Lewis acid promoted double bond migration in O-allyl to Z-products by Ru-H complexes

Article abstract of DOI:10.1016/j.mcat.2019.02.007

In catalytic double bond migration reaction, E-configuration olefins were normally generated as the dominant product because E-configuration was thermodynamically favored. However, Z-configuration products are sometimes desired in pharmaceutical chemistry owing to the structure-activity relationship. In this paper, we have demonstrated a new strategy that Lewis acid promoted an widely employed and convenient ruthenium(II) complex for the catalytic isomerization of O-allylethers, leading to thermodynamic-unfavored Z-product under mild conditions. The model substrate of allyl phenyl ether can be simply scaled up to 20 mmol to produce Z-product with TON of 2453 and TOF of 13,430 h^?1 at 40–60 °C. The system of Ru(II)/Lewis Acid catalysts was suitable for various substituted O-allylethers and other types of substrates. Through mechanism study including kinetic study, ligand inhibition effect and molecular spectroscopy, the dissociation of PPh₃ ligand by the addition of Lewis acid, and the formation a five-membered Ru complex from anchimeric assistance were both recognized as essential steps to improve the reactivity and to control the stereoselectivity of catalytic double bond migration reaction through metal hydride addition-elimination mechanism. This new strategy may provide a new opportunity to produce thermodynamic-unfavored product in heterocyclic compounds for pharmaceutical chemistry.

Full text of DOI:10.1016/j.mcat.2019.02.007

MolecularCatalysis469(2019)10–17
Contents lists available at ScienceDirect
Molecular Catalysis
journal homepage: www.elsevier.com/locate/mcat
Lewis acid promoted double bond migration in O-allyl to Z-products by Ru-
H complexes
Haibin Wang, Shaodong Liu, Tingting Sun, Zhanao Lv, Zhen Zhan, Guochuan Yin⁎, Zhuqi Chen⁎
Key Laboratory of Material Chemistry for Energy Conversion and Storage, Ministry of Education, Hubei Key Laboratory of Material Chemistry and Service Failure, School
of Chemistry and Chemical Engineering, Huazhong University of Science and Technology, Wuhan 430074, PR China
A R T I C L E I N F O
A B S T R A C T
Keywords:
In catalytic double bond migration reaction, E-configuration olefins were normally generated as the dominant
product because E-configuration was thermodynamically favored. However, Z-configuration products are
sometimes desired in pharmaceutical chemistry owing to the structure-activity relationship. In this paper, we
have demonstrated a new strategy that Lewis acid promoted an widely employed and convenient ruthenium(II)
complex for the catalytic isomerization of O-allylethers, leading to thermodynamic-unfavored Z-product under
mild conditions. The model substrate of allyl phenyl ether can be simply scaled up to 20 mmol to produce Z-
product with TON of 2453 and TOF of 13,430 h−1 at 40–60 °C. The system of Ru(II)/Lewis Acid catalysts was
suitable for various substituted O-allylethers and other types of substrates. Through mechanism study including
kinetic study, ligand inhibition effect and molecular spectroscopy, the dissociation of PPh3 ligand by the addition
of Lewis acid, and the formation a five-membered Ru complex from anchimeric assistance were both recognized
as essential steps to improve the reactivity and to control the stereoselectivity of catalytic double bond migration
reaction through metal hydride addition-elimination mechanism. This new strategy may provide a new op-
portunity to produce thermodynamic-unfavored product in heterocyclic compounds for pharmaceutical chem-
istry.
Isomerization
Ruthenium complexes
Stereoselectivity
Lewis acid
Z-product
1. Introduction
glycoside), the main component of polygonum multiflorum plant,
which has the effects of anti-oxidation, hypolipidemic action, inhibition
Double bond migration has attracted great interests due to its ex-
tensive applications [1–3]. For instance, isomerization is an initial step
for the synthesis of vinyl ethers, which are important intermediates of
heterocyclic compounds [4], and also a vital procedure of the protec-
tion and deprotection of free OH groups [5]. Comparing with acid-base
catalysts that maybe caused the break of double bond or led to the
polymerization or olefin or isomerization [6–9], transition metals in-
cluding Pt [10], Cr [11], Fe [12], Co [13,14], Ni [15], Ru [16], Rh [17],
Pd [18,19] and Ir [20,21] complexes were widely studied as catalysts to
achieve the desired reactivity in the olefin isomerization. In most cases,
E-configuration olefins were normally generated as the dominant pro-
duct because E-configuration was thermodynamically favored [22].
In relation to this, Z-configuration products are sometimes desired
in pharmaceutical chemistry owing to the structure-activity relation-
ship. For instance, 3,5-dihydroxystilbene have the inhibitory effects on
the activity of mushroom tyrosinase, whereas the inhibitory effects of Z-
isomer was much stronger than that of corresponding E-isomer [23].
Another example is THSG (2,3,5,4′-tetrahydroxystilbene-2-O-β-D-
of osteoporosis and excessive proliferation of vascular endothelial cells.
