Evaluation of the mechanism of heterogeneous hydrogenation of α,β-unsaturated carbonyl compounds via pairwise hydrogen addition

Article abstract of DOI:10.1021/cs500426a

Heterogeneous hydrogenation of α,β-unsaturated carbonyl compounds was addressed using the parahydrogen-induced polarization (PHIP) technique. PHIP effects were observed in hydrogenation of C = C bond of acrolein and crotonaldehyde over different supported metal catalysts, demonstrating the existence of a pairwise route of hydrogen addition to the substrate. Hydrogenation of acrolein over Pd-Sn/Al₂O₃, Pd-Sn/TiO ₂, Pd-Zn/TiO₂, and Pd/TiO₂ catalysts with parahydrogen also led to the polarization of the proton of CHO group of propanal. This was explained by C(O)-H bond dissociation which represents a side process on the catalyst surface. Formation of polarized cis- and trans-2-butenes was detected in hydrogenation of acrolein with parahydrogen over several Rh-based catalysts. This observation is made possible only due to the high NMR signal enhancement provided by PHIP. It was also found that hydrogenation of acetone and propanal with parahydrogen leads to polarized propane formation as a result of C-O bond hydrogenolysis.

Full text of DOI:10.1021/cs500426a

Research Article
pubs.acs.org/acscatalysis
Evaluation of the Mechanism of Heterogeneous Hydrogenation of
α,β-Unsaturated Carbonyl Compounds via Pairwise Hydrogen
Addition
Oleg G. Salnikov,†,‡ Kirill V. Kovtunov,*,†,‡ Danila A. Barskiy,†,‡ Alexander K. Khudorozhkov,‡,§
Elizaveta A. Inozemtseva,‡,§ Igor P. Prosvirin,§ Valery I. Bukhtiyarov,‡,§ and Igor V. Koptyug†,‡
†International Tomography Center, SB RAS, 3A Institutskaya St., Novosibirsk 630090, Russia
‡Novosibirsk State University, 2 Pirogova St., Novosibirsk 630090, Russia
§Boreskov Institute of Catalysis, SB RAS, 5 Acad. Lavrentiev Pr., Novosibirsk 630090, Russia
S
* Supporting Information
ABSTRACT: Heterogeneous hydrogenation of α,β-unsaturated carbonyl
compounds was addressed using the parahydrogen-induced polarization
(PHIP) technique. PHIP effects were observed in hydrogenation of C
C bond of acrolein and crotonaldehyde over different supported metal
catalysts, demonstrating the existence of a pairwise route of hydrogen
addition to the substrate. Hydrogenation of acrolein over Pd−Sn/Al2O3,
Pd−Sn/TiO2, Pd−Zn/TiO2, and Pd/TiO2 catalysts with parahydrogen
also led to the polarization of the proton of CHO group of propanal. This
was explained by C(O)−H bond dissociation which represents a side
process on the catalyst surface. Formation of polarized cis- and trans-2-
butenes was detected in hydrogenation of acrolein with parahydrogen
over several Rh-based catalysts. This observation is made possible only
due to the high NMR signal enhancement provided by PHIP. It was also
found that hydrogenation of acetone and propanal with parahydrogen
leads to polarized propane formation as a result of C−O bond hydrogenolysis.
KEYWORDS: parahydrogen-induced polarization (PHIP), heterogeneous hydrogenation, α,β-unsaturated carbonyl compounds,
reaction mechanism, pairwise hydrogen addition
processes is by far preferable for chemical industry. It is notable
that selective hydrogenation of unsaturated ketones to
unsaturated alcohols is hardly possible because in this case
the rate of CC bond hydrogenation is much higher than that
of the CO bond.3 Therefore, practically all such studies
address hydrogenation of unsaturated aldehydes.
In general, three major products can be formed upon
hydrogenation of an α,β-unsaturated aldehyde: saturated
aldehyde, unsaturated alcohol (which can isomerize to
saturated aldehyde), and saturated alcohol (Figure 1). In
addition, hydrocarbons can be produced as byproducts via
decarbonylation and C−O bond hydrogenolysis.3,8,9 Reaction
selectivity is affected by many different factors including the
nature of the catalyst, the structure of the substrate, and
reaction conditions.3,10 Nowadays, selective hydrogenation
catalysts commonly are supported metal nanoparticles. Differ-
ent metals can be used as an active component, and the
selectivity to unsaturated alcohol formation is known to
INTRODUCTION
■
Selective hydrogenation of unsaturated carbonyl compounds is
an industrially important reaction which is widely used in
production of pharmaceuticals, flavorings, and perfumes. In
most cases, the CC bonds in such compounds are more
readily hydrogenated than the CO bonds.1−5 Therefore,
selective hydrogenation of the CO bond in the presence of a
CC bond is a challenging task. It remains an important
problem in catalysis, and a lot of efforts are concentrated on
research in this field.2−5 Selective reduction of a CO bond is
a common reaction in fine organic synthesis. Metal hydrides
such as NaBH4 and LiAlH4 are usually used as reducing agents.
