Increasing the availability of active sites in Zn-Co double metal cyanides by dispersion onto a SiO2 support

Article abstract of DOI:10.1016/j.jcat.2017.08.008

Zn-Co double metal cyanides (DMCs) were dispersed on silica by dry and liquid-assisted (LAG) grinding to improve the catalytic performance of the DMC phase. The characterization of the resulting materials by transmission electron microscopy, N₂ physisorption, and powder X-ray diffraction indicated that the presence of water during the grinding step, not only helped to preserve the structural and textural properties of the DMC, but also facilitated the dispersion of the active phase onto the silica support. The accessibility of the Lewis acid sites was increased, as evidenced by FTIR spectroscopy with pyridine as probe molecule. The catalytic activity of the DMC phase for both the hydroamination of phenylacetylene with 4-isopropylaniline and the ring opening polymerization of 1,2-epoxyhexane was enhanced upon dispersion on silica due to the greater availability of the active Zn(II) sites. As a result, higher turnover frequencies were obtained upon decreasing the amount of DMC in the sample. The positive effect on the catalytic activity was the most pronounced for the catalysts prepared via LAG.

Full text of DOI:10.1016/j.jcat.2017.08.008

Journal of Catalysis 354 (2017) 92–99
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Increasing the availability of active sites in Zn-Co double metal cyanides
by dispersion onto a SiO support
2
a
b
b
b
Carlos Marquez , Miguel Rivera-Torrente , Pasi P. Paalanen , Bert M. Weckhuysen ,
a
a,
⇑
a,c,
⇑
Francisco G. Cirujano , Dirk De Vos , Trees De Baerdemaeker
a
Centre for Surface Chemistry and Catalysis, KU Leuven, Celestijnenlaan 200F, P.O. Box 2461, 3001 Leuven, Belgium
Inorganic Chemistry and Catalysis, Debye Institute for Nanomaterials Science, Utrecht Universiteit, Universiteitsweg 99, 3584 CG Utrecht, The Netherlands
Department of Chemistry, University of Oxford, Inorganic Chemistry Laboratory, South Parks Road, Oxford OX1 3QR, UK
b
c
a r t i c l e i n f o
a b s t r a c t
Article history:
Zn-Co double metal cyanides (DMCs) were dispersed on silica by dry and liquid-assisted (LAG) grinding
to improve the catalytic performance of the DMC phase. The characterization of the resulting materials by
transmission electron microscopy, N physisorption, and powder X-ray diffraction indicated that the
2
Received 20 April 2017
Revised 7 August 2017
Accepted 9 August 2017
presence of water during the grinding step, not only helped to preserve the structural and textural prop-
erties of the DMC, but also facilitated the dispersion of the active phase onto the silica support. The acces-
sibility of the Lewis acid sites was increased, as evidenced by FTIR spectroscopy with pyridine as probe
molecule. The catalytic activity of the DMC phase for both the hydroamination of phenylacetylene with 4-
isopropylaniline and the ring opening polymerization of 1,2-epoxyhexane was enhanced upon dispersion
on silica due to the greater availability of the active Zn(II) sites. As a result, higher turnover frequencies
were obtained upon decreasing the amount of DMC in the sample. The positive effect on the catalytic
activity was the most pronounced for the catalysts prepared via LAG.
Keywords:
Double metal cyanides
Liquid-assisted grinding
Ball milling
Catalyst support
Hydroamination
Ring opening epoxide polymerization
Ó 2017 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
1
. Introduction
2
copolymerization of CO and epoxides [11,12], esterification and
transesterification [13–17], hydroamination [18,19], Prins condensa-
tion [20] and oxidation type reactions [21]. For epoxide ring-opening
polymerization (ROP) reactions and hydroamination reactions,
Double metal cyanides (DMCs) are coordination polymers in
which two different metals are connected through a cyanide group
2
(
C„N). DMCs have the general formula M1
u
[M (CN)
n
v
] , where the
3 6 2
Zn [Co(CN) ] based materials (Zn-Co DMCs) show the best perfor-
most common u-v combinations are 3–2, 1–1 and 4–3. For the 3–2
mance and the most promise in catalysis [18,22].
The synthesis of double metal cyanide compounds is relatively
easy and normally involves a fast precipitation reaction after
1
combination, M is often a divalent metal, such as Zn(II), Fe(II), Co
2
(
II), or Ni(II); and M (CN)
n
is a hexacyanometallate anion of a triva-
1
lent metal such as Co(III), Fe(III) or Cr(III). The structure of these
combining solutions of a water-soluble metal salt (M X
v
) and a
1
v+
2
compounds is often cubic with a rock salt type ordering of [M ]
water-soluble metal cyanide salt (Y
increase the activity of the as-synthesized material, an excess of
u n
[M (CN) ]) [23]. In order to
2
uÀ
2
uÀ
6
and [M (CN)
6
]
, in which the [M (CN)
]
positions are only par-
1
tially occupied to maintain charge neutrality [1–4]. Recently, DMCs
have been investigated in diverse applications, for instance, in gas
and energy storage [5–8]. In catalysis, they have been used since
the 1960s, when the General Tire & Rubber company discovered
their potential as catalyst for the ring opening polymerization of
epoxides [9,10]. Since then, many studies have focused on new
catalytic applications of DMCs in numerous reactions, notably,
the metal salt (M X
agents (co-CAs) are added to the synthesis mixture, especially in
the case of Zn [Co(CN) based materials [22]. CAs are typically
v
), and complexing (CAs) and co-complexing
3
6 2
]
low molecular weight alcohols (e.g. ethanol, isopropanol,
n-butanol, with tert-butanol being the most frequently used).
Co–CAs are usually polyethers like poly(tetramethylene ether)
Ò
glycol (PTMEG), polyethylene glycol (PEG) or Pluronic P123
(
P123) [19,24]. While the effect of these agents on the catalyst
structure and activity is not entirely clear, co-CAs are believed to
weakly coordinate to Zn(II) and Co(III) ions via hydroxyl groups
during the nucleation process of the DMC, providing steric
stabilization and preventing excessive particle growth [25].
⇑
Corresponding authors at: Centre for Surface Chemistry and Catalysis, KU
Leuven, Celestijnenlaan 200F, P.O. Box 2461, 3001 Leuven, Belgium (T. De
Baerdemaeker and D. De Vos).
