Copper-catalyzed reduction of alkyl triflates and iodides: An efficient method for the deoxygenation of primary and secondary alcohols

Article abstract of DOI:10.1002/anie.201307697

We describe an effective method for catalytic reduction of 1°alkyl sulfonates, and 1°and 2°iodides in the presence of a wide range of functional groups. This Cu-catalyzed reaction provides a means for the effective deoxygenation of alcohols, as demonstrated by the highly selective reduction of 1°alcohols using a triflation/reduction sequence. A preliminary study of the reaction mechanism suggests that the reduction does not involve free-radical intermediates. Primarily reduced: The copper-catalyzed reduction of 1°alkyl sulfonates, and 1°and 2°iodides, which is effective in the presence of a wide range of functional groups, provides a means for the effective deoxygenation of alcohols. A preliminary study of the reaction mechanism suggests that the reduction does not involve free-radical intermediates. Copyright

Full text of DOI:10.1002/anie.201307697

.
Angewandte
Communications
DOI: 10.1002/anie.201307697
Alcohol Deoxygenation
Copper-Catalyzed Reduction of Alkyl Triflates and Iodides: An
Efficient Method for the Deoxygenation of Primary and Secondary
Alcohols**
Hester Dang, Nick Cox, and Gojko Lalic*
Abstract: We describe an effective method for catalytic
reduction of 18 alkyl sulfonates, and 18 and 28 iodides in the
presence of a wide range of functional groups. This Cu-
catalyzed reaction provides a means for the effective deoxyge-
nation of alcohols, as demonstrated by the highly selective
reduction of 18 alcohols using a triflation/reduction sequence.
A preliminary study of the reaction mechanism suggests that
the reduction does not involve free-radical intermediates.
À
T
he catalytic transformation of C O bonds has garnered
À
much attention owing to the difficulties associated with C O
bond activation,[1] and a growing interest in the ability to
convert over-oxygenated biomass into chemical feedstocks.[2]
Furthermore, the selective reduction of alcohols to alkanes
remains a difficult transformation that is important in organic
synthesis.[3] These challenges provide motivation for the
development of practical techniques for the deoxygenation
of alcohols.
Scheme 1. Cu-catalyzed approach to the deoxygenation of alcohols.
In contrast to the reduction of secondary or tertiary
alcohols, there are few practical approaches to primary
alcohol deoxygenation (Scheme 1).[11] The most commonly
used technique involves the reduction of halides or sulfonates
using borohydride reagents.[12] Even the mildest variant of this
approach requires prolonged heating with multiple equiva-
lents of NaBH4 in DMSO,[13] making this strategy an
impractical option for the reduction of molecules bearing
sensitive functional groups.[14] Overall, the deoxygenation of
complex primary alcohols remains a particularly challenging
task.
We were surprised to find that relatively few attempts
have been made to address these challenges through the use
of transition-metal catalysis, especially given its role in the
selective reduction of alkenes, alkynes, and carbonyls. Thus
far, investigations into the reduction of halides and sulfonates
using Pd,[15] Ni,[16] and Ir[17] have mostly been focused on
unfunctionalized substrates. Recently, we reported the use of
copper hydride in a chemoselective catalytic semireduction of
alkynes.[18] Considering that copper hydride complexes have
been shown to reduce unfunctionalized alkyl halides and
sulfonates in moderate yield when used in stoichiometric
quantities,[19] we reasoned that the catalytic system developed
in our laboratory could provide a new approach to the
deoxygenation of alcohols. In this paper, we describe
a copper-catalyzed reduction of alkyl triflates and iodides
that is practical, efficient, and functional-group tolerant.
As sulfonate esters can be readily accessed from primary
alcohols, and their reactivity can be easily tuned,[20] our initial
efforts were aimed at the reduction of alkyl sulfonates. We
found that primary triflates can be reduced using IPrCuOtBu
as the catalyst, tetramethyldisiloxane 2 (TMDSO) as the
hydride source, and CsF as an additive that aids catalyst
Current strategies for alcohol deoxygenation can be
categorized into single-step procedures that reduce the
alcohol directly, and two-step procedures that convert the
alcohol into a reactive group prior to the reduction step.