However, these reagents are not appropriate for industrial
processes as they are expensive, require rather complicated
after-treatments, and sometimes do not provide sufficient
selectivity.6 Another possibility for CO bond reduction is the
Meerwein−Ponndorf−Verley process which is highly selective
toward production of alcohols.7 However, its use in industrial
applications also encounters several problems such as poor
activity of the catalyst (aluminum isopropoxide) and
production of large amounts of waste. Therefore, selective
hydrogenation of a CO bond using heterogeneous catalytic
Received: March 31, 2014
Revised: May 13, 2014
© XXXX American Chemical Society
2022
dx.doi.org/10.1021/cs500426a | ACS Catal. 2014, 4, 2022−2028

ACS Catalysis
Research Article
heterogeneous reactions can be beneficial for modern
biomedical MRI as well.
Practically all heterogeneous PHIP studies reported to date
addressed hydrogenation of multiple C−C bonds in unsatu-
rated hydrocarbons,19 with the only exceptions being the
hydrogenation of the CC bond in methyl propiolate (HC
C−C(O)OCH3)30 and of the CC bond in acrylamide
(CH2CH−C(O)NH2).31 It would be highly desirable to
extend such studies to a much broader variety of substrates to
facilitate the production of biocompatible hyperpolarized
contrast agents as well as to aid in the investigation of the
mechanisms of a much broader range of important catalytic
reactions. To this end, we have chosen heterogeneous gas-
phase catalytic hydrogenation of α,β-unsaturated carbonyl
compounds, acrolein and crotonaldehyde. This choice makes
it possible to explore the feasibility of pairwise hydrogenation of
the CO bond, to compare the results with those for the C
C bond hydrogenation, and in addition to produce oxygen-
containing products which would be more suitable for bioMRI
applications as compared to hydrocarbons. In addition,
hydrogenation of α,β-unsaturated carbonyl compounds is
known to be accompanied by C−O bond hydrogenolysis,
and thus such studies could address the feasibility of extending
the PHIP technique to more complex catalytic processes.
Hydrogenation of saturated carbonyl compounds such as
acetone and propanal was addressed as well in order to further
study the selectivity toward alcohols formation and the
feasibility of PHIP by pairwise hydrogen addition to the C
O bond. The purpose of this work is thus to extend PHIP
studies toward a potentially much broader range of processes
and products while not, at this initial stage, going too far from
the hydrogenations of multiple C−C bonds that are known to
produce PHIP effects.
Figure 1. Scheme of possible reaction pathways of α,β-unsaturated
aldehydes hydrogenation.
decrease in the sequence Os > Ir > Pt > Ru > Rh > Pd.3,10
Bimetallic catalysts in which one of these metals is combined
with a more electropositive metal such as Sn, Ge, Fe, Co, Ni,
Zn, and Mn demonstrate higher selectivity of CO bond
hydrogenation.3,10 Another possibility to enhance selectivity is
to increase the metal particles size.3,4,10,11 The nature of the
support also has an important influence on the selectivity
toward CO bond hydrogenation.3,4 High-temperature treat-
ment of metal particles supported on reducible oxides such as
TiO2 or Nb2O5 in a hydrogen atmosphere leads to formation of
TiOx and NbOx species which may migrate onto the metal
surface. This effect, called SMSI (strong metal−support
interaction),12,13 as well as the use of electron-donating
supports (e.g., graphite) increase the selectivity toward
formation of unsaturated alcohols.2−4,10 The substrate structure
is also important; for instance, an alkyl substituent at the CC
bond leads to a steric hindrance which destabilizes adsorption
via the CC bond, making adsorption via the CO bond and
therefore its hydrogenation preferable.3,10 The situation is
further complicated by the fact that the mechanism of the
reaction is not fully established. Therefore, better under-
standing and control of the factors which impact the selectivity
of the catalytic systems require further studies of the reaction
mechanism.
EXPERIMENTAL SECTION
■
Catalysts Preparation and Characterization. Commer-
cially available rhodium(III) chloride was used as the precursor
for the preparation of rhodium catalysts. Rh(III) oxide was
prepared by thermal decomposition of RhCl3·3H2O at 800 °C
for 2 h. The powder was then washed with hot water until the
filtrate gave a negative reaction when tested for chloride ions
Parahydrogen-induced polarization (PHIP) is an efficient
tool for the studies of mechanisms of multiple C−C bond
hydrogenation with H2.14−18 It provides significant NMR signal
enhancements if addition of the parahydrogen (p-H2) molecule
(AgNO3 + Cl− = AgCl + NO3 ). Rh black was prepared by a
−
1
is molecular, or pairwise, i.e., if both H nuclei of p-H2 end up
slow reduction of RhCl3·3H2O aqueous solution with sodium
formate at 80 °C. The obtained black precipitate was filtered
and washed with hot water until the filtrate gave a negative
reaction when tested for chloride ions. Rhodium(III) hydroxide
Rh(OH)3 was prepared from RhCl3·3H2O solution by the
addition of sodium hydroxide at 80 °C until pH 10 was
reached. The obtained yellowish precipitate was filtered and
washed. Rhodium(III) nitrate solution (concentration ∼10 wt
%) used as a direct precursor for supported Rh catalysts
preparation was obtained by dissolution of Rh(OH)3 in nitric
acid (density 1.35 g/mL) with subsequent dilution by a
required amount of water. All supported rhodium catalysts
contained ∼1% Rh by weight. Palladium(II) nitrate solution
with 10.5 wt % palladium concentration was used for supported
palladium catalysts preparation. All supported palladium
catalysts contained ∼0.9% Pd by weight. In the case of
supported platinum catalysts, platinic acid H2[Pt(OH)6] was
prepared via premixing of H2PtCl6 × 6H2O aqueous solution
with sodium hydroxide with concentrations [OH−]/Pt = 12:1
at 80 °C for 6 h until pH 13 was reached. The resulting
yellowish precipitate was filtered and washed with significant
in the same product molecule. High sensitivity and unconven-
tional NMR signal patterns provided by PHIP allow one to
study the mechanisms of the catalytic hydrogenations and
observe unstable intermediates at low concentrations by
detecting the pairwise p-H2 addition. PHIP was originally
considered exclusively in homogeneous reactions catalyzed by
transition metal complexes. However, it was proven recently
that PHIP effects are also detectable in heterogeneous
hydrogenations19 catalyzed by supported metal nanopar-
ticles,20−22 immobilized metal complexes,23 bulk metals, metal
oxides,24 and noble metal-containing zeolite catalysts.25 There-
fore, the PHIP technique can be applied to study the
mechanisms of heterogeneous catalytic processes.
