Regio- and Chemoselective Hydrogenation of Dienes to Monoenes Governed by a Well-Structured Bimetallic Surface

Article abstract of DOI:10.1021/jacs.7b08792

Unprecedented surface chemistry, governed by specific atomic arrangements and the steric effect of ordered alloys, is reported. Rh-based ordered alloys supported on SiO₂ (Rh_xM_y/SiO₂, M = Bi, Cu, Fe, Ga, In, Pb, Sn, and Zn) were prepared and tested as catalysts for selective hydrogenation of trans-1,4-hexadiene to trans-2-hexene. RhBi/SiO₂ exhibited excellent regioselectivity for the terminal C=C bond and chemoselective hydrogenation to the monoene, not to the overhydrogenated alkane, resulting in a high trans-2-hexene yield. Various asymmetric dienes, including terpenoids, were converted into the corresponding inner monoenes in high yields. This is the first example of a regio- and chemoselective hydrogenation of dienes using heterogeneous catalysts. Kinetic studies and density functional theory calculations revealed the origin of the high selectivity: (1) one-dimensionally aligned Rh arrays geometrically limit hydrogen diffusion and attack to alkenyl carbons from one direction and (2) adsorption of the inner C=C moiety to Rh is inhibited by steric repulsion from the large Bi atoms. The combination of these effects preferentially hydrogenates the terminal C=C bond and prevents overhydrogenation to the alkane.

Full text of DOI:10.1021/jacs.7b08792

Subscriber access provided by La Trobe University Library
Article
Regio- and Chemoselective Hydrogenation of Dienes to
Monoenes Governed by a Well-Structured Bimetallic Surface
Masayoshi Miyazaki, Shinya Furukawa, and Takayuki Komatsu
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b08792 • Publication Date (Web): 28 Nov 2017
Downloaded from http://pubs.acs.org on November 28, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 30
Journal of the American Chemical Society
1
2
3
4
5
Regioꢀ and Chemoselective Hydrogenation of Dienes to Monoenes
Governed by a WellꢀStructured Bimetallic Surface
6
7
8
9
Masayoshi Miyazaki,† Shinya Furukawa,‡,§,* and Takayuki Komatsu†,*
†Department of Chemistry, School of Science, Tokyo Institute of Technology, 2ꢀ12ꢀ1ꢀE1ꢀ10
Ookayama, Meguroꢀku, Tokyo 152ꢀ8551, Japan
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
‡
Institute for Catalysis, Hokkaido University, N10 W21, Kitaꢀku, Sapporo 001ꢀ0021, Japan
§ ESICB, Kyoto University, Nishikyoꢀku, Kyoto 519ꢀ5510, Japan
Corresponding authors
Shinya Furukawa
Institute for Catalysis, Hokkaido University, N10 W21, Kitaꢀku, Sapporo 001ꢀ0021, Japan
Eꢀmail: furukawa@cat.hokudai.ac.jp, Tel: +81ꢀ11ꢀ706ꢀ9162,
Takayuki Komatsu
Department of Chemistry, School of Science, Tokyo Institute of Technology, 2ꢀ12ꢀ1ꢀE1ꢀ10
Ookayama, Meguroꢀku, Tokyo 152ꢀ8551, Japan
Eꢀmail: komatsu.t.ad@m.titech.ac.jp, Tel: +81ꢀ3ꢀ5734ꢀ3532, Fax: +81ꢀ3ꢀ5734ꢀ2758
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 30
1
2
3
Abstract
4
5
6
7
8
9
Unprecedented surface chemistry, governed by specific atomic arrangements and the steric effect of
ordered alloys, is reported. Rhꢀbased ordered alloys supported on SiO2 (RhxMy/SiO2, M = Bi, Cu, Fe,
Ga, In, Pb, Sn, and Zn) were prepared and tested as catalysts for selective hydrogenation of
transꢀ1,4ꢀhexadiene to transꢀ2ꢀhexene. RhBi/SiO2 exhibited excellent regioselectivity for the
terminal C=C bond and chemoselective hydrogenation to the monoene, not to the overhydrogenated
alkane, resulting in a high transꢀ2ꢀhexene yield. Various asymmetric dienes, including terpenoids,
were converted into the corresponding inner monoenes in high yields. This is the first example of a
regioꢀ and chemoselective hydrogenation of dienes using heterogeneous catalysts. Kinetic studies
and density functional theory calculations revealed the origin of the high selectivity: (1)
oneꢀdimensionally aligned Rh arrays geometrically limit hydrogen diffusion and attack to alkenyl
carbons from one direction and (2) adsorption of the inner C=C moiety to Rh is inhibited by steric
repulsion from the large Bi atoms. The combination of these effects preferentially hydrogenates the
terminal C=C bond and prevents overhydrogenation to the alkane.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Keywords: regioselective, chemoselective, diene, hydrogenation, intermetallic compound
2
ACS Paragon Plus Environment

Page 3 of 30
Journal of the American Chemical Society
1
2
3
Introduction
4
5
6
7
8
9
Selective hydrogenation of unsaturated hydrocarbons, such as alkynes and dienes, is an important
chemical transformation in various disciplines such as organic synthesis, pharmacology, and
industrial chemistry.1–6 Semihydrogenation of an alkyne to the alkene has been extensively studied
using Pdꢀbased bimetallic materials, as represented by Lindlar catalysts.7,8 To this day, a number of
efficient catalytic systems for alkyne semihydrogenation have been reported.9–15 The common
strategy for obtaining high alkene selectivity is the promotion of alkene desorption to inhibit
overhydrogenation to the alkane, which is typically attained by reducing the adsorption energy of the
alkene.16 Similar approaches are also available for diene hydrogenation to the monoene.17 However,
success in selective diene hydrogenation using heterogeneous catalysts has been limited for
symmetric dienes such as 1,3ꢀbutadiene.18–22 This is because regioselectivity between the two
symmetrically distinct C=C bonds is required for hydrogenation of asymmetric dienes. Traditional
solid metal catalysts are hardly capable of distinguishing such differences.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In a typical approach, regioselective molecular recognition ability can be provided to
heterogeneous systems using microꢀ or mesoporous materials. Physical restrictions of the adsorption
geometry or transition states by the pore walls aid regioselective conversions. For instance, Pt
nanoparticles encapsulated in a ZIFꢀ8 metal–organic framework23 or Pd complexes immobilized in
MCMꢀ41 mesoporous silica24 have shown higher regioselectivity for terminal C=C bond
hydrogenation than that of the inner C=C bond. However, no heterogeneous catalysts affording high
monoene yields from asymmetric diene have been reported in the literature. This is partly because
undesired overhydrogenation to the alkane seems to be unavoidable over such active metal sites.
That is to say, not only regioselectivity but also chemoselectivity is mandatory for achieving highly
efficient conversions of dienes to monoenes. For this purpose, a metallic surface itself should
recognize the position of the C=C bond and distinguish diene and monoene chemically. It remains a
tremendous challenge to construct such a sophisticated reaction environment for high monoene
yields using conventional inorganic materials.
