Epoxides as Precursors to 1-Hydroxymethyl Indolizidine and Pyrrolizidine

Article abstract of DOI:10.1055/s-0036-1588173

Herein we describe the synthesis of 1-hydroxymethyl pyrrolizidine and indolizidine alkaloids from epoxy-tethered pyrrole precursors. Two titanium-based methods are evaluated for the construction of the bicyclic scaffold: the Lewis acid mediated cyclizatio

Full text of DOI:10.1055/s-0036-1588173

SYNTHESIS
0
0
3
9
-
7
8
8
1
1
4
3
7
-
2
1
0
X
©
Georg Thieme Verlag Stuttgart · New York
2017, 49, A–F
paper
A
de
Synthesis
S. Hildebrandt et al.
Paper
Epoxides as Precursors to 1-Hydroxymethyl Indolizidine and
Pyrrolizidine
Sven Hildebrandt
Jonathan H. Schacht
Andreas Gansäuer*
OH
OH
O
TiIII
*
[Pt]
H2
*
*
N
TiIV
N
N
n
n
n
Kekulé-Institut für Organische Chemie
und Biochemie, Universität Bonn,
Gerhard-Domagk-Straße 1, 53121 Bonn,
Germany
andreas.gansaeuer@uni-bonn.de
Received: 20.02.2017
OH
Accepted after revision: 17.03.2017
Published online: 18.04.2017
DOI: 10.1055/s-0036-1588173; Art ID: ss-2017-z0104-op
OH
OH
H
N
H
N
H
N
Abstract Herein we describe the synthesis of 1-hydroxymethyl pyr-
rolizidine and indolizidine alkaloids from epoxy-tethered pyrrole pre-
cursors. Two titanium-based methods are evaluated for the construc-
tion of the bicyclic scaffold: the Lewis acid mediated cyclization
developed by Tanis and Raggon and the titanocene(III)-catalyzed reac-
tion proceeding via radical intermediates. The former method, after hy-
drogenation of the pyrrole moiety, ultimately gives rise to enantiopure
trachelantamidine
O
isoretronecanol
tashiromine
OH
OH
[M], H2
*
N
N
N
E
n
n
n
A
D
(–)-tashiromine, being the most concise and atom-economic proce-
dure reported to date.
2
[Cp Ti(III)Cl]
– [Cp2Ti(III)Cl]
Key words pyrrolizidines, indolizidines, epoxides, catalysis, hydroge-
nation
O[Ti(IV)Cp2Cl]
2
O[Ti(IV)Cp Cl]
H
N
N
n
n
Hydroxylated and hydroxymethylated pyrrolizidines
and indolizidines are common motifs in natural products.1
These substituted bicyclic alkaloids have attracted consid-
erable attention in the synthetic community resulting in
numerous total synthetic approaches and new methods for
B
C
Scheme 1 Members of the hydroxymethyl pyrrolizidine and indol-
izidine natural product class. Synthetic access from epoxy-tethered pyr-
roles A via titanocene(III) catalysis and subsequent hydrogenation
2
the stereoselective assembly of bicyclic systems. Some of
the simplest members of the above-mentioned class are the
hydroxymethyl pyrrolizidine trachelantamidine, as well as
its diastereomer isoretronecanol, and the hydroxymethyl
then liberates the catalyst and product D by electron trans-
fer and rearomatization (Scheme 1). In contrast to Lewis
acid mediated epoxide opening that delivers tetrahydroin-
dolizine products from both precursors A (n = 1, 2), the rad-
ical pathway would yield 1-hydroxymethyl-substituted di-
hydropyrrolizine/tetrahydroindolizine, respectively, and
subsequently the natural products after catalytic hydroge-
nation of the pyrrole moiety.
1
indolizidine tashiromine (Scheme 1). In this article we in-
vestigate the use of epoxides as precursors to these 1-hy-
droxymethyl-substituted pyrrolizidines and indolizidines.
Our recent efforts toward the C–H functionalization of
pyrroles led us to investigate the applicability of the titano-
cene(III)-catalyzed epoxide arylation to the synthesis of hy-
droxymethyl-substituted pyrrolizidines and indolizidines
in comparison to the Lewis acid mediated cyclizations de-
From literature precedent it is known that epoxide 1,
upon activation by a Lewis acid, reacts in an S 2-type reac-
N
tion to form tetrahydroindolizine 3 (Scheme 2).3a Initial in-
vestigations to transform epoxide 1 into the isomeric hy-
droxymethyl dihydropyrrolizine 2 under titanocene(III) ca-
talysis showed mixed results: Experiments with 5 or 10
mol% catalyst loading did not show full conversion of the
3
vised by Tanis and Raggon. The atom-economic nature of
the titanocene(III)-catalyzed reaction would allow for a
rapid and catalytic construction of the bicyclic scaffold.4
Mechanistically, the respective epoxy-tethered pyrrole pre-
3b
cursor A is opened by Cp Ti(III)Cl to furnish the secondary
substrate. This is in contrast to previous examples. Full
conversion of the substrate could be achieved with a cata-
2
radical B. This undergoes cyclization to intermediate C that
©
Georg Thieme Verlag Stuttgart · New York — Synthesis 2017, 49, A–F