Although there are several strategies for single-step alcohol
deoxygenation,[4] the only widely used technique is ionic
hydrogenation.[5] This procedure is best suited for the
reduction of alcohols that can generate a stabilized carbenium
ion, such as benzylic, allylic, or tertiary alcohols. For the
reduction of unactivated primary and secondary alcohols,
two-step strategies are predominantly used. The most
common strategy involves the formation of halides or
thiocarbonyl esters[6] followed by radical reduction with tin
hydride reagents.[7] Although other reducing reagents have
been successfully used,[8] in practice, alkylstannanes remain
the reagent of choice despite their toxicity and problems
associated with purification.[9] Furthermore, even with exten-
sive development of the Barton-McCombie reaction, the
reduction of primary alcohols using this procedure is still
difficult to achieve.[10]
[*] H. Dang,[+] N. Cox,[+] G. Lalic
Department of Chemistry, University of Washington
Seattle, WA 98195 (USA)
E-mail: lalic@chem.washington.edu
[+] These authors contributed equally to this work.
[**] NSF (CAREER Award 1254636) and University of Washington (RRF
award A70681) are acknowledged for financial support.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201307697.
752
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 752 –756

Angewandte
Chemie
Scheme 2. Conditions for the Cu-catalyzed reduction of 18 triflates.
Table 1: Development of the Cu-catalyzed reduction of 18 triflates.
Entry Change from optimized
conditions
Yield [%][a] Yield [%][a]
after 1 h
after 48 h
1
2
3
4
5
None
96
100
10
0
0
59
ROTs instead of ROTf
RONs instead of ROTf
No Catalyst
Cy3PCuI instead of
IPrCuOtBu
Scheme 3. Two-step deoxygenation of 18 alcohols by the Cu-catalyzed
6
7
8
XantphosCuI instead of
IPrCuOtBu
IMesCuOtBu instead of
IPrCuOtBu
1.3 equiv Et3SiH instead of
TMDSO
52
54
71
reduction of 18 alkyl triflates. Yields of isolated products are shown;
overall yields for the two-step triflation/reduction sequence are given
in parentheses. [a] 1 mol% of catalyst was used. [b] Reaction was
performed on a 20 mmol scale and yielded 3.6 g of the product.
9
1.3 equiv PMHS instead of TMDSO 10
98
100
16[b]
13[b]
methods, primary triflates can be efficiently prepared with
minimal purification. As shown in Scheme 3, the deoxygena-
tion of alcohols occurs in high yield over two steps and is
compatible with many functional groups. High yields were
obtained in the presence of substituents that are susceptible to
reduction, including cyano, carbomethoxy, nitro, and alkenyl
groups (10–12 and 21).[23] The reaction is compatible with aryl
chlorides, bromides, and iodides, all of which are typically
reactive in both transition-metal-catalyzed and free-radical
reductions (14–17). The reduction of compounds 18 and 19
shows that the selective reduction of triflates can be
accomplished even in the presence of alkyl bromides or
tosylates. No deprotection of silyl-protected alcohols was
observed, even though a stoichiometric amount of CsF is used
in the reaction (20a and 20b). Finally, the reduction of 20b
can successfully be performed on a 20 mmol sale. Overall, the
two-step procedure shown in Scheme 3 is a practical approach
to primary alcohol deoxygenation that addresses the key
limitations of existing methods.
We also found that the reaction conditions developed for
triflate reduction were effective in the reduction of primary
alkyl iodides. The results in Scheme 4 show that iodides can
be reduced by extending the reaction time to 24 h, while still
maintaining the functional-group compatibility observed in
the reduction of triflates. Nitriles, esters, carbamates, amides,
aryl halides, tosylates, silyl ethers, and alkenes are all well
tolerated.
10
11
12
Ph2SiH2 instead of TMDSO
NaOtBu instead of CsF
LiOtBu instead of CsF
89
[a] Yield was determined by GC analysis based on an internal standard.
