A rapid and efficient Sonogashira protocol and improved synthesis of free fatty acid 1 (FFA1) receptor agonists

Article abstract of DOI:10.1021/jo902533p

(Chemical Equation Presented) Aprotocol for rapid and efficient Pd/Cu-catalyzed coupling of aryl bromides and iodides to terminal alkynes has been developed with use of 2-(di-tert-butylphosphino)-N-phenylindole (cataCXium PIntB) as ligand inTMEDA and wate

Full text of DOI:10.1021/jo902533p

pubs.acs.org/joc
catalytic systems and environmentally friendly reaction
A Rapid and Efficient Sonogashira Protocol and
Improved Synthesis of Free Fatty Acid 1 (FFA1)
Receptor Agonists
conditions, including ligand-free, copper-free, amine-free,
and aqueous conditions.2,3 Nonetheless, unreactive sub-
strates and problems related to alkyne homocoupling are
still frequently encountered. We recently reported a series of
free fatty acid 1 (FFA1 or GPR40) receptor agonists with
potential antidiabetic properties.4 The original synthesis of
these compounds, which relied on the Sonogashira coupling
in two central steps, frequently resulted in low or no yields,
precluding upscaling and access to new analogues (Scheme 1).
Thus, we set out to optimize the general synthetic route to
these compounds.
Elisabeth Christiansen, Maria E. Due-Hansen, and
Trond Ulven*
Department of Physics and Chemistry, University of Southern
Denmark, Campusvej 55, DK-5230 Odense M, Denmark
ulven@ifk.sdu.dk
Received November 29, 2009
SCHEME 1. Original Synthetic Route to Alkyne FFA1 Agonistsa
aReagents and conditions: (a) I2, KIO3, H2SO4, H2O, AcOH, reflux, 5 h (33%);
(b) MeOH, HCl (cat.), rt, 2 h (98%); (c) Pd(PPh3)2Cl2, CuI, Et3N, TMSA
(added at 70 °C), DMF, 70f90 °C, 3 h; (d) K2CO3, MeOH, rt, 2.5 h (90%
over2steps);(e) arylhalide, Pd(PPh3)2Cl2,CuI, Et3N, DMF, 50 °C; (f) LiOH,
1,4-dioxane, H2O, rt.
A protocol for rapid and efficient Pd/Cu-catalyzed coupl-
ing of aryl bromides and iodides to terminal alkynes has
been developed with use of 2-(di-tert-butylphosphino)-
N-phenylindole (cataCXium PIntB) as ligand in TMEDA
and water. The new protocol successfully couples substrates
which failed with standard Sonogashira conditions, and
enables an efficient general synthetic route to free fatty
acid 1 (FFA1) receptor ligands from 3-(4-bromophenyl)-
propionic acid.
The low-yielding iodination step resulting in synthesis of
the common alkyne intermediate 3 in only 29% yield repre-
sented a general problem in the original synthetic route
(Scheme 1). We therefore decided to substitute the iodo
intermediate by the readily available 3-(4-bromophenyl)-
propionic acid (1). Unfortunately, coupling of 1 with tri-
methylsilylacetylene (TMSA) by the standard Sonogashira
protocol resulted in only 48% conversion (Table 1, entry 1).