Ordered alloys, also known as intermetallic compounds, exhibit specific crystal structures and
wellꢀdefined surface atomic arrangements of their metal components, making them attractive
candidates for inorganic materials that enable selective diene conversion.25,26 Their highly ordered
and electronically modified bimetallic surfaces would be appropriate reaction environments for
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 30
1
2
3
4
5
6
7
regioꢀ and chemoselective molecular recognition. We recently reported that specific atomic
arrangements of some intermetallic compounds controlled stereoselective alkene isomerization27 and
chemoselective molecular recognition for nitroarene hydrogenation quite well.28
8
9
Motivated by these discoveries, in the present study, we developed a highly efficient catalytic
system for hydrogenation of asymmetric dienes to monoenes using intermetallic catalysts. The
reaction mechanism was investigated in detail using kinetic studies and density functional theory
(DFT) calculations to clarify the surface reaction dynamics. Herein, we report the first example of
heterogeneous regioꢀ and chemoselective diene hydrogenation and a novel concept using ordered
alloys for wellꢀcontrolled surface chemistry.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Experimental Section
Catalyst preparation
A series of monometallic catalysts was prepared by poreꢀfilling impregnation using silica as a
support. Aqueous solutions of H2PtCl6, Pd(NO3)2, Rh(NO3)3, and Ni(NO3)2 were added to dried
silica gel (CARiACT Gꢀ6, Fuji Silysia, SBET = 470 m2 g−1) so that the solutions filled the silica pores
(pore volume: 0.76 cm3g−1, metal concentration: 185 mM for the synthesis of Rh/SiO2). The
concentration of Rh(NO3)3 aq was 4.53 wt%. The mixtures were sealed overnight at room
temperature and dried over a hot plate, followed by reduction under flowing H2 at 673 K for 1 h.
Silicaꢀsupported Rhꢀbased intermetallic catalysts (RhM/SiO2, M = Bi, Fe, Ga, In, Pb, Sb, Sn, and Zn)
were prepared by poreꢀfilling coꢀimpregnation. Mixed aqueous solutions of Rh(NO3)3 and a second
metal salt (Fe(NO3)3ꢁ9H2O (Wako, 99%), Zn(NO3)2ꢁ6H2O (Kanto, 99%), Ga(NO3)3ꢁnH2O (Wako,
99.9%), In(NO3)3ꢁ9H2O (Kanto, 99.9%), SnCl2 (Kanto, 97%), SbCl3 (Kanto, 98%),
Pb(CH3COO)2ꢁ3H2O (Kanto, 99.5%), and Bi(NO3)3ꢁ5H2O (Kanto, 99.5%)) were used in a manner
similar to that of the monometallic catalysts. HNO3 or HCl was used to dissolve Bi(NO3)3ꢁ5H2O or
SnCl2 and SbCl3, respectively. The metal loading and the atomic ratio of Rh/M were adjusted to 3
wt% and 1, respectively. For bimetallic catalysts, the reduction procedure was performed at 1073 K.
RhBi catalysts using other supports were prepared by conventional coꢀimpregnation with an excess
amount of water (ca. 15 mL of ion exchanged water/g of support).
Catalytic reaction
4
ACS Paragon Plus Environment

Page 5 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Diene hydrogenations were carried out in 50 mL threeꢀnecked roundꢀbottomed flasks equipped with
a silicone rubber septum and a gas storage balloon (2 L). The catalyst (10 or 20 mg) was reduced
under flowing H2 (60 mL min−1) at 673 K for 45 min in the reactor followed by cooling to room
temperature. The reaction was initiated by adding a tetrahydrofuran (THF) solution (5 mL) of diene
(0.2 mmol) into the reactor under a 1 atm H2 atmosphere at room temperature. The products were
quantitated by a flame ionization detection gas chromatograph (FID–GC, Shimadzu, GCꢀ14B)
equipped with a capillary column (GL Sciences, TCꢀWAX, 0.25 mm × 30 m). The partial pressure of
H2 was adjusted using Ar as a diluent for the kinetic study. The reaction rates for all kinetic analyses
were obtained at conversions of 1,4ꢀhexadiene and H2 lower than 20% and 1%, respectively. For
recycling tests, the catalyst (100 mg) after reaction was washed by decantation in the reactor with
THF. The THF rinse was completely removed by decantation and subsequent drying under an Ar
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
flow (40 mL min− ) at room temperature. The washed catalyst was reused after the pretreatment step
outlined above.
Characterization
The crystal structure of the prepared catalyst was examined by powder Xꢀray diffraction (XRD) with
a Rigaku RINT2400 using an Xꢀray source of Cu Kα. Difference XRD patterns were obtained by
subtraction of the silica support pattern from those of the supported catalysts so that the base line was
flattened. Transmission electron microscopy (TEM) was conducted using a JEOL JEMꢀ2010F
microscope. High angle annular dark field scanning TEM microscopy (HAADFꢀSTEM) was carried
out using a JEOL JEMꢀARM200M microscope equipped with an energy dispersive Xꢀray (EDX)
analyzer (EX24221M1G5T). Both TEM analyses were performed at an accelerating voltage of 200
kV. To prepare the TEM specimen, all samples were sonicated in ethanol and then dispersed on a Cu
grid supported by an ultrathin carbon film. Fourierꢀtransformed infrared (FTꢀIR) spectra of adsorbed
CO were obtained with a JASCO FT/IRꢀ430 spectrometer in transmission mode. A selfꢀsupporting
wafer (50 mg cm−2) of catalyst was placed in a quartz cell with CaF2 windows and attached to a glass
circulation system. The catalyst was reduced under flowing H2 at 400 °C for 1 h, evacuated at the
same temperature for 0.5 h, and cooled to room temperature. After the pretreatment, a spectrum was
recorded as the base line for subsequent measurements. A pulse of CO, typically of 100–103 Pa order,
was introduced in a stepwise manner at room temperature. All spectra were recorded at 1 cm−1
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 30
1
2
3
4
5
6
7
resolution. CO pulse chemisorption was performed to estimate the Pd, Pt, Rh, and Ni dispersion in
the prepared catalysts. Prior to chemisorption, the catalyst was pretreated under a H2 flow (60 mL
1
min− ) for 0.5 h. After the reduction pretreatment, He was introduced at the same temperature for 10
8
9
min to remove the chemisorbed hydrogen, followed by cooling to room temperature. A 5% CO/He
pulse was introduced into the reactor, and the passed supplied CO was quantified downstream by a
thermal conductivity detector. For example, for RhBi/SiO2, Rh dispersion was estimated as 0.26%.
Inductively coupled plasma atomic emission spectroscopy (ICPꢀAES) analysis of the catalysts was
performed using a Shimadzu ICPEꢀ9000 spectrometer.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
DFT calculations
Periodic DFT calculations were performed using the CASTEP code29 with Vanderbiltꢀtype ultrasoft
pseudopotentials30 and the revised version of the Perdew–Burke–Ernzerhof exchange–correlation
functional31,32 based on the generalized gradient approximation. The planeꢀwave basis set was
truncated at a kinetic energy of 350 eV. A Fermi smearing of 0.1 eV was utilized. DFTꢀD correlations
were considered using the Tkatchenko–Scheffler method33 with parameters of sR = 0.94 and d = 20.0.
