Tandem hydroformylation/hydrogenation of alkenes to normal alcohols using Rh/Ru dual catalyst or Ru single component catalyst

Article abstract of DOI:10.1021/ja307998h

The catalyst system for tandem hydroformylation/hydrogenation of terminal alkenes to the corresponding homologated normal alcohol was developed. The reaction mechanism for the Rh/Ru dual catalyst was investigated by real-time IR monitoring experiments and ³¹P NMR spectroscopy, which proved the mutual orthogonality of Rh-catalyzed hydroformylation and Ru-catalyzed hydrogenation. Detailed investigation about Ru-catalyzed hydrogenation of undecanal under H₂/CO pressure clarified different kinetics from the hydrogenation under H₂ and gave a clue to design more active hydrogenation catalysts under H₂/CO atmosphere. The solely Ru-catalyzed normal selective hydroformylation/hydrogenation is also reported.

Full text of DOI:10.1021/ja307998h

Article
pubs.acs.org/JACS
Tandem Hydroformylation/Hydrogenation of Alkenes to Normal
Alcohols Using Rh/Ru Dual Catalyst or Ru Single Component Catalyst
Kohei Takahashi, Makoto Yamashita,† and Kyoko Nozaki*
Department of Chemistry and Biotechnology, Graduate School of Engineering, University of Tokyo 7-3-1 Hongo, Bunkyo-ku,
113-8656, Tokyo, Japan
S
* Supporting Information
ABSTRACT: The catalyst system for tandem hydroformylation/hydro-
genation of terminal alkenes to the corresponding homologated normal
alcohol was developed. The reaction mechanism for the Rh/Ru dual
catalyst was investigated by real-time IR monitoring experiments and 31P
NMR spectroscopy, which proved the mutual orthogonality of Rh-
catalyzed hydroformylation and Ru-catalyzed hydrogenation. Detailed
investigation about Ru-catalyzed hydrogenation of undecanal under H2/
CO pressure clarified different kinetics from the hydrogenation under H2
and gave a clue to design more active hydrogenation catalysts under H2/
CO atmosphere. The solely Ru-catalyzed normal selective hydro-
formylation/hydrogenation is also reported.
investigations, it was shown to be difficult to achieve the
tandem reaction by using only one catalyst.
INTRODUCTION
■
Linear 1-alkanols (n-alcohols) are widely used in industry as
solvents and precursors of detergents.1 Direct and selective
conversion of terminal alkene into n-alcohol by anti-
Markovnikov hydration would be an ideal process.2 In reality,
current industrial production of n-alcohols mostly employs a
multiple-step process consisting of hydroformylation of
terminal alkenes, purification of n-aldehydes, and then hydro-
genation of n-aldehydes to n-alcohols. One-pot hydroformyla-
tion/hydrogenation process would be advantageous because it
simplifies the process operation. As another alternative to the
anti-Markovnikov hydration, Grubbs et al. reported a tandem
Wacker oxidation/hydrogenation of 1-alkenes to n-alcohols,
very recently.3
The tandem hydroformylation/hydrogenation has been
investigated for a long time using Co-,4 Rh-,5 Ru-,6 and Pd-
based7 systems (Scheme 1). Although these tandem systems
gave a mixture of n- and i-alcohols in good yields (mostly
>90%), a significant amount of alkane was often given as a
byproduct. In addition, another problematic issue is the low
normal/iso selectivities (n/i < 10) in the hydroformylation step,
causing low n-alcohol yield (up to 81%). Through those
Instead of expecting one catalyst to play multiple roles, the
use of multiple catalysts in one pot may be a more effective
approach to the tandem reaction. Recently, Cole-Hamilton et
al.8 and Breit et al.9 reported a tandem hydroformylation/
hydrogenation catalyst system employing Rh precursor with
two ligands giving high n-alcohol yield (∼90%) with high n/i
(>30). In these systems, one ligand is in charge for the Rh-
catalyzed n-selective hydroformylation, and the other ligand
mediates the Rh-catalyzed hydrogenation, respectively. Vogt et
al. reported that Rh/XANTPHOS,10 which was originally
reported as n-selective hydroformylation catalyst and does not
catalyze hydrogenation, can catalyze this tandem reaction (n-
alcohol 86%, n/i = 11) in a 1:9 mixture of polar organic solvent
and water at high temperatures.11
In 2010, we reported a Rh/Ru dual system for tandem
hydroformylation/hydrogenation, converting 1-decene to n-
undecanol in over 90% yield (n/i = 22).12 In the system, we
employed Rh/XANTPHOS10 as a n-selective hydroformylation
catalyst and Shvo’s catalyst13 as a chemo-selective hydro-
genation catalyst. The key to our success could be attributed to
the “orthogonality” between each catalyst. Shvo’s catalyst was
relatively inert in the hydroformylation step compared to the
rapid hydroformylation by Rh. Possible side reactions like
hydrogenation of 1-decene to decane or isomerization to 2-
decenes were not problematic. Also, Shvo’s complex main-
tained hydrogenation activity to aldehyde in the presence of Rh,
XANTPHOS, and even CO. A control experiment analyzing
catalyst solution by 31P NMR spectroscopy showed no evidence
Scheme 1. Tandem hydrformylation/hydrogenation
Received: August 14, 2012
Published: November 1, 2012
© 2012 American Chemical Society
18746
dx.doi.org/10.1021/ja307998h | J. Am. Chem. Soc. 2012, 134, 18746−18757

Journal of the American Chemical Society
Article
for the presence of Rh−Ru cluster, which also supported the
orthogonality of these catalysts.
catalyst system was successfully applied to the diols syntheses.
The results are summarized in the Table 1. The previously
reported data for the 1-decene are cited as run 1 in ref 12.
When allyl alcohol was treated under the same conditions, the
corresponding homologated alcohol, that is, 1,4-butandiol was
obtained in a low yield (31%, run 2). A significant amount of
propanol was given due to the direct hydrogenation of allyl
alcohol (roughly estimated as 20%). While Shvo’s complex is
less active for the hydrogenation of a CC double bond,
isomerization of allyl alcohol to propanal allows rapid
hydrogenation of the resulting CO double bond affording
1-propanol.15 Moreover, the hydroformylation product 4-
hydroxybutanal formed a five-membered ring cyclic hemiacetal,
which underwent dehydrogenation to produce thermodynami-
cally stable γ-butyrolactone. On the other hand, allyl acetate,
which corresponds to the protected form of ally alcohol was
successfully converted to 4-hydroxybutyl acetate in higher yield
(78%, run 3) because both isomerization of CC to CO
and formation of hemiacetal were suppressed. In the same way,
homoallyl alcohol, which is susceptible to formation of six-
membered ring cyclic acetal gave 1,5-pentandiol in 75% and δ-
valerolactone in 11% yields (run 4), and homoallyl acetate gave
5-hydroxypentyl acetate in 87% yield without any significant
byproducts (run 5). An even longer alcohol 4-pentene-1-ol was
converted to 1,6-hexanediol with excellent yield (95%) without
any byproduct (run 6). Functional group tolerance was
demonstrated in runs 7−11. Alkenes having THPO (80%,
run 7), benzyloxy (81%, run 8), TBSO (80%, run 9), 1,3-
dioxolan-2-yl (79%, run 10), and phenylcarbamate (75%, run
11) groups gave corresponding homologated n-alcohol with
In this work, we report the scope and limitation of the Rh/
Ru dual catalyst system aiming at the production of linear α,ω-
diols. Terminal alkenes having hydroxyl or various functional
groups were converted to the corresponding homologated
linear alcohols in good yields. Also, a kinetic study using a real-
time IR monitoring system and analysis of catalyst solution by
31P NMR spectroscopy proved the mutual orthogonality of our
catalyst system and gave a clue to design a new hydrogenation
catalyst more tolerant to CO. Furthermore, here we report a
Ru-catalyzed tandem normal-selective hydroformylation/hydro-
genation. Although the catalytic activity is much lower than the
Rh/Ru system, the Shvo’s catalyst itself mediated both
hydroformylation of 1-alkene and the subsequent hydro-
genation of aldehyde to alcohol.
RESULTS AND DISCUSSION
■
Substrate Scope. Among various n-alcohols, linear α,ω-
diols are of industrial products utilized as monomers for
polyurethanes, polyesters, etc.14 Our Rh/XANTPHOS/Shvo’s
a
Table 1. Hydrofromylation/Hydrogenation of Various 1-Alkenes Catalyzed by Rh/Ru
alcohols
b
c
run
R1, R2, R3
C8H17, H, H
n (%)
i (%)
n/i
alkane (%)
1.4
others (%)
1
2
90
31
78
75
87
95
4.1
3.5
22
8.9
>100
32
internal alkenes (1.9)
d
d
de
,
HOCH2, H, H
20
γ-butyrolactone (10) high boiling products (4)
b
3
AcOCH2, H, H
HO(CH2)2, H, H
AcO(CH2)2, H, H
HO(CH2)3, H, H
trace
2.4
1.9
4.5
4.5
4.0
nd
nd
nd
nd
nd
nd
nd
nd
0
butanol (9) isobutanol (10)
f
f
4
cyclic acetals (3) δ-valerolactone (11)
5
5.6
16
nd
6
2.9
33
none
g
f
f
f
f
7
THPO(CH2)4 , H, H
80
5.0
16
n-aldehyde (4) internal alkenes (2)
f
d
f
f
8
PhCH2O(CH2)4, H, H
81
4.1
20
internal alkenes (2) formates (6)
h
f
f
f
9
TBSO(CH2)4 , H, H
80
3.7
22
formate (4)
f
f
f
10
11
12
13
14
15
(1,3-dioxolan-2-yl)(CH2)8, H, H
PhNHCO2(CH2)4, H, H
cyclohexyl, H, H
79
4.2
19
formate (3)
f
f
75
4.9
15
nd
f
f
87
4.9
18
nd
f
f
f
C7H15, CH3, H,
62
trace
34
>50
0.6
1.5
starting material (15) internal alkenes (8)
internal alkenes (34) aldehydes (4.2)
none
i
C7H15, H, CH3
22
Ph, H, H
60
39
a
Reaction conditions: alkene, 2.0 mmol; Rh(acac)(CO)2, 20 μmol; XANTPHOS, 40 μmol; Shvo’s cat, 50 μmol (based on mol of Ru atom); DMA,
4.0 mL; H2, 1.0 MPa; CO, 1.0 MPa; 12.5 h. The yields in the table were determined by GC analysis with dodecane or tridecane as internal standard
otherwise mentioned. n/i = (mol of n-alcohol)/(mol of i-alcohols). The yields of aldehydes were trace otherwise mentioned. nd = not determined
b
c
d
Yields were determined by using calibration curve for n-alcohol. Number in the parentheses is the yield of the products. Yields were roughly
e
estimated by GC, comparing the integration of the peak with that of n-alcohol and corrected based on the number of carbon. Probably acetals or
aldol products. Yield was determined by H NMR with 1,3,5-trimethoxybenzene as internal standard. THP = 2-tetrahydropyranyl. TBS = tert-
butyldimethylsilyl Yield of n-undecanol.
f
g
h
1
i
18747
dx.doi.org/10.1021/ja307998h | J. Am. Chem. Soc. 2012, 134, 18746−18757