E-mail addresses: dirk.devos@kuleuven.be (D. De Vos), trees.debaerdemaeker@
kuleuven.be (T. De Baerdemaeker).
http://dx.doi.org/10.1016/j.jcat.2017.08.008
0
021-9517/Ó 2017 The Authors. Published by Elsevier Inc.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

C. Marquez et al. / Journal of Catalysis 354 (2017) 92–99
93
]3 sites are
À
responds to the stoichiometric ratio of 1.5, while for the cubic
modification, a Zn/Co of 2.2 is used (Supporting information).
In the cubic Zn
3
[Co(CN)
6
]
2
, one third of the [Co(CN)
6
vacant to maintain framework neutrality [3]. While these vacan-
cies increase the micropore volume, the pores are usually too small
(
ꢀ5 Å) to allow diffusion of reactants and products, resulting in
2.1.2. Preparation of silica-supported DMC catalysts by physical
mixing (PM)
organic reactions mostly taking place on the external surface of
the catalyst [19]. Several strategies have been suggested in order
to increase the number of active sites available per mass of DMC.
For example, Yi et al. [26] reported the synthesis of double and
multi-metal cyanide nanoparticles (nano-sized DMCs) using a
reverse emulsion technique, which increased the external surface
area of the particles and further increased the catalytic activity of
the materials for the copolymerization of CO
oxide. More recently, the synthesis of hierarchically mesoporous
M -M DMCs (M = Cu(II), Co(II) and Ni(II); and M = Co(III)) in an
ionic liquid/water/MgCl system has been reported and the result-
Specific amounts of DMC-PTMEG and 1 g of silica gel (Sigma-
Aldrich, high purity grade, pore size 150 Å, 200–425 mesh) were
ground for 15 minutes with an agate mortar and pestle set, with
(LAG) and without water (0.6 ml). The solids were dried by heating
under vacuum at 353 K overnight to obtain two series of silica-
supported DMC catalysts: PM-X%wet and PM-X%dry where X refers
to the wt.% of DMC in the solid. Additional supported DMCs were
prepared under LAG conditions, using titanium oxide (AEROXIDE
2
TiO P25) or zirconium (IV) oxide (Sigma-Aldrich, powder, 5 mm) as
support.
2
and cyclohexene
Ò
1
2
1
2
2
ing materials were used for the electrochemical reduction of CO
formic acid [27].
2
to
2.1.3. Preparation of silica-supported DMC catalysts by ball milling
(BM)
Another widely investigated strategy to improve the efficiency
of catalytic processes is the immobilization of a catalytically active
phase on a support, which can be useful to enhance the long-term
performance of the catalyst while reducing the required amount of
the catalytically active material [28–31]. For copolymerization
reactions, SiO and TiO supported DMC catalysts prepared by
2 2
co-precipitation of the DMC precursor salts and tetraethyl
orthosilicate or titanium ethoxide were found to be active
In order to understand the effect of the grinding process on the
physicochemical properties of the material, a series of catalyst
were prepared using ball milling. DMC-PTMEG was subjected to
ball milling in a stainless steel grinding jar with two stainless steel
grinding balls (7 mm diameter) without support for different times
(2, 5, 10, 15 and 20 min) using a Retsch MM 400 ball mill at a fre-
quency of 30 Hz. Additionally, specific amounts of DMC-PTMEG
and 0.5 g silica gel, both with and without water (0.3 ml), were ball
milled for 15 min under the same conditions to obtain two series of
silica-supported DMC catalysts: BM-X%wet and BM-X%dry where X
refers to the wt.% of DMC in the solid.
[
32,33]. Here, we report the formation of supported Zn-Co DMC
catalysts by simple grinding or ball milling in presence of a sup-
port. Our goal is to increase the accessibility of the active sites –
i.e., the Zn(II) sites in the case of hydroamination and epoxide
polymerization [19,22] – to achieve a higher turnover frequency
(
TOF). The effects of adding water during the grinding process (liq-
2.2. Catalyst characterization
uid assisted grinding, LAG), the support type and the DMC/support
ratio are carefully investigated. The catalysts are characterized
using transmission electron microscopy (TEM), Fourier transform
infrared spectroscopy with pyridine as probe molecule (Py-FTIR),
The metal content of the catalysts was determined by ICP-OES
analysis using a Varian 720-ES equipped with a double-pass glass
cyclonic spray chamber, a Sea Spray concentric glass nebulizer
and a high solids torch. The samples were digested using HF (aq).
PXRD patterns were collected on a STOE Stadi MP diffractometer
operating in transmission mode, using an image plate detector
2
powder X-ray diffraction (PXRD) and N physisorption and their
catalytic performance was investigated in the hydroamination of
phenylacetylene with 4-isopropylaniline and in the polymeriza-
tion of 1,2-epoxyhexane.
and focusing Ge(1 1 1) monochromator (Cu
k = 1.54060 Å), or on a Bruker D8 Advance eco diffractometer in
Bragg-Brentano geometry with a LYNXEYE XE-T detector (CuK
over a 10–60° 2h range. N physisorption isotherms were collected
K
a
1
radiation,
a
12),
2
. Experimental section
2
at 77 K on a Micromeritics 3Flex Surface Analyzer after evacuating
2
2
.1. Catalyst preparation
the samples at 423 K for 16 h. The specific surface area (SBET) was
determined using the BET method (0.05–0.3 p/p ), the specific
0
.1.1. Preparation of reference DMCs
A reference DMC catalyst, referred to as DMC-PTMEG, was syn-
external surface area (Sext) and micropore volume (Vmicro) were
obtained using t-plot analysis and the median pore width was
determined using the HorvathÀKawazoe model. TEM allowed the
determination of DMC crystallite size and dispersion onto the sup-
port. Bright field measurements were obtained in a Tecnai T12
(FEI) microscope with a field emission gun operating at 120 kV.