The reciprocal space was sampled using a kꢀpoint mesh with a spacing of typically 0.05 Å−1, as
generated by the Monkhorst−Pack scheme.34
Geometry optimizations and transition state (TS) searches were performed on supercell structures
using periodic boundary conditions. The surfaces were modeled using metallic slabs with a thickness
of typically four atomic layers with 13 Å of vacuum spacing. Unit cells were (2 × 6) for RhZn, RhGa,
RhIn, and RhFe, and (2 × 3) for RhBi. The convergence criteria for structure optimization and energy
calculation were set to (a) an SCF tolerance of 2.0 × 10−6 eV per atom, (b) an energy tolerance of
1.0 × 10−5 eV per atom, (c) a maximum force tolerance of 0.05 eV Å−1, and (d) a maximum
displacement tolerance of 1.0 × 10−3 Å.
The adsorption energy was defined as follows:
Ead = EAꢀS–(ES + EA)
where EAꢀS is the energy of the slab together with the adsorbate, EA is the total energy of the free
adsorbate, and ES is the total energy of the bare slab. The adsorption energy for a
hydrogenꢀpreadsorbed slab was calculated using ESH, the total energy of the hydrogenꢀadsorbed slab,
instead of using ES.
6
ACS Paragon Plus Environment

Page 7 of 30
Journal of the American Chemical Society
1
2
3
Surface energy calculations were conducted for densely packed lowꢀindex planes of RhBi
4
5
((100), (101), (001), (110), (102), (103), and (211)) and RhZn ((100), (110), (111), (210), (211),
6
7
and (221)), where Xꢀray diffraction were observed. Surface energies of stoichiometric surfaces
8
9
were estimated using
1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
γ = lim
[ES − NEB]
n→∞
2A
where EB is the energy of bulk unit cell, A is the surface area, n is the number of layers, and N is
the number of unit cells in the slab. The surface energies converged typically within 6 or 7
atomic layers for the considered surfaces. Surface energies of nonꢀstoichiometric surfaces were
calculated using a method in literature.27
The TS search was performed using the complete linear synchronous transit/quadratic
synchronous transit (QST) method.35,36 Linear synchronous transit maximization was performed
followed by energy minimization in the directions conjugating to the reaction pathway. The
approximated TS was used to perform QST maximization with conjugate gradient minimization
refinements. The cycle was repeated until a stationary point was found. Convergence criteria for the
TS calculations were set to rootꢀmeanꢀsquare forces on an atom tolerance of 0.10 eV Å−1.
Frequency calculations of adsorbed CO molecules were conducted from geometry optimized
structures using the DMol3 code.37 These calculations involved the RPBE functional, a
doubleꢀnumeric quality basis set with polarization functions except that no p functions are used on
hydrogen (DND, comparable to Gaussian 6ꢀ31G*)38 with a realꢀspace cutoff of 4.2 Å, DFT
semiꢀcore pseudopotential core treatment,39 and a Fermi smearing of 0.1 eV. The SCF convergence
was accelerated using the iterative scheme proposed by Kresse and Furthmüller.40 The partial
Hessian matrix including C and O atoms was computed to evaluate the harmonic frequencies for
adsorbed CO. All computed harmonic frequencies were scaled by an empirical factor of 1.0386,
which corresponds to the ratio of experimental41 and asꢀcalculated values for gasꢀphase CO (2143
cm−1/2063.4 cm−1).
Results
Catalyst characterization
A series of Rhꢀbased intermetallic catalysts was prepared using a coꢀimpregnation method with SiO2
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 30
1
2
3
4
5
6
7
8
9
as the support (RhM/SiO2, M = Bi, Fe, Ga, In, Pb, Sb, Sn, and Zn). Crystallite phases of the prepared
catalysts were analyzed using XRD (Figure 1). For all the catalysts, the desired intermetallic phases
were formed with high phase purities. The crystallite sizes of the intermetallic phases, estimated by
Scherrer’s equation, were 2–10 nm. Figures 2a and 2b show the transmission electron microscopy
(TEM) image of RhBi/SiO2 and the size distribution of the nanoparticles, respectively. Particle sizes
ranged from 2 to 6 nm (mean diameter: 4.1 nm), which agreed with the crystallite size estimated
using XRD (4 nm). Figures 2c and 2d show the highꢀresolution (HR)ꢀTEM image of a single
nanoparticle. Lattice fringes with 0.222 nm spacing were clearly observed, which is consistent with
the interplanar distance of the RhBi(102) plane (0.221 nm). The fast Fourier transform of the
HRꢀTEM image (Figure 2e) indicates that the nanoparticle is of a RhBi single crystal oriented along
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢀꢀ
the [221] direction. These results indicate that the observed nanoparticles are single crystals of the
intermetallic RhBi phase. We also performed highꢀangle annular dark field scanning TEM
(HAADFꢀSTEM) analysis for RhBi/SiO2. The HAADFꢀSTEM image of RhBi/SiO2 was shown in
Figure 2f. The Z contrast between metal nanoparticles, SiO2 support, and the background was clearly
observed, showing that the RhBi nanoparticles are dispersed on SiO2. Elemental maps of Rh and Bi
acquired using energy dispersive Xꢀray (EDX) analysis on this field revealed that Rh and Bi atoms
composing the nanoparticles were homogeneously dispersed (Figures 2g, 2h, and 2j). This suggests
that the observed nanoparticles comprise both Rh and Bi without segregation. The presence of Rh
and Bi on SiO2 in 1:1 ratio was also confirmed by ICPꢀAES analysis (Rh/Bi = 1.04). The EDX
mapping with Si allows elemental identification on the HAADFꢀSTEM image more clearly (Figures
2i and 2k). Figure 2l shows the HRꢀHAADFꢀSTEM image of a single RhBi nanoparticle.
Hexagonally aligned Rh and/or Bi atoms were observed, which can be assigned to intermetallic RhBi
(space group: P63/mmc) viewed along [001] direction (Figure 2m). Note that superposition of
Figures 2l and 2m shows a good agreement (Figure S1). At the edge region of this nanoparticle,
bulkꢀterminated structures were observed, implying that {110} planes are exposed. This result is
consistent with our recent experimental and theoretical study suggesting that (110) plane dominates
the surface of RhBi.28
Diene hydrogenation
We first tested the catalytic performances of monometallic catalysts (M/SiO2, M = Pt, Pd, Rh, and
8
ACS Paragon Plus Environment

Page 9 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Ni) in the hydrogenation of dienes (Table 1). transꢀ1,4ꢀHexadiene (hereafter 1,4ꢀhexadiene) was used
as a simple model diene with both terminal and internal C=C bonds. In all cases, transꢀ2ꢀhexene
(hereafter, 2ꢀhexene), 1ꢀhexene, and nꢀhexane were formed by hydrogenation of terminal, internal,
and both C=C bonds of 1,4ꢀhexadiene, respectively. The metal dispersion measured by CO
chemisorption and the corresponding turnover frequency (TOF) for each catalyst are also listed in
Table 1. Although Pt/SiO2 showed high catalytic activity, the alkene selectivity was decreased by
significant overhydrogenation to the alkane. Pd/SiO2 gave high alkene selectivity, and substantial
amounts of 1ꢀhexene were generated, which decreased the regioselectivity. When Rh/SiO2 was used,
moderate catalytic activity was obtained and the 2ꢀhexene selectivity was higher than that observed
with Pt and Pd. The use of Ni/SiO2 resulted in very low catalytic activity. This is probably due to
insufficient reduction of Ni in this condition (pretreated by H2 at 400°C), which is also reflected in
the very low Ni dispersion. Thus, based on these results, we chose Rh as the main active metal for
developing bimetallic catalysts that would be regioꢀ and chemoselective.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Then, a series of Rhꢀbased intermetallic catalysts was tested in the same reaction. Figure 3 shows
the conversion–selectivity curves obtained in 1,4ꢀhexadiene hydrogenation over Pt and some
Rhꢀbased catalysts. Bimetallic catalysts showed 2ꢀhexene selectivities higher than that of
monometallic Rh and Pt. RhBi showed the highest 2ꢀhexene selectivity over the whole conversion
region. For most catalysts, the selectivity decreased as the conversion increased, corresponding to
overhydrogenation of the hexenes to nꢀhexane. In contrast, surprisingly, RhBi retained the initial
high selectivity until complete conversion. This indicated that overhydrogenation of 2ꢀhexene to
nꢀhexane was well inhibited, even at the high conversion regions. Moreover, no 1ꢀhexene was
detected during the hydrogenation over RhBi (Figure S2), suggesting selective hydrogenation of the
terminal C=C bond. Thus, RhBi showed high regioꢀ and chemoselectivities in diene hydrogenation.