In addition to mechanistic studies in catalysis, significant
NMR signal enhancement is very important for applications of
the MRI technique.26−29 Substantial progress has been made in
this field based on PHIP produced in homogeneous hydro-
genations. However, polarized products intended for use as
contrast agents must be both catalyst-free and safe for in vivo
administration. Therefore, further progress with PHIP in
2023
dx.doi.org/10.1021/cs500426a | ACS Catal. 2014, 4, 2022−2028

ACS Catalysis
Research Article
amounts of glacial acetic acid up to pH 4.5 until the filtrate gave
a negative reaction when tested for chloride ions. Platinum(IV)
nitrate solution was obtained by dissolution of a weighed
amount of H2[Pt(OH)6] in nitric acid (density 1.35 g/mL) and
subsequent dilution of resulting solution with water until the 20
wt % Pt concentration was reached. All supported platinum
catalysts contained ∼0.8% Pt by weight.
Hydrogenation Reactions and NMR Experiments. All
hydrogenation experiments were carried out using commer-
cially available acrolein (Fluka, >95%), crotonaldehyde (Sigma-
Aldrich, >99.5%), acetone, propanal, and hydrogen. For PHIP
experiments, H2 was enriched with parahydrogen up to 50% by
passing it through FeO(OH) (Sigma-Aldrich) powder
maintained at liquid nitrogen temperature.20 After the enrich-
ment stage, hydrogen was collected in a gas cylinder. During
the experiments, it was bubbled through the flask containing
liquid substrate (acrolein, crotonaldehyde, acetone, or
propanal). In the case of PASADENA14 experiments (hydro-
genation in the high magnetic field of the NMR spectrometer),
the mixture of hydrogen and the substrate vapor was supplied
through a Teflon capillary to the bottom of a 10 mm NMR
tube where the catalyst was placed. For ALTADENA15
experiments (hydrogenation in Earth’s magnetic field with
subsequent transfer of reaction products to the NMR
spectrometer for analysis), the mixture of H2 and the substrate
vapor was supplied to the reactor comprising a straight piece of
a copper tube with the catalyst packed between two pieces of
fiberglass tissue, and then to an empty NMR tube. In all
experiments, the NMR tube residing in the NMR instrument
was heated to 100−130 °C to prevent condensation of the
reaction products and, in the case of PASADENA experiments,
to initiate the in situ hydrogenation. In ALTADENA experi-
ments, the copper reactor was heated with an electric tubular
furnace. All hydrogenations were carried out at atmospheric
pressure. The gas flow rate (from ∼1 to ∼17 mL/s) was
measured with a rotameter. The experimental conditions such
as the catalyst mass and the reactor temperature are
summarized in the Supporting Information SI (Table S1 in
the SI). The experimental setup is also presented schematically
in SI (Figure S3). For the liquid phase hydrogenation, the
mixture with a H2/acrolein vapor ratio of 8:1 was prepared in a
gas cylinder and supplied to the NMR tube containing 0.03 g of
Pd/TiO2 catalyst and approximately 5 mL of d8-toluene. NMR
spectra were recorded immediately after stopping the gas flow.
1H NMR spectra were recorded on a 300 MHz Bruker AV
300 NMR spectrometer. The 45° (PASADENA) and 90°
(ALTADENA) radiofrequency pulses were used for the NMR
signal detection.