High angle annular dark field (HAADF) images and EDX maps were
obtained in a Talos F200x (FEI), operated at 300 kV and equipped
with high-brightness field emission gun (X-FEG) and a Super-X
G2 EDX detector. Prior to imaging, the samples were gently ground,
suspended in methanol and dropped onto a Cu grid (200 mesh)
with holey carbon film. Pyridine adsorption followed by FTIR spec-
troscopy (Py-FTIR) was used to determine the acid site nature and
density of the samples using a Nicolet 6700 FTIR spectrometer. A
thesized with CA (tert-butanol) and co-CA (PTMEG) according to
literature procedures [18,34]. Solution A was prepared by dissolv-
ing 15 mmol of ZnCl
average Mn ꢀ1000) in 150 ml of distilled water. Solution B was
prepared by the addition of 1.5 mmol of K [Co(CN) ] to 15 ml of
distilled water. Solution B was then added dropwise to solution A
under vigorous stirring followed by the addition of 37.5 mL of
tert-butanol ( BuOH). The final mixture was stirred for 3 h at room
temperature. The obtained solids were recovered by centrifugation
and washed three times with a 50:50 mixture of water: BuOH.
After drying at 333 K overnight, the solid product was ground to
obtain a fine powder.
2
and 1.5 mmol of PTMEG (Sigma-Aldrich,
3
6
t
t
À2
DMCs with high phase purity were synthesized following the
procedures reported by Kuyper and Boxhoorn [35]. Both synthesis
methods involve precipitation reactions between aqueous solu-
self-supporting wafer (ꢀ10 mg Á cm ) was placed in a cell under
vacuum and activated at 523 K for 1 h. The cell was cooled down
and the probe molecule (25 mbar) was allowed to adsorb onto
the wafer at 323 K until saturation. The physisorbed pyridine
was removed by evacuation for 30 min before reheating to 423 K
to record the IR spectrum. The acid site density was determined
from the areas of the absorption bands corresponding to pyridine
2 3 6
tions of ZnCl and K [Co(CN) ], without the presence of CA or co-
CA. The rhombohedral phase is preferably obtained at higher tem-
perature (373 K), from slightly more dilute solutions. The molar
Zn/Co ratio in the synthesis of the rhombohedral modification cor-

9
4
C. Marquez et al. / Journal of Catalysis 354 (2017) 92–99
À1
coordinated to Lewis acid sites (1450 cm ) and using the inte-
grated molar extinction coefficients of Emeis [36].
reflections at 14.5° and 23.6° 2h indicating the presence of a small
amount of the monoclinic phase [2,37]. This mixture of phases is
consistent with previous reports of DMCs prepared using CAs
and co-CAs [19].
PXRD patterns of ball milled DMC-PTMEG were collected after
specific time points to follow structural changes and amorphiza-
tion of the DMC during the grinding process without a support pre-
sent (Fig. 1). As can be observed in Fig. 1a, without the presence of
water during the ball milling process, changes in the structure of
the material occur almost immediately (2 min), with amorphiza-
tion taking place after only a few minutes of grinding. Moreover,
a closer inspection of the patterns (Fig. S1, Supporting Information)
shows the appearance of additional peaks at 16.4°, 19.8° and 21.8°
2
2
.3. Catalytic reactions
.3.1. Intermolecular hydroamination
Before reaction, the catalyst samples were activated at 353 K
under vacuum overnight. Glass crimp cap reaction vials were
loaded under N atmosphere with 50 mg of catalyst, pheny-
lacetylene (0.5 mmol), 4-isopropylaniline (1 mmol), dry toluene
1 ml) as solvent and tetradecane (1 mmol) as internal standard.
The vials were placed in a heated copper block at 383 K and stirred
at 500 rpm using a magnetic stirring bar. After reaction, the cata-
lyst was removed by centrifugation and the liquid supernatant
was analyzed by GC (Shimadzu 2014 GC equipped with a FID
detector and a CP-Sil 5 CB column) and GC-MS (Agilent 6890 gas
chromatograph, equipped with a HP-5MS column, coupled to a
2
(
2h, which are typical of a denser, more hydrophobic, rhombohedral
phase of Zn [Co(CN) [38,39]. In contrast, under LAG conditions
3
6 2
]
the crystalline structure of the solid remains more stable, with
the pattern showing sharp reflections even after 20 min of grinding
5
973 MSD mass spectrometer). Additional reactions with lower
(Fig. 1b). No additional peaks are observed, indicating that no
amounts of DMC-PTMEG were performed. After the reaction, the
sample PM-9%wet was dried under vacuum and further character-
ized by XRD. Recycling test were performed after recovery and
re-activation of the sample by drying at 353 K under vacuum over-
night before each run.
phase transformation occurs when water is present during the
grinding process. The formation of the rhombohedral phase is pos-
sibly inhibited because under LAG conditions, there are always
water molecules available that could readily coordinate to the zinc
atoms, keeping the DMC in its cubic, hydrated form.
Comparable behavior is observed in the samples ball milled
together with SiO2 (Fig. 2). For both series, reflections assigned to
the cubic DMC phase can be seen superimposed over the broad
contribution of the amorphous silica. Once again, dry ball milling
resulted in a notable destruction of the DMC structure and a partial
phase transformation. In the case of the silica supported DMCs pre-
pared by physical mixing (Fig. S2, Supporting Information), the
effect of neat grinding is less pronounced, though for the sample
PM-29%wet, the reflections are still sharper than those observed
in the PM-29%dry pattern. For the latter, the peaks assigned to the
rhombohedral phase are also more intense.
The N2 physisorption isotherms obtained for the samples BM-
9%wet, BM-29%wet and BM-29%dry are shown in Fig. 3. All isotherms
are a combination of the two starting materials: a type I isotherm,
typical of microporous materials for DMC-PTMEG (median pore
width = 0.6 nm) and a type IV isotherm for silica gel (Fig. S3, Sup-
porting Information). A clear effect of the use of water during the
grinding process is observed. In fact, liquid-assisted grinding
resulted in a higher specific surface area and external surface area
solid when compared to those prepared under dry ball milling con-
ditions, as shown in Table 1. These results are in good agreement
with the XRD characterization and suggest that with the presence
of a solvent during the ball milling the crystalline structure of the
material is preserved to a larger extent. Additionally, for a more
diluted catalyst, BM-9%wet, the values obtained for SBET and Sext
2
.3.2. Epoxide polymerization
Before reaction, the catalyst samples were activated at 353 K
under vacuum overnight. Glass crimp cap reaction vials were
loaded with 20 mg of catalyst, 1,2-epoxyhexane (2 mmol) and
diethyleneglycol (0.2 mmol), under N
was carried out in a heated copper block at 383 K and stirred at
00 rpm using a magnetic stirring bar. After reaction, the product
2
atmosphere. The reaction
5
mixture was diluted in carbon disulfide (2.5 ml), and the catalyst
was removed by centrifugation. The liquid supernatant was then
collected in a 10.00 mm liquid cell for near-infrared measurements
on a Cary 5000 UV-VIS-NIR spectrophotometer. The absorbance of
the sample at 2220 nm was used to determine the epoxide conver-
sion by correlation to the concentration of 1,2-epoxyhexane left in
the reaction mixture using a calibration curve.
