Well-Defined Noble Metal Single Sites in Zeolites as an Alternative to Catalysis by Insoluble Metal Salts

Article abstract of DOI:10.1021/jacs.5b07304

Insoluble precious metal chlorides in polymeric form (i.e., PtCl₂, PdCl₂, AuCl, RhCl₃) are commonly used as catalysts for a plethora of organic reactions in solution. Here we show that only the minor soluble fraction of these precious metal chlorides (typically 5-30%) is catalytically active for the hydroamination, hydroalkoxylation, hydrosilylation, and cycloisomerization of alkynes and alkenes, and that the resting insoluble metal is catalytically useless. To circumvent this waste of precious metal and follow a rational design, we generate here well-dispersed Pt(II) and Pd(II) single sites on zeolite Y, with an exquisite control of the Lewis acidity, to catalyze different hydroaddition reactions to alkynes and alkenes with up to 10⁴ catalytic cycles (at least 2 orders of magnitude superior to precious metal chlorides) and with high isolated yields (82-99%, >15 examples).

Full text of DOI:10.1021/jacs.5b07304

Article
pubs.acs.org/JACS
Well-Defined Noble Metal Single Sites in Zeolites as an Alternative to
Catalysis by Insoluble Metal Salts
Paula Rubio-Marques,† Miguel A. Rivero-Crespo,† Antonio Leyva-Perez,*,† and Avelino Corma*,†,‡
́
́
†Instituto de Tecnología Química, Universidad Politec
Naranjos s/n, 46022 Valencia, Spain
́
nica de Valencia−Consejo Superior de Investigaciones Científicas, Avda. de los
‡King Fahd University of Petroleum and Minerals, Dhahran, 31261 Saudi Arabia
S
* Supporting Information
ABSTRACT: Insoluble precious metal chlorides in polymeric
form (i.e., PtCl2, PdCl2, AuCl, RhCl3) are commonly used as
catalysts for a plethora of organic reactions in solution. Here
we show that only the minor soluble fraction of these precious
metal chlorides (typically 5−30%) is catalytically active for the
hydroamination, hydroalkoxylation, hydrosilylation, and cyclo-
isomerization of alkynes and alkenes, and that the resting
insoluble metal is catalytically useless. To circumvent this
waste of precious metal and follow a rational design, we
generate here well-dispersed Pt(II) and Pd(II) single sites on
zeolite Y, with an exquisite control of the Lewis acidity, to
catalyze different hydroaddition reactions to alkynes and
alkenes with up to 104 catalytic cycles (at least 2 orders of magnitude superior to precious metal chlorides) and with high isolated
yields (82−99%, >15 examples).
INTRODUCTION
RESULTS AND DISCUSSION
■
■
Correlation between the Solubility of the Precious
Catalysis by late transition metals is a cornerstone of modern
organic synthesis, and for instance, the polymeric precious
metal chlorides PtCl2, PdCl2, AuCl3, and RhCl3 (0.5−5 mol %
respect to reactants) are the catalysts of choice for many
hydroaddition reactions to alkenes and alkynes (see Supporting
Information, Table S1),1−3 with industrial implementation in
some cases.4 Typically, these reactions are carried out in
common organic solvents, despite a significant part of the
precious metal chloride remaining at the bottom of the flask,
apparently unaltered. This fact makes one wonder if the metal
chloride acts as a homogeneous or heterogeneous catalyst, a
nontrivial issue considering the high price of precious metals
and, furthermore, the potential mechanistic implications for
these reactions.
Metal Chloride and Its Catalytic Activity for the
Hydroaddition Reaction to Alkynes and Alkenes. Figure
1 shows the solubility of PtCl2, PdCl2, AuCl, and RhCl3·nH2O
(0.1 mmol, 20−30 mg) with some representative organic
compounds in toluene (5 mol % in Pt, 0.25 M solution), after 2
h at 80 °C.
