Ruthenium hydride-promoted dienyl isomerization: Access to highly substituted 1,3-dienes

Article abstract of DOI:10.1021/ja4011207

Ruthenium hydrides were found to promote the positional isomerization of 1,3-dienes into more highly substituted 1,3-dienes in a stereoconvergent manner. The reaction can be conducted in one pot starting with terminal alkynes and alkenes by triggering decomposition of the Grubbs catalyst into a ruthenium hydride, which promotes the dienyl isomerization. The presence of an alcohol additive plays a helpful role in the reaction, significantly increasing the chemical yields. Mechanistic studies are consistent with hydrometalation of the geminally substituted alkene of the 1,3-diene and transit of the ruthenium atom across the diene framework via a π-allylruthenium intermediate.

Full text of DOI:10.1021/ja4011207

Communication
pubs.acs.org/JACS
Ruthenium Hydride-Promoted Dienyl Isomerization: Access to Highly
Substituted 1,3-Dienes
Joseph R. Clark, Justin R. Griffiths, and Steven T. Diver*
Department of Chemistry, University at Buffalo, the State University of New York, Buffalo, New York 14260-3000, United States
S
* Supporting Information
through a tandem metathesis/dienyl isomerization sequence
ABSTRACT: Ruthenium hydrides were found to pro-
mote the positional isomerization of 1,3-dienes into more
highly substituted 1,3-dienes in a stereoconvergent
manner. The reaction can be conducted in one pot
starting with terminal alkynes and alkenes by triggering
decomposition of the Grubbs catalyst into a ruthenium
hydride, which promotes the dienyl isomerization. The
presence of an alcohol additive plays a helpful role in the
reaction, significantly increasing the chemical yields.
Mechanistic studies are consistent with hydrometalation
of the geminally substituted alkene of the 1,3-diene and
transit of the ruthenium atom across the diene framework
via a π-allylruthenium intermediate.
starting from Grubbs catalyst 3. Dienyl isomerization of 1,3-
dienes obtained through ene−yne metathesis provides stereo-
convergent access to diene substitution patterns that cannot be
produced by ene−yne metathesis.
The synthesis of highly substituted 1,3-dienes through
tandem catalytic reactions3 extends the scope of diene
substitution patterns accessible in a stereoconvergent process.
Ene−yne metathesis is a useful catalytic method for 1,3-diene
synthesis, but it cannot be used to produce 1,1,4-trisubstitution
or 1,1,4,4-tetrasubstitution on the resulting diene. Because
these diene substitution patterns appear in a variety of bioactive
natural products, a catalytic method that could access them
from simple alkene and alkyne reactants would be useful.
Induced decomposition of Grubbs complex 3 into a ruthenium
hydride allows these reactions to be coupled to give dienyl
isomerization (Scheme 1), which could potentially use a single
source of ruthenium catalyst.3a Though dienyl isomerization
was observed previously in one case, the catalyst was undefined
and the reactivity was limited because the E and Z isomers
could not each be converted to product.4 Overcoming the
reactivity difference of individual stereoisomers poses a
challenge but would provide a stereoconvergent catalytic
process. Last, though isomerization of alkenes is well-known,5
that of conjugated dienes is not.
Initial studies were directed to the optimization of catalysts
to effect the conversion of both E and Z dienes to the dienyl
isomerization product. Typically, mixtures of E and Z dienes are
obtained from ene−yne cross-metathesis. Earlier studies
suggested that a metal hydride may promote dienyl positional
isomerization, but each geometric isomer did not react to give
the product.4 Since primary alcohols are known to result in
conversion of Grubbs ruthenium carbenes into ruthenium
hydrides,6 our initial screening efforts focused on some
common ruthenium hydrides that are known to bring about
alkene positional isomerization (Table 1).
he catalytic ene−yne cross-metathesis promoted by the
T
Grubbs ruthenium carbenes has emerged as a useful
carbon−carbon coupling to access the 1,3-diene motif.1
Conjugated 1,3-dienes appear in a variety of bioactive natural
products and lend themselves to further synthetic elaboration.
Ene−yne metathesis is a powerful method for coupling nearly
equimolar amounts of alkyne and alkene, fulfilling the criteria of
an atom-economical cross-coupling reaction.2 However, there
are limitations of the intermolecular ene−yne cross-metathesis.
