Gold-catalyzed oxidative cross-coupling of terminal alkynes: Selective synthesis of unsymmetrical 1,3-diynes

Article abstract of DOI:10.1021/ja5078365

Gold-catalyzed oxidative cross-coupling of alkynes to unsymmetrical diynes has been achieved for the first time. A N,N-ligand (1,10-Phen) and PhI(OAc)₂ were identified as crucial factors to promote this transformation, giving the desired cross-

Full text of DOI:10.1021/ja5078365

Communication
pubs.acs.org/JACS
Gold-Catalyzed Oxidative Cross-Coupling of Terminal Alkynes:
Selective Synthesis of Unsymmetrical 1,3-Diynes
Haihui Peng, Yumeng Xi, Nima Ronaghi, Boliang Dong, Novruz G. Akhmedov, and Xiaodong Shi*
C. Eugene Bennett Department of Chemistry, West Virginia University, Morgantown, West Virginia 26506, United States
S
* Supporting Information
Scheme 1. Gold-Catalyzed Oxidative Alkyne Cross-Coupling
ABSTRACT: Gold-catalyzed oxidative cross-coupling of
alkynes to unsymmetrical diynes has been achieved for the
first time. A N,N-ligand (1,10-Phen) and PhI(OAc)2 were
identified as crucial factors to promote this transformation,
giving the desired cross-coupled conjugated diynes in
excellent heteroselectivity (>10:1), in good to excellent
yields, and with large substrate tolerability.
onjugate diynes and polyynes are valuable building blocks
1
Cin chemical, pharmaceutical, and material sciences.
Methods for the synthesis of conjugate diynes, especially
unsymmetrical diynes, are thus of great importance. Typical
ways of making conjugate diynes include Glaser−Hay coupling2
and Cadiot−Chodkiewicz coupling.3,4 However, Glaser−Hay
coupling (catalyzed by Cu or Ni) suffers from undesired
competing homocoupling and often requires a large excess of one
alkyne, and Cadiot−Chodkiewicz coupling involves alkyne
prefunctionalization, which significantly increases the overall
cost of the reaction and is impractical due to the low stability of 1-
bromoalkyne. Thus, simple heteroselective alkyne coupling
remains very challenging yet desirable. Here we report a highly
selective heterocoupling of terminal alkynes for the synthesis of
1,3-unsymmetric diynes5 enabled by gold catalysis (Scheme 1),6
which requires neither one coupling counterpart in large excess7
nor alkyne prefunctionalization.8
discrimination of alkynes through metal acetylide formation.
As a superior π-acid, Au salt could form the σ-acetylide more
facilely than Cu, even without the presence of a strong base.15 We
hypothesized that gold cations might provide this “discrim-
ination effect” toward different alkynes, which was supported by
our investigations on the gold acetylide formation between
PPh3AuOAc and different types of alkynes (Figure 1).
In general, undesired alkyne homocoupling is the main
problem associated with any direct C−H alkynylation.9 To
develop a direct alkyne cross-coupling, homocouplings of both
alkynes must be suppressed to a minimum level. This has been a
big challenge for Cu and Pd systems, possibly due to the
disproportionation of [LM(CCR)2]n+ intermediates.10 Thus,
to achieve selective alkyne cross-coupling, rapid reductive
elimination (to avoid [LM(CCR)2]n+ transmetalation/dispro-
portionation) and selective formation of a specific metal acetylide
are crucial. It is well known that Au(I)/Au(III) redox potential is
high,11 and recent studies suggested that strong oxidants such as
PIDA and Selectfluor can facilely oxidize Au(I) to Au(III).12
Recently, Toste et al. demonstrated that highly oxidative Au(III)
complexes might undergo rapid reductive elimination.13 This
process could be further accelerated with proper choice of
ancillary ligand (e.g., π-acceptor and favored cis-reductive
elimination).8,14 On the basis of these analyses, we postulated
that, with fast reductive elimination, Au catalysis might open new
avenues to achieve the challenging alkyne cross-coupling.