Metal acetates Mx(CH3CO2)y·zH2O (M = Pd2+, Mn3+, Pb2+,
Zn2+, Ag+) from Sigma-Aldrich and Au(CH3CO2)3 (Alfa Aesar)
were used as precursors for the production of supported
bimetallic catalysts. The supports, γ-Al2O3 (Sasol, SBET = 215
m2/g), TiO2 (Hombifine, SBET = 70 m2/g), ZrO2 (TSP,
Ekaterinburg, Russia, SBET = 100 m2/g), and SiO2 (Sigma-
Aldrich, SBET = 350 m2/g), were ground to powder, calcined at
500 °C for 5 h, and dried at 120 °C for 2 h. Palladium(II)
nitrate, platinum(IV) nitrate, and rhodium(III) nitrate were
used as precursors for the preparation of supported
monometallic catalysts. Bimetallic acetic salts PdxMy(OAc)z
(M = Pb2+, Zn2+, Mn3+, Sn2+, Ag+, Au3+) used for the
preparation of bimetallic catalysts were prepared according to
the technique proposed by Akhmadullina et al.32 Catalyst
samples were prepared by wet impregnation of supports with
corresponding liquid precursor with subsequent drying at 120
°C for 3 h and calcination at 400 °C for 4 h. Then all supported
metal catalysts were reduced in hydrogen at 300 °C for 3 h. As
a result, bimetallic supported metal catalysts PdSn/Al2O3 (Pd ∼
1.0%, Sn ∼ 0.9%), PdSn/TiO2 (Pd ∼ 1.0%, Sn ∼ 0.9%), PdZn/
Al2O3 (Pd ∼ 1.0%, Zn ∼ 1.6%), PdZn/TiO2 (Pd ∼ 1.0%, Zn ∼
1.6%), PdPb/Al2O3 (Pd ∼ 1.0%, Pb ∼ 0.5%), PdPb/TiO2 (Pd
∼ 1.0%, Pb ∼ 0.5%), PdAg/Al2O3 (Pd ∼ 0.9%, Ag ∼ 0.9%),
PdAu/Al2O3 (Pd ∼ 0.9%, Au ∼ 1.7%), and PdMn3+/Al2O3 (Pd
∼ 0.9;Mn ∼ 0.5) were obtained.
All supported metal catalysts were studied with the XPS
technique by recording photoelectron spectra using SPECS
spectrometer with PHOIBOS-150-MCD-9 hemispherical en-
ergy analyzer (Al Kα, irradiation, hν = 1486.6 eV, 200 W). The
samples were supported onto double-sided conducting copper
Scotch tape. The binding energy (BE) scale was preliminarily
calibrated by the position of the peaks of Au 4f7/2 (BE = 84.0
eV) and Cu 2p3/2 (BE = 932.67 eV) core levels. The binding
energy of peaks was calibrated by the position of the C 1s peak
(BE = 284.8 eV) corresponding to the surface hydrocarbon-like
deposits (C−C and C−H bonds). Survey spectra were
recorded at a pass energy of the analyzer of 50 eV and the
narrow spectral regions at 20 eV. Rh and Pd foils, and also PdO
and Rh2O3 powders, were used as reference materials.
Figure S1(a) presents Pd 3d core-level spectra obtained for
Pd−Pb/TiO2 and Pd−Sn/TiO2 catalysts. The Pd 3d5/2 peak at
335.3 eV can be attributed to metallic Pd.33 Figure S1(b)
presents Pd 3d core-level spectra obtained for Pd−Zn/TiO2
and Pd−Au/Al2O3 catalysts, with Pd foil and PdO used as
references. For correct identification of the palladium chemical
state in the studied samples, we performed curve-fitting for Pd
3d spectra using XPS Peak 4.1 software.34 The Pd 3d5/2 peak at
335.2 eV can be attributed to metallic Pd (Pd0), while the peak
at 336.7 eV to Pd2+ (PdO).33 Figure S1(c) presents Pd 3d core-
level spectra obtained for Pd/TiO2 and Pd−Ag/Al2O3 catalysts.
The Pd 3d5/2 peak at 336.7 eV can be attributed to Pd2+
(PdO).33 Figure S2 presents Rh 3d5/2 core-level spectra
obtained for the Rh/TiO2 catalyst and rhodium foil and
Rh2O3 as the references. The Rh 3d5/2 peak at 308.2 eV can be
attributed to metallic Rh (Rh0) and the peak at 308.7 eV to
Rh3+ (Rh2O3).33
RESULTS AND DISCUSSION
■
Acrolein was hydrogenated with p-H2 over several different
supported Pd- and Pt-based catalysts (Pt/Al2O3, Pt/SiO2, Pt/C,
Pd/Al2O3, Pd/SiO2, Pd/TiO2, Pd−Sn/Al2O3, Pd−Sn/TiO2,
Pd−Zn/Al2O3, Pd−Zn/TiO2, Pd−Pb/Al2O3, Pd−Pb/TiO2,
Pd−Ag/Al2O3, Pd−Au/Al2O3, Pd−Mn3+/Al2O3, Pt−Y−Ce−
Zr/Al2O3) using the PASADENA14 technique. In all cases, the
polarized NMR signals of CH3 and CH2 protons of propanal
were observed, with some representative examples shown in
Figure 2. The Pd−Sn/Al2O3 catalyst was also tested in
ALTADENA15 experiments, and the same NMR signals
exhibited polarization (see Figure S4). The observation of
PHIP effects shows that H2 addition upon hydrogenation of the
CC bond of acrolein can be pairwise. It is notable that the
intensities of PHIP patterns were much higher for the Pd-based
catalysts than for the Pt-based catalysts. The main reason is the
higher activity of the former in hydrogenation of the CC
bond of α,β-unsaturated carbonyl compounds, which is
expected from both experimental and theoretical research.3,10
Moreover, in the case of supported palladium catalysts, Pd2+
and Pd0 show similar activities in terms of the pairwise
2024
dx.doi.org/10.1021/cs500426a | ACS Catal. 2014, 4, 2022−2028

ACS Catalysis
Research Article
1
Figure 2. H NMR PASADENA spectra of CH3 and CH2 protons of
propanal recorded in situ during hydrogenation of acrolein over Pt/
Al2O3, Pd/SiO2, and Pd/TiO2 catalysts.
