Efficient, stereodivergent access to 3-piperidinols by traceless P(OEt)3 cyclodehydration

Article abstract of DOI:10.1021/ol4026588

A stereodivergent and highly diastereoselective (dr up to >19:1 for both isomers), step economic (5-6 steps), and scalable synthesis (up to 14 g) of cis- and trans-2-substituted 3-piperidinols, the core motif of numerous bioactive compounds, providing efficient access to the NK-1 inhibitor L-733,060 is presented. Additionally, a "traceless" (referring to the simplified byproduct separation) cyclodehydration realizing simple P(OEt)₃ as a substitute for PPh₃ is developed.

Full text of DOI:10.1021/ol4026588

ORGANIC
LETTERS
2013
Vol. 15, No. 20
5178–5181
Efficient, Stereodivergent Access
to 3‑Piperidinols by Traceless
P(OEt)3 Cyclodehydration
Peter H. Huy* and Ari M. P. Koskinen*
Aalto University, School of Chemical Technology, Laboratory of Organic Chemistry,
Kemistintie 1, 00076 Espoo, Finland
ari.koskinen@aalto.fi; peter.huy@uni-koeln.de
Received August 8, 2013
ABSTRACT
A stereodivergent and highly diastereoselective (dr up to >19:1 for both isomers), step economic (5ꢀ6 steps), and scalable synthesis (up to 14 g)
of cis- and trans-2-substituted 3-piperidinols, the core motif of numerous bioactive compounds, providing efficient access to the NK-1 inhibitor
L-733,060 is presented. Additionally, a “traceless” (referring to the simplified byproduct separation) cyclodehydration realizing simple P(OEt)3 as
a substitute for PPh3 is developed.
1,2-Amino alcohols are a frequent motif found in many
pharmacologically active natural products,1,2 chiral
auxiliaries,3 and catalysts for asymmetric synthesis.4 In-
deed, numerous natural products and other bioactive
compounds are derived from a 2-substituted 3-hydroxy
piperidine scaffold (as one type of an 1,2-amino alcohol)
(Figure 1).1aꢀd,2a,2b Representative examples are the
selective nonpeptidic human neurokinin-1 (NK-1) sub-
stance P receptor antagonists L-733,0605 and CP-99,9946
and the natural products febrifugine (antimalarial)7 and
halofuginone (antiprotozoal, commercial trade names
Halocur (lactate salt) and Stenorol (hydrobromide salt)).8
Other prominent examples are 3-hydroxy pipecolic
acids, which serve as (conformationally restricted) substi-
tutes of proline and serine9 and have been incorporated in
various bioactive peptidomimetics,10 and the iminosugar
(1) For reviews on natural products based on a 1,2-amino alcohol
scaffold: (a) Karjalainen, O. K.; Koskinen, A. M. P. Org. Biomol. Chem.
2012, 10, 4311–4326. (b) Wijdeven, M. A.; Willemsen, J.; Rutjes, F. P. J.
T. Eur. J. Org. Chem. 2010, 2831–2844. (c) Huang, P.-Q. Synlett 2006,
1133–1149. (d) Nemr, A. L. Tetrahedron 2000, 56, 8579–8629.
(e) Bergmeier, S. C. Tetrahedron 2000, 56, 2561–2576.
(2) For reviews on imino sugars and sphingolipids: (a) Horne, G.;
Wilson, F. X.; Tinsley, J.; Williams, D. H.; Storer, R. Drug Discovery
Today 2011, 16, 107–118. (b) Winchester, B. G. Tetrahedron: Asymmetry
2009, 20, 645–651. (c) Pruett, S. T.; Bushnev, A.; Hagedorn, K.; Adiga,
M.; Haynes, C. A.; Sullards, M. C.; Liotta, D. C.; Merrill, A. H., Jr.
J. Lipid Res. 2008, 49, 1621–1639. (d) Brunner, M.; Koskinen, A. M. P.
Curr. Org. Chem 2004, 8, 1629–1645.
(6) (a) Desai, M. C.; Lefkowitz, S. L.; Thadeio, P. F.; Longo, K. P.;
Snider, R. M. J. Med. Chem. 1992, 35, 4911–4913. (b) Rosen, T.; Seeger,
T. F.; McLean, S.; Desai, M. C.; Guarino, K. J.; Bryce, D.; Pratt, K.;
Heym, J. J. Med. Chem. 1993, 36, 3197–3201. (c) Data, P.; Srivastava, S.;
Coutinho, E.; Govil, G. Curr. Top. Med. Chem. 2004, 4, 75–103.