Journal of the American Chemical Society
Article
a
Scheme 2. Summary of Observed Reaction Rate As a Function of Substrate for Each Step
a
[S]: concentration of substrate in each reaction.
Figure 1. Time course of substrate and products concentration in the hydroformylation catalyzed by Rh/XANTPHOS in the presence of Shvo’s
catalyst (a) and in the absence (b) monitored by real-time IR spectra. Black dot: 1-decene, red dot: n- and i-aldehyde, and blue dot: n- and i-alcohol.
Decay of 1-decene until 95% conversion was fitted with first-order equation. Formation of n- and i-alcohol was fitted with zero-order reaction. Black
line shows the fitted functions. Common conditions: DMA, 9 mL; H2, 1.0 MPa; CO, 1.0 MPa. (a) 1-decene, 3 mmol; Rh(acac)(CO)2, 50 μmol;
XANTPHOS, 100 μmol; Shvo’s cat, 125 μmol (based on mol of Ru atom). (b) 1-decene, 3 mmol; Rh(acac)(CO)2, 50 μmol; XANTPHOS, 100
μmol.
good yield and similar n/i with 1-decene. As other examples,
vinylcyclohexane and 2-methylnonene gave n-alcohol in 87 and
62% yields, respectively (run 12 and 13). An internal alkene,
(Z)-2-decene, was converted to n-undecanol (22% run 14) via
isomerization to 1-decene, n-selective hydroformylation and
hydrogenation. However the formation of i-alcohol via direct
hydroformylation of the internal CC bond flowed by
hydrogenation was still dominant (34% run 14). Styrene was
quantitatively converted to alcohols, but low n/i (1.5) was
observed because the formation of iso-aldehyde is intrinsically
preferable in the hydroformylation of styrene (run 15).10
When compared with precedents of the tandem hydro-
formylation/hydrogenation using Rh as a singular catalyst by
Cole-Hamilton8 and by Breit,9 the present system showed
18748
dx.doi.org/10.1021/ja307998h | J. Am. Chem. Soc. 2012, 134, 18746−18757

Journal of the American Chemical Society
Article
Figure 2. (a) Time course of 1-decene consumption in the presence of Shvo’s catalyst monitored by real-time IR spectra. (b) Time course of
undecanal and undecanol concentration in the presence of Shvo’s catalyst monitored by real-time IR spectra. Black dot: 1-decene, red dot: n- and i-
aldehyde, and blue dot: n- and i-alcohol. Decay of 1-decene until 50% conversion was fitted with first-order equation. Formation of n- and i-alcohol
was fitted with zero-order reaction. Common conditions: DMA, 9 mL; H2, 1.0 MPa; CO, 1.0 MPa. (a) 1-decene, 3 mmol; Shvo’s cat, 125 μmol
(based on mol of Ru atom). The rate constant was determined from the initial 50% conversion. (b) Undecanal, 3 mmol; Shvo’s cat, 125 μmol (based
on mol of Ru atom).
comparable results in product yields, reaction rates, and the
substrate scope except that here formation of internal alkenes
and formates as byproducts was observed in some cases.
Orthogonality. In our previous communication, we
proposed that the orthogonality between Rh/XANTPHOS
system and Shvo’s hydrogenation catalyst was the key of our
system. In order to prove it, we performed the following
experiments. First, the reaction rates of both step of tandem
hydroformylation/hydrogenation catalyzed by Rh/XANT-
PHOS/Shvo’s system were compared with the separately
performed each reaction under the same condition (in DMA, at
120 °C, under 2.0 MPa of H2/CO). The compared reactions
are summarized in Scheme 2: hydroformylation/hydrogenation
of 1-decene by Rh/XANTPHOS/Shvo’s cat. (reaction a),
hydroformylation of 1-decene by Rh/XANTPHOS (reaction
b), isomerization of 1-decene by Shvo’s cat. (reaction c), and
hydrogenation of undecanal by Shvo’s cat. (reaction d). The
experimental results are summarized in Figures 1 and 2.
The rate of hydroformylation of 1-decene by Rh/
XANTPHOS system was not affected by the presence of
Shvo’s catalyst: The rate in the presence (reaction a, derived
from Figure 1a) and in the absence (reaction b from Figure 1b)
of Shvo’s catalyst were both first order to the concentration of
1-decene until 95% conversion, and the observed rate constants
were (6.4 0.8) and (6.6 0.8) × 10−3 s−1, respectively. In
contrast, increases of i-alcohol and isomerized alkenes were
observed in the presence of Shvo’s catalyst (n/i = 16 and 24 in
reactions a and b, respectively). The lower n/i in reaction a can
be explained by the isomerization of 1-decene to internal
alkenes mediated by Shvo’s catalyst and successive hydro-
formylation of the internal alkenes by Rh/XANTPHOS.16
When isomerization of 1-decene by Shvo’s catalyst was
independently performed, the reaction rate obeyed the first
order kinetics on 1-decene concentration until 50% con-
version,17 and the rate constant was (3.6
0.4) × 10−4 s−1
(reaction c, derived from Figure 2a), which was 6% of the
observed rate constant of hydroformylation of 1-decene by Rh/
XANTPHOS (reaction b, Figure 2a). Slow hydrogenation of 1-
decene to decane and hydroformylation of 1-decene to
aldehyde was confirmed by the low yields of decane (3.9%)
and alcohols (6.2%) in reaction c. When the rates of
hydrogenation of undecanal by Shvo’s catalyst in the presence
and absence of Rh/XANTPHOS were compared (reactions a
and d, Figures 1a and 2b), they were both zero order on the
concentration of undecanal, and the observed reaction rates
were (8.4 0.8) and (9.1 0.9) × 10−5 mol/L·s, respectively.
The decrease of the reaction rate was within the margin of
error. Selectivity of aldehyde to alcohol was >95% in both cases.
In this context, we could conclude the presence of Shvo’s
catalyst did not affect the rate of hydroformylation by Rh/
XANTPHOS but slightly decreased the selectivity. On the
other hand, the presence of Rh/XANTPHOS might have
decreased the rate of hydrogenation, but the difference is
almost negligible.
Orthogonality was also demonstrated by comparing the
above one-pot, one-step reaction with the one-pot, stepwise
reaction shown in Scheme 3. First, hydroformylation of 1-
decene was performed with Rh/XANTPHOS under H2/CO.
After the completion of the reaction, Shvo’s catalyst was added
to the mixture, and H2/CO was purged by H2. The yield and n/
18749
dx.doi.org/10.1021/ja307998h | J. Am. Chem. Soc. 2012, 134, 18746−18757