hydrogen addition, which may be explained by reduction of
Pd2+ to Pd0 during the highly exothermic hydrogenation
processes (Figure S1 in the SI). The hydrogenation of the C
O bond is hardly possible with many catalysts, and only for Pd/
Al2O3, Pd/TiO2, and Pd−Zn/TiO2 catalysts the low-intensity
nonpolarized NMR signals of propanol were detected.
Interestingly, in the case of Pd−Sn/Al2O3, Pd−Sn/TiO2, Pd−
Zn/TiO2, and Pd/TiO2 catalysts, in addition to the CH3 and
CH2 protons, the proton of the CHO group of propanal
exhibited the PHIP effect (Figure 3b). A possible mechanism
which can lead to polarization of the CHO proton is presented
in Figure 3a. The crucial step is the formation of surface Pd−
C(O)−CHCH2 species which are hydrogenated with p-H2
via the pairwise route. The observation of Pt−C(O)CH3
intermediates in ethanol adsorption on Pt has been reported.35
Moreover, the formation of Ti−OC−CH3 species was
theoretically predicted in acetaldehyde adsorption on TiO2.36
Therefore, we assume that the C(O)−H bond can be broken
on the metal as well as on the support surface as shown in
Figure 3a. Moreover, in this case, the hydrogenation reaction
can proceed on the interface between the metal and the
support.37 The process leading to CHO-proton polarization is
the minor route of the reaction, but strong signal enhancement
which is produced by PHIP allows us to detect it.
Figure 3. Proposed reaction mechanism which explains the
observation of polarization on the CHO proton of propanal (a) and
1H NMR spectrum acquired in hydrogenation of acrolein with p-H2
over Pd−Sn/Al2O3 (b).
For the Rh-based catalysts, the performance was significantly
different from that of Pd- and Pt-based catalysts. In the case of
ALTADENA experiments with Rh/Al2O3 catalyst and
PASADENA experiments with Rh/Al2O3, Rh/TiO2, Rh2O3,
and bulk Rh metal catalysts, in addition to the polarized
propanal signals, two other strongly polarized NMR signals
were observed at 5.5 and 1.6 ppm (Figure 4). The chemical
shifts along with the absence of any additional polarized lines in
the spectra imply the formation of a mixture of cis- and trans-2-
butenes. From the equilibrium NMR spectra, their yield was
estimated as <4%. For an additional direct verification of 2-
butene formation in these experiments, we compared the
PASADENA experiment for acrolein with that for 2-butyne. It
was shown that the shapes of PASADENA-polarized lines and
1
Figure 4. Scheme of acrolein hydrogenation over Rh/Al2O3 (a). H
NMR ALTADENA spectra acquired during hydrogenation of acrolein
over Rh/Al2O3 with gas flow stopped (b) and with flowing gas (c).
1
their positions (5.5 and 1.6 ppm) in the H NMR spectra are
the same for the two experiments (Figure S5 in the SI),
confirming beyond a doubt that 2-butenes are formed in both
cases.
It is well-known that supported Rh nanoparticles are able to
decarbonylate aldehydes.38,39 In the case of unsaturated
aldehyde, acrolein, breaking of the C−C(O) bond should
release a vinyl species CH2CH. Other conceivable leaving
groups that may be formed at the Rh surface include those with
a larger number of hydrogen atoms, e.g., C2H5, CH3CH or
CH2CH2, an isomer of the vinyl CH3C, and CH2C or HC
CH.39 Moreover, Barteau and Brown reported that vinyl and
2025
dx.doi.org/10.1021/cs500426a | ACS Catal. 2014, 4, 2022−2028

ACS Catalysis
Research Article
acetylene are the most likely candidates for ligands that are
eliminated in acrolein decarbonylation.39 However, adsorbed
acetylene undergoes hydrogenation to ethylene and ethylidyne
species on the Rh surface.40 The formation of ethylene and
ethane in acrolein hydrogenation was successfully detected in
our experiments with several heterogeneous catalysts including
the Rh-based ones (Rh/Al2O3, Rh2O3, Rh, Pd/Al2O3, Pt/Al2O3,
Pd−Sn/Al2O3, Pd−Sn/TiO2, Pd/TiO2, Pd−Ag/Al2O3, Pd−
Au/Al2O3, Pd−Zn/TiO2; Figure 4). Therefore, one of the
plausible explanations of polarized 2-butene formation during
acrolein hydrogenation over Rh nanoparticles is the oligome-
rization of vinyl species with subsequent addition of p-H2 in a
pairwise manner. While 2-butenes are minor products of the
reaction, the strong signal enhancement allows one to observe
their formation. Our observations once again confirm the fact
that PHIP is a very informative method for studying the
mechanisms of both major and minor routes of hydrogenation
reactions. The observation of PHIP effects in reaction products
in many cases can provide valuable information about the
mechanism even if the structures of intermediates are not
known, as in our experiments with 2-butene formation during
acrolein hydrogenation.