B
Synthesis
S. Hildebrandt et al.
Paper
OH
OH
OH
OH
O
OH
O
a
a
N
N
N
N
N
N
N
1
2
3
31
4
29
5
6
7
10
40
:
:
90
:
Lewis
acid
100% conv., 52% combined yield
Lewis
acid
1
00% conv., 83% combined yield
Rsec
Rprim
O[Ti]
O[Ti]
Rsec
Rprim
O[Ti]
OH
O[Ti]
OH
+
N
N
+
N
N
N
major
slow cyclization
minor
fast cyclization
3
N
major
fast cyclization
minor
no cyclization
6
Scheme 2 Optimized conditions for the conversion of substrate 1 into
dihydropyrrolizine 2. Reagents and conditions: (a) Cp TiCl (20 mol%),
Mn (40 mol%), Coll·HCl (40 mol%), THF, reflux, 24 h. Coll = 2,4,6-colli-
dine.
2
2
Scheme 3 Optimized conditions for the conversion of substrate 5 into
tetrahydroindolizine 6. Reagents and conditions: (a) Cp TiCl (5 mol%),
Mn (20 mol%), Coll·HCl (20 mol%), THF, reflux, 3 h; Coll = 2,4,6-colli-
dine.
2
2
lyst loading of 20 mol% and prolonged reaction times. This
finding already indicated the formation of by-products pro-
hibiting catalyst turnover.
Indeed, several products were observed: dihydropyr-
rolizine 2 as the major product along with tetrahydroindol-
izine 3 and the uncyclized alcohol 4 in almost equal
amounts. This observation can be explained by the forma-
tion of two different β-titanoxyradicals from the epoxide
90:10 in favor of the secondary radical. The difference in se-
lectivity of opening of 1 can be attributed to the destabiliz-
ing effect of the pyrrole nitrogen on the secondary radical,
which is stronger in the case of 1 than in the case of 5.
While the reaction of 5 is efficient under Cp Ti(III)Cl ca-
2
talysis, it does not allow the stereoselective formation of 6
and derived natural products. For this purpose, we pre-
pared enantiomerically pure epoxide 5 and investigated its
reaction with Lewis acids (Scheme 4). To establish an over-
all concise synthesis, we started from unsaturated alcohol
8. After mesylation and epoxidation to furnish ep-
oxymesylate 9, reaction with pyrrole led to racemic epox-
ide 5. Jacobsen’s hydrolytic kinetic resolution then afforded
substrate upon reaction with Cp Ti(III)Cl. The major sec-
2
sec
5
ondary radical R undergoes slow cyclization to ultimate-
ly form product 2, while at the same time it is prone to side
reactions such as deoxygenation or hydrogen atom abstrac-
6
tion, presumably from THF. Product 4 arises from these
side reaction pathways as well as unquantified amounts of
the respective volatile alkene. The combined yield of the
products is significantly lowered due to this product distri-
a
O
prim
HO
MsO
bution. On the other hand, the minor radical R
is more
7
0%
reactive toward cyclization via a six-membered transition
state that leads to the formation of product 3. The selectivi-
ty of epoxide opening of 1 can be estimated from the prod-
uct mixture to be >70:<30 in favor of the secondary radical.
The possibility of 3 arising from Lewis acid reaction path-
ways can be disregarded on the basis of previously reported
8
9
b
6
5%
O
N
O
N
c
4
2%
(93% of max. 45%)
(R)-5
rac-5
3b
control experiments.
99:1 er
The formation of hydroxymethyl tetrahydroindolizines,
in contrast to dihydropyrrolizines, proved to be less prob-
lematic (Scheme 3). Reaction of 5 with a 5 mol% catalyst
loading led to the formation of the desired product 6 along
with alcohol 7 in an otherwise clean reaction.
iPrO OPr
i
d
6%
8
i
Cl Ti O Pr
OH
O
H
N
N
The higher combined yield and the formation of only
two products indicates altered reactivities of the formed in-
(
R)-6
enantiospecific
99:1 er
sec
termediate radicals. The secondary radical R undergoes
Scheme 4 Preparation of enantiopure 6. Reagents and conditions: (a)
MsCl, Et N, CH Cl , 0 °C, 1 h then mCPBA, CH Cl , 0 °C to r.t., 15 h; (b)
prim
fast cyclization while the primary radical R
will not cy-
3
2
2
2
2
clize via a seven-membered transition state. Now, with less
non-productive reaction pathways, less catalyst is needed
and a higher combined yield is obtained. The selectivity of
epoxide opening can be estimated from the products to
pyrrole, KOH, DMSO, r.t., 24 h; (c) (R,R)-N,N′-bis(3,5-di-tert-butylsali-
cylidene)-1,2-cyclohexanediaminocobalt acetate (2 mol%), H O (0.55
2
equiv), THF, r.t., 18 h; (d) TiCl(Oi-Pr) (2.9 equiv), CH Cl , 0 °C, 1 h. De-
3
2
2
termination of enantiomeric excess via HPLC.