These selectivities should be evaluated quantitatively so that individual contributions to the overall
2ꢀhexene selectivity are distinguished more clearly. The intrinsic regioselectivity was expected to
appear at the initial stage of hydrogenation, where overhydrogenation of monoene to alkane would
be little. In this study, we used the relative formation rate of 1ꢀhexene to that of 2ꢀhexene (r1h/r2h) at
low conversions (<30%) as a scale for the regioselectivity. Chemoselectivity was described as the
decrease in monoene (1ꢀ and 2ꢀhexenes) selectivity during the reaction (ꢂS1+2), which reflects the
degree of overhydrogenation of monoenes to alkane. The values for all the tested catalysts are
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 30
1
2
3
4
5
6
7
8
9
summarized in Figure 4. For regioselectivity (Figure 4, red circles), the intermetallic catalysts gave
lower values than the monometallic Rh and other metal catalysts. This demonstrates that selectivity
for the terminal C=C bond (r1h/r2h) improved when intermetallic phases were used. The lowest values
were obtained with RhBi and RhSn (r1h/r2h = 0), highlighting the outstanding terminal selectivity.
Here one should confirm the absence of fast isomerization and overhydrogenation of 1ꢀhexene to
2ꢀhexene and nꢀhexane, respectively. The presence of such pathways would lead to underestimation
of the regioselectivity. Therefore, we examined the competitive hydrogenation of 1,4ꢀhexadiene and
1ꢀoctene using an equimolar mixture. Note that 1ꢀoctene was used instead of 1ꢀhexene to distinguish
the products. The hydrogenation of 1ꢀoctene was much slower than that of 1,4ꢀhexadiene and no
2ꢀoctene was formed (Figure S3), which supports the absence of the fast side reactions mentioned
above. Next, for chemoselectivity (Figure 4, blue bars), the ꢂS1+2 values differed significantly
depending on the second metal. Among them, RhBi showed the lowest ꢂS1+2 value, close to zero
(0.6), indicating minimum overhydrogenation of 2ꢀhexene. This feature was clearly reflected in the
conversion–selectivity curve mentioned above. Interestingly, only a small amount of nꢀhexane was
detected, with a constant fraction, throughout the reaction (Figure S2). This implies the involvement
of direct conversion of 1,4ꢀhexadiene to nꢀhexane. The remarkable chemoselectivity of the
RhBi/SiO2 catalyst was also demonstrated in the hydrogenation of 2ꢀhexene in the presence and
absence of 1,4ꢀhexadiene. As shown in Figure 5a, the reaction rate of 2ꢀhexene was much slower
than that of 1,4ꢀhexadiene. Moreover, 2ꢀhexene was not converted in the presence of 1,4ꢀhexadiene
(Figure 5b). These results suggest that, on RhBi, both adsorption and hydrogenation of 1,4ꢀhexadiene
are more favored than those of 2ꢀhexene. Thus, RhBi showed excellent regioꢀ and chemoselectivity
in the hydrogenation of 1,4ꢀhexadiene to 2ꢀhexene.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The RhBi/SiO2 catalyst also exhibited good selectivity toward yielding inner alkenes in the
hydrogenation of various dienes and diene derivatives (Table 2). Conjugated dienes, such as
1,3ꢀhexadiene and 3ꢀmethylꢀ1,3ꢀpentadiene, were also converted into the corresponding inner
alkenes in good selectivities (Table 2, entries 2 and 3). 3ꢀMethylꢀ1,3ꢀpentadiene showed relatively
lower selectivity than those of the other dienes, probably because a stable alkyl intermediate having
tertiary carbon is formed during hydrogenation of the inner C=C bond. 4ꢀVinylcyclohexene was
hydrogenated to 3ꢀethylcyclohexene with high selectivity (Table 2, entry 4). Hydrogenation of
linalool (3,7ꢀdimethylꢀ1,6ꢀoctadienꢀ3ꢀol) gave dihydrolinalool (3,7ꢀdimethylꢀ6ꢀoctenꢀ3ꢀol), which is
10
ACS Paragon Plus Environment

Page 11 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
a synthetic perfume, in high yield (Table 2, entry 5). The catalytic performance of RhBi/SiO2 was
also examined in intermolecular competitive hydrogenation. Here, a 1:1 mixture of styrene and
βꢀmethylstyrene was used as the terminal and internal alkene, respectively (Table 2, entry 6). When
styrene conversion to ethylbenzene reached 83%, only 9% of βꢀmethylstyrene was hydrogenated to
propylbenzene. This indicated that the selectivity for styrene hydrogenation against βꢀmethylstyrene
was 90%. Thus, RhBi/SiO2 is capable of preferential hydrogenation of terminal C=C bonds not only
intramolecularly but also intermolecularly. We also checked the recyclability of RhBi/SiO2 catalyst
for hydrogenation of 1,4ꢀhexadiene. The catalyst could be separated from the reaction mixture by
simple decantation and reused at least three times without significant change in the initial catalytic
performances (Figure S4). The concentration of Rh and Bi in the liquid phase after the reaction were
below the detection limit of ICPꢀAES (< 0.02 ppm), indicating that leaching of the metallic species
can be negligible. The effects of catalyst support and solvent on selectivity were also tested using the
hydrogenation of 1,4ꢀhexadiene (Tables S1 and S2). However, no significant difference in selectivity
was observed, suggesting that the regioꢀ and chemoselectivity depend strongly on the properties of
the metallic surface.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
We subsequently performed kinetic studies for 1,4ꢀhexadiene hydrogenation over RhBi/SiO2.