3
. Results and discussion
3
.1. Characterization
ICP analysis of DMC-PTMEG shows that the Zn/Co ratio is higher
than the stoichiometric ratio of 1.5 expected for a phase pure cubic
Zn [Co(CN) (Table S1, Supporting information). However, the
PXRD pattern shows that DMC-PTMEG still consists predominantly
3
6 2
]
of the cubic phase (e.g., peaks at 15°, 17.3° and 24.6° 2h, Fig. 1) with
Fig. 1. XRD patterns of DMC-PTMEG after different grinding times: (a) dry grinding and (b) LAG. Reflections corresponding to the monoclinic phase are denoted by (r) and to
the rhombohedral phase by (.).

C. Marquez et al. / Journal of Catalysis 354 (2017) 92–99
95
Fig. 2. XRD patterns of ball milled samples: (a) BM-9% prepared by (j) LAG and (h) dry grinding and (b) BM-29% prepared by (j) LAG and (h) dry grinding.
600
500
400
300
200
100
0
This suggest that the use of a solvent during the ball milling not
only helps preserve the structural and textural properties of the
DMC, but also facilitates the segregation of conglomerated DMC
crystals and their dispersion onto the silica support. Furthermore,
with parallel EDX elemental mapping (Fig. 4d and e), the close
proximity between the support and the active phase for the sample
BM-29%wet is clearly observed, with a considerable amount of DMC
homogeneously supported and well dispersed onto the silica.
Pyridine adsorption followed by FTIR spectroscopy was used to
probe the surface acidity of the synthesized materials. Fig. 5 shows
the difference spectra obtained for the reference DMC and the
silica-supported DMCs prepared by physical mixing under LAG
conditions. The appearance of bands at 1610, 1492 and
À1
1
451 cm
indicates the presence of Lewis acid sites, whereas
À1
the absence of bands at 1639 and 1545 cm denotes the lack of
Brønsted acidity [40], showing that Zn-Co DMCs are purely Lewis
acid catalysts. Furthermore, the low intensity of the 1575 cm
0
0.2
0.4
0.6
0.8
1
À1
Relative pressure (p/p°)
band means that the majority of the hydrogen-bonded pyridine
is desorbed after heating at 423 K. The amount of pyridine
adsorbed on Lewis acid sites per mass of DMC increases with the
degree of dilution of DMC in the mixture (Table 2): from
Fig. 3. Nitrogen physisorption isotherms of BM-9%wet (j/h), BM-29%wet (d/s) and
BM-29%dry (▲/h). Filled symbols represents adsorption and open symbols,
desorption.
0
.11 mmol/g for DMC-PTMEG to 0.66 mmol/g for PM-9%wet. This
Table 1
Textural properties determined from
materials and of selected samples.
N
2
physisorption at 77 K of the starting
is a direct indication of the enhanced availability and accessibility
of the Zn(II) Lewis acid sites upon dispersion of the DMC phase
onto the support. The formation of Lewis-coordinated pyridine
species is facilitated because of the greater exposure of surface
Zn(II) sites, which is also suggested by the increase in Sext observed
with the decrease in the amount of DMC in the sample (Table 1).
2
2
Sample
S
BET (m /g)
S
ext (m /g)
Vmicro (ml/g)
DMC-PTMEG
Silica gel
652
365
112
0.284
340
0.012
BM-9%wet
BM-29%wet
BM-29%dry
400 (391)
354 (448)
292 (448)
332 (320)
223 (274)
155 (274)
0.033 (0.037)
0.070 (0.091)
0.066 (0.091)
3
3
.2. Intermolecular hydroamination
Theoretically predicted weight averages from DMC-PTMEG and the support are
showed in parenthesis.
.2.1. Silica supported DMCs
The catalytic activity of the silica-supported DMCs was first
are slightly higher than the weighted average of DMC-PTMEG and
the support for these values, which in the case of Sext, is indicative
of the formation of smaller crystalline particles (Table 1). In con-
trast, for the samples containing a higher concentration of DMC –
BM-29%wet and BM-29%dry – these values are lower than the
weighted average, demonstrating the effect the ball milling process
has on the textural properties of the supported catalysts.
investigated for the intermolecular hydroamination of 4-
isopropylaniline with phenylacetylene. The hydroamination
results in the formation of an enamine that tautomerizes to a more
stable
imine
(4-isopropylphenyl)(1-phenylethylidene)amine
(Scheme 1) [18]. Hydroamination represents a green route towards
the production of substituted amines [41,42]. Besides the fact that
it is a 100% atom efficient reaction, the formation of side-products
is limited: under these reaction conditions, only the formation of
acetophenone as a side product was observed.
TEM bright field images (Figs. 4 and S4, Supporting Information)
show mostly supported, but also more isolated DMC crystals in
close proximity to SiO
2
. Even though size and morphology are
Fig. 6 shows the yield of the desired imine product for the
unsupported DMC-PTMEG and the catalysts prepared by physical
mixing under LAG conditions after 24 h reaction time. The corre-
sponding space-time yield (STY) of the reaction can be found in
Fig. S5, Supporting Information. As can be observed, with DMC-
PTMEG (100 mg DMC per mmol of phenylacetylene) a high yield
can be achieved (81%). Unsurprisingly, when the DMC loading is
barely modified during the ball milling process, the larger aggre-
gates present in the DMC-PTMEG (>500 nm) were not detected
after the ball milling process. For the samples prepared under
LAG conditions, the DMC particles are in close contact with the
support (Fig. 4a and b), whereas for the dry ball milled sample
(Fig. 4c) a higher fraction of DMC crystals seems to be unsupported.

