A copper-catalyzed petasis reaction for the synthesis of tertiary amines and amino esters

Article abstract of DOI:10.1021/jo3003503

We have developed a copper-catalyzed process for the coupling of aldehydes, amines, and boronic acids. This allows greater reactivity with simple aryl boronic acids and allows coupling reactions to proceed that previously failed. Initial mechanistic studi

Full text of DOI:10.1021/jo3003503

Note
pubs.acs.org/joc
A Copper-Catalyzed Petasis Reaction for the Synthesis of Tertiary
Amines and Amino Esters
Robin Frauenlob, Carlos García, Gary A. Bradshaw, Helen M. Burke, and Enda Bergin*
School of Chemistry, Trinity Biomedical Sciences Institute, University of Dublin, Trinity College, Dublin 2, Ireland
S
* Supporting Information
ABSTRACT: We have developed a copper-catalyzed process
for the coupling of aldehydes, amines, and boronic acids. This
allows greater reactivity with simple aryl boronic acids and
allows coupling reactions to proceed that previously failed.
Initial mechanistic studies support a process involving
transmetalation from boron to copper.
he three component coupling of an amine, aldehyde, and
boronic acid (known as the Petasis reaction) is a
convenient route to the synthesis of highly functionalized
tertiary amines.1 In general, the reaction tends to proceed
exceptionally well when there is a free hydroxy group on the
aldehyde fragment, allowing the formation of an activated
boronate salt (see Scheme 1).2 Other heteroatoms may also
initial hope was to form a more reactive aryl palladium species
capable of arylating the transient iminium intermediate, in a
similar manner to the arylation of N-tosyl imines by boronic
acids and palladium (see Scheme 2).13−17 Unfortunately, these
T
Scheme 2. Initial Attempts at Catalysis
Scheme 1. Expansion of Scope of Petasis Reaction
attempts were uniformly unsuccessful, and no product was
observed. We next attempted to use copper(I) bromide. This
had previously proven effective for the addition of terminal
alkynes to iminium salts18 as well as activating boronic acids in
the Suzuki coupling.19 To our delight, this allowed the reaction
to proceed, albeit with a 9% yield. To ensure this was due to the
presence of copper, we repeated the reaction with no catalyst,
and again no product was observed.
With this result in hand, we set about trying to increase the
yield of amino ester obtained. A solvent screen indicated that
DMF was the optimal solvent out of those tested, giving an
improved yield of 48%. This increased to 63% upon increasing
the temperature to 70 °C, but no product was observed at 100
°C (possibly because of the volatility of the amine employed).
Next, the effect of the copper salt was examined (see Table
1). Copper(I) halides all worked in roughly comparable yields,
although copper(I) acetate gave a dramatically reduced amount
of product. Copper(II) salts were in general ineffective, and no
trace of product was observed. The only exception to this was
copper(II) acetate, which to our surprise gave the desired
compound in 50% yield. The reason for this discrepancy is
currently being investigated. Finally, in order to test the
sensitivity of the reaction to oxygen and moisture, we ran it
fulfill this role, but reactivity is invariably reduced; in general
without a hydroxy group on the aldehyde, unactivated aryl
boronic acids give little or no product, and the more reactive
vinyl boronic acids, or trifluoroalkenyl boronates activated with
Lewis acids, are required.2−4
Some catalytic variants of the Petasis reaction have been
developed. Notably an enantioselective organocatalytic variant
was reported by Schaus and co-workers, using allyl boronate
esters and chiral diols5,6 and another organocatalytic system by
Yuan and co-workers using salicylaldehydes.7 Catalytic variants
have also been developed for Petasis-type reactions such as the
addition of organoboron compounds to N-acyl quinolinium
ions,8−10 N,O-aminals,11 and very recently the three component
coupling of imines, acid chlorides, and tetraaryl boron
compounds, catalyzed by copper salts and a Lewis base.12
In an attempt to increase the reactivity of aryl boronic acids,
we investigated the use of palladium salts as catalysts. Our
Received: February 17, 2012
Published: April 11, 2012
© 2012 American Chemical Society
4445
dx.doi.org/10.1021/jo3003503 | J. Org. Chem. 2012, 77, 4445−4449

The Journal of Organic Chemistry
Note
Next, we examined the generality of the reaction by altering
the amine, boronic acid, and aldehyde (see Table 2). First, after