Recent studies suggested that Z-THSG was much more effective than E-
THSG [24–26]. Therefore, in the catalytic isomerization reactions,
manipulating the selectivity of Z and E products in a desired way is of
great importance. As far as we know, to gain kinetically controlled Z-
alkenes is still a challenge. Only limited cases were reported including
neutral Rh(I) complex catalyzing isomerization of β,γ-unsaturated ke-
tones to α,β-unsaturated Z-ketones [27], rhodium-catalyzed pathway
from allylaziridines to stable Z-enamines [28], and in-situ formed Ru
complex for Z-selective isomerization of allylamides [29]. However,
effective catalysts with satisfying efficiency, selectivity, stability and
wide range of substrates were still in great demand, and the mechanism
based on which transition metal catalysts favored Z-product in iso-
merization needs to be disclosed.
Recently, the promotional effect to the catalytic property of tran-
sition metal complex through the addition of non-redox metal ions as
Lewis acid (L.A.) has attracted considerable attentions in diverse types
of reactions. In particular, we found the acceleration effect from Al(III)
⁎ Corresponding authors.
E-mail addresses: gyin@hust.edu.cn (G. Yin), zqchen@hust.edu.cn (Z. Chen).
https://doi.org/10.1016/j.mcat.2019.02.007
Received 28 November 2018; Received in revised form 31 January 2019; Accepted 12 February 2019
2468-8231/©2019ElsevierB.V.Allrightsreserved.

H. Wang, et al.
MolecularCatalysis469(2019)10–17
2.2. reaction procedure
2.2.1. General procedure for isomerization by RuH2(CO)(PPh3)3 / Lewis
acid
In a typical experiment, 0.002 mmol RuH2(CO)(PPh3)3 complex,
0.004 mmol non-redox metal salts and 0.4 mmol allyl phenyl ethers (or
more concentration for different ethers) in 2 mL solvent was mixed in a
glass tube. Then the reaction mixture was magnetically stirred at 60 ℃
in an oil bath for 30 min. After the reaction, the analysis was conducted
by GC using the internal standard method. GC data were colledted on
FL-5090 with XE-60. The colum temperature was programmed from
100 ℃ held for 5 min to 130 ℃ held for 3 min at a rate of 30 ℃/min.
The injector temperature was 250 ℃. Naphthalene was chosen as the
internal standard. Chloroform-d or toluene-d8 was selected as the re-
action solvent for qualitative analysis in 1H NMR. Control experiments
using RuH2(CO)(PPh3)3 or non-redox metal salts individually as the
catalyst were carried out in parallel. Reactions were performed at least
in triplicate, and average data were used in the results and discussion
section.
Scheme 1. (a) Chemical structure of RuH2(CO)(PPh3)3 catalyst and (b) general
isomerization reaction investigated in this work.
in oxidovanadium(IV) complexes catalyzed hydrogen atom abstraction
[30]. We also observed that Sc(III) can promote Pd(II)-catalyzed
Wacker-type oxidations even better than Cu(II) [31], enhance oxidative
coupling of indoles with olefins by the palladium(II) acetate catalyst
[32], and accelerate the dehydrogenation of saturated CeC bond by a
ruthenium catalyst [33]. Further mechanism studies revealed that the
Lewis acid can change the coordination of metal complex to form new
active intermediates or transient states, then enhance the activity of the
catalyst or even change the reaction pathway. Inspired by these studies,
transition metals/L.A. may be considered as an alternative approach to
modulate the desired product selectivity in isomerization reactions.
In this paper, Ru(II)/LA catalyst was reported for the isomerization
of O-allyl compounds (Scheme 1). We surprisingly found that kineti-
cally controlled Z-configuration was the main product for certain sub-
strates. as far as we know, this is the first report on Lewis acid promoted
isomerization that lead to the kinetically controlled product by Ru-H
catalyst. The toleration for extensive unsaturated compounds was ex-
plored, and remarkable performances could be found under mild con-
ditions. The influence of different redox-inactive metals and the ratio of
Ru(II)/L.A. was investigated, and gram scale experiments were tested.
Moreover, isomerization mechanism was disclosed through kinetic
study, HMR and spectroscopic characterization.
2.2.2. General procedure for kinetics study
In a typical experiment, 0.002 mmol RuH2(CO)(PPh3)3, 0.004 mmol
non-redox metal salts and 3 mmol ethers were added in 2 mL of solvent
in a glass tube. Then the reaction solution was magnetically stirred at
60 ℃ in an oil bath. The quantitative product analysis was conducted by
GC using the internal standard method at set intervals. Control ex-
periments using RuH2(CO)(PPh3)3 or non-redox metal salt alone as
catalyst were also carried out in parallel. Reactions were performed at
least in triplicate, and average data were used in the catalytic kinetics
results.
2.2.3. General procedure for FT-IR study
In a typical experiment, 0.002 mmol RuH2(CO)(PPh3)3, 0.004 mmol
non-redox metal salt were added in 5 mL of chloroform in a glass tube.
Then the mixture solution was magnetically stirred at 60 ℃ in the oil
bath for 10 min. After the reaction, the solvent was removed and the
product was dried for several minutes. Then the solid sample was de-
tected by FT-IR spectrum. Control experiments using RuH2(CO)(PPh3)3
or non-redox metal salts individually were also carried out in parallel.
2.2.4. General procedure for UV–vis study
In a typical experiment, in 10 mL of chloroform containing 0.1 mM
RuH2(CO)(PPh3)3, a certain amount of Lewis Acid were added. Then
the mixture were monitored by Agilent UV 8454 at 30 ℃ after mixed in
ten minutes.
