthesis
The nanoparticles were synthesized via microemulsions and stabilized with Polymer C. Then
the nanoparticles were added to the reaction mixture prepared for the synthesis of FDU-12
together with either CAB (cetyltrimethylammonium bromide) or PEA (1-phenylethan-1-
amine) as stabilizer (Figure 6b). In the sample name, in addition to the Pd content and the
water content of the microemulsion, the respective abbreviation of the stabilizer, i.e. CAB or
PEA, is given. The template was removed at 550 °C in a nitrogen atmosphere. Figure 7b and
7
c show TEM pictures of the two catalysts produced by in situ synthesis. Obviously, also here
the nanoparticles do not get access to the large cavities here either. Instead, mainly large
aggregates of about 100 to 200 nm size as well as some smaller clusters are formed, which are
also located at the outer surface of the support. Note that the removal of the template is a
sintering process which is known to lead to the formation of NP aggregates. The diameter of
the spherical cavities of FDU-12 remains virtually unchanged by the presence of the
nanoparticles and the stabilizer. At this point, it is important to point out that agglomerates
such as those shown in Figure 7a are only a collection of weakly bound particles, whereas
aggregates illustrated in Figures 7b and 7c consist of strongly bound and fused particles.[31]
3
.5. Impregnation of Support with Pd Salt
Classical incipient wetness impregnation of the support FDU-12 was carried out by adding
dropwise a palladium salt solution under constant slewing (Figure 6c). The catalyst was
reduced with hydrogen in a fixed-bed reactor. “Imp” in the sample name indicates the
classical impregnation method. A TEM picture of this material (Figure 7d) shows that the
1
1

classical impregnation leads to rather well dispersed Pd particles with sizes between 2 and 15
nm, which seem to be mainly located in the spherical cavities of the support.
Table 3 summarizes the properties of all the different Pd-containing catalysts. The
NanoSelectTM LF200 and the Pd/FDU-12(0.6)_NP_0.02 catalysts have metal contents of
around 0.5 wt.-%, while the Pd content of all other catalysts is between 3.7 and 5 wt.-%. The
different metal contents enable comparisons of our own catalysts with the NanoSelectTM and
the Lindlar catalyst. The supports are non-porous, microporous or mesoporous. The sizes of
the palladium particles range from a few nanometers to large aggregates up to 200 nm.
3
.6. Catalytic Performance of Novel Catalysts
Figure 8 shows the catalytic performance of both the novel catalysts and the commercial
catalysts. It illustrates the molar hydrogen consumption as a function of the reaction time. The
hydrogen consumption for the complete hydrogenation of 3-hexyn-1-ol to 1-hexanol amounts
to 0.018 mol, while the hydrogenation to the hexenols needs only half the amount of
hydrogen, i.e. 0.009 mol.
Let us first discuss the performance of the nonselective catalysts Pd/C, Pd/FDU-12(4.6)_Imp,
and Pd/FDU-12(0.6)_NP_0.02, which completely hydrogenated the reactant to 1-hexanol.
After a reaction time of 30 min and a hydrogen consumption of 0.0085 mol, there is a bend in
the hydrogen consumption curve of the catalysts Pd/FDU-12(4.6)_Imp and Pd/FDU-
1
2(0.6)_NP_0.02. Afterwards, the reaction proceeds more slowly. Before the bend is reached,
the selectivity S3-Z-hexen-1-ol is 94% for Pd/FDU-12(4.6)_Imp and 97% for Pd/FDU-
2(0.6)_NP_0.02. Shortly after the bend the selectivities drop to 71% and 57%, respectively,
1
because the double bond is further hydrogenated to yield 1-hexanol. The alkyne is
preferentially reduced because of its higher heat of adsorption, which enables it to displace the
alkene from the catalyst surface.[32,33] Once all the alkyne molecules have been reduced to the
alkene, the alkene can be further reduced to the alkane. The rates of alkyne reduction and
subsequent alkene reduction are usually within the same order of magnitude.[34]
By contrast, the hydrogen consumption on Pd/C is linear up until the end of the reaction after
about 60 min, when the hydrogen consumption is 0.018 mol. This is typical of monometallic
Pd, which needs modification in order to suppress the formation of hydrides and carbides,
thereby stabilizing a high alkene yield.[22] In general, modifiers can (a) increase the difference
in the adsorption strength of the alkyne versus the alkene, (b) directly compete with the alkene
1
2