Figure 5. Scheme of crotonaldehyde hydrogenation over Pd/TiO2 (a).
1H NMR PASADENA spectra acquired during hydrogenation of
crotonaldehyde over Pd/TiO2 with gas flow stopped (b) and with
flowing gas (c).
Next, liquid phase heterogeneous hydrogenation of acrolein
was performed by bubbling p-H2 through the solution of
acrolein in toluene-d8 in the presence of the solid catalyst, and
the pairwise hydrogen addition to the CC bond could be
detected, similar to the gas-phase experiments. In PASADENA
experiments with the Pd/TiO2 catalyst, we observed polarized
NMR signals of CH3 and CH2 groups of propanal (see Figure
S6 in the SI). However, intensities of the polarized lines were
lower than for gas-phase hydrogenation due to the low activity
of the catalyst in the liquid phase.
TiO2 catalysts, the polarized NMR signals of propane were
observed. Propane is formed as a result of decarbonylation,
which is a common side process in the hydrogenation of
aldehydes on metals.3,8,37
In our experiments with acrolein and crotonaldehyde, only
protons added to CC bonds were polarized. Therefore,
substrates possessing only the CO but not the CC group
were used in the next attempts. Heterogeneous hydrogenation
of acetone vapor over Pt/Al2O3, Pt/TiO2, Pd/Al2O3, Pd/TiO2,
Rh/TiO2, Rh/SiO2, and Pd−Sn/Al2O3 catalysts yielded
For more substituted carbonyl compounds, the selectivity
toward CO bond hydrogenation is known to be higher.3,10
Therefore, in order to observe CO bond hydrogenation,
such compounds as crotonaldehyde, prenal (3-methylbut-2-
enal), and cinnamaldehyde can be more suitable than acrolein.
For prenal and cinnamaldehyde, only liquid-phase experiments
can be efficiently performed because of their low volatility.
Therefore, crotonaldehyde was used next as a substrate for
hydrogenation in the gas phase over Pt/Al2O3, Pt/TiO2, Pt/
SiO2, Pt/ZrO2, Pd/TiO2, Pd/SiO2, Pd/ZrO2, Rh/Al2O3, Rh/
TiO2, Pd−Sn/Al2O3, and Pd−Sn/TiO2 catalysts using the
PASADENA protocol. Polarized NMR signals of CH2 groups
of butanal were observed in all cases. This, without a doubt,
confirms the existence of a reaction pathway where the CC
double bond may add hydrogen molecularly and, importantly,
with the preservation of the spin coherence of two protons
originating from the same p-H2 molecule. Importantly, when
Pd/TiO2, Pd/ZrO2, and Pd−Sn/Al2O3 catalysts were used for
crotonaldehyde gas-phase heterogeneous hydrogenation, not
only the CC but also the CO bond was hydrogenated
(Figure 5). It should be noted that the CO bond reacts only
after the CC bond hydrogenation because only signals of
butanal and butanol, but not but-2-en-1-ol, were observed in
the 1H NMR spectra. Moreover, butanol is formed in significant
quantities only at the low flow rates of a gas mixture through
the catalyst layer when conversion is sufficiently high. However,
slow gas flow makes it difficult to observe PHIP effects because
of the relatively fast relaxation of polarization (Figure 5b).
1
isopropanol, but in all cases its H NMR signals did not
exhibit any PHIP effects. At the same time, the Rh/TiO2 and
Rh/Al2O3 catalysts in addition produced polarized propane
(Figure S7). A similar situation was observed for propanal
hydrogenation catalyzed by Pt/TiO2 and Pd/TiO2. Propanol
and propane were formed in the reaction, while PHIP was
observed only for the CH3 and CH2 groups of propane (Figure
S8). Propane could be produced from the corresponding
alcohol (isopropanol or propanol) by dehydration/hydro-
genation processes. In this case, hydrogenation of propanol
with p-H2 under the same conditions should lead to PHIP
observation. However, when propanol was used in heteroge-
neous hydrogenation with p-H2, no reaction products were
detected. Thus, propane is formed directly from the carbonyl
compounds, most probably as a result of C−O bond
hydrogenolysis.