Journal of the American Chemical Society
Article
catalyzed by Shvo’s catalyst under H2/CO is very slow
compared to under H2.18 Considering industrial application,
it is a significant drawback to our catalyst system. Therefore,
understanding the hydrogenation step in detail is important for
further improvement of our system.
Scheme 3. Stepwise Hydroformylation/Hydrogenation
First, the effects of H2 and CO pressure and Ru and
XANTPHOS concentration on the reaction rate were
determined by real time IR monitoring (Figure 3a−d). The
varied parameters were CO and H2 pressure (Figure 3a,b,
respectively) and concentration of Shvo’s cat and XANTPHOS
(Figure 3c,d, respectively).
i of undecanol were about the same as the one-pot reaction
(run 1 in Table 1), which indicates the presence of Shvo’s
catalyst did not affect the yield of hydroformylation by Rh/
XANTPHOS at all. On the other hand, hydrogenation of
undecanal under H2 was much faster than under H2/CO (<1 h
versus ∼10 h). As discussed above, the presence of Rh/
XANTPHOS did not change the rate of hydrogenation by
Shvo’s catalyst. Therefore, poisoning of Shvo’s catalyst by CO
was confirmed.
Based on the data obtained in Figures 2b and 3, the rate
equation in the absence of XANTPHOS was expressed to be
−d[aldehyde]/dt = k1[aldehyde]0 PH P −1[Ru]
CO
(1)
2
In the presence of XANTPHOS, the rate of hydrogenation
decreased. But the effect of XANTPHOS concentration could
not be simply described. The rate eq 1 is different from the
previously reported dependency on the concentration of
Kinetics of Hydrogenation Catalyzed by Shvo’s
Catalyst under H2/CO. The reaction rate of hydrogenation
Figure 3. Rate of hydrogenation of undecanal catalyzed by Shvo’s catalyst using H2/CO under varying CO pressure (a), H2 pressure (b), Ru
concentration (c), and XANTPHOS concentration (d). Standard condition: DMA, 10 mL; H2, 1.0 MPa; CO, 1.0 MPa; undecanal, 5 mmol;
dodecane, 2.5 mmol (total 11.6 mL); Shvo’s catalyst, 0.125 mmol (based on the mol of Ru atom). Selectivity from undecanal to undecanol is >95%
in all cases. Rate constants were determined from time course of alcohols in initial 200 min by fitting with zero-order reaction. Obtained rate
constants in each figure were fitted with inverse proportion to CO pressure in (a), direct proportion to H2 pressure in (b), direct proportion to Ru
concentration in (c), and direct proportion to XANTPHOS concentration in (d). In (d) two different lines are drawn for XANTPHOS
concentration from 0 to 1.1 × 10−2 M and 1.1 × 10−2 to 2.2 × 10−2 M, respectively.
18750
dx.doi.org/10.1021/ja307998h | J. Am. Chem. Soc. 2012, 134, 18746−18757

Journal of the American Chemical Society
Article
aldehyde in the hydrogenation of aldehyde by Shvo’s catalyst
under H2.13d−g Accordingly, the change of reaction mechanism
caused by poisoning of the catalyst by CO is suggested.19
In the previous communication12 we performed 31P NMR
spectroscopic analysis of the catalyst solution prepared by the
treatment of Rh(acac)(CO)2/XANTPHOS/Shvo’s catalyst in
DMA at 120 °C under 1.0 atm of H2/CO for 15 min. Observed
signals were assigned to be Rh(CO)2H(xantphos) (δP 19 (d),
4.3 μmol), free XANTPHOS (δP −19 (s), 4.0 μmol),
Ru(CO)2(cyclopentadienone)(xantphos-κ1P) (1) (δP −22 (s),
39 (s), 8.8 μmol), and Rh2(CO)4(xantphos)2 (δP 1 (d), 8 (d),
1.5 μmol) (Scheme 4).20 On the other hand, here we found
Scheme 6. Proposed Mechanism of Hydrogenation of
Aldehyde
Scheme 4. Observed Species by 31P NMR in Ref 12
electron rich compared to 2 to strengthen the coordination of
CO to Ru, thus it makes the formation of active species 6 from
1 via loss of CO to form 5 and successive metal−ligand
cooperative activation of H2 less favorable.21 As XANTPHOS
concentration was increased from 0 to 1 equiv to Ru, the
reaction rate decreased because the ratio of XANTPHOS
ligated Ru species was increased (Figure 3d, diagonal line).
In the presence of more than 1 equiv XANTPHOS to Ru, all
the active species were XANTPHOS ligated Ru species,
resulting in a constant reaction rate (Figure 3d, horizontal
line) as a function of XANTPHOS concentration. In the
presence of XANTPHOS, the reaction was gradually slowed
down especially after 500 min.22 It means there is a catalyst
decomposition pathway caused by XANTPHOS. One explan-
ation may be dissociation of cyclopentadienone by the steric
repulsion with XANTPHOS. It should be noted however, the
actual hydroformylation/hydrogenation was carried out with
only a slight excess of XANTPHOS (the Rh/XANTPHOS/Ru
ratio was 1/2/2.5 in the standard condition), and thus the
deceleration of Ru-catalyzed hydrogenation by XANTPHOS
should not have been problematic.
that the Shvo’s catalyst was quantitatively converted to
Ru(CO)3(cyclopentadienone) (2) by the treatment under
higher H2/CO pressure of 2.0 MPa (Scheme 5). Since the
Scheme 5. Treatment of Shvo’s Catalyst under H2/CO
Pressure
Comparison of Hydrogenation Activity of Ru Catalyst
under H2/CO. The hydrogenation activity of Shvo’s catalyst
was compared with other Ru complexes under H2/CO. In the
presence of CO, a very strong coordinating ligand to Ru, the
hydrogenation rates were significantly decelerated in all of the
examined Ru catalysts when compared to their originally
reported rates under pure H2 without CO. Interestingly, Shvo’s
catalyst is far more active than Ru3(CO)12, Ru(CO)H2(PPh3)3,
and Cp*Ru(cod)Cl/Ph2PCH2CH2NH2/tBuOK23 (Table 2
runs 1−4). Under H2/CO, Shvo’s catalyst exists as 2 (Scheme
5), and Ru3(CO)12 and Ru(CO)H2(PPh3)3 are thought to be
exist as RuxHyLz (L = CO or PPh3).24 The rate of dissociation
of CO might be comparable between them because IR
absorption band of νCO for 2 (2081, 2026, 2005 cm−1) and
Ru4H4(CO)12 (2081, 2067, 2030, 2024, 2008 cm−1) is similar
to each other (Scheme 7). Thus, the difference should be
attributed to the rate difference in the subsequent steps: Shvo’s
catalyst hydrogenates aldehyde in outer sphere mechanism13d−g
(Scheme 7, eq 2), while Ru3(CO)12 and Ru(CO)H2(PPh3)3
hydrogenate aldehydes via coordination of aldehyde to Ru and
insertion of the carbonyl group to Ru−H followed by
hydrogenolysis (Scheme 7, eq 3). When Shvo’s catalyst is
compared to Cp*Ru(cod)Cl/Ph2PCH2CH2NH2/tBuOK, elec-
tron density on Ru center might be the origin for the difference
(Scheme 7, eqs 2 and 4). As discussed before, dissociation of
one CO from 2 is necessary to generate hydrogenation active
catalytic reaction employs high H2/CO pressure, the major
quantity of the Ru atoms most likely exists as 2. From these
observations, the reaction mechanism of hydrogenation under
H2/CO by Shvo’s catalyst may be proposed as described in
Scheme 6. Under high H2/CO pressure, Ru species mainly
exists as the tricarbonyl species 2. Equilibrium to form a
dimer13g was not involved in the rate-determining step
considering that the rate equation is first order on Ru
concentration. From 2, dissociation of one CO molecule
gives 3, and successive metal−ligand cooperative activation of
H2 affording the Ru−H species 4 is the rate-determining step,
which explains the fact that the reaction rate is first order on H2
pressure and inverse first order on CO pressure. Once formed,
4 immediately reacts with aldehyde to give alcohol and return
to 2 by rapid coordination of CO, resulting in zero-order
contribution of the aldehyde concentration to the reaction rate.
In the presence of XANTPHOS, it coordinates to 3 to give 1,
which was detected by 31P NMR experiment.12 When Ru is
coordinated by XANTPHOS (to form 1), it becomes more
18751
dx.doi.org/10.1021/ja307998h | J. Am. Chem. Soc. 2012, 134, 18746−18757