for reaction sites and/or (c) decrease the rate of the surface reaction of the adsorbed alkene,
which can explain why the hydrogenation rate slows down once all the alkyne has been
consumed.[33] The nitrogen content obtained from the CHN analysis presented in Table 3 is an
indication for the modifier content of the catalysts, since HHDMA, PEA and CAB contain
nitrogen. However, the synthesis of Pd/C does not usually involve a modifier.[35] Hence, the
high nitrogen content of Pd/C is most probably due to the nitrogen content of the source
material for the activated carbon[36], but it is no indication for a modification of the Pd
surface.
For the catalysts Pd/FDU-12(4.6)_Imp and Pd/FDU-12(0.6)_NP_0.02 we did not use any
modifier either. However, it has been shown that despite extensive cleaning of the
nanoparticles it is common to find traces of the stabilizing and capping agents, which modify
the catalytic behavior of the metal.[4] Hence, the structure of the palladium surface created
during the surfactant-mediated nucleation and growth should be different from the one of
palladium in surfactant-free syntheses. The preparation of Pd/FDU-12(4.6)_Imp involved the
use of tetraaminepalladium(II)chloride as a Pd precursor. Since the catalyst was only reduced
and not oxidized, nitrogen from the Pd precursor can remain on the sample, which explains
the relatively high nitrogen content of 0.1 wt.-% (Table 3). Moreover, Pd(NO
3
2
) was used for
the production of the nanoparticles, and we assume that after reduction of the palladium,
sodium nitrate is formed. Traces of sodium nitrate which are not removed by washing
presumably lead to the observed nitrogen content of Pd/FDU-12(0.6)_NP_0.02 of 0.03 wt.-%
(Table 3). In summary, both samples contain traces of modifiers, which can explain the bend
in the hydrogen consumption curves (Figure 8). However, the amount of modifier is not
sufficient to achieve selectivities similar to the Lindlar or the NanoSelectTM catalysts. In
addition, since concentration profiles in favor of a higher concentration of alkene at the Pd
sites in porous catalysts reduce the selectivity[33], the selectivity of the Pd/FDU-12(4.6)_Imp
catalyst, where the Pd seems to be located in the pores, is limited.
Having discussed the nonselective catalysts, we will now focus on the selective ones. As
expected, the hydrogen consumption is only 0.009 mol for all of them. With the
NanoSelectTM catalyst, the reaction starts to slow down after 80 min; after 159 min, an alkyne
conversion of 100% and a selectivity S3-Z-hexen-1-ol of 88% are reached. The catalyst Pd/FDU-
1
2(3.7)_PEA_0.02 shows a nearly identical performance, only the selectivity to 3-Z-hexen-1-
ol is slightly higher (Table 4). The reaction on the Lindlar catalyst is not fully finished after
74 min with a conversion of 95%, but at a very high selectivity S3-Z-hexen-1-ol of 98% (Table
1
1
3

