Acceleration effect of an allylic hydroxy group on ring-closing enyne metathesis of terminal alkynes: Scope, application, and mechanistic insights

Article abstract of DOI:10.1002/chem.200801439

An interesting acceleration effect of an allylic hydroxy group on ring-closing enyne metathesis has been found. Ring-closing enyne metathesis of terminal alkynes possessing an allylic hydroxy group proceeded smoothly without the ethylene atmosphere generally necessary to promote the reaction. The synthesis of (+)-isofagomine with the aid of this efficient reaction has been demonstrated. Mechanistic studies of the acceleration effect were also carried out. Results of NMR studies suggested that the reaction proceeded via an "ene-then-yne" pathway. Kinetic studies indicated switching of the rate-determining step as a consequence of the presence of an allylic hydroxy group. These results suggest acceleration of the reentry step of Ru-carbene species by the allylic hydroxy group.

Full text of DOI:10.1002/chem.200801439

DOI: 10.1002/chem.200801439
Acceleration Effect of an Allylic Hydroxy Group on Ring-Closing Enyne
Metathesis of Terminal Alkynes: Scope, Application, and Mechanistic
Insights
Tatsushi Imahori,* Hidetomo Ojima, Yuichi Yoshimura, and Hiroki Takahata*[a]
Abstract: An interesting acceleration
effect of an allylic hydroxy group on
ring-closing enyne metathesis has been
found. Ring-closing enyne metathesis
of terminal alkynes possessing an allyl-
ic hydroxy group proceeded smoothly
without the ethylene atmosphere gen-
erally necessary to promote the reac-
tion. The synthesis of (+)-isofagomine
with the aid of this efficient reaction
has been demonstrated. Mechanistic
studies of the acceleration effect were
also carried out. Results of NMR stud-
ies suggested that the reaction proceed-
ed via an “ene-then-yne” pathway. Ki-
netic studies indicated switching of the
rate-determining step as a consequence
of the presence of an allylic hydroxy
group. These results suggest accelera-
tion of the reentry step of Ru-carbene
species by the allylic hydroxy group.
Keywords: (+)-isofagomine · kinet-
ics
·
reaction mechanisms
·
·
ring-closing enyne metathesis
substituent effects
Introduction
In particular, the presence of a hydroxy group often
shows interesting effects.[4,6–8] Hoye and Zhao found an ac-
celeration effect of an allylic hydroxy group in ring-closing
olefin metathesis with the first-generation Grubbs catalys-
t.[4a–c] Modulation of the ring size of a ring-closing olefin
metathesis by the activation effects of an allylic hydroxy
group has also been demonstrated.[4d,e,5] On the other hand,
there are several reported instances of negative effects of an
allylic hydroxy group decreasing the efficiency of olefin
metathesis.[6,7] A few reports of substituent effects of a hy-
droxy group in enyne metathesis have also been reported.[8,9]
We have reported an acceleration effect of an allylic hy-
droxy group in ring-closing enyne metathesis (Scheme 1).[8a]
Very recently, Diver and co-workers also reported an accel-
eration effect of an allylic hydroxy group in cross-metathe-
sis.[8b]
Metathesis catalyzed by Ru–alkylidene catalysts has
emerged as a powerful synthetic method.[1] The inherent
characteristic properties of Ru–alkylidene catalysts—partic-
ularly their remarkable functional group compatibility, air-
and moisture-insensitivity, and thermal stability, even in tol-
uene at reflux—expand the utility of metathesis. The reac-
tion has contributed greatly to organic synthesis.[2] Despite
its popularity, however, our knowledge of the interplay of
structure and efficiency in metathesis is limited. Substituents
on a substrate often activate or deactivate the substrate and
greatly affect reactivity and/or selectivity in metathesis, al-
though not usually in a readily controllable manner.[3–8]
Better fundamental understanding and application of those
substituent effects are important for the execution of effi-
cient and selective molecular transformations. Extensive
works have been devoted to study of the substituent effects
of metathesis.[3–8]
In this report we give details of a study of the acceleration
effect of an allylic hydroxy group on ring-closing enyne
[a] Dr. T. Imahori, H. Ojima, Dr. Y. Yoshimura, Prof. Dr. H. Takahata
Faculty of Pharmaceutical Sciences
Tohoku Pharmaceutical University, 4–4–1 Komatsushima
Aoba-ku, Sendai 981–8558 (Japan)
Fax : (+81)22-727-0144
E-mail: imahori@tohoku-pharm.ac.jp
takahata@tohoku-pharm.ac.jp
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200801439.
Scheme 1. Allylic hydroxy group-accelerated enyne metathesis.[8a]
10762
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 10762 – 10771

FULL PAPER
metathesis. In general, ring-closing enyne metathesis of ter-
minal alkynes in the presence of the first-generation Grubbs
catalyst is known to be a slow reaction. An ethylene atmos-
phere is necessary to promote the reaction efficiently.[10] On
the other hand, the ring-closing enyne metathesis of enyne
substrates possessing an allylic hydroxy group proceeded
smoothly in the absence of an ethylene atmosphere (under
Ar). The substrate scope of this acceleration effect of an al-
lylic hydroxy group and the effects of hydroxy groups at
other positions were investigated. As an application of the
efficient reaction based on the acceleration effect of an allyl-
ic hydroxy group, a synthesis of (+)-isofagomine was accom-
plished.
hand, the enyne metathesis of the substrate with an allylic
hydroxy group (1b) proceeded rapidly (Table 1, entry 3).
After a short time, the reaction had afforded the desired
product (2b) quantitatively (1.5 h, >99%). This acceleration
effect of an allylic hydroxy group is comparable to the accel-
eration effect of an ethylene atmosphere, which is a known
accelerator of ring-closing enyne metathesis of terminal al-
kynes (Table 1, entry 2).
In addition, the acceleration is specific to an allylic hy-
droxy group. Protection of the allylic hydroxy group by a Bn
or TBDPS group decreased the reaction efficiency (Table 1,
entries 4 and 5). These results clearly indicate an accelera-
tion effect of the allylic hydroxy group.
Mechanistic studies of substituent effects on metathesis
are relatively rare.[3h,11] Improved knowledge of mechanisms
should lead to extended application of substituent effects
and the design of new reactions. Mechanistic insights into
the acceleration effect are also described.
We then investigated the scope of this acceleration effect
of an allylic hydroxy group. Ring-closing metathesis of vari-
ous N-, O-, and C-tethered enynes, each containing an allyl-
ic hydroxy group, was investigated in the presence of the
first-generation Grubbs catalyst (4 mol%) in CH2Cl2 at RT
(Table 2). The enyne metathesis of an O-tethered enyne
Results and Discussion
Table 2. Scope of the acceleration effect of an allylic hydroxy group.
Entry
Substrate
Con- Time
ditions[a] [h]
Product
Yield
[%][b,c]
Ring-closing enyne metathesis is a powerful tool for con-
struction of carbo- and heterocyclic compounds.[1c,d] In con-
junction with our ongoing studies on aza-sugars,[12] we inves-
tigated enyne metathesis of N-containing enyne substrates
for the construction of 1,2,5,6-tetrahydropyridine frame-
works. During these studies, we identified an interesting sub-
stituent effect on ring-closing enyne metathesis. N-Contain-
ing enyne substrates (1a–d) were treated with the first-gen-
eration Grubbs catalyst (4 mol%) in CH2Cl2 at room tem-
perature. The results are summarized in Table 1.
1
2
A
1.5
>99
A
1.5
99
3
4
A
B
44.5
1
74 (17)
66
5
6
A
B
32
5
>99
66
Table 1. Acceleration effect of allylic hydroxy group.
7
8
A
B
48.5
3
76
77
Entry
Substrate
R
Time [h]
Product
Yield[a,b] [%]
9
A
4
79
1
1a
1a
1b
1c
1d
H
H
OH
OBn
OTBDPS
41
2a
2a
2b
2c
2d
32 (41)
96
>99
44 (32)
7 (73)
2[c]
3
1.5
1.5
41
10
11
12
A
B
C
44
16.5
1.5
15
45
61
4
5
41
[a] Yield of isolated product. [b] Values in parentheses are yields of re-
covery estimated from the 1H NMR spectrum of a crude reaction mix-
ture. [c] The reaction was performed under ethylene atmosphere. Boc=
tert-butoxycarbonyl, Bn=benzyl, TBDPS=tert-butyldiphenylsilyl.
