A ruthenium-catalyzed approach to the friedlaender quinoline synthesis

Article abstract of DOI:10.1002/ejoc.200701001

In a modification of the Friedlaender reaction, 2-aminobenzyl alcohol is oxidatively cyclized with a variety of ketones to yield substituted quinolines. Of all the ruthenium catalysts that were tested for this reaction, the second-generation Grubbs' catalyst gives the highest quinoline yield, in combination with KOtBu as a base. The presence of a hydrogen acceptor is required to regenerate the catalyst. Also the reaction mechanism is discussed, and the results show that there are possibly two different pathways towards quinolines. Wiley-VCH Verlag GmbH & Co. KGaA, 2008.

Full text of DOI:10.1002/ejoc.200701001

FULL PAPER
DOI: 10.1002/ejoc.200701001
A Ruthenium-Catalyzed Approach to the Friedländer Quinoline Synthesis
Hans Vander Mierde,[a] Pascal Van Der Voort,[a] Dirk De Vos,[b] and Francis Verpoort*[a]
Keywords: Nitrogen heterocycles / Ruthenium / Oxidative cyclization / Friedländer reaction
In a modification of the Friedländer reaction, 2-aminobenzyl
alcohol is oxidatively cyclized with a variety of ketones to
yield substituted quinolines. Of all the ruthenium catalysts
that were tested for this reaction, the second-generation
acceptor is required to regenerate the catalyst. Also the reac-
tion mechanism is discussed, and the results show that there
are possibly two different pathways towards quinolines.
Grubbs’ catalyst gives the highest quinoline yield, in combi- (© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
nation with KOtBu as a base. The presence of a hydrogen
Germany, 2008)
mediated quinoline synthesis. A modified Friedländer pro-
tocol (Scheme 1) has been developed by Cho and co-
workers.[25]
Introduction
The synthesis of nitrogen-containing heterocycles, such
as quinoline, is the subject of extensive research in organic
chemistry, because the quinoline scaffold is present in many
biologically active compounds. Applications of quinolines
in medicinal chemistry include the use as antimalarial,[1,2]
antiinflammatory,[3] antiasthmatic,[4] antibacterial,[5] anti-
hypertensive[6] and tyrosine kinase inhibitory agents.[7]
Quinoline based polymers are currently investigated for ap-
plications as thermally stable transparent materials in the
fields of electronics, optoelectronics and nonlinear op-
tics.[8,9]
Many traditional methods, such as the Skraup, Doebner–
von Miller, Conrad–Limpach, and Pfitzinger syntheses, suf-
fer from harsh reaction conditions, low stereoselectivity or
consist of multiple steps, resulting in low overall yields, lim-
iting their applicability.[1] The Friendländer method is gen-
erally considered to be the most versatile method of synthe-
sis although its full potential is inhibited due to the use of
unstable aminobenzaldehydes.
Besides these conventional named reactions, several or-
ganometal-catalyzed approaches have recently been devel-
oped for the synthesis of the quinoline nucleus. Ruthenium
complexes catalyze the reaction of aniline with allyl
alcohols,[10] triallylamines,[11] allylammonium chlorides[11]
and even alkylamines.[12] Substituted quinolines were syn-
thesized by Arisawa, Theeraladanon et al. via ring closing
metathesis of α,ω-dienes derived from 2-isopropenylani-
line.[13–15] Other researchers reported the use of Pd,[16–19]
Ni,[20] Rh,[21,22] and Co[23,24] complexes for transition-metal
Scheme 1. The modified Friedländer quinoline synthesis.
Instead of using the unstable 2-aminobenzaldehyde as a
starting product it is generated in situ via a catalytic trans-
fer hydrogenation reaction, involving the oxidation of 2-
aminobenzyl alcohol. RuCl2(=CHPh)(PCy3)2, commonly
known as the first-generation Grubbs’ catalyst, was re-
ported to be the best catalyst for this reaction.[25] Other
metal complexes that have been successfully employed for
the modified Friedländer reaction include CuCl2/O2,[26] Pd/
C,[27] RuCl2(DMSO)4,[28–31] IrCl3 and [IrCl(cod)]2.[32]
Given our experience in ruthenium-based catalysis, this
remarkable activity for hydrogen transfer reactions of the
first-generation Grubbs’ catalyst prompted us to further in-
vestigate the potential of similar ruthenium complexes.
Here we present a comprehensive overview of our results.