[b] Complete conversion of the starting material was observed. Cy=cy-
clohexyl, Ns=nosyl, Ts=tosyl, Xantphos=4,5-bis(diphenylphos-
phanyl)-9,9-dimethylxanthene.
turnover (Scheme 2). The reduction occurs in high yield in
under 1 hour at ambient temperature, and without competing
elimination. Tosylates and nosylates were found to be
significantly less reactive, illustrating the mild nature of the
reducing agent (Table 1, entries 2–3). No reaction took place
in the absence of catalyst (entry 4). Furthermore, IPrCuOtBu
performed significantly better than catalysts supported by
phosphines or less sterically demanding NHC ligands
(entries 5–7). In general, we observed a higher rate of
reduction when more reactive silanes such as poly(methylhy-
drosiloxane) (PMHS) or Ph2SiH2 were used. Surprisingly, the
highest rate of reduction was achieved with TMDSO
(entries 8–10).[21] Although alkoxide additives are commonly
employed in copper-catalyzed reactions of silanes to aid in
catalyst turnover,[22] we found that their use resulted in
unproductive side-reactions of the triflate starting material
(entries 11,12).
With the copper-catalyzed reduction of alkyl triflates as
a starting point, we turned our attention to the development
of a two-step procedure for the deoxygenation of primary
alcohols. We found that, with slight modifications to existing
We next turned our attention to the reduction of
secondary iodides, which is a common transformation in
synthesis.[24] We found that under the standard reaction
Angew. Chem. Int. Ed. 2014, 53, 752 –756
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
753

.
Angewandte
Communications
Scheme 6. Proposed catalytic cycle of the Cu-catalyzed reduction of
triflates and iodides.
metallation,[18,26] we chose to focus our investigation of the
mechanism on the proposed copper hydride reduction step.
Whereas copper hydride reductions of carbonyls, alkenes, and
alkynes have been studied,[22] copper hydride reduction of
alkyl halides and sulfonates has not been explored. We
reasoned that the reduction step is likely to occur by one of
two pathways: 1) the formation of radical intermediates, or
2) an ionic mechanism, which would involve either nucleo-
philic substitution or oxidative addition followed by reductive
elimination.
Scheme 4. Cu-catalyzed reduction of 18 alkyl iodides. Yields of isolated
products are shown. [a] 1.4 equiv Ph2SiH2 and 1.1 equiv 37 were used
instead of TMDSO and CsF. Ac=acetyl, Boc=tert-butoxycarbonyl.
conditions, a competing elimination reaction produced trace
amounts of an undesired alkene. To address this problem, we
explored the use of alternative hydride sources and additives.
Ultimately, we found that switching from TMDSO to
Ph2SiH2, and from CsF to potassium 2-tert-butylphenoxide
(37) allowed us to achieve high yield in the reduction of
secondary iodides without alkene formation.
As shown in Scheme 5, these changes to the reaction
conditions allowed the successful reduction of secondary
iodides bearing a variety of functional groups, including ester,
silyl ether, cyano, methoxy, carbamate, and aryl halide
groups.[25]
The reduction of triflates by copper hydride is unlikely to
proceed by a radical mechanism, as it would involve
[27]
À
homolytic cleavage of the strong C O bond.
However,
À
considering the facile homolytic cleavage of the C I bond, it
is possible that the reduction of iodides involves the formation
of free-radical intermediates. To explore this possibility, we
conducted the experiments shown in Scheme 7a. We found
that, in the copper-catalyzed reduction of iodide 33, the ratio
of reduced product 47 to cyclization product 48 was 9:1. In the
presence of the radical trap 2,2,6,6-tetramethyl-1-piperidiny-
loxy free radical (TEMPO), formation of the cyclization
product 48 was completely suppressed, and reduction product
47 was formed in 81% yield. This suggests that a free-radical
reaction is a minor pathway.
We propose that the copper-catalyzed reduction of
triflates and iodides proceeds according to the mechanism
depicted in Scheme 6. Given the precedence for the forma-
tion of copper(I) hydrides through silicon-to-copper trans-
Scheme 5. Cu-catalyzed reduction of 28 alkyl iodides. Unless otherwise
specified, yields of isolated products are shown. [a] Yield was deter-
mined by GC analysis based on an internal standard. [b] PMHS was
used instead of Ph2SiH2, and NaOtBu instead of 37.
Scheme 7. Experiments to test for free-radical formation in the copper-
catalyzed reduction of alkyl halides.