Hydrogenation of 1,4ꢀhexadiene using D2 instead of H2 resulted in a small kinetic isotope effect
(kH/kD = 1.6), indicating that the dissociation of H2 was not rate determining. Firstꢀorder (1.04) and
negativeꢀorder (−0.41) dependences were observed for H2 pressure and 1,4ꢀhexadiene concentration,
respectively, on the reaction rate (Figures S5a and S5b). Considering a Langmuir–Hinshelwood
mechanism involving H2 and diene, the observed reaction orders suggest that (1) a surface reaction is
the rateꢀdetermining step and (2) the strong adsorption of 1,4ꢀhexadiene inhibits H2 adsorption
(details are shown in Supporting Information, P. S13, Kinetic model). Therefore, another surface
reaction may be the rateꢀlimiting step, such as hydrogen diffusion or hydride attack on the alkenyl
carbon. We also performed a kinetic analysis for monometallic Rh/SiO2, resulting in reaction orders
close to those for RhBi; 1.26 and −0.26 for H2 pressure and diene concentration, respectively
(Figures S5c and S5d). This implies that a kinetic model analogous to that for RhBi also holds for
monometallic Rh. We also considered the effect of intraparticle diffusion on the reaction rates to
confirm that the obtained kinetic parameters only reflect catalytic processes. For this purpose, a
series of RhBi/SiO2 catalysts with different Rh loadings (2, 3, 4, and 5 wt%) were used with the total
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 30
1
2
3
4
5
6
7
8
9
amount of Rh constant. This allows change in the number of RhBi nanoparticles per silica
nanoparticles, which reflects the intraparticle diffusion length. We confirmed that all the catalysts
had same crystallite sizes (3 nm) and no apparent difference in catalytic activity was observed
(Figure S6). These results demonstrate that the intraparticle diffusion processes do not contribute to
the overall reaction rates.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
DFT calculations
Based on the experimental results, we believe that the reaction dynamics of hydrogen diffusion
and alkenyl carbon attack on the surface are drastically modified by the second metal. Therefore, we
carried out DFT calculations using some slab models to clarify the effect of the second metal on the
kinetics of hydrogenation and the catalytic performance. For each model, we chose the crystal plane
most likely to be exposed as a surface. In this context, we previously reported a brief study on the
surface energies and vibrational frequencies of chemisorbed CO on the surface of Rhꢀbased
intermetallic compounds.27,28 For instance, the (110) planes of RhBi (Figures 6a and 6b), RhZn, and
RhGa gave low surface energies, and the theoretical vibrational frequencies of CO were close to the
experimental values. In the present study, we performed deeper calculations for RhBi and RhZn with
a variety of lowꢀindex planes as candidates for the exposed surface. Here, RhZn was chosen as a
representative material for RhM with CsClꢀtype structures such as RhGa, RhIn, and RhZn. The
surface energies and frequencies of linearly adsorbed CO on various planes of RhBi ((100), (101),
(001), (110), (102), (103), and (211)) and RhZn ((100), (110), (111), (210), (211), and (221)) were
considered. For both materials, the (110) planes showed the lowest surface energies (Table S3),
demonstrating that the (110) planes are most stable and likely to be exposed. To confirm this, we
conducted CO adsorption on RhBi/SiO2 and compared the experimental and theoretical vibrational
frequencies of adsorbed CO. Figure 7 shows the inꢀsitu FTꢀIR spectrum of CO adsorbed on
RhBi/SiO2. An intense peak with a shoulder appeared in the region of atop CO, whereas no signal
assigned to bridge CO was observed.42,43 This clearly indicates sufficient dilution of Rh–Rh
ensembles, hence the coꢀpresence of both Rh and Bi atoms at the outermost layers. The observed
band could be deconvoluted into two symmetric curves. The peak position of the main component
well agreed with the theoretical frequencies of CO adsorbed on (110), which support that the (110)
plane dominates the surface of RhBi. The minor component could be assigned to CO adsorbed on
12
ACS Paragon Plus Environment

Page 13 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
(102) plane, which possesses the second lowest surface energy (γ(110) = 0.75 Nm−2, γ(102) = 0.87 Nm−2,
Table S3). Therefore, it is thermodynamically reasonable that the surface was dominated by these
planes. These two planes display surface atomic arrangements different to each other (Figure 7). Rh
atoms are aligned oneꢀdimensionally on (110) plane (Rh–Rh distance: 2.83 Å), whereas they are
isolated on (102) plane (Rh–Rh distance: 4.08 Å). It has been reported that isolation of an active
metal by an inert metal significantly reduces the catalytic activity for C=C bond hydrogenation.14,44
Therefore, (102) plane seems to be less active for diene hydrogenation than (110) plane. We hereafter
consider RhBi(110) plane as the main active reaction site for diene hydrogenation. For RhZn(110)
plane, we also confirmed that a good agreement was obtained between the experimental and the
theoretical frequency values (Table S3).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Then, we focused on the adsorption configuration of 1,4ꢀhexadiene at the surface of RhBi(110). In
general, diꢀσ adsorption of alkenyl moieties on monometallic surfaces is known to give adsorption
energies more negative than πꢀadsorption.45 However, recent experimental and theoretical studies
have shown that πꢀadsorbed species were more favored for hydrogenation than σꢀadsorbed
species.46,47 In this study, various possible adsorption configurations were considered and compared;
1,2ꢀdiꢀσ (terminal C=C), 4,5ꢀdiꢀσ (internal C=C), 1,2,4,5ꢀtetraꢀσ (both C=C), 1,2ꢀπ, 4,5ꢀπ,
1,2,4,5ꢀdiꢀπ, alkylidene, and alkylidyne. The optimized structures and the corresponding adsorption
energies of these configurations are provided in Supporting Information Figure S7 and Table S4,
respectively. The 1,2,4,5ꢀdiꢀπ structure, where the diene molecule aligned with the Rh lows (Figure
6c), showed the most negative adsorption energy (−170 kJmol−1; more negative than all other
configurations at least by 21 kJmolꢀ1). Thus, the DFT calculation revealed that the diꢀπ adsorption
was most stable and likely at the surface. Similar diꢀπ adsorption of 1,4ꢀhexadiene on RhGa(110) and
RhZn(110) planes was also considered and optimized (Figure S8). On such configurations, hydrogen
diffusion and attack to the alkenyl carbons are restricted and only allowed along the Rh rows owing
to the extremely high energy potentials of the second metals, inert to hydrogen.24 The Rh–C distance
at the internal C5=C4 moiety was longer than that at the terminal C2=C1 moiety, probably because of
the steric repulsion between the C6 methyl group and the surface. This feature was prominent for
surfaces having second metals with a large atomic radius,48 such as Bi (Rh: 1.34 Å; Bi: 1.82 Å;
Figure 6c; see Figure S8 and Table S5 for details of all surfaces). Because the longer Rh−C distance
generally results in a less favored TS, hydrogen attack to the terminal and inner alkenyl carbons
13
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 30
1
2
3
4
5
6
7
8
9
might be kinetically distinguished by the steric hindrance. We therefore calculated the energy profiles
of hydrogen diffusion and attack to the alkenyl carbons of adsorbed 1,4ꢀhexadiene. Figure 8 shows
the corresponding energy diagrams for the RhBi(110) surface and side views of the slab models. The
first step is hydrogen diffusion from the stationary Rh–Rh bridge site, adjacent to the diene, to the
site under the alkenyl moiety via the intermediate Rh top site. The second step is hydrogen attack on
the alkenyl carbon to form an alkyl species. For the initial diffusion step, similar energy barriers were
observed for terminal and internal C=C bonds. Conversely, the energy barriers for hydrogen attack
significantly differed: the terminal C2=C1 bond (blue line) showed a much lower barrier than the
internal C4=C5 bond (red line). The difference in the highest activation energies (ꢂEA) was 15.8 kJ
mol−1, which corresponds to an approximately 600ꢀfold higher rate constant for the terminal C=C
than that for the internal one (from the Arrhenius equation at 25 °C). This is consistent with the high
regioselectivity of RhBi. At the TSs, the distances between Rh and the attacking H are 2.055 and
1.798 Å for the internal and terminal moieties, respectively (Figure S9). The unstable TS and the
resulting higher Ea for the inner moiety can be attributed to the longer Rh–H distance. Thus, as
expected, the steric hindrance between large Bi atoms and the inner alkenyl moiety makes the TS
significantly unfavorable; hence, regioselective hydrogenation is allowed. We also considered similar
energy profiles for some Rhꢀbased intermetallic surfaces having second metals of different atomic
radii: RhZn(110) and RhGa(110) (Rh: 1.34 Å; Zn: 1.35 Å; Ga: 1.53 Å; Bi: 1.82 Å; Figure S10). For
the RhZn(110) surface, where Rh and Zn are similar in size, the Rh–C distance at the inner alkenyl
moiety is close to that at the terminal one because of less steric repulsion from the surface (Figure S7
and Table S5). The reaction barriers of hydrogen diffusion and attack for the inner moiety are also
close to those of the terminal one, providing a small ꢂEA value of 4.3 kJ mol−1. For the RhGa(110)
surface, an intermediate ꢂEA value of 12.8 kJ mol−1 was observed. The obtained ꢂEA values from
these bimetallic surfaces were plotted with respect to the experimental regioselectivity (Figure S11).