For instance, it produces limited substitution patterns on the
1,3-diene. Normally the 1,3-diene has 1,3-disubstitution (see
the reactant in Scheme 1), and there is no direct isomerization
Scheme 1. Dienyl Isomerization
Attempts to isomerize 4A without any additive led to
incomplete reaction (Table 1, entry 1). The product 5A was
contaminated with aldehyde-containing side products. On the
basis of this result, and knowing that dehydrogenation of
alcohols occurs by ruthenium hydrides,7 we reasoned that an
alcohol additive might suppress the oxidation of alcohol 5A
under the catalytic conditions. Improved conversion and
chemical yield were found using the alcohol additive 1-butanol
(entry 2). For diene reactant 4B, which does not possess a free
of the 1,3-diene into a more highly substituted 1,3-diene. In this
report, we describe a formal 1,5-hydride shift mediated by a
ruthenium hydride complex that accomplishes “dienyl isomer-
ization”, that is, the conversion of a 1,3-diene into a new 1,3-
diene (Scheme 1). In addition, conversion of alkynes and
alkenes into dienyl isomerization products can be achieved
Received: November 20, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja4011207 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
Table 1. Screening of Catalysts and Alcohol Additives
Table 2. Dienyl Isomerization of 1,3-Dienes
%
a
entry reactant
Ru cat.
alcohol
none
conv.
%5
40
1
2
4A
4A
4B
4B
4B
4B
4B
4B
4B
4B
4B
4B
1
1
1
1
1
1
1
1
2
60
100
100
80
1-butanol
1-octanol
2-propanol
cyclohexanol
tert-butanol
phenol
75
65
70
80
10
10
80
70
3
4
5
100
10
6
7
10
8
1-butanol
1-butanol
1-butanol
1-butanol
1-butanol
97
b
9
100
10
bc
,
10
11
12
RuHCl(CO)(PPh3)3
RuHCl(CO)(PCy3)2
RuH2(PPh3)4
n.d.
n.d.
n.d.
c
c
25
10
a
Conditions: 0.05 M diene in toluene, 95 °C, 6−12 h. NMR yields vs
b
mesitylene as an internal standard. Semihydrogenation was also
c
observed. Not detected.
a
Unless otherwise noted, products were isolated as pure E isomers. 5
b
c
d
mol % 1 was used. 5.5:1 E/Z. NMR yield. Isolated in 56% isolated
yield after saponification of the acetate.
hydroxyl group, several alcohols were investigated (entries 3−
7). Primary and secondary alcohols were found to improve the
yield of the reaction, whereas tert-butanol and phenol did not
(no catalyst turnover; entries 6 and 7). With 1-butanol as the
additive, several ruthenium hydrides were found to promote the
isomerization (entries 8−12). Complexes 1 and 2 gave high
conversion and reasonable yield, though some semihydrogena-
tion product was detected with 2 (entries 8 and 9).
RuHCl(CO)(PPh3)3 did not induce dienyl isomerization
(entry 10), and two other common hydrides gave low
conversions (entries 11 and 12). In these latter cases, mostly
semihydrogenation products were observed in the crude
reaction mixtures.
The scope of the diene positional isomerization was
investigated using ruthenium hydride catalyst 1 and 1-butanol
as an additive. This catalyst was selected because alcohol-
induced decomposition of Grubbs catalyst 3 forms this
ruthenium hydride.6b The reaction scope was studied for an
assortment of 1,3-dienes obtained directly from catalytic ene−
yne metathesis as E/Z mixtures8 (Table 2; for brevity, the
starting 1,3-dienes are not shown). Dienyl isomerization
proceeded efficiently in cases where R1 lacked a hydrogen
(entries 1 and 2). The R1 group (derived from the alkyne in the
ene−yne metathesis step) could be linear or branched (entries
3 and 4), and free hydroxyl groups were tolerated without
undergoing dehydrogenation by the ruthenium catalyst (entries
2−6). In only two cases (entries 3 and 4) were minor amounts
of the Z isomer produced. Further positional isomerization
occurring down the alkane chain was not observed. Remote
oxygen functionality was well-tolerated (entries 7−10). The
integrity of the propargylic chiral center was conserved
throughout the process.8,9 The isomerization occurred into
secondary alkyl groups to produce the 1,1,4,4-tetrasubstitution
pattern on the resulting 1,3-diene (entry 11). It is presumed
that the isomerizations are driven to produce the most
thermodynamically stable 1,3-diene. In all cases, the dienyl
isomerization was highly regioselective, occurring away from
the R1 position.8,10 In entries 3−12, the hydrogen at the
propargylic position could have been eliminated to give a
regioisomer, but this was not observed.