2. Experimental section
2.1. Chemicals
2.3. Characterization equipment
Allyl phenyl ethers with substitutes were synthesized according to
literatures [34–36]. ((2-methylallyl)oxy)benzene, N,N-diallyl-2,2-di-
chloroacetamide and allyl glycidyl ether were purchased from adamas.
Allyl(phenyl)sulfane, N-allylaniline and pent-4-enoic acid were pur-
chased from Energy Chemistry. Allylbenzene was purchased from
Meryer. Methyl undec-10-enoate was purchased from Heowns. Ruthe-
nium(III) chloride hydrate, Yb(OTf)3 and Y(OTf)3 were purchased from
Accela ChemBio Co., Ltd. Sodium NaOTf, Mg(OTf)2, In(OTf)3, Sc(OTf)3
came from Aladdin. Other trifluoromethanesulfonates including Ca
(OTf)2, Zn(OTf)2, Ba(OTf)2 and Al(OTf)3 were purchased from Shanghai
DiBai Chemical Co. Triphenylphosphine came from Alfa Aesar.
Common solvents and inorganic magnesium salts were purchased from
Sinopharm Chemical Reagent Co., Ltd. Other chemical reagents were
commercially available and used without further purify unless other-
wise stated.
Gas chromatography-mass spectrometry (GC–MS) analysis was
performed on an Agilent 7890 A/5975C spectrometer. FT-IR spectra
were obtained on Bruker VERTEX70. 1H NMR spectra were collected on
a Bruker AV-400 using TMS as an internal reference. UV–vis analysis
was proceed on Agilent UV 8454.
3. Results and discussion
3.1. Lewis acid promoted isomerization by the Ru(II) catalyst
At the outset, allyl phenyl ether was employed as the model com-
pound for RuH2(CO)(PPh3)3-catalyzed isomerization. As shown in
Table 1, with the presence of 0.25% of RuH2(CO)(PPh3)3 alone as the
catalyst, the isomerization of allyl phenyl ether in toluene provided
only 36.7% conversion with 27.4% Z-products and 8.3% E-products
within 0.5 h (Table 1, entry 1). Introducing 0.5 mol% NaOTf showed no
influence on the conversion (Table 1, entry 2), and a further increase to
1.5 mol% NaOTf (Table 1, entry 3) was still sluggish, providing only
The ruthenium catalyst RuH2(CO)(PPh3)3 was synthesized ac-
cording to the literatures [20,37–41]. The detailed procedures for cat-
alyst synthesis was exhibited in the supporting information.
11

H. Wang, et al.
MolecularCatalysis469(2019)10–17
Table 1
a
Screening of Lewis acids for the isomerization of allyl phenyl ether
.
Entry
Lewis acid
Conv. (%)
Yield (%)
2a-Z
2a-E
1
–
36.7
27.4
33.8
24.9
27.6
34.5
52.9
47.8
49.6
90.9
92.6
99.8
trace
27.4
20.1
24.7
17.8
18.6
25.8
41.9
37.6
40.0
76.4
77.4
84.5
trace
8.3
6.4
8.0
5.8
6.0
7.7
10.1
9.3
8.6
12.1
12.5
13.2
trace
2
NaOTf
NaOTf
3b
4
Ca(OTf)2
Ba(OTf)2
Mg(OTf)2
Zn(OTf)2
Yb(OTf)3
Y(OTf)3
Sc(OTf)3
Al(OTf)3
In(OTf)3
In(OTf)3
5
6
7
8
9
10
11
12
13c
33.8% conversion. This phenomenon clearly supported that the addi-
tion of OTf− anion had no influence on the catalytic efficiency of the
isomerization of allyl phenyl ether by RuH2(CO)(PPh3)3. On the other
hand, the addition of 2 equiv. of Zn2+ obviously accelerates the iso-
merization, with 52.9% conversion, 41.0% Z-products and 10.1% E-
products (Table 1, entry 7). In this case Z/E ratio was increased to 4.1.