The fact that PHIP effects corresponding to pairwise
hydrogen addition to CO bond of unsaturated and saturated
aldehydes were not observed could be the result of a specific
reaction mechanism. In a recent publication of Kennedy et al.,37
the authors studied hydrogenation of crotonaldehyde, crotyl
alcohol, and butanal over Pt/TiO2 and Pt/SiO2 catalysts using
in situ sum-frequency generation vibrational spectroscopy. They
concluded that in the case of the Pt/TiO2 catalyst, the CO
bond is hydrogenated by H atoms spilled over from the metal
to the support, whereas for the Pt/SiO2 catalyst this reaction
pathway is hardly possible due to the inert nature of the SiO2
support. We attempted to hydrogenate propanal with p-H2
using Pt/TiO2 and Pt/SiO2 as the catalysts. No hyper-
polarization of reaction products was observed in these
1
When the gas is flowing (17 mL/s), the H NMR signals of
butanol are not detected (Figure 5c). However, in the case of
Pt/Al2O3, Pt/TiO2, Pt/SiO2, Pd/TiO2, Rh/TiO2, and Pd−Sn/
2026
dx.doi.org/10.1021/cs500426a | ACS Catal. 2014, 4, 2022−2028

ACS Catalysis
Research Article
experiments. In the case of Pt/TiO2, this could be the result of
participation of spillover hydrogen in the reaction37 because
incorporation of random H atoms is not expected to produce
any PHIP effects. The activity of the Pt/SiO2 catalyst was
significantly lower than that of Pt/TiO2: no reaction was
observed at 120 °C; at 150 °C (in ALTADENA conditions)
decarbonylation leading to ethane began, and only at 300 °C
did the hydrogenation become apparent, but without any PHIP
effects.
REFERENCES
■
(1) Narayanan, S. Bull. Catal. Soc. Ind. 2003, 2, 107−121.
(2) Claus, P. Top. Catal. 1998, 5, 51−62.
(3) Gallezot, P.; Richard, D. Catal. Rev.: Sci. Eng. 1998, 40, 81−126.
(4) Maki-Arvela, P.; Haj
́
ek, J.; Salmi, T.; Murzin, D. Yu. Appl. Catal.,
̈
A 2005, 292, 1−49.
(5) Ponec, V. Appl. Catal., A 1997, 149, 27−48.
(6) Daimon, A.; Kamitanaka, T.; Kishida, N.; Matsuda, T.; Harada, T.
J. Supercrit. Fluids 2006, 37, 215−219.
(7) de Graauw, C. F.; Peters, J. A.; van Bekkum, H.; Huskens, J.
Synthesis 1994, 10, 1007−1017.
(8) Touroude, R. J. Catal. 1980, 65, 110−120.
(9) Birchem, T.; Pradier, C. M.; Berthier, Y.; Cordier, G. J. Catal.
1994, 146, 503−510.
CONCLUSION
■
Parahydrogen-induced polarization (PHIP) was used for
studying the heterogeneous hydrogenation of several unsatu-
rated (acrolein, crotonaldehyde) and saturated (acetone,
propanal) carbonyl compounds. PHIP effects were successfully
observed in liquid and gas phase hydrogenations of the CC
bond of acrolein and crotonaldehyde over several supported
metal catalysts. These observations confirm the existence of the
route of pairwise hydrogen addition to the CC bond of
unsaturated aldehydes in the reaction mechanism. At the same
time, no PHIP effects that would correspond to pairwise
hydrogen addition to the CO bond could be observed for
unsaturated and saturated substrates tested. However, the
conversion to alcohols was very low in all cases. Therefore, to
address the feasibility of PHIP observation in CO bond
hydrogenation it would be valuable to come up with catalytic
systems and reaction conditions that provide a much higher
activity in this reaction. Upon hydrogenation of acrolein over
several of the supported metal catalysts, the NMR signal of the
CHO group of propanal was also polarized, possibly as a result
of a minor process in which the C(O)−H bond is broken.
Furthermore, the formation of polarized cis- and trans-2-
butenes was detected in hydrogenation of acrolein over several
Rh-based catalysts, and polarized propane was formed in
heterogeneous hydrogenation of acetone and propanal as a
result of C−O bond hydrogenolysis. This demonstrates that the
applicability of PHIP can be successfully extended from
hydrogenations to other more complex catalytic reactions,
which generally improves the availability of PHIP as the
technique for uncovering fine details of catalytic processes
including identification of the structure and the mechanism of
formation of less abundant reaction products.
(10) Delbecq, F.; Sautet, P. J. Catal. 1995, 152, 217−236.
(11) Wei, H.; Gomez, C.; Liu, J.; Guo, N.; Wu, T.; Lobo-Lapidus, R.;
Marshall, C. L.; Miller, J. T.; Meyer, R. J. J. Catal. 2013, 298, 18−26.
(12) Tauster, S. J.; Fung, S. C.; Garten, R. L. J. Am. Chem. Soc. 1978,
100, 170−175.
(13) Tauster, S. J. Acc. Chem. Res. 1987, 20, 389−394.
(14) Bowers, C. R.; Weitekamp, D. P. J. Am. Chem. Soc. 1987, 109,
5541−5542.
(15) Pravica, M. G.; Weitekamp, D. P. Chem. Phys. Lett. 1988, 145,
255−258.
(16) Duckett, S. B.; Mewis, R. E. Acc. Chem. Res. 2012, 45, 1247−
1257.
(17) Natterer, J.; Bargon, J. Prog. NMR Spectrosc. 1997, 31, 293−315.
(18) Blazina, D.; Duckett, S. B.; Dunne, J. P.; Godard, C. Dalton
Trans. 2004, 2601−2609.
(19) Kovtunov, K. V.; Zhivonitko, V. V.; Skovpin, I. V.; Barskiy, D.