(7) (a) Koepfli, J. B.; Mead, J. F.; Brockman, J. A., Jr. J. Am. Chem.
Soc. 1947, 69, 1837. (b) Koepfli, J. B.; Mead, J. F.; Brockman, J. A., Jr.
J. Am. Chem. Soc. 1949, 71, 1048–1054.
(8) Linder, M. R.; Heckeroth, A. R.; Najdrowski, M.; Daugschies,
A.; Schollmeyer, D.; Miculka, C. Bioorg. Med. Chem. Lett. 2007, 17,
4140–4143.
(3) (a) Gacs-Baitz, E.; Delle Monache, G.; Misiti, D.; Nevolam, L.;
Botta, B.; Zappia, G.; Cancelliere, G. Curr. Org. Synth. 2007, 4, 238–307.
(b) Prakash, I.; Schaad, D. R. Chem. Rev. 1996, 96, 835–875.
(4) For reviews on asymmetric reactions with amino alcohol derived
chiral ligands: (a) Singh, P. K.; Singh, V. K. Pure Appl. Chem. 2012, 84,
1651–1657. (b) Nishiyama, H.; Ito, J. Chem. Commun. 2010, 46, 203–
212. (c) Desimoni, G.; Faita, G.; Jørgensen, K. A. Chem. Rev. 2006, 106,
3561–3651.
(9) For reviews on constrained amino acid analogs: (a) Grauer, A.;
€
Konig, B. Eur. J. Org. Chem. 2009, 5099–5111. (b) Hanessian, S.;
McNaughton-Smith, G.; Lombart, H.-G.; Lubell, W. D. Tetrahedron
1997, 53, 12789–12854.
(10) (a) Maillard, M. C.; Brookfield, F. A.; Courtney, S. M.;
Eustache, F. M.; Gemkow, M. J.; Handel, R. K.; Johnson, L. C.; Johnson,
~
ꢀ
P. D.; Kerry, M. A.; Krieger, F.;Meniconi, M.; Munoz-Sanjuan, I.; Palfrey,
J. J.; Park, H.; Schaertl, S.; Taylor, M. G.; Weddell, D.; Dominguez, C.
Bioorg. Med. Chem. 2011, 19, 5833–5851. (b) Quibell, M.; Benn, A.; Flinn,
N.; Monk, T.; Ramjee, M.; Wang, Y.; Watts, J. Bioorg. Med. Chem. 2004,
12, 5689–5710.
(5) (a) Baker, R.; Harrison, T.; Swain, C. J.; Williams, B. J. J. Eur.
Patent, 0528495A1, 1993. (b) Harrison, T.; Williams, B. J.; Swain, C. J.;
Ball, R. G. Bioorg. Med. Chem. Lett. 1994, 4, 2545–2550.
r
10.1021/ol4026588
Published on Web 10/03/2013
2013 American Chemical Society

swainsonine, which is inter alia a new potential chemother-
apeutic agent.11 The high importance of further derivatives
is underlined by recent patents on analogs of halofuginone
as inhibitors for tRNA synthetases.12
The majority of the reported1bꢀd,13,14 syntheses suffer
from some drawbacks: They are elaborate (far more than
10 steps), specific on one of the above-mentioned targets
(either in cis- or trans-configuration), and not proven to
be scalable. Considering the versatile pharmacological
activities of compounds based on the 3-piperidinol scaf-
fold, the development of a stereodivergent, scalable, and
efficient synthetic access is highly desirable.
Herein, we report a step economic (5ꢀ6 steps), scalable,
and stereodivergent synthesis of trans- and cis-2-substituted
3-piperidinols A in high diastereoselectivities (up to >19:1)
and enantiopurities (ee = 90ꢀ99%) originating from the
ketone intermediates C (Figure 1).
