Alternatives to vinyl triflates for cross-coupling with arylboronic acids

Article abstract of DOI:10.1055/s-1999-2643

Vinyl phosphates, mesylates, and tosylates derived from 1,3-dicarbonyl compounds are coupled with phenylboronic acid using nickel or palladium catalysts. An application to the area of 2-aryl carbapenem antibiotics is demonstrated. Alternative conditions for the nickel-catalyzed coupling of aryl mesylates are also reported.

Full text of DOI:10.1055/s-1999-2643

LETTER
471
Alternatives to Vinyl Triflates for Cross-Coupling with Arylboronic Acids
Mark A. Huffman*, Nobuyoshi Yasuda
Department of Process Research, Merck Research Laboratories, P.O. Box 2000, Rahway, New Jersey 07065
E-mail: mark_huffman@merck.com
Received 8 September 1998
Abstract: Vinyl phosphates, mesylates, and tosylates derived from
1,3-dicarbonyl compounds are coupled with phenylboronic acid us-
ing nickel or palladium catalysts. An application to the area of 2-
aryl carbapenem antibiotics is demonstrated. Alternative conditions
for the nickel-catalyzed coupling of aryl mesylates are also report-
ed.
Key words: cross-coupling, boronic acid, carbapenem, palladium,
nickel
The Suzuki-Miyaura cross-coupling of arylboronic acids
with vinyl triflates1 provides convenient access to a vari-
ety of 2-aryl carbapenem antibiotics.2 Unfortunately, the
high cost of triflic anhydride and the instability of some
vinyl triflates present disadvantages for large-scale syn-
thesis. In order to avoid these drawbacks, we sought con-
ditions for cross-coupling arylboronic acids with other
vinylic functional groups derived from 1,3-dicarbonyl
compounds.3 Reported here are results for coupling phen-
ylboronic acid with some simple vinyl mesylates, tosy-
lates, and diphenylphosphates, along with an application
to a carbapenem vinyl tosylate.
A more convenient version of this reaction would employ
an air-stable pre-catalyst which is reduced to Ni(0) in situ
by the boronic acid. The Ni(I) complexes NiCl(PPh3)3 and
NiBr(PPh3)3,8 in the presence of additional PPh3, are ap-
proximately equal in cross-coupling reactivity to
Ni(PPh3)4, but they provide little improvement in air-sta-
bility. While air-stable Ni(II) chloride phosphine com-
plexes are not reduced by phenylboronic acid, the
analogous bromides and iodides are. The most active
Ni(II) complex we have found is NiI2(dppp),9 combined in
a 1 : 2 ratio with PPh3. Its reduction by phenylboronic acid
is too slow to give a high yield with the unstable phos-
phate 2, but it performs well with more robust substrates
(vide infra).
Nickel Catalysts. Initial efforts focused on the enol phos-
phate 2, since the carbapenem enol phosphates are well
known.4 The report by Percec et al. on nickel-catalyzed
aryl mesylate/arylboronic acid coupling5 encouraged us to
try nickel catalysis with this substrate. The reported con-
ditions (NiCl2(dppf), Zn dust, K3PO4)6 did induce cou-
pling of 2 with phenylboronic acid. However, we found
the reaction to be inconsistent and very sensitive to con-
centration and the order of reagent addition. The difficul-
ties presumably arise in the heterogeneous zinc reduction
of the Ni(II) precatalyst. By starting with Ni(0) in the form
of Ni(PPh3)4, the heterogeneous reduction is eliminated
and reliable high-yielding cross-coupling occurs (see Ta-
ble).7 The drawback of the Ni(0) catalyst is its extreme air
sensitivity, even in the solid state.
The nickel catalysts were not successful in carbapenem
enol phosphate couplings, nor were they very effective
with the vinyl sulfonates 3 and 4. On the other hand, cross-
coupling with aryl mesylate 5 performed extremely well,
even using the air-stable Ni(II) iodide catalyst. These cat-
alysts provide significant yield improvement over the
Ni(II) chloride/Zn method,5 and without the zinc, reduc-
tive homocoupling of the aryl mesylate is not an issue. 10
Palladium Catalysts. Phosphine-free palladium catalysts
such as Pd(dba)26 and Pd(OAc)2 can be very effective for
mild Suzuki-Miyaura couplings with unsaturated triflates,
iodides and bromides,2,11 yet these catalysts are complete-
ly ineffective with the less reactive enol phosphate 2 and
Equation 1
Synlett 1999, No. 4, 471–473 ISSN 0936-5214 © Thieme Stuttgart · New York

472
M. A. Huffman et al.
LETTER
sis. Roth, G. P.; Sapino, C. Tetrahedron Lett. 1991, 32, 4073-
4076.
(4) (a) Salzmann, T. N.; Ratcliffe, R. W.; Christensen, B. G.;
Bouffard, F. A. J. Am. Chem. Soc. 1980, 102, 6163-6165.
(b) Melillo, D. G.; Shinkai, I.; Liu, T.; Ryan, K.; Sletzinger,
M. Tetrahedron Lett. 1980, 21, 2783-2786.
(5) Percec, V.; Bae, J.-Y.; Hill, D. H. J. Org. Chem. 1995, 60,
1060-1065.
(6) dppf = 1,1´-Bis(diphenylphosphino)ferrocene; dppp = 1,3-
Bis(diphenylphosphino)propane; dppb = 1,4-Bis(diphenyl-
phosphino)butane; dba = Dibenzylideneacetone.
(7) Characterization of new compounds. 2: colorless oil; 1H NMR
(250 MHz, CDCl3) d 7.37 (m, 4H), 7.26 (m, 6H), 6.04 (s, 1H),
2.57 (app. t, J = 6.2 Hz, 2H), 2.38 (app. t, J = 6.7 Hz, 2H), 2.04
(app. quint, J = 6.5 Hz, 2H); 13C NMR (63 MHz, CDCl3) d
198.9, 168.7 (d, J = 9.1 Hz), 150.0 (d, J = 7.4 Hz), 130.0,
126.0, 120.0 (d, J =4.7 Hz), 115.0 (d, J = 4.8 Hz), 36.3, 28.6
(d, J = 5.6 Hz), 20.8; IR (thin film) 1676, 1633, 1589, 1489
cm-1. Anal. Calcd for C18H17O5P: C, 62.79; H, 4.98;. Found:
C, 62.53; H, 4.96. 4: white solid; 1H NMR (250 MHz, CDCl3)
d 7.83 (dm, J = 8.4 Hz, 2H), 7.33 (dm, J = 8.7 Hz, 2H), 4.07
(q, J = 7.1 Hz, 2H), 2.45 (s, 3H), 2.37 (m, 2H), 2.22 (m, 2H),
1.62 (m, 2H), 1.23 (t, J = 7.1 Hz, 3H); 13C NMR (63 MHz,
CDCl3) d 165.9, 151.1, 145.1, 134.1, 129.7, 128.0, 121.8,
60.8, 29.0, 26.0, 22.2, 21.7, 21.3, 14.0; IR (thin film)
1721 cm-1. Anal. Calcd for C16H20O5S: C, 59.24; H, 6.21;.
Found: C, 59.16; H, 6.28. The data for compounds 6 and 7 are
listed in note 12. All other compounds gave data in accordance
with values reported in the literature.
sulfonates 3 and 4. Much better are Pd(II) phosphine com-
plexes PdCl2(PPh3)2 and PdCl2(diphosphine). Of these,
PdCl2(dppb) gave the most rapid reactions, but commer-
cially available PdCl2(dppf) was a close second.6 Using
aqueous K2CO3 in DMF, the latter catalyst worked well
with the enol phosphate and both simple vinyl sulfonates.
Like the nickel catalysts, palladium catalysts failed in at-
tempted carbapenem enol phosphate couplings. On the
other hand, PdCl2(dppf) catalyzed rapid reaction of phe-
nylboronic acid with the analogous vinyl tosylate 6. Un-
fortunately, both 6 and the product 7 are unstable to the
reaction conditions in aqueous basic DMF. In THF, stabil-
ity was better but the reaction slowed when a second liq-
uid phase appeared part way through. By using 10 %
palladium catalyst and added phase transfer catalyst
Bu4N+Cl-, the balance of reactivity and stability allowed a
67 % isolated yield.12
(8) D’Aniello, M. J. Jr.; Barefield, E. K. J. Am. Chem. Soc. 1978,
100, 1474-1481.
(9) Prepared from NiI2 and dppp in 1-propanol. Van Hecke, G. R.;
Horrocks, W. D. Jr. Inorg. Chem. 1966, 5, 1968-1974. Depen-
ding on the amount of water present during preparation, we
obtained either light purple or dark red crystals which were
equal in activity.
(10) An alternative in-situ formation of Ni(0) from Ni(II) and n-
BuLi has been reported for cross-coupling of aryl chlorides.
(a) Saito, S.; Sakai, M.; Miyaura, N. Tetrahedron Lett. 1996,
37, 2993-2996. For related nickel-catalyzed couplings, see al-
so: (b) Kobayashi, Y.; Mizojiri, R. Tetrahedron Lett. 1996, 37,
8531-8534. (c) Indolese, A.F.; Tetrahedron Lett. 1997, 38,
3153-3516.
Equation 2
(11) Wallow, T. I.; Novak, B. M. J. Org. Chem. 1994, 59, 5034-
5037.
(12) Synthesis of 6: The b-keto ester (eq.1 and ref 4) (3.48 g, 10
mmol) was suspended in 50 mL of dry CH2Cl2 at -60 °C.
Triethylamine (1.46 mL, 10.5 mmol) was added dropwise fol-
lowed by p-toluenesulfonic anhydride (3.26 g, 10 mmol). The
suspension was warmed slowly to 0 °C, then poured into a se-
peratory funnel and washed once with water and once with aq.
NaHCO3. The CH2Cl2 solution was kept cold while drying
over Na2SO4, followed by solvent evaporation. The residual
oil was redissolved in 50 mL dry CH2Cl2 and cooled below
-70° C. Triethylamine (1.95 mL, 14 mmol) was added slowly,
followed by t-butyldimethylsilyltrifluoromethanesulfonate.
The solution was stirred below -70°C for 12h. After washing
twice with water, drying over Na2SO4 and solvent evaporati-
on, the crude product was chromatographed on silica with 4 :
1 hexanes/ethyl acetate. Concentration and addition of hexa-
nes gave a white solid which was isolated by filtration, washed
with hexanes and dried under vacuum (3.17g, 51 %). 1H NMR
(250 MHz, CDCl3) d 8.18 (dm, J = 8.7 Hz, 2H), 7.83 (dm, J =
8.4 Hz, 2H), 7.52 (dm, J = 8.7 Hz, 2H), 7.33 (dm, J = 8.5 Hz,
2H), 5.29 (d, J = 14.0 Hz, 1H), 5.17 (d, J = 14.0 Hz, 1H), 4.23
(m, 2H), 3.21 (m, 3H), 2.44 (s, 3H), 1.23 (d, J = 6.2 Hz, 3H),
0.87 (s, 9H), 0.76 (s, 3H), 0.67 (s, 3H); 13C NMR (63 MHz,
In summary, we have identified conditions with which ac-
tivated vinyl triflates can be replaced by more economical
and at times more stable vinyl phosphates, mesylates and
tosylates for Suzuki-Miyaura cross-coupling with arylbo-
ronic acids. This was achieved by employing nickel or
palladium phosphine catalysts which will insert into these
less reactive substrates.
References and Notes
(1) (a) Oh-e, T.; Miyaura, N.; Suzuki, A. J. Org. Chem. 1993, 58,
2201-2208. (b) Fu, J.-m.; Snieckus, V. Tetrahedron Lett.
1990, 31, 1665-1668.
(2) (a) Yasuda, N.; Xavier, L.; Rieger, D. L.; Li, Y.; DeCamp, A.
E.; Dolling, U.-H. Tetrahedron Lett. 1993, 34, 3211-3214.
(b) Decamp, A.; Dolling, U.-H.; Li, Y.; Rieger, D. L.; Yasuda,
N.; Xavier, L. C. U.S. Patent 5,338,875, 1994.
(3) The vinyl fluorosulfonate group has been reported as a triflate
replacement in Stille-type couplings for cephalosporin synthe-
Synlett 1999, No. 4, 471–473 ISSN 0936-5214 © Thieme Stuttgart · New York

LETTER
Alternatives to Vinyl Triflates for Cross-Coupling with Arylboronic Acids
473
CDCl3) d 177.3, 158.3, 150.7, 147.5, 146.4, 142.5, 132.0,
(ethyl acetate/water) the crude product was chromatographed
on silica with 4:1 hexanes/ethyl acetate. Solvent evaporation
gave a white crystalline solid (174 mg, 67 %). 1H NMR (250
MHz, CDCl3) d 8.14 (dm, J = 8.8 Hz, 2H), 7.44 (dm, J = 8.8
Hz, 2H), 7.35 (s, 5H), 5.35 (d, J = 13.9 Hz, 1H), 5.19 (d, J =
13.9 Hz, 1H) 4.30 (m, 2H), 3.25 (m, 3H), 1.28 (d, J = 6.2 Hz,
3H), 0.88 (s, 9H), 0.10 (s, 6H); 13C NMR (63 MHz, CDCl3) d
176.6, 160.6, 147.5, 145.8, 142.7, 133.3, 129.0, 128.2, 128.03,
128.00, 126.6, 123.6, 67.3, 65.9, 65.3, 52.1, 42.5, 25.7, 22.5,
18.0, -4.2, -4.9; IR (thin film) 1778, 1725 cm-1. Anal. Calcd
for C28H34N2O6Si: C, 64.34; H, 6.56; N, 5.36. Found: C,
64.35; H, 6.46; N, 5.32.
130.0, 128.4, 127.8, 123.6, 121.6, 68.3, 65.5, 65.2, 50.6, 36.5,
25.7, 22.3, 21.7, 17.9, -4.3, -5.0; IR (thin film) 1785, 1732 cm-1.
Anal. Calcd for C29H36N2O9SiS: C, 56.48; H, 5.88; N, 4.54.
Found: C, 56.50; H, 5.73; N, 4.47.
Synthesis of 7: Phenylboronic acid (80 mg, 0.66 mmol) and
tetrabutylammonium chloride hydrate (6 mg, 0.25 mmol)
were combined in 3 mL THF. Aqueous 3.0 M potassium car-
bonate (0.44 mL, 1.32 mmol) was added and the solution was
sparged with nitrogen. Vinyl tosylate 5 (308 mg, 0.50 mmol)
was added followed by brief nitrogen sparging and addition of
PdCl2(dppf)•CH2Cl2 (41 mg, 0.05 mmol). The mixture was
stirred in a 30 °C bath for 100 min. After extractive workup
Synlett 1999, No. 4, 471–473 ISSN 0936-5214 © Thieme Stuttgart · New York