Another important feature for successful cross-coupling is the
selective coordination of electronically biased alkynes, i.e.,
As revealed by reaction NMR studies (1H and 19F), under both
stoichiometric (1 equiv of Au) and substoichiometric conditions
(15% Au), Ph3PAuOAc16 reacted selectively with aliphatic
alkyne 2 over aromatic alkyne 1 to form the corresponding gold
acetylides (3:1 selectivity; see SI). Although the selectivity was
modest, this result implied the intrinsically distinct response of
aliphatic and aromatic alkynes toward Au catalysts, which might
be used to “discriminate” alkynes for heterocoupling.
Figure 1. Substrate-dependent gold acetylide formation.
Received: July 31, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja5078365 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
which ruled out the possibility that trace amounts of Cu or Pd in
commercial Au salts catalyzed the reaction. Improved reactivity
was observed with Au(I) precatalyst (entry 6). Furthermore,
compared with chloride salt, the gold bromide gave better results
for both reaction rate and selectivity (entries 6 and 7).20
Application of a digold catalyst21 further improved the reaction
performance, giving the desired cross-coupling product in good
yield (72%, entry 11). Pyridine as the ligand could not promote
this reaction (entry 12), while bidentate ligand bipyridine (bpy)
gave diyne 3a in 43% yield (entry 13). This result suggested the
importance of bidentate N,N-ligands for this transformation.
Through screening of various ligands, Phen was identified as the
optimal choice (entry 14). Finally, application of a slight excess of
Phen (10%) and MeCN/dioxane (3:1) cosolvents gave the
cross-coupling product in excellent selectivity (12:1, entry 15).
Notably, increasing the 1:2 ratio to 1:3 gave the cross-coupling
diyne 3a as the dominant product with >95% isolated yield.
Surprisingly, based on our screening, PhI(OAc)2 was the only
effective oxidant to promote this transformation (see SI). Other
oxidants, including the Selectfluor, which was previously used in
alkyne homocoupling,5 yielded only a trace amount of diyne.
As illustrated in Table 2, this new system indicated excellent
substrate compatibility, giving the desired cross-coupling
products in good to excellent isolated yields. First, for aromatic
alkynes, both electron-rich and electron-deficient alkynes
furnished the conjugated diynes in good yields (3a−3j). The
halogen-substituted aromatic alkynes (3c, 3e, 3g) gave the
desired products in excellent yields, highlighting the orthogonal
reactivity of Au catalyst over Pd, Cu, and Ni catalysts, which
undergo oxidative addition. Heteroaromatic alkynes, including
pyridine (3m, 3n), thiophene (3l), and indole (3k), also
proceeded efficiently. Great functional group tolerability was also
observed with aliphatic alkynes. Substrates containing unpro-
tected alcohol (3o), carboxylic acid (3p), amide (3q), silyl (3v),
ester (3r and 3t), and 1,2,3-triazole (3u and 3w) were all suitable
for this reaction. In particular, some highly reactive functional
groups, such as vinyl ether (3s), propargyl ester (3r), and
conjugated alkyne ester (3t), could also tolerate the reaction
conditions, given the desired diynes in good yields. Moreover, to
further evaluate the synthetic practicability of this method, we
attempted to modify alkyne-derived natural product-like
molecules, such as estrone (3za and 3zb), sugar (3y), and
amino acid (3x) derivatives. Fortunately, satisfactory yields were
obtained with all these substrates, which not only highlighted the
excellent substrate compatibility but also implied the great
potential of this new method for complex molecule synthesis/
modification.
Figure 2. Kinetics comparison of Cu and Au systems. Conditions: (left)
1 (0.1 mmol), 2 (0.13 mmol), CuCl (15 mol%), TMEDA (15 mol%),
acetone (0.3 M), 50 °C; (right) 1 (0.1 mmol), 2 (0.13 mmol), AuCl3 (5
mol%), Phen (5 mol%), CH3CN (0.3 M), 50 °C.