In the syntheses of potentially new drug candidates
scalability is a significant factor to provide sufficient sub-
stance amounts for clinical testing.15 Furthermore, alter-
natives in reactions driven by the formation of phosphine
oxides from phosphines (e.g., the Appel and Mitsunobu
reactions) are highly desired to improve atom economy
(reduced waste amounts) and to circumvent difficulties in
the separation of these byproducts.16 Numerous protocols
have been developed to improve these issues, mostly based
on polymer bound or otherwise modified (more complex)
phosphines.16 Surprisingly, in this context simple and
inexpensive phosphites (P(OR)3) have only been applied
as phosphine substitutes in one single and specific
example.17
Figure 1. Selected examples for bioactive piperidine derivatives
and retrosynthetic analysis of 3-hydroxyl piperidines A (X = OH
or leaving group, [Hꢀ] = hydride reducing agent, TS = transition
state).
(11) (a) Guengerich, F. P.; DiMari, S. J.; Broquist, H. P. J. Am.
Chem. Soc. 1973, 95, 2055–2056. (b) Elbein, A. D. FASEB J. 1991, 5,
3055–3063.
(12) (a) Teitelbaum, S. L.; Deselm, C. J. U.S. Pat. Appl. Publ. (2011),
US20110311519A1. (b) Keller, T.; Mazitschek, R.; Whitman, M.; Lee, J. U.
S. patent 2011/0263532A1. (c) Keller, T.; Mazitschek, R.; Whitman, M. PCT
Int. Appl. (2010), WO 2010019210 A2.
We intended to control the diastereoselectivity in A
(cis/trans) through a targeted protecting group (PG) ma-
nipulation resulting in the retrosynthetic analysis shown in
Figure 1:18 Reduction of the common precursor ketone C
(derived from amino acids) should deliver the syn amino
alcohol B according to the FelkinꢀAnh model (due to the
sterically demanding ꢀNBnPG carbamate function).
Further PG cleavage and cyclization should give cis-A.
On the other hand, initial deprotection of C (to liberate the
Lewis basic ꢀNHBn amino moiety) and subsequent re-
duction toward a Cram chelate transition state should
deliver the anti-amino alcohol B. After subsequent cycliza-
tion trans-A would result.
At the outset L-alanine 1a, L-phenylalanine 1b, and
L-phenylglycine 1c were converted to their N-benzyl-N-
Cbz protected derivatives 2aꢀ2c in a novel practical one-
pot procedure through the combination of Quitt’s reductive
benzylation protocol19 and SchottenꢀBaumann acylation
in 70ꢀ79% yield (Scheme 1). Thereby not only one workup
was spared, but also the overall yield was improved signifi-
cantly (e.g., 40% (two steps) f 70% for 2a).
(13) For leading references, see: (a) Pansare, S. A.; Paul, E. K. Org.
Biomol. Chem. 2012, 10, 2119–2125. (b) Bilke, J. L.; Moore, S. P.;
O’Brien, P.; Gilday, J. Org. Lett. 2009, 11, 1935–1938. (c) Cochi, A.;
Burger, B.; Navarro, C.; Pardo, D. G.; Cossy, J.; Zhao, Y.; Cohen, T.
Synlett 2009, 2157–2161. (d) Lemire, A.; Grenon, M.; Pourashraf, M.;
Charette, A. B. Org. Lett. 2004, 6, 3517–3520.
(14) For a review on the synthesis of L-733,060 and L-733,061, see:
(a) Cochi, A.; Pardo, D. G.; Cossy, J. Heterocycles 2012, 86, 89–116. For
reviews on the synthesis of (3-hydroxy-) pipecolic acid derivatives, see:
(b) Cochi, A.; Pardo, D. G.; Cossy, J. Eur. J. Org. Chem. 2013, 809–829.
(c) Cant, A. A.; Sutherland, A. Synthesis 2012, 44, 1935–1950.
(d) Kadouri-Puchot, C.; Comesse, S. Amino Acids 2005, 29, 101–130.
(15) (a) Shioiri, T.; Izawa, K.; Konoike, T.; Karpf. M. Pharmaceu-
tical Process Chemistry; Wiley VCH: Weinheim, 2010; pp 1ꢀ37.
(b) Roberge, D. M. Org. Process Res. Dev. 2004, 8, 1049–1053.
(16) For reviews, see: (a) Constable, D. J. C.; Dunn, P. J.; Hayler,
J. D.; Humphrey, G. R.; Leazer, J. L., Jr.; Linderman, R. J.; Lorenz, K.;
Manley, J.; Pearlman, B. A.; Wells, A.; Zaks, A.; Zhang, T. Y. Green
Chem. 2007, 9, 411–420. (b) Dandapani, S.; Curran, D. P. Chem.;Eur.