Furthermore, both TMS-alkyne 2 and the deprotected alkyne
intermediate 3 turned out to be very difficult to separate from 1,
thus, complete conversion was required. The outcome was only
marginally influenced by the exchange of DMF and Et3N by
TMEDA (entry 2). Other catalytic systems like Pd(OAc)2 with
Xantphos (entry 3) or tri-tert-butylphosphine (entry 4) resulted
in further reduced conversion.5 Recently, Beller and co-workers
Since its discovery more than three decades ago, the
Sonogashira reaction has attained the position as the most
important method for synthesis of substituted alkynes.1,2
Originally referring to the cross-coupling of terminal alkynes
with vinyl or aryl halides cocatalyzed by palladium and
copper, the name has lately been used broadly for any
metal-catalyzed C-C bond-forming cross-coupling with
terminal alkynes. The original reaction conditions are gen-
erally efficient and tolerant to a wide variety of functional
groups and are still widely used, but a large number of
modified protocols have also appeared which have been
aimed at solving various limitations, such as broadening
the scope of possible sp2-partners to include unactivated
aryl bromides and chlorides, and developing more efficient
(3) Recent improved Sonogashira protocols: (a) Schulz, T.; Torborg, C.;
Enthaler, S.; Schaeffner, B.; Dumrath, A.; Spannenberg, A.; Neumann, H.;
Boerner, A.; Beller, M. Chem.;Eur. J. 2009, 15, 4528–4533. (b) Torborg, C.;
€
€
Huang, J.; Schulz, T.; Schaffner, B.; Zapf, A.; Spannenberg, A.; Borner, A.;
Beller, M. Chem.;Eur. J. 2009, 15, 1329–1336. (c) Lipshutz, B. H.; Chung,
D. W.; Rich, B. Org. Lett. 2008, 10, 3793–3796. (d) Mori, S.; Yanase, T.;
Aoyagi, S.; Monguchi, Y.; Maegawa, T.; Sajiki, H. Chem.;Eur. J. 2008, 14,
6994–6999. (e) Huang, H.; Liu, H.; Jiang, H.; Chen, K. J. Org. Chem. 2008,
73, 6037–6040. (f) Finke, A. D.; Elleby, E. C.; Boyd, M. J.; Weissman, H.;
Moore, J. S. J. Org. Chem. 2009, 74, 8897–8900. (g) Bolligera, J. L.; Frech, C.
M. Adv. Synth. Catal. 2009, 351, 891–902.
(1) Sonogashira, K.; Tohda, Y.; Hagihara, N. Tetrahedron Lett. 1975, 16,
4467–4470.
(2) For recent reviews of the Sonogashira reaction, see: (a) Doucet, H.;
Hierso, J.-C. Angew. Chem., Int. Ed. 2007, 46, 834–871. (b) Chinchilla, R.;
Najera, C. Chem. Rev. 2007, 107, 874–922. (c) Negishi, E.; Anastasia, L.
Chem. Rev. 2003, 103, 1979–2017. (d) Sonogashira, K. J. Organomet. Chem.
2002, 653, 46. (e) Plenio, H. Angew. Chem., Int. Ed. 2008, 47, 6954–6956. (f)
Nicolaou, K. C.; Bulger, P. G.; Sarlah, D. Angew. Chem., Int. Ed. 2005, 44,
4442–4489. (g) Tykwinski, R. R. Angew. Chem., Int. Ed. 2003, 42, 1566–1568.
(4) Christiansen, E.; Urban, C.; Merten, N.; Liebscher, K.; Karlsen,
K. K.; Hamacher, A.; Spinrath, A.; Bond, A. D.; Drewke, C.; Ullrich, S.;
Kassack, M. U.; Kostenis, E.; Ulven, T. J. Med. Chem. 2008, 51, 7061–7064.
(5) (a) Hundertmark, T.; Littke, A. F.; Buchwald, S. L.; Fu, G. C. Org.
€
Lett. 2000, 2, 1729–1731. (b) Kollhofer, A.; Plenio, H. Adv. Synth. Catal.
2005, 347, 1295–1300. (c) Bohm, V. P. W.; Herrmann, W. A. Eur. J. Org.
€
Chem. 2000, 3679–3681.
DOI: 10.1021/jo902533p
r
Published on Web 01/25/2010
J. Org. Chem. 2010, 75, 1301–1304 1301
2010 American Chemical Society

Christiansen et al.
JOCNote
TABLE 1. Optimizing the Coupling of 1 with Trimethylsilylacetylene
TABLE 2. Coupling of 3 with Aryl Bromides
solvent/base
(equiv)
TMSA
(equiv)
temp
(°C)
conv
(%)b
entry
catalysta
1
2
3
4
A
B
C
D
DMF/Et3N (3)
TMEDA
DMF/Et3N (22)
dioxane/Et3N
(1.2)
2c
90
80
25
25
48
54
3
2c
1.2
1.2
18
5
6
7
8
9
E
E
E
E
F
TMEDA
TMEDA
TMEDA
TMEDA
TMEDA
1.2
2
80
80
80
80
80
67
80
1.2c
2c
98
100d
24
2c
aCatalyst A: Pd(PPh3)2Cl2 (1 mol %), CuI (2 mol %). Catalyst B:
Pd(PPh3)2Cl2 (0.5 mol %), CuI (1 mol %). Catalyst C: Pd(OAc)2
(3 mol %), Xantphos (6.5 mol %), CuI (2 mol %). Catalyst D:
Pd(OAc)2 (3 mol %), P(t-Bu)3 (6.5 mol %), CuI (2 mol %). Catalyst
E: Na2PdCl4 (0.5 mol %), PIntB (1 mol %), CuI (1 mol %). Catalyst F:
Pd(OAc)2 (0.5 mol %), PIntB (1 mol %), CuI (1 mol %). bDetermined
by HPLC. cAdded at 70 °C. d3 was isolated in 86% overall yield
from 1.