However, the promotional effect from other bivalents metal ions like
Ca2+, Ba2+ and Mg2+ was not that obvious (Table 1, entry 4–6), which
was attributed to their very limited solubility in toluene. Meanwhile,
adding trivalent metal ions like Y3+, Yb3+, Al3+ and Sc3+ can greatly
promote the catalytic efficiency of RuH2(CO)(PPh3)3, achieving 47.8%,
49.6%, 90.9% and 92.6% conversion, respectively (Table 1, entry
8–11). Particularly, in the case of In(OTf)3, the conversion of allyl
phenyl ether can be improved up to 99.8%. It was worthy to mention
that, 84.5% yield of Z-configuration product was collected in the Ru
(II)/In3+ system. Meanwhile, the yield of E-configuration, which was
the thermodynamically controlled product, was only 13.2%. Accord-
ingly the ratio of Z/E was found to be 6.4. In control experiments, In
(OTf)3 alone as the catalyst demonstrated no activity for isomerization
under identical conditions (Table 1, entry 13). Taking together, all
these experiments illustrated that Ru(II) or In3+ alone was sluggish for
the catalytic isomerization of ally phenyl ethers, but the grateful effi-
ciency and selectivity on Z-product can be obtained in In(OTf)3/
RuH2(CO)(PPh3)3 system. Additionally, the screening of solvent, tem-
perature and more control experiments under Ar were listed in Table
functional group tolerance for the present bimetallic Ru(II)/In3+ cata-
lyst were next examined for isomerization, and the results were sum-
marized in Scheme 2. Different substituents such as methyl, methoxyl,
halogens, nitryl, trifluoromethyl, aldehyde were tested. For the model
substrate allyl phenyl ether, the catalyst loading can be reduced to
0.04 mol% while 93.7% conversion could be achieved at 60 ℃ within
1 h, with the major product in Z-configuration (2a). It is worth em-
phasizing that allyl phenyl ether with both electron-donating (2b) and
withdrawing (2c, 2d, 2e, 2f) substituents in the para position could
proceed smoothly to provide the corresponding O-(1-propenyl) pro-
ducts at 50 ℃ within 0.5 or 1 h. For example, isomerization of allyl 4-
(tert-butyl)phenyl ether offered 93.5% conversion with 80.0% Z-pro-
ducts and 14.9% E-products at 50 ℃ for 0.5 h (2b), allyl 4-nitrylphenyl
ether offered 94.9% conversion with 80.9% Z-products and 10.5% E-
products at 50 ℃ for 1 h (2e). The similar phenomenon were found in
the ortho and meta position. For instance, isomerization of allyl 2-
chlorophenyl ether and allyl 2-bromophenyl ether gave 93.1% and
88.9% conversion at 50 ℃ for 1 h (2 h, 2i). Allyl 3-methylphenyl ether
and allyl 3-methoxyphenyl ether also proceed smoothly at 40 or 50 ℃
for 1 h (2 j, 2k). While for allyl 4-formylphenyl ether as substrates,
higher catalyst loading of 1% was needed to provide 99.1% conversion
(2 g). It was noteworthy that when multiple substituents were present
on the phenyl ring, the reaction was also effective, while a high se-
lectivity of Z-product was still observed (2l-2q). It was interesting to
find that when 1-(allyloxy)naphthalene was chosen for isomerization,
the catalyst loading can be reduced to even 0.1 mol%, and 91.2%
conversion with a Z/E ratio of 5.3was still achieved.
S1-S3 in supporting information.
a
Reaction conditions:
allyl phenyl ether (0.8 mmol), RuH2(CO)
(PPh3)3 (0.25 mol%), Lewis acid (0.5 mol%), toluene (2 mL) at 60 ℃ for
0.5 h. NaOTf (1.5 mmol%). In(OTf)3 alone as the catalyst.
Reaction conditions: RuH2(CO)(PPh3)3 (0.002 mmol), In(OTf)3
(0.004 mmol), toluene (2 mL). The value under the substrates were GC
yield and the value in parentheses were the yield of Z-product. The
detail condition were on Table S4 in supporting information.
In particular, the isomerization of 2a was further performed in a
scale of substrate/catalyst = 2500:1, in which a yield of 98.1% with
70.6% yield of Z-product would be found. The maximum TON reached
2453 with TOF at 13,431 h−1 (Table S5), which additionally demon-
strate the efficiency and selectivity of Ru(II)-H/LA in isomerization
reaction.
b
c
The effect of In(OTf)3/RuH2(CO)(PPh3)3 ratio on the isomerization
of allyl phenyl ether was further investigated, and results were showed
in Figure S1. One can see that increasing the ratio from 0 to 1 sharply
accelerate the reaction, evidenced by the increase of conversion from
36.7% to 89.8%. A maximum conversion of 99.8% were achieved when
2 equiv. of In(OTf)3 was introduced, while a further increase the ratio
expressed a negative effect, and this phenomenon was similar to our
previous reports of L.A. accelerated oxidation processes [31,42,43]. It
was worthy to mention that, the effect of In(OTf)3/RuH2(CO)(PPh3)3
ratio on the reactivity of catalytic isomerization was highly consistent
with the effect of this ratio on UV–vis spectra, which would be dis-
cussed in mechanism section (vide infra).
3.2. Reaction mechanism of isomerization by the Ru(II)-Lewis acid
In our previous study [30–33,42–44], it was observed that non-
redox metal ions as Lewis acid can modulate the reactivity of transition
With the optimized conditions in hand, the substrate scope and the
12

H. Wang, et al.
MolecularCatalysis469(2019)10–17
Scheme 2. Substrate scope for the Ru(II)/In3+-catalyzed isomerization.
metal catalysis in versatile homogeneous reactions because the pre-
sence of non-redox metal ions can substantially manipulate the co-
ordination structure of reactive species, thus further regulate the cata-
lytic activities, or even alter the reaction pathway. However, as far as
we know, this is the first report on Lewis acid promoted isomerization
that lead to the kinetically controlled product by Ru-H catalyst. Herein,
the mechanism study was conducted to disclose the reaction pathway.
Firstly, the interaction between RuH2(CO)(PPh3)3 and olefin sub-
strate was investigated. We previously found that the dissociation of
PPh3 ligand played the essential role due to the fact that this dis-
sociation in turn offered some certain coordination sites that free alkene
could interact with the ruthenium hydride species [45]. In this case, the
existence of extra PPh3 could inhibit the reaction. Herein, the kinetic of
allyl phenyl ether isomerization by RuH2(CO)(PPh3)3/In3+ system were
recorded through the externally addition of free PPh3, and clearly a
negative influence on the reaction rate was observed. More im-
portantly, the reciprocal of the rate constant and the concentration was
found as a linear relationship to the concentration of extra PPh3
(Fig. 1), which was in good agreement with our previous result and
other reports [44,46].