Results and Discussion
In a previous communication,[33] we have compared sev-
eral ruthenium complexes based on the first and second
generation Grubbs’ catalyst and the ruthenium dimer
[RuCl2(p-cymene)]2 8 (Figure 1) for the reaction between 2-
aminobenzyl alcohol (1) and acetophenone (2a, R1 = Ph,
R2 = H) which was chosen as a model reaction. The most
important results are summarized in Table 1. Although the
ruthenium dimer precursor 8 was ineffective, the inclusion
of a phosphane ligand (complexes 9a and 9b) proved to be
[a] Centre for Ordered Materials, Organometallics and Catalysis,
Department of Inorganic and Physical Chemistry,
Ghent University,
Krijgslaan 281 (S3), 9000 Ghent, Belgium
Fax: +32-9-2644983
E-mail: Francis.Verpoort@UGent.be
[b] Centre for Surface Chemistry and Catalysis, K. U. Leuven,
Kasteelpark Arenberg 23, 3001 Heverlee Leuven, Belgium
Eur. J. Org. Chem. 2008, 1625–1631
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1625

H. Vander Mierde, P. Van Der Voort, D. De Vos, F. Verpoort
FULL PAPER
Figure 1. Catalysts tested for the modified Friedländer synthesis.
beneficial. The conversion increased from 15% to 53% for
When the reaction is monitored over time (Figure 2), it
9b, whereas the use of an N,O-bidentate Schiff base ligand is revealed that the yields after one hour are not necessarily
in complex 10 did not improve the catalytic activity.
final yields. The catalysts are still active after one hour of
reaction. Complex 5 reaches full conversion after 60 min-
utes, 4 after 90 minutes and also 9a and 9b eventually reach
full conversion after 6 and 5 hours, respectively.
Table 1. Ruthenium-catalyzed synthesis of quinolines from 1 and
acetophenone (2a, R1 = Ph, R2 = H).[a]
Entry
Catalyst
% Yield
1
2
3
4
5
6
7
8
9
4
5
74
100
53
32
15
40
53
14
26
6
7
8
9a
9b
10
11
[a] Reaction conditions:
(0.01 mmol) and KOH (1 mmol) in dioxane (3 mL) at 80 °C for 1
1
(1 mmol), 2a (2 mmol), catalyst
h. Yields determined by GC.
We have shown that catalyst 5 was superior to 4. The
replacement of a phosphane ligand by an N-heterocyclic
carbene (NHC) ligand clearly improved catalytic activity,
resulting in 100% conversion for 5 after 1 hour compared
to 74% for 4. This might be attributed to the higher σ-
donating ability of the NHC ligand, making it more suit-
Figure 2. Monitoring of the reaction over time with different Ru
complexes. Reaction conditions: 1 (1 mmol), 2a (2 mmol), catalyst
(0.01 mmol) and KOH (1 mmol) in dioxane (3 mL) at 80 °C.
From the results in Table 2, it becomes clear that not
able to stabilize the [RuH2] species with a presumably only the catalyst determines the reaction rate, also the base
higher oxidation state compared to the original catalyst. is important. The role of the base is to abstract an α-proton
Variation of the NHC ligand through replacement of one of the ketone, so it can undergo a cross-aldol reaction with
mesityl group by aliphatic groups, such as methyl or cyclo- 2-aminobenzaldehyde [see steps (a), (b) and (c) in
hexyl, decreased the quinoline yield.[33]
Scheme 2]. The pKa value of the α-proton is approximately
Based on the conclusions of our previous report, a series 16 for 2a. Typically, KOH is used by most other researchers
of new complexes (6, 7 and 11) was screened for the modi- and although KOH is insoluble in dioxane, it is seldomly
fied Friedländer reaction. It is remarkable that, when the specified under which conditions it is added.[25,31,32] When
phosphane ligand in 9a or 9b is replaced by an NHC ligand, large pellets are used, the conversion after one hour is only
in complex 11, the conversion does not further increase but 8% but this can be increased to 67% with KOH powder.
drops to 26%. Even more remarkable is the relatively low An even higher conversion of 74% is achieved when KOH
conversion achieved with 6 while it is similar to 5 in struc- is added as a 4  solution in methanol. This is probably
ture. We are currently unable to explain this peculiar behav- caused by the increased solubility. As an added advantage,
iour but apparently, the presence of at least one phosphane this not only results in a higher yield, it is also much more
ligand is required to achieve good yields. Replacing the practical. Unless otherwise stated, for all experiments in
benzylidene ligand of 4 with a bulkier indenylidene ligand this manuscript, KOH was added as a 4  solution in meth-
in complex 7 decreases the conversion.
anol (1 mmol, 250 µL).
1626
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 1625–1631

A Ru-Catalyzed Approach to Friedländer’s Quinoline Synthesis
Scheme 2. Proposed reaction mechanism.
Table 2. Influence of the base on quinoline synthesis with 4.[a]
mately 26, but only 27% quinoline yield is obtained. Both
bases have non-nucleophilic character, but maybe
a
[b]
Entry
Base (1 mmol)
pKa
% Yield
LiHMDS is too aggressive and deactivates the catalyst.