reported an efficient catalyst system for the Sonogashira reac-
tion using Na2PdCl4, 2-(di-tert-butylphosphino)-N-phenyl-
indole (cataCXium PIntB), and CuI in TMEDA.6,7 Applying
these conditions to our system brought the conversion up
to a promising 67% (entry 5). Alkyne homocoupling was a
suspected competing reaction, and the amount of TMSA was
increased to 2 equiv, resulting in a moderate increase in
conversion (entry 6). It was found for the original procedure
(Scheme 1) that addition of TMSA at 70 °C was crucial to
obtain good yield. Addition of TMSA at 70 °C with Beller’s
procedure increased the conversion to 98% (entry 7). Full
conversion was realized by using 2 equiv of TMSA, and the
pure central intermediate 3 was isolated in 86% yield over
two steps (entry 8). The palladium source is critical in this
reaction, as exchange of Na2PdCl4 for Pd(OAc)2 resulted in a
dramatic drop in conversion (entry 9).
The second problem in the original synthesis of the alkyne
FFA1 agonists was that the central intermediate 3 turned out
to be a surprisingly unwilling Sonogashira substrate, even
though the 3-propionic ester side chain is believed to be only
moderately electron donating. Sonogashira reactions are
usually performed with excess alkyne because of the compet-
ing Glaser homocoupling. However, since remaining 3
frequently coelutes with the product, we chose to rather
perform the reactions with a small excess of the aryl halide
to ensure complete conversion of the alkyne, which made
suppression of alkyne homocoupling essential. Although most
aryl iodides produced the desired products in satisfactory
aMethod I: ArBr (0.55 mmol), 3 (0.5 mmol), Pd(PPh3)2Cl2 (1 mol %),
CuI (2 mol %), Et3N (2.4 equiv), DMF (1 mL), 50 °C. Method II: ArBr
(0.55 mmol), 3 (0.5 mmol), Na2PdCl4 (1 mol %), PIntB (2 mol %), CuI
(2 mol %), TMEDA (1 mL), 80 °C. bIsolated yields.
yields, the aryl bromides performed poorly, and in several
cases failed completely (e.g., Table 2, entries 1 and 3). The
optimized conditions from the TMSA coupling were investi-
gated with the aryl bromides which had failed or performed
poorly with the original catalyst system. The new conditions
proved efficient for aryl bromide with both electron donating
and electron withdrawing substituents (Table 2), including
yields above 80% in cases where the original protocol had
failed completely (entries 1-4).
The potent FFA1 agonist TUG-424 is synthesized via
intermediate 5h, which was obtained in only 31% yield from
coupling of o-iodotoluene (4h) with 3 by the original Sono-
gashira protocol (Table 3, entry 1).4 Applying the optimized
coupling conditions with this reaction surprisingly led to
further decreased yield (entry 2). Substantial amounts of
alkyne dimerization were observed with both the new and the
original conditions. Since alkyne dimerization is known to be
promoted by Cu(I), a copper-free protocol was investigated.8
This resulted in a yield only comparable to the original
procedure (entry 3). Other aryl iodides also produced un-
satisfactory results. This was surprising, since aryl iodides
normally are significantly more reactive substrates than aryl
bromides in the Sonogashira reaction. Noticing that the
original report by Beller and co-workers only included aryl
bromides, we replaced 4h by the bromo-analogue 4i, which
raised the yield to 67% (entry 4). To investigate the idea that
the coupling conditions might prefer aryl bromides to aryl
(6) Torborg, C.; Zapf, A.; Beller, M. ChemSusChem 2008, 1, 91–96.
(7) Other studies involving PIntB (cataCXium P): (a) Rataboul, F.; Zapf,
A.; Jackstell, R.; Harkal, S.; Riermeier, T.; Monsees, A.; Dingerdissen, U.;
Beller, M. Chem.;Eur. J. 2004, 10, 2983–2990. (b) Choi, Y. L.; Yu, C.-M.;
Kim, B. T.; Heo, J.-N. J. Org. Chem. 2009, 74, 3948–3951. (c) Schulz, T.;
€
Torborg, C.; Enthaler, S.; Schaffner, B.; Dumrath, A.; Spannenberg, A.;
Neumann, H.; Borner, A.; Beller, M. Chem.;Eur. J. 2009, 15, 4528–4533.