J. 2004, 10, 3130–3138. (c) Dembinski, R. Eur. J. Org. Chem. 2004, 2763–
2772.
(17) For a Mitsunobu condensation of a guanine derived nucleoside
analog with activated benzylic alcohols utilizing P(OiPr)3 providing
simplified byproduct separation through improved water solubility
ꢀ
(of (OdP(OiPr)3), see: (a) Veliz, E. A.; Beal, P. A. Tetrahedron Lett.
2006, 47, 3153–3156. In our case we were not able to remove stoichio-
metric amounts of OP(OEt)3 (which is more hydrophilic than OP(OiPr)3)
through an aqueous workup (without saponification). P(OEt)5, prepared
from P(OEt)3 with diethyl peroxide and ethyl benzenesulfenate, respec-
tively, in an additional step, was reported to induce cyclization of diols to
furans and pyrans: (b) Chang, B. C.; Conrad, W. E.; Denney, D. B.;
Denney, D. Z.; Edelman, R.; Powell, R. L.; White, D. W. J. Am. Chem.
Soc. 1971, 93, 4004–4009. (c) Denney, D. B.; Denney, D. Z.; Gigantino,
J. J. J. Org. Chem. 1984, 49, 2831–2832. Thereby, the volatile products
were separated from OdP(OEt)3 through distillation. For a recent
cyclodehydration of diols, see: (d) Kelly, B. D.; Lambert, T. H. Org. Lett.
2011, 13, 740–743.
(18) Four examples following a related strategy have been reported:
(a) Reyes, E.; Ruiz, N.; Vicario, J. L.; Badia, D.; Carrillo, L. Synthesis
2011, 443–450. (b) Fraser, D. S.; Park, S. B.; Chong, J. M. Can. J. Chem.
2004, 82, 87–101. (c) Chung, S.-K.; Lee, J.-M. Tetrahedron: Asymmetry
1999, 1441–1444. (d) Dondoni, A.; Perrone, A. Synthesis 1993, 1162–
1176. For an early example of diastereodiscriminating reductions of
amino ketones, see: (e) Stevens, C. L.; TerBeek, K. J.; Pillai, P. M. J. Org.
Chem. 1974, 26, 3943–3946. For the FelkinꢀAnh model, see: (f) Cherest,
M.; Felkin, H.; Prudent, N. Tetrahedron Lett. 1968, 9, 2199–2204.
(19) Quitt, P.; Hellerbach, J.; Vogler, K. Helv. Chim. Acta 1963, 46,
327–333.
Org. Lett., Vol. 15, No. 20, 2013
5179

Straightforward amidation of carboxylic acids 2 with
DCC and conversion of the resulting Weinreb amides
(not shown) with Grignard reagent 3, which was found
to be superior in the final concentration to the reported
ClMgnPrOMgCl derivative,20 gave the ketones 4 in good
to excellent yields (70ꢀ88%, over 2 steps). Notably, with
this strategy we saved additional protection and deprotec-
tion steps of the free hydroxyl group of 3.
triethyl phosphate and chromatographic purification in
68ꢀ82% yield and high enantiomeric excess (90ꢀ99%).23,24
Interestingly, we observed the strong influence of the ratio
of NEt3/DCM on the yield: For instance, piperidinol 6c was
obtained in 68% yield when using NEt3/DCM in a 1:1.3
ratio, while a 1:3 solvent mixture gave 6c in only 22% yield
(noteworthily NEt3 is as cheap as common solvents such as
DCM and THF). This can be attributed to the lower
solubility of iodine in NEt3 (which leads to a slower and
thus more selective reaction) and general base catalysis:
simultaneous deprotonation (through NEt3) in the cycliza-
tion step strongly favors the desired reaction pathway to
piperidines 6.