With minor variations, the amount of precious metal chloride
dissolved in aromatic, ether, nitrile, silane, alkene, alcohol,
acetate, and alkyne solutions is always <20%, and only carbon
monoxide, the strong chelating ligand 1,5-cyclooctadiene (1,5-
COD), and the primary alkyl amine cyclohexylamine are able to
dissolve >40% of the starting precious metal, at the expense of
forming the well-known metal complexes MCl2(L)2 (L = 1,5-
COD, cHex-NH2) for the latter.6
Here we show that the catalytically active species in a variety
of hydroaddition reactions to alkynes and alkenes is a minor
soluble fraction (<5−30%) of the starting precious metal
chloride. With the exact nature of the catalytically active species
in hand, we design an atomically efficient Pt(II) catalyst based
on well-dispersed precious metal single sites5 on zeolite Y, after
fine-tuning the Lewis acidity of Pt(II) with the zeolite
framework. This “ligand-free” Pt(II) (or Pd(II)) solid catalyst
is, at least, 2 orders of magnitude more active than the state-of-
the art precious metal salt catalysts, not only for the
hydroamination of alkynes but also for a variety of hydro-
addition reactions to alkynes and alkenes. Besides that, the
catalyst can be recovered and reused at least 10 times.
Figure 2 shows that the intramolecular hydroamination of o-
(phenylacetylen)aniline 1 in the presence of PtCl2 (0.5 mol %)2
proceeds with the same initial reaction rate, final yield, and
selectivity to product 2 regardless of the reaction mixture being
filtered after 3 min (conversion 2%) or not, despite most of the
solid precious salt remaining undissolved at the bottom of the
flask. This effect also occurs for other precious metal chlorides
(PdCl2, AuCl, and PtCl4), other solvents (dichloromethane and
1,4-dioxane), and other hydroaddition reactions to alkynes and
alkenes (hydroalkoxylation of alkynes, cycloisomerization of
Received: July 23, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.5b07304
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
in o-xylene], despite the latter requiring additional synthetic
steps and being more expensive than any Pt chloride. We
wondered here if the treatment of PtCl2 with the alkyne or the
silane, just for a few minutes, could generate the active species
in solution, as it occurs for the hydroamination, hydro-
alkoxylation, and cycloisomerization reactions, thus circum-
venting the use of Speier’s or Karstedt’s catalysts. Compara-
tively, PtCl2 has been obviated as a catalyst for the
hydrosilylation of alkynes although it is an easy-to-handle
solid. The results showed that treatment of a 5 mol % of PtCl2
with triethylsilane in toluene, for just 3 min, gives a solution
that catalyzes the hydrosilylation of 1-dodecyne with better
initial catalytic activity than the insoluble PtCl2 or H2PtCl6
solids, without the typical induction time found in hydro-
sylilation reactions (Figure S8). Notice that this solution
contains only 4% of the starting Pt (0.2 mol % of Pt respect to
the reactants) according to ICP-AES and differential gravim-
etry. The same solubility effect was found for H2PtCl6 and
when the reaction was carried out in other solvents (dichloro-
methane and tetrahydrofuran), when alkenes were used as
substrates (Figure S9), and when RhCl3 was used as the
precious metal salt catalyst (Figure S10). These results suggest
that the formation of catalytically active metal species in
solution for the hydrosilylation reaction is general regardless of
the substrate or metal used.
Figure 1. Percentage of precious metal dissolved from PtCl2, PdCl2,
AuCl, or RhCl3·nH2O (5 mol %) in 0.25 M toluene solutions,
calculated in two ways: (1) by inductively coupled plasma atomic
emission spectroscopy (ICP-AES) of the filtered solution and (2) by
differential gravimetry of the MCln solid. The results by both methods
agree well within an estimated error of 2% after an average of two
experiments. In most of the cases, the amount of metal dissolved in
toluene after 3 min was nearly the same after 2 h.
The amount of metal dissolved for the different hydro-
addition reactions in toluene versus the amount of metal
expected to be dissolved for the substrates (see Figure 1 and
Figure S11) correlates well; thus, it seems that the solubility of
the precious metal chloride ultimately controls the catalytic
activity for each reaction. For instance, Pt analysis in solution
during the hydroamination of 1 gives 30% of soluble Pt, which
fits the 27% of PtCl2 dissolved with 1-hexyne in toluene (see
Figure 1).