The Ru(II)-catalyzed isomerization of allyl phenyl ether were de-
tected by 1H NMR. In Fig. 2A, RuH2(CO)(PPh3)3 or In(OTf)3 alone
13

H. Wang, et al.
MolecularCatalysis469(2019)10–17
Fig. 1. The influence of extra PPh3 on isomerization. Reaction conditions: allyl
phenyl ether (0.6 mmol), RuH2(CO)(PPh3)3 (0.002 mmol), In(OTf)3
(0.004 mmol), PPh3 (0-0.02 mmol) in 2 mL toluene at 60 ℃ for 10 min.
Scheme 3. Generally accepted mechanisms for isomerization of olefins.
showed no signals in the region from 6.5 ppm to 4.2 ppm, while allyl
phenyl ether alone gave four clear signals, that is, 6.05 ppm for me-
thine, 5.42 and 5.27 ppm for terminal alkene, and 4.53 ppm for me-
thylene (Fig. 2A (c)). Meanwhile, with the addition of 1 equiv. of
RuH2(CO)(PPh3)3, these four typical signals remained without any
change, indicating the isomerization could not be initiated with
RuH2(CO)(PPh3)3 alone. On the other hand, the presence of 2 equiv. In
(OTf)3 led to several new signals, which once again demonstrated the
promotional effect from Lewis acid, and was highly consistent with
results in Table 1.
As shown in Fig. 2B, RuH2(CO)(PPh3)3 also gave two characterized
resonances with the same integral area in the region from -4.0 to
-10.0 ppm. These two resonances were attributed to the fact that two
hydride coordinated to the Ru center were not equal in chemical en-
vironment. One hydride was in the contraposition of CO (chemical shift
at -6.9 ppm) while the other one (chemical shift at -8.8 ppm) was
against PPh3. With the addition of In(OTf)3, a new Ru-H signal located
at -4.46 ppm was found, indicating the formation of a new species as
the adduct of Ru(II)-H/In(III). Meanwhile, signals at -8.8 ppm and
-6.9 ppm could still be found, suggesting that part of RuH2(CO)(PPh3)3
complex remained in the reaction system. Additionally, when allyl
phenyl ether (1a) was added, all resonances from Ru-hydride species
quickly vanished, which clearly indicated that the Ru(II)-H/In(III)
species was responsible for the accelerated isomerization reaction. This
process was supposed to be through the classic metal hydride addition-
elimination mechanism, during which free alkene firstly coordinated to
ruthenium hydride species, then insertion into the metal-hydride bond
yielded a secondary metal-alkyl, subsequent β-hydride elimination
yielded the isomerized product and regenerated the initial ruthenium
hydride catalyst (scheme 3 ). Therefore, the addition of allyl phenyl
ether would be responsible for the disappearance of the resonances
from ruthenium hydride complex.
To explore the interaction of Lewis acid and RuH2(CO)(PPh3)3,
UV–vis spectra were carried out and shown in Fig. 3A. RuH2(CO)
(PPh3)3 alone showed a characteristic shoulder band centered at around
335 nm in CHCl3. Noticeably, when different Lewis acid was added, the
shifting of this characteristic band from RuH2(CO)(PPh3)3 was different
in each case. 2 equiv. of NaOTf or Mg(OTf)2 as the additive slightly
lower the intensity of this peak (Fig. 3(B)), while the location of
shoulder band did not shift. On the contrast, the addition of 2 equiv. In
(OTf)3 significantly changed the adsorption spectra of RuH2(CO)
(PPh3)3, in which the intensity of 335 nm band decreased dramatically
while an obvious larger adsorption band raised in the region lower than
330 nm. This trend was highly consistent with previous results that the
enhancement on isomerization from Na+ or Mg2+ was neglect, while
the promotional effect from In(III) was remarkable (Table 1). Once
again, the distinctions in UV–vis spectra demonstrated the interaction
between In3+ and Ru-H complex.
Furthermore, the interaction between In(III) and Ru-H complex was
further disclosed by adding different amount of In(III) from 0.2 to 4
equiv. to RuH2(CO)(PPh3)3 (Fig. 3(C)). Obviously an isobestic point at
305 nm could be found, which further suggested that the addition of In
(III) led to the change of coordination sphere of Ru center, and a
transformation from RuH2(CO)(PPh3)3 to another reactive species of Ru
(III)/In(III) adduct was revealed. This observation was highly consistent
with previous 1H-NMR tests, in which a new Ru-hydride signal at
-4.46 ppm could be found with the addition of In(III). Additionally, this
change in UV–vis spectra became stable after the concentration of In
Fig. 2. 1H NMR studies on the reaction of allyl phenyl ether (1a) with RuH2(CO)(PPh3)3 and In(OTf)3.
14

H. Wang, et al.
MolecularCatalysis469(2019)10–17
Fig. 3. UV–vis spectra of RuH2(CO)(PPh3)3 and In(OTf)3 mixture in CHCl3 (Ru(II) 0.1 mM).