observing the need to maintain an oxygen-free atmosphere (see
above), we noticed that yield increased dramatically after
degassing the DMF prior to use and finely powdering the
molecular sieves. With regard to the amine component,
pyrrolidine worked better than piperidine or diethylamine,
presumably because of its smaller steric profile. The reaction
could tolerate both electron-rich and electron-poor boronic
acids, with perhaps a slight increase in yields observed with
more electron poor substrates (Table 2, entry 1 vs 2, entry 3 vs
4, entry 7 vs 8). In common with all Petasis couplings, it was
necessary to use aldehydes with coordinating groups. In our
case, pyridine-2-carboxaldehyde gave higher yields than ethyl
glyoxalate (Table 2, entry 3 vs entry 7). With regard to sterics,
while heterocylic and secondary amines with primary alkyl
groups gave product, changing to dicyclohexylamine shut down
the reaction (Table 2, entry 6), as did use of ortho-substituents
on the aryl boronic acid (Table 2, entry 9) (of course, these
reactions also failed to provide any product in the uncatalyzed
reaction). Finally, to confirm that some coordinating group was
necessary on the aldehyde, we repeated the reaction using 2-
nitrobenzaldehyde (Table 2, entry 10). Because this would be
approximately as electron-poor (and hence electrophilic) as
Table 1. Effect of Various Metal Salts on Reaction Yield
a
entry
metal salt
yield (%)
1
2
3
4
5
6
7
8
9
CuBr
63
54
64
4
CuCl
CuI
CuOAc
CuBr2
CuCl2
CuSO4
Cu(OTf)2
Cu(OAc)2
CuBr
0
0
0
0
50
42
b
10
a
Reaction times 24 h. Determined by 1H NMR spectroscopy.
b
Reaction run under a noninert atmosphere.
under a noninert atmosphere. In this case, we observed a
significant drop in yield (Table 1, entry 10), indicating the
importance of excluding oxygen and moisture.
Table 2. Substrate Scope
a
Reaction times 24 h. DMF degassed prior to use.
4446
dx.doi.org/10.1021/jo3003503 | J. Org. Chem. 2012, 77, 4445−4449

The Journal of Organic Chemistry
Note
Figure 1. Proposed mechanism for copper catalysis.
Figure 2. 11B NMR analysis of phenyl boronic acid with copper salts in DMF.
pyridine-2-carboxaldehyde, but without any suitable site for
coordination, it would allow us to examine these two
requirements separately. As expected, no product was formed,
indicating the importance of some coordinating functionality
on the aldehyde component.
These results led us to tentatively assign a mechanism for the
process (see Figure 1). We postulated the initial formation of a
boronate salt through the interaction of the boronic acid 8 with
the counterion of the copper salt. This activated species (12)
could then transmetallate copper, leading to the formation of
an organocuprate (14) along with byproduct 13. Carbon−
carbon bond formation could then occur through nucleophilic
attack at any iminium ions in solution. The requirement of a
coordinating atom on the aldehyde fragment would seem to
imply that there is a coordination event, activating the cuprate
and bringing it in close proximity to the electrophilic center.
This then allows the regeneration of the copper salt.
If the above mechanism is broadly correct, it would be
expected to see three major boron-containing compounds in
solution: 11, 12, and 13. To examine this, we monitored a 1:1
ratio of bpy/CuBr with phenyl boronic acid (see Figure 2).
Pleasingly, we observed 3 main peaks by 11B NMR, with shifts
corresponding to those expected for compounds 11, 12, and 13
(on the basis of literature values for related compounds20).
Phenyl boronic acid appears at 29.3 ppm. Upon addition of
CuBr, we observe first the formation of a boron-containing
species at 2.0 ppm. We postulate that this is the boronate salt
12 (where X is bromide or hydroxide). The shift broadly
4447
dx.doi.org/10.1021/jo3003503 | J. Org. Chem. 2012, 77, 4445−4449

The Journal of Organic Chemistry
Note
Compound 2. Isolated as a yellow oil after column chromatography
(CH2Cl2/EtOAc 98:2): H NMR (400 Hz, CDCl3) δ 7.40 (dd, 1H, J
corresponds to that observed for the related reaction with
alkenyl boronate esters20 (to the best of our knowledge, there
have been no previously reported shifts for such compounds
derived from aryl boronic acids). As the reaction progresses, a
third species appears, which we assigned as 13. The
concentration of this side product should correspond to the
total amount of aryl copper species 14 produced. Within 4 h,
the starting phenyl boronic acid is completely consumed, and
13 is the major species observed, presumably in equilibrium
with salt 12.