754
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 752 –756

Angewandte
Chemie
Barrett, P. A. Prokopiou, D. H. R. Barton, J. Chem. Soc. Chem.
Commun. 1979, 1175.
In further support of the conclusion that a two-electron
pathway is the dominant mechanism, we found that when
bromide 49 was submitted to the reaction conditions,
reduction product 47 was formed in quantitative yield, and
cyclization product 48 was not observed (Scheme 7b). Finally,
when we carried out the reduction of iodide 50 in the presence
of a stoichiometric amount of TEMPO, the reduction product
51 was observed in quantitative yield and the TEMPO-alkyl
adduct 52 was not detected (Scheme 7c). While it is
inherently difficult to rule out the possibility of free-radical
formation, we believe that, taken collectively, the results
presented in Scheme 7 suggest that a radical mechanism is not
the major pathway, and that an ionic mechanism is more
likely.
In conclusion, we have developed a copper-catalyzed
deoxygenation of primary and secondary alcohols that is
convenient, versatile, and compatible with many functional
groups. We have demonstrated that this transformation can
be achieved through a practical two-step procedure in which
the alcohol is converted either into a triflate or an iodide prior
to a highly efficient copper-catalyzed reduction step. Initial
investigation of the mechanism suggests that free-radical
formation is a minor pathway, and that the reduction step is
more likely to occur by a two-electron process. Overall, this
technique is particularly well-suited to the deoxygenation of
primary alcohols, which has traditionally been a challenging
transformation.
[7] a) L. W. Menapace, H. G. Kuivila, J. Am. Chem. Soc. 1964, 86,
3047 – 3051; b) H. G. Kuivila, Acc. Chem. Res. 1968, 1, 299 – 305;
c) N. J. Lawrence, M. D. Drew, S. M. Bushell, J. Chem. Soc.
Perkin Trans. 1 1999, 3381 – 3391.
[8] a) A. R. Pinder, Synthesis 1980, 425 – 452; b) F. Alonso, I. P.
Beletskaya, M. Yus, Chem. Rev. 2002, 102, 4009 – 4091; c) J. K.
Kochi, P. E. Mocadlo, J. Am. Chem. Soc. 1966, 88, 4094 – 4096;
d) J. R. Hanson, Synthesis 1974, 1 – 8; e) D. P. Curran, T. L.
Fevig, C. P. Jasperse, M. J. Totleben, Synlett 1992, 943 – 961; f) R.
Hekmatshoar, S. Sajadi, M. M. Heravi, J. Chin. Chem. Soc. 2008,
55, 616 – 618.
[9] B. S. Edelson, B. M. Stoltz, E. J. Corey, Tetrahedron Lett. 1999,
40, 6729 – 6730.
[10] See Ref. [6a] for details. The use of this procedure in the
deoxygenation of primary alcohols is complicated by the high
energy of the primary radical intermediate.
[11] In a recent publication, an iridium catalyst was used to reduce
primary alkyl iodides through a visible-light-mediated process;
see: J. D. Nguyen, E. M. DꢀAmato, J. M. R. Narayanam, C. R. J.
Stephenson, Nat. Chem. 2012, 4, 854 – 859.
[12] a) R. O. Hutchins, D. Kandasamy, C. A. Maryanoff, D. Masila-
mani, B. E. Maryanoff, J. Org. Chem. 1977, 42, 82 – 91; b) F.
Rolla, J. Org. Chem. 1981, 46, 3909 – 3911; c) E. Santaniello, A.
Fiecchi, A. Manzocchi, P. Ferraboschi, J. Org. Chem. 1983, 48,
3074 – 3077; d) S. Krishnamurthy, H. C. Brown, J. Org. Chem.
1983, 48, 3085 – 3091.
[13] R. O. Hutchins, D. Kandasamy, F. Dux III, C. A. Maryanoff, D.
Rotstein, B. Goldsmith, W. Burgoyne, F. Cistone, J. Dalessandro,
J. Puglis, J. Org. Chem. 1978, 43, 2259 – 2267.
[14] This procedure is known to reduce esters; for an example, see:
M. Ono, K. Suzuki, S. Tanikawa, H. Akita, Tetrahedron:
Asymmetry 2001, 12, 2597 – 2604.