The alkyne or alkene does not always exclusively control the
metal solubility since the other coupling partner, or even the
product, can play a major role in some cases. For instance, the
cycloisomerization of a 1,6-diene proceeds with a clear
induction time that corresponds, according to differential
gravimetric measurements, to the time needed to dissolve the
PdCl2 catalyst (Figure S7). The sigmoidal curve comes from
better dissolution of the metal in the presence of the cyclic
product, probably by a preferred coordination with the 1,3-
unsaturated ester. In accordance, the addition of increasing
amounts of product progressively eliminates the induction time,
and when so, the final filtration of the solid catalyst has no
influence on the reaction outcome. Another remarkable
example where the dissolution of the precious metal salt
seems to not depend on the alkyne or alkene is the
hydrosilylation reaction, since the solubility value of PtCl2
with Et3SiH (4%) matches the Pt dissolved during the PtCl2-
catalyzed hydrosilylation of alkynes (see Figure S11); if the
PtCl2 catalyst is treated with Et3SiH in toluene for 3 min and
filtered, and then the alkyne is added to the filtrate, the
hydrosilylation reaction proceeds exactly with the same rate and
selectivity. Thus, we can say that the silane and not the alkyne is
the main species responsible for the formation of the
catalytically active soluble Pt species, which is consistent with
the redox nature of the reaction.
Figure 2. Yield−time plot for the hydroamination of o-
(phenylacetylen)aniline 1 catalyzed by PtCl2 under the indicated
reaction conditions. Each point is an average of three different
measurements, and bars represent an estimated error of 5%. The top
left inset shows a photograph of the PtCl2 remaining, and the bottom
right inset illustrates the equilibrium-controlled dissolution of PtCl2
with the substrates.
propargyl acetates, and cycloisomerization of 1,6-dienes, see
Supporting Information Figures S1−S10). Notice that the small
amounts of PtCl2 or PdCl2 that can dissolve in neat toluene
(∼3% for Pt according to ICP-AES and the gravimetric test)
are catalytically competent for the hydroamination of 1 (Figure
S3).
The solubility effect is not restricted to Lewis-catalyzed
reactions but also occurs for reactions following an oxidative-
addition reductive-elimination mechanism7 such as the hydro-
silylation of alkynes and alkenes, which is perhaps the most
important process in organic synthesis catalyzed by Pt salts.4,8
The first Pt catalyst reported for the hydrosilylation of C−C
multiple bonds was Speier’s catalyst H2PtCl6, but the difficult
handling and the low solubility of this highly hygroscopic salt
lead to the rapid implementation in organic synthesis of
Karstedt’s catalyst [Pt2(divinyltetramethyldisiloxane)3 solution
Increasing the Solubility of PtCl2 with Carbon
Monoxide. Since the catalytic effect is directly proportional
to the metal dissolved, it should be possible to increase the
B
DOI: 10.1021/jacs.5b07304
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
conversion level by increasing the solubility of the metal during
the reaction.
linearly proportional (Figure S12). For instance, when 26.5
0.5 mg of PtCl2 is added as a catalyst (5 mol %), the ICP-AES
analysis shows 27% of the starting Pt in solution, and the
gravimetric method gives 5.9 0.5 mg of a brown solid residue
for the filtrates and 20.9 0.5 mg for the resting PtCl2 solid,
after exhaustive drying. At the end of the reaction (conversion
>95%), the recovered PtCl2 solid still accounts for ∼75 wt % of
that initially added, so the dissolved Pt remains in solution even
when no more 1 is present, which could be due to coordination
with the indole product. Indeed, 22% of Pt is dissolved after
treating a dispersion of PtCl2 in toluene with indole 2. These
results together suggest that Pt is in equilibrium with the alkyne
and/or the indole, as shown in Figure 2. Accordingly, Figure 4
In principle, a simple way to increase the catalytic activity of
the metal chloride would consist of the use of external agents
that dissolve more metal without poisoning the catalyst. For
instance, CO dissolves nearly double PtCl2 as compared to the
alkyne in toluene (∼50%, see Figure 1); thus, a CO atmosphere
may be beneficial for the hydroamination of alkynes. This is
indeed what occurs,2c and Figure 3 (top left) shows that the
Figure 4. Yield−time plots for successive runs of the hydroamination
of o-(phenylacetylen)aniline 1 catalyzed by the recovered PtCl2 (0.5
mol %) under the reaction conditions indicated in Figure 2. Bars
represent an estimated error of 5%.