9
6
C. Marquez et al. / Journal of Catalysis 354 (2017) 92–99
500 nm
500 nm
500 nm
a)
b)
c)
d)
e)
Fig. 4. TEM bright field images of samples BM-9%wet (a), BM-29%wet (b) and BM-29%dry (c); HAADF-STEM (d) and EDX composition map (e) of BM-29%wet. The green arrows
indicate the SiO
2
support, the blue arrows the presence of DMC crystallites supported on SiO
2
2
and the red ones, unsupported DMC crystallites in close proximity to SiO .
Table 2
L
L+H
Lewis acid properties as determined by FTIR analysis
with pyridine as probe molecule.
Sample
LAS (mmol/g)a
L
DMC-PTMEG
PM-29%wet
PM-17%wet
PM-9%wet
0.11
0.14
0.28
0.66
H
d
c
b
Silica gel
n.d
a
Amount of pyridine adsorbed on Lewis acid sites
(
LAS) per gram of DMC.
Not detected.
b
a
b
ilar to the one obtained with the samples prepared by physical
mixing. Furthermore, the DMC ground without the presence of sil-
ica, even under LAG conditions (PM-100%dry and PM-100%wet),
showed the same activity as the original DMC-PTMEG (Table S2,
Supporting Information).
To evaluate the activity of the DMC phase in the catalysts, the
initial TOF of the Zn(II) sites (mol imine produced per mol Zn(II)
per hour) was calculated. As can be observed in Fig. 7, the TOF
increases with the degree of dilution of DMC in the sample, espe-
cially for the samples prepared under LAG conditions, which is in
good agreement with the increased amount of pyridine adsorbed
on Lewis acid sites as indicated by Py-FTIR spectroscopy. This fur-
ther indicates that the availability of the Zn(II) sites is improved
upon dispersion of the active phase onto the support. The Zn(II)
sites – the active sites for this reaction [19] – are in this case more
accessible to the reactant molecules causing the rise in catalytic
activity.
1
650
1600
1550
1500
1450
1400
-1
Wavenumber (cm )
Fig. 5. Difference IR spectra of adsorbed pyridine (normalized to 10 mg of DMC/
cm ) for: (a) DMC-PTMEG, (b) PM 29%wet, (c) PM-17%wet and (d) PM 9%wet. Peaks
corresponding to pyridine coordinated to Lewis acid sites (L) and hydrogen-bonded
pyridine (H) are marked.
2
reduced, the yield decreases; however, when the active phase is
supported on silica, the imine production is largely maintained –
even if the overall STY decreases compared to the unsupported
DMC-PTMEG in the highest loading (100 mg DMC per mmol of
phenylacetylene). For all investigated DMC:substrate ratios the
imine production is substantially higher than the one achieved
with the unsupported solid. For the ball milled samples
(
Table S2, Supporting Information), the yield obtained is very sim-

C. Marquez et al. / Journal of Catalysis 354 (2017) 92–99
97
Scheme 1. Intermolecular hydroamination of 4-isopropylaniline with phenylacetylene.
1
00
catalysts prepared with DMC-PTMEG (PM-9%wet) (Table S2, Sup-
porting Information). While the sample PM-9%wet – Cubic showed
an increased activity relative to its unsupported counterpart, the
activity was still inferior to the one exhibited by PM-9%wet. The
results are in agreement with previous reports that claim that
90
80
70
60
50
40
30
20
10
0
t
the use of complexing and co-complexing agents, such as BuOH
and PTMEG, increase the activity of DMCs by affecting active site
accessibility and particle size, while maintaining phase purity
and crystallinity [19,25].
Recycling tests show that the catalyst largely maintains its
activity after five runs (Fig. S6, Supporting Information), with no
notable loss of crystallinity or change in the phase composition
of the sample when compared to the pristine one (Fig. S7, Support-
ing Information). Furthermore, the heterogeneity of the catalyst
was studied by a hot filtration test by removing the catalyst after
0
20
40
60
80
100
DMC loading (mg DMC / mmol of phenylacetylene)
Fig. 6. Yield of the hydroamination product after 24 h reaction time at 383 K for the
silica supported DMCs prepared by physical mixing under LAG conditions (j), and
for unsupported DMC-PTMEG (d).
4
h of reaction time. Following this removal, no increase in conver-
sion was observed after additional stirring for 20 h at 383 K
Fig. S8, Supporting Information). These results indicate that no
catalytically active species leach from the catalyst.
(
0
0
0
.6
.4
.2
0
3
.2.2. Support scope
To appreciate the role of silica as support, the hydroamination
2 2
reaction was repeated using ZrO and TiO as supports. These oxi-
des were chosen because they are commonly used supports that
normally exhibit catalytic properties on their own related to their
Lewis acidity and their interaction with the active phase [33,43–
4
5]. The supports showed poor catalytic activity without the pres-
ence of the DMC phase, producing the desired product only in low
yields (Table S2, Supporting Information) even after 24 h of reac-
tion. This again shows the important role the DMC plays as the
active phase for the intermolecular hydroamination reaction.
Table 3 presents the initial TOFs (Zn(II) based) obtained for the
DMC on different supports. The activity increases in the order
9
17
23
29
50
100
DMC content (wt.%)
Fig. 7. Initial TOF (mol imine produced per mol Zn(II) per hour) for hydroamination
of phenylacetylene at 383 K of DMC-PTMEG (j) and silica supported catalyst
prepared by physical mixing under LAG ( ) and dry ( ) conditions.
2 2 2
TiO < ZrO < SiO ; however, for all supports, a higher hydroamina-
tion activity is achieved for the more diluted sample, suggesting
that the accessibility of the Zn(II) sites can be increased, regardless
of the oxide.
A clearer difference in the activity of the supported samples is
observed when the total imine product yield after 24 h reaction
time is examined (Fig. 8). The yield of the desired imine product
The activity of the samples prepared by LAG is higher than the
activity of samples prepared by dry grinding, especially at low
DMC/support ratios (Fig. 7). XRD and N sorption data showed that
2
the presence of water helps to preserve the structural and the tex-
tural properties of the solid, in particular, the external surface area.
Likewise, TEM images showed that a better dispersion of the DMC
phase was obtained under LAG conditions, which would increase
the accessibility of the Zn(II) sites. Moreover, the dry mixed cata-
lysts also exhibited reflections attributed to the rhombohedral
phase. In order to assess the relative activity of the phases, DMC
catalysts with high phase purity (rhombohedral and cubic) were
synthesized following the procedures reported by Kuyper and Box-
hoorn [35]. The TOF for the cubic modification was found to be ele-
ven times higher (Table S2, Supporting Information) than for the
rhombohedral counterpart, which could be related to the lower
density and hence, slightly higher accessibility of the cubic phase.
This could partially contribute to the higher activity obtained for
the liquid-assisted ground samples.
2
can be increased by 10% when SiO is used as support instead of
TiO or ZrO . This could be an indication that the acidic properties
2
2
of these oxides do not play an important role in the overall
hydroamination activity of the catalyst and demonstrate that the
silica gel, with its higher surface area, represents the best alterna-
tive of support. Further characterization of the catalysts and pure
supports (Table S3, Supporting Information) shows that the
Table 3
Initial TOF for hydroamination of DMC supported on different oxides.
Support
SiO
2
TiO
2
ZrO
2
PM-9%wet
PM-29%wet
0.60
0.34
0.33
0.17
0.58
0.27
The pure-phased cubic material was also supported on silica
(
PM-9%wet – Cubic), and its activity was compared to the supported
TOF = mol imine produced per mol Zn(II) per h.

9
8
C. Marquez et al. / Journal of Catalysis 354 (2017) 92–99
60
50
40
30
20
10
0
25
2
1
0
5
10
5
0
5
10
15
20
25
30
35
DMC content (wt.%)
Fig. 8. Yield of the hydroamination product after 24 h reaction time at 383 K for the
DMC samples supported on SiO
mixing under LAG conditions.
2
(j), ZrO
2
(▲) and TiO
2
(r) prepared by physical
0
9
17
23
29
100
DMC content (wt.%)
dispersion of the acid sites, expressed as surface Lewis acid sites
density, is increased when the DMC phase is supported on TiO
and SiO , while both the pure ZrO and the ZrO -supported DMC
present a higher concentration of Lewis acid sites on the external
surface in comparison to the unsupported DMC-PTMEG. This low
dispersion can be attributed to the low external surface area of
the support.
Fig. 9. Initial TOF (mol epoxyhexane consumed per mol Zn(II) per hour) for ring
opening polymerization of 1,2-epoxyhexane at 383 K of DMC-PTMEG (j) and silica
supported catalyst prepared by physical mixing under LAG ( ) and dry (
conditions.
2
2
2
2
)
4
. Conclusions
The results presented here demonstrate that a highly active,
These results are in contrast with previous studies [33], where
the dispersion of DMC over TiO
activity (superior chemoselectivity) relative to the DMC dispersion
onto SiO in the copolymerization of CO with propylene oxide.
2
had a better effect on the overall
stable and reusable supported DMC catalyst can be prepared by
simple physical mixing methods. In the two reactions probed, the
activity of the DMC phase was enhanced by supporting it onto sil-
ica, resulting in up to five times higher TOFs with a higher degree
of dilution of the active phase. This catalytic result shows that the
availability and accessibility of the Zn(II) sites are improved upon
dispersion onto the support as was also indicated by TEM and FTIR
characterization.
2
2
However, in this case, the supports were formed simultaneously
with the precipitation of the DMC phase via condensation of alkox-
ide precursors, likely resulting in different textural properties and
surface species.
Both methods of preparation of the supported catalysts – phys-
ical mixing and ball milling – have a similar effect on the catalytic
performance. In the hydroamination of phenylacetylene with 4-
isopropylaniline, silica proved to be the best choice of support,
although the accessibility of the active sites was increased regard-
less of the support used. These results could probably be related to
its higher specific surface and external surface area, allowing a bet-
ter dispersion of the active DMC phase.
The samples prepared under LAG conditions showed a higher
activity when compared to the samples prepared under dry grind-
ing conditions. The effect of water during the grinding process
seems to be threefold: it helps to preserve the structural and tex-
tural properties of the DMC, promotes the dispersion of the DMC
onto the support, and inhibits phase transformation from the cubic
phase to the less active rhombohedral phase.
3
.3. Epoxide polymerization
Ring opening polymerization of epoxides is one of the most
important chemical processes catalyzed by double metal cyanides
46,47]. reaction scheme of the polymerization of 1,2-
[
A
epoxyhexane is presented in Scheme 2.
The initial TOF (mol epoxyhexane consumed per mol Zn(II) per
hour) of the Zn(II) sites was calculated to gain information on the
catalytic activity of the materials for this particular reaction. The
dispersion of the DMC phase generated an increase in the catalytic
activity, especially when the physical mixing process was carried
out under LAG conditions (Fig. 9), reaching a TOF of 21.1 h for
PM-9%wet
For this reaction, the effect of the solvent during the dispersion
process on the activity of the samples is once again very pro-
À1
.
Overall, the results confirm that the number of active sites
available in the DMC can easily be increased through physical mix-
ing or ball milling with a suitable and simple support, such as sil-
ica. This implies that in order to achieve a high activity, lower
amounts of the catalytically active material are required. These
results will have direct implications for large scale applications,
as further mechanical processing and shaping of the heterogeneous
DMC catalyst with binders could seriously affect – and improve –
the catalytic performance.
nounced, especially at low DMC loadings, with the sample PM-
À1
9
%dry showing a TOF of only 13.8 h . This value is, however, still
much higher than the one exhibited by the DMC-PTMEG (TOF of
À1
4
.1 h ), which is a clear indication that the availability and acces-
sibility of the active sites can be increased by simple dispersion of
the active phase onto a SiO . The activity shown by the samples
2
prepared by ball milling is very similar to that exhibited by the
samples prepared by physical mixing (Table S4, Supporting
Information).
Acknowledgment
This project has received funding from the European Union’s
Horizon 2020 research and innovation programme under the
Marie Sklodowska-Curie grant agreement No 641887 (project
acronym: DEFNET). T.D.B. acknowledges F.W.O. – Vlaanderen
2 3 3
Scheme 2. Ring opening polymerization of 1,2-epoxyhexane. R: -(CH ) CH .

C. Marquez et al. / Journal of Catalysis 354 (2017) 92–99
99
[21] A. García-Ortiz, A. Grirrane, E. Reguera, H. García, Mixed (Fe2+ and Cu2+) double
metal hexacyanocobaltates as solid catalyst for the aerobic oxidation of
oximes to carbonyl compounds, J. Catal. 311 (2014) 386–392.
(
Research Foundation Flanders) for a Postdoctoral Fellowship. D.D.