€
(d) Harkal, S.; Kumar, K.; Michalik, D.; Zapf, A.; Jackstell, R.; Rataboul,
F.; Riermeier, T.; Monsees, A.; Beller, M. Tetrahedron Lett. 2005, 46, 3237–
3240. (e) Ebran, J.-P.; Hansen, A. L.; Gogsig, T. M.; Skrydstrup, T. J. Am.
Chem. Soc. 2007, 129, 6931–6942.
ꢀ
(8) Komaromi, A.; Novak, Z. Chem. Commun. 2008, 40, 4968–4970.
ꢀ
1302 J. Org. Chem. Vol. 75, No. 4, 2010

Christiansen et al.
JOCNote
TABLE 3. Reactivity Differences of Aryl Iodides and Bromidesa
TABLE 4. Screening of Cosolventsa
entry
cosolvent
7 (%)c
dimer (%)c
1
2
3
4
5
6
none
84
46
11
54
42
19
12
2
DMSO
EtOAc
PEG-400
DMF
46
77b
78b
98
H2O
aReaction conditions: 4h (0.55 mmol), 6(0.5 mmol), Na2PdCl4 (1mol%),
PIntB (2 mol %), CuI (2 mol %), TMEDA/cosolvent (1 mL, 1:1, v/v), 80 °C.
bReaction not complete. cDetermined by HPLC.
TABLE 5. Optimization of Water Contenta
aReaction conditions: 4h-k (0.55 mmol), 3 (0.5 mmol), 80 °C, volume
(1 mL). bCatalyst G: Pd(PPh3)4 (1 mol %), CuI (2 mol %). Catalyst E:
Na2PdCl4 (1 mol %), PIntB (2 mol %), CuI (2 mol %). Catalyst H: Pd/C
(1 mol %), SPhos (1 mol %). Catalyst I: Na2PdCl4 (1 mol %), PIntB
(2 mol %). cPerformed at 90 °C. dPerformed at 110 °C. e10 vol % of DMF.
iodides, the substrates 4j and 4k were employed, resulting in
isolation of the bromo-substituted products in 40% and 15%
yield, respectively, accompanied by the dicoupled products
(entries 5-6). Thus, aryl iodides react preferentially, but
yields were invariably poor. It was observed that a solid
precipitate was formed in all reactions with aryl bromides,
while a sticky syrup was formed in the reactions with aryl
iodides. We hypothesized that the poor yields with the
aryl iodides were a consequence of the catalyst complex or
essential components being trapped in the syrup, and we
initiated a search for conditions that would produce a
homogeneous reaction mixture. Addition of 10% DMF to
the reaction increased the yield from 22% to 40% (entry 7),
but only reduced the alkyne dimerization slightly. Removal
of CuI to suppress competing alkyne homocoupling was
fatal for the reaction (entry 8).
Phenylacetylene (6) was chosen as a less precious substrate
for rapidly screening different conditions that would dissolve
the syrup without having detrimental effects on the reaction.
This usually excellent Sonogashira substrate only gave 84%
conversion in coupling with 4h in pure TMEDA. Adding
50% DMSO or ethyl acetate as cosolvents reduced the
conversion considerably (Table 4, entries 2 and 3), while
50% PEG-400 and DMF led to a slight decrease in conver-
sion compared to pure TMEDA (entries 4 and 5). On the
other hand, addition of water (entry 6) resulted in a homo-
geneous reaction mixture and almost full conversion to the
desired coupling product.
aReaction conditions: ArX (0.55 mmol), 3 (0.5 mmol), Na2PdCl4
(1 mol %), PIntB (2 mol %), CuI (2 mol %), solvent/base (1 mL), 80 °C.
bThe double-coupled product was isolated in 48% yield on the basis
of 3 together with the 3-bromo-substituted 5k.
whereas 10% water increased the yield further to 79%
(entry 3). The corresponding bromide 4i gave 86% yield
with the same conditions (Scheme 2). Coupling with 4k
(entry 4) resulted in isolation of the iodo-coupled and the
double-coupled products in equal amounts, indicating that a
preference for coupling of iodides over bromides still exists.