Figure 3. Top left: Yield−time plot for the hydroamination of o-
(phenylacetylen)aniline 1 catalyzed by PtCl2 (1.5 mol %) under an
atmosphere of carbon monoxide. Top right: 13C NMR of the extracts
with an atmosphere of isotopically labeled 13CO in toluene-d8, before
(down) and after (up) adding isotopically labeled 2-13C-phenyl-
acetyene. Bottom left: Increasing of the initial rate with the amount of
soluble Pt, in the presence or absence of a carbon monoxide
atmosphere. Bottom right: Yield−time plot after extracting PtCl2 in
toluene with an atmosphere of carbon monoxide during 30 min,
filtering, and then adding the alkyne. Bars in all graphics represent an
estimated error of 5%.
shows that the amount of solid PtCl2 recovered after 4
successive hydroamination reactions and, also, the initial rates
systematically decrease ∼1/4 with each use (see also Figures
S12 and S13), which confirms that the hydroamination of 1
follows an equilibrium with 20−30% of Pt dissolved along the
reaction.
Determination of the Soluble Species in the PtCl2-
Catalyzed Hydroamination of Alkynes. In order to
determine the soluble Pt active species for the hydroamination
of 1 (5 mol % PtCl2), electrospray ionization high-resolution
mass spectrum (ESI-HRMS) of the filtrates at 7% conversion
time were carried out (Figure S14). The results unequivocally
showed that the single Pt species present in solution is the
PtCl2(1)2 complex, according to the isotopic pattern and the
expected fragmentations. Treatment of this complex in solution
with a stoichiometric amount of the bis-alkene COD gave the
PtCl2(COD) complex after ligand displacement (Figure S15),
and a gravimetric test of the filtered solution with AgNO3
confirmed the presence of two chloride atoms per Pt atom.
These results indicate that 1 dislodges monomeric PtCl2 units
from the insoluble chloride during reaction and that these
species are catalytically competent for the hydroamination
reaction.
PtCl4 is also considered an active and selective catalyst for
the Lewis-catalyzed intramolecular hydroamination of alkynes
under homogeneous reaction conditions,1 since the high
catalytic efficiency of Pt for this reaction is ascribed to its
high alkynophilicity, which stemmed from relativistic effects,1f
together with the expected high Lewis acidity of Pt4+. When
PtCl4 (0.5 mol %) is used as a catalyst instead of PtCl2 under
the reaction conditions indicated in Figure 2, and the solution is
filtered after 3 min of reaction, the kinetic profile is similar to
that when the insoluble PtCl4 is still present in the reaction
catalytic activity is similar regardless of the presence or absence
of the resting PtCl2. Control experiments showed that the same
Pt is extracted with CO in the presence or absence of the
alkyne.
13C NMR measurements of the toluene extracts with an
isotopically labeled 13CO atmosphere confirm the formation of
CO−Pt complexes (∼150 ppm, 33% active 195Pt, J = 1567.4
Hz) and that the CO molecules coordinated to Pt are rapidly
displaced by the alkyne (disappearance of the CO−Pt signal,
new 13C signal corresponding to the alkyne at ∼80 ppm, see
Figure 3 top right). With these results in hand, we can suggest
that the enhanced rate observed under an atmosphere of CO
for some PtCl2-catalyzed reactions (see Table S1) might be
explained by the enhanced solubility of Pt beyond any
electronic considerations. This proposal is supported by the
good correlation found during the hydroamination of 1
between catalytic activity versus solubility of Pt in solution,
regardless of a CO atmosphere being set or not, and also by the
fact that the reaction proceeds well with the Pt extracted under
an atmosphere of carbon monoxide before adding the alkyne
(see bottom Figure 3).
Analysis of Pt in solution for different amounts of starting
PtCl2 shows that the amount of Pt dissolved is systematically
1/4 of the originally added, and accordingly, the initial rate is
C
DOI: 10.1021/jacs.5b07304
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
medium. However, in contrast to PtCl2, an induction time of 15
min is found (Figure S4). If the amount of PtCl4 is decreased 3
times, the induction time lengthens to 4 h, and then the
hydroamination starts with an initial reaction rate 3 times lower.
Ultraviolet−visible spectroscopy (UV−vis) of the filtrates of the
PtCl4-catalyzed reaction does not show any bands during the
induction time, but a new band appears when the reaction
starts and proceeds, a band that corresponds to PtCl2 in
solution. These results, although not conclusive, strongly
suggest that the induction time found with PtCl4 corresponds
to the formation of PtCl2 in solution, so the same soluble
monatomic PtCl2−alkyne complex found for PtCl2 seems to be
the one catalyzing the reaction when starting with PtCl4 which
catalyzes the reaction.