Journal of the American Chemical Society
Article
Normal Selective Tandem Hydroformylation/Hydro-
genation using Ru as a Single Component Catalyst.
Rather to our surprise, the normal-selective tandem hydro-
formylation/hydrogenation of 1-decene did take place in the
absence of Rh precursor; namely, a combination of Shvo’s
catalyst and XANTPHOS was active to mediate the tandem
reaction, although the hydroformylation activity was lower
compared to the dual catalyst.
Table 2. Hydrogenation of Undecanal under H2/CO with
Various Catalysts
a
time
(h)
conv
(%)
alcohol
(%)
run
cat
1
2
3
Shvo’s cat
11
12
23
12
99
<1
4.7
21
99
0
Ru3(CO)12
Ru(CO)H2(PPh3)3
4.4
b
c
4
Cp*Ru(cod)Cl/Ph2PCH2CH2NH2
<1
/tBuOK
The tandem hydroformylation/hydrogenation of 1-decene to
n-alcohol was successfully accomplished when catalyzed by a
combination of Shvo’s catalyst or cyclopentadienone ligated Ru
tricarbonyl complexes, and the results are summarized in Table
3. The examined cyclopentadienone Ru complexes were having
phenanthrene-fused 2,5-diphenylcyclopentadienone (7),26 3,4-
diphenyl-2,5-bis(ethoxycarbonyl)cyclopentadienone (8), and
cyclopentane-fused 2,5- bis(trimethylsilyl)cyclopentadienone
(9) (Figure 4). Since Shvo’s catalyst was proven to have low
activity for the hydroformylation of 1-decene in the above
experiment, higher reaction temperature was employed for this
investigation. When 1-decene was treated with a combination
of Shvo’s catalyst and XANTPHOS under H2/CO at 160 °C, it
gave n-alcohol in 47% yield as the major product with n/i of 32
(Table 3, run 1). It should be noted that the n/i value was even
higher than Rh/XANTPHOS at 120 °C (n/i = 24 in Figure
1b). A major byproduct was isomerized alkenes (13%). The
activity for hydroformylation was significantly affected by the
cyclopentadinenone ligand (runs 2−4). A combination of 7/
XANTPHOS gave n-alcohol with similar yield and n/i to Shvo’s
catalyst. The activity and n/i were lower with 8/XANTPHOS.
The highest selectivity in the hydroformylation of 1-decene to
n-undecanal was accomplished with 9/XANTPHOS resulting
in the highest yield of n-alcohol (73%, run 4). Although the
reaction rates were quite low compared to Rh or Co, n/i
selectivity is highest level among the reported tandem
hydroformylation/hydrogenation catalysts.
d
5
Shvo’s cat
13
10
99
85
98
d
c
6
Cp*Ru(cod)Cl/Ph2PCH2CH2NH2
16
/tBuOK
a
Reaction conditions: DMA, 10 mL; H2, 1.0 MPa; CO, 1.0 MPa;
undecanal, 5 mmol; dodecane, 2.5 mmol; Ru complex, 0.125 mmol
b
(based on the mol of Ru atom). The mol ratio of Cp*Ru(cod)-
c
Cl:Ph2PCH2CH2NH2:tBuOK = 1:1:1. High-boiling products were
d
observed by GC, which are considered to be dimers. iPrOH was used
as solvent.
species 4 in the case of Shvo’s catalyst. On the other hand,
Cp*Ru(Ph2PCH2CH2NH2)H was reported to be the hydro-
genation active species formed from Cp*Ru(cod)Cl/
Ph2PCH2CH2NH2/tBuOK under H2.23 After loss of dihydro-
gen, it forms coordinatively unsaturated 16e species, which
would be trapped by CO to give Cp*Ru(CO)-
(Ph2PCH2CH2NH). Regeneration of active species requires
dissociation of CO from it. Since the Ru−CO bond is
considered to be weaker in Shvo’s system than Cp*RuPN
system, (as reported values, IR absorption band of νCO
=
2081, 2026, and 2005 cm−1 for 2 and 1904 cm−1 for
Cp*Ru(CO)(NHPh)(PiPrPh2)25), the regeneration of active
species should be much easier to take place in Shvo’s catalyst.
i
When PrOH, which was reported as the best solvent for
Cp*Ru(cod)Cl/Ph2PCH2CH2NH2/tBuOK, was used for com-
parison, Shvo’s catalyst was still more active and selective than
Cp*RuPN system (runs 5 and 6).
Here we assume that (hydroxycyclopentadienyl)RuH(κ2-
XANTPHOS) (C in Figure 5) as the active species for both
Scheme 7. Comparison with Hydrogenation Mechanism with Various Ru Catalysts
18752
dx.doi.org/10.1021/ja307998h | J. Am. Chem. Soc. 2012, 134, 18746−18757

Journal of the American Chemical Society
Article
a
Table 3. Tandem hydroformylation/hydrogenation Catalyzed by Ru(CO)3(cyclopentadienone) /XANTPHOS Systems
run
cat.
conv. (%)
aldehydes (%)
n/i
alcohols (%)
n/i
internal alkenes (%)
b
1
Shvo’s cat/XANTPHOS
7/XANTPHOS
71
100
60
0.2
17
−
29
32
−
47
50
<1
73
32
26
−
13
24
50
12
2
3
4
8/XANTPHOS
7.0
1.2
9/XANTPHOS
98
29
a
Reaction condition: 1-decene, 1.0 mmol; Ru complex, 25 μmol (based on Ru atom); XANTPHOS, 50 μmol; toluene, 2.0 mL; H2, 1.0 MPa; CO,
1.0 MPa; 24 h. The yields in the table were determined by GC analysis with dodecane as internal standard, and n/i = (mol of n-product)/(mol of i-
b
products). XANTPHOS 25 μmol.
Figure 4. Investigated Ru(CO)3(cyclopentadienone) complexes.
Figure 6. ORTEP drawing of Cp*Ru(xantphos)Cl (11). Thermal
ellipsoids set at 50% probability; hydrogen atoms and solvent
molecules are omitted for clarity.
Figure 5. Conceptual explanation for Ru-based hydroformylation/
hydrogenation catalyst and Ru catalyzed hydroformylation/hydro-
genation.
CO. For hydrogenation of aldehyde, either B or C should be
responsible, although the details are still unknown.
CONCLUSION
hydroformlylation and hydrogenation. Recently, we reported
■
the Ru-catalyzed normal-selective hydroformylation using
(cyclopentadienyl)Ru/bisphosphine or bisphosphite systems.26
In this paper, we isolated Cp*Ru(Xantphos)H27 (10,
corresponds to A in Figure 5) and confirmed that it worked
as a catalyst precursor for hydroformylation of 1-decene with
high n/i selectivity. The fact that the combination of Shvo’s
catalyst/XANTPHOS was active not only for hydrogenation of
aldehyde but also for hydroformylation can be nicely explained
by considering complex C as a bifunctional catalyst playing the
roles of both A and B (Figure 5).
Although one might wonder if the bidentate coordination of
XANTPHOS in a three-legged piano stool structure would
cause a serious steric repulsion with the Cp* ligand, we isolated
Cp*Ru(xantphos)Cl (11) and characterized it by X-ray single
crystal analysis as shown in Figure 6.27,28 Thus, for hydro-
formylation, either dissociation of one phosphorus atom or a
partial dissociation of the cyclopentadienyl ligand in C would
provide a vacant site for coordination−insertion of alkene and
In summary, we reported that our system could offer a new
pathway to industrially important linear α,ω-diols from
corresponding alkenyl alcohols or acetates. Other functional
groups were also tolerate under our catalytic condition. The
mutual orthogonality of Rh catalyzed hydroformylation and Ru
catalyzed hydrogenation was proven by control experiments
using real-time IR monitoring. Also, reaction mechanism of
hydrogenation of aldehyde by Shvo’s catalyst under H2/CO
pressure was investigated by real-time IR monitoring and 31P
NMR spectroscopy. Shvo’s catalyst was found to be more active
than the conventional Ru hydrogenation catalyst because of its
robustness under CO pressure. Based on the above
consideration, we propose the following three points for
designing more active hydrogenation catalyst under H2/CO
pressure: (1) bifunctional type hydrogenation catalyst; (2) less
electron-donating ligand on Ru; and (3) more basic functional
group to activate H2 on the ligand. Furthermore, here we found
a Ru-based tandem n-selective hydroformylation/hydrogena-
18753
dx.doi.org/10.1021/ja307998h | J. Am. Chem. Soc. 2012, 134, 18746−18757