Received: September 2, 2013
Published online: December 4, 2013
[15] a) R. Boukherroub, C. Chatgilialoglu, G. Manuel, Organome-
tallics 1996, 15, 1508 – 1510; b) M. S. Viciu, G. A. Grasa, S. P.
Nolan, Organometallics 2001, 20, 3607 – 3612; c) R. E. Maleczka,
R. J. Rahaim, R. R. Teixeira, Tetrahedron Lett. 2002, 43, 7087 –
7090; d) R. J. Rahaim, Jr., R. E. Maleczka, Jr., Tetrahedron Lett.
2002, 43, 8823 – 8826; e) A. Arcadi, G. Cerichelli, M. Chiarini, R.
Vico, D. Zorzan, Eur. J. Org. Chem. 2004, 3404 – 3407; f) S. K.
Pandey, A. E. Greene, J.-F. Poisson, J. Org. Chem. 2007, 72,
7769 – 7770; g) A. Kunai, T. Sakurai, E. Toyoda, M. Ishikawa, Y.
Yamamoto, Organometallics 1994, 13, 3233 – 3236; h) J. Ohshita,
R. Taketsugu, Y. Nakahara, A. Kunai, J. Organomet. Chem.
2004, 689, 3258 – 3264; i) P. Srithanakit, W. Chavasiri, Tetrahe-
dron Lett. 2011, 52, 2505 – 2507.
[16] a) M. Weidauer, E. Irran, C. I. Someya, M. Haberberger, S.
Enthaler, J. Organomet. Chem. 2013, 698-698, 53 – 59; b) N. M.
Yoon, H. J. Lee, J. H. Ahn, J. Choi, J. Org. Chem. 1994, 59, 4687 –
4688.
[17] a) J. Yang, M. Brookhart, J. Am. Chem. Soc. 2007, 129, 12656 –
12657; b) J. Yang, M. Brookhart, Adv. Synth. Catal. 2009, 351,
175 – 187.
Keywords: alcohols · copper · deoxygenation · reduction ·
silanes
.
[1] a) S. Kundu, J. Choi, D. Y. Wang, Y. Choliy, T. J. Emge, K.
Krogh-Jespersen, A. S. Goldman, J. Am. Chem. Soc. 2013, 135,
5127 – 5143; b) D.-G. Yu, B.-J. Li, Z.-J. Shi, Acc. Chem. Res. 2010,
43, 1486 – 1495.
[2] a) M. Schlaf, Dalton Trans. 2006, 4645 – 4653; b) F. W. Lich-
tenthaler, S. Mondel, Pure Appl. Chem. 1997, 69, 1853 – 1866;
c) A. G. Sergeev, J. D. Webb, J. F. Hartwig, J. Am. Chem. Soc.
2012, 134, 20226 – 20229.
[3] J. M. a. K. B. Herrmann, Eur. J. Org. Chem. 2013, 31, 7017 –
7027.
[4] a) J. B. Hendrickson, M. Singer, M. S. Hussoin, J. Org. Chem.
1993, 58, 6913 – 6914; b) M. Yasuda, Y. Onishi, M. Ueba, T.
Miyai, A. Baba, J. Org. Chem. 2001, 66, 7741 – 7744; c) M.
Movassaghi, O. K. Ahmad, J. Org. Chem. 2007, 72, 1838 – 1841;
d) M. Mirza-Aghayan, R. Boukherroub, M. Rahimifard, Tetra-
hedron Lett. 2009, 50, 5930 – 5932; e) K. Kobiro, K. Sumoto, Y.
Okimoto, P. Wang, J. Supercrit. Fluids 2013, 77, 63 – 69.
[5] a) M. Schlaf, M. E. Thibault, D. Di Mondo, D. Taher, E. Karimi,
D. Ashok, Int. J. Chem. React. Eng. 2009, 7, A34; b) V.
Gevorgyan, J.-X. Liu, M. Rubin, S. Benson, Y. Yamamoto,
Tetrahedron Lett. 1999, 40, 8919 – 8922; c) V. Gevorgyan, M.
Rubin, S. Benson, J.-X. Liu, Y. Yamamoto, J. Org. Chem. 2000,
65, 6179 – 6186; d) M. Dobmeier, J. M. Herrmann, D. Lenoir, B.
Koenig, Beilstein J. Org. Chem. 2012, 8, 330 – 336.
[18] A. M. Whittaker, G. Lalic, Org. Lett. 2013, 15, 1112 – 1115.
[19] a) T. Yoshida, E.-i. Negishi, J. Chem. Soc. Chem. Commun. 1974,
762 – 763; b) E. C. Ashby, A. B. Goel, J. J. Lin, Tetrahedron Lett.
1977, 18, 3695 – 3698; c) E. C. Ashby, J.-J. Lin, A. B. Goel, J. Org.
Chem. 1978, 43, 183 – 188.
[20] a) S. D. Lepore, D. Mondal, Tetrahedron 2007, 63, 5103 – 5122;
b) P. J. Stang, M. Hanack, L. R. Subramanian, Synthesis 1982,
85 – 126.
[21] In general, the expected reactivity of the silanes used would be as
follows: Et3SiH < TMDSO < PMHS < Ph2SiH2; for details, see:
E. Keinan, Pure Appl. Chem. 1989, 61, 1737 – 1746.
[6] a) S. W. McCombie, W. B. Motherwell, M. J. Tozer, Org. React.
2012, 77, 161 – 591; b) D. H. R. Barton, S. W. McCombie, J.
Chem. Soc. Perkin Trans. 1 1975, 1574 – 1585; c) A. G. M.
Angew. Chem. Int. Ed. 2014, 53, 752 –756
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
755

.
Angewandte
Communications
[22] a) C. Deutsch, N. Krause, B. H. Lipshutz, Chem. Rev. 2008, 108,
2916 – 2927; b) S. Rendler, M. Oestreich, Angew. Chem. 2007,
119, 504 – 510; Angew. Chem. Int. Ed. 2007, 46, 498-504.
3220; e) T. B. Durham, N. Blanchard, B. M. Savall, N. A. Powell,
W. R. Roush, J. Am. Chem. Soc. 2004, 126, 9307 – 9317; f) T. B.
Durham, W. R. Roush, Org. Lett. 2003, 5, 1875 – 1878; g) F. Peng,
S. J. Danishefsky, J. Am. Chem. Soc. 2012, 134, 18860 – 18867;
h) D. Zurwerra, J. Gertsch, K.-H. Altmann, Org. Lett. 2010, 12,
2302 – 2305; i) E. A. Crane, T. P. Zabawa, R. L. Farmer, K. A.
Scheidt, Angew. Chem. 2011, 123, 9278 – 9281; Angew. Chem. Int.
Ed. 2011, 50, 9112 – 9115.
[23] Although the reaction was found to be tolerant of alkenes,
alkynes are known to undergo rapid hydrocupration in the
presence of copper hydride. Similarly, aldehydes are not
compatible with the reaction conditions, and ketones are
partially reduced. See the Supporting Information for details.
[24] a) Secondary iodide reduction has a number of applications in
natural product synthesis outside of alcohol deoxygenation; for
recent examples, see: b) P. Burch, M. Binaghi, M. Scherer, C.
Wentzel, D. Bossert, L. Eberhardt, M. Neuburger, P. Scheiffele,
K. Gademann, Chem. Eur. J. 2013, 19, 2589 – 2591; c) S. M.
Canham, D. J. France, L. E. Overman, J. Am. Chem. Soc. 2010,
132, 7876 – 7877; d) F. Fueloep, M. Palko, E. Forro, M. Dervarics,
T. A. Martinek, R. Sillanpaeae, Eur. J. Org. Chem. 2005, 3214 –
[25] The reaction conditions used in Scheme 5 can also be used in the
reduction of primary alkyl iodides (see Scheme 4, compound 26).
However, alkyl triflates are not compatible with the phenoxide
37.
[26] N. P. Mankad, D. S. Laitar, J. P. Sadighi, Organometallics 2004,
23, 3369 – 3371.
[27] M. B. Smith, J. March, Marchꢀs Advanced Organic Chemistry,
5th ed., Wiley, New York, NY, 2001.
756
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 752 –756