V. thanks F.W.O. for project funding (grant no. G.0256.14N) and KU
Leuven for the Methusalem grant CASAS. The authors thank J. Ze cˇ e-
vi c´ (UU) and W. S. Lamme (UU) for their help with the TEM and ele-
mental maps, prof. K. Binnemans (KUL) for the use of the UV-VIS-
NIR spectrophotometer and prof. A.L. Goodwin (Oxford) for the use
of the ball mill and discussions.
[
22] X.-H. Zhang, Z.-J. Huab, S. Chenc, F. Liua, X.-K. Suna, G.-R. Qi, Role of zinc
chloride and complexing agents in highly active double metal cyanide
catalysts for ring-opening polymerization of propylene oxide, Appl. Catal. A
325 (2007) 91–98.
[
23] H. Liu, X. Wang, Y. Gu, W. Guo, Preparation and characterization of double
metal cyanide complex catalysts, Molecules 8 (2003) 67–73.
[
24] I. Kim, K.J. Lee, D.W. Park, B.U. Kim, C.S. Ha, Aliphatic polycarbonate synthesis
by copolymerization of carbon dioxide with epoxides over double metal
cyanide catalysts prepared by using ZnX2 (X = F, Cl, Br, I), Catal. Today 111
Appendix A. Supplementary material
(
2006) 292–296.
[
25] I.K. Lee, J.Y. Ha, C. Cao, D.-W. Park, C.-S. Ha, I. Kim, Effect of complexing agents
of double metal cyanide catalyst on the copolymerizations of cyclohexene
oxide and carbon dioxide, Catal. Today 148 (2009) 389–397.
Detailed explanation of the synthesis of high phase purity
DMCs, additional characterization data (ICP, XRD patterns, TEM
images, N
2
physisorption isotherms) and tables containing cat-
[26] M.J. Yi, S.H. Byun, C.-S. Ha, D.-W. Park, I. Kim, Copolymerization of cyclohexene
oxide with carbon dioxide over nano-sized multi-metal cyanide catalysts,
Solid State Ionics 172 (2004) 139–144.
alytic data for the intermolecular hydroamination and the ring
opening polymerization reactions can be found in the online ver-
sion. Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.jcat.2017.08.
[
27] X. Kang, X. Sun, Q. Zhu, X. Ma, H. Liu, J. Ma, Q. Qian, B. Han, Synthesis of
hierarchical mesoporous Prussian blue analogues in ionic liquid/water/MgCl
and application in electrochemical reduction of CO , Green Chem. 18 (2016)
869–1873.
2
2
1
0
08.
[
28] T. Blasco, A. Corma, A. Martínez, P. Martínez-Escolano, Supported
heteropolyacid (HPW) catalysts for the continuous alkylation of isobutane
with 2-butene: the benefit of using MCM-41 with larger pore diameters, J.
Catal. 177 (1998) 306–313.
References
[
[
[
[
29] T. De Baerdemaeker, M. Feyen, U. Müller, B. Yilmaz, F.-S. Xiao, W. Zhang, T.
Yokoi, X. Bao, H. Gies, D.E. De Vos, Bimetallic Zn and Hf on silica catalysts for
the conversion of ethanol to 1,3-butadiene, ACS Catal., 5 (2015) 3393–3397.
30] I. Hermans, E. Breynaert, H. Poelman, R. De Gryse, D. Liang, G. Van Tendeloo, A.
Maes, J. Peeters, P. Jacobs, Silica-supported chromium oxide: colloids as
building blocks, Phys. Chem. Chem. Phys. 9 (2007) 5382–5386.
31] G. Vilé, D. Baudouin, I.N. Remediakis, C. Copéret, N. López, J. Pérez-Ramírez,
Silver nanoparticles for olefin production: new insights into the mechanistic
description of propyne hydrogenation, ChemCatChem 5 (2013) 3750–3759.
32] Y. Dienes, W. Leitner, M.G.J. Müller, W.K. Offermans, T. Reier, A. Reinholdt, T.E.
Weirich, T.E. Müller, Hybrid sol-gel double metal cyanide catalysts for the
[
[
[
1] P. Valvekens, D. De Vos, in: New Materials for Catalytic Applications, Elsevier,
Amsterdam, 2016, p. 1.
2] D.F. Mullica, W.O. Milligan, G.W. Beall, W.L. Reeves, Crystal structure of Zn
CN) O, Acta Crystallogr., Sect. B 34 (1978) 3558–3561.
Á12H
3] A. Ludi, H.U. Guedel, M. Ruegg, The structural chemistry of Prussian blue
analogs. single-crystal study of manganese(II) hexacyanocobaltate(III),
Mn [CO(CN) O, Inorg. Chem. 9 (1970) 2224–2227.
ÁxH
4] G.W. Beall, W.O. Milligan, J. Korp, I. Bernal, Crystal structure of Mn
CN) O and Cd [Co(CN) Á12H O by neutron and X-ray diffraction,
Á12H
Inorg. Chem. 16 (1977) 2715–2718.
5] F.-M. Yang, Y. Liu, L. Chen, C.-T. Au, S.-F. Yina, Synthesis of amine-modified
solid Fe-Zr adsorbents for CO adsorption, J. Chem. Technol. Biotechnol. 91
2016) 2340–2348.
6] S.S. Kaye, J.R. Long, Hydrogen storage in the dehydrated Prussian blue
analogues M [Co(CN) (M = Mn, Fe Co, Ni, Cu, Zn), J. Am. Chem. Soc. 127
2005) 6506–6507.
3
[Co
(
6
]
2
2
A
3
6
]
2
2
[
[
[
[
3
[CO
(
6
]
2
2
3
]
6 2
2
copolymerisation of styrene oxide and CO
177.
33] M.A. Subhani, C. Gürtler, W. Leitner, T.E. Müller, Nanoparticulate TiO
2
, Green Chem. 14 (2012) 1168–
2
1
(
[
[
2
-
supported double metal cyanide catalyst for the copolymerization of CO
with propylene oxide, Eur. J. Inorg. Chem. 13–14 (2016) 1944–1949.
2
3
6 2
]
(
34] I. Kim, J.T. Ahn, C.-S. Ha, C.S. Yang, I. Park, Polymerization of propylene oxide by
using double metal cyanide catalysts and the application to polyurethane
elastomer, Polymer 44 (2003) 3417–3428.