Grubbs et al. have also shown the exchange of the chloride
ligands with the tert-butoxy group, which implies that,
when KOtBu is used, a different catalytic center may be
formed.[34,35]
Organic bases such as triethylamine (entry 9) or 1,8-di-
azabicyclo[5.4.0]undec-7-ene (DBU, entry 10) have the ad-
vantage of being readily soluble in dioxane, but their low
basicity prevents them from being effective bases for this
reaction. Although the proton abstraction is base-catalyzed,
an equimolar ratio of base and 1 gives the best results, as
can be deducted from entries 3, 11 and 12. A higher concen-
tration of base does not further improve the yield. Figure 3
1
2
3
4
5
6
7
8
KOH (pellets)
KOH (powder)
15.7
15.7
15.7
15.7
15.7
15.9
17.0
≈26[c]
10.6
12.8
8
64
74
38
48
74
98
27
0
KOH (4  in MeOH)
NaOH (powder)
NaOH (4  in MeOH)
NaOEt (powder)
KOtBu (powder)
LiHMDS (0,5  in toluene)
Triethylamine
9
10
11
12
DBU
0
67
73
0.4 mmol KOH (4  in MeOH) 15.7
2.0 mmol KOH (4  in MeOH) 15.7
[a] Reaction conditions: 1 (1 mmol), 2a (2 mmol), 4 (0.01 mmol)
and base (1 mmol, except for entries 11 and 12) in dioxane (3 mL)
at 80 °C for 1 h. Yields determined by GC. [b] pKa values of the
protonated form of the base. [c] in THF.
The yield is substantially lower when NaOH is used,
either as powder (entry 4) or as a 4  solution in MeOH
(entry 5). This may be explained by the smaller size of the
sodium cation, resulting in a lower solubility. It is, however,
surprising that a cation change from potassium to sodium
leads to such a big difference. Sodium ethoxide (entry 6)
has approximately the same base strength as NaOH, yet a
higher yield, comparable to KOH, is achieved. Again, the
higher solubility of NaOEt because of the aliphatic ethyl
group can explain these results. Another common base,
KOtBu, has a higher basic strength, which is reflected in
the higher quinoline yield (entry 7). Not only the basic
strength is important, as is evidenced by entry 8. Lithium
Figure 3. KOH vs. KOtBu. Reaction conditions: 1 (1 mmol), 2a
(2 mmol), catalyst 4 or 5 (0.01 mmol) and base (1 mmol) in dioxane
bis(trimethylsilyl)amide (LiHMDS) has a pKa of approxi- (3 mL) at 80 °C.
Eur. J. Org. Chem. 2008, 1625–1631
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1627

H. Vander Mierde, P. Van Der Voort, D. De Vos, F. Verpoort
FULL PAPER
illustrates the effect of the base even better. With KOtBu
To assess the scope of the modified Friedländer reaction,
and 5, full conversion is reached already after 20 minutes, 1 was treated with a variety of ketones in the presence of 4
compared to 60 minutes for KOH. and 5 and both KOH and KOtBu were used as base. The
The determination of the turn-over number (TON) for results are shown in Table 3. From these results, it is obvi-
the model reaction of 1 with 2a in the presence of KOtBu ous that the second generation outperforms the first genera-
and 5 was carried out by lowering the catalyst concentra- tion Grubbs’ catalyst. For almost all ketone substrates, a
tion and measuring the maximum yield. A catalyst loading higher quinoline yield was obtained for 5 compared to 4.
of 0.1% still results in full conversion within 1 h. With a The stronger base KOtBu gives better results than KOH for
catalyst loading of 0.01%, a maximum yield of 85% is ob- almost all ketones. The only two exceptions are acetone (en-
served after 5 h, meaning a TON as high as 8500 was try 8) and 1-indanone (entry 14) were KOH is the preferred
achieved, showing the potential of this catalytic system. The base in combination with 5. The reaction is inhibited by
calculation of the turn-over frequency (TOF) at the begin- substrates with strong electron withdrawing groups like
ning of the reaction (after the first 5 minutes), fully quanti- NO2 (entry 7). Entries 9 and 10 illustrate that a mixture of
fies the difference between the catalytical systems. With two quinolines is formed when two α-protons are available
KOH, complex 4 has a TOF of 1.7 min–1 (measured after in an asymmetric ketone, a problem that is avoided with
20 min because of the observed induction period), while symmetric ketones (entries 12 and 13). The ratio of 3i/3j
that of 5 is twice as large (3.8 min–1). Using KOtBu the and 3k/3l is the same for 4 and 5 as can be expected since
TOF increases spectacularly to 14.0 and 17.0 respectively the only action of the catalyst is the oxidation of 1 (this is
for 4 and 5.
not entirely true, as will be explained in the discussion of
Table 3. Ruthenium-catalyzed synthesis of quinolines from 1 and ketones.[a]
[a] Reaction conditions: 1 (1 mmol), 2 (2 mmol), catalyst (0.01 mmol) and base (1 mmol) in dioxane (3 mL) at 80 °C for 1 h. Yields
determined by GC. [b] From ref.[33] [c] Isolated yield: 65%. [d] Isolated yield: 94%. [e] Value in parentheses is the yield of 3-butyl-2-
methylquinoline (3j). [f] Value in parentheses is the yield of 2-ethyl-3-propylquinoline (3l).
1628
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 1625–1631

A Ru-Catalyzed Approach to Friedländer’s Quinoline Synthesis
the reaction mechanism). There is however a small but be an alternative pathway that allows for catalyst regenera-
notable difference between KOH and KOtBu. With 2-hep- tion.
tanone, the ratio of 3i/3j is 4.2 for KOH and 2.9 for KOtBu,
meaning there is a higher selectivity with KOH. With 3- länder synthesis is presented in Scheme 2. First, in step (a),
heptanone the difference is less pronounced (3 vs. 2.6).
1 is oxidized to 2-aminobenzaldehyde (12) by the ruthenium
A plausible reaction mechanism for the modified Fried-
The GC spectra did not only show unreacted starting catalyst that is hydrogenated to a hydrido-ruthenium com-
products 1 and 2 and the quinoline, but also one peak plex. Under basic conditions, the aldehyde and the ketone
which in many cases almost completely overlapped with the undergo a cross aldol reaction to form 14 in steps (b) and
ketone peak. This peak was identified as the alcohol equiva- (c). Step (j) shows how the catalyst is regenerated by a hy-
lent 2Ј of the ketone 2. This alcohol is the result of hydro- drogen transfer reaction in which ketone 2 is reduced to the
genation of the ketone by the [RuH2] complex, resulting in alcohol 2Ј. This role can also be fulfilled by another hydro-
the regeneration of the catalyst. Exactly for this reason, two gen acceptor, e.g. benzophenone. In step (h), the aldol prod-
equivalents of ketone were used to perform the reaction: uct 14 can cyclize via imine condensation (“imination”) and
one equivalent for the reaction and the other equivalent as subsequent H2O elimination in step (i) leads to the quino-
a hydrogen acceptor for the regeneration of the catalyst. line.
In an alternative pathway, shown in steps (d)–(g), a base-
catalyzed H2O elimination of 14 results in the trans enone
15. The cis product is not likely to be formed due to steric
hindrance. Compound 15 is then hydrogenated by a [RuH2]
species, representing a second method to regenerate the cat-
alyst. Imine condensation and subsequent dehydrogenation
lead to the desired quinoline 3.
Proof that the conversion of 15 to 16 occurs, is found
when benzyl alcohol is reacted with acetophenone in diox-
ane in the presence of 4 and KOH (Scheme 3). Instead of
the expected chalcone 19, 3-phenylpropiophenone 20 is
formed, which means that the double bond of chalcone is
hydrogenated by the [RuH2] complex, regenerating the cata-
lyst in the process. Similar coupling reactions have been per-
formed by Cho et al. and it were in fact these findings that
led them to the modified Friedländer method.[25,36,37]
One could argue that also the oxidation of methanol to
formaldehyde could be the reason of the reduction of
ketone 2. When the reaction is carried out with KOH pow-
der in the absence of methanol, the alcohol 2Ј is still
formed, albeit in slightly smaller quantities. This means that
methanol oxidation certainly contributes to the formation
of 2Ј, but not exclusively and not to a major extent.
Theoretically, the order of steps (a) and (f) could be re-
The use of other hydrogen acceptors was examined by using
only one equivalent of the ketone 2b vs. 1, instead of 2
equivalents (Table 4). With 1 equivalent of benzophenone
(compared to 1), full conversion is reached after 70 minutes.
Increasing the amount of benzophenone to 2 equivalents,
is slightly counterproductive. An other common hydrogen
acceptor, 1-dodecene, was less effective and the use of nitro-
benzene resulted in unwanted side-products, but no quino-
line.
Table 4. Effect of a hydrogen acceptor on quinoline synthesis.[a]
Entry
Additive
% Yield[b]
1
2
3
4
5
6
7
–
51 (71)
91 (100)
83 (100)
0[c]
benzophenone (1 mmol)
benzophenone (2 mmol)
nitrobenzene (1 mmol)
nitrobenzene (2 mmol)
1-dodecene (1 mmol)
1-dodecene (2 mmol)
0[c]
52
72
[a] Reaction conditions: 1 (1 mmol), 2b (1 mmol), 5 (0.01 mmol)
and KOH (1 mmol) in dioxane (3 mL) at 80 °C for 1 h. Yields de-
termined by GC. [b] The value of the maximum yield, achieved
after 90 minutes, is indicated in parentheses. [c] No quinoline was
formed, but the GC spectrum showed many other unidentified
compounds.
When the reaction is carried out without hydrogen ac- versed, i.e. first a condensation reaction between the amine
ceptor, a maximum yield of 71% is achieved. The ketone and the ketone to form an imine, followed by the catalytic
peak has completely disappeared on the GC spectrum and oxidation and cross aldol reaction. This is, however, not
a new peak of the alcohol has appeared, accounting for observed. The reaction of 1 with 2a in basic media did not
approximately 0.30 mmol. This is, however, in contradiction lead to imines. To exclude the possibility of [Ru]-catalyzed
with the previous statement of catalyst regeneration, as with imine formation, aniline was treated with 2a in the presence
a 1:1 ratio, 0.50 mmol of the alcohol and a maximum yield of 4 under standard reaction conditions used for the experi-
of only 50% is to be expected. This means that there must ments, but again, no imines were formed. The interested
Scheme 3. Ru-catalyzed coupling between benzyl alcohol and acetophenone.
Eur. J. Org. Chem. 2008, 1625–1631
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1629

H. Vander Mierde, P. Van Der Voort, D. De Vos, F. Verpoort
FULL PAPER
the mixture was stirred at room temperature. A grey precipitate of
TlCl started to form almost immediately. After 4 h, the solvent
volume was reduced to 1 mL and the mixture was purified by col-
umn chromatography (Silica gel 60, 70–230 mesh, Merck). The red
band containing the catalyst was collected and the solvent volume
was reduced to 1 mL. Upon addition of hexane (10 mL), a red
precipitation formed, that was filtered, washed with hexane
(2ϫ5 mL) and dried in vacuo to afford the pure compound 10 in
good overall yield (1.17 g, 70%). 1H NMR (300 MHz, CDCl3,
25 °C): δ = 7.76 (s, 1 H, CH=N), 7.15 (t, 1 H, SB aryl H), 6.96 (d,
1 H, SB aryl H), 6.92 (d, 1 H, SB aryl H), 6.43 (t, 1 H, SB aryl
H), 5.46 (d, 1 H, p-cymene aryl H), 5.40 (d, 1 H, p-cymene aryl
H), 5.30 (d, 1 H, p-cymene aryl H), 5.07 (d, 1 H, p-cymene aryl
H), 4.23 (m, 1 H, N-CH), 2.80 [m, 1 H, p-cymene CH(CH3)2], 2.50–
1.25 (m, 10 H, cyclohexyl H), 2.16 (s, 3 H, ar-CH3), 1.23, 1.15 [both
d, 3 H, CH(CH3)2] ppm. 13C NMR (300 MHz, CDCl3, 25 °C): δ =
165.4, 161.6, 134.5, 134.3, 123.0, 120.6, 114.3, 102.4, 97.0, 83.9,
83.3, 82.7, 80.7, 76.3, 36.0, 35.1, 30.9, 26.5 (2 C), 25.9, 22.8, 22.1,
18.6 ppm. C23H30ClNORu (473.01): calcd. C 58.40, H 6.39, N 2.96;
found C 58.44, H 6.25, N 2.84.
reader can find an excellent article by Muchowski and
Maddox, dealing with the mechanism of the Friedländer
synthesis.[38]
Conclusions
In summary, we have shown that the first and second
generation Grubbs’ catalysts are efficient catalysts for the
oxidative cyclization of 2-aminobenzyl alcohol with a vari-
ety of ketones, leading to quinolines. The best results are
achieved with the second generation catalyst, in combina-
tion with KOtBu as base. A TON as high as 8500 was ob-
served. A sacrificial hydrogen acceptor is required for the
regeneration of the catalyst. Benzophenone is a fine choice,
but an extra equivalent of ketone also works perfectly. In
the discussion on the reaction mechanism, the experimental
results seem to indicate that there are probably two different
pathways involved in this modified Friedländer reaction.
General Procedure for Quinoline Synthesis:
A mixture of 1
(1 mmol), 2 (2 mmol, unless otherwise stated), base (1 mmol, KOH:
250 µL of a 4  solution in MeOH, other bases in their original
form) and Ru catalyst (0.01 mmol) in 3 mL of 1,4-dioxane (Aldra-
SORBTM, Aldrich) was placed in a 7 mL screw-capped vial and
allowed to react at 80 °C for 1 h. The catalyst and inorganic salts
were removed from the reaction mixture by filtration through a
short silica gel column (ethyl acetate). The reported quinoline
yields were determined by GC. To isolate the quinoline, the re-
sulting solution was concentrated and passed through a second sil-
ica gel column (ethyl acetate/hexane mixture, 1:4). The solvent was
evaporated and the resulting product was dissolved again in a mini-
mal amount of ethyl acetate. A pale yellow precipitate formed upon
addition of HCl (4  solution in dioxane, Aldrich), which was fil-
tered and suspended in an aqueous 1  NaOH solution (15 mL).
The aqueous phase was extracted with CH2Cl2 (2ϫ15 mL) and
after evaporation of the combined CH2Cl2 phases, the quinoline
was obtained in good yield (typically 5–10% lower than GC yields).
Experimental Section
All synthetic procedures were performed under argon atmosphere
on a vacuum line using standard Shlenck techniques. Solvents were
dried and distilled prior to use. Ethyl acetate for column
chromatography was of Rotisolv® Pestilyse® quality (Roth). Com-
pounds 4 (Aldrich), 5 (Aldrich), 7 (Umicore), 8 (Aldrich) and all
other chemicals were obtained from commercial sources and used
as received. Compounds 6,[35] 9a+b,[39,40] and 11[41] were prepared
according to literature procedures. H and 13C NMR spectra were
1
recorded on a Varian Unity-300 Spectrometer. GC measurements
were performed on a Finnigan TraceGC Ultra with an Ultra Fast
Column Module (UFC-1, 100% dimethyl polysiloxane,
0.32 mmϫ5 m, 0.25 µm film thickness, helium carrier gas, 5 mL/
min).
The synthesis of 10 is achieved in three steps: (a) synthesis of the
Schiff base (SB), (b) formation of the thallium salt of the Schiff
base, and (c) reaction of the Tl salt with [RuCl2(p-cymene)]2 to
afford 10. The procedure is analogous to that of similar Ru–Schiff
base complexes previously published by our group.[42]
All quinolines were fully characterized with 1H and 13C NMR spec-
troscopy. Spectroscopic data of 3a–d, 3f–j, 3m and 3n has pre-
viously been reported[30,43] and our data was in accordance with
those results. Yields were determined from the optimized reaction
with 5 and KOtBu.
(a) Cyclohexylamine (5.4 mL, 47 mmol) and salicyladehyde
(5.0 mL, 47 mmol) were dissolved in 25 mL THF. The solution was
allowed to reflux at 60 °C for 4 h, then cooled down to room tem-
perature and dried on MgSO4. After filtration, the solvent was
evaporated in vacuo, affording the Schiff base as a viscous yellow
oil in good yields (6.7 g, 70%). 1H NMR (300 MHz, CDCl3,
25 °C): δ = 13.81 (s, 1 H, OH), 8.31 (s, 1 H, CH=N), 7.25 (t, 1 H,
arom. H), 7.20 (d, 1 H, arom. H), 6.93 (d, 1 H, arom. H), 6.83 (t,
1 H, arom. H), 3.19 (m, 1 H, N-CH), 1.80–1.20 (m, 10 H, cyclo-
hexyl H) ppm. 13C NMR (300 MHz, CDCl3, 25 °C): δ = 162.5,
161.8, 132.2, 131.4, 119.2, 118.6, 117.3, 67.7, 34.5 (2 C), 25.8, 24.6
(2 C) ppm.
2-(4-Methoxyphenyl)quinoline (3e): Pale yellow oil, 0.2094 g, 89%.
1H NMR (300 MHz, CDCl3, 25 °C): δ = 8.15 (m, 2 H), 7.90–7.80
(m, 3 H), 7.71 (m, 1 H), 7.52 (m, 1 H), 7.42 (m, 1 H), 7.13 (m, 1
H), 7.03 (m, 1 H), 3.87 (s, 3 H) ppm. 13C NMR (300 MHz, CDCl3,
25 °C): δ = 157.5, 157.4, 148.6, 135.3, 131.8, 130.6, 130.0, 129.9,
129.5, 127.7, 127.3, 126.4, 123.7, 121.5, 111.7, 55.9 ppm.
C16H13NO (235.28): calcd. C 81.68, H 5.57, N 5.95; found C 81.74,
H 5.85, N 5.75.
1
2-Butyl-3-methylquinoline (3k): Pale yellow oil, 0.1116 g, 56%. H
NMR (300 MHz, CDCl3, 25 °C): δ = 8.01 (m, 1 H), 7.82 (d, 1 H),
7.70 (m, 1 H), 7.60 (m, 1 H), 7.44 (m, 1 H), 2.98 (m, 2 H), 2.48 (s,
3 H), 1.76 (m, 2 H), 1.50 (m, 2 H), 1.04 (m, 3 H) ppm. 13C NMR
(300 MHz, CDCl3, 25 °C): δ = 161.5, 145.6, 134.7, 128.5, 127.2,
126.4, 126.2, 125.6, 124.5, 35.2, 30.0, 22.0, 18.2, 13.0 ppm.
C14H17N (199.29): calcd. C 84.37, H 8.60, N 7.03; found C 84.09,
H 8.60, N 6.68.
2-Ethyl-3-propylquinoline (3l): Pale yellow oil, 0.0417 g, 21%. 1H
NMR (300 MHz, CDCl3, 25 °C): δ = 8.01 (m, 1 H), 7.82 (d, 1 H),
7.70 (m, 1 H), 7.60 (m, 1 H), 7.44 (m, 1 H), 2.97 (m, 2 H), 2.66
(b) The Schiff base (0.72 g, 3.53 mmol) was treated with thallium(I)
ethoxide (250 µL, 3.53 mmol) in 1 mL THF at room temperature.
A yellow precipitate of the Tl salt of the Schiff base started to form
after a few minutes. After 1 h, the solvent was evaporated under
reduced pressure and the crude product was used in the next step
without further purification.
(c) A solution of [RuCl2(p-cymene)]2 8 (1.08 g, 1.76 mmol) in
25 mL THF was added to the Schiff base (SB) thallium salt and
1630
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 1625–1631

A Ru-Catalyzed Approach to Friedländer’s Quinoline Synthesis
(m, 2 H), 1.76 (m, 2 H), 1.39 (m, 2 H), 0.98 (m, 3 H) ppm. 13C
NMR (300 MHz, CDCl3, 25 °C): δ = 162.0, 145.5, 133.8, 132.6,
127.8, 127.4, 127.3, 126.2, 125.8, 33.3, 27.8, 22.5, 13.0, 12.6 ppm.