The ene−yne metathesis and dienyl isomerization were
promoted consecutively in one pot using Grubbs catalyst 3.
The dienyl isomerization product was thought to arise from a
ruthenium hydride formed in situ by decomposition of 3 in the
presence of added alcohol. To evaluate this hypothesis, the
ene−yne metathesis of 2-butynyl benzoate and 4-pentenol was
examined using Grubbs catalyst 3.8 After 1 h of heating at 85
1
°C, a new signal was observed in the H NMR spectrum at
−24.8 ppm (d, JHP = 21.6 Hz). After continued heating, the
hydride was still present. A spiking experiment with authentic 1
showed only a single peak (no new hydride signals), verifying
that the decomposition of 3 had produced 1 under the
metathesis conditions. In the preparative runs shown in Scheme
2, the suspected hydride complex 1 was intentionally formed in
situ at the end of the ene−yne metathesis upon the addition of
vinyl trimethylsilyl ether.6d This resulted in dienyl isomer-
ization. The results of the tandem sequential process are shown
in Scheme 2. Advantages of this procedure include its catalyst
Scheme 2. Tandem Ene−Yne Metathesis/Dienyl
Isomerization
B
dx.doi.org/10.1021/ja4011207 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
economy and the fact that the products of dienyl isomerization
can be accessed starting directly from alkyne and alkene
precursors.
Taken together, the observations are consistent with a π-allyl
mechanism. The labeling experiments are consistent with slow
hydroruthenation of the geminal alkene, leading to π-complex
B (eq 6). Previous studies showed facile π-allylruthenium
Subjecting pure E and Z dienes individually to the conditions
of Table 2 revealed that Z-to-E isomerization preceded dienyl
isomerization. Pure (Z)-4A and (E)-4A were allowed to react
individually under the standard conditions, and the reactions
were monitored by 1H NMR spectroscopy over a 12 h period.11
The E isomer was converted to the isomerization product 5A
without the accumulation of any detectable intermediates. In
contrast, the Z isomer did not directly provide product 5A;
instead, it underwent rapid isomerization to give the E isomer.
As geometric isomerization occurred, conversion to the dienyl
isomerization product 5A began. The stoichiometric reaction of
diene (E,Z)-17 with d-1 gave ca. 25% deuterium incorporation
at C4 (eq 4), suggesting reversible hydrometalation of the 1,2-
disubstituted alkene, which would account for the observed Z-
to-E isomerization and explain the observed stereoconvergence.
complex formation for certain 1,3-dienes.14 Both Ryu14 and
Krische7 showed that an intermediate π-allylruthenium species
can be intercepted by an aldehyde.15,16 Subsequently,
intermediate B would undergo β-elimination to give the
product of dienyl isomerization and a ruthenium hydride.
Relatedly, π-allylpalladium intermediates are known to undergo
elimination to give dienes,17 generally a minor side product in
the Tsuji−Trost reaction. The low D incorporation is
consistent with the proposed mechanism. Each turnover of a
ruthenium deuteride produces a ruthenium hydride through β-
elimination of B (eq 6), thereby introducing protium in the
activated catalyst. Ligand exchange with diene reactant would
carry the catalytic cycle forward.
The mechanistic studies do not rule out other possible roles
of the alcohol that may also account for the higher chemical
yields. For instance, with the alcohol additive present, the
hydride catalyst was found to persist and was unchanged at the
end of the reaction; without an alcohol, the ruthenium hydride
was found to be decomposed.8,18 The alcohol can regenerate a
hydride species: dienyl isomerization in the presence of
catalytic benzoquinone failed,8,19 but in the presence of an
alcohol, dienyl isomerization occurred with benzoquinone
present. No spectroscopic evidence for an alcohol-activated
catalyst was found, and no new hydride species were observed
besides complex 1, though it is possible that highly reactive
intermediates would not be detected. Further mechanistic
studies are in progress.