©
Georg Thieme Verlag Stuttgart · New York — Synthesis 2017, 49, A–F

C
Synthesis
S. Hildebrandt et al.
Paper
the practically enantiopure epoxide (R)-5.7 Tanis and
Raggon reported that treatment of rac-5 with TiCl(Oi-Pr)3
led to a selective and high yielding cyclization to form 6.3a
Our cyclization with enantiopure starting material re-
veals that the reaction is fully enantiospecific leading to en-
antiopure cyclization product (R)-6 from (R)-5 with inver-
sion of configuration. Therefore, we have unequivocally
bond migration has been determined to decrease from Pd >
Rh > Pt but still, when enantiopure substrate (R)-6 was sub-
jected to hydrogenation with Rh/Al O in HFIP at ambient
hydrogen pressure, high diastereoselectivity (88:12 dr) but
almost complete racemization was observed (Scheme 6).
2
3
OH
OH
H
N
proven an S 2 mechanism for the cyclization. This consti-
[Rh], H2
(a)
N
tutes a significant addition to the initial investigation by
N
8
Tanis and Raggon. A report by Banwell showed partial ero-
(
R)-6
(–)-10, 62%
62:38 er
sion of diastereoselectivity for a 1,2-trans-disubstituted ep-
99:1 er
oxide during a related InCl -mediated pyrrole cyclization.
Therefore, care has to be taken in the choice of reagents for
3
[
Rh], air
(
R)-6
(R)-6, >95%
83:17 er
8
5:15 er
pyrrole functionalization via an S 2 reaction in order to ob-
(b)
N
<5% oxidation products
tain complete inversion of configuration and hence stereo-
chemically homogeneous products.
OH
With established routes for the synthesis of 1-hy-
droxymethyl-substituted dihydropyrrolizine 2 and enantio-
pure tetrahydroindolizine 6, hydrogenation of the pyrrole
core constitutes access to the respective natural products.
Heterogeneous Rh catalysts have found widespread use
throughout the literature for the hydrogenation of aromatic
compounds at ambient hydrogen pressure and appear es-
[Rh]
[Rh]
rac-10
H2
(R)-6
N
racemization
through
M–H migration
11
Scheme 6 Hydrogenation of (R)-6 to tashiromine 10. Reagents and
conditions: (a) 5% Rh/Al O (100% w/w), H (1 atm), HFIP, r.t., 4 h, 88:12
dr; (b) 5% Rh/Al O (100% w/w), air, HFIP, r.t., 2 h. Enantiomeric excess
2
3
2
2 3
9
was determined by Mosher ester analysis.
pecially convenient for this purpose. Although the prepa-
ration of dihydropyrrolizine 2 did not proceed to satisfac-
tion, minor quantities of pure material could be obtained to
examine the subsequent transformation (Scheme 5). To
date, only one report on the hydrogenation of 2 is known.10
However, missing experimental and spectroscopic details
led us to examine the transformation under conditions
analogously reported by Banwell.8 Full conversion was
achieved at ambient hydrogen pressure using 5% rhodium
on alumina in hexafluoro-2-propanol (HFIP). Analysis of
the crude product showed the clean but unselective forma-
tion of the diastereoisomeric natural products trachelant-
amidine and isoretronecanol (60:40 dr). Retrospectively,
this serves as a structural confirmation of 2.
To examine the underlying racemization process, we in-
vestigated the potential dehydrogenation of 6 to indolizine
11. The hydrogenation of indolizines to indolizidines by
Rh/Al O has recently been reported by Coelho, and thus we
2
3
reasoned that an equilibrium of 6 and 11 might exist in
solution that over time leads to racemization of the sub-
strate.12 Treatment of an enantioenriched sample of 6 with
Rh/Al O under air and analysis of the crude reaction mix-
2
3
ture revealed that only minor amounts (<5%) of 6 were con-
verted into indolizine structures and that no erosion of en-
antiomeric excess takes place. Thus, racemization indeed
occurs by means of double bond migration of intermediate
structures during the hydrogenation process.
Following the established trend for double bond migra-
tion processes, heterogeneous platinum catalysts would
ideally lead to a conservation of enantiomeric excess during
hydrogenation. Only recently, a hydrogenation of indol-
izines was reported that in the presence of trifluoroacetic
OH
OH
H
N
[
Rh], H2
N
(a)
rac-2
trachelantamidine/
isoretronecanol
(
>95%)
acid enables PtO , being otherwise unreactive toward pyr-
2
Scheme 5 Hydrogenation of rac-2. Reagents and conditions: (a) 5%
Rh/Al O (100% w/w), H (1 atm), HFIP, r.t., 4 h, 60:40 dr.
roles at ambient hydrogen pressure, for the formation of in-
2
3
2
12
dolizidines. Under these conditions substrate 6 under-
went smooth conversion (66:34 dr) into the desired prod-
uct (–)-tashiromine in 57% yield and with complete
conservation of the enantiomeric excess (Scheme 7). Pre-
sumably, initial substrate protonation/dearomatization by
the strong acid leads to intermediate 12 that facilitates Pt-
catalyzed hydrogenation.