13
A
1
>99[d]
[a] Conditions A: first-generation Grubbs catalyst (4 mol%, 0.002m),
CH2Cl2, RT; conditions B: first-generation Grubbs catalyst (8 mol%,
0.002m), CH2Cl2, RT; conditions C: first-generation Grubbs catalyst
(12 mol%, 0.002m), CH2Cl2, RT. [b] Yield of isolated product. [c] Values
in parentheses are yields of recovery estimated from 1H NMR spectrum
of the crude reaction mixture. [d] Since 2k is easily transformed into 2-vi-
nylnaphthalene, the yield was calculated from the sum of 2k and 2-vinyl-
naphthalene (2k/2-vinylnaphthalene 74:26). After 14 h, 2k had been
completely converted into 2-vinylnaphthalene.
An interesting effect of the allylic hydroxy group was
highlighted by the results. Enyne metathesis of the substrate
without an allylic substituent (1a) proceeded slowly, and
only a 32% yield of product (2a) had been obtained after
41 h, with 41% recovery of the starting material (Table 1,
entry 1). The efficiency of the reaction is low. On the other
Chem. Eur. J. 2008, 14, 10762 – 10771
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10763

H. Takahata, T. Imahori et al.
(1e) to form a six-membered cyclic 1,3-diene (2e) proceed-
ed with an excellent yield (99%) and in a short time
(Table 2, entry 2). C-Tethered enynes with and without sub-
stituents on the tethered chain (1 f–j) smoothly promoted
ring-closing enyne metathesis to afford five- and six-mem-
bered cyclic products (2 f–j) (Table 2, entries 3–12). Al-
though some reactions took a long time and had low effi-
ciency, almost all of these reactions were complete in short
times with higher catalyst loadings, giving cyclic products in
good yields. The reaction of a benzene ring-tethered enyne
(1k) also proceeded smoothly to yield a bicyclic product
(2k) in an excellent yield (Table 2, entry 13).
In contrast, no acceleration effect of an allylic hydroxy
group was observed in ring-closing enyne metathesis to con-
struct seven-membered ring. No seven-membered ring prod-
ucts were obtained from the corresponding N-, O-, and C-
tethered enynes with allylic hydroxy groups (1l–1n;
Scheme 2). At this point, the reason for this shutdown is un-
clear.
Scheme 3. Influence of position of the hydroxy group.
Scheme 2. Lack of ring-closing enyne metathesis by potential seven-mem-
bered ring precursor substrates.
In a utilization of this efficient ring-closing enyne meta-
thesis based on the acceleration effect of an allylic hydroxy
group, a synthesis of (+)-isofagomine was demonstrated
(Scheme 4). (+)-Isofagomine is a potent selective b-glucosi-
dase inhibitor that has recently received much attention in
Gaucherꢁs disease therapy.[13] The enyne substrate 1b, with
an allylic hydroxy group, was synthesized by addition of
propargylamine to butadiene monoxide, followed by protec-
tion of the imino group with a Boc group (71% over two
steps). The allylic hydroxy group-accelerated ring-closing
enyne metathesis of 1b efficiently provided cyclic product
2b (>99%) in a short reaction time. The hydroxy group of
2b was then protected with a TBDPS group (99%), and the
TBDPS-protected product 2d was treated with AD-mix-a.ꢂ
Highly regioselective dihydroxylation of terminal olefin pro-
ceeded to provide diol 3 (78%). Oxidative cleavage of the
diol 3 with NaIO4, followed by reduction with NaBH4, gave
allyl alcohol 4 (98% over two steps), and protection of the
hydroxy group of 4 with a benzyl group (92%) and subse-
quent removal of the TBDPS group (98%) then provided
benzylated product 5. Kinetic transesterification of 5 with
vinyl acetate in the presence of lipase PS-C accomplished
excellent resolution of enantiomers, and almost enantiomer-
ically pure (+)-5 was obtained (46%, 99% ee). Hydrobora-
tion of the internal olefin of (+)-5, followed by oxidation
with NaOH/H2O2, afforded diols 7 (67% over two steps).
After removal of both the benzyl group (99%) and the Boc
group, (+)-isofagomine (78%) and (À)-3-epi-isofagomine
(20%) were obtained. (+)-Isofagomine was thus synthesized
in an 11.5% total yield from commercially available buta-
1,3-diene monoxide.[14,15]
Though there are some limitations, the enyne metathesis
of terminal alkynes possessing an allylic hydroxy group pro-
ceeds smoothly in the absence of an ethylene atmosphere.
The acceleration effect is applicable to a wide range of
enyne substrates.
We next investigated the influence of the position of the
hydroxy group on the acceleration effect (Scheme 3).[9] We
were interested in knowing whether a hydroxy group at an-
other position would also function as an accelerator. A ho-
moallylic hydroxy group showed an acceleration effect
weaker than that of an allylic hydroxy group. An enyne sub-
strate with a homoallylic hydroxy group—1q—promoted
ring-closing enyne metathesis slowly in the presence of the
first-generation Grubbs catalyst, although the desired cyclic
1,3-diene (2q) was obtained in good yield (74%) after 14 h.
The corresponding enyne substrate with an allylic hydroxy
group—1 f—promoted rapid ring-closing enyne metathesis
(1 h, 66%) under the same conditions. The effect of a prop-
argylic hydroxy group was also investigated. An enyne sub-
strate with a propargylic hydroxy group—1r—showed poor
reactivity for ring-closing enyne metathesis. The reaction of
1r under the same conditions as used for the reaction of 1i
(corresponding enyne substrate with an allylic hydroxy
group; first-generation Grubbs catalyst, 4 mol%, RT, 4 h,
79%) provided only a 6% yield of the ring-closure product
with 50% recovery of the starting material (NMR yields).
These results indicated that an allylic hydroxy group has a
strong acceleration effect.
10764
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 10762 – 10771

Allylic Hydroxy Group Acceleration of Ring-Closing Enyne Metathesis
FULL PAPER
Scheme 4. Application to a synthesis of (+)-isofagomine: DMAP=4-N,N-dimethylaminopyridine.
The acceleration effect of an allylic hydroxy group on
ring-closing enyne metathesis is interesting. A similar accel-
eration effect of an allylic hydroxy group has been observed
in olefin metathesis in the presence of the first-generation
Grubbs catalyst, though details
nals are assigned as Ru-(a-hydroxy)carbene species (IM-1,
18.92 ppm; Scheme 5) and Ru-vinylcarbene species (IM-2,
19.53 ppm), respectively. While these observations do not
rule out the possibility of the “yne-then-ene” pathway, they
are not clear.[4] To examine the
acceleration effect further, we
investigated the mechanism of
the allylic hydroxy group-accel-
erated enyne metathesis. We
first checked the course of the
reaction: “ene-then-yne” or
“yne-then-ene” (Scheme 5).[17]
1H NMR analysis of the enyne
metathesis of oct-1-en-7-yn-3-ol
(1 f) in the presence of the first-
generation Grubbs catalyst
(Table 2, entry 3) showed new
two signals around 19 ppm,
which are attributed to Ru-car-
bene species (Figure 1,a).[17a,18]
Through comparison with relat-
ed Ru-carbene species generat-
ed by straightforward routes
(Figure 1,b and c), the two sig- Scheme 5. Reaction pathways: “ene-then-yne” or “yne-then-ene”.
Chem. Eur. J. 2008, 14, 10762 – 10771
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10765

H. Takahata, T. Imahori et al.
(Scheme 5, IM-2 to IM-1)
seems to be the rate-determin-
ing step.
A kinetic study of ring-clos-
ing enyne metathesis of a sub-
strate with an allylic hydroxy
group, on the other hand,
showed different results. Be-
cause the reaction of 1b is too
fast for kinetic study to be car-
ried out, we chose 1j as the
substrate for the kinetic study
(Figure 3). The kinetic study in-
dicated no initial rate depend-
ency on concentration of the
enyne substrate (zero order in
enyne substrate). This result
means that the enyne substrate
is not associated with the rate-
Figure 1. NMR study: “ene-then-yne” versus “yne-then-ene”.
provide direct evidence that the “ene-then-yne” pathway is
viable. In addition, styrene was detected quickly in the
1H NMR spectrum of the reaction mixture. This observation
is further suggestive of initial generation of Ru-(a-hydroxy)-
carbene species (IM-1).
Selecting the “ene-then-yne” pathway, we speculated that
the reentry step of the Ru-carbene species (IM-1) from the
Ru-vinylcarbene intermediate (IM-2; Scheme 5, “ene-then-
yne” pathway, IM-2 to IM-1) would be accelerated by an al-
lylic hydroxy group on the next substrate (Scheme 6). The
rate-determining step of ring-closing enyne metathesis of a
terminal alkyne is thought to be the reentry step (Scheme 5,
IM-2 to IM-1).[10,17b,c,19] The reaction rate would be increased
if the rate-determining step is accelerated.