With these data in hand, it seems that the key issue to
catalyze the hydroamination of 1 is just having discrete Pt(II)
species; however, the absence of catalytic activity for soluble
monatomic Pt chloride complexes, such as, for instance,
PtCl2(COD) or PtCl2(NH3)2 (conversion <5% under the
conditions in Figure 2, 9 h reaction time), indicates that the
limitation for the ligand-promoted complete solubility of Pt(II)
in toluene is the lack of catalytic activity for the hydroamination
of 1, at least for complexes having stronger ligands than 1, that
avoid ligand exchange in the Pt coordination sphere.9
Solid Catalysts with Isolated Pt(II) Based on a Zeolite.
Since the Pt catalyst for the hydroamination of 1 consists of
isolated Pt(II) atoms attached to weak ligands, we thought of
zeolite Y to introduce Pt(II).10 In principle this solid has some
properties that can help to achieve a relatively large amount of
catalytically active Pt(II) species on its surface: (1) Zeolite Y
has a very large surface area (∼800 m2 g−1) and exchange
capacity that enable high dispersion of the Pt(II) atoms.11 (2)
The oxygen atoms (the counteranions) on the zeolite walls act
as weak ligands. (3) The Lewis acidity of Pt(II) can be
modulated by varying the electronic density of the crystalline
framework. (4) Incorporation of Pt(II) by cationic exchange is
easy and nearly quantitative. (5) Potential generation of
mesopores in the zeolite may allow the reaction of relatively
large molecules.12,13 Thus, Pt(II)−NaY (∼1 wt % Pt according
to ICP−AES) was prepared by cationic exchange with an
aqueous solution of Pt(NH3)4(NO2)2 followed by aerobic
thermal elimination of the NH3 ligands at 100−400 °C (see
preparation and characterization details in Figures S16−S21),
preserving the microstructure and total surface area of the
zeolite.
The amount of NH3-free Pt(II) sites was determined by the
loss of NH3 in temperature-programmed infrared spectroscopy
and by the ratio of Pt(II) to Pt(0) atoms with X-ray
photoelectron spectroscopy (XPS), the results being double-
checked by low-temperature IR experiments with carbon
monoxide as a probe. It was found that the amount of Pt(II)
remaining after activation depends very much on the
decomposition conditions of the cationic amino complex. A
maximum number of NH3-free Pt(II) sites (77%, Table S2) are
obtained at a calcination temperature of 200 °C over 48 h,
which is in good agreement with the decomposition temper-
ature of the Pt(II) ammonia complex precursor (∼250 °C, see
thermogravimetric analysis in Figure S22). Larger calcination
times and/or higher temperatures decrease the amount of NH3-
free Pt(II) sites since further reduction to Pt(0) occurs, as
observed by dark-field scanning transmission electron micros-
copy (DF STEM) and diffuse reflectance ultraviolet−visible
spectroscopy (DR-UV−vis, Pt plasmon bands as a broad
downhill band covering from 230 to 700 nm, Figures S23 and
S24).
Figure 5 shows that the catalytic activity of Pt(II)−NaY for
the hydroamination of 1 (1 mol % Pt) correlates linearly with
Figure 5. Initial conversion rates for the hydroamination of 1, at 110
°C, catalyzed by different Pt(II)−Y zeolites (1 mol %) where the
number of NH3-free Pt(II) sites for NaY (right) or charge
compensating cation ionic radius (samples calcined at 200 °C for 48
h, left) has been varied. Error bars account for 5% uncertainty. The
electronic density of the zeolite framework is represented as the
Madelung potential curve; notice that the initial rate for the electron
rich CsY zeolite is lower than expected because significant reduction of
Pt(II) to Pt(0) occurs after calcination.
the number of NH3-free Pt(II) sites (see also Figure S25), with
a calculated TOF0 per NH3-free Pt(II) atom of 273 h−1. Notice
that the TOF0 with PtCl2 under the same reaction conditions is
66 h−1 (∼200 h−1 per dissolved Pt atom). This result indicates
that the supported Pt(II) atoms on NaY are intrinsically some
more active than dissolved PtCl2 monomers, in accordance with
the weaker electron donation of the NaY framework compared
to chloride anions.