4
). Very similar results are obtained on Pd/FDU-12(4.5)_CAB_0.02 (Table 4). Table 4 also
shows the slope of the hydrogen consumption per reaction time and per Pd mass which is a
measure for the hydrogenation rate. The much higher “hydrogenation rate” on the
NanoSelectTM catalyst as compared to these two catalysts is very remarkable, since we used
the same amount of catalyst but the Pd content of the NanoSelectTM catalyst is only about
1
/10 of the Pd content of the Lindlar catalyst. It has been stated that on a quinoline-modified
Lindlar catalyst, less than 10% of the surface palladium sites are active for the hydrogenation
reaction.[37] On the other hand, it has been described that the action of the ligand in the
NanoSelectTM catalyst also effectively reduces the number of active sites and the adsorption
energy of H in the first layer in the same way as Pd alloying does in the Lindlar catalyst, even
though at high hydrogen pressures overhydrogenation cannot be prevented on the
NanoSelectTM catalyst.[22] The ligand blocks about 75% of the metal surface of the
NanoSelectTM catalyst, thus leading to ensembles with a size of approximately 0.8 nm,
whereas in the Lindlar catalyst Pb isolates the Pd sites, thus leading to ensembles with a size
of approximately 1 nm.[7,22] Therefore, the difference in reactivity between the NanoSelectTM
and the Lindlar catalyst must be due to the higher activity of the specially structured Pd
particles in the NanoSelectTM catalyst as compared to the Pd in the Lindlar catalyst, which is
strongly poisoned by Pb. This is even more remarkable since the comparatively high nitrogen
content of the NanoSelectTM catalyst of 0.49 wt.-% (Table 3) indicates considerable amounts
of modifier (corresponding to around 17 wt.-% HDDMA[38]), the presence of which imposes
severe access and/or adsorption constraints since the hexadecyl chains are arranged
perpendicularly to the palladium surface. Therefore, alkynes with an adsorption length greater
than 0.8 nm are not hydrogenated by the three-dimensional NanoSelectTM catalyst with its
hydrophobic “pores”, whereas the two-dimensional Lindlar catalyst provides an optimal
conversion.[22]
With regard to the catalysts Pd/FDU-12(4.5)_CAB_0.02 and Pd/FDU-12(3.7)_PEA_0.02, it is
noticeable that the selectivity is significantly increased as compared to Pd/FDU-
1
2(0.6)_NP_0.02, even though all three catalysts were made using the same nanoparticles
produced via microemulsions. This can be explained by the fact that on palladium
nanocrystals stabilized with polyvinylpyrrolidone overhydrogenation of an alkynol occurred
solely on the edge atoms, which is presumably due to increased adsorption strength of the
alkene, whereas the edge atoms are much less active in semi-hydrogenation as compared to
the plane sites.[39] Since our nanoparticles made via microemulsions are multicrystalline and
dispose of many edge atoms, the low selectivity of Pd/FDU-12(0.6)_NP_0.02 can be
1
4

explained quite easily. As Figure 7a clearly shows, on Pd/FDU-12(0.6)_NP_0.02
agglomerates of palladium nanoparticles are formed which still have many unselective edge
atoms. For the catalysts Pd/FDU-12(4.5)_CAB_0.02 and Pd/FDU-12(3.7)_PEA_0.02,
removal of the stabilizer and the surfactant leads to sintering of the nanoparticles to large
aggregates, which significantly reduces the number of edge atoms. Since very large Pd
aggregates are formed on the catalysts Pd/FDU-12(4.5)_CAB_0.02 and Pd/FDU-
1
2(3.7)_PEA_0.02, the palladium aggregates are obviously not located in the pore system of
the mesoporous support and are thus freely accessible to the reactants. We assume that the
residual nitrogen contents of 0.08 and 0.07 wt.-%, respectively, (Table 3) are again due to the
sodium nitrate being not entirely removed (from production of nanoparticles with Pd(NO
since the modifiers CAB and PEA were completely removed by treatment with nitrogen at
50 °C, as we were able to prove with thermogravimetric measurements: Whereas CAB and
3
)
2
)
5
PEA decompose completely at temperatures as low as just under 300 °C, sodium nitrate only
slowly starts to decompose at 550 °C. In conclusion, both catalysts are highly selective
without the need for lead as modifier, but at higher Pd contents than the NanoSelectTM
catalyst. The faster reduction rate on Pd/FDU-12(3.7)_PEA_0.02 as compared to Pd/FDU-
1
2(4.5)_CAB_0.02 can be explained by the slightly smaller aggregates on the catalyst
mentioned first (100-150 nm versus 150-200 nm).
4
. Conclusions & Outlook
In this study, we present novel catalysts for the selective hydrogenation of alkynes to alkenes.
The synthesis of the novel catalysts involves (1) the synthesis of Pd nanoparticles via
microemulsions and (2) the use of highly-ordered mesoporous silica with a 3-D pore network
(FDU-12) as support. The nanoparticles are deposited on FDU-12 in two different ways.
Firstly, we simply impregnated the support with a dispersion of the nanoparticles. The
resulting catalyst was not selective at all but fully hydrogenated our model alkyne, namely 3-
hexyn-1-ol. We argue that this is due to the fact that the palladium particles formed weakly
bound agglomerates with many unselective edge atoms. Secondly, we synthesized the FDU-
1
2 in the presence of the nanoparticles (in-situ synthesis). In this case, we obtained two
selective catalysts. One catalyst performed as well as the Lindlar catalyst although our
catalyst had a slightly lower metal content and contained no Pb. Another catalyst of the same
series, prepared in the presence of another stabilizer, performed as well as the NanoSelectTM
catalyst, but at a 7 times higher metal content. However, our catalysts are free of stabilizers
(amines), which could significantly improve the long-term performance of our catalysts
because no stabilizer leaching can take place as opposed to the NanoSelectTM catalyst with its
1
5