7] Y. Lu, L. Wang, J. Cheng, J.B. Goodenough, Prussian blue: a new framework of
electrode materials for sodium batteries, Chem. Commun. 48 (2012) 6544–
6
546.
[
35] J. Kuyper, G. Boxhoorn, Hexacyanometallate salts used as alkene-oxide
polymerization catalysts and molecular sieves, J. Catal. 105 (1987) 163–174.
36] C.A. Emeis, Determination of integrated molar extinction coefficients for
infrared absorption bands of pyridine adsorbed on solid acid catalysts, J. Catal.
[
8] C.D. Wessells, R.A. Huggins, Y. Cui, Copper hexacyanoferrate battery electrodes
with long cycle life and high power, Nat. Commun. 2 (2011) 550.
9] J. Milgrom, Method of making a polyether using a double metal cyanide
complex, US Patent 3 278 457 (1966) to General Tire and Rubber Company.
[
[
141 (1993) 347–354.
[
[
10] R.J. Herold, Method of making a polyether using a double metal cyanide
complex. US Patent 3 278 457 (1966) to General Tire and Rubber Company.
11] X.H. Zhang, S. Chen, X.M. Wu, X.K. Sun, F. Liu, G.R. Qi, Highly active double
metal cyanide complexes: effect of central metal and ligand on reaction of
[
[
[
37] J.C. Wojdel, S.T. Bromley, F. Illas, J.C. Jansen, Development of realistic models
for Double Metal Cyanide catalyst active sites, J. Mol. Model. 13 (2007) 751–
756.
38] J. Roque, E. Reguera, J. Balmaseda, J. Rodrıguez-Hernandez, L. Reguera, L.F. del
Castillo, Porous hexacyanocobaltates(III): role of the metal on the framework
properties, Micropor. Mesopor. Mater. 103 (2007) 57–71.
epoxide/CO
2
, Chin. Chem. Lett. 18 (2007) 887–890.
[
[
[
[
[
[
[
[
[
12] J. Sebastian, D. Srinivas, Effects of method of preparation on catalytic activity
of Co-Zn double-metal cyanide catalysts for copolymerization of CO
2
and
39] C.P. Krap, B. Zamora, L. Reguera, E. Reguera, Stabilization of cubic and
rhombohedral phases of zinc hexacyanocobaltate (III), Micropor. Mesopor.
Mater. 120 (2009) 414–420.
epoxide, Appl. Catal. A 482 (2014) 300–308.
13] R. Srivastava, D. Srinivas, P. Ratnasamy, Fe-Zn double-metal cyanide
complexes as novel, solid transesterification catalysts, J. Catal. 241 (2006)
[
[
[
40] M. Orazov, M.E. Davis, Catalysis by framework zinc in silica-based molecular
3
4–44.
sieves, Chem. Sci. 7 (2016) 2264–2274.
14] P.S. Sreeprasanth, R. Srivastava, D. Srinivas, P. Ratnasamy, Hydrophobic, solid
acid catalyts for production of biofuels and lubricants, Appl. Catal. A 314
41] T.E. Müller, K.C. Hultzsch, M. Yus, F. Foubelo, M. Tada, Hydroamination: direct
addition of amines to alkenes and alkynes, Chem. Rev. 108 (2008) 3795–3892.
42] J. Penzien, T.E. Müller, J.A. Lercher, Hydroamination of 6-aminohex-1-yne over
zinc based homogeneous and zeolite catalysts, Micropor. Mesopor. Mater. 48
(
2006) 148–159.
15] J.K. Satyarthi, D. Srinivas, P. Ratnasamy, Influence of surface hydrophobicity on
the esterification of fatty acids over solid catalysts, Energy Fuels 24 (2010)
(
2001) 285–291.
2
154–2161.
[
43] G.R. Bamwenda, S. Tsubota, T. Nakamura, M. Haruta, The influence of the
preparation methods on the catalytic activity of platinum and gold supported
16] J.K. Satyarthi, S. Radhakrishnan, D. Srinivas, Factors influencing the kinetics of
esterification of fatty acids over solid acid catalysts, Energy Fuels 25 (2011)
2
on TiO for CO oxidation, Catal. Lett. 44 (1997) 83–87.
4
106–4112.
[
44] S.J. Tauster, S.C. Fung, R.T.K. Baker, J.A. Horsley, Strong interactions in
supported-metal catalysts, Science 211 (1981) 1121–1125.
17] M. Kotwal, S.S. Deshpande, D. Srinivas, Esterification of fatty acids with
glycerol over Fe-Zn double-metal cyanide catalyst, Catal. Commun. 12 (2011)
[
45] P.D.L. Mercera, J.G. van Ommen, E.B.M. Doesburg, A.J. Burggraaf, J.R.H. Ross,
Zirconia as a support for catalysts Influence of additives on the thermal
stability of the porous texture of monoclinic zirconia, Appl. Catal. 71 (1991)
1
302–1306.
18] A. Peeters, P. Valvekens, F. Vermoortele, R. Ameloot, C. Kirschhock, D. De Vos,
Lewis acid double metal cyanide catalysts for hydroamination of
phenylacetylene, Chem. Commun. 47 (2011) 4114–4116.
19] A. Peeters, P. Valvekens, R. Ameloot, G. Sankar, C.E.A. Kirschhock, D. De Vos, Zn-
Co double metal cyanides as heterogeneous catalysts for hydroamination: a
structure-activity relationship, ACS Catal. 3 (2013) 597–607.
20] M.V. Patil, M.K. Yadav, R.V. Jasra, Prins condensation for synthesis of nopol
from b-pinene and paraformaldehyde on novel Fe–Zn double metal cyanide
solid acid catalyst, J. Mol. Catal. A Chem. 273 (2007) 39–47.
363–391.
[
[
46] G. Trott, P.K. Saini, C.K. Williams, Catalysts for CO /epoxide ring-opening
2
copolymerization, Phil. Trans. R. Soc. A 374 (2016) 20150085.
47] S. Lee, S.T. Baek, K. Anas, C.-S. Ha, D.-W. Park, J.W. Lee, I. Kim, Tuning of activity,
induction period and polymer properties of double metal cyanide catalyzed
ring-opening polymerizations of propylene oxide by using quaternary
ammonium salts, Polymer 48 (2007) 4361–4367.