It should be remarked that the Pt(II)−NaY samples calcined
at temperatures ≤200 °C do not show any Pt nanoparticles
before or after use, according to ultramicotome-cut DF STEM
and HR-TEM (up to the limit of detection of the 400 keV
instrument, ∼1 nm) and RD-UV−vis spectroscopy (since only
the band of Pt(II) oxide at ∼215 nm is detected, see Figures
S17 and S24). On the other hand, the samples calcined at
temperatures ≥300 °C lose their higher initial catalytic activity
during reaction due to further Pt aggregation under reaction
conditions, as confirmed by the increase of the number and
average diameter size of Pt nanoparticles formed during
calcination, from ∼3 nm in the fresh sample to ∼6 nm after
reaction, as also confirmed by the increase of the Pt(0) plasmon
band intensity in the sample after reaction (see Figures S23−
S25). Thus, we can conclude that the Pt(II) zeolite must be
prepared by calcining the exchanged zeolite at a temperature
and time sufficient to eliminate ammonia from the Pt complex
without reducing and agglomerating Pt. In our case, the optimal
catalytic activity for the hydroamination of 1 in toluene at 110
°C corresponds to a Pt(II)−NaY sample calcined at 200 °C for
48 h.
The possible leaching of catalytically active Pt species into
solution was evaluated for Pt(II)-NaY by the hot filtration test,
showing that the liquid phase accounts for 11% of the catalytic
activity after filtration of the solid catalyst at ∼35% conversion,
which is within the range of the noncatalyzed thermal
cyclization (blank experiment, 8% conversion, Figure S26).
ICP-AES analyses showed that <0.01% of Pt was present in the
liquid phase, and elemental analysis of the recovered Pt(II)-
NaY solid confirmed that Pt remains on the zeolite. These
results are in accordance with previous reports on the low level
of leaching in zeolite-supported containing noble metals when
toluene is the solvent.14 For the sake of comparison, we also
supported by impregnation the soluble PtCl2−alkyne species
D
DOI: 10.1021/jacs.5b07304
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
dislodged from PtCl2 on zeolite NaY, and we found a complete
leaching of the active Pt catalyst to the solution during the
hydroamination reaction. These results illustrate the stability of
the Pt(II) species supported on Pt(II)−NaY during reaction,
without losses of precious metal.
Other Pt-supported catalysts including Pt on carbon and Pt
supported on alumina or silica−alumina gave TOF values <10
h−1.
Influence of the Zeolite. The Lewis acidity of the Pt(II)
atoms bound to the oxygen atoms of the zeolite walls is
controlled by the electronic density of the zeolite network;15
thus, in principle we could increase the Lewis acidity of the
Pt(II) sites in electronically deficient zeolite frameworks.12
Figure 6 shows a slight shift of the peaks corresponding to the
Figure 7. Hydroadditions to alkynes catalyzed by Pt(II)−NaY.
a
Isolated yields, 1 mol% of Pt otherwise indicated.
Indicates
b
Pt(II)−NaY reused 3 times. Indicates Pt(II)−HY used as a catalyst.
Indicates Pd(II)−HY reused 8 times, TOF0 for the first use.
c
3−6, 9−12, and 14), amidoalkynes (7−8 and 13), alkynols
(15−16), and one diene (17) cyclize well in excellent isolated
yields with a TON and TOF0 up to 10 000 and 32 700 h−1,
respectively, the highest reported as far as we know for these
hydroaddition reactions.3 Particularly relevant are indoles 9−
12, which are key intermediates during the synthesis of the
widely used drug indomethacin and analogues (see Figure S28
for the complete syntheses). The Pt(II)−NaY can be recycled
up to 10 times with high yields after 24 h reaction time (Figure
S29 for kinetics). After each use, the zeolite was extracted with
dichloromethane under microsoxhlet conditions to give ∼5 wt
% of retained product, giving a mass balance >95%.
Notice that Pt−HY was not used as a general catalyst for the
hydroamination of alkynes since the Brønsted-catalyzed
hydration to form the ketone occurs at some extent, and this
problem could not be alleviated by further drying the zeolite
since, then, reduction of Pt(II) occurs. In the case of 1,6-dienes,
Pd−HY was prepared in the same way that Pt−HY was,
characterized (Figure S30) and used as a catalyst to give a good
yield of 17 after 8 reuses (Figure S31). In this case, the
expected primary asymmetric cyclized product17 gradually
isomerizes to the symmetric cycle by the action of the
Brønsted sites of the zeolite. This result demonstrates that the
synthesis of dispersed noble metal cations in zeolites for
hydroaddition reactions can be applied to other metals rather
than Pt.