To further test our hypothesis, a series of kinetics experiments
were conducted to compare the Cu and Au catalytic systems. As
shown in Figure 2, with 1a and 1b as the substrates (1:1.3
ratio),17 the standard CuCl/TMEDA conditions18 gave very
poor selectivity, close to the statistical distribution.
Interestingly, when AuCl3 (5 mol%) was used as the catalyst
and PhI(OAc)2 as the oxidant, no alkyne coupling products were
observed at either room temperature or 50 °C. Addition of Phen
ligand (5 mol%)19 dramatically improved the reactivity, giving
the desired conjugated diynes in good yield. Good hetero-/
homoselectivity (3a vs 4a) was obtained, and the selectivity was
further improved at higher temperature (50 °C, 3a:4a =
60%:13%). At both room temperature and 50 °C, only a small
amount of aliphatic alkyne homocoupling product 4a was
observed. With these exciting results, we initiated a conditions
screening with the focus on both Au precatalysts and ligands to
further improve the reaction performance. The results were
summarized in Table 1 (see detailed conditions screening in SI).
Under the standard reaction conditions (50 °C in CH3CN
with PhI(OAc)2 as the oxidant), various catalysts were tested.
While Cu catalyst gave very poor cross-coupling selectivity
(1.6:1, entry 4), Pd did not catalyze the reaction at all (entry 5),
a
Table 1. Optimization of Conditions
The resulting conjugated diynes are very useful synthons that
can be easily converted into other useful compounds. Scheme 2A
summarizes some known transformations of diynes to important
heteroaromatic molecules.22 A gram-scale synthesis was also
performed to verify the practical synthesis using this method, as
shown in Scheme 2B. In addition, considering the importance of
polyynes in chemical and material research, we explored the
feasibility of using this method to achieve polyyne synthesis. As
shown in Scheme 2C, using the terminal alkyne synthesized from
TIPS deprotection of 3v gave the conjugated triyne 5 in excellent
yield under standard conditions (85% in two steps). It is
anticipated that a variety of polyynes could be readily prepared
using this strategy, highlighting the broad synthetic application of
the reported method.
a
Reaction conditions: 1 (0.1 mmol), 2 (0.13 equiv), catalyst (5 mol%),
b
and ligand (5 mol%) in CH3CN (0.3 mL), 50 °C, 2 h. Determined by
c
19F NMR using benzotrifluoride as internal standard. Solvent:
CH3CN/1,4-dioxane (3:1). 5% ArC(O)CH2OAc was detected from
alkyne 1 oxidized by PhI(OAc)2 along with some unidentified
decomposition products.
Several examples involving Au-catalyzed C−C bond coupling
under oxidative conditions have been reported in the literature.
B
dx.doi.org/10.1021/ja5078365 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
ab
,
Table 2. Generality
a
General reaction conditions: R1-alkyne (0.2 mmol, 1.3 equiv), 2.5 mol% dppm(AuBr)2, 10 mol% Phen, and PhI(OAc)2 (0.4 mmol, 2 equiv) in
b
c
CH3CN/1,4-dioxane (0.6 mL/0.2 mL), 50 °C. Isolated yield. 1 (0.26 mmol, 1.3 equiv); the yield was based on the R1-alkyne.
However, there has been much uncertainty regarding the exact
mechanism, namely, whether Au(I)/Au(III) redox catalysis or
sole Au(I) catalysis through π-activation is operative.6a,e To
verify if this reaction indeed involves a Au(III) intermediate, we
conducted a set of single-turnover experiments using stoichio-
metric AuCl3 as the sole oxidant and gold source.23
Scheme 2. Synthetic Utility
As shown in Scheme 3A, when only stoichiometric AuCl3 was
reacted with alkynes 1 and 2 (without any additive and/or
ligand), no diynes were detected. Interestingly, addition of Phen
ligand gave only trace amounts of products (<5%). Further
addition of acetate ion (NaOAc) significantly improved the
reaction performance, with a 65% yield of 3a, despite low
heteroselectivity (4:1). These results suggest that (1) the
reaction can undergo Au(III) reductive elimination; (2) the
Phen ligand is vital to the reaction and presumably has a large
influence on the rate of reductive elimination and (3) the acetate
ion from PhI(OAc)2 may play an important role as base to
sequester free protons, consistent with the fact that PhI(TFA)2
does not promote this transformation.24 Based on these results,
two different pathways are proposed in Scheme 3B.