(III) reached 2 equiv., and the spectra of Ru(II) plus 2, 3 and 4 equiv. In
(III) overlapped with each other. To further testify this pattern, the
adsorption intensity at 335 nm was recorded along with the increasing
amount of In(III) (Fig. 3(D)), and a minimum adsorption intensity could
be reached with the addition of 2 equiv. In(III), and remained after the
further increase of In3+ dosage. From these data it was suggested that a
ratio of In(III):Ru(II) at 2:1 already achieved the equilibrium in the
transformation from RuH2(CO)(PPh3)3 to Ru(III)/In(III) adduct. Once
again, this observation was in good agreement with the fact that the
ratio of In(III):Ru(II) at 2:1 was determined as the optimized ratio in
catalytic isomerization reaction (Table 1 and Figure S1).
Aforementioned results indicated that the presence of Lewis acid
such as In(III) could induce the dissociation of PPh3 ligand and the
formation of Ru(III)/In(III) adduct as the reactive species, and then
catalyze the double bond migration through the classic metal hydride
addition-elimination mechanism. However, the binding site between
Ru-H center and Lewis acid was not clear yet. In FT-IR spectra, the
vibration carbonyl group in RuH2(CO)(PPh3)3 gave a characteristic
adsorption band around 1940 cm−1, and two week absorption from Ru-
H, which theoretically located at 1960 and 1900 cm−1 but was nor-
mally obscured by the strong signal from the carbonyl group. As shown
in Figure S2, along with the addition of increasing amount of In(OTf)3,
the original absorption of carbonyl at 1943 cm−1 was found shifting to
a new location at 1973 cm−1. This shifting suggested that Lewis acid
may bind and influence the carbonyl group in Ru-H complex through
the linkage of OC-Ru-H2—In(III) or H2-Ru−CO—In(III). If the addition
of the electron acceptor (In3+) was through the coordination with
oxygen in carbonyl group, it would imply an increase in electron den-
sity on the bridging carbonyl ligand, and then cause the decrease in
frequency. Alternatively, if this binding of electron acceptor was
through Ru-H, it would decrease the electron density on the terminal
carbonyl ligand and then cause the increase in frequency [44,47–49].
Thus, it was speculated that the Lewis acid interacted with ruthenium
species through the linkage of OC-Ru-H2—In(III), which led to the in-
creased wavenumbers of carbonyl group, rather than H2-Ru−CO— In
(III). This speculation of OC-Ru-H2—In(III) as the Ru(II)/In(III) adduct
was also in accordance with the 1H-NMR discussed before.
The catalytic kinetics of allyl phenyl ether isomerization gave fur-
ther information for the isomerization process. As revealed in Fig. 4B,
with RuH2(CO)(PPh3)3 and 2 equiv. In(OTf)3 as the catalyst, 96.7%
conversion with 74.6% Z-products and 20.8% E-products was observed
using allyl phenyl ether 2a as the substrate, and longer reaction time
did not change Z-E stereoselectivity, during which Z/E ratio remained
at 4.0 after 1 h. This phenomena revealed that the generation of Z and E
products were simultaneous, and no transformation between two pro-
ducts was observed in the isomerization of allyl phenyl ether. However,
under similar conditions, when RuH2(CO)(PPh3)3 plus Lewis acid were
conducted as the catalyst, the isomerization of eugenol (Fig. 4A) was
found with E products as the major one, while the transformation from
Z to E product was clearly observed from the kinetic observation: the
yield of Z products increased initially, but decreased slowly after 15 min
[44]. Therefore, distinguished stereoselectivity was found between allyl
phenyl ether and eugenol, which accordingly suggested that the dif-
ference in the structure of the substrate may be essential to this ste-
reoselectivity.
To further explore the mechanism of the reaction, other substrates
were also tested and the results were summarized in Scheme 4. In the
case of methyl substituted substrate 3, the isomerization was much
slower comparing with the case of model compound 2a, which was
probably attributed to the steric disturbing of the coordinative sphere to
Ru center. Therefore, longer reaction time and higher reaction tem-
perature were required to obtain 94.1 conversion and 93% yield. On the
other hand, when terminally di-substituted substrate 11 was tested
under the same conditions, isomerization product could not be found
due to the much stronger steric hindrance. To further testify the role of
oxygen in 2a, N-allyl phthalimide (5) was used as the substrate, and the
15

H. Wang, et al.
MolecularCatalysis469(2019)10–17
Fig. 4. Lewis acid accelerated isomerization kinetics by the ruthenium complex. Reaction condition: (A) eugenol (2 mmol), RuH2(CO)(PPh3)3 (0.002 mmol), Mg
(OTf)2 (0.004 mmol), EtOH (2 mL) at 80 ℃. (B) allyl phenyl ether (3 mmol), RuH2(CO)(PPh3)3 (0.002 mmol), In(OTf)3 (0.004 mmol), toluene (2 mL) at 60 ℃.