Scheme 1. Synthesis of Hydroxy Ketones 4
Scheme 2. Synthesis of cis-Piperidinols 6
Diastereoselective reduction of ketones 4aꢀc with
L-Selectride (4aþb) and Superhydride (4c), respectively,
delivered the amino alcohols I (Scheme 2). Without isola-
tion of I the Cbz-moiety was cleaved through chemoselec-
tive hydrogenolysis (in preference over the Bn-group)
to result in diols 5aꢀc in useful to excellent diastereoselec-
tivities (4:1 to >19:1 syn/anti in accordance to a Felkinꢀ
Anh transition state). Subsequently, Appel cyclization21
(PPh3/I2) with careful temperature control (ꢀ40 °C to rt)
delivered the piperidinols 6aꢀc in good yields (69ꢀ77%).
Unfortunately, the byproduct triphenyl phosphine oxide
was only separable by chromatography requiring increased
amounts of silica gel.
In general, the reaction of alcohols with alkyl phosphites
(P(OR)3), activated through oxidants such as iodine, have
been reported to give the corresponding phosphates.22 In
order to improve atom economy and side product separa-
tion, werationalized that inthe phosphonium intermediate
II the intramolecular substitution by the amino function
(delivering the desired piperidinols 6) should be signifi-
cantly fasterthanthe bimolecular reaction of the iodide ion
with intermediate II as indicated (resulting in the forma-
tion of the corresponding undesired phosphates).
a R0 = sBu for 4aþb (L-Selectride); R0 = Et for 4c (Superhydride).
b dr g 19:1. c dr = 4:1. d dr g 19:1, ee g 99%. e dr = 5.3:1, ee = 90%.
Significantly, 14 g of alanine derived piperidinol cis-6a
was obtained (cyclodehydration with I2/P(OEt)3) in one
batch with no purification of the intermediates (2a, 4a and
syn-5a) and an overall yield of 44%, demonstrating the
scalability and high practicability of our sequence. The
relative configuration of compounds 5 and 6 was proven
by NOE-spectroscopy of piperidinols cis-6aꢀc, trans-6a,
cis- and trans-10, L-733,060 HCl, and oxazolidinones
3
derived from the aminoalcohols syn-5a and 5c.25
We wondered if the novel I2/P(OEt)3 cyclodehydration
is also applicable to the synthesis of other heterocycles:
Indeed, not only pyrrolidine 8a (Table 1, entry 1) but also
furans 8b to 8d (bearing inter alia sterically demanding
mesityl or two Ph substituents) were obtained in good to
excellent yields and purities >90% without any additional
purification (Table 1, entries 2ꢀ6).
Indeed, under optimized conditions the piperidinols 6
were isolated after saponification (during work up) of
(20) For the preparation of ClMgnPrMgCl, see: Cahiez, G.; Alexakis,
A.; Normant, J. F. Tetrahedron Lett. 1978, 33, 3013–3014.
€
(21) (a) Appel, R.; Kleinstuck, R. Chem. Ber. 1974, 107, 5–12. (b)
Having developed a short 5-step route leading to cis-
piperidinols of type 6, we turned our attention to the trans
diastereomers of 6 (Scheme 3). Initial reduction studies
with the amino ketone (not shown) obtained through Cbz-
cleavage of 4a and basic workup revealed L-Selectride as
the best reducing agent again in agreement with a Cram
chelate transition state (dr g25:1 according to 1H NMR).
Disappointingly, the hydroxy piperidine trans-6a isolated
after subsequent Appel cyclization showed a significantly
Lange, G. L.; Gottardo, C. Synth. Commun. 1990, 20, 1473–1479.
(22) (a) Stowell, J. K.; Widlanski, T. S. Tetrahedron Lett. 1995, 36,
1825–1826. (b) Oza, V. B.; Corcoran, R. C. J. Org. Chem. 1995, 60, 3680–
3684. (c) Watanabe, Y.; Inada, E.; Jinno, M.; Ozaki, S. Tetrahedron Lett.
1993, 34, 497–500.
(23) The optical purity was determined with HPLC on a chiral
stationary phase and comparison with racemic samples (alanine and
phenylglycine derived substrates) or in analogy to the aforementioned
amino acid derivatives (phenylalanine deduced substrates). The slight
decrease in the optical purity of the phenyl glycine derived piperidinol 6c
occurred in the amidation step (95% ee) and the subsequent addition of
Grignard reagent 3 (92% ee, transformation 2cf4c).
(24) While for the substrates 6a and 6b cyclodehydration with
P(OEt)3 improved the yield about 5% (compated to PPh3), cyclization
with PPh3 provided the piperidinol 6c in a higher yield (77%).