Scheme 3. Mechanistic Implication
With Au(I) as the precatalyst (the optimal condition), the key
mechanistic concern is the timing of gold monoacetylide
formation.25 As shown in Scheme 3B, oxidation might occur
either at the stage of Au(I) salt (prior to the formation of gold
acetylide, path A) or at the stage of gold(I) acetylide (path B).
The reaction with a stoichiometric amount of Au(III) shown in
Scheme 3A gave lower selectivity (4:1) compared with the
reaction using Au(I) precatalyst (12:1). This result provides
strong evidence against pathway A. In addition, the selective
formation of aliphatic gold(I) acetylide and fast reaction rates
shown in Figure 1 provide strong support for the late oxidation.
Thus, based on current information, we favor a mechanism the
same as or similar to pathway B. Obviously, the ligand (Phen)
C
dx.doi.org/10.1021/ja5078365 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
Am. Chem. Soc. 2010, 132, 2522−2523. (c) Gao, Y.; Wang, G.; Chen, L.;
Xu, P.; Zhao, Y.; Zhou, Y.; Han, L.-B. J. Am. Chem. Soc. 2009, 131, 7956.
Silver salt was also recently used: (d) He, C.; Guo, S.; Ke, J.; Hao, J.; Xu,
H.; Chen, H.; Lei, A. J. Am. Chem. Soc. 2012, 134, 5766.
(10) Weng, Y.; Cheng, B.; He, C.; Lei, A. Angew. Chem., Int. Ed. 2012,
51, 9547.
plays a very important role in tuning the rate of reductive
elimination of three distinct dialkynylgold(III) species.26 In
addition, improved performance from bis-gold catalysts offers
some possibilities for alternative mechanisms. Overall, the exact
mechanism and origin of selectivity await further study and will
be reported in due course.
(11) AuI/AuIII, E0 = +1.40 V. See: CRC Handbook of Chemistry and
Physics, 84th ed.; Lide, D. R., Ed.; CRC Press: Boca Raton, FL, 2004.
(12) For some leading reviews, see: (a) Hopkinson, M. N.; Gee, A. D.;
Gouverneur, V. Chem.Eur. J. 2011, 17, 8248. (b) Wegner, H. A.;
Auzias, M. Angew. Chem., Int. Ed. 2011, 50, 8236. (c) Boorman, T. C.;
Larrosa, I. Chem. Soc. Rev. 2011, 40, 1910 and references cited therein.
For elegant use of Au redox catalysis in cross-coupling reactions, see:
(d) Ball, L. T.; Lloyd-Jones, G. C.; Russell, C. A. Science 2012, 337, 1644.
(e) Ball, L. T.; Lloyd-Jones, G. C.; Russell, C. A. J. Am. Chem. Soc. 2014,
136, 254. (f) Zhang, G.; Peng, Y.; Cui, L.; Zhang, L. Angew. Chem., Int.
Ed. 2009, 48, 3112. (g) Peng, Y.; Cui, L.; Zhang, G.; Zhang, L. J. Am.
Chem. Soc. 2009, 131, 5062. (h) Zhang, G.; Cui, L.; Wang, Y.; Zhang, L.
J. Am. Chem. Soc. 2010, 132, 1474. (i) Zhang, G.; Luo, Y.; Wang, Y.;
Zhang, L. Angew. Chem., Int. Ed. 2011, 50, 4450.