high stabilizer content. We think that the high selectivity of our in situ-synthesized catalysts is
due to the formation of large palladium aggregates during the removal of the template at high
temperatures. This sintering leads to the removal of most of the unselective edge atoms.
Let us conclude by suggesting some reasonable next steps. (1) Firstly, the NPs should be
TM
deposited on the support that is used for NanoSelect , e.g. TiSi
2
6
O . The aim is to choose
conditions which are as similar as possible to those used for the synthesis of the
TM
NanoSelect : if we deposited 0.5 wt.% of 4-8 nm Pd particles on mesoporous TiSi
2
6
O as
TM
support and if this catalyst performed much better (or worse) than NanoSelect , we would
have clear experimental evidence that the way the nanoparticles are synthesized does matter
after all! (2) The controlled sintering of nanoparticles should be investigated in more depth in
order to mask unselective edge or corner sites. (3) Finally, the reaction should be carried out
with more complex reactants which contain carbon-carbon double and triple bonds to see (a)
whether the selectivity for alkyne hydrogenation is maintained in the presence of a double
bond and (b) whether the accessibility of the selective sites is sufficient.
Acknowledgements
The authors thank Dorothea Häussermann for valuable practical contributions to this work.
1
6

References
1] http://www.suschem.org/documents/document/20120208153057-suschem_sra_final.pdf,
accessed: 8th of september 2016.
[
[2] G. L. Parker, L. K. Smith, I. R. Baxendale, Tetrahedron 2016, 72, 1645–1652.
[3] O. Isler, H. Lindlar, M. Montavon, R. Rüegg, P. Zeller, Helv. Chim. Acta 1956, 39, 249–
2
59.
[4] M. Crespo-Quesada, F. Cárdenas-Lizana, A.-L. Dessimoz, L. Kiwi-Minsker, ACS Catal.
2
012, 2, 1773-1786.
[5] G. Vilé, S. Wrabetz, L. Floryan, M. E. Schuster, F. Girgsdies, D. Teschner, J. Pérez-
Ramírez, ChemCatChem. 2014, 6, 1928-1934.
[6] F. Schwab, N. Weidler, M. Lucas, P. Claus, Chem. Commun. 2014, 50, 10406-10408.
[7] G. Vilé, D. Albani, M. Nachtegaal, Z. Chen, D. Dontsova, M. Antonietti, N. López, J.
Pérez-Ramírez, Angew. Chem. Ind. Ed. 2015, 54, 11265-11269.
[
8] C. W. A. Chan, A. H. Mahadi, M. M.-J. Li, E. C. Corbos, C. Tang, G. Jones, W. C. H.
Kuo, J. Cookson, C. M. Brown, P. T. Bishop, S. C. E Tsang, Nature Communications 2014,
787, 1-9.
6
[9] S. Vernuccio, R. Goy, P. R. von Rohr, J. Medlock, W. Bonrath, React. Chem. Eng. 2016,
1
, 445-453.
[10] M. Boutonnet, J. Kizling, P. Stenius, Colloid. Surface. 1982, 5, 209-225.
[11] L. M. Magno, W. Sigle, P. A. van Aken, D. G. Angelescu, C. Stubenrauch, Chem. Mat.
2
010, 22, 6263-6271.
[12] L. M. Magno, W. Sigle, P. A. van Aken, D. Angelescu, C. Stubenrauch, Phys. Chem.
Chem. Phys. 2011, 13, 9134-9136.
[13] L. M. Magno, D. Angelescu, W. Sigle, C. Stubenrauch, Phys. Chem. Chem. Phys. 2011,
1
3, 3048-3058.
[
14] S. Eriksson, U. Nylén, S. Rojas, M. Boutonnet, Appl. Catal. A 2004, 265, 207–219.
15] M. Boutonnet, S. Lögdberg, E. E. Svensson, Curr. Opin. Colloid Interface Sci. 2008, 13,
[
2
70–286.
[16] S. Järås, S. Lögdberg, M. Boutonnet: Method for depositing metal particles on a support.
European Patent Application Number EP07106863.9 2007.
[17] S. Li, A. Tuel, J.-L. Rousset, F. Morfin, M. Aouine, L. Burel, F. Meunier, D. Farrusseng,
ChemNanoMat 2016, 2, 534-539.
[18] T. Sottmann, C. Stubenrauch in Microemulsions (Eds.: C. Stubenrauch), Wiley and Sons,
Hoboken, NJ, 2009, pp. 1-83.
1
7