The higher the ꢂEA value, the lower the r1h/r2h value (higher regioselectivity), supporting the
assertion that regioselectivity is under kinetic control for the hydrogen diffusion or attack step.
The reaction barrier of diene hydrogenation was also considered for RhBi(102) surface to
demonstrate our assumption that (102) surface is less active than (110) surface as mentioned above.
Indeed, hydrogen attack to the terminal alkenyl carbon of 1,4ꢀhexadiene adsorbed on RhBi(102) with
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
diꢀπ configuration showed a very high activation energy (128 kJmol−1, Figure S12), which is much
14
ACS Paragon Plus Environment

Page 15 of 30
Journal of the American Chemical Society
1
2
3
4
higher than the corresponding barrier for RhBi(110) surface (33 kJmol−1, Figures 7 and S12). This is
probably due to the absence of stable Rh–Rh bridge adsorption site adjacent to the alkenyl carbon.
5
6
7
8
9
Discussion
On the basis of the obtained results, we now discuss the essential factors governing the
regioselectivity. Firstly, as mentioned above, the adsorption affinity of the terminal and inner C=C
moieties (Rh–C distance) relies largely on the steric hindrance from the second metal atoms.
Therefore, the size of the second metal atom, namely the atomic radius, should play a crucial role in
the regioselectivity. Secondly, the atomic arrangement of Rh at the surface may also contribute to the
selective hydrogenation. As observed for RhBi(110), RhGa(110), and RhZn(110), some oneꢀtoꢀone
intermetallic compounds display oneꢀdimensionally aligned Rh arrays separated by the inert second
metal. Hydrogen diffusion and attack to the alkenyl carbons on such surfaces are restricted and
allowed only on the Rh arrays as described in Figures 6 and 8. However, other types of Rh
arrangements are also observed. RhPb exhibits the (001) plane as the most stable surface, where
Rhꢀrich composition (Rh/Pb = 3) and large surface Rh–Rh coordination number (4 at maximum) are
observed. On such a surface, hydrogen diffusion and attack are no longer restricted
oneꢀdimensionally and are allowed from multiple directions, which makes regioselective
hydrogenation more difficult. Conversely, for RhSn, Rh atoms at the most likely (012) surface are
isolated by the surrounding Sn atoms. Hydrogen diffusion on this surface is tightly restricted,
perhaps showing higher regioselectivity. Figure 9 shows the correlation between regioselectivity
r1h/r2h and atomic radius of the second metal. Overall, the trend for regioselectivity increases,
depending roughly on the increase in atomic radius of the second metal. This trend is more clearly
observed among catalysts exposing planes with oneꢀdimensionally aligned Rh arrays. RhPb and
RhSn fall out of this strong correlation, probably because of the difference in surface Rh alignment
and hydrogen diffusion mentioned above. Thus, in this catalytic system, the bulkiness of the second
metal and the atomic arrangement of Rh determine the regioselectivity. Highly ordered intermetallic
RhBi provides a wellꢀstructured reaction environment for the excellent terminal regioselectivity,
where hydrogen diffusion and attack are controlled geometrically and sterically. This specific surface
structure may also contribute to the high chemoselectivity to the monoene. The weakened adsorption
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
of the inner C=C moiety should promote desorption of the product inner alkene and make the TS for
15
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 30
1
2
3
4
5
6
7
8
9
hydrogen attack unfavorable, thereby effectively preventing overhydrogenation to the alkane.
We also investigated the effect of the Rh electronic state on catalytic performance. The
vibrational frequency of linearly adsorbed CO on Rh, obtained from FTꢀIR measurements, was used
as an indicator of the Rh electron density. However, no correlation was observed between the
selectivities and the electronic factor (Figure S13), excluding the electronic effects on the observed
modification in selectivity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
It should be noted that the surface chemistry observed in the present study differs completely
from those reported for conventional alloy catalysts. There are two wellꢀknown effects of alloying on
catalysis: ligand and ensemble effects, which are simple electronic and geometric modifications in
the state of the active metal, respectively. In contrast, the chemistry at the surface of RhBi relies on
the highly and sterically controlled molecular dynamics governed by the ordered atomic arrangement
of the two metal components. As mentioned at the beginning, we recently reported similar effects in
stereoselective alkene isomerization and chemoselective molecular recognition for nitroarene
hydrogenation, where we coined the term “steric effect” for the novel alloying effect. The
observation in this study provides a third example of the steric effect, which expands the horizons of
catalytic chemistry, surface science, and material science.
Conclusion
In this study, monometallic and Rhꢀbased bimetallic ordered alloys were tested as catalysts for the
hydrogenation of 1,4ꢀhexadiene to 2ꢀhexene. RhBi/SiO2 showed complete regioselectivity for
hydrogenation of the terminal C=C bond. This catalyst also exhibited high hydrogenation
chemoselectivity to monoene, where overhydrogenation of monoene to alkane was effectively
prevented. Various asymmetric dienes could be converted into the corresponding inner monoenes in
high yields. This is the first report of a heterogeneous system achieving regioꢀ and chemoselective
hydrogenation of dienes to monoenes. Detailed kinetic studies and DFT calculations revealed the
reaction mechanism and the origin of high regioꢀ and chemoselectivity. Intermetallic RhBi displays
oneꢀdimensionally aligned Rh arrays at the stable (110) surface that geometrically limits hydrogen
diffusion and attack to the alkenyl carbons to one direction. Moreover, adsorption of the inner C=C
moiety to the Rh site is inhibited by steric repulsion from the large Bi atoms. The combination of
these two effects preferentially hydrogenates the terminal C=C bond and prevents overhydrogenation
16
ACS Paragon Plus Environment

Page 17 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
to the alkane. The surface chemistry observed in this study, which we term the “steric effect,” differs
completely from the wellꢀknown ligand and ensemble effects. This work provides not only an
efficient heterogeneous catalytic system but also important insights into the design and construction
of sophisticated inorganic reaction environments.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Acknowledgment
This work was supported by JSPS KAKENHI Grant Number 26820350 and 17H04965. We deeply
appreciate the Center for Advanced Materials Analysis Tokyo Institute of Technology and technical
staffs of Research Institute for Electronic Science, Hokkaido University for help with TEM and
HAADFꢀSTEM observations, respectively. Computation time was provided by the supercomputer
systems in Tokyo Institute of Technology (TSUBAME2.5) and in Institute for Chemical Research,
Kyoto University.
Associated Content
Supporting information
The Supporting Information is available free of charge on the ACS Publications website at DOI:
Detailed catalytic performance; kinetics of hydrogenation on RhBi; surface energy and frequency
calculation; adsorption structure of 1,4ꢀhexadiene over RhBi, RhGa, and RhZn; energy diagram of
hydrogenation of 1,4ꢀhexadiene over RhGa and RhZn.
17
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 30
1
2
3
4
5
6
7
References
(1)
(2)
Bond, G. C.; Wells, P. B. Adv. Catal. 1965, 15, 91–226.
Jia, J.; Haraki, K.; Kondo, J. N.; Domen, K.; Tamaru, K. J. Phys. Chem. B 2000, 104, 11153–
11156.
8
9
(3)
(4)
(5)
(6)
(7)
(8)
(9)
Mohammadi, N. A.; Rempel, G. L. Macromolecules 1987, 20, 2362–2368.
Molnár, Á.; Sárkány, A.; Varga, M. J. Mol. Catal. A Chem. 2001, 173, 185–221.
Borodziński, A.; Bond, G. C. Catal. Rev. 2006, 48, 91–144.
Chinchilla, R.; Nájera, C. Chem. Rev. 2014, 114, 1783–1826.
Lindlar, H. Helv. Chim. Acta 1952, 35, 446–450.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Rajaram, J.; Narula, A. P. S.; Chawla, H. P. S.; Dev, S. Tetrahedron 1983, 39, 2315–2322.
Mitsudome, T.; Yamamoto, M.; Maeno, Z.; Mizugaki, T.; Jitsukawa, K.; Kaneda, K. J. Am.
Chem. Soc. 2015, 137, 13452–13455.
(10) Aduriz, H. R.; Bodnariuk, P.; Coq, B.; Figueras, F. J. Catal. 1989, 119, 97–107.
(11) Shin, E. W.; Choi, C. H.; Chang, K. S.; Na, Y. H.; Moon, S. H. Catal. Today 1998, 44, 137–
143.
(12) Furlong, B. K.; Hightower, J. W.; Chan, T. Y. L.; Sarkany, A.; Guczi, L. Appl. Catal. A, Gen.
1994, 117, 41–51.
(13) González, S.; Neyman, K. M.; Shaikhutdinov, S.; Freund, H. J.; Illas, F. J. Phys. Chem. C
2007, 111, 6852–6856.
(14) Armbrüster, M.; Behrens, M.; Cinquini, F.; Föttinger, K.; Grin, Y.; Haghofer, A.; Klötzer, B.;
KnopꢀGericke, A.; Lorenz, H.; Ota, A.; Penner, S.; Prinz, J.; Rameshan, C.; Révay, Z.;
Rosenthal, D.; Rupprechter, G.; Sautet, P.; Schlögl, R.; Shao, L.; Szentmiklósi, L.; Teschner,
D.; Torres, D.; Wagner, R.; Widmer, R.; Wowsnick, G. ChemCatChem 2012, 4, 1048–1063.
(15) McCue, A. J.; McRitchie, C. J.; Shepherd, A. M.; Anderson, J. A. J. Catal. 2014, 319, 127–
135.
(16) Studt, F.; AbildꢀPedersen, F.; Bligaard, T.; Sørensen, R. Z.; Christensen, C. H.; Nørskov, J. K.
Science 2008, 320, 1320–1322.
(17) Lucci, F. R.; Liu, J.; Marcinkowski, M. D.; Yang, M.; Allard, L. F.; FlytzaniꢀStephanopoulos,
M.; Sykes, E. C. H. Nat. Commun. 2015, 6, 8550.
(18) Zhang, X.; Shi, H.; Xu, B. Q. Angew. Chemie ꢀ Int. Ed. 2005, 44, 7132–7135.
(19) Lonergan, W. W.; Vlachos, D. G.; Chen, J. G. J. Catal. 2010, 271, 239–250.
(20) Baiker, A.; Baris, H.; Güntherodt, J. H. J. Chem. Soc., Chem. Commun. 1986, No. 12, 930–
932.
(21) Yardimci, D.; Serna, P.; Gates, B. C. ChemCatChem 2012, 4, 1547–1550.
(22) Yardimci, D.; Serna, P.; Gates, B. C. ACS Catal. 2012, 2, 2100–2113.
(23) Stephenson, C. J.; Hupp, J. T.; Farha, O. K. Inorg. Chem. Front. 2015, 2, 448–452.
(24) Shimazu, S.; Baba, N.; Ichikuni, N.; Uematsu, T. J. Mol. Catal. A Chem. 2002, 182–183, 343–
350.
(25) Komatsu, T.; Hyodo, S.; Yashima, T. J. Phys. Chem. B 1997, 101, 5565–5572.
18
ACS Paragon Plus Environment

Page 19 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
(26) CastillejosꢀLópez, E.; Agostini, G.; Di Michel, M.; IglesiasꢀJuez, A.; BachillerꢀBaeza, B. ACS
Catal. 2017, 7, 796–811.
(27) Furukawa, S.; Ochi, K.; Luo, H.; Miyazaki, M.; Komatsu, T. ChemCatChem 2015, 7, 3472–
3479.
(28) Furukawa, S.; Takahashi, K.; Komatsu, T. Chem. Sci. 2016, 7, 4476–4484.
(29) Segall, M. D.; Lindan, P. J. D.; Probert, M. J.; Pickard, C. J.; Hasnip, P. J.; Clark, S. J.; Payne,
M. C. J. Phys. Condens. Matter 2002, 14, 2717–2744.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(30) Vanderbilt, D. Phys. Rev. B 1990, 41, 7892–7895.
(31) Hammer, B.; Hansen, L. B.; Nørskov, J. K. Phys. Rev. B 1999, 59, 7413–7421.
(32) Ernzerhof, M.; Scuseria, G. E. J. Chem. Phys. 1999, 110, 5029–5036.
(33) Tkatchenko, A.; Scheffler, M. Phys. Rev. Lett. 2009, 102, 73005.
(34) Monkhorst, H.; Pack, J. Phys. Rev. B 1976, 13, 5188–5192.
