Copper-catalyzed arylation of beta-amino alcohols.

Article abstract of DOI:10.1021/ol026655h

[reaction: see text] Methods for synthesizing N-aryl beta-amino alcohols and O-aryl beta-amino alcohols are described. The presence of a neighboring hydroxyl or amino group, respectively, is thought to activate beta-amino alcohols toward these transformations. These protocols significantly increase access to a variety of arylated beta-amino alcohols.

Full text of DOI:10.1021/ol026655h

ORGANIC
LETTERS
2002
Vol. 4, No. 21
3703-3706
Copper-Catalyzed Arylation of â-Amino
Alcohols
Gabriel E. Job and Stephen L. Buchwald*
Department of Chemistry, Massachusetts Institute of Technology,
Cambridge, Massachusetts 02139
sbuchwal@mit.edu
Received August 1, 2002
ABSTRACT
Methods for synthesizing N-aryl â-amino alcohols and O-aryl â-amino alcohols are described. The presence of a neighboring hydroxyl or
amino group, respectively, is thought to activate â-amino alcohols toward these transformations. These protocols significantly increase access
to a variety of arylated â-amino alcohols.
â-Amino alcohols are of interest in medicinal chemistry1 and
in catalytic organometallic chemistry,2 where their use as
ligands for enantioselective addition of organozinc reagents
to carbonyls is well-known. The ubiquity of the â-amino
alcohol motif has inspired several synthetic pathways to the
N-aryl derivatives of â-amino alcohols;3 perhaps the most
practical route is the reaction of epoxides with anilines. The
low reactivity of anilines with respect to aminolysis has
prompted the development of metal amides and Lewis acid
activators to overcome this limitation.4 The use of tin(II)
triflate and copper(II) triflate to catalyze the reaction of
epoxides with even weak nucleophiles such as p-nitroaniline
exemplifies this approach.5 Another noteworthy route is the
Ir-catalyzed asymmetric reduction of N-aryl imines, which
has found use in the synthesis of the pesticide metalochlor.6
The Rh-catalyzed aminolysis of vinyl epoxides devised by
Lautens provides an effective route to many N-aryl â-amino
alcohols; a minor limitation is the need to employ an excess
of amine.7 Lautens has also developed a highly enantio-
selective version for meso-oxabicyclic alkenes.8 Conditions
for selective arylation of â-amino alcohols by SNAr9 have
been discovered, but this approach has a limited substrate
scope.
Ma has reported a mild and convenient method for the
N-arylation of the structurally analogous R-amino acids.10
The high reactivity of R-amino acids, as compared to that
of simple amines, is attributed to the ability of these
compounds to form a chelate with Cu(I). One might
anticipate â-amino alcohols to react similarly to R-amino
acids, but Ma found valinol to be no more reactive than
benzylamine in his system. In contrast to this finding, Hida
observed that â-amino alcohols are more reactive than simple
amines in the Ullmann condensation with bromoanthro-
quinones;11 however, the 100:1 molar ratio of amino alcohol
to aryl bromide used by Hida generally makes the reaction
impractical.
(1) In addition to epinephrine and â-blockers: Rogers, G. A.; Parsons,
S. M.; Anderson, D. C.; Nilsson, L. M.; Bahr, B. A.; Kornreich, W. D.;
Kaufman, R.; Jacobs, R. S.; Kirtman, B. J. Med. Chem. 1989, 32, 1217.
(2) Bolm, C.; Hildebrand, J. P.; Mun˜iz, K.; Hermanns, N. Angew. Chem.,
Int. Ed. 2001, 40, 3284.
(3) (a) Sasaki, T.; Mori, K.; Ohno, M. Synth. Commun. 1985, 280. (b)
Barluenga, J.; Alonso-Cires, L.; Asensio, G. Synth. Commun. 1981, 376.
(c) Barluenga, J.; Fan˜anas, F. J.; Yus, M. J. Org. Chem. 1979, 44, 4798.