Figure 6. Comparison of the acidity of Pt(II)−HY, Pt(II)−NaY, and
Pt(II)−KY by pyridine-desorption IR spectroscopy at 150 °C, under
vacuum, for samples previously calcined at 200 °C.
Pt(II) Lewis sites to higher wavenumber values (more acidic)
in the different Pt(II)-zeolites, in the order HY > NaY > KY,
using pyridine as a probe molecule together with infrared
spectroscopy measurements (for detailed temperature-pro-
grammed spectra see Figure S27), with a relatively similar
number of total Pt(II) Lewis sites. Thus, it seems that the acid
strength of the NH3-free Pt(II) sites increases after decreasing
the electron density of the zeolite by changing the charge-
compensating cation (from K+ to H+). Notice that Brønsted
sites are significant only in Pt(II)−HY (see Figure S27).
If the energy of the lower unoccupied molecular orbital
(LUMO) of Pt(II) is decreased by modifying the average
Madelung electronegativity of the zeolite,16 the catalytic activity
of Pt(II) in an orbital-controlled reaction such as the Lewis-
catalyzed hydroamination of alkynes should improve. Figure 5
shows that, indeed, the catalytic activity increases for
electronically deficient zeolites, with the maximum catalytic
activity found for zeolite Pt(II)−HY (TOF = 375 h−1). Blank
experiments showed that the zeolite HY without Pt is inactive
toward the hydroamination reaction.
Scope of Pt(II)− and Pd(II)−Zeolite Y in Hydroaddition
Reactions. Considering the hydroamination of 1 catalyzed by
Pt(II)−Y zeolite as a proof of concept for other hydroaddition
reactions, it could be expected that Pt(II)−Y would be a good
catalyst not only for other alkynes but also for other
nucleophiles, alkenes, and metals. Figure 7 shows the scope
for some Lewis-acid-catalyzed hydroadditions to alkenes and
alkynes, and indeed, different aromatic aminoalkynes (products
CONCLUSIONS
■
Finely dispersed Pt(II) or Pd(II) within zeolite Y circumvents
the low efficiency of precious metal chlorides in a variety of
hydroaddition reactions. Electrophilic alkyne activation has
enormous potential in the synthesis of heterocycles, and given
the importance of those in agrochemistry and pharma, the
results shown here should greatly help to make this versatile
chemistry more practical and bring it closer to applications.
E
DOI: 10.1021/jacs.5b07304
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
of stable amino Pt(II) chlorides is the reason why the intermolecular
hydroamination of alkynes does not proceed with PtCl2.
(10) Corma, A. J. Catal. 2003, 216, 298.
(11) de Graaf, J.; van Dillen, A. J.; de Jong, K. P.; Koningsberger, D.
C. J. Catal. 2001, 203, 307.
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.5b07304.
(12) Oudenhuijzen, M. K.; van Bokhoven, J. A.; Miller, J. T.;
Ramaker, D. E.; Koningsberger, D. C. J. Am. Chem. Soc. 2005, 127,
1530.
General information, experimental procedures, character-
ization of starting materials and products with NMR
spectra, and additional Figures S1−S31 with comments
(PDF)
́
(13) Cabrero-Antonino, J. R.; Leyva-Perez, A.; Corma, A. Angew.
Chem., Int. Ed. 2015, 54, 5658.
(14) (a) Dams, M.; Drijkoningen, L.; Pauwels, B.; Van Tendeloo, G.;
De Vos, D. E.; Jacobs, P. A. J. Catal. 2002, 209, 225. (b) Corma, A.;
Garcia, H.; Leyva, A. Appl. Catal., A 2002, 236, 179. (c) Corma, A.;
Garcia, H.; Leyva, A.; Primo, A. Appl. Catal., A 2003, 247, 41. (d) Li,
Y.; Liu, J. H.−C.; Witham, C. A.; Huang, W.; Marcus, M. A.; Fakra, S.
C.; Alayoglu, P.; Zhu, Z.; Thompson, C. M.; Arjun, A.; Lee, K.; Gross,
E.; Toste, F. D.; Somorjai, G. A. J. Am. Chem. Soc. 2011, 133, 13527.