metal ions such as In(III) influenced the coordination sphere of Ru and
formed an adduct as OC-Ru-H2—In(III). The interaction between Ru(II)
and In(III) twisted of the first coordination sphere of ruthenium center,
and induced the dissociation of PPh3 ligand that favored the formation
of unoccupied coordination site for alkene; On the other hand, the
strain associated with Ru-H bond could be relieved by invoking a In(III)
Ru(II)-bridged species that benefitted the isomerization of alkene.,
Therefore, significantly higher activity of OC-Ru-H2—In3+ species
could be found comparing with RuH2(CO)(PPh3)3. Complete iso-
merization reaction was through metal hydride addition-elimination
mechanism, including the free alkene coordinates to ruthenium hydride
species, then insertion into the metal-hydride bond yields a secondary
metal-alkyl, subsequent β-hydride elimination yields the isomerized
product and regenerates the initial ruthenium hydride catalyst. During
this process, allyl phenyl ether could offer one extra oxygen atom,
which was also involved into the first coordination sphere of central Ru
(pathway A). In this case, because the oxygen atom would turn to O-(1-
propenyl) phenyl ether in the secondary metal-alkyl. Therefore, Z-
configuration olefin would be the main product afterβ-hydride elim-
ination. However, in the case of other olefins such as allylbenzene, E-
Scheme 4. Substrates scope of RuH2(CO)(PPh3)3/In(OTf)3 catalyzed iso-
merization.
configuration olefin as the thermodynamic favored one would become
the dominant product (pathway B).
isomerization reaction generated 6-E as main product. When allylben-
zene (7) was employed as substrate, the selectivity still followed the
trend that thermodynamically favored E-configuration as the main
product. Additionally, substrate 9 and 10 with S and N instead of a
oxygen atom also failed in the isomerization reaction. These experi-
ments thus demonstrated the essential role of the oxygen atom for the
selectivity. During the reaction, the oxygen atom may coordinate to the
central Ru to form five- or six-membered ring, and this anchimeric
assistance can not only accelerate reaction, but also stable the inter-
mediate configuration to form thermodynamic-unfavored product.
Further more we added little EtOH in Ru(II)-In3+ system, we found
EtOH delayed the reaction and reduced the Z-E ratio. When the amount
of EtOH and toluene almost the same, the Z-product and E-product are
very close, also the conversation is reduced (Table S6). These results
were proportional to the O-Ru coordination in this system.
4. Conclusions
In summary, we have demonstrated a new strategy that Lewis acid
promoted the efficiency and the stereoselectivity of widely employed
and convenient ruthenium(II) catalysts for the isomerization of O-allyl
ethers. More importantly, thermodynamic-unfavored Z-products were
generated dominantly. The model substrate of allyl phenyl ether can be
simply scaled up to 20 mmol to achieve TON of 2453 and TOF of 13430
h−1 under mild conditions. This Ru(II)/L.A. catalyst system was sui-
table for various substituted O-allyl ethers and different types of sub-
strates. Due to the addition of L.A., the dissociation of PPh3 ligand was
recognized as the initial step, followed by which the formation a five-
membered Ru complex was suggested as the key intermediate in the
isomerization process. The anchimeric assistance in this five-membered
intermediate was essential in controlling the configuration of the pro-
duct. Through metal hydride addition-elimination mechanism, ther-
modynamic-unfavored Z-configuration product was observed as the
dominant product.
Taken together all the data from catalytic reactions and character-
izations, a possible reaction pathway was proposed in Scheme 5.
RuH2(CO)(PPh3)3 possessed a saturated coordination sphere in its first
coordination sphere, which caused the steric hindrance for free alkene
to coordinate, and was accordingly demonstrated with limited catalytic
reactivity in double bond migration by itself. The addition of non-redox
16

H. Wang, et al.
MolecularCatalysis469(2019)10–17
Scheme 5. Proposed mechanism for Ru(II)/Mn+-catalyzed isomerization.
Acknowledgements
[18] S. Kang, H. Lee, S. Jang, T. Kim, S. Pyun, J. Org. Chem. 61 (2017) 2604–2605.
[19] D. Ojha, K. Gadde, K. Prabhu, J. Org. Chem. 82 (2017) 4859–4865.
[20] B. Neugnot, J. Cintrat, B. Rousseau, Tetrahedron. 60 (2004) 3575–3579.
[21] K. Singh, S. Staig, J. Weaver, J. Am. Chem. Soc. 136 (2014) 5275–5278.
[22] K. Nakashima, T. Hirota, K. Obara, M. Shimizu, S. Doi, K. Fujita, T. Shirakawa,
T. Enomoto, S. Yoshihara, M. Ebisawa, K. Matsumoto, H. Saito, Y. Suzuki,
Y. Nakamura, M. Tamari, Biochem. Biophys. Res. Commun. 344 (2006) 300–307.
[23] T. Yi, K. Leung, G. Lu, H. Zhang, K. Chan, Phytochem. Anal. 18 (2007) 181–187.
[24] C. Larsen, D. Grotjahn, J. Am. Chem. Soc. 134 (2012) 10357–10360.
[25] S. Krompiec, M. Krompiec, R. Penczek, H. Ignasiak, Coord. Chem. Rev. 252 (2008)
1819–1841.
This work was supported by the National Natural Science
Foundation of China (No. 21671072 and 21573082). NMR and GC–MS
tests were performed in the Analytical and Testing Center of Huazhong
University of Science and Technology.
Appendix A. Supplementary data
[26] S. Yip, C. Aissa, Angew. Chem, Int. Ed. 54 (2015) 6870–6873.
[27] L. Zhuo, Z. Yao, Z. Yu, Org. Lett. 15 (2013) 4634–4637.
[28] F. Alphonse, A. Yudin, J. Am. Chem. Soc. 128 (2006) 11754–11755.
[29] S. Krompiec, M. Pigulla, M. Krompiec, S. Baja, J. Mrowiec-Bialonb, J. Kasperczyk,
Tetrahedron Lett. 35 (2004) 5257–5261.