[19] T. Sottmann, R. Strey in Soft Colloids V – Fundamentals in Interface and Colloid
Science (Eds.: J. Lyklema), Elsevier, Amsterdam, 2005, pp. 5.1-5.95.
[20] M. Magno, D. G. Angelescu, C. Stubenrauch, Colloid. Surface A 2009, 348, 116–123.
[21] R. Schlögl, K. Noack, H. Zbinden, A. Reller, Helv. Chim. Acta 1987, 70, 627-679.
[22] G. Vilé, N. Almora-Barrios, S. Mitchell, N. López, J. Pérez-Ramírez, Chem. Eur. J.
2
014, 20, 5926-5937.
[23] P.N. Rylander, Catalytic Hydrogenation in Organic Synthesis, Academic Press, New
York, 1979.
[24] P. T. Witte, S. Boland, F. Kirby, R. van Maanen, B. F. Bleeker, D. A. Matthijs de
Winter, J. A. Post, J. W. Geus, P. H. Berben, ChemCatChem 2013, 5, 582-587.
[25] P. T. Witte, P. H. Berben, S. Boland, E. H. Boymans, D. Vogt, J. W. Geus, J. G.
Donkervoort, Top Catal. 2012, 55, 505–511.
[26] H.-U. Blaser, A. Indolese, A. Schnyder, H. Steiner, M. Studer, J. Molec. Catal. A: Chem.
2
001, 173, 3–18.
[27] A. Jung, A. Jess, T. Schubert, W. Schütz, Appl. Catal. A 2009, 362, 95-105.
[28] Y. A. Ryndin, M. V. Stenin, A. I. Boronin, V. I. Bukhtiyarov, V. I. Zaikovskii, Appl.
Catal. 1989, 54, 277-288.
[29] T. Yu, H. Zhang, X. Yan, Z. Chen, X. Zou, P. Oleynikov, D. Zhao, J. Phys. Chem. B
2
006, 110, 21467-21472.
[30] M. Kruk, C. M. Hui, Micropor. Mesopor. Mater. 2008, 114, 64-73.
[31] P. Albers, M. Maier, M. Reisinger, B. Hannebauer, R. Weinand, Cryst. Res. Technol.
2
015, 50, 1-20.
[32] J. Rajaram, A. P. S. Narula, H. P. S. Chawla, S. Dev, Tetrahedron 1983, 39, 2315-2322.
[33] H.-U. Blaser, A. Schnyder, H. Steiner, F. Rössler, P. Baumeister in Handbook of
Heterogeneous Catalysis, Vol. 7 (Eds.: G. Ertl, H. Knözinger, F. Schüth, J. Weitkamp)
WILEY-VCH, Weinheim, 2008, pp. 3284-3308.
[
[
[
34] A. Molnár, A. Sárkány, M. Varga, J. Molec. Catal. A 2001, 173, 185-221.
35] R. Mozingo, Org. Synth. 1946, 26, 77-79.
36] S. Hermans, C. Diverchy, O. Demoulin, V. Dubois, E. M. Gaigneaux, M. Devillers, J.
4
Catal. 2006, 243, 239-25 .
[37] M. García-Mota, J. Gómez-Díaz, G. Novell-Leruth, C. Vargas-Fuentes, L. Bellarosa, B.
Bridier, J. Pérez-Ramírez, N. López, Theor. Chem. Acc. 2011, 128, 663-673.
1
8

[38] D. Albani, G. Vilé, S. Mitchell, P. T. Witte, N. Almora-Barrios, R. Verel, N. López, J.
Pérez-Ramírez, Catal. Sci. Technol. 2016, 6, 1621-1631.
[39] M. Crespo-Quesada, A. Yarulin, M. Jin, Y. Xia, L. Kiwi-Minsker, J. Am. Chem. Soc.
2
011, 133, 12787-12794.
1
9