Although Rh/Al O constitutes a highly active catalyst,
2
3
previous investigations into its double bond hydrogenation
reactivity revealed racemization of adjacent stereocenters
through double bond migration.1 The tendency for hetero-
geneous catalysts to racemize substrates through double
1
©
Georg Thieme Verlag Stuttgart · New York — Synthesis 2017, 49, A–F

D
Synthesis
S. Hildebrandt et al.
Paper
[Pt], H2, TFA
rac-3-(Oxiran-2-yl)propyl Methanesulfonate (9)
(
R)-6
(–)-10, 57%
99:1 er
(a)
99:1 er
To a solution of 4-penten-1-ol (8) (4.307 g, 50.0 mmol, 1.00 equiv) in
CH Cl (100 mL) at 0 °C were added Et N (10.4 mL, 75.0 mmol, 1.50
equiv) and slowly methanesulfonyl chloride (5.8 mL, 75.0 mmol, 1.50
equiv). The mixture was stirred for 1 h at 0 °C and then quenched by
2
2
3
OH
TFA
[Pt-H]
(
R)-6
(–)-10
the addition of sat. NH Cl solution (100 mL). The separated organic
4
N
phase was cooled to 0 °C and treated with mCPBA (70% in H O, 18.490
H
2
H
12
no racemization
g, 75.0 mmol, 1.50 equiv) portionwise. The mixture was stirred for 15
h at r.t. and then quenched by the addition of Na SO (9.450 g, 75.0
2
3
Scheme 7 Conservation of enantiomeric excess during the hydroge-
nation of (R)-6 with Adam’s catalyst in the presence of TFA. (a) PtO (10
mol%), TFA (1 equiv), H (1 atm), MeOH, r.t., 14 h, 66:34 dr. Enantio-
meric excess was determined by Mosher ester analysis
mmol, 1.50 equiv) and H O (100 mL). The organic phase was separat-
2
2
ed and washed with sat. NaHCO solution (3 × 100 mL) and 2 M NaOH
3
2
solution (100 mL). The resulting organic phase was filtered through a
short plug of silica gel and the solvent was removed under reduced
pressure to afford 9 (6.337 g, 35.2 mmol, 70%) as a colorless oil.
IR (neat): 2940, 1345, 1170, 915 cm–1
.
In conclusion, we have shown that epoxy-tethered pyr-
roles can be transformed into hydroxymethyl pyrrolizidines
and indolizidines via cyclization and subsequent hydroge-
nation. Hydroxymethyl dihydropyrrolizines are accessible
through titanocene(III) catalysis, albeit in low yield. Howev-
er, hydroxymethyl tetrahydroindolizines were formed effi-
ciently. In particular, the enantiospecific Lewis acid mediat-
ed cyclization to furnish tetrahydroindolizines and subse-
quent hydrogenation under non-racemizing conditions
represents the most efficient access to enantiopure tashiro-
mine to date.
1
H NMR (500 MHz, CDCl ): δ = 4.32–4.21 (m, 2 H), 2.99 (s, 3 H), 2.92
3
(dtd, J = 6.8, 4.1, 2.6 Hz, 1 H), 2.75 (dd, J = 4.9, 3.9 Hz, 1 H), 2.48 (dd,
J = 4.9, 2.7 Hz, 1 H), 1.91 (dddd, J = 12.4, 6.3, 5.0, 3.4 Hz, 2 H), 1.83–
1.74 (m, 1 H), 1.52 (ddt, J = 14.0, 8.4, 6.9 Hz, 1 H).
13
C NMR (125 MHz, CDCl ): δ = 69.5, 51.5, 46.9, 37.4, 28.5, 25.9.
3
HRMS (ESI): m/z [M + Na]+ calcd for C H O SNa: 203.0354; found:
6
12
4
203.0349.
rac-1-[3-(Oxiran-2-yl)propyl]-1H-pyrrole (5)
To a mixture of powdered KOH (0.561 g, 10.0 mmol, 1.00 equiv) in an-
hydrous DMSO (25 mL) was added pyrrole (0.671 g, 10.0 mmol, 1.00
equiv) and the resulting mixture was stirred for 30 min at r.t. Next,
sulfonate 9 (1.802 g, 10.0 mmol, 1.00 equiv) was added and the mix-
All reactions involving air- or moisture-sensitive compounds were
carried out in oven-dried glassware under argon using standard
Schlenk and vacuum line techniques. All solvents were either dried
and deoxygenated by distillation (THF over Na/K-alloy) before use, or
purified inside an M-Braun MB-SPS-800 solvent purification system
and used after degasification. Commercially available chemicals were
used without further purification. Column chromatography was car-
ried out on neutral alumina (Brockmann grade) or silica gel (230–400
mesh) supplied by Merck and Macherey-Nagel. TLC was performed on
silica gel on aluminum plates (Merck silica gel 60 F254 plates) and the
compounds were detected with the Seebach staining mixture. Melt-
ing points were recorded using a melting point apparatus DigiMelt
MPA 161 from SRS. Optical rotations were determined on a Krüss
ture was stirred for 24 h at r.t. Sat. NH
and the aqueous phase extracted with EtOAc (3 × 25 mL). The com-
bined organic phase was washed with H O, dried over anhydrous
MgSO and the solvent was removed under reduced pressure. The
crude residue was purified by silica gel column chromatography to
yield product 5 (0.980 g, 6.5 mmol, 65%) as a colorless oil.
4
Cl solution (25 mL) was added
2
4
Rf = 0.4 (cyclohexane/EtOAc = 80:20).
