An asymmetric approach towards (-)-aphanorphine and its analogues

Article abstract of DOI:10.1002/ejoc.200801175

A short enantioselective approach towards the alkaloid (-)-aphanorphine and its substituted analogues is described. The enantiomerically pure starting material for the synthesis was obtained by asymmetric hydrogenation in the presence of the Rh-(S)-PipPho

Full text of DOI:10.1002/ejoc.200801175

SHORT COMMUNICATION
DOI: 10.1002/ejoc.200801175
An Asymmetric Approach towards (–)-Aphanorphine and Its Analogues
Pavel A. Donets,[a] Jan L. Goeman,[b] Johan Van der Eycken,[b] Koen Robeyns,[c]
Luc Van Meervelt,[c] and Erik V. Van der Eycken*[a]
Keywords: Alkaloids / Asymmetric catalysis / Heck reaction / Radical reactions / Cyclization
A short enantioselective approach towards the alkaloid
(–)-aphanorphine and its substituted analogues is described.
The enantiomerically pure starting material for the synthesis
was obtained by asymmetric hydrogenation in the presence
of the Rh–(S)-PipPhos complex. Microwave-assisted Heck
cyclization selectively provided the seven-membered 3-
benzazepine ring. Further, the assembly of the tricyclic core
of the alkaloid was accomplished by intramolecular radical
cyclization. The X-ray structure confirming the absolute con-
figuration of the obtained products is described.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2009)
reductive cyclization.[4] Herein we propose an application
of the elaborated reaction for the formal asymmetric syn-
thesis of (–)-aphanorphine and its structural analogues.
Introduction
The 3-benzazepine alkaloid (–)-aphanorphine, first iso-
lated by Shimizu and Clardy,[1] shares structural features of
benzomorphane analgesics such as pentazocine and eptazo-
cine[2] and has been chosen as a synthetic target by many
research groups[3] (Figure 1). Recently, we reported a novel
approach towards the selective and facile formation of 1-
substituted 3-benzazepines through intramolecular Heck
Results and Discussion
Several syntheses[3b–3e] of aphanorphine have been ac-
complished via intermediate 1 (R = H) (Scheme 1). Our ap-
proach towards 1 is based on three crucial stages. Intramo-
lecular radical cyclization provides the required tricyclic
framework. Heck reductive cyclization selectively creates
the seven-membered 3-benzazepine ring. As we described
previously,[4] the substituent R may be either an aromatic
or an aliphatic group. Catalytic enantioselective hydrogena-
tion of the appropriate N-formyl dehydroaminoester pro-
vides enantiomerically pure starting material.
The synthesis starts with readily available bromoaldehyde
2 (Scheme 2) obtained by directed ortho-metalation of p-
anisaldehyde[3b,5] followed by treatment with CBr4. Subse-
quent condensation of 2 with methyl isocyanoacetate pro-
ceeds through the formation of the Schöllkopf 2-oxazol-
ine[6] and provides after workup N-formyl dehydroaminoes-
ters 3 as a mixture of two geometrical E/Z isomers. Thus,
(Z)-3 was isolated in 57% yield with the overall reaction
yield of 78%, which was satisfactory for further elabora-
tion. The geometry of the double bond in (Z)-3 was unam-
Figure 1. (–)-Aphanorphine and its structural analogues.
[a] Laboratory for Organic & Microwave-Assisted Chemistry (LO-
MAC), Department of Chemistry,
Katholieke Universiteit Leuven, Celestijnenlaan 200F, 3001 biguously confirmed by the NOE effect between the amide
Leuven, Belgium
proton and the ortho proton of the aryl ring.
[b] Laboratory of Organic and Bioorganic Synthesis, Department
It is necessary to mention that Gallagher and coworkers
did not succeed[3b] to obtain enantiomerically pure 4 by hy-
drogenation of the (E)-3/(Z)-3 mixture working on the syn-
thesis of aphanorphine. In contrast, a literature survey[7] re-
vealed that dehydroaminoesters similar to 3 have been suc-
cessfully hydrogenated with a Rh catalyst and the readily
available[7b] monodentate phosphoramidite ligand PipPhos.
of Chemistry, Ghent University,
Krijgslaan 281(S4), 9000 Gent, Belgium
[c] Biomolecular Architecture, Department of Chemistry, Kathol-
ieke Universiteit Leuven,
Celestijnenlaan 200F, 3001 Leuven, Belgium
Fax: +32-16-32-79-90
E-mail: erik.vandereycken@chem.kuleuven.be
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
Eur. J. Org. Chem. 2009, 793–796
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
793