On the basis of this speculation, we predicted switching of
the rate-determining step. If the reentry step of Ru–carbene
species (Scheme 5, IM-2 to IM-1) is accelerated by the allyl-
ic hydroxy group, the rate-determining step might be re-
placed by another step of the catalytic cycle because the re-
entry step would be faster. In such cases, a different step of
the catalytic cycle would be the rate-determining step rather
than the reentry step. Kinetic study of the ring-closing
enyne metathesis[11a,b] was carried out to confirm the change
of the rate-determining step. We first investigated the reac-
tion behavior of 1a, an enyne substrate without an allylic hy-
droxy group, to check the usual rate-determining step of
ring-closing enyne metathesis (Figure 2). Generally, the
rate-determining step of ring-closing enyne metathesis is be-
lieved to be the reentry step of the Ru-carbene species
(Scheme 5, IM-2 to IM-1).[10,17b,c,19] The results of a kinetic
study of 1a support the general assumption. Ring-closing
enyne metathesis of an enyne substrate without an allylic
hydroxy group indicated nearly first-order initial rate de-
pendency on concentration of the substrate. This result indi-
cates involvement of a substrate in the rate-determining
step, so the reentry step of the Ru-carbene species
Figure 2. Kinetic study: rate dependency on concentration of a substrate
without an allylic hydroxy group.
determining step. From this result, the reentry step of the
Ru–carbene species (Scheme 5, IM-2 to IM-1), which in-
volves an enyne substrate, would not be the rate-determin-
ing step.
The results of kinetic studies thus indicate a change in the
rate-determining step of ring-closing enyne metathesis re-
sulting from the presence or absence of an allylic hydroxy
group. These results are consistent with our speculation that
the reentry step of the Ru–carbene species from a Ru–vinyl-
carbene intermediate would be accelerated by the allylic hy-
droxy group on the next substrate (Scheme 6).[20] In addi-
tion, switching of the rate-determining step by an allylic hy-
droxy group demonstrates that ethylene is not necessary in
10766
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 10762 – 10771

Allylic Hydroxy Group Acceleration of Ring-Closing Enyne Metathesis
FULL PAPER
(+)-isofagomine with the aid of allylic hydroxy group-accel-
erated ring-closing enyne metathesis. Mechanistic studies on
the acceleration effect have suggested acceleration of the re-
entry step of Ru–carbene species from the Ru–vinylcarbene
species in the catalytic cycle (Scheme 6). The olefin with an
allylic hydroxy group on the substrate would be reactive
enough to react with the stable Ru–vinylcarbene intermedi-
ate to bring about the reentry of the reactive Ru-carbene
species into next catalytic cycle. However, details of the ac-
celeration, especially why the acceleration occurs, are still
unclear.[23] Further investigations into the mechanism of the
acceleration are currently underway. In addition, application
of this acceleration effect to other systems and the develop-
ment of selective molecular transformations using this accel-
eration effect are proceeding in our laboratory.
Experimental Section
Figure 3. Kinetic study: rate dependence on concentration of a substrate
with an allylic hydroxy group.
General: NMR spectra were recorded on JNM-EX270 (270 MHz), JEOL
JNM-AL400 (400 MHz), JNM-EX400 (400 MHz), and JNM-EX600
(600 MHz) spectrometers in CDCl3, C6D6, D2O, and CD2Cl2. 13C NMR
spectra were recorded with use of broad-band proton decoupling. The re-
sidual CHCl3 signal or tetramethylsilane were used as internal standards
for 1H and 13C NMR in CDCl3. The C6D6 itself was used as an internal
standard for 13C NMR in C6D6. The residual non-deuterated H2O signal
was used as an internal standard for 1H NMR, and acetonitrile was used
as an internal standard in 13C NMR in D2O. CD2Cl2 was used in NMR
studies and 1H NMR spectroscopic kinetic studies. In NMR studies, the
chemical shift of the residual non-deuterated CH2Cl2 was used as an in-
ternal standard. In kinetic studies, CH2ClCH2Cl was used as internal
standard both for chemical shift and for concentration. Chemical shifts
are expressed in d (ppm) values, and coupling constants are expressed in
hertz (Hz). The following abbreviations are used: s=singlet, d=doublet,
m=multiplet, brs=broad singlet, brd=broad doublet, dd=double dou-
blet, and ddd=double double doublet. Mass spectra were recorded on
JEOL JMN-DX303 or JEOL JMA-DA5000 spectrometers. IR spectra
were measured with a Perkin–Elmer 1725 X series FT-IR spectrometer.
CH2Cl2 was bubbled with Ar well before use in ring-closing enyne meta-
thesis.
Scheme 6. Proposed acceleration effect of an allylic hydroxy group.
General procedure for allylic hydroxy group-accelerated ring-closing
enyne metathesis: The first-generation Grubbs catalyst (4, 8, or
12 mol%) was added at room temperature under Ar to a CH2Cl2 solution
of an enyne substrate[24] containing an allylic hydroxy group. The concen-
tration of the first-generation Grubbs catalyst was kept at 0.002m. The
mixture was stirred for the indicated reaction time. The reaction mixture
was then concentrated in vacuo, and the residue was purified by silica gel
column chromatography to provide a cyclic 1,3-diene.
ring-closing enyne metathesis in the presence of an allylic
hydroxy group.[17b,c]
Conclusion
N-tert-Butoxycarbonyl-3-vinyl-1,2,5,6-tetrahydropyridine (2a): 1H NMR
(400 MHz, CDCl3): d=1.47 (s, 9H), 2.22 (s, 2H), 3.48 (t, J=5.6 Hz, 2H),
4.04 (s, 2H), 4.97 (d, J=11.1 Hz, 1H), 5.07 (d, J=17.9 Hz, 1H), 5.84 (s,
1H), 6.30 ppm (dd, J=17.9, 11.1 Hz, 1H); 13C NMR (100 MHz, CDCl3):
d=25.4, 27.4, 28.4, 39.6, 40.8, 42.3, 42.8, 79.6, 85.1, 111.0, 126.7, 127.4,
133.4, 136.9, 146.7, 154.9 ppm; IR (neat): n˜ =1697 cmÀ1; EI-MS: m/z: 209
[M]+; HRMS: m/z: calcd for C12H19NO2: 209.1416; found: 209.1424.
We have found an interesting acceleration effect of an allylic
hydroxy group on ring-closing enyne metathesis.[21] The ring-
closing enyne metathesis of various terminal alkynes, each
containing an allylic hydroxy group, proceeded smoothly in
the absence of an ethylene atmosphere. The ethylene-free
reaction is convenient and atom-economical. In addition,
the metatheses with the first-generation Grubbs catalyst
often demonstrate more chemoselective transformation than
the more reactive newer-generation Ru–carbene catalysts.[22]
We believe that enyne metathesis utilizing this acceleration
effect could be a more helpful and familiar tool in organic
synthesis. Indeed, we have accomplished the synthesis of
N-tert-Butoxycarbonyl-5-hydroxy-3-vinyl-1,2,5,6-tetrahydropyridine (2b):
1H NMR (600 MHz, CDCl3, TMS): d=1.48 (s, 9H), 2.23 (s, 2H), 3.55
(brs, 2H), 3.95 (d, J=17.6 Hz, 1H), 4.10–4.26 (m, 2H), 5.12 (d, J=
11.0 Hz, 1H), 5.22–5.25 (m, 1H), 5.87 (brs, 1H), 6.30 ppm (dd, J=17.6,
11.0 Hz, 1H); 13C NMR (67.5 MHz, C6D6, 608C): d=28.5, 43.0, 48.2,
64.2, 79.8, 113.1, 130.3, 135.7, 136.7, 155.1 ppm; IR (neat): n˜ =1683,
3406 cmÀ1; EI-MS: m/z: 225 [M]+; HRMS: m/z: calcd for C12H19NO3:
225.1365; found: 225.1370.