In summary, we report the first example of Au-catalyzed alkyne
oxidative cross-coupling for the synthesis of unsymmetrical
conjugated diynes. This method is straightforward, efficient, and
highly selective. Compared with the literature methods, this new
approach does not require a large excess of one coupling partner
or prefunctionalization of alkyne. Broad substrate scope and
excellent functional group tolerance are demonstrated. Prelimi-
nary investigations reveal that both the Phen ligand and the
distinct nature of alkynes are crucial to the high selectivity. The
selectivity picture with any two random alkynes can be
complicated. In fact, brief screening of two different aromatic
alkynes showed no selectivity (forming 1:2:1 coupling mixtures).
However, the fact that highly selective coupling is achieved
through the effective alkyne alkyne discrimination suggests the
strong potential of using Au(I/III) redox catalysis to develop
other selective C(sp)−C bond-forming reactions.
(13) (a) Wolf, W. J.; Winston, M. S.; Toste, F. D. Nat. Chem. 2014, 6,
159. (b) Winston, M. S.; Wolf, W. J.; Toste, F. D. J. Am. Chem. Soc. 2014,
136, 7777.
(14) Hartwig, J. F. Inorg. Chem. 2007, 46, 1936.
(15) Brown, T. J.; Widenhoefer, R. A. Organometallics 2011, 30, 6003.
(16) Weber, D.; Jones, T. D.; Adduci, L. L.; Gagne,
Int. Ed. 2012, 51, 2452.
́
M. R. Angew. Chem.,
ASSOCIATED CONTENT
* Supporting Information
Experimental details and characterization data. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
S
(17) A 1:1.3 mixture of 1a:1b was used for practical reasons. Kinetic
studies with a 1:1 mixture were less accurate due to the experimental
errors when trying to match the exact amount of all samples.
(18) CuCl/TMEDA gave slower conversion.
(19) For previous use of Au(III) and N,N-ligand, see: (a) Xie, J.; Li, H.;
Zhou, J.; Cheng, Y.; Zhu, C. Angew. Chem., Int. Ed. 2012, 51, 1252.
(b) O’Neill, J. A. T.; Rosair, G. M.; Lee, A.-L. Catal. Sci. Technol. 2012, 2,
1818. (c) Li, J.; Hu, J.; Li, G. Catal. Commun. 2011, 12, 1401. (d) Casini,
AUTHOR INFORMATION
Corresponding Author
xiaodong.shi@mail.wvu.edu
Notes
■
A.; Diawara, M. C.; Scopelliti, R.; Zakeeruddin, S. M.; Gratzel, M.;
̈
The authors declare no competing financial interest.
Dyson, P. J. Dalton Trans. 2010, 39, 2239. (e) Cocco, F.; Cinellu, M. A.;
Minghetti, G.; Zucca, A.; Stoccoro, S.; Maiore, L.; Manassero, M.
Organometallics 2010, 29, 1064. (f) Ref 6d.
(20) Reactions with different concentrations of n-Bu4NCl were
performed to evaluate the anion influence on reactivity and selectivity.
The results were shown in Figures S8 and S9. Reduced reactivity and
selectivity were observed in the presence of Cl−, consistent with the
anion effect reported in Russell’s work in ref 12e. Lower selectivity was
also observed with LAuOAc as the catalyst (3:1), suggesting gold
bromide salts as optimal precatalysts.
(21) (a) Tkatchouk, E.; Mankad, N. P.; Benitez, D.; Goddard, W. A.;
Toste, F. D. J. Am. Chem. Soc. 2011, 133, 14293. (b) Levin, M. D.; Toste,
F. D. Angew. Chem., Int. Ed. 2014, 53, 6211.
(22) (a) Duan, H.; Sengupta, S.; Petersen, J. L.; Akhmedov, N. G.; Shi,
X. J. Am. Chem. Soc. 2009, 131, 12100. (b) Chalk, A. J. Tetrahedron Lett.
1972, 13, 3487. (c) Jiang, H.; Zeng, W.; Li, Y.; Wu, W.; Huang, L.; Fu,
W. J. Org. Chem. 2012, 77, 5179. (d) Wang, L.; Yu, X.; Feng, X.; Bao, M.
Org. Lett. 2012, 14, 2418. (e) Wang, L.; Yu, X.; Feng, X.; Bao, M. J. Org.