(d) Sanna, P.; Savelli, F. J. Heterocycl. Chem. 1984, 21, 297. (d) Rampalli,
S.; Chaudhari, S. S.; Akamanchi, K. G. Synthesis 2000, 1, 78.
(4) (a) Harrak, Y.; Pujol, M. D. Tetrahedron Lett. 2002, 43, 819. (b)
Chini, M.; Crotti, P.; Macchia, F. Tetrahedron Lett. 1990, 31, 4661. (c)
Overman, L. E.; Flippin, L. A. Tetrahedron Lett. 1981, 22, 195.
(5) Sekar, G.; Singh, V. K. J. Org. Chem. 1999, 64, 287.
(7) Fagnou, K.; Lautens, M. Org. Lett. 2000, 2, 2319.
(8) Lautens, M.; Fagnou, K. J. Am. Chem. Soc. 2001, 123, 7170.
(9) Beugelmans, R.; Bigot, A.; Zhu, J. Tetrahedron Lett. 1994, 35, 5649.
(10) Ma, D.; Zhang, Y.; Yao, J.; Wu, S.; Tao, F. J. Am. Chem. Soc.
1998, 120, 12459.
(6) Spindler, F.; Pugin, B.; Blaser, H.-U. Angew. Chem., Int. Ed. Engl.
1990, 29, 558.
(11) Arai, S.; Takamichi, Y.; Ototake, S.; Hida, M. Bull. Chem. Soc.
Jpn. 1977, 50, 547.
10.1021/ol026655h CCC: $22.00 © 2002 American Chemical Society
Published on Web 09/18/2002

Recently, there has been renewed interest in Cu-catalyzed
coupling reactions.12 As a continuation of our long-standing
interest in Pd-catalyzed C-N and C-O bond formation,13
our group has contributed to the expansion of the analogous
Cu-catalyzed chemistry.14 In keeping with these efforts, we
sought to develop a practical arylation reaction that would
exploit the increased reactivity of â-amino alcohols.
Table 1. N-Arylation Using Sodium Hydroxide as Basea
A major challenge in the arylation of â-amino alcohols
lies in controlling the site of reaction. During a preliminary
survey of bases and solvents, the use of sodium tert-butoxide
in DMSO was found to give moderately selective N-arylation
(N/O ∼5). Greater N/O selectivity on a humid day led to
the deliberate addition of water and subsequent improvement
of selectivity, culminating in a DMSO/H2O (2:1) solvent
system, with sodium hydroxide as base, yielding N/O >50.
Eventually, 2-propanol was found to be essentially equivalent
to the DMSO/H2O system for N-arylation.
To assess the effect of the proximal hydroxyl group for
the N-arylation of â-amino alcohols, the reactions shown in
Scheme 1 were performed. While 2,6-dimethyl iodobenzene
Scheme 1. Control Experiments for the Putative Chelation
Effect
a Conditions: 1 mL of iPrOH or DMSO/H2O under N2 in a sealed tube.
b Reactions 1-3. c Reactions 4-7. d Average of two isolated yields;
satisfactory combustion analyses were obtained for all compounds. No
evidence for the O-aryl isomers was seen in the 1H NMR spectra of the
isolated products. e Temperature ) 100 °C. f The amino alcohol was used
as the HCl salt, requiring 3 mmol of base. g The stoichiometries of the aryl
iodide and the amino alcohol were reversed. h Three equivalents of the amino
alcohol and 0.8 mL of iPrOH were used. i Reaction was run in neat amino
alcohol.
was used successfully in entry 1 of Table 1, the analogous
reaction with n-hexylamine yielded only a trace amount of
the desired product. Furthermore, the reactions of simple
secondary amines yielded no product in contrast to entries
6a and 6b in Table 1.