Chem. Eur. J. 2008, 14, 10762 – 10771
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10767

H. Takahata, T. Imahori et al.
N-tert-Butoxycarbonyl-5-benzyloxy-3-vinyl-1,2,5,6-tetrahydropyridine
(2c): 1H NMR (400 MHz, CDCl3): d=1.48 (s, 9H), 3.48–3.85 (m, 2H),
4.06 (d, J=15.0 Hz, 3H), 4.58 (d, J=11.6 Hz, 1H), 4.66–4.68 (m, 1H),
5.10 (d, J=11.1 Hz, 1H), 5.14–5.31 (m, 1H), 5.89 (s, 1H), 6.31 ppm (dd,
J=17.9, 11.1 Hz, 1H); 13C NMR (100 MHz, CDCl3): d=28.4, 42.9, 44.8,
70.4, 70.6, 80.0, 113.4, 113.9, 126.5, 127.7, 128.4, 136.0, 136.1, 138.2,
154.8 ppm; IR (neat): n˜ =1696 cmÀ1; EI-MS: m/z: 315 [M]+; HRMS:
m/z: calcd for C19H25NO3: 315.1834; found: 315.1829.
3-Vinyl-1,4-dihydronaphthalen-1-ol (2k): This compound is not stable
and easily converted into 2-vinylnaphthalene. Compound 2k was there-
1
1
fore not isolated. H NMR data for 2k were extracted from the H NMR
spectrum of the mixture of 2k and 2-vinylnaphthalene. No other charac-
terization data for 2k could be measured. 1H NMR (400 MHz, CDCl3):
d=1.77 (d, J=9.2 Hz, 1H), 3.47 (dd, J=21.3, 3.9 Hz, 1H), 3.60 (brd, J=
19.8 Hz, 1H), 5.20 (d, J=10.6 Hz, 1H), 5.21–5.30 (m, 1H), 5.39 (d, J=
17.4 Hz, 1H), 6.08 (m, 1H), 6.52 (dd, J=17.4, 10.6 Hz, 1H), 7.24–7.33
(m, 3H), 7.59–7.61 ppm (m, 1H).
N-tert-Butoxycarbonyl-5-tert-butyldiphenylsilyloxy-3-vinyl-1,2,5,6-tetra-
hydropyridine (2d): 1H NMR (400 MHz, CDCl3, TMS): d=1.00 (s, 9H),
1.33 (s, 9H), 3.13 (dd, J=12.4, 6.8 Hz, 1H), 3.58–3.61 (m, 1H), 3.82 (d,
J=16.6 Hz, 1H), 4.04 (d, J=17.1 Hz, 1H), 4.24 (brs, 1H), 4.96 (d, J=
11.2 Hz, 1H), 5.08 (d, J=17.6 Hz, 1H), 5.56 (brs, 1H), 6.13 (dd, J=17.8,
11.0 Hz, 1H), 7.28–7.38 (m, 6H), 7.59–7.63 ppm (m, 4H); 13C NMR
(100 MHz, CDCl3): d=19.2, 26.9, 28.4, 42.0, 48.0, 65.5, 79.8, 127.6, 127.6,
127.7, 129.7, 129.8, 135.5, 135.7, 135.8, 154.7 ppm; IR (neat): n˜ =
1699 cmÀ1; EI-MS: m/z: 463 [M]+; HRMS: m/z: calcd for C28H37NO3Si:
463.2543; found: 463.2522.
5-Vinyl-3,6-dihydro-2H-pyran-3-ol (2e): Colorless oil; 1H NMR
(400 MHz, CDCl3): d=2.04 (brs, 1H), 3.70 (dd, J=11.6, 2.9 Hz, 1H),
3.85 (dd, J=11.6, 1.9 Hz, 1H), 4.04 (brs, 1H), 4.18 (d, J=15.9 Hz, 1H),
4.38 (d, J=15.5 Hz, 1H), 5.09 (d, J=18.4 Hz, 1H), 5.10 (d, J=10.6 Hz,
1H), 5.93 (d, J=5.9 Hz, 1H), 6.25 ppm (dd, J=18.4, 11.6 Hz, 1H);
13C NMR (100 MHz, CDCl3): d=63.0, 65.1, 70.8, 114.0, 126.3, 135.3,
137.9 ppm; IR (neat): n˜ =1007, 1123, 1606, 1648, 2849, 3390 cmÀ1; EI-MS:
m/z: 126 [M]+; HRMS: m/z: calcd for C7H10NO2: 126.0681; found:
126.0676.
3-Vinylcyclohex-2-en-1-ol (2 f): Colorless oil; 1H NMR (400 MHz,
CDCl3): d=1.56–1.64 (m, 2H), 1.71 (brs, 1H), 1.79–1.90 (m, 2H), 2.11–
2.15 (m, 2H), 4.29 (brs, 1H), 5.04 (d, J=10.7 Hz, 1H), 5.20 (d, J=
17.6 Hz, 1H), 5.75 (brs, 1H), 6.34 ppm (dd, J=17.6, 10.7 Hz, 1H);
13C NMR (100 MHz, CDCl3): d=18.8, 23.7, 32.0, 66.2, 112.9, 130.8, 138.6,
139.2 ppm; IR (neat): n˜ =909, 991, 1049, 1607, 2863, 2935, 3308 cmÀ1; EI-
MS: m/z: 124 [M]+; HRMS: m/z: calcd for C8H12O: 124.0888; found:
124.0889.
6,6-Dimethyl-3-vinylcyclohex-2-en-1-ol (2g): Colorless oil; 1H NMR
(400 MHz, CDCl3): d=0.90 (s, 3H), 0.97 (s, 3H), 1.42–1.47 (m, 2H),
1.56–1.62 (m, 1H), 2.10–2.17 (m, 1H), 3.86 (brs, 1H), 5.03 (d, J=
10.6 Hz, 1H), 5.18 (d, J=17.9 Hz, 1H), 5.65 (brs, 1H), 6.35 ppm (dd, J=
17.9, 11.1 Hz, 1H); 13C NMR (100 MHz, CDCl3): d=20.9, 21.5, 26.4, 32.7,
33.9, 74.6, 112.7, 130.5, 137.4, 138.9 ppm; IR (neat): n˜ =990, 1029, 1606,
2920, 3369 cmÀ1; EI-MS: m/z: 152 [M]+; HRMS: m/z: calcd for C10H16O:
152.1201; found: 152.1197.
2-Vinylnaphthalene: White solid; 1H NMR (400 MHz, CDCl3): d=5.35
(d, J=10.6 Hz, 1H), 5.89 (d, J=17.4 Hz, 1H), 6.90 (dd, J=17.4, 10.6 Hz,
1H), 7.44–7.50 (m, 2H), 7.65 (dd, J=8.7, 1.4 Hz, 1H),7.77 (brs, 1H),
7.80–7.85 ppm (m, 1H); 13C NMR (100 MHz, CDCl3): d=114.2, 123.2,
125.9, 126.2, 126.4, 127.7, 128.0, 128.1, 133.1, 133.5, 135.0, 136. 9 ppm; IR
(KBr): n˜ =3382, 3058, 2933, 1688, 1630, 1600, 1508, 1467,1356, 1284, 1230,
1196, 1125 cmÀ1; EI-MS: m/z: 154 [M]+; HRMS: m/z: calcd for C12H10
:
115.0783; found: 115.0779.
Ring-closing enyne metathesis of enynes containing a homoallylic or a
propargylic hydroxy group: The reactions were carried out by the Gener-
al Procedure for allylic hydroxy group-accelerated ring-closing enyne
metathesis.
4-Vinylcyclohex-3-enol (2q): 1H NMR (400 MHz, CDCl3, TMS): d=
1.60–1.75 (m, 2H), 1.89–1.97 (m, 1H), 2.10–2.24 (m, 2H), 2.32–2.38 (m,
1H), 2.48 (brd, J=17.4 Hz, 1H), 3.97 (m, 1H), 4.94 (d, J=10.6 Hz, 1H),
5.08 (d, J=17.4 Hz, 1H), 5.64 (brs, 1H), 6.35 ppm (dd, J=17.4, 10.6 Hz,
1H); 13C NMR (100 MHz, C6D6): d=21.9, 30.5, 34.8, 66.9, 110.9, 126.2,
135.6, 139.1 ppm; IR (neat): n˜ =1606, 1644, 3346 cmÀ1; EI-MS: m/z: 124
[M]+; HRMS: m/z: calcd for C12H19NO3: 124.0888; found: 124.0887.
2-Vinylcyclopent-2-enol (2r): This compound was obtained in quantities
too small to isolate and we could not measure spectroscopic data of a
pure sample. The 1H NMR data were extracted from a spectrum of a
mixed sample with reference to reported data.[25] 1H NMR (400 MHz,
CDCl3, TMS): d=1.82–1.88 (m, 1H), 2.17–2.35 (m, 2H), 2.54–2.61 (m,
1H), 4.90–5.10 (m, 1H), 5.15 (d, J=10.6 Hz, 1H), 5.42 (d, J=17.9 Hz,
1H), 5.89 (t, J=2.4 Hz, 1H), 6.47 ppm (dd, J=17.9, 10.6 Hz, 1H). The
yield was estimated from the 1H NMR spectrum of the crude reaction
mixture based on internal standard.