The beneficial effect of 10% water was found to be general
for both aryl iodides and bromides (entries 5-10). Likewise,
addition of 10% water to the coupling of 1 to TMSA further
increased the yield to 89% over two steps, and the reaction
time decreased from 2 h to 5 min (Scheme 2).
In conclusion, we have identified an efficient general
protocol for Pd/Cu-catalyzed coupling of aryl bromides
and iodides to alkynes, using the recently reported ligand
cataXCium PIntB6 and TMEDA/water as reaction medium.
Addition of 50% water as cosolvent with TMEDA more
than doubled the yield of 5h (Table 5, entry 1). Reducing
the amount of water to 20% brought the yield up to 76%,
J. Org. Chem. Vol. 75, No. 4, 2010 1303

Christiansen et al.
JOCNote
SCHEME 2. Optimized Synthesis of TUG-424a
Experimental Section
General Procedure for Sonogashira Coupling. A Schlenk flask
charged with Na2PdCl4 (1 mol %), 2-(di-tert-butylphosphino)-
N-phenylindole (PIntB, 2 mol %), CuI (2 mol %), alkyne
(1 equiv), aryl halide (1.1 equiv), H2O (0.2 mL/mmol), and
TMEDA (1.8 mL/mmol) was evacuated and backfilled with
argon three times, then heated to 80 °C. After consumption of
the starting material (<30 min), the reaction mixture was cooled
to rt, then water was added and extracted with EtOAc. The
combined extracts were washed with brine, dried over MgSO4,
and concentrated, and the residue was purified by flash chro-
matography.
aReagents and conditions: (a) MeOH, HCl (cat.), rt, 2 h (99%); (b) TMSA,
Na2PdCl4 (1 mol %), PIntB (2 mol %), CuI (2 mol %), TMEDA, H2O,
80 °C, 5 min; (c) K2CO3, MeOH, rt, 2.5 h (89% over 2 steps); (d) 2-bromoto-
luene (4i), Na2PdCl4 (1 mol %), PIntB (2 mol %), CuI (2 mol %), TMEDA,
H2O, 80 °C, 30 min (86%); (e) LiOH, 1,4-dioxane, H2O, rt (100%).
Methyl 3-(4-((2,6-Dichloropyridin-4-yl)ethynyl)phenyl)-
propanoate (5m). 5m was prepared from 3 (190 mg, 1.01 mmol)
and 4m (302 mg, 1.10 mmol) according to the general procedure
to give 301 mg (89%) of a white solid after purification by flash
chromatography (SiO2, EtOAc:petroleum ether, 1:7): Rf 0.11
1
(EtOAc:petroleum ether, 1:7); H NMR (CDCl3) δ 7.48-7.45
The reactions proceed cleanly, usually with no detectable
alkyne homocoupling, and are complete in less than 30 min.
It was observed in previous reactions that longer reaction
times often resulted in more of the homocoupled product,
and it appears likely that suppression of homocoupling is at
least partly a consequence of the high Sonogashira cross-
coupling rate obtained with the new conditions. Whereas the
original protocol produces discolored products, even if pure
by NMR and HPLC, the new protocol generally produces
white or yellow products. Implementation of the new Sono-
gashira protocol resulted in a vastly improved synthesis of
the FFA1 receptor ligands by enabling efficient synthesis of
the central intermediate 3 from 3-(4-bromophenyl)propionic
acid, and by providing a rapid, reliable, practical, and
efficient method for coupling of this intermediate with
diverse aryl bromides and iodides.
(m, 2H), 7.34 (s, 1H), 7.24-7.22 (m, 2H), 3.68 (s, 3H), 2.99 (t,
J = 7.7 Hz, 2H), 2.65 (t, J = 7.7 Hz, 2H); 13C NMR (CDCl3) δ
172.9, 150.7, 142.9, 136.6, 132.3, 128.7, 124.4, 119.1, 96.9, 81.4,
51.7, 35.2, 30.9; ESI-HRMS calcd for C17H13Cl2NO2 (M þ
Naþ) 356.0216, found 356.0222.
Acknowledgment. We thank the Danish Natural Science
Research Council, the Danish Medical Research Council,
and The Danish Council for Independent Research/
Technology and Production Sciences (grant 09-070364) for
financial support.
Supporting Information Available: Additional experimental
1
procedures, compound characterization data, and H and 13C
NMR spectra. This material is available free of charge via the
Internet at http://pubs.acs.org.
1304 J. Org. Chem. Vol. 75, No. 4, 2010