Since the costs of â-amino alcohols and aryl iodides are
frequently comparable, the use of either as the limiting
reagent was explored. With sodium hydroxide as base, the
yield is somewhat better using a moderate excess of aryl
iodide rather than amino alcohol (entries 4a and 4b in Table
1) owing to competitive conversion of the aryl iodide to the
phenol and, in 2-propanol solvent, the isopropyl aryl ether.
When the amino alcohol is used as the limiting reagent,
0-5% of the amino alcohol remains after approximately 16
h. In general, the use of 1.4 equiv of aryl iodide or the use
of 10% CuI could eliminate the residual amino alcohol,
although no method was completely general. These measures
usually led to a reduced N/O ratio and/or increased amounts
of diaryl byproducts. Consequently, no significant improve-
ment in yield was realized and the residual amino alcohol
was tolerated in our cases since it is easily separated from
the N-aryl product.15
As sodium hydroxide curtails functional group tolerance,
the use of mild bases such as K3PO4 and Cs2CO3 was
investigated. The selectivity observed in reactions using mild
bases was initially low. Subsequently, in our laboratory,
ethylene glycol was identified as a ligand for the Cu-
catalyzed arylation of simple aliphatic amines,14 and so we
investigated its use in the arylation of amino alcohols.
(12) (a) Gujadhur, R. K.; Bates, C. G.; Venkataraman, D. Org. Lett. 2001,
3, 4315. (b) Gujadhur, R.; Venkataraman, D.; Kintigh, J. T. Tetrahedron
Lett. 2001, 42, 4791. (c) Bates, C. G.; Gujadhur, R. K.; Venkataraman, D.
Org. Lett. 2002, 4, 2803. (d) Fagan, P. J.; Hauptman, E.; Shapiro, R.;
Casalnuovo, A. J. Am. Chem. Soc. 2000, 122, 5043. (e) Lang, F.; Zewge,
D.; Houpis, I. N.; Volante, R. P. Tetrahedron Lett. 2001, 42, 3251.
(13) (a) Muci, A. R.; Buchwald, S. L. Practical Palladium Catalysis for
C-N and C-O Bond Formation. In Topics in Current Chemistry; Miyaura,
N., Ed.; Springer-Verlag: Berlin, 2002; Vol. 219, p 133.
(14) Amides and heterocycles: (a) Klapars, A.; Antilla, J. C.; Huang,
X.; Buchwald, S. L. J. Am. Chem. Soc. 2001, 123, 7727. Amines: (b)
Kwang, F. Y.; Klapars, A.; Buchwald, S. L. Org. Lett. 2002, 4, 581.
Alcohols: (c) Wolter, M.; Nordmann, G.; Job, G. E.; Buchwald, S. L. Org.
Lett. 2002, 4, 973. Hydrazines: (d) Wolter, M.; Klapars, A.; Buchwald, S.
L. Org. Lett. 2001, 3, 3803.
(15) For example, the compound in entry 2 of Table 1 could be purified
by extraction. After concentration, analytically pure material was obtained.
3704
Org. Lett., Vol. 4, No. 21, 2002

compromise between selectivity and reaction time was
realized using butyronitrile as the solvent, although toluene
was better in one case (entry 4 in Table 3). In the reactions
Table 2. N-Arylation Using K3PO4 as Basea
Table 3. O-Arylation of Amino Alcoholsa
a Conditions: 1.1 mL of 2-propanol under N2 in a sealed tube. b Average
of two isolated yields; satisfactory combustion analyses were obtained for
1
all products. No evidence of the O-aryl isomers was seen in the H NMR
spectra of the isolated products. c The stoichiometries of aryl iodide and
amino alcohol were reversed. d Temperature ) 80 °C
Addition of 1 equiv of ethylene glycol (Table 2) suppresses
formation of the O-aryl isomer, although the N/O selectivities
observed are usually lower than those seen with the
hydroxide method. While nitro and ketone functional groups
were well-tolerated, low yields were obtained in reactions
using ethyl 2-iodobenzoate16 and 3-iodobenzonitrile. As with
the hydroxide method, residual amino alcohol is observed
with its use as the limiting reagent in the mild base protocol.
However, the use of the aryl iodide as the limiting reagent,
in combination with K3PO4, results in the better yield (entries
1a and 1b in Table 2).
A trait common to both N-arylation protocols is the
diminished reactivity of â-amino alcohols containing a
secondary amine (secondary amino alcohols). Although the
hydroxide method could effect the N-arylation of 2-(ethy-
lamino)ethanol in acceptable yield, N-phenyl ephedrine was
obtained in less than 50% yield under the conditions shown
in Table 1 owing to the incomplete conversion of ephedrine.
These conditions were modified (2 equiv of iodobenzene,
56 h, 100 °C) in an attempt to overcome this difficulty;
however, approximately 30% of the ephedrine still remained
at the end of the reaction. The reactivity of ephedrine
compared with that of the amino alcohols in Table 1 suggests
that the presence of R-branching or a secondary amine may
be tolerated individually but not together.
Low N/O ratios were obtained in reactions using mild
bases and secondary â-amino alcohols. At times, moderately
selective O-arylation was observed for these compounds, thus
raising the possibility of complementary O-arylation. To
assess this possibility, a variety of conditions were screened.
Cesium carbonate proved to be the best base for C-O bond
formation, consistent with previous experience.14 The best
a Conditions: 1 mL of butyronitrile under N2 in a sealed tube. b Average
of two isolated yields. Satisfactory combustion analyses were obtained for
entries 1-4. Entries 5 and 6 determined to be >95% pure by 1H NMR;
water was the chief impurity. No evidence of the N-aryl isomers was seen
in the 1H NMR spectra of the isolated products. c The amount of aryl iodide
was 1.5 mmol. d Two equivalents of amino alcohol and 1 equiv of aryl
iodide were used. e Toluene was used as the solvent, and the oil bath
temperature was 115 °C.
of 2-(ethylamino)ethanol, the amino alcohol was used in
excess since it is extremely cheap and the high water
solubility of this amino alcohol simplified the isolation of
its O-aryl derivatives. In the other examples shown in Table
3, though, the amino alcohol was used as the limiting reagent
and was fully consumed, thus allowing isolation of the
arylated products by precipitation as their hydrochloride salts.
Reactions of primary â-amino alcohols, such as those used
in entries 5 and 6 of Table 3, exhibited higher N/O ratios
and/or greater amounts of diaryl byproducts compared to the
secondary â-amino alcohols. The complete lack of reactivity
of simple alcohols under the conditions of Table 317 provides
(17) Nordmann, G.; Wolter, M.; Job, G. E.; Buchwald, S. L. Unpublished
results.
(16) The reaction was run in ethanol.
Org. Lett., Vol. 4, No. 21, 2002
3705

evidence for activation by the neighboring amine group. The
yields in Table 3 are better than or similar to those obtained
with the competitive routes. For example, the SNAr displace-
ment of fluoride from arene-Cr(CO)3 complexes has been
used to generate, in moderate yields, a series of mexiletine
analogues similar to the compound in entry 3 of Table 3.18
Both the Mitsunobu and Williamson reactions using phenols
have been applied to the synthesis of mexiletine analogues,
but the highest yield reported was 48%.19 The opening of a
phthalimidoaziridine with phenol, followed by reduction, has
been reported to occur in good yield,20 although the need
for a phthalimidoaziridine intermediate somewhat reduces
the desirability of this method.
In our study, enantiopure or enantioenriched amino alco-
hols were used when possible. Valinol, phenylglycinol,
norephedrine, and ephedrine were all used in enantiopure
form, whereas trans-2-amino cyclohexanol was used as the
racemate. For all products bearing a single stereocenter,
comparison of the product with its racemate or antipode by
chiral HPLC21 demonstrated complete retention of stereo-
chemistry in the coupling reactions. For those products
bearing two stereocenters, the lack of evidence for formation
of diastereomers, as evidenced by the NMR spectra of the
products and the gas chromatographs of the crude reaction
mixtures, also demonstrated retention of stereochemistry.