Application to the synthesis of (+)-isofagomine
tert-Butyl 2-hydroxybut-3-enylprop-2-ynylcarbamate (1b): Butadiene
monoxide (18 mmol) was added at 158C to a solution of propargylamine
(54 mmol) and H2O (0.25 mL), and the mixture was stirred at 1008C for
6 h. The mixture was then concentrated in vacuo, and the residue was dis-
solved in 1,4-dioxane (25 mL) and H2O (5 mL). NaOH (1n, 20 mL,
20 mmol) and Boc2O (20 mmol) were added to the solution at ambient
temperature, and the mixture was stirred overnight. The solvent was
evaporated in vacuo, and the residue was diluted with Et2O. The mixture
was washed with aqueous citric acid solution (20%) and brine. The or-
ganic layer was dried over Na2SO4. The solvent was evaporated in vacuo,
and the residue was purified by silica gel column chromatography (n-
hexane/AcOEt 10:1–6:1) to provide 1b (2.87 g, 71% yield) and tert-butyl
2-hydroxybut-3-enylprop-2-ynylcarbamate (132 mg). Compound 1b:
1H NMR (400 MHz, CDCl3): d=1.46 (s, 9H), 2.24 (t, J=2.4 Hz, 1H),
3.37 (brs, 1H), 3.44 (dd, J=14.6, 3.8 Hz, 2H), 4.07 (brs, 2H), 4.35–4.39
(m, 1H), 5.16 (d, J=10.6 Hz, 1H), 5.33 (brd, J=16.4 Hz, 1H), 5.84 ppm
(ddd, J=16.4, 10.6, 5.8 Hz, 4H); 13C NMR (100 MHz, CDCl3): d=28.3,
38.3, 52.9, 71.7, 72.2, 79.6, 81.0, 115.9, 138.1, 156.3 ppm; IR (neat): n˜ =
Di-tert-butyl 5-hydroxy-3-vinylcyclohex-3-ene-1,1-dicarboxylate (2h):
Colorless, viscous oil; 1H NMR (400 MHz, CDCl3): d=1.42 (s, 9H), 1.46
(s, 9H), 2.20–2.28 (m, 2H), 2.53 (d, J=16.9 Hz, 1H), 2.68 (d, J=16.9 Hz,
1H), 3.07 (d, J=9.2 Hz, 1H), 4.31 (brs, 1H), 5.10 (d, J=10.6, 1H), 5.30
(d, J=17.9 Hz, 1H), 5.76 (brs, 1H), 6.53 ppm (dd, J=17.4, 11.1 Hz, 1H);
13C NMR (100 MHz, CDCl3): d=27.7, 27.8, 29.3, 36.6, 53.8, 64.2, 81.7,
82.1, 113.5, 130.1, 134.4, 138.4, 170.3 ppm; IR (neat): n˜ =1147, 1257, 1369,
1608, 1716, 1729, 2934, 2978, 3522 cmÀ1; EI-MS: m/z: 324 [M]+; HRMS:
m/z: calcd for C18H28O5: 324.1937; found: 324.1947.
3-Vinylcyclopent-2-en-1-ol (2i): Colorless oil; 1H NMR (400 MHz,
CDCl3): d=1.72–1.78 (m, 1H), 1.86 (brs, 1H), 2.29–2.39 (m, 2H), 2.55–
2.70 (m, 1H), 4.89 (brd, J=5.3 Hz, 1H), 5.17 (d, J=11.1 Hz, 1H), 5.21
(d, J=18.8 Hz, 1H), 5.75 (s, 1H), 6.56 ppm (dd, J=17.9, 10.6 Hz, 1H);
13C NMR (100 MHz, CDCl3): d=28.94, 33.43, 77.31, 116.70, 131.92,
1683, 3441 cmÀ1
;
EI-MS: m/z: 225 [M]+; HRMS: m/z: calcd for
133.08, 145.80 ppm; IR (neat): n˜ =905, 1044, 1591, 2853, 2939, 3339 cmÀ1
;
C12H19NO3: 225.1365; found: 225.1361.
EI-MS: m/z: 110 [M]+; HRMS: m/z: calcd for C7H10O: 110.0732; found:
1-tert-Butoxycarbonyl-3-vinyl-1,2,5,6-tetrahydropyridin-3-ol (2b): Com-
pound 2b was synthesized by the allylic hydroxy group-accelerated ring-
closing enyne metathesis (Table 2, entry 1).
110.0732.
5,5-Dimethyl-3-vinylcyclopent-2-en-1-ol (2j): Colorless oil; 1H NMR
(400 MHz, CDCl3): d=1.08 (s, 3H), 1.09 (s, 3H), 1.32 (brs, 1H), 2.18 (d,
J=15.9 Hz, 1H), 2.36 (d, J=15.9 Hz, 1H), 4.25 (brs, 1H), 5.15 (d, J=
10.6 Hz, 1H), 5.16 (d, J=17.4 Hz, 1H), 5.68 (brs, 1H), 6.53 ppm (dd, J=
17.4, 10.6 Hz, 1H); 13C NMR (100 MHz, CDCl3): d=22.7, 28.4, 41.7, 44.1,
84.5, 116.4, 131.0, 133.5, 145.0 ppm; IR (neat): n˜favor993, 1036, 2926,
2956, 3361 cmÀ1; EI-MS: m/z: 138 [M]+; HRMS: m/z: calcd for C9H14O:
138.1045; found: 138.1047.
1-tert-Butoxycarbonyl-5-tert-butyldiphenylsilyloxy-3-vinyl-1,2,5,6-tetrahy-
dropyridine (2d): Imidazole (7.02 mmol), DMAP (0.094 mmol), and
TBDPSCl (5.15 mmol) were added at ambient temperature to a solution
of 2b (4.68 mmol) in CH2Cl2 (20 mL), and the mixture was stirred for
1 h. The mixture was then filtered with a pad of celite. The filtrate was
washed with brine and dried over Na2SO4. The solvent was evaporated in
10768
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 10762 – 10771

Allylic Hydroxy Group Acceleration of Ring-Closing Enyne Metathesis
vacuo, and the residue was purified by silica gel column chromatography
FULL PAPER
(À)-N-tert-Butoxycarbonyl-5-acetoxy-3-(benzyloxy)methyl-3-piperidine
(6)
and
(S)-N-tert-Butoxycarbonyl-5-hydroxy-3-(benzyloxy)methyl-
(n-hexane/AcOEt 8:1) to provide 2d (2141 mg, 99% yield).
1,2,5,6-tetrahydropyridine: Vinyl acetate (96.8 mmol) and lipase PS-C
(628 mg) were added at ambient temperature to solution of
1-tert-Butoxycarbonyl-5-tert-butyldiphenylsilyloxy-3-(1,2-dihydroxyeth-
yl)-1,2,5,6-tetrahydropyridine (3): AD-mix-a and methanesulfonamide
(0.30 mmol) were added at ambient temperature to a solution of 2d
(0.30 mmol) in tBuOH/H2O (1:1), and the mixture was stirred for 15 h.
The mixture was then diluted with AcOEt and washed with saturated
aqueous NH4Cl and H2O. The organic layer was dried over Na2SO4. The
solvent was evaporated in vacuo, and the residue was purified by silica
gel column chromatography (n-hexane/AcOEt 1:1) to provide 3 (116 mg,
a
5
(1.21 mmol) in diisopropyl ether (5.8 mL). The mixture was warmed to
308C and stirred for 15 h. The mixture was then filtered through a pad of
celite. The filtrate was concentrated in vacuo, and the residue was puri-
fied by silica gel column chromatography (n-hexane/AcOEt 10:1) to pro-
vide 6 (179 mg, 46% yield) and (+)-5 (233 mg, 53% yield).