[30] J. Zhang, H. Yang, T. Sun, Z. Chen, G. Yin, Inorg. Chem. 56 (2017) 834–844.
[31] S. Qin, L. Dong, Z. Chen, S. Zhang, G. Yin, Dalton Trans. 44 (2015) 17508–17515.
[32] S. Zhang, Z. Chen, S. Qin, C.Lou.A. Senan, R. Liao, G. Yin, Org. Biomol. Chem. 14
(2016) 4146–4157.
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.mcat.2019.02.007.
References
[1] R. Drift, E. Bouwman, E. Drent, J. Organomet. Chem. 650 (2002) 1–24.
[2] G. Long, E. Weitz, J. Am. Chem. Soc. 122 (2000) 1431–1442.
[3] J. Lummiss, K. Oliveira, A. Pranckevicius, A. Santos, E. Santos, D. Fogg, J. Am.
Chem. Soc. 134 (2012) 18889–18891.
[4] B. Schmidt, S. Hauke, N. Mühlenberg, Synthsis. 46 (2014) 1648–1658.
[5] F. Guibé, Tetrahedron 54 (1998) 2967–3042.
[6] F. Asinger, B. Fell, G. Collin, Chem. Ber. 96 (1963) 716–735.
[7] G. Tasic, M. Simic, S. Popovic, S. Husinec, V. Maslak, V. Savic, Tetrahedron Lett. 54
(2013) 4536–4539.
[8] C. Su, P. Williard, Org. Lett. 12 (2010) 5378–5381.
[9] S. Martinez-Erro, A. Sanz-Marco, A. Cómez, A. Vázquez-Romero, M. Ahlquist,
B. Martín-Matute, J. Am. Chem. Soc. 138 (2016) 13408–13414.
[10] A. Scarso, M. Colladon, P. Sgarbossa, C. Santo, R. Michelin, G. Strukul, Organomet.
29 (2010) 1487–1497.
[11] H. Yamada, M. Sodeoka, M. Shibasaki, J. Org. Chem. 56 (1991) 4569–4574.
[12] R. Corriu, V. Huynh, J. Moreau, M. Pataud-sat, J. Organomet. Chem. 255 (1983)
359–364.
[13] A. Schmidt, A. Nödling, G. Hilt, Angew. Chem. Int. Ed. 54 (2015) 801–804.
[14] H. Kumobayashi, S. Akutagawa, S. Otsuka, J. Am. Chem. Soc. 100 (1978)
3949–3950.
[33] Z. Lv, W. Zheng, Z. Chen, Z. Tang, W. Mo, G. Yin, Dalton Trans. 45 (2016)
11369–11383.
[34] H. Murakami, T. Minami, F. Ozawa, J. Org. Chem. 69 (2004) 4482–4486.
[35] R. Bujok, M. Bieniek, M. Masnyk, A. Micgtowska, A. Sarosiek, H. Stepowska, D. Arlt,
K. Grela, J. Org. Chem. 69 (2004) 6894–6896.
[36] B. Trost, A. Martos-Redruejo, Org. Lett. 11 (2009) 1071–1074.
[37] N. Ahmad, J. Levison, S. Robinson, M. Uttley, Inorg. Synth. 15 (1974) 45–63.
[38] N. Owston, A. Parker, J. Williams, Chem. Commun. (Camb.) 39 (2008) 624–625.
[39] H. Du, Q. Liu, S. Shi, S. Zhang, J. Organomet. Chem. 627 (2001) 127–131.
[40] C. Yue, Y. Liu, R. He, J. Mol. Catal. A Chem. 259 (2006) 17–23.
[41] W. Morris, M. Shair, Org. Lett. 11 (2009) 9–12.
[42] A. Senan, S. Qin, S. Zhang, C. Lou, Z. Chen, R. Liao, G. Yin, ACS Catal. 6 (2016)
4144–4148.
[43] C. Lou, S. Qin, S. Zhang, Z. Lv, A. Senan, Z. Chen, G. Yin, Catal. Commun. 90
(2017) 5–9.
[44] Z. Lv, Z. Chen, Z.Y. Hu, W. Zheng, H. Wang, W. Mo, G. Yin, ChemCatChem 9 (2017)
3849–3859.
[45] S. Muthaiah, S. Hong, Adv. Synth. Cataly. 354 (2012) 3045–3053.
[46] A. Ledger, P. Slatford, J. Lowe, M. Mahon, M. Whittlesey, J. Williams, Dalton Trans.
4 (2009) 716–722.
[47] N. Nelson, N. Kime, D. Shriver, J. Am. Chem. Soc. 91 (1969) 5173–5174.
[48] A. Alich, N. Nelson, D. Shriver, J. Chem. Soc. D: Chem. Commun. 6 (1971) 254–255.
[49] J. Kristoff, N. Nelson, D. Shriver, J. Organomet. Chem. 49 (1973) C82–C84.
[15] L. Wang, C. Liu, R. Bai, Y. Pan, A. Lei, Chem. Commun. (Camb.) 49 (2013)
7923–7925.
[16] S. Manzini, D. Nelson, S. Nolan, ChemCatChem 5 (2013) 2848–2851.
[17] R. Gorman, M. Little, J. Morris, V. Sridharan, Chem.Commun. 48 (2012)
9537–9539.
17