Chem. 2013, 78, 1693.
ACKNOWLEDGMENTS
We thank the NSF (Career CHE-1362057 and CHE-1228336)
and NSFC (21228204) for financial support.
■
REFERENCES
■
(1) Shi, W.; Lei, A. Tetrahedron Lett. 2014, 55, 2763.
(2) Siemsen, P.; Livingston, R. C.; Diederich, F. Angew. Chem., Int. Ed.
2000, 39, 2632.
(3) Chodkiewicz, W.; Cadiot, P. Compt. Rend. 1955, 241, 1055.
(4) Other methods also can form 1,3-diynes. FBW rearrangement:
(a) Jahnke, E.; Tykwinski, R. R. Chem. Commun. 2010, 3235.
Decarbonylation: (b) Dermenci, A.; Whittaker, R. E.; Dong, G. Org.
Lett. 2013, 15, 2242. Metathesis: (c) Li, S. T.; Schnabel, T.; Lysenko, S.;
Brandhorst, K.; Tamm, M. Chem. Commun. 2013, 7189. Elimination:
(d) Negishi, E.; Okukado, N.; Lovich, S. F.; Luo, F. T. J. Org. Chem.
1984, 49, 2629.
(5) Leyva-Per
Catal. 2012, 2, 121.
́ ́
ez, A.; Domenech, A.; Al-Resayes, S. I.; Corma, A. ACS
(23) Russell et al. demonstrated the rapid oxidation of PPh3 ligand in
Au(I)/Au(III) reactions using 31P NMR in ref 12e. This was also
confirmed in our case, where rapid dppm decomposition occurred under
the reaction conditions, revealed by 31P NMR shown in the SI.
(24) We observed a dramatic color change from yellow to orange-red
when NaOAc was added to a solution of AuCl3, presumably leading to
(6) Gold-catalyzed oxidative cross-couplings of alkynes with other
nucleophiles have been reported: (a) de Haro, T.; Nevado, C. J. Am.
Chem. Soc. 2010, 132, 1512. (b) Hopkinson, M. N.; Ross, J. E.; Giuffredi,
G. T.; Gee, A. D.; Gouverneur, V. Org. Lett. 2010, 12, 4904. (c) Qian, D.;
Zhang, J. Beilstein J. Org. Chem. 2011, 7, 808. (d) Ma, Y.; Zhang, S.; Yang,
S.; Song, F.; You, J. Angew. Chem., Int. Ed. 2014, 53, 7870. For a recent
review, see: (e) Brand, J. P.; Li, Y.; Waser, J. Isr. J. Chem. 2013, 53, 901.
(7) (a) Yin, W.; He, C.; Chen, M.; Zhang, H.; Lei, A. Org. Lett. 2009,
11, 709. (b) Balaraman, K.; Kesavan, V. Synthesis 2010, 3461.
(8) Shi, W.; Luo, Y.; Luo, X.; Chao, L.; Zhang, H.; Wang, J.; Lei, A. J.
Am. Chem. Soc. 2008, 130, 14713.
the formation of species like AuCln(OAc)3−n or NaAuCln(OAc)4−n
.
(25) It is also possible that the formation of gold(III) bisacetylide is
enabled by transmetalation of gold(I) acetylide and gold(III)
monoacetylide; see ref 5.
(26) Shekhar, S.; Hartwig, J. F. J. Am. Chem. Soc. 2004, 126, 13016.
(9) Copper system: (a) Hamada, T.; Ye, X.; Stahl, S. S. J. Am. Chem.
Soc. 2008, 130, 833. (b) Wei, Y.; Zhao, H.; Kan, J.; Su, W.; Hong, M. J.
D
dx.doi.org/10.1021/ja5078365 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX