A highly regioselective Friedl?nder reaction mediated by lanthanum chloride

Article abstract of DOI:10.1016/j.tetlet.2012.04.038

A highly efficient and regioselective Friedl?nder reaction of unsymmetrical 1,3-diketones with 2-aminoaryl aldehydes (ketones) is described. The methodology leads to the synthesis of a broad scope of substituted quinolines in high yield and excellent regioselectivity.

Full text of DOI:10.1016/j.tetlet.2012.04.038

Tetrahedron Letters 53 (2012) 3237–3241
Contents lists available at SciVerse ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
A highly regioselective Friedländer reaction mediated by lanthanum chloride
⇑
⇑
Ying Chen , Jinkun Huang , Tsang-Lin Hwang, T.J. Li, Sheng Cui, Johann Chan, Matthew Bio
Chemical Process Research & Development, Amgen Inc., One Amgen Center Dr., Thousand Oaks, CA 91320, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 16 February 2012
Revised 4 April 2012
Accepted 5 April 2012
Available online 25 April 2012
A highly efficient and regioselective Friedländer reaction of unsymmetrical 1,3-diketones with 2-aminoa-
ryl aldehydes (ketones) is described. The methodology leads to the synthesis of a broad scope of substituted
quinolines in high yield and excellent regioselectivity.
Ó 2012 Elsevier Ltd. All rights reserved.
Keywords:
Friedländer reaction
Regioselective
Quinolines
Quinoline and its derivatives are among the most important
synthetic or natural heterocycles with interesting biological activ-
ities showing anti-malarial, anti-inflammatory, anti-bacterial, and
anti-asthmatic properties.1 As a result, the quinoline scaffolds have
spurred the development of a great deal of synthetic methodolo-
gies. Among those variations, the Friedländer reaction2 is one of
the simplest, straightforward, and most widely used approaches.3
This classical method condenses 1 equiv of 2-aminoaromatic
carbonyl compound with 1 equiv of carbonyl species bearing a
reactive methylene group to provide the quinoline product and for-
mation of 2 equiv of water (Scheme 1). The reaction is typically
carried out in the presence of an acid or base, or otherwise harsh
conditions are required.
As part of our continuing efforts toward the development of
efficient syntheses of quinolines,4 we were interested in the
regio-selective synthesis of 2-substituted-3-acylquinolines. How-
ever, practical and general approaches to this type of substituted
quinolines are rare in the literature. Conceivably, the desired
substituted quinoline could be synthesized from Friedländer con-
densation as shown in Scheme 2. Recently Lewis acids such as
ZnCl2, FeCl3, Sc(OTf)3, and Nd(NO3)3ꢀ6H2O have been shown to be
effective on Friedländer reaction for the substituted quinoline syn-
thesis under mild conditions but without discussion of regioselec-
tivity issues.5 Although the attractive Friedländer reaction has the
potential to form highly substituted quinolines, limited attention
has been paid to specifically address the regio-selectivity issue of
the resulting products.6 In 2007, Masciadri and coworkers reported
a gold-catalyzed Friedländer condensation of 2-aminoaryl ketone
with 1,3-diones to form substituted quinolines.7 However the
reaction was not general and typically gave low to moderate yields
even under forcing conditions. To overcome the problems with the
current art such as harsh conditions, limited scope, and poor selec-
tivity, herein, we report our development of a mild, efficient, and
regio-selective synthesis of substituted quinolines via lanthanum
chloride—mediated Friedländer reaction.
Synthesis of 3-(2-methylquinolinyl) phenyl ketone 3a was our
initial target. As shown in Scheme 3, the Friedländer annulation
of 2-aminobenzaldehyde 1a with 1-phenylbutane 1,3-dione 2a
may provide the most straightforward synthesis of 3a. Under
typical Friedländer conditions in a variety of solvents and under
different temperatures, the best conditions (Table 1, entry 1) gave
modest yield (43%) and low selectivity (3a/4a = 64:36) of the de-
sired compound. We envisioned that Lewis acids, which have been
shown effective in many carbonyl-related organic transformations,
might lead to a different reaction profile in terms of reactivity and
selectivity.8 A broad scope of Lewis acids and conditions was
R2
N
O
R3
R4
R4
R2
NH2
acid or base
or heat
+
R3
R1
R1
O
Scheme 1. General scheme of Friedländer reaction.