E. V. Van der Eycken et al.
SHORT COMMUNICATION
drogen pressure with 2 mol-% of catalyst. We failed to hy-
drogenate the (E)-3 isomer with any appreciable enantio-
selectivity. However, hydrogenation of (E)-3 in the presence
of Wilkinson’s (5 mol-%) catalyst at 25 bar provided race-
mic (Ϯ)-4, which was further used as a standard for the ee
analysis. Compound (Ϯ)-4 was also employed in the synthe-
sis of racemic 1 with R = Ph in the early stages of the syn-
thesis while exploring the general feasibility of the ap-
proach. Subsequent reduction of (–)-4 with an excess
amount of BH3·THF complex yielded amino alcohol 5[8]
without any significant racemization, as confirmed by chi-
ral HPLC analysis after conversion into a cyclic carbamate.
Following the synthetic plan, we further prepared amides
7a–d in good yields by coupling amino alcohol 5 with NHS
derivatives of corresponding propynoic acids 6 (Scheme 3).
Obtained propynoic acid amides 7a–d were converted into
3-benzazepinones 8a–d in accordance with the previously
elaborated approach by using microwave-assisted Heck re-
ductive cyclization.[4] Experiments were carried out in a
multimode Milestone MicroSYNTH microwave reactor in
a sealed vial by applying a maximum power level of 300 W
in the presence of Hermann’s palladacycle[9] (2.5 mol-%) as
a catalyst and HCO2Na (1.5 equiv.) as the reducing reagent.
For compounds 7a–c, the microwave-assisted protocol
worked well to provide 8a–c in good yields, but 7d contain-
ing a bulky TIPS group underwent mainly reduction with
a loss of bromine; 8d was isolated in only 7% yield. How-
ever, when subjected to the same reaction conditions at a
lower temperature of 65 °C and under conventional heating,
7d was successfully transformed into 8d without formation
of the direct reduction product (Scheme 3, yield in paren-
thesis). Under Appel reaction conditions, 8a–d were
smoothly converted into bromides 9a–d, which thus served
as the precursors for the subsequent intramolecular radical
cyclization.
Scheme 1. Synthetic strategy towards compounds 1.
Scheme 2. Enantioselective synthesis of amino alcohol 5. Reagents
and conditions: (i) 1) methyl isocyanoacetate, Cu2O, Et2O; 2)
tBuOK; 3) AcOH; (ii) H2 (1 atm), Rh(cod)2BF4, (S)-PipPhos,
DCM, room temp., 24 h; (iii) BH3·THF, THF, reflux, overnight.
It was also reported that unlike the (Z) isomers, the (E)
isomers are generally difficult to reduce and give low ee
values. Therefore, we tested the (S)-PipPhos ligand, which
is slightly cheaper than the (R) form, and Rh(nbd)2BF4
(nbd = 2,5-norbornadiene) or Rh(cod)2BF4 (cod = 1,5-cy-
clooctadiene) complexes as Rh sources for the asymmetric
hydrogenation of (Z)-3. Indeed, the Rh–PipPhos complex
Scheme 3. Synthesis of bromides 9a–d. Reagents and conditions:
(i) 1) NHS, DCC, DCM; 2) 5, DCM; (ii) Hermann’s palladacycle,
HCO2Na, 110 °C, 20 min, MW. [a] for 7d: 65 °C and conventional
heating for 5 h resulted in the formation of 8d in 73% yield; (iii)
2,6-lutidine, PPh3, CBr4, DCM.
Unfortunately, during our work on Heck cyclization we
exhibited excellent selectivity and provided (–)-4 with high found out that unprotected propynoic acid amides like 7
ee and quantitative yield in both cases. However, Rh(cod)2- with R = H undergo decomposition under the given reac-
BF4 proved to be more reliable and less pressure de- tion conditions.[4] We hoped to find conditions for protode-
manding, as it allowed (–)-4 to be obtained at 1 atm of hy- silylation of 8d or 9d to obtain 3-benzazepinone 8e or 9e
794
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 793–796