A.; Koptyug, I. V. In Topics in Current Chemistry; Kuhn, L. T., Ed.;
Springer-Verlag: Berlin, 2013; Vol. 338, pp 123−180.
(20) Kovtunov, K.; Beck, I.; Bukhtiyarov, V.; Koptyug, I. Angew.
Chem., Int. Ed. 2008, 47, 1492−1495.
(21) Zhivonitko, V. V.; Kovtunov, K. V.; Beck, I. E.; Ayupov, A. B.;
Bukhtiyarov, V. I.; Koptyug, I. V. J. Phys. Chem. C 2011, 115, 13386−
13391.
(22) Kovtunov, K. V.; Beck, I. E.; Zhivonitko, V. V.; Barskiy, D. A.;
Bukhtiyarov, V. I.; Koptyug, I. V. Phys. Chem. Chem. Phys. 2012, 14,
11008−11014.
(23) Koptyug, I. V.; Kovtunov, K. V.; Burt, S. R.; Anwar, M. S.; Hilty,
C.; Han, S.; Pines, A.; Sagdeev, R. Z. J. Am. Chem. Soc. 2007, 129,
5580−5586.
(24) Kovtunov, K. V.; Barskiy, D. A.; Salnikov, O. G.; Khudorozhkov,
A. K.; Bukhtiyarov, V. I.; Prosvirin, I. P.; Koptyug, I. V. Chem.
Commun. 2014, 50, 875−878.
(25) Henning, H.; Dyballa, M.; Scheibe, M.; Klemm, E.; Hunger, M.
Chem. Phys. Lett. 2013, 555, 258−262.
ASSOCIATED CONTENT
■
́
̊
(26) Golman, K.; Axelsson, O.; Johannesson, H.; Mansson, S.;
S
* Supporting Information
Olofsson, C.; Petersson, J. S. Magn. Reson. Med. 2001, 46, 1−5.
(27) Bhattacharya, P.; Harris, K.; Lin, A. P.; Mansson, M.; Norton, V.
A.; Perman, W. H.; Weitekamp, D. P.; Ross, B. D. Magn. Reson. Mater.
Phys. Biol. Med. 2005, 18, 245−256.
Catalysts used, experimental setup and conditions, and
additional NMR spectra. This material is available free of
charge via the Internet at http://pubs.acs.org.
(28) Reineri, F.; Viale, A.; Giovenzana, G.; Santelia, D.; Dastru, W.;
Gobetto, R.; Aime, S. J. Am. Chem. Soc. 2008, 130, 15047−15053.
(29) Dechent, J. F.; Buljubasich, L.; Schreiber, L. M.; Spiess, H. W.;
AUTHOR INFORMATION
■
Munnemann, K. Phys. Chem. Chem. Phys. 2012, 14, 2346−2352.
Corresponding Author
*E-mail: kovtunov@tomo.nsc.ru.
̈
(30) Balu, A. M.; Duckett, S. B.; Luque, R. Dalton Trans. 2009,
5074−5076.
Notes
(31) Koptyug, I. V.; Zhivonitko, V. V.; Kovtunov, K. V.
ChemPhysChem 2010, 11, 3086−3088.
The authors declare no competing financial interest.
(32) Akhmadullina, N. S.; Cherkashina, N. V.; Kozitsyna, N. Yu.;
Stolarov, I. P.; Perova, E. V.; Gekhman, A. E.; Nefedov, S. E.; Vargaftik,
M. N.; Moiseev, I. I. Inorg. Chim. Acta 2009, 362, 1943−1951.
(33) Moudler, J.; Stickle, W.; Sobol, P.; Bomben, K. Handbook of X-
ray Photoelectron Spectroscopy; Perkin-Elmer Corp.: Eden Priarie, MN,
1992.
ACKNOWLEDGMENTS
■
This work was partially supported by the RAS (5.1.1), RFBR
(12-03-00403-a, 14-03-00374-a, 14-03-31239-mol-a), SB RAS
(60, 61), the Ministry of Education and Science of the Russian
Federation, and the Council on Grants of the President of the
Russian Federation (MK-4391.2013.3).
(34) XSPEAK 4.1 Download (Free). http://xpspeak.software.
informer.com/4.1/.
2027
dx.doi.org/10.1021/cs500426a | ACS Catal. 2014, 4, 2022−2028

ACS Catalysis
Research Article
(35) Iwasita, T.; Pastor, E. Electrochim. Acta 1994, 39, 531−537.
(36) Gaba, R.; Bhandari, M.; Kakkar, R. Adv. Mater. Lett. 2013, 4,
769−778.
(37) Kennedy, G.; Baker, L. R.; Somorjai, G. A. Angew. Chem., Int. Ed.
2014, 53, 3405−3408.
(38) Houtman, C. J.; Barteau, M. A. J. Catal. 1991, 130, 528−546.
(39) Brown, N. F.; Barteau, M. A. J. Am. Chem. Soc. 1992, 114,
4258−4265.
(40) Dubois, L. H.; Castner, D. G.; Somorjai, G. A. J. Chem. Phys.
1980, 72, 5234−5240.
2028
dx.doi.org/10.1021/cs500426a | ACS Catal. 2014, 4, 2022−2028