1
H NMR (400 MHz, CDCl ): δ = 6.66 (t, J = 2.1 Hz, 2 H), 6.15 (t, J = 2.1
3
Hz, 2 H), 4.08–3.84 (m, 2 H), 2.90 (dtd, J = 6.8, 4.1, 2.7 Hz, 1 H), 2.75
(dd, J = 4.9, 4.0 Hz, 1 H), 2.46 (dd, J = 5.0, 2.7 Hz, 1 H), 2.10–1.84 (m, 2
H), 1.74–1.57 (m, 1 H), 1.44 (ddt, J = 13.7, 8.8, 6.7 Hz, 1 H).
13
C NMR (100 MHz, CDCl ): δ = 120.6, 108.2, 51.9, 49.2, 46.9, 29.7,
3
2
–1
28.2.
P8000-PT polarimeter; [α]D values are given in units of deg·cm ·g
The data are in agreement with the literature.3a
and concentration values in g/100 mL. IR spectra were measured on
1
an ATR-IR Spectrometer NicoletTM 380 instrument as neat films. H,
1
3
19
C and F NMR spectra were recorded on a Bruker AMX 300 Hz or a
(R)-1-[3-(Oxiran-2-yl)propyl]-1H-pyrrole [(R)-5]
1
Bruker AM 400 MHz instrument. H NMR chemical shifts are given in
ppm and calibrated by using the residual undeuterated solvent as in-
ternal reference [CHCl3 (7.26 ppm), C HD (7.16 ppm)]. C NMR
According to the literature,7 (R,R)-(–)-N,N′-bis(3,5-di-tert-butylsali-
cylidene)-1,2-cyclohexanediaminocobalt(II) (60 mg, 2 mol%) was re-
acted with AcOH (60 mg) in toluene (1 mL) under air for 1 h. The vola-
tile components were removed and rac-5 (763 mg, 5.0 mmol, 1.00
13
6
5
chemical shifts are given in ppm and calibrated by using the solvent
peak as internal reference [CHCl3 (77.16 ppm), C D (128.06 ppm)].
6
6
equiv), H O (49 mg, 2.7 mmol, 0.55 equiv) and THF (0.5 mL) were add-
2
High-resolution mass spectroscopic analysis data were obtained on a
Thermoquest MAT 95 CL instrument. The enantiomeric ratio (er) of
the products was determined by chiral stationary phase HPLC on
Knauer equipment using a Daicel Chiralpak AS-3 column and n-hex-
ane/2-propanol as eluent. The racemic compounds were used to de-
termine optimal HPLC conditions.
ed. The mixture was stirred for 18 h at r.t. and the volatiles were re-
moved under reduced pressure. The crude residue was purified by sil-
ica gel column chromatography to yield product (R)-5 [317 mg, 2.1
mmol, 42% (93% of maximum yield)] as a colorless oil.
20
Rf = 0.4 (cyclohexane/EtOAc = 80:20); [α]
D
+14.8 (c 1.0, CHCl3).
HPLC: Daicel Chiralpak AS-3; n-hexane/2-propanol (90:10); flow rate
.0 mL/min; t = 8.6 min (major, R), t = 10.1 min (minor, S); 99:1 er.
1
R
R
©
Georg Thieme Verlag Stuttgart · New York — Synthesis 2017, 49, A–F

E
Synthesis
S. Hildebrandt et al.
Paper
13
(R)-(5,6,7,8-Tetrahydroindolizin-8-yl)methanol [(R)-6]
C NMR (75 MHz, C D ): δ = 137.4, 114.1, 113.0, 99.8, 65.3, 45.3, 40.5,
6
6
31.5.
A solution of (R)-5 (315 mg, 2.1 mmol, 1.00 equiv) in CH Cl (20 mL)
2
2
HRMS (ESI): m/z [M + H]+ calcd for C H12NO: 138.0919; found:
was cooled to 0 °C and TiCl(Oi-Pr) (6 mL, 1 M in hexane, 2.86 equiv)
3
8
was added. The mixture was stirred for 1 h at 0 °C and quenched by
138.0917.
the addition of H O (20 mL). The aqueous phase was extracted with
2
CH Cl (3 × 20 mL) and the combined organic phase dried over MgSO .
2
2
4
rac-(Hexahydro-1H-pyrrolizin-1-yl)methanol(Trachelantamidine,
Isoretronecanol)
The solvent was removed under reduced pressure. The crude residue
was purified by alumina column chromatography
cyclohexane/EtOAc/MeOH = 90:10:1) to yield product (R)-6 (273 mg,
To a solution of rac-2 (8 mg) in hexafluoro-2-propanol (0.5 mL) was
(
added 5% Rh/Al O (8 mg). The mixture was degassed and backfilled
2
3
1.8 mmol, 86%) as a colorless solid.
with H gas four times and then stirred under H (1 atm) for 4 h. The
2
2
20
Mp 40–42 °C; R = 0.2 (cyclohexane/EtOAc = 80:20); [α]
–43.1 (c
f
D
solution was filtered through a short plug of silica gel and the remain-
ing solid was washed with MeOH (5 mL). The solvent was removed
under reduced pressure to yield a mixture of trachelantamidine (A)
and isoretronecanol (B) (8 mg, 60:40 dr).