Journal of the American Chemical Society
Article
1
proton (δ 8.0, s, CH2OCHO). The yields determined by H NMR
were consistent with those determined by GC.
tion, which affords n-alcohol with low reaction rate but high n/i
ratio. If a more active and recyclable catalyst is developed, these
systems will be attractive alternative for industrial oxo-
processes.
Real-Time IR Monitoring of Hydroformylation/Hydrogena-
tion of 1-Decene by Rh(acac)(CO)2/XANTPHOS/Shvo’s Cata-
lyst. An autoclave (100 mL) equipped with IR probe, high-pressure
dropping funnel and magnetic stir bar was charged with Rh(acac)-
(CO)2 (13.0 mg, 50 mmol), XANTPHOS (57.8 mg, 100 mmol). After
flushed with Ar, DMA (2 mL) was added via syringe to the autoclave.
Shvo’s catalyst (67.8 mg, 125 μmol (Ru)) was charged into 20 mL
Shlenck under Ar and was dissolved in DMA (3.0 mL). Then the
solution was transferred to the autoclave by cannulation, the Schlenk
was washed two times with DMA (total 1.0 mL), and they were
transferred to the autoclave. At the same time, the dropping funnel was
charged with 1-decene (1.0 mL, ∼5.3 mmol) and DMA (3.0 mL). The
autoclave was pressurized with 2.0 MPa of H2/CO and stirred at 120
°C and 800 rpm for 1,5 h. Then the mixture of 1-decene and DMA in
dropping funnel was pressed in to the autoclave with 3 MPa of H2/
CO, and the gas pressure was partially released to the value before
substrate injection. The integration of the characteristic peaks for 1-
decene (912 cm−1, terminal CC), undecanal (1726 cm−1, CO),
and undecanol (1058 cm−1, C−O) were monitored during the
reaction time. After appropriate reaction time, the autoclave was
cooled with water/ice bath for 30 min, and the pressure was released.
Dodecane (500.0 mg, 2.945 mmol) was added to the crude solution.
Then the solution was analyzed by GC.
EXPERIMENTAL SECTION
■
General. All the manipulations involving the air- and moisture-
sensitive compounds were carried out by using standard Schlenk
technique or glovebox under argon purified by passing through a hot
column packed with BASF catalyst R3-11. H2/CO mixed gas (H2:CO
= 49.1:50.9) was purchased from Suzuki-Shoukan and used without
further purification. Commercially available anhydrous N,N-dimethy-
lacetamide, methanol, and 2-propanol were distilled and degassed by
freeze−pump−thaw before use. Commercially available anhydrous
toluene was passed through solvent purification column before use. 1-
decene, dodecane, tridecane, allyl alcohol, allyl acetate, 3-butenyl
alcohol, 3-butenyl acetate, and 4-pentenyl alcohol were purchased
from TCI and distilled and degassed by freeze−pump−thaw before
use. Undecanal styrene, vinylcyclohexane, 2-methyl-1-nonene, and
(Z)-2-decene were purchased from TCI and degassed by freeze−
pump−thaw before use. Ru(CO)H2(PPh3)3 was purchased from TCI.
Cp*Ru(cod)Cl was purchased from Strem. Rh(acac)(CO)2 and
Ph2PCH2CH2NH2 were purchased from Aldrich. Shvo’s catalyst was
prepared according to literature method from Ru3(CO)12 and
tetraphenylcyclopentadienone and purified by recrystallization from
toluene/hexane. 2-(5-hexen-1-yloxy)-tetrahydropyran,28 (5-hexen-1-
yloxy)methylbenzene,29 (5-hexen-1-yloxy)-tert-butyldimethylsilane,30
2-(9-decen-1-yl)-1,3-dioxolan,31 (5-hexen-1-yl)-N-phenylcarbamate,32
XANTPHOS,10 and 733 were prepared by the literature method.
Cp*Ru(xantphos)H (10) and Cp*Ru(xantphos)Cl (11) were
prepared as a previously reported procedure by us.26 Product yields
were determined by Shimadzu GC-2014 equipped with InertCap
5MS/Sil capillary column (0.25 ID, 0.25 μm df, 30 m) using
calibration curve made with dodecane or tridecane as an internal
standard. Real-time IR measurement was performed by using Mettler
Toledo ReactIR 45 and analyzed by icIR. NMR spectra were recorded
on a JEOL JIN-ECP500 or JEOL-ECS400 spectrometers. Chemical
shifts are reported in ppm relative to the residual protiated solvent for
1H and external 85% H3PO4 for 31P nuclei. Data are presented in the
following space: chemical shift, multiplicity (s = singlet, d = doublet, t
= triplet, q = quartet, m = multiplet, br = broad), coupling constant in
hertz (Hz), and signal area integration in natural numbers. NMR yields
The actual amount of substrate injected into the autoclave was
estimated as sum of the observed product with GC analysis. The actual
liquid volume was estimated with the following equation:
(actual liquid volume)
= (initial charge of solvent) + (mixture of solvent and substrate
charged via dropping funnel)
= 7 + 4 × (mmol of the substrate injected into the autoclave)
/(mmol of the substrate charged into the dropping funnel)
Data treatment of IR was as follows: Background was measured
before experiment under air. During the reaction, the peak area for 1-
decene (912 cm−1, terminal CC), undecanal (1726 cm−1, CO),
and undecanol (1058 cm−1, C−O) were plotted versus time (t) every
15 s (64 scans were integrated) for initial 1.5 h and every 5 min (256
scans were integrated) after that time. Signal to noise ratio of these
peaks of compounds at concentration of 0.32 M in DMA was ∼40, 60,
and 70 respectively, which supports the accuracy of the integral value.
The consumption of 1-decene until 95% conversion was monitored to
confirm the first-order kinetics. The obtained pseudofirst-order rate
constant was multiplied by the selectivity to aldehyde to calculate rate
constant for hydroformylation. Since the increase of 1-undecanol was
linear versus time, the observed rate constant was calculated from the
slope.
1
were determined by H experiment with 15 s relaxation delay using
1,3,5-trimethoxybenzene as internal standard. IR spectra of solid
sample were recorded by Shimadzu FTIR-8400. X-ray crystallographic
analyses were performed on Rigaku Mercury CCD or Valimax Saturn
diffractometer. Elemental analysis was performed by the Micro-
analytical Laboratory, Department of Chemistry, Graduate School of
Science, the University of Tokyo.
General Procedure for Hydroformylation/Hydrogenation of
Alkene. DMA (1.0 mL) was added to a stainless autoclave (50 mL)
charged with Rh(acac)(CO)2 (5.2 mg, 20 μmol), XANTPHOS (23.1
mg, 40.0 μmol), and magnetic stir bar under Ar, and the resulting
mixture was stirred for 5 min at room temperature. Shvo’s catalyst
(27.1 mg, 50.0 μmol (Ru)) was weighed and dissolved in DMA (2.0
mL) under Ar, which was transferred to the autoclave by cannulation.
A 2:1 mol ratio mixture of alkene (2.0 mmol) and internal standard
(1.0 mmol) was added via syringe. The autoclave was pressurized with
2.0 MPa of H2/CO and stirred at 120 °C, at 800 rpm for 12.5 h. When
the autoclave was cooled with water/ice bath for 30 min, the pressure
was released. 1,3,5-trimethoxybenzene (100 mg, 0.590 μmol) was
added to the crude solution. Then the solution was analyzed by GC
As experimental error, the amount of injected substrate 5%, H2/
CO pressure 2.5%, volume of liquid 1.0%, the amount of weighed
catalyst <0.8%, was considered ( 9.6% in total). Statistical error was
respectively determined as standard deviation from obtained data and
its least-squares fitting curve. The total error (%) was calculated as
multiple of experimental and statistical error.
Real-Time IR Monitoring of Hydroformylation of 1-Decene
by Rh(acac)(CO)2/XANTPHOS. An autoclave (100 mL) equipped
with IR probe and high pressure dropping funnel was charged with
Rh(acac)(CO)2 (13.0 mg, 50 μmol) and XANTPHOS (57.8 mg, 100
μmol), and magnetic stir bar was flushed with Ar. The IR monitoring
was started at this point. DMA (7.0 mL) was added, and the resulting
mixture was stirred for 5 min at room temperature. At the same time,
the dropping funnel was charged with 1-decene (1.0 mL, 5.3 mmol)
and DMA (3.0 mL). The autoclave was pressurized with 2.0 MPa of
H2/CO and stirred at 120 °C and 800 rpm for 1.5 h. Then the mixture
of 1-decene and DMA in dropping funnel was pressed into the
autoclave with 3 MPa of H2/CO, and the gas pressure was partially
released to the value before substrate injection. The concentration of
1
and H NMR. NMR yield of n- or i-aldehydes was determined from
the integration of corresponding formyl proton (δ 9.8, t, −CH2CHO,
and δ 9.6, d, −CHRCHO, respectively). NMR yield of n- or i-alcohols
were determined from the alpha-proton of hydroxyl group (δ 3.6, t,
−CH2CH2OH, and δ 3.4−3.5, m, −CHRCH2OH). NMR yield of
formats was determined by the integration of corresponding formyl
18754
dx.doi.org/10.1021/ja307998h | J. Am. Chem. Soc. 2012, 134, 18746−18757