C14H17N (199.29): calcd. C 84.37, H 8.60, N 7.03; found C 84.09,
H 8.60, N 6.68.
[10] Y. Watanabe, Y. Tsuji, Y. Ohsugi, Tetrahedron Lett. 1981, 22,
2667–2670.
[11] C. S. Cho, J. S. Kim, B. H. Oh, T. J. Kim, S. C. Shim, N. S.
Yoon, Tetrahedron 2000, 56, 7747–7750.
[12] C. S. Cho, B. H. Oh, J. S. Kim, T. J. Kim, S. C. Shim, Chem.
Commun. 2000, 1885–1886.
2-Methyl-1,2,3,4-tetrahydroacridine (3o): White solid, 0.1815 g,
[13]
M. Arisawa, Y. Terada, C. Theeraladanon, K. Takahashi, M.
Nakagawa, A. Nishida, J. Organomet. Chem. 2005, 690, 5398–
5406.
1
92%. H NMR (300 MHz, CDCl3, 25 °C): δ = 7.98 (d, 1 H), 7.9
(s, 1 H), 7.61 (d, 1 H), 7.52 (t, 1 H), 7.34 (t, 1 H), 3.20–2.90 (m, 3
H), 2.50 (dd, 1 H), 1.98 (m, 1 H), 1.90 (m, 1 H), 1.52 (m, 1 H)
ppm. 13C NMR (300 MHz, CDCl3, 25 °C): δ = 159.3, 146.9, 135.2,
130.8, 128.7, 127.4, 127.1, 125.8, 38.0, 33.4, 31.7, 29.3, 21.9 ppm.
C14H15N (197.28): calcd. C 85.24, H 7.66, N 7.10; found C 84.98,
H 7.88, N 7.03.
11H-Indeno[1,2-b]quinoline (3p): White solid, 0.0326 g, 15%. 1H
NMR (300 MHz, CDCl3, 25 °C): δ = 8.31 (s, 1 H), 8.19 (m, 2 H),
7.83 (m, 1 H), 7.70 (m, 1 H), 7.61 (m, 1 H), 7.51 (m, 3 H), 4.05 (s,
2 H) ppm. 13C NMR (300 MHz, CDCl3, 25 °C): δ = 161.9, 148.3,
145.3, 140.6, 134.9, 131.4, 130.2, 129.9, 129.4, 129.1, 128.0, 127.8,
125.9, 125.7, 122.3, 34.3 ppm. C16H11N (217.27): calcd. C 88.45, H
5.10, N 6.45; found C 88.55, H 5.12, N 6.31.
[14]
[15]
[16]
C. Theeraladanon, M. Arisawa, A. Nishida, M. Nakagawa,
Tetrahedron 2004, 60, 3017–3035.
M. Arisawa, C. Theeraladanon, A. Nishida, M. Nakagawa,
Tetrahedron Lett. 2001, 42, 8029–8033.
L. S. Hegedus, G. F. Allen, J. J. Bozell, E. L. Waterman, J.
Amer. Chem. Soc. 1978, 100, 5800–5807.
R. C. Larock, M. Y. Kuo, Tetrahedron Lett. 1991, 32, 569–572.
C. S. Cho, J. Organomet. Chem. 2005, 690, 4094–4097.
J. S. Mahanty, M. De, P. Das, N. G. Kundu, Tetrahedron 1997,
53, 13397–13418.
[17]
[18]
[19]
[20]
[21]
R. P. Korivi, C. H. Cheng, J. Org. Chem. 2006, 71, 7079–7082.
S. E. Diamond, A. Szalkiewicz, F. Mares, J. Am. Chem. Soc.
1979, 101, 490–491.
[22]
M. Beller, O. R. Thiel, H. Trauthwein, C. G. Hartung, Chem.
Eur. J. 2000, 6, 2513–2522.
Synthesis of 3-Phenylpropiophenone (20): 4 (0.0082 g, 0.01 mmol),
benzyl alcohol (0.1081 g, 1 mmol), acetophenone (0.1201 g,
1 mmol) and KOH powder (0.0561 g, 1 mmol) in 3 mL of dioxane
were placed in a screw-capped vial and allowed to react for 1 h at
80 °C. The mixture was passed through a silica gel column (ethyl
acetate) and the solvent was evaporated. 19 was obtained as an
orange solid (0.1984 g, 94%). 1H NMR (300 MHz, CDCl3, 25 °C):
δ = 7.95 (d, 2 H), 7.55 (t, 1 H), 7.44 (t, 2 H), 7.24–7.36 (m, 5 H),
3.30 (t, 2 H), 3.07 (t, 2 H) ppm. 13C NMR (300 MHz, CDCl3,
25 °C): δ = 199.5, 141.6, 137.1, 133.3, 128.9 (2 C), 128.8 (2 C),
128.7 (2 C), 128.3 (2 C), 126.4, 40.7, 30.4 ppm.