(35) Halgren, T. A.; Lipscomb, W. N. Chem. Phys. Lett. 1977, 49, 225–232.
(36) Govind, N.; Petersen, M.; Fitzgerald, G.; KingꢀSmith, D.; Andzelm, J. Comput. Mater. Sci.
2003, 28, 250–258.
(37) Delley, B. J. Chem. Phys. 1990, 92, 508–517.
(38) Inada, Y.; Orita, H. J. Comput. Chem. 2008, 29, 225–232.
(39) Delley, B. Phys. Rev. B 2002, 66, 155125.
(40) Kresse, G.; Furthmüller, J. Phys. Rev. B 1996, 54, 11169–11186.
(41) Scarano, D.; Bordiga, S.; Lamberti, C.; Spoto, G.; Ricchiardi, G.; Zecchina, a; Area, C. O.
Surf. Sci. 1998, 411, 272–285.
(42) Yang, C.; Garl, C. W. J. Phys. Chem. 1957, 61, 1504–1512.
(43) Tomishige, K.; Asakura, K.; Iwasawa, Y. J. Catal. 1994, 149, 70–80.
(44) Kovnir, K.; Armbruter, M.; Teschner, D.; Venkov, T. V; Jentoft, F. C.; KnopꢀGericke, A.;
Grin, Y.; Schlol, R. Sci. Technol. Adv. Mater. 2007, 8, 420–427.
(45) Horiuti, I.; Polanyi, M. Trans. Faraday Soc. 1934, 30, 1164–1172.
(46) Cremer, P. S.; Su, X. C.; Shen, Y. R.; Somorjai, G. a. J. Am. Chem. Soc. 1996, 118, 2942–
2949.
(47) Valcárcel, A.; Clotet, A.; Ricart, J. M.; Delbecq, F.; Sautet, P. Surf. Sci. 2004, 549, 121–133.
(48) Housecroft, C. E.; Sharpe, A. G. Inorganic Chemistry, 4th edition; Pearson, 2012.
19
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 30
1
2
3
Tables
4
5
6
7
8
9
Table 1. Catalytic performance of monometal catalysts supported on SiO2
in 1,4ꢀhexadiene hydrogenationa
Conv.
(%)
TOF
(s−1)c
Selectivity (%)d
Disp. Time
(%)b (min)
Metal
2ꢀene
1ꢀene
10
nꢀane
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Pt
22.6
8.4
15
5
94
97
95
94
169
432
50
59
76
70
73
31
7
Pd
Rh
Ni
17
35.1
0.26
12
270
8
21
15
177
12
aReaction conditions: catalyst, 10 mg; substrate, 0.2 mmol; temperature, 25
˚C; solvent, THF (5 mL); atmosphere, 1 atm H2. bMetal dispersion
determined by CO adsorption. c Estimated at conversion below 30%. d2ꢀene:
2ꢀhexene; 1ꢀene: 1ꢀhexene; nꢀane: nꢀhexane.
Table 2. Selective hydrogenation of dienes on the RhBi/SiO2 catalyst
Time Conv. Sel.
(h)
Entry Substrate
Product
(%)
95
99
85
90
97
83
(%)
1
2
3
4
5
1,4ꢀHexadiene
2ꢀHexene
5.5
90
1,3ꢀHexadiene
4ꢀVinylcyclohexene
Linalool
3ꢀHexene
20
6
88
4ꢀEthylcyclohexene
Dihydrolinalool
89
9
95
3ꢀMethylꢀ1,3ꢀpentadiene 3ꢀMethylꢀ2ꢀpentene
18
80
>99
Styrene +
Ethylbenzene +
propylbenzene
6
10
βꢀmethylstyrene
9
>99
Reaction conditions: catalyst, 10 mg (RhBi/SiO2); substrate, 0.2 mmol; temperature, 25
˚C; solvent, THF (5 mL); atmosphere, 1 atm H2.
20
ACS Paragon Plus Environment

Page 21 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. X-ray diffraction (XRD) patterns of Rh-based bimetallic catalysts supported on SiO2 and
corresponding references (ICDD-PDF, open and filled symbols). For clarity, diffraction of the SiO2 support
was subtracted from the original patterns.
75x74mm (600 x 600 DPI)
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. (a) Transmission electron microscopy (TEM) image of RhBi/SiO2 and (b) size distribution of the
nanoparticles. (c) High-resolution TEM image of a single nanoparticle. (d) Expansion of the region
designated by the orange square in (c). (e) Fast Fourier transform (FT) of the single nanoparticle in (c). (f)
HAADF-STEM image of RhBi/SiO2. Elemental maps of (g) Rh (L-edge), (h) Bi (M-edge), (i) Si (K-edge), (j)
Rh + Bi, and (k) Rh + Bi + Si acquired using EDX. (l) HR-HAADF-STEM image of one RhBi nanoparticle. (m)
Crystal structure of intermetallic RhBi viewed along [001] direction as a structural model of (l).
152x115mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 23 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Conversion–selectivity curves obtained from 1,4-hexadiene hydrogenation over Pt and selected Rh-
based catalysts.
66x55mm (300 x 300 DPI)
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 24 of 30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Regioselectivity (r1h/r2h) and chemoselectivity (∆S1+2) for monometal catalysts (Pd, Pt, Ni, Rh) and
Rhꢀbased intermetallic catalysts. Regioselectivity (r1h/r2h) is the relative formation rate of 1ꢀhexene to that
of 2ꢀhexene at low conversions (<30%), and chemoselectivity (∆S1+2) is the decrease in monoene (1ꢀ and
2ꢀhexenes) selectivity during the reaction.
48x29mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 25 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Time courses of 1,4-hexadiene and 2-hexene conversion when these substrates were used (a)
alone or (b) together.
42x22mm (300 x 300 DPI)
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 26 of 30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. (a) Crystal structure of RhBi using the ball model (space group: P63/mmc). (b) Atomic
arrangement of the RhBi(110) surface with the space-filling model. (c) Optimized adsorption structure of
1,4-hexadiene on the RhBi(110) surface. Carbon atoms are numbered for distinction.
80x57mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 27 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. FT-IR spectrum of CO adsorbed on RhBi/SiO2. Bold dashed lines indicate the theoretical vibrational
frequencies of CO adsorbed on atop sites of (110) and (102) planes. Upper left inset shows change in the
FT-IR spectra upon the increase in CO pressure from 2 Pa to 170 Pa. The initial spectrum designated by bold
line, of which peak intensity is lower than 20% of that for saturation coverage, is shown as the main
spectrum.
73x67mm (300 x 300 DPI)
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 28 of 30
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 8. Energy diagrams for C=C hydrogenation of 1,4ꢀhexadiene over a RhBi(110) surface. All energies
are relative to gasꢀphase 1,4ꢀhexadiene plus the slab with hydrogen. For each model, the C=C bonds were
adsorbed onto Rh atoms through diꢀπ adsorption.
72x65mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 29 of 30
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 9. Relation between atomic radius and regioselectivity. Filled circles indicate catalysts exposing planes
with one-dimensionally aligned Rh arrays, and open circles indicate those exposing planes with other types
of Rh arrangement.
50x32mm (600 x 600 DPI)
ACS Paragon Plus Environment