In order to confirm the identity of 12, we heated phenyl
boronic acid with NaOAc in DMF. This would be expected to
show only two peaks by 11B NMR: boronic acid 11 and a
boronate salt 12 (transmetalation, and hence the formation of
13, would not be possible without copper). Again, this is what
was observed: some phenyl boronic acid was consumed, but
this then stayed constant throughout the duration of the
experiment (4 h). The only other boron-containing species in
solution was salt 12.
1
= 1.9, 0.9 Hz), 6.41−6.33 (m, 2H), 4.31−4.17 (m, 3H), 2.74−2.47 (m,
4H), 1.89−1.74 (m, 4H), 1.27 (t, 3H, J = 7.1 Hz); 13C NMR (100 Hz,
CDCl3) δ 169.6 (q), 150.0 (q), 142.6, 110.3, 109.0, 65.4 (CH), 61.3,
51.6, 23.4, 14.2 (CH3); IR (NaCl disk) νmax/cm−1 2966, 2809, 1736 (s,
CO), 1150, 1012; HRMS calcd for C12H18NO3 [M + H]+ 224.1287,
found 224.1285.
Compound 3.23 Isolated as an orange oil after column
1
chromatography (CH2Cl2/EtOAc 92:8): H NMR (400 Hz, CDCl3)
δ 7.50−7.43 (m, 2H), 7.40−7.31 (m, 3H), 4.28−4.09 (m, 2H), 3.98 (s,
1H), 2.47−2.37 (m, 4H), 1.67−1.56 (m, 4H), 1.52−1.41 (m, 2H),
1.23 (t, 3H, J = 7.1 Hz); 13C NMR (100 Hz, CDCl3) δ 171.9 (q),
136.4 (q), 128.8, 128.4, 128.1, 75.0, 60.7, 52.4, 25.8, 24.4, 14.2; IR
(NaCl disk) νmax/cm−1 2933, 2853, 2300, 1741 (s, CO), 1261,
1155, 1122, 801, 696; HRMS calcd for C15H22NO2 [M + H]+
248.1651, found 248.1660.
Compound 4. Isolated as a yellow oil after column chromatography
(CH2Cl2/EtOAc 90:10): 1H NMR (400 Hz, CDCl3) δ 7.51−7.37 (m,
2H), 7.12−6.96 (m, 2H), 4.30−4.06 (m, 2H), 3.94 (s, 1H), 2.46−2.30
(m, 4H), 1.69−1.53 (m, 4H), 1.52−1.41 (m, 2H), 1.23 (t, 3H, J = 7.1
Hz); 13C NMR (100 Hz, CDCl3) δ 171.7 (q), 162.6 (d, J = 246.4 Hz,
q), 132.2 (q), 130.4 (d, J = 8.0 Hz), 115.3 (d, J = 21.5 Hz), 74.2, 60.8
(CH2), 52.3 (CH2), 25.8 (CH2), 24.3 (CH2), 14.2; IR (NaCl disk)
νmax/cm−1 2934, 1733 (s, CO), 1604, 1507, 1223, 1150, 805;
HRMS calcd for C15H21FNO2 [M + H]+ 266.1556, found 266.1566.
Compound 5. Isolated as an amorphous green solid after column
To show that our active nucleophile was generated by the
interaction of copper salt with the boronic acid, we repeated the
experiment, again monitoring the process by 11B NMR. Once
the phenyl boronic acid was completely consumed, we added
the solution to a mixture of ethyl glyoxalate and piperidine
under our standard conditions: the desired compound was
formed rapidly in essentially quantitative yield.21
1
chromatography (hexane/EtOAc/Et3N 94:1:5): H NMR (400 Hz,
CDCl3) δ 8.52 (d, 1H, J = 4.8 Hz), 7.68−7.58 (m, 2H), 7.53 (d, 2H, J
= 7.5 Hz), 7.34−7.27 (m, 2H), 7.21 (apt t, 1H), 7.15−7.05 (m, 1H),
4.90 (s, 1H), 2.63 (q, 4H, J = 7.0 Hz), 1.01 (t, 6H, J = 7.0 Hz); 13C
NMR (100 Hz, CDCl3) δ 163.2 (q), 149.0, 142.1 (q), 136.4, 128.4,
128.3, 127.0, 121.8, 73.6, 42.8 (CH2), 10.6; IR (NaCl disk) νmax/cm−1
2969, 2933, 1588, 1432, 746, 730, 699; HRMS calcd for C16H21N2 [M
+ H]+ 241.1705, found 241.1712.
Finally, to examine the cause for a complete lack of reactivity
with a number of Cu(II) salts, we monitored a solution
containing CuBr2 and phenyl boronic acid. No change was
observed by NMR, indicating that in these cases, it is the
absence of any interaction between the boronic acid and the
metal salt that results in no formation of product.
In conclusion, we have developed a copper-catalyzed process
for the coupling of an amine, aldehyde, and boronic acid. This
allows a much greater flexibility in the choice of reagents in the
Petasis reaction, thereby removing a significant limitation with
the original process. The catalytic system is convenient and
inexpensive, and a likely mechanism of action has been
described.
Compound 7. Isolated as an amorphous green solid after column
1
chromatography (hexane/EtOAc/Et3N 80:15:5): H NMR (400 Hz,
CDCl3) δ 8.51 (d, 1H, J = 5.1 Hz), 7.70−7.58 (m, 2H), 7.50 (d, 2H, J
= 7.3 Hz), 7.36−7.26 (m, 2H), 7.25−7.17 (m, 1H), 7.15−7.05 (m,
1H), 4.42 (s, 1H), 2.61−2.26 (m, 4H), 1.70−1.55 (m, 4H), 1.53−1.40
(m, 2H); 13C NMR (100 Hz, CDCl3) δ 162.9 (q), 149.0, 141.8 (q),
136.5, 128.4, 128.3, 127.0, 122.1, 121.8, 78.7, 53.3 (CH2), 26.1 (CH2),
24.6 (CH2); IR (NaCl disk) νmax/cm−1 2931, 2753, 1587, 1431, 746,
699; HRMS calcd for C17H21N2 [M + H]+ 253.1705, found 253.1713.
Compound 8. Isolated as an amorphous green solid after column
EXPERIMENTAL SECTION
■
1
General Methods. Ethyl glyoxalate was purchased as a 50%
solution in toluene and freshly distilled before use. DMF was dried
over calcium hydride and degassed by successive freeze−pump−thaw
cycles. Molecular sieves were powdered and activated by heating under
vacuum prior to use. All reactions were carried out in oven-dried
glassware.
chromatography (hexane/EtOAc/Et3N 80:15:5): H NMR (400 Hz,
CDCl3) δ 8.48 (d, 1H, J = 4.6 Hz), 7.65−7.52 (m, 2H), 7.35 (d, 2H, J
= 7.7 Hz), 7.15−7.00 (m, 3H), 4.34 (s, 1H), 2.56−2.10 (m, 7H),
1.63−1.53 (m, 4H), 1.48−1.40 (m, 2H); 13C NMR (100 Hz, CDCl3)
δ 163.1 (q), 149.0, 138.8 (q), 136.6 (q), 136.5, 129.1, 128.2, 122.0,
121.7, 78.4, 53.3 (CH2), 26.1 (CH2), 24.6 (CH2), 21.1; IR (NaCl
disk) νmax/cm−1 2930, 2855, 1587, 1432, 797, 753; HRMS calcd for
C18H23N2 [M + H]+ 267.1861, found 267.1871.
General Procedure for the Catalyzed Petasis Reaction. Dry,
degassed DMF (12 mL) was added to a mixture of copper salt (0.142
mmol) and 2,2′-bipyridine (26.4 mg, 0.17 mmol) under nitrogen, and
the solution was stirred at 60 °C for 1 h. After this time, aldehyde
(1.42 mmol), amine (1.42 mmol), and boronic acid (2.84 mmol) were
charged to the flask along with powdered 3 Å molecular sieves. The
solution was stirred at 70 °C for 24 h and subsequently filtered on a
short silica pad. After evaporation of the solvent, the residue was
purified by column chromatography on silica gel.
ASSOCIATED CONTENT
■
S
* Supporting Information
NMR spectra for all compounds. This material is available free
of charge via the Internet at http://pubs.acs.org.
Compound 1.22 Isolated as a yellow oil after column chromatog-
raphy (CH2Cl2/EtOAc/MeOH 90:9:1): 1H NMR (400 Hz, CDCl3) δ
7.50 (d, 2H, J = 7.6 Hz), 7.40−7.30 (m, 3H), 4.27−4.08 (m, 2H), 3.93
(s, 1H), 2.68−2.54 (m, 2H), 2.52−2.40 (m, 2H), 1.91−1.72 (m, 4H),
1.22 (t, 3H, J = 7.1 Hz); 13C NMR (100 Hz, CDCl3) δ 171.8 (q),
137.5 (q), 128.5, 128.4, 128.2, 74.1, 60.9, 52.5, 23.3, 14.1; IR (NaCl
disk) νmax/cm−1 2968, 2791, 1743 (s, CO), 1153, 1024; HRMS
calcd for C14H20NO2 [M + H]+ 234.1494, found 234.1496.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: ebergin@tcd.ie.