Compound 6: [a]2D7 =À79.6 (c=1.7 in CHCl3); 1H NMR (400 MHz,
CDCl3): d=1.47 (s, 9H), 2.05 (s, 3H), 3.40–3.44 (m, 1H), 3.74–3.86 (m,
2H), 3.98 (s, 2H), 4.19 (brs, 1H), 4.50 (s, 2H), 5.19 (brs, 1H), 5.88 (brs,
1H), 7.27–7.37 ppm (m, 5H); 13C NMR (100 MHz, CDCl3): d=21.0, 28.1,
43.3, 44.8, 65.5, 71.1, 72.4, 79.9, 122.3, 127.6, 128.1, 128.3, 137.6, 154.6,
1
78% yield). H NMR (400 MHz, CDCl3): d=1.06 (s, 9H), 1.41 (brs, 9H),
2.49 (brs, 2H), 3.31–3.93 (m, 6H), 4.06 (brs, 1H), 4.27 (brs, 1H), 5.65
(brs, 1H), 7.35–7.45 (m, 6H), 7.65–7.70 ppm (m, 4H); 13C NMR
(100 MHz, CDCl3): d=19.2, 26.9, 28.4, 42.5, 48.1, 65.0, 65.3, 73.4, 80.0,
127.6, 127.7, 127.7, 129.8, 135.7, 135.7, 154.9 ppm; IR (neat): n˜ =1661,
3407 cmÀ1; EI-MS: m/z: 496 [M+H]+, 497 [M]+; HRMS: m/z: calcd for
C28H39NO5Si: 497.2597; found: 497.2587.
170.4 ppm; IR (neat): n˜ =1701, 1735 cmÀ1 EI-MS: m/z: 361 [M]+;
;
HRMS: m/z: calcd for C20H27NO5: 361.1889; found: 319.1883.
Compound (+)-5: [a]2D7 =+45.6 (c=1.0 in CHCl3); HPLC (Chiralpak IA,
hexane/AcOEt 90:10, 1.5 mLminÀ1
, 254 nm): tminor =23.4 min, tmajor =
N-tert-Butoxycarbonyl-5-tert-butyldiphenylsilyloxy-3-hydroxymethyl-
1,2,5,6-tetrahydropyridine (4): NaIO4 (2.04 mmol) was added at ambient
temperature to a solution of 3 (1.70 mmol) in EtOH/H2O (1:1, 30 mL),
and the mixture was stirred for 3 h. NaBH4 (3.40 mmol) was then added
to the mixture at ambient temperature. After the mixture had been
stirred for 3 h, the reaction mixture was concentrated in vacuo, and the
residue was diluted with CH2Cl2. The mixture was washed with Na2S2O3
(10%), saturated aqueous NaHCO3, and brine. The organic layer was
dried over Na2SO4. The solvent was evaporated in vacuo, and the residue
was purified by silica gel column chromatography (n-hexane/AcOEt 2:1)
to provide 4 (776 mg, 98% yield). 1H NMR (400 MHz, CDCl3): d=1.07
(s, 9H), 1.39–1.45 (m, 9H), 2.19 (brs, 1H), 3.27 (dd, J=12.8, 6.6 Hz, 1H),
3.59 (d, J=9.9 Hz, 1H), 3.76–3.98 (m, 4H), 4.29 (s, 1H), 5.64 (s, 1H),
7.36–7.45 (m, 6H), 7.67–7.70 ppm (m, 4H); 13C NMR (100 MHz, CDCl3):
d=19.2, 26.9, 28.3, 43.0, 48.0, 64.2, 65.1, 80.0, 127.6, 127.7, 128.3, 129.8,
135.7, 135.7, 154.7 ppm; IR (neat): n˜ =1689, 3419 cmÀ1; EI-MS: m/z: 467
[M]+; HRMS: m/z: calcd for C27H37NO4Si: 467.2492; found: 467.2491.
25.9 min: 99% ee. The absolute configuration of (+)-5 was determined as
(S) after conversion into (+)-isofagomine as described below.
The ee of 6 was determined after cleavage of the benzyl group to convert
it into (À)-5. K2CO3 (0.06 mmol) was added at ambient temperature to a
solution of 6 (0.12 mmol) in dry MeOH (1.5 mL). After stirring for 1
hour, the reaction mixture was diluted with AcOEt and washed with
water. The aqueous layer was extracted twice with AcOEt, and the com-
bined organic extract was dried over Na2SO4. The solvent was evaporated
in vacuo, and the residue was purified by silica gel column chromatogra-
phy (n-hexane/AcOEt 1:1) to provide (À)-5 (36 mg, 94% yield). [a]D27
=
À43.0 (c=1.1 in CHCl3). The absolute configuration of (À)-5 was deter-
mined as (R) based on (+)-5.
5-Benzyloxymethyl-1-tert-butoxycarbonylpiperidine-3,4-diol
(7):
BH3·THF (1.0m THF solution, 2.1 mL, 2.1 mmol) was added at 08C to a
solution of (+)-5 (0.26 mmol) in THF (1.0 mL). The reaction mixture was
then warmed to 708C and stirred for 9.5 h. After the stirring, the mixture
was cooled to 08C. NaOH (3n, 3.0 mL) and H2O2 (30%) were added
dropwise to the mixture, which was stirred for 6 h at ambient tempera-
ture. Then organic and aqueous layers were separated, and the aqueous
layer was extracted five times with CHCl3. The combined organic ex-
tracts were washed with brine and then dried over Na2SO4. The solvent
was evaporated in vacuo, and the residue was purified by silica gel
column chromatography (n-hexane/AcOEt 1:2) to provide 7 (59 mg,
3-Benzyloxymethyl-1-tert-butoxycarbonyl-1,2,5,6-tetrahydropyridin-5-ol
(5): NaH (1.68 mmol) was added at 08C to a solution of 4 (0.84 mmol) in
THF (8.0 mL), and the mixture was stirred for 1 hour. Benzyl bromide
(2.52 mmol) and tetrabutylammonium iodide (12 mg) were then added to
the mixture at 08C, and the mixture was stirred for 6.5 h at ambient tem-
perature. The resulting mixture was diluted with Et2O and was washed
with saturated aqueous NH4Cl, water, and brine. The organic layer was
dried over Na2SO4. The solvent was evaporated in vacuo, and the residue
was purified by silica gel column chromatography (n-hexane/AcOEt
10:1) to provide benzylated product (431 mg, 92% yield). 1H NMR
(400 MHz, CDCl3): d=1.08 (s, 9H), 1.41 (brs, 9H), 3.36–3.27 (m, 1H),
3.48–3.96 (m, 5H), 4.28 (brs, 1H), 4.41 (s, 2H), 5.67 (brs, 1H), 7.27–7.49
(m, 11H), 7.67–7.71 ppm (m, 4H); 13C NMR (100 MHz, CDCl3): d=19.7,
26.8, 28.3, 43.3, 47.8, 65.0, 71.3, 72.0, 79.5, 126.3, 127.5, 127.6, 127.7, 128.2,
129.6, 129.7, 135.6, 135.6, 137.8, 154.5 ppm; IR (neat): n˜ =1701 cmÀ1; EI-
MS: m/z: 557 [M]+; HRMS: calcd for C34H43NO4Si: 557.2961; found:
557.2958.
1
67% yield). H NMR (400 MHz, CDCl3): d=1.44 (s, 9H), 1.84 (brs, 1H),
2.52 (brs, 2H), 3.37–3.48 (m, 2H), 3.58 (brs, 3H), 3.83 (brs, 1H), 4.11–
4.19 (m, 2H), 4.51 (s, 2H), 7.27–7.36 ppm (m, 5H); 13C NMR (100 MHz,
CDCl3): d=28.3, 44.1, 44.1, 45.3, 47.6, 64.8, 67.2, 70.1, 71.5, 73.3, 80.2,
127.2, 127.5, 127.7, 127.8, 137.7, 140.3, 154.6, 155.7 ppm; IR (neat): n˜ =
1671, 3395 cmÀ1
;
EI-MS: m/z: 337 [M]+; HRMS: m/z: calcd for
C18H27NO5: 337.1889; found: 337.1898.