Journal of the American Chemical Society
Article
1-decene and undecanal was monitored by the integration of the area
at 1-decene (912 cm−1, terminal CC) and undecanal (1726 cm−1,
CO) . After appropriate reaction time, the autoclave was cooled
with water/ice bath for 30 min, and the pressure was released.
Dodecane (500.0 mg, 2.945 mmol) was added to the crude solution.
Then the solution was analyzed by GC. Following data treatments
were similar to that mentioned above.
amount of Ru catalyst (125, 62.5, or 31.3 μmol) was charged into 20
mL Shlenck under Ar and dissolved in solvent (5.0 mL). An autoclave
(100 mL) equipped with IR probe and magnetic stir bar was charged
with appropriate amount of XANTPHOS (0, 62.5, 125, or 250 μmol)
and flushed with Ar. IR monitoring was started at this point. Then the
solution of Shvo’s catalyst was added to the autoclave by cannulation,
the Schlenk was washed two times with solvent (total 5.0 mL), and
they were transferred to the autoclave. A mixture of undecanal and
dodecane (2:1 mol ratio, 1.6 mL, 5.0 and 2.5 mmol) was introduced
into the autoclave via syringe and was immediately pressurized with
2.0 MPa of H2/CO and stirred at 120 °C and 800 rpm. The
integration of the characteristic peaks for undecanol (1058 cm−1, C−
O) was monitored during the reaction time. After appropriate reaction
time, the autoclave was cooled with water/ice bath for 30 min, and the
pressure was released. Dodecane (500.0 mg, 2.945 mmol) was added
to the crude solution. Then the solution was analyzed by GC.
Following data treatments were similar to that mentioned above
except that the rate constants were determined from the time course of
alcohol in initial 200 min.
Real-Time IR Monitoring of Isomerization of 1-Decene by
Shvo’s Catalyst. Shvo’s catalyst (67.8 mg, 125 μmol) was charged
into 20 mL Shlenck under Ar and dissolved in DMA (5.0 mL). An
autoclave (100 mL) equipped with IR probe and magnetic stir bar was
flushed with Ar. IR monitoring was started at this point. Then the
solution of Shvo’s catalyst was added to the autoclave by cannulation,
the Schlenk was washed two times with DMA (total 5.0 mL), and they
were transferred to the autoclave. At the same time, the dropping
funnel was charged with 1-decene (1.0 mL, c.a. 5.3 mmol) and DMA
(3.0 mL). The autoclave was pressurized with 2.0 MPa of H2/CO and
stirred at 120 °C and 800 rpm for 1.5 h. Then the mixture of 1-decene
and DMA in dropping funnel was pressed in to the autoclave with 3
MPa of H2/CO, and the gas pressure was partially released to the value
before substrate injection. The integration of the characteristic peaks
for 1-decene (912 cm−1, terminal CC) was monitored during the
reaction time. After appropriate reaction time, the autoclave was
cooled with water/ice bath for 30 min, and the pressure was released.
Dodecane (500.0 mg, 2.945 mmol) was added to the crude solution.
Then the solution was analyzed by GC. Initially the reaction rate was
first order on substrate concentration. The rate constant for the
consumption of 1-decene until 50% conversion was calculated from
the plot of ln(1 − [1-decene]/[1-decene]0) versus time. The rate
constant for isomerization was calculated as (rate constant for the
consumption of1-decene) × (selectivity to internal alkenes)
As experimental errors, the amount of injected substrate 1.0%,
H2/CO pressure 2.5%, volume of liquid 1.0%, the amount of
weighed catalyst <0.8% were considered ( 5.4% in total). Statistical
error was respectively determined as standard deviation from obtained
data and its least-squares fitting curve.
Treatment of Shvo’s Catalyst under H2/CO. Shvo’s catalyst (50
mg, 92 μmol) was charged into autoclave under Ar and dissolved in
toluene (2.0 mL). The autoclave was pressurized with 2.0 MPa of H2/
CO and stirred at 120 °C for 2 h. After cooled to room temperature,
the pressure was released, and the solution was transferred to glass vial
in groove box. Evaporation of the solvent yielded slightly yellowish
powder, which was confirmed to be almost pure Ru(CO)3(2,3,4,5-
1
Real-Time IR Monitoring of Hydrogenation of Undecanal by
Shvo’s Catalyst. Shvo’s catalyst (67.8 mg, 125 μmol) was charged
into 20 mL Shlenck under Ar and dissolved in DMA (5.0 mL). An
autoclave (100 mL) equipped with IR probe and magnetic stir bar was
flushed with Ar. IR monitoring was started at this point. Then the
solution of Shvo’s catalyst was added to the autoclave by cannulation,
the Schlenk was washed two times with DMA (total 5 mL), and they
were transferred to the autoclave. At the same time, the dropping
funnel was charged with undecanal (1.1 mL, c.a. 5.3 mmol) and DMA
(3.0 mL). The autoclave was pressurized with 2.0 MPa of H2/CO and
stirred at 120 °C and 800 rpm for 1.5 h. Then the mixture of 1-decene
and DMA in dropping funnel was pressed in to the autoclave with 3
MPa of H2/CO, and the gas pressure was partially released to the value
before substrate injection. The integration of the characteristic peaks
for undecanal (1726 cm−1, CO), and undecanol (1058 cm−1, C−O)
was monitored during the reaction time. After appropriate reaction
time, the autoclave was cooled with water/ice bath for 30 min, and the
pressure was released. Dodecane (500.0 mg, 2.945 mmol) was added
to the crude solution, and then the solution was analyzed by GC.
Following data treatments were similar to that mentioned above.
Stepwise Hydroformylation/Hydrogenation of 1-Decene.
DMA (2.0 mL) was added to a stainless autoclave (50 mL) charged
with Rh(acac)(CO)2 (5.2 mg, 20 μmol), XANTPHOS (23.1 mg, 40.0
μmol), and magnetic stir bar under Ar, and the resulting mixture was
stirred for 5 min at room temperature. A 2:1 mol ratio mixture of 1-
decene (2.0 mmol) and dodecane (1.0 mmol) was added via syringe.
The autoclave was pressurized with 2.0 MPa of H2/CO and stirred at
120 °C and 800 rpm for 1 h. Then the autoclave was cooled with
water/ice bath for 10 min, and the pressure was released. Shvo’s
catalyst (27.1 mg, 50.0 μmol (Ru)) was weighed and dissolved in
DMA (2.0 mL) under Ar, which was transferred to the autoclave by
cannulation. The autoclave was pressurized with 1.0 MPa of H2 and
stirred at 120 °C and 800 rpm for 1 h. Then the autoclave was cooled
with water/ice bath for 30 min, the pressure was released, and the
solution was analyzed by GC. Obtained products were n-alcohol 90%,
i-alcohol 3.6%, n-aldehyde 1.9%, i-alcehyde 0.5%, decane 1.0%, undecyl
formate 0.6%.
tetraphenylcyclopentadienone) by H NMR and IR spectroscopy.
Hydroformylation/Hydrogenation of 1-Decene by Ru
Singular Catalyst. To a stainless autoclave (50 mL) charged with
Ru complex (25.0 μmol), XANTPHOS (28.9 mg, 50.0 μmol) and
magnetic stir bar under Ar, toluene (2.0 mL) and 2:1 mol ratio mixture
of 1-decene and dodecane (total 300 μL, 1-decene 1.0 mmol,
dodecane 0.5 mmol) were added via syringe. The autoclave was
pressurized with 2.0 MPa of H2/CO and stirred at 160 °C and 800
rpm for 24 h. Then the autoclave was cooled with water/ice bath for
30 min, and the pressure was released. Then the solution was analyzed
by GC.
Preparation of Tricarbonyl(2,5-bis(ethoxycarbonyl)3,4-
diphenylcycopentadienone)ruthenium (8). To a 50 mL double-
necked round-bottomed flask containing Ru3(CO)12 (506.9 mg, 2.379
mmol(mol Ru)) and 2,5-bis(methoxycarbonyl)3,4-diphenylcyclopen-
tadienone (878.9 mg, 2.335 mmol), toluene 17 mL was added and
refluxed until starting materials were consumed as confirmed by TLC.
After cooled, the reaction mixture to room temperature, and the
solvent was evaporated. Desired product was recrystallized from
1
CHCl3/hexane to give yellow crystals (766.3 mg, yield 58.5%). H
NMR (CD2Cl2, 500 MHz) δ 0.97 (t, J = 7 Hz, 6H) 4.03 (dq, J = 16, 7
Hz, 2H), 4.05 (dq, J = 16, 7 Hz, 2H), 7.21−7.33 (m, 10H); 13C NMR
(CDCl3, 101 MHz) δ 13.4 (CH3), 61.3 (CH2), 70.8 (4°), 109.5 (4°),
128.0 (CH), 129.1 (CH), 131.1 (CH), 164.7 (4°), 170.8 (4°), 192.1
(4°); mp 165−169 °C (decomp.); IR (KBr, cm−1):1653 (s), 1709 (s),
1722 (s), 2002 (s), 2029 (s), 2100 (s). Anal. calcd for C26H20O8Ru: C,
55.61; H, 3.59. Found: C, 55.38; H, 3.61.
Preparation of Tricarbonyl(2,4-bis(trimethylsilyl)bicycle-
[3,3,0]octa-1,4-dien-3-one)ruthenium (9). To a 50 mL stainless
autoclave, 1,7-bis(trimethylsilyl)-hepta-1,6-diyne (970 μL, 3.3 mmol)
and trirutheniumdodecacarbonyl (700 mg, 1.095 mmol) are charged
with acetonitrile 50 mL. Then the autoclave was pressurized with CO
0.5 MPa, and the resulting mixture was stirred at 120 °C for 12 h. After
evaporation of the solvent, the residue was dissolved in CH2Cl2 and
passed through a short silica gel column. The volatiles of the filtrate
were evaporated, and then the residue was recrystallized from toluene
at −35 °C (1.011 g, yield 68.0%). 1H NMR (CDCl3, 500 MHz) δ 0.26
(s, 18H) 1.75−1.89 (m, 1H), 2.33 (m, 1H), 2.50−2.67 (m, 4H); 13C
NMR (CDCl3, 101 MHz) δ 0.07 (CH3), 25.8 (CH2), 70.8 (4°), 109.5
Real-Time IR Monitoring of Hydrogenation of Undecanal by
Various Ru Catalysts under Various Conditions. Appropriate
18755
dx.doi.org/10.1021/ja307998h | J. Am. Chem. Soc. 2012, 134, 18746−18757