Figure 1: Hydrogenation of 3-hexyn-1-ol to the desired product 3-Z-hexen-1-ol and the
byproducts 3-E-hexen-1-ol, its isomers and hexan-1-ol.
w/o
oil
2
water
excess
1
) Pd(NO3)2
aq.2
T
3
1

2) NaBH4
aq.1
oil
excess
2
metallic
nanoparticles
o/w

~ 5 - 30 nm
wA
Figure 2: Mixing two water-in-oil microemulsions (w/o-µe) of same structure but different
composition (left), at the water emulsification failure boundary, wefb (middle), one obtains
metallic nanoparticles (right). The aqueous phase of one µe contains the metal salt (here:
aq.1 = Pd(NO
3
)
2
), while the other contains the reducing agent (here: aq.2 = NaBH ). The
4
reduction of the metal salt leads to nanoparticles the size of which can be controlled by the
size of the templating microemulsion droplets. See text for further details.
5
4
3
2
1
0
0
0
0
-
2
1
0
0
.00
0.04
0.08
0.12
wA
Figure 3: Water emulsification failure boundary (wefb) of the quaternary system H O –
2
n-octane – Brij30 – 1-octanol in the presence of 20 mM Pd(NO at  = 0.114 and  = 0.39.
3
)
2
B
2
0

Figure 4: TEM pictures of Pd nanoparticles synthesized at different w
A
, namely NP_0.02 (w
A
=
0.02), NP_0.04 (w = 0.04), NP_0.06 (w = 0.06), and NP_0.08 (w = 0.08).
A
A
A
Figure 5: (top, left) Schematic drawing of the porous structure of FDU-12 where the red
surfaces point toward the pores and the green surfaces toward the walls[29]; (top, right) TEM
picture of the support FDU-12; (bottom) nitrogen adsorption isotherm.
2
1

a. Impregnation of support with Pd nanoparticles

Pd/FDU-12(X.X)_NP_(0.0Y)
Synthesis of the
support
Catalyst
Synthesis of the
nanoparticles
b. In-situ synthesis

Pd/FDU-12(X.X)_stabilizer_(0.0Y)
Synthesis of the
nanoparticles
Synthesis of the
support
Catalyst
Catalyst
c. Impregnation of support with Pd salt

Pd/FDU-12(X.X)_Imp
Synthesis of the
support
Pd salt
Figure 6: Scheme of the three routes via which the different catalysts were synthesized.
Catalyst names: X.X = Palladium content in wt.-% related to the dry catalyst; NP_(0.0Y) =
water content of the microemulsion used for the synthesis of the nanoparticles; stabilizer =
PEA (1-phenylethan-1-amine) or CAB (cetyl trimethylammonium bromide); Imp = traditional
impregnation of FDU-12 with Pd salt.
Figure 7: TEM pictures of (a) Pd/FDU-12(0.6)_NP_0.02, (b) Pd/FDU-12(4.5)_CAB_0.02,
(c) Pd/FDU-12(3.7)_PEA_0.02, (d) Pd/FDU-12(4.6)_Imp.
2
2

Figure 8: Catalytic performance of the different catalysts in the selective hydrogenation of 3-
hexyn-1-ol compared to the commercial catalysts Lindlar, NanoSelectTM LF 200 and Pd/C
(
mcatalyst = 10 mg, except for Pd/FDU-12(0.6)_NP_0.02 where mcatalyst = 89 mg to compensate
for the lower Pd content).
2
3

Table 1: Chemicals used in this work, their purification grade and the respective
manufacturer. All materials were used without further purification.
Chemical
Abbreviation /
Purification Manufacturer
Molecular Formula
1
1
3
-Octanol
≥99%
99%
98%
n.a.
Sigma Aldrich
Sigma Aldrich
Sigma Aldrich
Sigma Aldrich
Aldrich
-phenylethan-1-amine
-Hexyn-1-ol
PEA
Brij30*
cetyl (= hexadecyl) trimethyl- CAB
ammonium bromide
n.a.
Ethanol
EtOH
HCl
Abs.
37%
Various
Hydrochloric Acid
Lindlar catalyst
Bernd Kraft
Sigma Aldrich
Acros
Mesitylene
Mes
99%
98%
NanoSelectTM LF200
Strem Chemicals
Sigma Aldrich
Heraeus
n-Octane
Palladium on activated carbon
Palladium(II) nitrate
Pluronic F-127
Pd/C
Pd(NO
3
)
2
99.99%
n.a.
chemPur
Sigma Aldrich
BASF
Polymer C
PolyC
KCl
n.a.
Potassium chloride
Sodium borohydride
Tetraaminepalladium(II)chloride
Tetraethyl orthosilicate
Water
99%
98%
n.a.
Riedel-deHaen
Sigma Aldrich
ChemPur
NaBH
(NH
TEOS
4
3
)
4
PdCl
2
98%
demin.
Sigma Aldrich
2
H O
*
The surfactant Brij 30, tetraethyleneglycol mono-dodecylether (C12
E~4), is a non-ionic
commercially available surfactant containing an average of four ethylene oxide groups in the
head group.
2
4