O
CHO
NH2
Scheme 2. Retrosynthetic analysis of 2-substituted-3-acylquinoline.
O
O
R"
R
R
+
R"
R'
N
R'
⇑
Corresponding authors.
E-mail addresses: Ying@amgen.com (Y. Chen), jhuang@amgen.com (J. Huang).
0040-4039/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetlet.2012.04.038

3238
Y. Chen et al. / Tetrahedron Letters 53 (2012) 3237–3241
O
O
O
O
CHO
NH2
Me
+
+
Me
N
N
Me
1a
2a
3a, desired
4a, undesired
Scheme 3. Model reaction.
screened for the desired transformation. Based on the initial
results, we further narrowed the screening to the more promising
metal salts of the secondary group and lanthanum series in the
periodic table. Our initial attempt with Ca(OTf)2 improved the
selectivity from 64/36–78/22, however the yield was not satisfac-
tory (63%, Table 1, entry 2). To our delight, further screening indi-
cated that lanthanum chloride was among the best to increase the
reaction yield as well as the selectivity (Table 1, entry 14). It is well
known that lanthanum salts enhance the reactivity and selectivity
of many types of reactions, such as reduction, carbon–carbon bond
formation, aldol condensation, cycloaddition, ring-opening, and
polymerization.9 To the best of our knowledge, a lanthanum salt
mediated Friedländer has not been previously reported. In this
case, the readily available, inexpensive, and non-hygroscopic LaCl3
hydrate gave high selectivity and good yield of desired 3a (Table 1,
entry 15).10 Other Lewis acids tested either gave less selectivity or
much lower yield al beit higher selectivity in one case (Table 1,
entry 12). Alternatively, the basic conditions5a essentially gave no
product (Table 1, entry 16).
Table 1
Lewis acid screeninga
Entry
Lewis Acid
Yieldb (3a, %)
Ratio 3a/4ac
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16d
None
43
63
57
44
52
31
52
66
44
61
80
30
64
60
85
Trace
64/36
78/22
76/24
72/28
64/36
76/24
73/27
79/21
81/19
83/17
79/21
89/11
85/15
86/14
86/14
n/d
Ca(OTf)2
Sm(OTf)3
Dy(OTf)3
Ba(OTf)2
Er(OTf)3
La(OTf)3
Yb(OTf)3
DyCl3
ZnCl2
ErCl3
CoCl2
CeCl3
LaCl3
LaCl3ꢀ7H2O
None
a
Reaction conditions: 1a (0.2 mmol), 2a (0.24 mmol), and additive (0.2 mmol)
were mixed in 0.4 mL acetic acid and heated to 60 °C for 18 h. The resulting mixture
The reaction was further optimized by examining the amount of
LaCl3,11 solvents,12 and reaction temperature.13 The optimal reac-
tion was performed in acetic acid in the presence of lanthanum
chloride heptahydrate (1.0 equiv) at 60 oC for 3–5 h.14
was analyzed by HPLC.
b
Yield was determined by HPLC with standard calibration of the pure sample.
The ratio was determined by HPLC using both isomers as standard.
Reaction conditions: 1a (0.2 mmol), 2a (0.24 mmol), and KOH (0.24 mmol) were
c
d
With the optimal conditions in hand, we extended the reaction
scope to a variety of unsymmetrical 1,3-diketones with different
mixed in 0.4 mL ethanol and heated to 60 °C for 26 h. The resulting mixture was
analyzed by HPLC.
Table 2
Scope of 2-aminobenzaldehyde (or ketone)
R2
N
O
R2
O
O
R3
R4
R3
R4
R2
LaCl3, HOAc
R1
R1
+
R1
+
R4
R3
N
O
O
NH2
Entry
1
Aniline
1,3-Diketone
Major isomer
Yielda (%)
Selectivity major/minorb
86/14
CHO
NH2
Me
Ph
Ph
Ph
O
Ph
O
O
O
O
O
O
72
70
68
N
3a
Me
2a
2a
2a
1a
CHO
Me
Me
Me
O
Ph
2
3
86/14
91/9
F
NH2
F
N
Me
1b
3b
Me
O
O
Ph
Me
N
3c
Me
NH2
Br
1c
Ph
Br
O
CHO
NH2
O
O
Ph
Me
4
62
93/7
Br
N
Br
2a
3d
1d

Y. Chen et al. / Tetrahedron Letters 53 (2012) 3237–3241
3239
Table 2 (continued)
Entry
Aniline
1,3-Diketone
Major isomer
Yielda (%)
Selectivity major/minorb
CHO
NH2
Me
Ph
Ph
O
O
O
O
O
Ph
5
90
94/6
N
N
N
3e
Me
2a
2a
1e
CHO
NH2
Me
O
Ph
6
69
100/0
MeO
MeO
N
3f
Me
1f
a
Yield of isolated major isomer.
Ratio determined by NMR of the crude mixture after workup.
b
types of 2-aminobenzaldehydes (or ketone). As shown in Table 2,
the reaction of 1-phenylbutane 1,3-dione with several substituted
2-aminobenzaldehydes (Table 2, entries 2 and 4) worked as well as
the parent substrate (Table 2, entry 1). The ketone was found to be
an excellent substrate in this condensation to give 2,3,4-substitued
quinoline in appreciable yield (Table 2, entry 3). The 2-aminonico-
tinaldehyde showed superior selectivity (Table 2, entry 5) to give
the corresponding azaquinoline in an excellent isolated yield of
Table 3
Electronic effect of 1,3-diketones
R2
N
O
R2
O
O
R2
R3
R4
R3
R4
LaCl3, HOAc
R1
R1
R1
+
+
R4
R3
NH2
N
O
O
Entry
1
Aniline
1,3-Diketone
Major isomer
Yielda (%)
Selectivity major/minorb
100/0
CHO
NH2
O
O
O
O
O
F3C
F3C
Ph
81
83
80
77
N
3g
CF3
O
1a
2g
2h
CHO
NH2
S
S
2
3
100/0
89/11
90/10
N
CF3
1a
3h
CHO
NH2
O
O
O
O
O
F3C
F3C
F3C
CH3
CH3
Me
CF3
F
2i
F
N
3i
1b
CHO
O
O
Me
30
NH2
2i
N
3i'
CF3
O
1a
1a
CHO
NH2
O
O
4
5
6
85
78
85
100/0
100/0
96/4
N
CF3
2j
3j
CHO
NH2
O
O
F3C
O
tBu
N
3k
CF3
O
1a
1a
2k
CHO
NH2
O
O
Me
OMe
O
N
2l
3l
O
a
Yield of isolated major isomer.
Ratio determined by NMR of the crude mixture after workup.
b

3240
Y. Chen et al. / Tetrahedron Letters 53 (2012) 3237–3241
3i
4i
Figure 1. The X-ray structure of 3i and 4i.
Table 4
Steric effect of 1,3-diketones
R2
O
R2
O
O
R3
R2
R4
R3
R4
R1
R1
+
R1
LaCl3, HOAc
+
R3
N
R4
NH2
N
O
O
Entry
1
Aldehyde
1,3-Diketone
Major isomer
Yielda (%)
Selectivity major/minorb
CHO
NH2
O
O
O
O
O
Me
85
81
72
85
96/4
N
3m
Me
O
1a
1a
2m
CHO
NH2
O
O
tBu
2
3
4
100/0
80/20
93/7
N
Me
2n
3n
CHO
NH2
O
O
O
tBu
N
3o
iPr
1a
1a
2o
CHO
NH2
O
O
O
iPr
Me
2p
N
3p
a
Yield of isolated major isomer.
Ratio determined by NMR of the crude mixture after workup.
b
the product. Electron-rich aminoaldehyde provided exclusively
one isomer (Table 2, entry 6). The major products consistently
showed the same substitution pattern as their parent compound
in all cases.
When trifluoromethyl group is replaced by carboxylic methyl
ester, another electron withdrawing group, the regioselectivity
reached 96/4 and the same selectivity trend holds true that the
electronic withdrawing group (ester) is attached to the quinoline
ring (Table 3, entry 6).
As electronic effect of the unsymmetrical 1,3-diketones was
established in terms of regioselectivity, the steric effect was also
examined. When the sterically unsymmetrical 1,3-diketones were
used, in all the cases examined, the less hindered group tends to stay
on the aromatic quinoline ring as the major products (Table 4). The
larger the steric difference the better the regioselectivity is shown as
comparing entry 2 (Me to t-butyl) versus entry 4 (i-Pr to t-Bu). When
less steric difference is present, the selectivity declines (Table 4,
entry 4).
Two different pathways are proposed for Friedländer reaction as
generally recognized mechanisms. The first step is either an inter-
molecular Schiff base formation17 or aldol condensation.18 Both
intermediates can then go through a ring-closing process to form
the quinoline products. The origin of the regioselectivities of the
current lanthanum chloride mediated Friedländer condensation
Next, we turned our attention to the scope of unsymmetrical
1,3-diketones. Due to steric and electronic differences of the two
ending substitution groups on 1,3-diketones, the product outcome
would be difficult to predict. To differentiate the two carbonyl
groups electronically, 1,3-diketone with an electron withdrawing
group such as trifluoromethyl (CF3) at one side was tested
(Table 3). It is shown that the resulting major products share the
same feature with the CF3 directly attached to the quinoline ring
(Table 3, entries 1–5).15 It is clear that the electronic difference
dominates the outcome of the selectivity as most of the reactions
exclusively give one product (Table 3, entries 1, 2, 4, and 5). Even
for the sterically similar CF3 versus CH3, the regioisomeric ratio is
still as high as 89:1116 and 90:10 (Table 3, entries 3 and 30).
Furthermore, in the case of entry 3, both of the regioisomers were
isolated and structurally confirmed by NMR and X-ray diffraction
analyses (Fig. 1, 3i, and 4i).

Y. Chen et al. / Tetrahedron Letters 53 (2012) 3237–3241
6. Muchowski, J. M.; Maddox, M. L. Can. J. Chem. 2004, 82, 461.
3241
is not well understood. One rationale is that lanthanum chloride
facilitated the Schiff base formation as the first step,19 and the
resulting Schiff base(s) dictate the outcome of the regioselectivity.
Although the exact role of lanthanum chloride is not clear, this pro-
posal explains both the electronic and steric effects of the 1,3-dike-
tones on the selectivities: the nucleophilic amine favors to attack
either the electron-poor or the sterically less hindered carbonyl
group of the 1,3-diketone to form the corresponding enaminone
as the major intermediate.20
In conclusion, we have developed a regioselective and efficient
protocol for substituted quinolines based on the lanthanum chlo-
ride mediated Friedländer reaction. The Friedländer reaction
between an unsymmetrical 1,3-diketone with the corresponding
2-carbonyl aniline gives high regioselective quinoline in moderate
to excellent yields.21 The protocol presented here is effective for a
broad range of both substituted 2-carbonyl aniline and 1,3-
diketones.
7. Atechian, S.; Nock, N.; Norcross, R. D.; Ratni, A. W.; Verron, J.; Masciadri, R.
Tetrahedron 2007, 63, 2811.
8. Lewis Acids in Organic Synthesis, Yamamoto, H. Eds.; Wiley-VCH: Weinheim;
Vol. 1 and 2.
9. Lu, J.; Bai, Y.; Wang, Z.; Yang, B.; Ma, H. Tetrahedron Lett. 2000, 41, 9075.
10. Seyedi, N.; Saidi, K.; Khabazzadeh, H. Synth. Commun. 1864, 2009, 39.
11. Substoichiometric amount of LaCl3 is effective to give comparable selectivity
but leads to a slower reaction rate and a lower yield.
12. Other solvents tested such as THF, toluene, and acetonitrile were less effective
as acetic acid.
13. The reaction can be run at room temperature but a longer reaction time is
required.
14. A typical experimental procedure. To a reaction vial loaded with substituted 1,3-
diketone (0.24 mmol), amino benzaldehyde, (0.2 mmol) and lanthanum
chloride heptahydrate was added acetic acid (3.5 mL). The resulting
suspension was heated to 60 °C for 3–5 h. The reaction progress was
monitored by LC–MS. At the completion, the reaction mixture was allowed
to cool to room temperature. Ethyl acetate (5 mL) was added and the resulting
mixture was washed with water (5 mL) and 1 N NaOH (5 mL), the organic layer
was dried on MgSO4 and concentrated to give a crude product. The major
quinoline product was isolated by column chromatography.
15. Interestingly, the selectivity observed is opposite to that reported for the gold-
catalyzed Friedländer reaction results wherein the CF3 was found exclusively
on the carbonyl group in a slightly different setting (Ref. 6).
Acknowledgment
16. When no LaCl3 was used, the reaction gave an overall yield of 52% (3h/4h = 54/
46). The typical KOH/EtOH conditions gave only a trace amount of product as in
the case of. Table 1, entry 16.
17. (a) Schofield, K.; Theobald, R. S. J. Chem. Soc. 1950, 395; (b) Fehnel, E. A.;
Deyrup, J. A.; Davidson, M. B. J. Org. Chem. 1996, 1958, 23; (c) Tamura, Y.;
Tsugoshi, T.; Mohri, S.-I.; Kita, Y. J. Org. Chem. 1985, 50, 1542.
18. Majewicz, T. J.; Caluwe, P. J. Org. Chem. 1975, 40, 3407.
We thank Ms. Maosheng Chen (HPLC), Mr. Mike Ronk (HRMS),
and Dr. Richard Staples (X-ray) for their analytical support.
References and notes
19. Lanthanum chloride catalyzed enaminone formation was recently reported
which may indirectly support this rationale: Lenin, R.; Raju, M. ARKIVOC 2007,
viii, 204.
1. (a) Michael, J. P. Nat. Prod. Rep. 2002, 19, 742; (b) Jones, G. In Comprehensive
Heterocyclic Chemistry II; Katritzky, A. R., Rees, C. W., Scriven, E. F. V., Eds.;
Pergamon: New York, 1996; Vol. 5, pp 167–243; (c) Egan, T. J. Quinoline
Antimalarials 2001, 11, 185.
20. Preliminary 1H and 19F NMR studies showed the imine formation when aniline
and 1-trifluromethylbutane 1,3-dione was mixed with LaCl3ꢀ7H2O in CD3CO2D,
although the ratio of the isomeric imines could not be determined. However,
the ultimate mechanism is yet to be probed in more details.
21. The current protocol of Friedländer reaction was optimized and scaled up to
synthesize a building block for our internal program.
2. Friedländer, P. Ber. Dtsch. Chem. Ges. 1882, 15, 2572.
3. (a) Marco-Contelles, J.; Perez-Mayoral, E.; Samadi, A.; Carreiras, M. C.; Soriano,
E. Chem. Rev. 2009, 109, 2652; (b) Martinez, R.; Ramon, D. J.; Yus, M.; Martinez,
R.; Ramin, J. R.; Yus, M. J. Org. Chem. 2008, 73, 9778.
4. Huang, J.; Bunel, E.; Faul, M. M. Org. Lett. 2008, 10, 2251.
5. (a) Hu, Y.; Zhang, Z. G.; Thummel, R. P. Org. Lett. 2003, 5, 2251; (b) Varala, R.;
Enugala, R.; Adapa, R. S. Synthesis 2006, 22, 3825–3830; (c) Bose, D. S.; Kumar,
R. K. Tetrahedron Lett. 2006, 47, 813.