Journal of the American Chemical Society
Article
(4°), 128.0 (CH), 129.1 (CH), 131.1 (CH), 164.7 (4°), 170.8 (4°),
192.1 (4°); mp;146−147 °C (decomp.), IR (KBr, cm−1):1609, 2006,
2070. Anal. calcd for C17H24O4RuSi2: C, 45.41; H, 5.38. Found: C,
45.25; H, 5.34.
Recrystallization of 11. Rutheniumu complex 11 was synthesized
according to our previous report.26 The solid material was dissolved in
THF, and hexane was allowed to slowly difusse into the solution to
give single crystals suitable for X-ray analysis.
(6) (a) Gordon, E. M.; Eisenberg, R. J. Organomet. Chem. 1986, 306,
C53. (b) Fukuoka, A.; Matsuzaka, H.; Hidai, M.; Ichikawa, M. Chem.
Lett. 1987, 941. (c) Mitsudo, T.; Suzuki, N.; Kobayashi, T.; Kondo, T.
J. Mol. Catal. A 1999, 137, 253. (d) Tominaga, K.-i.; Sasaki, Y. J. Mol.
Catal. A 2004, 220, 159. (e) Tominaga, K.-i.; Sasaki, Y. Chem. Lett.
2004, 33, 14. (f) Moreno, M. A.; Haukka, M.; J_skel_inen, S.; Vuoti,
S.; Pursiainen, J.; Pakkanen, T. A. J. Organomet. Chem. 2005, 690,
3803. (g) Moreno, M. A.; Haukka, M.; Turunen, A.; Pakkanen, T. A. J.
Mol. Catal. A 2005, 240, 7.
(7) (a) Drent, E.; Budzelaar, P. H.M. J. Organomet. Chem. 2000,
593−594, 211. (b) Konya, D.; Almeida Lenero, K. Q.; Drent, E.
Organometallics 2006, 25, 3166.
(8) Boogaerts, I. I. F.; White, D. F. S.; Cole-Hamilton, D. J. Chem..
Commun. 2010, 46, 2194.
(9) Fuchs, D.; Rousseau, G.; Diab, L.; Gellrich, U.; Breit, B. Angew.
Chem., Int. Ed. 2012, 51, 2178.
(10) Kranenburg, M.; Vanderburgt, Y. E. M.; Kamer, P. C. J.; van
Leeuwen, P.W. N. M.; Goubitz, K.; Fraanje, J. Organometallics 1995,
14, 3081.
(11) Diebolt, O.; Muller, C.; Vogt, D. Catal. Sci. Technol. 2012, 2,
773.
(12) Takahashi, K.; Yamashita, M.; Ichihara, T.; Nakano, K.; Nozaki,
ASSOCIATED CONTENT
■
S
* Supporting Information
Figures of plot of ln(1 − [1-decene]/[1-decene]0) versus time
for hydroformylation/hydrogenation, hydroformylation, and
isomerization of 1-decene, time course of formation of
undecanol by hydrogenation of undecanal by Shvo’s catalyst
under various conditions, and details for 8−11 by X-ray
crystallographic analysis. This material is available free of charge
via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
K. Angew. Chem., Int. Ed. 2010, 49, 4488.
Corresponding Author
nozaki@chembio.t.u-tokyo.ac.jp
(13) (a) Blum, Y.; Shvo, Y. Inorg. Chim. Acta 1985, 97, L25.
(b) Shvo, Y.; Czarkie, D. J. Organomet. Chem. 1986, 315, C25.
(c) Shvo, Y.; Czarkie, D.; Rahamim, Y.; Chodosh, D. F. J. Am. Chem.
Soc. 1986, 108, 7400. (d) Casey, C. P.; Singer, S.; Powell, D. R.;
Hayashi, R. K.; Kavana, M. J. Am. Chem. Soc. 2001, 123, 1090.
(e) Casey, C. P.; Johnson, J. B.; Singer, W. S.; Cui, Q. J. Am. Chem. Soc.
2005, 127, 3100. (f) Casey, C. P.; Strotman, N. A.; Beetner, S. E.;
Johnson, J. B.; Priebe, D. C.; Guzei, I. A. Organometallics 2006, 25,
1236. (g) Casey, C. P.; Beetner, S. E.; Johnson, J. B. J. Am. Chem. Soc.
2008, 130, 2285.
(14) (a) Grafje, H.; Kornig, W.; Werrz, H. M.; Stefan, G.; Diehi, H.;
Bosche, H.; Schneider, K.; Kieczka, H. Butanediols, Butenediol, and
Butynediol. In Ullmann’s Encyclopedia of Industrial Chemistry,
Electronic Release, 7 th ed.; Wiley-VCH: Weinheim, Germany,
2009. (b) Werle, P.; Morawietz, M.; Lundmark, S.; Sorensen, K.;
Karvinen, E.; Lehtonen, J. Alcohols, Polyhydric. In Ullmann’s
Encyclopedia of Industrial Chemistry, Electronic Release, 7 th ed.;
Wiley-VCH: Weinheim, Germany, 2009.
(15) Backvall, J. E.; Andreasson, U. Tetrahedron Lett. 1993, 34, 5459.
(16) Hydroformylation of internal alkene by Rh/XANTPHOS can
produce n-aldehyde via isomerization to terminal alkene and successive
n-selective hydroformylation. However the n/i was as low as 1.
(17) The consumption of 1-decene gradually got slower than first
order.
(18) For example in the ref 13f, hydrogenation of benzaldehyde (0.97
M) by Shvo’s catalyst (2.4−5.2 mM) under H2 (3.5 MPa) at 60 °C
gave the reaction rate of −d[benzaldehyde]/dt = −3.5 × 10−4 (mol/
L·s), which is 10 times faster than our case (−d[undecanal]/dt = −1.1
× 10−5 (mol/L·s), with undecanal 0.43 M, Shvo’s cat 11 mM, H2 1.0
MPa, CO 1.0 MPa at 120 °C). Considering the difference of
temperature (65 °C versus 120 °C), the reaction rate should 100−
1000 times slower in the presesnce of CO.
(19) Casey et al. reported negative effect of the presence of PPh3 on
the rate of hydrogenation of aldehyde at relatively high temperature
(>60 °C), but the rate was still first order on aldehyde concentration
(ref 13f). Therefore, the change of the rate equation in our experiment
ascribed to the presence of CO.
(20) The stoichiometry of the observed species does not represent
the actual stoichiometry under catalytic condition, which employs
higher H2/CO pressure (2.0 MPa compared to 0.1 MPa in NMR
experiment.)
(21) As a similar example to 1, νCO for Ru(tetraphenylcyclopenta-
dienone)(CO)2(PPh3) was reported. (2037, 2011, 1981, 1955 cm−1).
The lower νCO for this compound than 2 (2081, 2026, 2005 cm−1)
indicates the cooridination of triaryl phosphine makes Ru−CO bond
stronger. For Ru(tetraphenylcyclopentadienone)(CO)2(PPh3), see:
Present Address
†Department of Applied Chemistry, Faculty of Science and
Engineering, Chuo University 1−13−27, Kasuga, Bunkyo-ku,
112−8551, Tokyo (Japan)
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by Funding Program for Next
Generation World-Leading Researchers, Green Innovation, and
Mitsubishi foundation. K.T. is grateful to the Japan Society for
the Promotion of Science (JSPS) for a Research Fellowship for
Young Scientists.
REFERENCES
■
(1) Fable, J.; Bahrmann, B.; Lipps, W.; Mayer, D. Alcohols, Aliphatic.
In Ullmann’s Encyclopedia of Industrial Chemistry, Electronic Release, 7
th ed.; Wiley-VCH: Weinheim, Germany, 2009.
(2) Haggin, J. Chem. Eng. News 1993, 71, 23.
(3) Dong, G.; Teo, P.; Wickens, Z. K.; Grubbs, R. H. Science 2011,
333, 1609.
(4) (a) Slaugh, L. H.; Hill, P.; Mullineaux, R. D. Shell Oil Company,
U.S. Patent 3,239,569, 1966. (b) Slaugh, L. H.; Mullineaux, R. D. J.
Organomet. Chem. 1968, 13, 469. (c) van Winkle, J. L.; Lorenzo, S.;
Moris, R. C.; Mason, R. F. Shell Oil Company, U.S. Patent 3,420,898,
1969. (d) Alvila, L.; Pakkanen, T. A.; Pakkanen, T. T.; Krause, O. J.
Mol. Catal. 1992, 71, 281. (e) Bartik, T.; Bartik, B.; Hanson, B. E. J.
Mol. Catal. 1993, 85, 121. (f) Wong, P. K.; Moxey, A. A. Shell Oil
Company, U.S. Patent 6,114,588, 2000. (g) Crause, C.; Bennie, L.;
Damoense, L.; Dwyer, C. L.; Grove, C.; Grimmer, N.; Rensburg, W. J.
V.; Kirk, M. M.; Mokheseng, K. M.; Otto, S.; Steynberg, P. J. Dalton
Trans. 2003, 2036.
(5) (a) MacDougall, J. K.; Cole-Hamilton, D. J. J. Chem. Soc. Chem.
Commun. 1990, 165. (b) MacDougall, J. K.; Simpson, M. C.; Green,
M. J.; Cole-Hamilton, D. J. J. Chem. Soc. Dalton 1996, 1161.
(c) Sandee, A. J.; Reek, J. N. H.; Kamer, P. C. J.; van Leeuwen, P. W.
N. M. J. Am. Chem. Soc. 2001, 123, 8468. (d) Ropartz, L.; Morris, R.
E.; Foster, D. F.; Cole-Hamilton, D. J. J. Mol. Catal. A 2002, 182, 99.
(e) Solsona, A.; Suades, J.; Mathieu, R. J. Organomet. Chem. 2003, 669,
172. (f) Ichihara, T.; Nakano, K.; Katayama, M.; Nozaki, K. Chem.
Asian J. 2008, 3, 1722. (g) Diab, L.; Mejkal, K. M.; Geier, J.; Breit, B.
Angew. Chem., Int. Ed. 2009, 48, 8022.
18756
dx.doi.org/10.1021/ja307998h | J. Am. Chem. Soc. 2012, 134, 18746−18757