An Asymmetric Approach towards (–)-Aphanorphine and Its Analogues
(R = H) as precursors for aphanorphine. Even though the
Acrylamide 7e was found to be unstable on silica during
analogous cleavage of the Si–Csp2 bond in vinylsilanes is column purification and, therefore, was directly subjected
rather rare,[10] we first investigated this transformation on to Heck reaction. The cyclization was completed in 2 h at
the example of compound 10 (Scheme 4), which was pre- 110 °C with conventional heating in the presence of
pared by a similar procedure to that used for 8d.
Pd(PPh3)4 (5 mol-%) as a catalyst and Et3N as a base and
afforded 8e in 69% yield. However, the product in this case
was contaminated with inseparable Ph3PO (24 mol-%). Ap-
plication of Pd(dppf)Cl2 as a catalyst allowed purification
complications to be avoided, even though the yield dropped
down to 61%. Under microwave irradiation conditions, the
reaction time was greatly decreased (15 min) and the yield
was improved to 72%. Bromide 9e was subsequently ob-
tained in excellent yield through the already used protocol.
Finally, we had all precursors 9a–e at hand to proceed
with the intramolecular radical cyclization (Scheme 6).
During optimization of the reaction conditions on the ex-
ample of bromide 9a, the reaction was also applied to the
analogous chloride and methyl xanthate obtained from 8e.
However, complete conversion of the chloride could not be
reached even with a fivefold excess of Bu3SnH. The methyl
xanthate reacted similarly to the bromide, but the product
purification in the case of 9a was easier.
Scheme 4. Attempted protodesilylation of 9d.
Treatment of 10 with aqueous HI in toluene[9b] caused
only decomposition of the starting material. Having tried
TfOH and MeSO3H as strong acidic additives in a DCM/
TFA mixture for protodesilylation of 10, we obtained 11 in
26 and 51% yield, respectively. The yield of 11 was further
increased to 81% by using the HBr (33%)/AcOH reagent
in DCM at room temperature, and this result seemed to be
acceptable. Unfortunately, under these conditions we man-
aged to isolate only the acetylation product from the reac-
tion mixture with alcohol 8d. However, bromide 9d yielded
required 9e in very low yield (15%) together with dibromide
9f (25%).
Then, we decided to obtain compound 8e by conven-
tional intramolecular Heck reaction of acrylamide 7e
(Scheme 5). The latter was prepared by selective acylation
of 5 with acryloyl chloride in aqueous acetone in the pres-
ence of Na2CO3.
Scheme 6. Radical cyclization of bromides 9a–e.
By optimizing the reaction conditions for the cyclization
of 9a we found that the rate of reagent addition (Bu3SnH
and AIBN) strongly influenced the ratio of required cy-
clized 1a and unwanted reduced product.[11] The best ratio
(92:8) was observed with the gradual addition of Bu3SnH
(1.5 equiv.) together with AIBN (0.7 equiv.) over the course
of 3 h. The overall isolated yield was 77%, which corre-
sponds to 71% yield of 9a. The structure of 9a was fully
confirmed with a set of 2D NMR spectroscopic experi-
ments. We tried to improve the efficiency of the cyclization
by using (Me3Si)3SiH. It is known to provide a higher ratio
of cyclized and reduced products[11] than Bu3SnH. Unfortu-
nately, in our case, the use of (Me3Si)3SiH resulted in a
dramatic decrease in the overall yield and no expected im-
provement in the product ratio.
By applying the optimized reaction conditions, analogues
1b and 1c were obtained as the sole products in a high yield
of 81% for both. In the case of 9d, 16% of reduction prod-
ucts was observed, still 1d was isolated in a moderate yield
of 55%. Despite the absence of any significant amount of
the reduction product in the reaction of 9e, corresponding
1e was obtained in 45% yield. An X-ray structure[12] of 1e
was obtained to confirm the absolute configuration (Fig-
ure 2).
Scheme 5. Synthesis of bromide 9e. Reagents and conditions: (i)
acryloyl chloride, acetone/H2O, Na2CO3, 0 °C, 1 h; (ii) Pd(dppf)-
Cl2, Et3N, MeCN, 110 °C; (iii) 2,6-lutidine, PPh3, CBr4, DCM.
Eur. J. Org. Chem. 2009, 793–796
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
795