1-tert-Butoxycarbonyl-5-(hydroxymethyl)piperidine-3,4-diol (8): MeOH
(7 mL) was added under H2 atmosphere at ambient temperature to a
mixture of 7 (0.22 mmol) and Pd/C (30 mg). After stirring for 5 h, the re-
action mixture was filtered with a pad of celite. The filtrate was concen-
trated in vacuo, and the residue was purified by silica gel column chroma-
Tetrabutylammonium fluoride (2.1 mmol) was added at room tempera-
ture to a solution of the obtained benzylated product (1.7 mmol) in THF
(20 mL), and the mixture was stirred at room temperature for 1.5 h. The
solvent was then evaporated in vacuo, and the residue was diluted with
CHCl3. The mixture was washed with saturated aqueous NaHCO3, and
the resulting aqueous phase was extracted with CHCl3. The combined or-
ganic extracts were washed with brine and then dried over Na2SO4. The
solvent was evaporated in vacuo, and the residue was purified by silica
gel column chromatography (n-hexane/AcOEt 2:1) to provide 5 (535 mg,
98% yield). 1H NMR (400 MHz, CDCl3): d=1.47 (s, 9H), 1.80 (s, 1H),
3.52 (brs, 2H), 3.81 (d, J=18.4 Hz, 1H), 3.97–4.11 (s, 3H), 4.21 (brs,
1H), 4.50 (s, 2H), 5.90–5.91 (s, 1H), 7.26–7.39 ppm (m, 5H); 13C NMR
(100 MHz, CDCl3): d=28.3, 44.1, 47.2, 63.5, 71.5, 72.3, 80.2, 127.7, 127.7,
127.9, 128.2, 128.4, 137.8, 155.1 ppm; IR (neat): n˜ =1695, 3405 cmÀ1; EI-
MS: m/z: 319 [M]+; HRMS: m/z: calcd for C18H25NO4: 319.1784; found:
319.1776.
1
tography (CH3Cl:MeOH 10:1) to provide 8 (54 mg, 99% yield). H NMR
(400 MHz, CDCl3): d=1.42 (s, 9H), 1.67 (brs, 1H), 2.56 (brs, 2H), 3.34–
3.55 (m, 2H), 3.69–3.76 (m, 3H), 3.88–4.16 ppm (m, 3H); 13C NMR
(100 MHz, CDCl3): d=28.4, 43.4, 47.7, 61.8, 62.8, 71.4, 75.7, 80.5,
155.0 ppm; IR (neat): n˜ =1669, 3383 cmÀ1 EI-MS: m/z: 247 [M]+;
;
HRMS: m/z: calcd for C11H21NO5: 247.1420; found: 247.1416.
5-(hydroxymethyl)piperidine-3,4-diols ((+)-isofagomine and (À)-3-epi-
isofagomine): HCl (10%, 2.7 mL) was added to
a solution of 8
(0.20 mmol) in 1,4-dioxane (0.9 mL), and the mixture was heated to
reflux. After stirring for 2 h, the reaction mixture was allowed to cool to
ambient temperature. Aqueous NH4OH solution (28%) was then added
to basify the mixture, and the solvent was evaporated in vacuo. The resi-
due was purified by silica gel column chromatography (MeOH/10%
NH4OH 10:1) to afford (+)-isofagomine (23 mg, 78% yield) and (À)-3-
epi-isofagomine (6 mg, 20%).
Chem. Eur. J. 2008, 14, 10762 – 10771
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10769

H. Takahata, T. Imahori et al.
(+)-Isofagomine: [a]D27 =+17.6 (c=0.3 in EtOH) [lit.:[13] [a]D27 =+16.3 (c=
0.32 in EtOH)]; 1H NMR (600 MHz, D2O): d=1.48–1.57 (m, 1H), 2.21–
2.28 (m, 2H), 2.91–3.00 (m, 2H), 3.13 (t, J=9.7 Hz, 1H), 3.30–3.37 (m,
1H), 3.45 (dd, J=11.6, 6.8 Hz, 1H), 3.64 ppm (dd, J=11.6, 3.4 Hz, 1H);
13C NMR (100 MHz, D2O, CH3CN): d=44.4, 46.3, 49.3, 60.4, 71.9,
73.6 ppm; IR (neat): n˜ =3383 cmÀ1; EI-MS: m/z: 147 [M]+; HRMS: m/z:
calcd for C6H13NO3: 147.0895; found: 147.0889.
(À)-3-epi-Isofagomine: [a]2D8 =À19.3 (c=0.6 in EtOH); 1H NMR
(400 MHz, D2O): d=2.05–2.10 (m, 1H), 2.61 (t, J=12.6 Hz, 1H), 2.89 (d,
J=13.5 Hz, 1H), 3.16–3.25 (m, 2H), 3.59–3.65 (m, 2H), 3.72 (dd, J=11.6,
3.4 Hz, 1H), 3.98 ppm (s, 1H); 13C NMR (67.8 MHz, D2O, CH3CN): d=
37.8, 45.0, 48.3, 60.1, 66.2, 68.5 ppm; IR (neat): n˜ =3345 cmÀ1; EI-MS:
m/z: 147 [M]+; HRMS: m/z: calcd for C6H13NO3: 147.0895; found:
147.0894.
Chem. 2006, 471–482; k) Y. J. Kim, J. B. Grimm, D. Lee, Tetrahe-
dron Lett. 2007, 48, 7961–7964.
[4] For acceleration effects of an allylic hydroxy group on Ru–carbene-
catalyzed olefin metathesis, see: a) T. R. Hoye, H. Zhao, Org. Lett.
1999, 1, 1123–1125; b) R. M. Kanada, D. Itoh, M. Nagai, J. Niijima,
N. Asai, Y. Mizui, S. Abe, Y. Kotake, Angew. Chem. 2007, 119,
4428–4433; Angew. Chem. Int. Ed. 2007, 46, 4350–4355; c) K. P. Ka-
liappan, N. Kumar, Tetrahedron 2005, 61, 7461–7469. Modulation of
ring size of ring-closing metathesis by allylic hydroxy group substitu-
ent effects, see: d) B. Schmidt, S. Nave, Chem. Commun. 2006,
2489–2491; e) B. Schmidt, S. Nave, Adv. Synth. Catal. 2007, 349,
215–230.
[5] Modulation of ring size of ring-closing metathesis by steric hin-
drance has also been reported. See; a) K. J. Quinn, A. K. Isaacs,
R. A. Arvary, Org. Lett. 2004, 6, 4143–4145; b) K. J. Quinn, A. K.
Isaacs, B. A. DeChristopher, S. C. Szklarz, R. A. Arvary, Org. Lett.
2005, 7, 1243–1245.
Procedure for NMR studies to determine reaction pathway: The first-
generation Grubbs catalyst (0.025 mmol) was placed in a NMR tube and
the tube was equipped with a rubber stopper. Then the tube was dried in
vacuo for 30 min. After drying, the tube was placed under Ar. CD2Cl2
was added. An equimolar amount of a substrate (0.025 mmol) was added
to the mixture at ambient temperature. 1H NMR spectra of the resulting
mixtures were then measured.
[6] a) M. K. Gurjar, Tetrahedron Lett. 2001, 42, 3633–3636; b) L. A. Pa-
quette, I. Efremov, J. Am. Chem. Soc. 2001, 123, 4492–4501; c) T. K.
Maishal, D. K. Shinha-Mahapatra, K. Paranjape, A. Sarkar, Tetrahe-
dron Lett. 2002, 43, 2263–2267; d) L. Ackermann, D. E. Tom, A.
Fꢃrstner, Tetrahedron 2000, 56, 2195–2202.
Procedure for kinetic studies to confirm the change in rate-determining
step:[26] The first-generation Grubbs catalyst was dried in vacuo for
30 min, and an appropriate amount of CD2Cl2 was added under Ar to
prepare a 0.002m solution of first-generation Grubbs catalyst in CD2Cl2.
1,2-Dichloroethane was added to the solution as an internal standard (ca.
0.034m). This solution (0.6 mL) was then transferred to a dried NMR
tube. Just before 1H NMR measurement, a corresponding amount of
enyne substrate (1a, 1j) was added to the NMR tube and the tube was
shaken intensively. The 1H NMR spectrum was then measured continu-
ously. The concentration of the desired product (2a, 2j) was estimated by
integration with 1,2-dichloroethane as an internal standard. Initial reac-
tion rates were estimated from the approximated curves (first order) of
the first 5 minutes under each set of conditions.
[7] In reference [4a], it is reported that a decomposition mechanism
based on an allylic hydroxy group also exists for secondary allylic al-
cohols, although the ring-closing metathesis proceeds more rapidly.
This decomposition decreased the efficiency of the metathesis. For
tertiary allylic alcohols, such decomposition was not observed and
the metathesis proceeded rapidly and efficiently.
[8] a) T. Imahori, H. Ojima, H. Tateyama, Y. Mihara, H. Takahata, Tet-
rahedron Lett. 2008, 49, 265–268; b) D. A. Clark, J. R. Clark, S. T.
Diver, Org. Lett. 2008, 10, 2055–2058.
[9] For effects of a hydroxy group at other positions on Ru–carbene-
catalyzed enyne metathesis, see: a) M. Mori, K. Tonogaki, N. Nishi-
guchi, J. Org. Chem. 2002, 67, 224–226; b) J. A. Smulik, S. T. Diver,
Org. Lett. 2000, 2, 2271–2274.
[10] M. Mori, N. Sakakibara, A. Kinoshita, J. Org. Chem. 1998, 63,
6082–6083.
[11] a) A. J. Giessert, S. T. Diver, Org. Lett. 2005, 7, 351–354; b) A. C.
Tsipis, A. G. Orpen, J. N. Harvey, Dalton Trans. 2005, 2849–2858.
[12] For selected examples, see: a) H. Takahata, Y. Suto, E. Kato, Y.
Yoshimura, H. Ouchi, Adv. Synth. Catal. 2007, 349, 685–693; b) H.
Ouchi, Y. Mihara, H. Takahata, J. Org. Chem. 2005, 70, 5207–5214;
c) N. Asano, K. Ikeda, L. Yu, A. Kato, K. Takebayashi, I. Adachi, I.
Kato, H. Ouchi, H. Takahata, G. W. J. Feet, Tetrahedron: Asymmetry
2005, 16, 223–229; d) A. Kato, N. Kato, E. Kano, I. Adachi, K.
Ikeda, L. Yu, T. Okamoto, Y. Banba, H. Ouchi, H. Takahata, N.
Asano, J. Med. Chem. 2005, 48, 2036–2044; e) H. Takahata, Y.
Banba, H. Ouchi, H. Nemoto, Org. Lett. 2003, 5, 2527–2529; f) H.
Takahata, Y. Banba, H. Ouchi, H. Nemoto, A. Kato, I. Adachi, J.
Org. Chem. 2003, 68, 3603–3607.
[13] X. Z. Zhu, K. A. Sheth, S. Li, H.-H. Chang, J.-Q. Fan, Angew.
Chem. 2005, 117, 7616–7619; Angew. Chem. Int. Ed. 2005, 44, 7450–
7453.
[14] For other examples of (+)-isofagomine syntheses, see: a) H. Ouchi,
Y. Mihara, H. Watanabe, H. Takahata, Tetrahedron Lett. 2004, 45,
7053–7056; b) L. Banfi, G. Guanti, M. Paravidino, R. Riva, Org.
Biomol. Chem. 2005, 3, 1729–1737.
Acknowledgements
We are grateful for the financial support provided by a Grant-in-Aid for
Scientific Research from the Japan Society for the Promotion of Science
(JSPS) (no. 18570012) and the High Technology Research Program of
the Ministry of Education, Culture, Sports, Sciences and Technology,
Japan.
[1] a) R. H. Grubbs, Tetrahedron 2004, 60, 7117–7140; b) A. H. Hovey-
da, A. R. Zhugralin, Nature 2007, 450, 243–251; c) S. T. Diver, A.
Giessert, Chem. Rev. 2004, 104, 1317–1382; d) H. Villar, M. Frings,
C. Bolm, Chem. Soc. Rev. 2007, 36, 55–66; e) A. Fꢃrstner, Angew.
Chem. 2000, 112, 3140–3172; Angew. Chem. Int. Ed. 2000, 39, 3012–
3043; f) Handbook of Metathesis, Vol. 3 (Ed.: R. H. Grubbs), Wiley-
VCH, Weinheim, 2003.
[2] K. C. Nicolaou, P. G. Bulger, D. Sarlah, Angew. Chem. 2005, 117,
4564–4601; Angew. Chem. Int. Ed. 2005, 44, 4490–4527.
[3] For selected examples of substituent effects on Ru-carbene-cata-
lyzed metathesis, see: a) T. A. Kirkland, R. H. Grubbs, J. Org.
Chem. 1997, 62, 7310–7318; b) X. Guo, K. Basu, J. A. Cabral, L. A.
Paquette, Org. Lett. 2003, 5, 789–792; c) B. Kang, J. M. Lee, J.
Kwak, Y. S. Lee, S. Chang, J. Org. Chem. 2004, 69, 7661–7664;
d) N. A. Sheddan, V. B. Arion, J. Mulzer, Tetrahedron Lett. 2006, 47,
6689–6693; e) D. L. Wright, L. C. Usher, M. Estrella-Jimenez, Org.
Lett. 2001, 3, 4275–4277; f) S. Randl, N. Lucas, S. J. Connon, S. Ble-
chert, Adv. Synth. Catal. 2002, 344, 631–633; g) T. Kinoshita, Y.
Sato, M. Mori, Adv. Synth. Catal. 2002, 344, 678–693; h) B. R.
Galan, A. J. Giessert, J. B. Keister, S. T. Diver, J. Am. Chem. Soc.
2005, 127, 5762–5763; i) G. Vassilikogiannakis, L. Margaros, M. Tofi,
Org. Lett. 2004, 6, 205–208; j) F.-D. Boyer, I. Hanna, Eur. J. Org.
[15] The acetylated product 6 in the kinetic transesterification step also
had high enantiopurity (53%, 98% ee). This product could be uti-
lized in (À)-isofagomine synthesis.
[16] G. Pamdey, M. Kapur, Synthesis 2001, 1263–1267.
[17] Two pathways of ring-closing enyne metathesis—“ene-then-yne” or
“yne-then-ene”—are possible. Some mechanistic and computational
studies on ring-closing enyne metathesis support the “ene-then-yne”
pathway; see: a) T. R. Hoye, S. M. Donaldson, T. J. Vos, Org. Lett.
1999, 1, 277–279; b) G. C. Lloyd-Jones, R. G. Margue, J. G. de Vries,
Angew. Chem. 2005, 117, 7608–7613; Angew. Chem. Int. Ed. 2005,
44, 7442–7447; c) J. J. Lippstreu, B. F. Straub, J. Am. Chem. Soc.
2005, 127, 7444–7457. The “ene-then-yne” pathway is also proposed
10770
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 10762 – 10771

Allylic Hydroxy Group Acceleration of Ring-Closing Enyne Metathesis
FULL PAPER
in other reports; see: d) A. Kinoshita, M. Mori, Synlett, 1994, 1020–
1022; e) E. Vedrenne, F. Royer, J. Oble, L. E. Kaꢄm, L. Grimaud,
Synlett, 2005, 2379–2381; f) N. Dieltiens, K. Moonen, C. V. Stevens,
Chem. Eur. J. 2007, 13, 203–214.
2045–2048. DFT calculations have also predicted that NHC-Ru-cat-
alyzed olefin isomerization arising from substrate-induced decompo-
sition of the second-generation Grubbs catalyst should be more
facile than in the case of the first-generation catalyst; see: W. J. van
Rensburg, P. J. Steyberg, W. H. Meyer, M. M. Kirk, G. S. Forman, J.
Am. Chem. Soc. 2004, 126, 14332–14333.
[18] P. Schwab, R. H. Grubbs, J. W. Ziller, J. Am. Chem. Soc. 1996, 118,
100–110.
[19] The acceleration effect of an ethylene atmosphere (reference [7b,c])
also suggests that regeneration of reactive Ru–carbene species is the
rate-determining step.
[20] We cannot rule out the possibility of acceleration by the allylic hy-
droxy group in the case of Ru–vinylcarbene intermediate IM-4.
However, we think that the hydroxy group is too far from the Ru
center to interact and it seems unlikely.
[21] For effects of other substituents on the rate of Ru–carbene-cata-
lyzed enyne metathesis see references [3g, 3h, and 11a].
[22] While the Grubbs second-generation catalyst displays significantly
enhanced activity and thermal stability relative to the first-genera-
tion catalyst, overall selectivity is often compromised by competing
olefin isomerization; see: S. S. Kinderman, J. H. van Maarseveen,
H. E. Schoemaker, H. Hiemstra, F. P. J. T. Rutjes, Org. Lett. 2001, 3,
[23] An interesting acceleration effect of phenol as an additive for olefin
metathesis has been reported; see: G. S. Forman, A. E. McConnell,
R. P. Tooze, W. J. van Rensburg, W. H. Meyer, M. M. Kirk, C. L.
Dwyer, D. W. Serfontein, Organometallics 2005, 24, 4528–4542. In
the present enyne metathesis, similar activation of the catalyst by an
allylic hydroxy group might act in the same way as phenol.
[24] Details of the synthesis of enyne substrates are given in the Support-
ing Information.
[25] W. Herz, R.-R. Juo, J. Org. Chem. 1985, 50, 618–627.
[26] Detailed data of the kinetic studies are given in the Supporting In-
formation.
Received: July 16, 2008
Published online: October 10, 2008
Chem. Eur. J. 2008, 14, 10762 – 10771
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10771