Journal of the American Chemical Society
Article
Yamazaki, S.; Taira, Z. J. Organomet. Chem. 1999, 578, 61 For 2, see ref
13a..
(22) See Figure S7 in SI.
(23) Ito, M.; Hirakawa, M.; Osaku, A.; Ikariya, T. Organometallics
2003, 22, 4190.
(24) Multiple species are possibly formed under H2/CO. There is a
report isolating H4Ru4(CO)12 by treating Ru3(CO)12 under high H2/
CO pressure: Piacenti, F.; Bianchi, M.; Frediani, P.; Benedetti, E.
Inorg. Chem. 1971, 10, 2759. It was proposed that Ru(CO)H2(PPh3)3
exists as Ru(CO)2H2(PPh3)2 under H2/CO: Delgado, R. A. S.;
Bradley, J. S.; Wilkinson, G. J. Chem. Soc., Dalton Trans. 1976, 399.
(25) Johnson, T. J.; Folting, K.; Streib, W. E.; Martin, J. D.; Huffman,
J. C.; Jackson, S. A.; Eisenstein, O.; Caulton, K. G. Inorg. Chem. 1995,
34, 488.
(26) Takahashi, K.; Yamashita, M.; Tanaka, Y.; Nozaki, K. Angew.
Chem., Int. Ed. 2012, 51, 4383.
(27) We obtained single crystal of Cp*Ru(xantphos)H, however the
result of X-ray single crystal analysis of it was not fully solved due to
the disorder of solvent molecules. Nevertheless, the bidentate ligation
of XANTPHOS was confirmed (see SI).
(28) Casey et al. isolated Ru(tetraarylcyclopentadienone)
(PPh3)2(CO) (with X-ray single crystal analysis), which supports the
possibility of coordination of two triaryl phosphine ligand to Ru in
spite of steric repulsion with tetraarylcyclopentadienone: Casey, C. P.;
Stroman, N. A.; Beetner, S. E.; Johnson, J. B.; Priebe, D. C.; Guzei, I.
A. Organometallics 2006, 25, 1236.
(29) Sabitha, G.; Swapna, R.; Reddy, E. V.; Yadav, J. S. Synthesis
2006, 24, 4242.
(30) Rawat, V.; Chouthaiwale, P. V.; Suryavanshi, G.; Sdalai, A.
Tetrahedron Asym. 2009, 20, 2173.
(31) Rotulo-Sims, D.; Prunet, J. Org. Lett. 2002, 4, 4701.
(32) Lipshutz, B. H.; Ghorai, S.; Leong, W. W. Y.; Taft, B. R. J. Org.
Chem. 2011, 76, 5061.
(33) Breit, B.; Seiche, W. J. Am. Chem. Soc. 2003, 125, 6608.
(34) Mavrynsky, D.; Sillanpaa, R.; Leino, R. Organometallics 2009, 28,
598.
18757
dx.doi.org/10.1021/ja307998h | J. Am. Chem. Soc. 2012, 134, 18746−18757