[23]
[24]
J. Jacob, W. D. Jones, J. Org. Chem. 2003, 68, 3563–3568.
J. Jacob, C. M. Cavalier, W. D. Jones, S. A. Godleski, R. R. Va-
lente, J. Mol. Catal. A Chem. 2002, 182, 565–570.
C. S. Cho, B. T. Kim, T. J. Kim, S. C. Shim, Chem. Commun.
2001, 2576–2577.
C. S. Cho, W. X. Ren, S. C. Shim, Tetrahedron Lett. 2006, 47,
6781–6785.
C. S. Cho, W. X. Ren, S. C. Shim, Bull. Korean Chem. Soc.
2005, 26, 1286–1288.
R. Martinez, D. J. Ramon, M. Yus, Eur. J. Org. Chem. 2007,
1599–1605.
R. Martinez, D. J. Ramon, M. Yus, Tetrahedron 2006, 62,
8982–8987.
R. Martinez, D. J. Ramon, M. Yus, Tetrahedron 2006, 62,
8988–9001.
R. Martinez, G. J. Brand, D. J. Ramon, M. Yus, Tetrahedron
Lett. 2005, 46, 3683–3686.
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
Acknowledgments
The authors gratefully acknowledge the Research Fund of Ghent
University for financial support during the preparation of this ma-
nuscript and Olivier F. Grenelle and Dr. Marc G. Proot of Chevron
Technology, Ghent for elemental analyses. F. V. and D. De V. also
thank the FWO Flanders for a project grant. Catalyst 7 was gener-
ously supplied to us by Umicore.
K. Taguchi, S. Sakaguchi, Y. Ishii, Tetrahedron Lett. 2005, 46,
4539–4542.
H. Vander Mierde, N. Ledoux, B. Allaert, P. Van Der Voort,
R. Drozdzak, D. De Vos, F. Verpoort, New J. Chem. 2007, 31,
1572–1574.
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]
M. S. Sanford, L. M. Henling, M. W. Day, R. H. Grubbs, An-
gew. Chem. Int. Ed. 2000, 39, 3451–3453.
M. S. Sanford, J. A. Love, R. H. Grubbs, Organometallics 2001,
20, 5314–5318.
C. S. Cho, B. T. Kim, T. J. Kim, S. C. Shim, Tetrahedron Lett.
2002, 43, 7987–7989.
C. S. Cho, B. T. Kim, T. J. Kim, S. C. Shim, J. Org. Chem. 2001,
66, 9020–9022.
J. M. Muchowski, M. L. Maddox, Can. J. Chem. Revue Canad-
ienne de Chim. 2004, 82, 461–478.
M. A. Bennett, A. K. Smith, J. Chem. Soc. Dalton Trans. 1974,
233–241.
R. A. Zelonka, M. C. Baird, Can. J. Chem. 1972, 50, 3063–
3072.
L. Jafarpour, J. K. Huang, E. D. Stevens, S. P. Nolan, Organo-
metallics 1999, 18, 3760–3763.
B. De Clercq, F. Verpoort, J. Mol. Catal. A Chem. 2002, 180,
67–76.
C. S. Cho, B. T. Kim, H. J. Choi, T. J. Kim, S. C. Shim, Tetrahe-
dron 2003, 59, 7997–8002.
[1] T. L. Gildchrist in Heterocyclic Chemistry, 1st ed., Pitman Pub-
lishing LTD, London, 1985, p. 239–302.
[2] D. Lednicer, L. A. Mitscher in The Organic Chemistry of Drug
Synthesis, Wiley-Interscience, New York, 1977, vol. 1, p. 337–
347.
[3] G. Roma, M. Di Braccio, G. Grossi, F. Mattioli, M. Ghia, Eur.
J. Med. Chem. 2000, 35, 1021–1035.
[4] D. Dube, M. Blouin, C. Brideau, C. C. Chan, S. Desmarais, D.
Ethier, J. P. Falgueyret, R. W. Friesen, M. Girard, Y. Girard, J.
Guay, D. Riendeau, P. Tagari, R. N. Young, Bioorganic Med.
Chem. Lett. 1998, 8, 1255–1260.
[5] Y. L. Chen, K. C. Fang, J. Y. Sheu, S. L. Hsu, C. C. Tzeng, J.
Med. Chem. 2001, 44, 2374–2377.
[6] H. Ebisu, M. Nishikawa, M. Tanaka, T. Okazoe, Y. Morizawa,
H. Shinyama, N. Nakamura, J. Cardiovasc. Pharmacol. 1999,
34, 526–532.
[7] M. P. Maguire, K. R. Sheets, K. McVety, A. P. Spada, A. Zilb-
erstein, J. Med. Chem. 1994, 37, 2129–2137.
[8] S. Concilio, P. M. Pfister, N. Tirelli, C. Kocher, U. W. Suter,
Macromolecules 2001, 34, 3607–3614.
[9] H. S. Nalwa, M. Suzuki, A. Takahashi, A. Kageyama, Appl.
Phys. Lett. 1998, 72, 1311–1313.
Received: October 22, 2007
Published Online: February 1, 2008
Eur. J. Org. Chem. 2008, 1625–1631
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1631