In summary, a new and regioselective dienyl isomerization
pathway has been described. Isomerization of the 1,3-
conjugated dienes produced through ene−yne metathesis
provides a stereoconvergent synthesis of a 1,3-diene with a
new substitution pattern. The isomerization is promoted by a
ruthenium hydride, which was identified as a decomposition
product of the Grubbs catalyst in ene−yne metathesis when
alkenol substrates or alcohols were used. The ruthenium
hydrides may be produced directly from the Grubbs catalyst
used in the metathesis step for a one-pot tandem catalytic
reaction or can be added separately to trigger the isomerization.
Mechanistic studies are consistent with the intermediacy of a π-
allylruthenium species.
Deuterium-labeling experiments using diene 18 with or
without labeled external alcohol (Table 3) provided insight to
Table 3. Deuterium Labeling Studies
a
b
% yield (% D )
entry
additive
time (h)
% conv.
6/d-6
18/d-18
1
2
3
none
14
14
2.5
80
100
25
60 (15)
70 (30)
25 (20)
20 (30)
c
CyOD
nBuOD
n.d.
75 (0)
a
b
NMR yields vs mesitylene as an internal standard. % deuterium
c
incorporation. Not detected.
the reaction pathway and suggested a role for the added
alcohol. Both cases showed significant deuterium incorporation,
though greater D incorporation12 and a better chemical yield
were seen when the deuterated alcohol was present (entry 2).
With no alcohol additive (entry 1), the recovered reactant
showed 30% D incorporation in the geminal alkene positions,
suggesting that reversible hydroruthenation of the geminally
substituted end of the 1,3-diene had occurred. With an alcohol
present, as in entry 2, the reaction went to full conversion,
providing a 70% yield of 6 with 30% D incorporation. This is
consistent with hydrometalation of the geminally substituted
1,3-diene. With an alcohol present, we were interested in
identifying the D incorporation in the recovered 18. In entry 3,
stopping the reaction at 25% conversion revealed 20% D
incorporation in the product, but unlike entry 1, no D
incorporation was found in the recovered reactant. This finding
suggests that the hydroruthenation step of the geminally
substituted end of the 1,3-diene is not reversible under these
conditions. With an alcohol additive, hydrometalation
represents “high commitment” in the catalytic reaction, with
subsequent steps, such as π-allyl formation and β-hydride
elimination,13 occurring faster than the reverse reaction.
ASSOCIATED CONTENT
* Supporting Information
Experimental procedures and characterization data. This
material is available free of charge via the Internet at http://
pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
diver@buffalo.edu
■
Notes
The authors declare no competing financial interest.
C
dx.doi.org/10.1021/ja4011207 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
ACKNOWLEDGMENTS
■
The authors thank the NSF (CHE-0848560) for support of this
work and Materia for a generous gift of Grubbs’ complex 3.
J.R.C. is a recipient of a Speyer Graduate Fellowship (UB
Department of Chemistry).
REFERENCES
■
(1) Reviews: (a) Poulsen, C. S.; Madsen, R. Synthesis 2003, 1.
(b) Mori, M., Ene−Yne Metathesis. In Handbook of Metathesis;
Grubbs, R. H., Ed.; Wiley-VCH: Weinheim, Germany, 2003; Vol. 2, pp
176−204. (c) Diver, S. T.; Giessert, A. J. Chem. Rev. 2004, 104, 1317.
(d) Diver, S. T. Sci. Synth. 2009, 46, 97.
(2) Clark, J. R.; Diver, S. T. Org. Lett. 2011, 13, 2896.
(3) (a) Fogg, D. E.; dos Santos, E. N. Coord. Chem. Rev. 2004, 248,
2365. With metathesis: (b) Alcaide, B.; Almendros, P.; Luna, A. Chem.
Rev. 2009, 109, 3817. (c) Li, J.; Lee, D. Eur. J. Org. Chem. 2011, 4269.
(4) Clark, D. A.; Clark, J. R.; Diver, S. T. Org. Lett. 2008, 10, 2055.