Notes
The authors declare no competing financial interest.
4448
dx.doi.org/10.1021/jo3003503 | J. Org. Chem. 2012, 77, 4445−4449

The Journal of Organic Chemistry
Note
ACKNOWLEDGMENTS
■
Financial support from Trinity College and from the Science
Foundation Ireland SURE program (HMB) is gratefully
acknowledged.
REFERENCES
■
446.
(1) Petasis, N. A.; Zavialov, I. A. J. Am. Chem. Soc. 1997, 119, 445−
(2) Candeias, N. R.; Montalbano, F.; Cal, P. M. S. D.; Gois, P. M. P.
Chem. Rev. 2010, 110, 6169−6193.
(3) Schlienger, N.; Bryce, M. R.; Hansen, T. K. Tetrahedron Lett.
2000, 41, 1303−1305.
(4) Tremblay-Morin, J.-P.; Raeppel, S.; Gaudette, F. Tetrahedron Lett.
2004, 3471−3474.
(5) Lou, S.; Schaus, S. E. J. Am. Chem. Soc. 2008, 130, 6922−6923.
(6) Muncipinto, G.; Moquist, P. N.; Schreiber, S. L.; Schaus, S. E.
Angew. Chem., Int. Ed. 2011, 50, 8172−8175.
(7) Han, W.-Y.; Wu, Z.-J.; Zhang, X.-M.; Yuan, W.-C. Org. Lett. 2012,
14, 976−979.
(8) Yamaoka, Y.; Miyabe, H.; Takemoto, Y. J. Am. Chem. Soc. 2007,
129, 6686−6687.
(9) Graham, T. J. A.; Shields, J. D.; Doyle, A. G. Chem. Sci. 2011, 2,
980−984.
(10) Kodama, T.; Moquist, P. N.; Schaus, S. E. Org. Lett. 2011, 13,
6316−6319.
(11) Huang, Y.-Y.; Chakrabarti, A.; Morita, N.; Schneider, U.;
Kobayashi, S. Angew. Chem., Int. Ed. 2011, 50, 11121−11124.
(12) Morin, M. S. T.; Lu, Y.; Black, D. A.; Arndtsen, B. A. J. Org.
Chem. 2012, 77, 2013−2017.
(13) Dai, H.; Yang, M.; Lu, X. Adv. Synth. Catal. 2008, 350, 249−253.
(14) Zhang, Q.; Chen, J.; Liu, M.; Wu, H.; Cheng, J.; Qin, C.; Su, W.;
Ding, J. Synlett 2008, 935−939.
(15) Huixiong, D.; Lu, X. Tetrahedron Lett. 2009, 50, 3478−3481.
(16) Marques, C. S.; Burke, A. J. Eur. J. Org. Chem. 2010, 1639−1643.
(17) Marques, C. S.; Burke, A. J. ChemCatChem 2011, 3, 635−645.
(18) Gommermann, N.; Koradin, C.; Polborn, K.; Knochel, P. Angew.
Chem., Int. Ed. 2003, 43, 5763−5766.
(19) Deng, J. Z.; Paone, D. V.; Ginnetti, A. T.; Kurihara, H.; Dreher,
S. D.; Weissman, S. A.; Stauffer, S.; Burgey, C. S. Org. Lett. 2009, 11,
345−347.
(20) Wada, R.; Shibuguchi, T.; Makino, S.; Oisaki, K.; Kanai, M.;
Shibasaki, M. J. Am. Chem. Soc. 2006, 128, 7687−7691.
(21) This yield is higher than those observed in the catalytic
reactions, but it should be noticed that since it is a 1:1 mixture of
phenyl boronic acid with CuBr/bpy, in this case the copper is present
in stoichiometric quantities.
(22) Deng, Q.-H.; Xu, H.-W.; Yuen, A. W.-H.; Xu, Z.-J.; Che, C.-M.
Org. Lett. 2008, 10, 1529−1532.
(23) Haurena, C.; Le Gall, E.; Sengmany, S.; Martens, T.; Troupel,
M. J. Org. Chem. 2010, 75, 2645−2650.
4449
dx.doi.org/10.1021/jo3003503 | J. Org. Chem. 2012, 77, 4445−4449