These data are consistent with our experience in the Cu-
catalyzed arylation of simple amines and alcohols14 and the
Pd-catalyzed arylation of simple amines using bidentate
phosphine ligands.22
enantiopure â-amino alcohols and many stereoselective
syntheses of â-amino alcohols.24 Thus, in addition to
providing another bond disconnection to consider in a
retrosynthetic analysis, our methods provide increased flex-
ibility in planning a stereoselective synthesis. Furthermore,
our N-arylation methods enjoy an advantage over epoxide
aminolysis. A compound such as that in entry 3 of Table 2
could not be synthesized in good yield by epoxide aminolysis
since the regioisomer resulting from attack at the less
substituted epoxide carbon is expected to predominate. As
for the synthesis of enantiopure O-aryl â-amino alcohols,
the Williamson reaction can yield an enantiopure compound
such as that in entry 1 of Table 3; however, obtaining the
necessary enantiopure halide is not trivial. While the Mit-
sunobu reaction is a viable route to enantiopure O-aryl
â-amino alcohols, it typically gives no better yield than our
O-arylation protocol.
In conclusion, our methods greatly increase both the
number of arylated â-amino alcohols reported and the
synthetic paths to these compounds. The greatest advantage
to our methods, though, lies in their convenience. Com-
mercially available reagents were used without purification
and all reagents were weighed and handled in air. Addition-
ally, our protocols make use of widely available, often
enantiopure, â-amino alcohols.
Acknowledgment. We are grateful to the National
Institutes of Health (GM58160) and the National Cancer
Institute (Training Grant NCI CI T32CA09112) for financial
support. We also thank Pfizer, Merck, and Bristol-Myers-
Squibb for additional unrestricted support
Few enantioselective syntheses of N-aryl â-amino alcohols
have been reported,23 while there exist many natural,
Supporting Information Available: Experimental details
and characterization for all unknown compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(18) Loughhead, D. G.; Flippin, L. A.; Weikert, R. J. J. Org. Chem.
1999, 64, 3373.
(19) Carocci, A.; Catalano, A.; Corbo, F.; Duranti, A.; Amoroso, R.;
Franchini, C.; Lentini, G.; Tortorella, V. Tetrahedron: Asymmetry 2000,
11, 3619.
(20) von Egli, M.; Hoesch, L.; Dreiding, A. S. HelV. Chim. Acta 1985,
68, 220.
OL026655H
(21) Chromatography was performed on an (R,R)-Whelk-O analytical
column. The conditions are described in Supporting Information.
(22) Wolfe, J. P.; Wagaw, S.; Marcoux, J.-F.; Buchwald, S. L. Acc. Chem.
Res. 1998, 31, 805.
(23) In addition to refs 6 and 8: (a) Fu, X.-L.; Wu, S.-H. Synth. Commun.
1997, 27, 1677. (b) Hou, X.-L.; Wu, J.; Dai, L.-X.; Xia, L.-J.; Tang, M.-H.
Tetrahedron: Asymmetry 1998, 9, 1747.
(24) For example: (a) Chang, H.-T.; Sharpless, K. B. Tetrahedron Lett.
1996, 37, 3219. (b) Martinez, L. E.; Leighton, J. L.; Carsten, D. H.; Jacobsen,
E. N. J. Am. Chem. Soc. 1995, 117, 5897. (c) Reddy, K. C.; Sharpless, K.
B. J. Am. Chem. Soc. 1998, 120, 1207. (d) List, B.; Pojarliev, P.; Biller,
W. T.; Martin, H. J. J. Am. Chem. Soc. 2002, 124, 822. (e) Tramonti, M.
Synthesis 1982, 605. (f) Hoffman, R. V.; Maslouh, N.; Cervantes-Lee, F.
J. Org. Chem. 2002, 67, 1045.
3706
Org. Lett., Vol. 4, No. 21, 2002