0
.4, CHCl3).
1
H NMR (400 MHz, C D ): δ = 6.37 (t, J = 2.2 Hz, 1 H), 6.34 (dd, J = 3.5,
6
6
2.7 Hz, 1 H), 6.06 (dt, J = 3.1, 1.4 Hz, 1 H), 3.62–3.48 (m, 2 H), 3.35–
3.18 (m, 2 H), 2.79–2.66 (m, 1 H), 1.59 (ddt, J = 12.3, 6.8, 5.0 Hz, 1 H),
1.48–1.34 (m, 2 H), 1.26 (ddtd, J = 14.8, 12.0, 6.4, 5.8, 3.0 Hz, 2 H).
13
C NMR (125 MHz, CDCl ): δ = 68.1 (A), 66.6 (B), 65.6 (A), 63.7 (B),
3
55.7 (B), 54.9 (A), 54.6 (A), 54.1 (B), 48.4 (A), 44.1 (B), 32.0 (A), 30.2 (A),
13
C NMR (100 MHz, C D ): δ = 129.4, 119.4, 108.4, 104.5, 65.9, 45.2,
27.4 (B), 26.7 (B), 26.1 (B), 25.8 (A).
6
6
37.4, 24.6, 22.3.
The data are in agreement with the literature.13
HPLC: Daicel Chiralpak AS-3; n-hexane/2-propanol (90:10); flow rate
.0 mL/min; t = 9.5 min (minor, S), t = 11.1 min (major, R); 99:1 er.
1
[(8R,8aS)-Octahydroindolizin-8-yl]methanol [(–)-Tashiromine]
R
R
The data are in agreement with the literature.3a
(10)
rac-(5,6,7,8-Tetrahydroindolizin-8-yl)methanol (6)
2
Hydrogenation with Rh/Al O3
To solution of (R)-6 (99:1 er, 40 mg, 0.26 mmol, 1.00 equiv) in hexa-
A Schlenk tube was filled with a mixture of Cp TiCl (6 mg, 0.025
2
2
fluoro-2-propanol (1 mL) was added 5% Rh/Al O (40 mg). The mix-
mmol, 0.05 equiv), Mn powder (6 mg, 0.1 mmol, 0.20 equiv) and an-
hydrous 2,4,6-trimethylpyridinium chloride (Coll·HCl) (16 mg, 0.1
mmol, 0.20 equiv). The solids were evacuated and flushed with argon.
Anhydrous THF (5 mL) and substrate 5 (76 mg, 0.5 mmol, 1.00 equiv)
were added. The tube was equipped with a reflux condenser and the
mixture was stirred under reflux (95 °C bath temperature) for 3 h. Af-
ter cooling, the solvent was removed under reduced pressure and the
crude residue was purified by alumina column chromatography
2
3
ture was degassed and backfilled with H2 gas four times and then
stirred under H (1 atm) for 4 h. The solution was filtered through a
2
short plug of silica gel and the remaining solid was washed with
MeOH (5 mL). The solvent was removed under reduced pressure to
yield the crude product in an 88:12 diastereomeric ratio (determined
by NMR). The crude residue was purified by silica gel column chro-
matography [CH Cl /MeOH/NH (aq) = 85:15:1] to yield product (–)-
2
2
3
10 (25 mg, 0.16 mmol, 62%) as a yellow oil.
(cyclohexane/EtOAc/MeOH = 90:10:1) to yield an inseparable mix-
20
ture of 6 and 7 (63 mg, 83%) as a colorless oil.
R = 0.2 [CH Cl /MeOH/NH (aq) = 85:15:1]; [α] –8.4 (c 0.25, CHCl3).
f
2
2
3
D
Rf = 0.2 (cyclohexane/EtOAc = 80:20).
The enantiomeric ratio was determined by esterification of the prod-
uct with both (R)-MTPA-Cl and (S)-MTPA-Cl in the presence of pyri-
The products were identified by combined NMR and GC–MS analysis.
14
19
dine according to the literature. Comparison of the F NMR spectra
of the crude esterification products showed two signals for each ester
in the ratio 62:38 (see the Supporting Information).
rac-(2,3-Dihydro-1H-pyrrolizin-1-yl)methanol (2)
A Schlenk tube was filled with a mixture of Cp TiCl (25 mg, 0.1
2
2
mmol, 0.2 equiv), Mn powder (11 mg, 0.2 mmol, 0.40 equiv) and an-
hydrous 2,4,6-trimethylpyridinium chloride (Coll·HCl) (32 mg, 0.2
mmol, 0.40 equiv). The solids were evacuated and flushed with argon.
Anhydrous THF (5 mL) and substrate 1 (69 mg, 0.5 mmol, 1.00
equiv) were added. The tube was equipped with a reflux condenser
and the mixture was stirred under reflux (95 °C bath temperature) for
Investigation on the Formation of Indolizine 11
To solution of (R)-6 (85:15 er, 20 mg, 0.13 mmol, 1.00 equiv) in hexa-
3a
fluoro-2-propanol (0.5 mL) was added 5% Rh/Al O (20 mg). The mix-
2 3
ture was stirred under air for 2 h. The solution was filtered through a
short plug of silica gel and the remaining solid washed with MeOH (5
mL). The solvent was removed under reduced pressure and the crude
product (20 mg) was subjected to NMR and HPLC analysis [see data
for (R)-6].
24 h. After cooling, the solvent was removed under reduced pressure
and the crude residue was purified by alumina column chromatogra-
phy (cyclohexane/EtOAc/MeOH = 90:10:1) to yield an inseparable
mixture of 2, 3 and 4 (36 mg, 52%) as a colorless oil.
Hydrogenation with PtO2
Rf = 0.1 (cyclohexane/EtOAc = 80:20).
To a solution of (R)-6 (99:1 er, 50 mg, 0.33 mmol, 1.00 equiv) in MeOH
The products were identified by combined NMR and GC–MS analysis.
(2 mL) was added PtO2 (8 mg, 0.033 mmol, 0.10 equiv) and TFA (38
mg, 0.33 mmol, 1.00 equiv). The mixture was degassed and backfilled
with H O gas four times and then stirred under H (1 atm) for 14 h.
The solution was filtered through a short plug of silica gel and the re-
maining solid washed with MeOH (10 mL). The solvent was removed
under reduced pressure to yield the crude product with a 66:34 dia-
stereomeric ratio (determined by NMR). The crude residue was puri-
Data of Analytically Pure 2
IR (neat): 3355, 2925, 2880, 1465, 1285, 1030 cm–1
2
2
.
1
H NMR (300 MHz, C D ): δ = 6.42 (d, J = 2.3 Hz, 2 H), 5.98 (dt, J = 2.4,
6
6
1
3
1
.2 Hz, 1 H), 3.44–3.29 (m, 3 H), 3.20 (ddd, J = 10.2, 8.3, 5.9 Hz, 1 H),
.02–2.85 (m, 1 H), 1.98 (dtd, J = 12.8, 8.3, 5.7 Hz, 1 H), 1.90–1.75 (m,
H).
©
Georg Thieme Verlag Stuttgart · New York — Synthesis 2017, 49, A–F

F
Synthesis
S. Hildebrandt et al.
Paper
fied by silica gel column chromatography [CH Cl /MeOH/NH
3
(3) (a) Tanis, S. P.; Raggon, J. W. J. Org. Chem. 1987, 52, 819.
(b) Hildebrandt, S.; Gansäuer, A. Angew. Chem. Int. Ed. 2016, 55,
9719.
2
2
(aq) = 85:15:1] to yield product (–)-10 (29 mg, 0.19 mmol, 57%) as a
yellow oil.
Rf = 0.2 [CH Cl /MeOH/NH (aq) = 85:15:1]; [α]D20 –29.2 (c 0.36,
CHCl3).
(4) For examples of efficient titanocene(III)-catalyzed radical reac-
tions, see: (a) Gansäuer, A.; Rinker, B.; Ndene-Schiffer, N.;
Pierobon, M.; Grimme, S.; Gerenkamp, M.; Mück-Lichtenfeld, C.
Eur. J. Org. Chem. 2004, 2337. (b) Gansäuer, A.; Fleckhaus, A.;
Lafont, M. A.; Okkel, A.; Kotsis, K.; Anoop, A.; Neese, F. J. Am.
Chem. Soc. 2009, 131, 16989. (c) Gansäuer, A.; Worgull, D.;
Knebel, K.; Huth, I.; Schnakenburg, G. Angew. Chem. Int. Ed.
2
2
3
The enantiomeric ratio was determined by esterification of the prod-
uct with both (R)-MTPA-Cl and (S)-MTPA-Cl in the presence of pyri-
14
19
dine according to the literature. Comparison of the F NMR spectra
of the crude esterification products displayed no signals due to the
opposite diastereomer (see the Supporting Information).
1
2009, 48, 8882. (d) Gansäuer, A.; Behlendorf, M.; von Laufen-
H NMR (700 MHz, CDCl ): δ = 3.63 (dd, J = 10.8, 4.8 Hz, 1 H), 3.48 (dd,
berg, D.; Fleckhaus, A.; Kube, C.; Sadasivam, D. V.; Flowers, R. A.
II Angew. Chem. Int. Ed. 2012, 51, 4739. (e) Streuff, J.; Feurer, M.;
Bichovski, P.; Frey, G.; Gellrich, U. Angew. Chem. Int. Ed. 2012, 51,
8661. (f) Gansäuer, A.; Klatte, M.; Brandle, G. M.; Friedrich, J.
Angew. Chem. Int. Ed. 2012, 51, 8891. (g) Gansäuer, A.; Kube, C.;
Daasbjerg, K.; Sure, R.; Grimme, S.; Fianu, G.; Sadasivam, D. V.;
Flowers, R. A. II J. Am. Chem. Soc. 2014, 136, 1663. (h) Gansäuer,
A.; Hildebrandt, S.; Michelmann, A.; Dahmen, T.; von Laufen-
berg, D.; Kube, C.; Fianu, G. D.; Flowers, R. A. II Angew. Chem. Int.
Ed. 2015, 54, 7003. (i) Streuff, J.; Feurer, M.; Frey, G.; Steffani, A.;
Kacprzak, S.; Weweler, J.; Leijendekker, L. H.; Krakert, D.;
Plattner, D. A.; Dietmar, H. J. Am. Chem. Soc. 2015, 137, 14396.
3
J = 10.8, 6.5 Hz, 1 H), 3.14–3.06 (m, 2 H), 2.09 (q, J = 8.9 Hz, 1 H), 1.96
(dtt, J = 15.7, 6.2, 3.5 Hz, 2 H), 1.91–1.85 (m, 1 H), 1.81 (dddd, J = 16.2,
14.8, 8.1, 3.4 Hz, 1 H), 1.74 (ddq, J = 12.9, 5.3, 2.8 Hz, 1 H), 1.72–1.60
(m, 3 H), 1.56–1.46 (m, 2 H), 1.06 (tdd, J = 13.0, 11.8, 4.5 Hz, 1 H).
13
C NMR (175 MHz, CDCl ): δ = 66.6, 65.7, 54.2, 52.8, 44.6, 29.1, 27.6,
3
25.1, 20.8.
The data are in agreement with the literature.2b
Acknowledgment
(
j) Henriques, D. S. G.; Zimmer, K.; Klare, S.; Meyer, A.; Rojo-
We thank the DFG (SFB 813 ‘Chemistry at Spin Centers’) for support.
S.H. thanks the Jürgen Manchot Stiftung for a Ph.D. fellowship. We
gratefully acknowledge the support of Andreas J. Schneider in HPLC
measurements.
Wiechel, E.; Bauer, M.; Sure, R.; Grimme, S.; Schiemann, O.;
Flowers, R. A. II; Gansäuer, A. Angew. Chem. Int. Ed. 2016, 55,
7
671. (k) Funken, N.; Mühlhaus, F.; Gansäuer, A. Angew. Chem.
Int. Ed. 2016, 55, 12030.
(
(
(
5) (a) Gansäuer, A.; Pierobon, M. Synlett 2000, 1357. (b) Gansäuer,
A.; Pierobon, M.; Bluhm, H. Synthesis 2001, 2500.
6) Justicia, J.; Jimenez, T.; Morcillo, S. P.; Cuerva, J. M.; Oltra, J. E.
Tetrahedron 2009, 65, 10837.
7) Schaus, S. E.; Brandes, B. D.; Larrow, J. F.; Tokunaga, M.; Hansen,
K. B.; Gould, A. E.; Furrow, M. E.; Jacobsen, E. N. J. Am. Chem. Soc.
Supporting Information
Supporting information for this article is available online at
http://dx.doi.org/10.1055/s-0036-1588173.
S
u
p
p
ortioIgnfrm oaitn
S
u
p
p
ortioIgnfrm oaitn
2
002, 124, 1307.
8) Banwell, M. G.; Beck, D. A. S.; Smith, J. A. Org. Biomol. Chem.
004, 2, 157.
(
References
2
(
1) (a) Brambilla, M.; Davies, S. G.; Fletcher, A. M.; Thomson, J. E.
(9) Siegel S.; Rhodium on Alumina, In e-Eros Encyclopedia of Reagents
for Organic Synthesis; 2001; DOI: 10.1002/047084289X.rr003.
(10) Ortiz, C.; Greenhouse, R. Tetrahedron Lett. 1985, 26, 2831.
(11) Rylander, P. N. Catalytic Hydrogenation over Platinum Metals;
Academic Press: New York, 1967.
Tetrahedron: Asymmetry 2014, 25, 387. (b) Michael, J. P. Nat.
Prod. Rep. 2007, 24, 191.
(2) For selected examples of new methodology applied to the syn-
thesis of hydroxymethyl indolizidines and pyrrolizidines, see:
(
a) Nagao, Y.; Dai, W.-M.; Ochiai, M.; Tsukagoshi, S.; Fujita, E.
(12) Teodoro, B. V. M.; Correia, J. T. M.; Coelho, F. J. Org. Chem. 2015,
80, 2529.
(13) Brambilla, M.; Davies, S. G.; Fletcher, A. M.; Roberts, P. M.;
Thomson, J. E. Tetrahedron 2014, 70, 204.
J. Am. Chem. Soc. 1988, 110, 289. (b) Kim, S.-H.; Kim, S.-I.; Lai, S.;
Cha, J. K. J. Org. Chem. 1999, 64, 6771. (c) Bates, R. W.;
Boonsombat, J. J. Chem. Soc., Perkin Trans.
1 2001, 654.
(
d) Conrad, J. C.; Kong, J.; Laforteza, B. N.; Macmillan, D. W. C.
(14) (a) Dale, J. A.; Mosher, H. S. J. Am. Chem. Soc. 1973, 95, 512.
(b) Hoye, T. R.; Jeffrey, C. S.; Shao, F. Nat. Protoc. 2007, 2, 2451.
J. Am. Chem. Soc. 2009, 131, 11640. (e) Cutter, A. C.; Miller, I. R.;
Keily, J. F.; Bellingham, R. K.; Light, M. E.; Brown, R. C. D. Org.
Lett. 2011, 13, 3988. (f) Koley, D.; Krishna, Y.; Srinivas, K.; Khan,
A. A.; Kant, R. Angew. Chem. Int. Ed. 2014, 53, 13196.
©
Georg Thieme Verlag Stuttgart · New York — Synthesis 2017, 49, A–F