(25) See Supporting Information for further details.
5180
Org. Lett., Vol. 15, No. 20, 2013

diminished enantiomeric excess (32%).25 This result was
somewhat surprising, because reductions of structurally
related secondary amino ketones18c,d were previously re-
ported to proceed without a decrease in enantiopurity.
However, the observed racemization might be explained
through an intermolecular enamine formation.25
With 8 steps our sequence is one of the shortest syntheses
reported to date.13aꢀc Furthermore, with the carbamate
cis-10 (synthesized in 6 rather than 8 steps) we also
achieved a formal total synthesis of CP-99,994.26
Scheme 3. Synthesis of trans-Piperidinol 6a and of L-733,060
Table 1. Preliminary Substrate Scope of the I2/P(OEt)3-Cyclo-
dehydration
entry
substrate
Y
R1
R2
yielda
1
2
3
4
5
6
7a
NBn
O
H
H
H
H
H
H
Ph
81%b
92%
88%
83%
84%
82%
rac-7b
rac-7c
rac-7d
rac-7e
7f
Ph
O
οClPh
pBrPh
Mes
Ph
O
O
In conclusion, we have developed a stereodivergent and
highly diastereoselective (dr up to 19:1 for both isomers),
scalable, and practical (up to 14 g of 6a without any
purification of intermediates) synthesis of cis- and trans-
3-piperidinols 6, which represent a key structural motif in
various natural products and other bioactive target com-
pounds. Thereby, high step economy (5ꢀ6 steps) was
achieved by establishing several novel one-pot procedures
(1f2, 4f5, 9aftrans-6a) and avoiding any protection of
the OH-functions. From piperidinol 6c the NK-1 inhibitor
L-733,060 was prepared in three further steps. Addition-
ally, a unique cyclodehydration procedure replacing PPh3
through P(OEt)3 to improve atom economy (166 com-
pared to 262 g/mol) and to allow separation of the oxi-
dized side product (OP(OEt)3) through saponification
(no similar literature precedents known) was established.
Currently, we are exploiting our piperidinol 6 synthesis to
other pharmacologically relevant targets and are investi-
gating the substrate scope of the I2/P(OEt)3-cyclization
procedure.
O
a Isolated yields, purity >90% according to crude 1H NMR. b OP-
(OEt)3 was removed by washing with EtOAc.
To gain access to the piperidine trans-6a in high ee,
hydroxy ketone 4a was subjected to mesylation and Cbz-
cleavage to give the hydrochloride 9a, which cannot
racemize through an enamine equilibrium due the proto-
nation of the amino function. To our delight, subsequent
liberation of the free amine through DBU at low tempera-
ture, immediate L-Selectride reduction (giving intermediate
III), HCl quenching, and NEt3-induced cyclization af-
forded the piperidine trans-6a in an excellent ee (>99%)23
and as a single diastereomer according to crude 1H NMR.
Although the reduction is performed in the presence
of a free hydroxy function and 1 equiv of DBU-Hþ, only
1.5 equiv of L-Selectride were necessary for a quantitative
conversion. Thus we assume a Cram chelate transition state
is formed through an amine NꢀH proton rather than an
amide NꢀLi lithium cation (which would result from
deprotonation of the amino group by L-Selectride and would
therefore require at least 2 equiv of the reducing agent).
In order to probe the practicability of our sequence, we
synthesized L-733,060 as its hydrochloride salt as depicted
in Scheme 3. After cleavage of the Bn-group and Boc-
protection in one pot, the diastereomers cis- and trans-10
were easily separated by flash chromatography. The re-
sulting alcohol cis-10 was converted to the desired target
through etherification and cleavage of the Boc-group.
Acknowledgment. We want to thank the German Aca-
demic Exchange Service (DAAD) for a generous scholar-
ship for P.H., and Prof. H.-G. Schmalz for the very kind
opportunity to perform a part of the project at the Uni-
versity of Cologne. We are grateful to Dr. O. Karjalainen
(Aalto University) for helpful discussions.
Supporting Information Available. Experimental pro-
cedures and analytical data. This material is available free
of charge via the Internet at http://pubs.acs.org.
(26) Huang, P.-Q.; Liu, L.-X.; Wei, B.-G.; Ruan, Y.-P. Org. Lett.
2003, 5, 1927–1929.
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 20, 2013
5181