(15) Corma, A.; Garcia, H.; Leyva, A. J. Catal. 2004, 225, 350.
(16) Dempsey, E. J. Phys. Chem. 1969, 73, 3660.
AUTHOR INFORMATION
■
Corresponding Authors
*anleyva@itq.upv.es
*acorma@itq.upv.es
Notes
The authors declare no competing financial interest.
(17) Widenhoefer, R. A. Acc. Chem. Res. 2002, 35, 905.
ACKNOWLEDGMENTS
■
A.L.-P. thanks ITQ for a contract. P.R.-M. and M.A.R.-C. thank
MEC for a FPU scholarship. We thank Dr. Angel Cantin for the
́
preparation of three starting materials. We also thank the
Electron Microscopy Service of the UPV for TEM measure-
ments. Financial support by Consolider-Ingenio 2010 (proyec-
to MULTICAT), Prometeo from Generalitat Valenciana
(PROMETEOII/2013/011), and “Severo Ochoa” program is
acknowledged.
REFERENCES
■
(1) Selected reviews: (a) Muller, T. E.; Beller, M. Chem. Rev. 1998,
98, 675. (b) Pohlki, F.; Doye, S. Chem. Soc. Rev. 2003, 32, 104.
̈
(c) Furstner, A.; Davies, P. W. Angew. Chem., Int. Ed. 2007, 46, 3410.
̈
(d) Chianese, A. R.; Lee, S. J.; Gagne,
2007, 46, 4042. (e) Wu, X.−F.; Anbarasan, P.; Neumann, H.; Beller,
M. Angew. Chem., Int. Ed. 2010, 49, 9047. (f) Leyva-Perez, A.; Corma,
́
M. R. Angew. Chem., Int. Ed.
́
A. Angew. Chem., Int. Ed. 2012, 51, 614. (g) Furstner, A. Acc. Chem.
Res. 2014, 47, 925.
̈
(2) Selected publications concerning PtCl2-catalyzed reactions:
(a) Chatani, N.; Furukawa, N.; Sakurai, H.; Murai, S. Organometallics
1996, 15, 901. (b) Furstner, A.; Szillat, H.; Gabor, B.; Mynott, R. J.
̈
Am. Chem. Soc. 1998, 120, 8305. (c) Furstner, A.; Davies, P. W. J. Am.
̈
Chem. Soc. 2005, 127, 15024. (d) For one of the most active
molecular ligand modified PtCl2-catalyzed systems, see: Carreras, J.;
Patil, M.; Thiel, W.; Alcarazo, M. J. Am. Chem. Soc. 2012, 134, 16753.
(3) Muller, T. E.; Hultzsch, K. C.; Yus, M.; Foubelo, F.; Tada, M.
̈
Chem. Rev. 2008, 108, 3795.
(4) (a) Brook, M. A. Silicon in Organic, Organometallic, and Polymer
Chemistry; Wiley Interscience: New York, 2000. (b) Marciniec, B.
Hydrosilylation: A Comprehensive Review on Recent Advances; Springer
Science: The Netherlands, 2008; Vol. 1.
(5) Witham, C. A.; Huang, W.; Tsung, C.−K.; Kuhn, J. N.; Somorjai,
G. A.; Toste, F. D. Nat. Chem. 2010, 2, 36.
(6) (a) Angelici, R. J. Reagents for transition metal complex and
organometallic syntheses. In Inorganic Synthesis; John Wiley & Sons
Inc.: New York, 1990; Vol. 28, p 346. (b) Unger, D. Y. M.; Molt, O.;
Schildknecht, C.; Munster, I.; Wagenblast, G.; Strassner, T.; Meyer, D.
̈
Angew. Chem., Int. Ed. 2010, 49, 10214.
(7) Harrod, J. F.; Chalk, A. J. Nature 1965, 205, 280.
(8) Speier, J. L.; Webster, J. A.; Barnes, G. H. J. Am. Chem. Soc. 1957,
79, 974.
(9) When we tried the intermolecular hydroamination between
cyclohexylamine and 1-octyne catalyzed by PtCl2, under the indicated
reaction conditions in Figure 2, no coupling products were found, but
only the transformation of the brown PtCl2 solid into a white
precipitate. This precipitate corresponds to the PtCl2(CyNH2)2
complex (see Figure 1), and this result suggests that the formation
F
DOI: 10.1021/jacs.5b07304
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX