Dimethyl acetylenedicarboxylate and phospholes: A variety of reaction pathways

Article abstract of DOI:10.1002/ejoc.201000723

Correcting an earlier report, the reaction of 1-phenyl-3,4- dimethylphosphole (1a) with an excess amount of dimethyl acetylenedicarboxylate (DMAD) at room temperature affords the stable cyclopentadienylidenephosphorane 3a resulting from unexpected deoxygenation of initially formed [phosphole + 2DMAD] adduct 2a by the starting phosphole, The identity of 3a was established by X-ray crystal structure analysis. DFT calculations on a model [1+2] adduct similar to 2a shows a bent allenic structure and an ambident reactivity for the terminal carbon of the (DMAD)₂ chain. Cyclopentadienylidenephosphorane 3b is also obtained with 1-benzyl phosphole lb. With 1-stannylphosphole 1c, a 1:3 adduct with a seven-membered ring (i.e., 5) is obtained. It was also characterized by X-ray crystal structure analysis, Finally, with 1-benzyl-2-benzoylphosphole 1d, the P lone pair has lost its nucleophilicity and a [4+2] cycloaddition takes place with the dienol tautomer of the unsaturated ketone to give phosphindole 6, also characterized by X-ray.

Full text of DOI:10.1002/ejoc.201000723

FULL PAPER
DOI: 10.1002/ejoc.201000723
Dimethyl Acetylenedicarboxylate and Phospholes: A Variety of Reaction
Pathways
Congbao Huang,[a] Hui Liu,[a] Jianzhen Zhang,[a] Zheng Duan,*[a] and François Mathey*[a,b]
Keywords: Phosphorus heterocycles / Cycloaddition / Ylides / Structure elucidation
Correcting an earlier report, the reaction of 1-phenyl-3,4-di-
methylphosphole (1a) with an excess amount of dimethyl
acetylenedicarboxylate (DMAD) at room temperature affords
the stable cyclopentadienylidenephosphorane 3a resulting
from unexpected deoxygenation of initially formed [phos-
phole + 2DMAD] adduct 2a by the starting phosphole. The
identity of 3a was established by X-ray crystal structure
analysis. DFT calculations on a model [1+2] adduct similar to
2a shows a bent allenic structure and an ambident reactivity
for the terminal carbon of the (DMAD)2 chain. Cyclopen-
tadienylidenephosphorane 3b is also obtained with 1-benzyl-
phosphole 1b. With 1-stannylphosphole 1c, a 1:3 adduct with
a seven-membered ring (i.e., 5) is obtained. It was also char-
acterized by X-ray crystal structure analysis. Finally, with 1-
benzyl-2-benzoylphosphole 1d, the P lone pair has lost its
nucleophilicity and a [4+2] cycloaddition takes place with the
dienol tautomer of the unsaturated ketone to give phosph-
indole 6, also characterized by X-ray.
spectroscopic data and could not be considered as defini-
tively established. In view of our long-standing interest in
phosphole chemistry, we decided some time ago, to rein-
vestigate the reactions of some phospholes with DMAD,
trying whenever possible to establish the structures of the
products by X-ray crystal structure analysis. We have pub-
lished a first report[4] on the reaction of 1-phenyl-3,4-di-
methylphosphole (1a) with DMAD with results at variance
with those of Hughes. The 1:2 adduct thus produced
seemed to have the structure of ylide 2a on the basis of 13C
NMR spectroscopic data (Scheme 1). Unfortunately, we
Introduction
Since the pioneering work of Tebby on the reactions of tri-
phenylphosphane with dimethyl acetylenedicarboxylate
(DMAD),[1] the reactivity of tervalent phosphorus deriva-
tives with DMAD has been consistently investigated and
has yielded a lot of unpredictable results.[2] The phosphole
case has been already studied in some depth by Hughes.[3]
It appeared that these reactions yielded a variety of ylides,
some having ring-expanded or ring-fused structures. Unfor-
tunately, the assignments were made on the sole basis of
Scheme 1. The reaction of 1-phenyl- and 1-benzyl-3,4-dimethylphosphole with DMAD.
were unable to grow crystals of 2a, and this formula re-
[a] Chemistry Department, International Phosphorus Laboratory,
Key Lab of Chemical Biology and Organic Chemistry of Henan
Province, Zhengzhou University,
mained tentative because the analogous adduct with tri-
phenylphosphane is known to be highly unstable.[1] We de-
scribe herein additional results that supplement and correct
our earlier data while underlying the broad variety of struc-
tures that can be obtained through this apparently simple
chemistry. All of our structures have been definitively estab-
lished by X-ray crystal structure analysis.
Zhengzhou 450052, P. R. China
[b] Division of Chemistry & Biological Chemistry, School of
Physical & Mathematical Sciences, Nanyang Technological
University,
21 Nanyang Link, Singapore 637371
E-mail: duanzheng@zzu.edu.cn
fmathey@ntu.edu.sg
View this journal online at
wileyonlinelibrary.com
5498
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 5498–5502

Dimethyl Acetylenedicarboxylate and Phospholes
tween the two units has essentially single-bond character
(C15–C16 1.485 Å). The terminal carbon (C17) displays a
structure similar to that of a disymmetrical bent allene:[9]
C16–C17 1.312 Å, C17–C30 1.421 Å, C16–C17–C30
138.84°, C15–C16···C30–O32 88.3°. The planes of the two
π systems are indeed orthogonal. The HOMO (Figure 3) is
strongly localized at C17 and its shape resembles an in-
plane lone pair. Surprisingly, the Mulliken charge is
strongly negative at C16 (–1.56) and strongly positive at
C17 (+0.83). Thus, we expect that C17 will display both
nucleophilic and electrophilic reactivities. The resulting pic-
ture is more complex than what would be expected from the
classical mesomeric representation of these [1+2] adducts.
Results and Discussion
After numerous unsuccessful attempts, we were finally
able to grow crystals of the reaction product of 1a with
DMAD. The quality of the crystals was rather poor, so the
structural parameters are only approximate, but the col-
lected data establish beyond any doubt that the actual
structure of the product is that of 3a, as shown in Figure 1.
The (DMAD)2 substituent has cyclized to give a cyclopen-
tadienylidenephosphorane. The peak at δ = 159.06 ppm in
the 13C NMR spectrum that was formerly assigned to a
carbonyl carbon atom in fact corresponds to C9 [d, 2JC,P
=
7.5 Hz, =C(OMe)-]. The structural parameters of the cyclo-
pentadiene ring are closely similar to those of other cyclo-
pentadienylidenephosphoranes.[5] The unexpected reaction
that takes place is the deoxygenation of the methoxy-
carbonyl substituent on the α carbon of the (DMAD)2
chain of initially formed [1:2] adduct 2a (Scheme 1).The de-
oxygenation of 2a is probably effected by the starting phos-
phole itself. The mechanism likely involves a bisylide ([2+2]
adduct)[6] whose cyclization by intramolecular Wittig reac-
tion is possible due to the flexibility of the poorly conju-
gated chain (see later). Indeed, when mixing the phosphole
and the excess amount of DMAD in dichloromethane, two
peaks appear in the 31P NMR spectrum at δ = 19.6 and
40.7 ppm. The first one corresponds to 3a, whereas the sec-
ond one corresponds to the monomeric phosphole oxide[7]
that slowly disappears by [4+2] dimerization (δ = 52.5 and
74.1 ppm, JP,P = 39.6 Hz).
Figure 2. Computed structure of 4: Main bond lengths [Å] and
angles [°]: P–C14 1.802, C14–C15 1.390, C15–C16 1.485, C16–C17
1.312, C17–C30 1.421, C30–O31 1.224, C30–O32 1.370, P–C1
1.816, P–C4 1.812, P–C9 1.821, C1–C2 1.341, C2–C3 1.476, C3–
C4 1.342; C1–P–C4 92.43, C9–P–C14 113.55, C13–C14–C15
124.24, C14–C15–C16 123.68, C15–C16–C17 105.89, C16–C17–
C30 138.84, C17–C30–O31 124.57, C17–C30–O32 113.90; P–C14–
C15–C16 151.43, C14–C15–C16–C17 88.50, C15–C16–C17–C30
11.68, C16–C17–C30–O31 74.28.
Figure 1. X-ray crystal structure of cyclopentadienylidenephos-
phorane 3a. Main bond lengths [Å] and angles [°]: P1–C5 1.732(6),
P1–C8 1.780(6), P1–C20 1.800(7), P1–C21 1.773(7), C5–C1
1.452(8), C1–C6 1.394(8), C6–C12 1.436(8), C12–C9 1.394(8), C9–
C5 1.406(8); C8–P1–C21 92.7(3), C5–P1–C20 112.6(3), C1–C5–C9
105.8(5), P1–C5–C1 126.7(4), P1–C5–C9 127.4(4).
To better understand the chemistry of initially formed
[1+2] adduct 2a, we decided to perform a DFT study of
We then decided to investigate the influence of the phos-
model compound 4 at the B3LYP/6-311+G(d,p) level.[8] phole P-substituent on the course of its reaction with
The computed structure is shown in Figure 2. The DMAD. With 1-benzylphosphole 1b, we obtained a prod-
(DMAD)2 substituent is highly distorted, and the two uct whose 13C NMR spectrum is closely similar to that of
DMAD units are orthogonal (C14–C15–C16–C17 88.50°). 3a, and hence its formula 3b. Replacing 1a by stannylphos-
As a result, the delocalization along the chain is weakened phole 1c yielded an entirely different result. A 1:3 adduct
and a strong bond alternation is observed. The bond be- with a seven-membered ring 5 was produced (Scheme 2).
Eur. J. Org. Chem. 2010, 5498–5502
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5499

C. Huanga, H. Liua, J. Zhanga, Z. Duana, F. Mathey
FULL PAPER
5, which shows two α-CH signals at δ = 5.70 and 6.48 ppm.
The most logical mechanism for the formation of 5 is de-
picted in Scheme 3.
Figure 3. HOMO of adduct 4 as computed by DFT.
Scheme 3. Proposed mechanism for the formation of seven-mem-
bered ring 5.
The first step involves insertion of DMAD into the reac-
tive P–Sn bond, followed by the formation of a 1:2 adduct
between the P lone pair and DMAD, as usual. Then, the
nucleophilic carbon carrying tin attacks the electrophilic
carbon at the end of the (DMAD)2 chain, thus forming
Scheme 2. Formation of a seven-membered ring from 1-stannyl-
phosphole 1c.
The structure of 5 was established by X-ray analysis (Fig-
ure 4). The representation fits the data of the structure with
a long C16–C12 (1.443 Å), a short C7–C12 (1.386 Å), a
long C4–C7 (1.446 Å), and a short C1–C4 bond (1.352 Å).
All the carbon atoms of the seven-membered ring are
planar, except C13. The two sides of the phosphole ring
are sharply inequivalent due to the chirality of C13. This
1
inequivalence is quite visible in the H NMR spectrum of
Scheme 4. Formation of a phosphindole from a 2-acylphosphole.
Figure 4. X-ray crystal structure of seven-membered ring 5. Main
bond lengths [Å] and angles [°]: P–C20 1.809(12), P–C21 1.796(10),
C19–C20 1.292(17), C19–C22 1.505(17), C21–C22 1.383(15), P– Figure 5. X-ray crystal structure of phosphindole 6. Main bond
C16 1.743(10), C12–C16 1.443(15), C12–C7 1.386(16), C4–C7 lengths [Å] and angles [°]: P–O5 1.481(3), P–C1 1.823(3), P–C19
1.446(13), C1–C4 1.352(13), C1–C13 1.482(14), P–C13 1.812(10); 1.773(4), P–C20 1.798(4), C1–C2 1.399(5), C2–C17 1.495(4), C17–
C20–P–C21 95.0(5), C13–P–C16 108.4(5).
C19 1.338(5); C1–P–C19 92.14(17).
5500
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 5498–5502

Dimethyl Acetylenedicarboxylate and Phospholes
cess amount of lithium wire until P–Ph bond cleavage was com-
plete. After the remaining amount of lithium was removed, the
solution was treated with aluminum trichloride (95 mg, 0.7 mmol)
at 0 °C for 30 min. Then, tri-n-butyltin chloride (575 µL,
2.12 mmol) was added at –78 °C. After warming to room tempera-
ture, DMAD (1040 µL, 8.50 mmol) was added. The crude organic
product was purified by chromatography (silica gel; dichlorometh-
ane/ethyl acetate, 1:1). A yellow solid was collected and, once con-
centrated, 54 mg of a pure product was obtained (7% yield). 31P
NMR (CDCl3): δ = 38.6 ppm. 1H NMR (CDCl3): δ = 2.12 (s, 3 H,
the seven-membered ring. It is interesting to note that this
mechanism relies on the electrophilic nature of the terminal
carbon of the (DMAD)2 chain as predicted by our theoreti-
cal calculation on 4.
We then investigated the reaction of DMAD with
1-benzyl-2-benzoyl-3,4-dimethylphosphole (1d).[10] The
lone pair of 1d has almost completely lost its nucleophilic-
ity. As a result, DMAD does not react at phosphorus any
more. The product of the reaction is phosphindole 6
(Scheme 4). The structure of 6 was established by X-ray Me), 2.14 (s, 3 H, Me), 3.55, 3.64, 3.72, 3.78, 3.79, 3.81 (6 s, 6ϫ3
2
2
H, OMe), 4.23 (d, JH,P = 15.3 Hz, sp3 CH-P), 5.70 (d, JH,P
=
crystal structure analysis (Figure 5). The product results
from the reaction of DMAD with the dienol tautomer of
the 2-acylphosphole.
29.7 Hz, =CH-P), 6.48 (d, JH,P = 30.0 Hz, =CH-P) ppm. 13C
2
3
3
NMR (CDCl3): δ = 17.43 (d, JC,P = 18.5 Hz, Me), 17.96 (d, JC,P
= 19.0 Hz, Me), 42.40 (d, 1JC,P = 57.4 Hz, sp3 CH-P), 51.40, 51.81,
1
52.41, 52.66, 53.03, 53.7058.23 (d, JC,P = 119.5 Hz, ylide C=P),
1
1
117.23 (d, JC,P = 96.1 Hz, =CH-P), 121.41 (d, JC,P = 76.6 Hz,
2
2
Conclusions
=CH-P), 154.98 [d, JC,P = 22.8 Hz, =C(Me)], 158.92 [d, JC,P
20.9 Hz, =C(Me)], 165.86–169.061 (C=O) ppm.
=
As can be seen, the outcome of the reaction of DMAD
with phospholes heavily depends on the nature of the phos-
phole substituents. Some of the products of this very diverse
chemistry might have some synthetic interest.
Synthesis of 6: 1-Phenyl-3,4-dimethylphosphole (0.4 mL, 2.12
mmol) in dry THF (10 mL) was allowed to react with an excess
amount of lithium wire until P–Ph bond cleavage was complete.
After the remaining amount of lithium was removed, the solution
was treated with tert-butyl chloride (230 µL, 2.12 mmol) and
heated to 60 °C for 1.5 h. Then, benzoyl chloride (248 µL,
2.12 mmol) was added at –78 °C. After warming to 0 °C, tBuOK
was added (240 mg, 2.12 mmol). The resulting mixture was heated
at 60 °C for 2 h. Then, benzyl bromide (230 µL, 2.12 mmol) was
added at –78 °C. After warming to room temperature, DMAD
(246 µL, 2.00 mmol) was added. After the reaction was complete,
the crude organic product was purified by chromatography (silica
gel; dichloromethane/ethyl acetate, 5:2). A colorless solid was col-
lected and, once concentrated, 125 mg of pure product was ob-
tained (13% yield). 31P NMR (CDCl3): δ = 46.3 ppm. 1H NMR
(CDCl3): δ = 2.18 (br., 3 H, Me), 2.60–3.01 (m, 2 H, PhCH2), 3.59
Experimental Section
General Methods: All reactions were performed under nitrogen by
using standard Schlenk techniques. Nuclear magnetic resonance
spectra were obtained with a Bruker Avance 300 spectrometer
operating at 300.13 MHz for 1H, 75.45 MHz for 13C, and
121.496 MHz for 31P. Chemical shifts are expressed in ppm down-
field from internal TMS (1H and 13C) and external 85% H3PO4
(31P). All coupling constants (J) are reported in Hz. Mass spectra
were recorded with an LC-MSD-Trap-XCT instrument by electro-
spray ionization (ESI).
2
(s, 3 H, OMe), 3.93 (s, 3 H, OMe), 6.04 (d, JH,P = 25.2 Hz, 1 H,
PCH=), 6.84–7.79 (m, 11 H, PhH) ppm. Selected 13C NMR
Synthesis of 3b: 1-Phenyl-3,4-dimethylphosphole (0.4 mL,
2.12 mmol) in dry THF (10 mL) was allowed to react with an ex-
cess amount of lithium wire until P–Ph bond cleavage was com-
plete. After the remaining amount of lithium was removed, the
solution was treated with tert-butyl chloride (230 µL, 2.12 mmol)
and heated to 60 °C for 1.5 h. Then, benzyl bromide (230 µL,
2.12 mmol) was added at –78 °C. After warming to room tempera-
ture, DMAD (730 µL, 6.00 mmol) was added. The crude organic
product was purified by chromatography (silica gel; petroleum
ether/ethyl acetate, 1:4). A green liquid was collected and, once
concentrated, 125 mg of pure product was obtained (13% yield).
3
1
(CDCl3): δ = 16.61 (d, JC,P = 15.8 Hz, Me), 35.51 (d, JC,P
64.1 Hz, PhCH2), 52.45, 52.97 (2s, OMe), 122.03 (d, JC,P
111.7 Hz, PCH=), 165.45 (C=O), 168.21 (C=O) ppm.
=
=
1
Acknowledgments
The authors thank Professor John Tebby for a stimulating exchange
of ideas, the National Natural Science Foundation of China (No.
20702050), the Scientific Research Foundation for Returned Over-
seas Chinese Scholars, Zhengzhou University, and Nanyang Tech-
nological University for financial support.
1
31P NMR (CDCl3): δ = 24.0 ppm. H NMR (CDCl3): δ = 2.04 (s,
6 H, Me), 3.33 (s, 3 H, OMe), 3.69 (d, 2JH,P = 16.2 Hz, 2 H, CH2P),
3.73 (s, 3 H, OMe), 3.83 (s, 3 H, OMe), 3.91 (s, 3 H, OMe), 6.43
(d, 2JH,P = 27.9 Hz, 2 H, =CHP), 7.03–7.27 (m, 5 H, Ph) ppm. 13
C
[1] A. W. Johnson, J. C. Tebby, J. Chem. Soc. 1961, 2126; N. E.
Waite, J. C. Tebby, R. S. Ward, D. H. Williams, J. Chem. Soc.
C 1969, 1100; N. E. Waite, J. C. Tebby, R. S. Ward, M. A. Shaw,
D. H. Williams, J. Chem. Soc. C 1971, 1620.
[2] See, for example: K. Bieger, J. Tejeda, R. Réau, F. Dahan, G.
Bertrand, J. Am. Chem. Soc. 1994, 116, 8087; S. Kumaraswamy,
P. Kommana, N. Satish Kumar, K. C. Kumara Swamy, Chem.
Commun. 2002, 40; R. A. Aitken, S. J. Costello, A. M. Z. Sla-
win, N. J. Wilson, Eur. J. Org. Chem. 2003, 623; N. N. Bhu-
wan Kumar, M. Chakravarty, K. C. Kumara Swamy, New J.
Chem. 2006, 30, 1614.
3
1
NMR (CDCl3): δ = 17.74 (d, JC,P = 10.5 Hz, Me), 32.78 (d, JC,P
= 51.7 Hz, CH2P), 50.77 (s, OMe), 51.57 (s, OMe), 52.20 (s, OMe),
1
63.03 (s, OMe), 77.95 (d, JC,P = 93.8 Hz, C=P),107.05 (d, JC,P
=
=
1
14.2 Hz, C=C), 108.42 (d, JC,P = 9.0 Hz, C=C), 116.25 (d, JC,P
87.7 Hz, =CHP), 126.77 (d, 2JC,P = 12.0 Hz, Ph C ipso), 127.70 (d,
JC,P = 3.7 Hz, Ph CH para), 128.54 (d, JC,P = 3.0 Hz, Ph CH),
129.85 (d, JC,P = 6.0 Hz, Ph CH), 130.23 (d, JC,P = 8.2 Hz, C=C),
2
2
157.67 [d, JC,P = 18.0 Hz, =C(Me)],158.43 [d, JC,P = 6.7 Hz,
=C(OMe)], 163.77 (d, JC,P = 3.0 Hz, CO), 166.31 (s, CO), 169.90
(d, JC,P = 3.0 Hz, CO) ppm.
[3] A. N. Hughes, S. Uaboonkul, Tetrahedron 1968, 24, 3437;
A. N. Hughes, K. Amornraksa, S. Phisithkul, V. Reutrakul, J.
Heterocycl. Chem. 1976, 13, 65; D. G. Holah, A. N. Hughes, D.
Kleemola, J. Heterocycl. Chem. 1978, 15, 1319.
Synthesis of 5: 1-Phenyl-3,4-dimethylphosphole (0.4 mL,
2.12 mmol) in dry THF (10 mL) was allowed to react with an ex-
Eur. J. Org. Chem. 2010, 5498–5502
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5501

C. Huanga, H. Liua, J. Zhanga, Z. Duana, F. Mathey
FULL PAPER
[4] Z. Duan, J. Zhang, R. Tian, L. Bai, F. Mathey, Organometallics
2008, 27, 5169.
Hratchian, J. B. Cross, C. Adamo, J. Jaramillo, R. Gomperts,
R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pom-
elli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P.
Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich,
A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D.
Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui,
A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu,
A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J.
Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara,
M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W.
Wong, C. Gonzalez, J. A. Pople, Gaussian 03, Revision B.05,
Gaussian, Inc., Pittsburgh, PA, 2003.
[5] See, for example: L. J. Higham, J. Muldoon, P. G. Kelly, D. M.
Corr, H. Müller-Bunz, D. G. Gilheany, J. Org. Chem. 2007, 72,
8780; L. J. Higham, P. G. Kelly, H. Müller-Bunz, D. G.
Gilheany, Acta Crystallogr., Sect. C 2004, 60, o308; N. H.
Tran Huy, C. Compain, L. Ricard, F. Mathey, J. Organomet.
Chem. 2002, 650, 57.
[6] M. A. Shaw, J. C. Tebby, J. Ronayne, D. H. Williams, J. Chem.
Soc. C 1967, 944.
[7] L. D. Quin, X.-P. Wu, Heteroat. Chem. 1991, 2, 359.
[8] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, J. A. Montgomery Jr., T.
Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar,
J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N.
Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K.
Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y.
Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P.
[9]
C. A. Dyker, V. Lavallo, B. Donnadieu, G. Bertrand, Angew.
Chem. Int. Ed. 2008, 47, 3206.
[10] P. Toullec, F. Mathey, Synlett 2001, 1977.
Received: March 20, 2010
Published Online: August 16, 2010
5502
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 5498–5502