E. V. Van der Eycken et al.
SHORT COMMUNICATION
[1] N. Gulavita, A. Hori, Y. Shimizu, P. Laszlo, J. Clardy, Tetrahe-
dron Lett. 1988, 29, 4381.
[2] a) S. Shiotani, T. Kometani, K. Mitsuhashi, T. Nozawa, A.
Kurobe, O. Fitsukaichi, J. Med. Chem. 1976, 19, 803; b) D. C.
Palmer, M. J. Strauss, Chem. Rev. 1977, 77, 1.
[3] a) X. Yang, H. Zhai, Z. Li, Org. Lett. 2008, 10, 2457; b) J. F.
Bower, P. Szeto, T. Gallagher, J. Org. Biomol. Chem. 2007, 5,
143; c) M. Katoh, H. Inoue, A. Suzuki, T. Honda, Synlett
2005, 18, 2820; d) J. R. Fuchs, R. L. Funk, Org. Lett. 2001, 3,
3923; e) R. S. Grainger, E. J. Welsh, Angew. Chem. Int. Ed.
2007, 46, 5377; f) M. Li, P. Zhou, H. F. Roth, Synthesis 2007,
55; g) Z. Ma, H. Zhai, Synlett 2007, 161; h) Z. Ma, H. Hu, W.
Xionga, H. Zhaia, Tetrahedron 2007, 63, 7523 and references
cited therein.
[4] P. A. Donets, E. Van der Eycken, Org. Lett. 2007, 9, 3017.
[5] Y. Lear, T. Durst, Can. J. Chem. 1997, 75, 817.
[6] U. Schollkopf, F. Gerhart, R. Schroder, D. Hoppe, Justus Lie-
bigs Ann. Chem. 1972, 766, 116.
Figure 2. ORTEP representation of 1e with thermal displacement
ellipsoids drawn at the 50% probability level.
[7] a) L. Panella, A. M. Aleixandre, G. J. Kruidhof, J. Robertus,
B. L. Feringa, G. Johannes, J. G. de Vries, A. J. Minnaard, J.
Org. Chem. 2006, 71, 2026; b) H. Bernsmann, M. van den Berg,
R. Hoen, A. J. Minnaard, G. Mehler, M. T. Reetz, J. G.
de Vries, B. L. Feringa, J. Org. Chem. 2005, 70, 943.
[8] See Supporting Information: [α]2D0 = –20.2 (c = 0.6, CHCl3),
whereas the reported value has an opposite sign [α]2D0 = +20.2
(c = 0.6, CHCl3).[3b]
[9] a) W. A. Herrmann, C. Brossmer, C. P. Reisinger, T. H. Rierme-
ier, K. Ofele, M. Beller, Chem. Eur. J. 1997, 3, 1357; b) W. A.
Herrmann, C. Brossmer, K. Ofele, C. P. Reisinger, T. Prierme-
ier, M. Beller, H. Fischer, Angew. Chem. 1995, 107, 1989; c)
W. A. Herrmann, C. Brossmer, K. Ofele, C. P. Reisinger, T. Pri-
ermeier, M. Beller, H. Fischer, Angew. Chem. Int. Ed. Engl.
1995, 34, 1844.
Conclusions
In summary, we developed a short asymmetric catalytic
approach towards compounds 1. As the subsequent trans-
formations leading from 1e to aphanorphine have already
been explored,[3b–3d] our route constitutes a formal asym-
metric synthesis of aphanorphine and its analogues.
Supporting Information (see footnote on the first page of this arti-
cle): Experimental procedures, characterization data and NMR
spectra for all compounds.
[10] a) M. D. Bachi, E. Bosch, Tetrahedron Lett. 1988, 29, 2581; b)
K. Utimoto, W. Kitai, H. Nozaki, Tetrahedron Lett. 1975, 16,
2825.
[11] H. Togo in Advanced Free Radical Reactions for Organic Syn-
thesis, Elsevier, 2004.
[12] CCDC-705141 (for 1e) contains the supplementary crystallo-
graphic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
Acknowledgments
Support was provided by the research fund of the Katholieke Uni-
versiteit Leuven and the FWO (Fund for Scientific Research - Flan-
ders, Belgium). P.A.D. is grateful to the Katholieke Universiteit
Leuven for a PhD scholarship.
Received: November 26, 2008
Published Online: January 12, 2009
796
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 793–796