Table 2: Masses of used substrates in gram for the synthesis of 30 mg Pd NPs with  = 0.39
and  = 0.114 at different w
B
A
.
w
A
0.02
0.04
0.06
0.08
Pd(NO
3
)
2
· 2 H
2
O 0.0751 0.0751 0.0751 0.0751
14.067 14.067 14.067 14.067
48.188 23.603 15.407 11.310
613.961 300.716 196.300 144.093
2
H O
Brij 30
n-Octane
-Octanol
NaBH
1
30.810 15.090 9.851
7.231
4
0.0853 0.0853 0.0853 0.0853
Table 3: Properties of Pd-containing catalysts used for the selective hydrogenation of
-hexyn-1-ol.
3
metal
size of metal crystals / stabilizer Nitrogen
BET sur- support
/ face area
content particles
/ modifier content
wt.-%
2
-1
/ m g
Lindlar
5 wt.%
(1) platelets, several nm 2-3 wt.% 0.05
1
non-porous
CaCO
mesoporous
TiSi (pore
(2) microdomains
Pb
3
NanoSelectTM
LF200
0.5 wt.% particles ~ 4-8 nm
HHDMA
0.49*
0.36
231*
2
O
6
size ~ 3 nm)
microporous
carbon
Pd/C
5 wt.%
(1) mainly particles none
smaller than 10 nm
2) few bigger clusters
877
(
Pd/FDU-
4.6 wt.% particles ~ 2-15 nm
none
0.10
0.03
0.07
0.08
335
565
399
396
mesoporous
FDU-12
1
2(4.6)_Imp
Pd/FDU-
2(0.6)_NP_0.02
Pd/FDU-
2(3.7)_PEA_0.02
Pd/FDU-
2(4.5)_CAB_0.02
0.6 wt.% particles ~ 4.5 nm
none
mesoporous
FDU-12
1
3.7 wt.% aggregates 100-150 nm PEA
4.5 wt.% aggregates 150-200 nm CAB
mesoporous
FDU-12
1
mesoporous
FDU-12
1
*
see Ref. 22
2
5

Table 4: Performance of Pd-containing catalysts in the selective hydrogenation of 3-hexyn-1-
ol: hydrogen consumption related to the Pd content of the catalyst (the slope of the H
2
consumption after 20 min of reaction time was used for this purpose which represents for the
samples with a bend in the curve the hydrogenation of the alkyne to the alkene.) and
selectivities after reaction times  120 min (X is conversion, S is selectivity).
n
H2 t (mPd)-1
-
1
reaction
X
3-hexyn-1-ol
S
3-Z-hexen-1-ol
S
3-E-hexen-1-ol
S
isomers
S
hexan-1-ol
time
min
174
/
/ %
/ %
/ %
/ %
/ %
/ mmol min-1
mg-1
Lindlar
95.4
99.9
97.9
87.8
1.5
6.9
0
0.6
3.1
0.1
NanoSelectTM LF200
159
2.2
2.0
Pd/C
120
254
201
99.7
99.7
100.0
99.8
97.7
0.3
0.3
0
0.3
0.3
0.1
5.9
3.5
10.1
6.9
12.1
0
89.3
92.5
87.8
3.0
0.6
0.6
0.5
0.3
0.2
Pd/FDU-12(4.6)_Imp
Pd/FDU-12(0.6)_NP_0.02
Pd/FDU-12(3.7)_PEA_0.02 175
Pd/FDU-12(4.5)_CAB_0.02 173
91.1
95.2
0
1.3
2
6