(5) (a) Sutton, A. E.; Seigal, B. A.; Finnegan, D. F.; Snapper, M. L. J.
Am. Chem. Soc. 2002, 124, 13390. (b) Schmidt, B. Eur. J. Org. Chem.
2003, 816. Review: (c) Schmidt, B. Eur. J. Org. Chem. 2004, 1865.
(6) (a) Dinger, M. B.; Mol, J. C. Organometallics 2003, 22, 1089. (b)
From 3: Trnka, T. M.; Morgan, J. P.; Sanford, M. S.; Wilhelm, T. E.;
Scholl, M.; Choi, T.-L.; Ding, S.; Day, M. W.; Grubbs, R. H. J. Am.
Chem. Soc. 2003, 125, 2546. Via Fischer carbenes: (c) Louie, J.;
Bielawski, C. W.; Grubbs, R. H. J. Am. Chem. Soc. 2001, 123, 11312.
(d) Arisawa, M.; Terada, Y.; Takahashi, K.; Nakagawa, M.; Nishida, A.
J. Org. Chem. 2006, 71, 4255.
(7) (a) Shibahara, F.; Bower, J. F.; Krische, M. J. J. Am. Chem. Soc.
2008, 130, 6338. (b) Shibahara, F.; Bower, J. F.; Krische, M. J. J. Am.
Chem. Soc. 2008, 130, 14120. (c) Smejkal, T.; Han, H.; Breit, B.;
Krische, M. J. J. Am. Chem. Soc. 2009, 131, 10366. (d) Han, H.;
Krische, M. J. Org. Lett. 2010, 12, 2844.
(8) See the Supporting Information (SI) for details.
(9) ent-4B (99% ee) was converted to ent-5B (96% ee) with
negligible erosion of ee at the stereogenic center.
(10) With a free hydroxyl group in the propargylic position, the
alternate regioisomer was not detected in the crude NMR spectrum.
(11) See Figures S1 and S2 in the SI.
(12) Control studies revealed that H/D exchange between the 1 and
the alcohol occurred on the time scale of the experiment. Thus,
deuterated alcohol was expected to increase the deuterium content of
the active catalyst. The alcohols in entries 2 and 3 were also deuterated
at the methine carbon.
(13) Added base (DBU or Bu4NOAc, 1 equiv) under the standard
conditions of Table 2 had a deleterious effect on the yield.
(14) π-Allyl complex formation between (Ph3P)3RuCl(CO)H and 2-
substituted-1,3-dienes has been observed: Omura, S.; Fukuyama, T.;
Horiguchi, J.; Murakami, Y.; Ryu, I. J. Am. Chem. Soc. 2008, 130,
14094.
(15) The allylation chemistry is very powerful yet limited to simple
diene substitution, such as 1-substituted- and 2-substituted-1,3-
butadienes.
(16) Our attempts to produce a π-allyl intermediate using 1,3-dienes
bearing the 1,3-disubstitution pattern did not produce an π-
1
allylruthenium intermediate observable by H NMR spectroscopy.
(17) β-Hydride elimination in π-allylpalladium chemistry: (a) Trost,
B. M.; Tometzki, G. B. J. Org. Chem. 1988, 53, 915. (b) Keinan, E.;
Kumar, S.; Dangur, V.; Vaya, J. J. Am. Chem. Soc. 1994, 116, 11151.
Decarboxylation of a π-allyl intermediate to give 1,3-dienes: (c) Trost,
B. M.; Fortunak, J. M. J. Am. Chem. Soc. 1980, 102, 2841. Specific
base-catalyzed elimination to form 1,3-dienes: (d) Takacs, J. M.;
Lawson, E. C.; Clement, F. J. Am. Chem. Soc. 1997, 119, 5956.
(18) The presence of the alcohol has a negligible effect on the rate of
phosphine exchange in catalyst 1 with or without a diene present.
(19) Benzoquinone is known to destroy a related ruthenium hydride:
Hong, S. H.; Sanders, D. P.; Lee, C. W.; Grubbs, R. H. J. Am. Chem.
Soc. 2005, 127, 17160.
D
dx.doi.org/10.1021/ja4011207 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX