Quinoline formation via a modified Combes reaction: Examination of kinetics, substituent effects, and mechanistic pathways

Article abstract of DOI:10.1002/poc.1433

A comprehensive product regioselectlvity and kinetics study of the modified Combes quinoline synthesis shown below has been undertaken: This is the first reported investigation of the Combes condensation employing 19F NMR spectroscopy to monitor intermedi

Full text of DOI:10.1002/poc.1433

Research Article
Received: 7 April 2008,
Revised: 21 July 2008,
Accepted: 21 July 2008,
Published online in Wiley InterScience: 26 August 2008
(www.interscience.wiley.com) DOI 10.1002/poc.1433
Quinoline formation via a modified Combes
reaction: examination of kinetics, substituent
effects, and mechanistic pathways
Joseph C. Sloopa
*
A comprehensive product regioselectivity and kinetics study of the modified Combes quinoline synthesis shown
below has been undertaken:
This is the first reported investigation of the Combes condensation employing 19F NMR spectroscopy to monitor
intermediate consumption and product formation rates. The reaction was found to be first order in both the diketone
and aniline. Product regioselectivity and reaction rates were found to be influenced by substituents on the diketones
and anilines with rates varying as much as five fold. The consumption rate of key imine and enamine intermediates
mirrored quinoline formation rates, in accord with rate determining annulation. A r of S0.32 was determined for this
cyclization. While the sign of the reaction constant is consistent with rate limiting electrophilic aromatic substitution
(EAS), the magnitude is likely a composite value, resulting from opposing substituent effects in the nucleophilic
addition and EAS steps. Mechanistic details and reaction pathways supporting these findings are proposed. Copyright
ß 2008 John Wiley & Sons, Ltd.
Keywords: Hammett correlation; steric effects; substituent effects; kinetics; Combes quinoline synthesis; trifluoromethyl-1,3-
diketones
The problems attendant with early quinoline formation
methods which used concentrated sulfuric acid as a dehydrating
INTRODUCTION
Quinolines, which are important precursors to pharmaceuticals
and agrochemicals,[1–4] have been studied extensively from a
synthetic point of view.[5–14] The Combes quinoline synthesis
depicted in Scheme 1 uses a b-diketone substrate, setting it apart
from other quinoline preparations.[15] The reaction, which
involves nucleophilic addition to the carbonyl, imine, and
enamine intermediates and electrophilic aromatic annulation
has been qualitatively investigated in terms of substituent effects
on product regiochemistry and aniline reactivity, but a systematic
kinetics investigation of this condensation is lacking.[16–19]
reagent at high temperatures[12–14,17] can be ameliorated by
substituting mixtures of polyphosphoric acid (PPA) and selected
alcohols.[28,29] The resulting polyphosphoric ester (PPE) catalysts
are milder and more effective as dehydrating/cyclizing reagents
than concentrated H2SO4 or neat PPA. This work[16,19,28]
demonstrates that PPA in ethanol is an especially effective
catalyst/solvent system, giving satisfactory yields of trifluoro-
methylquinolines at 70 8C.[29]
The role of intermediates in quinoline product distribution
Multiple modes of aniline addition to b-diketones lead to
intermediates shown in Scheme 1, which have been isolated in
several cases.[11,13,16–19] The 1,2-addition of anilines to the
substrate keto group followed by a 1,2-elimination of water
produce an imine,[11,16] whereas the enamine results from either
Reactant/catalyst composition and reaction conditions
The nucleophilic aniline reactant may contain a variety of
substituents. Powerful electron-withdrawing groups (EWGs) have
been found to inhibit the reaction while electron-donating
groups (EDGs) can accelerate quinoline formation unless
substituted ortho or para to the amino group.[5–8,17,18]
The b-diketone electrophile exists as a tautomeric mixture of
two chelated, cis-enols and the diketo form in most solvent
systems.[20–25] However, 2-fluoro-b-diketones are known to exist
solely as the ketonic tautomers.[26] Both the enol and keto forms
of b-diketones generally react rapidly with amines in acidic
media.[27] These phenomena will be discussed in detail later.
*
Correspondence to: J. C. Sloop, Department of Chemistry and Life Science,
United States Military Academy, 646 Swift Rd., West Point, NY 10996, USA.
E-mail: joseph.sloop@usma.edu
a
J. C. Sloop
Department of Chemistry and Life Science, United States Military Academy,
646 Swift Rd., West Point, NY 10996, USA
J. Phys. Org. Chem. 2009, 22 110–117
Copyright ß 2008 John Wiley & Sons, Ltd.

QUINOLINE FORMATION VIA COMBES REACTION
Scheme 1. Combes quinoline synthesis
a 1,4-addition of the aniline to the enol species or 1,2-addition to
the carbonyl followed by a 2,3-elimination of water.[18,19]
Substitution patterns on the aniline also influence quinoline
regiochemistry and, in concert with intermediate topology,
complicate product distribution. Figure 1 shows the various
positions on the aniline ring at which cyclization could occur.
In earlier work, the syntheses of selectively fluorinated
quinolines derived from a modified Combes’ method were
reported.[15] While preparing these heterocycles, variations in
product formation rates and regioselectivity were noted,
prompting a comprehensive kinetics and substituent effects
study of this reaction.
With the exception of recent work by Clayton,[30] no other
substituent effects studies on aromatic annulations have been
conducted. The principal efforts in this work were directed at
quantifying the influences that substituted trifluoro-
methyl-b-diketones and substituted anilines exert on the rate
of quinoline formation. Finally, mechanistic pathways for this
condensation are proposed which are consistent with exper-
imental observations.
possible quinoline regioisomers (d ꢀ ꢁ62–67 ppm)[15,31] were
followed. The quinoline products are shown in Scheme 2.
Virtual quinoline formation rates were determined for the
condensation of selected trifluoromethyl-b-diketones with
anilines. The product distribution, initial rates of product
formation, observed rate constants and relative rates are
presented in Table 1. The initial rates were determined as the
slope of the best fit line to a plot of (quinoline) versus time which
consisted of 60 data points. All initial rates were obtained prior to
t1/2 for consumption of the imine and enamine intermediates
observed, with t1/2 ranging from ꢂ40 min to > 3 h.
The reaction order for quinoline formation was determined by
the isolation method. The results, recorded in Table 2, reflect first
order dependence of quinoline formation on both diketone and
aniline, where rate ¼ kobs[diketone][aniline] and kobs is a
second-order rate constant.
Effect of diketones and anilines on quinoline product
distribution
The modulation of quinoline product distribution by diketone
substituents is evident from the data in Table 1. Regioisomeric
quinolines were formed in most cases, in agreement with prior
investigations.[17,31] Exceptions to this trend include entries 6, 7,
12, 16, and 17, in which the 2-CF3 quinoline was the sole product.
Because 2-fluoro-1,3-diketones exist in ketonic form,[20,26] the
initial nucleophilic addition of the aniline in the cases of entries 6
and 16 would preferentially occur at the carbonyl with the least
electron density, for example, adjacent to the CF3 substituent. For
entry 17, the aniline preferentially attacks the carbonyl adjacent
to the functional group with the smaller van der Waals volume
(VW); —CF3 (20.49 cm3/mol) versus —CH(CH3)2 (34.12 cm3/
mol).[32] At present, it is unclear why the 2-CF3 quinoline was
the only product for entries 7 and 12.
QUINOLINE KINETICS STUDY FINDINGS
AND DISCUSSION
Synthesis
Figure 2 shows the trifluoromethyl-b-diketones used in this work.
Diketones 1a–f are available commercially; 1g was prepared by
literature methods.[15,25]
Once the trifluoromethyl-b-diketones were in hand, their
condensation with anilines in ethanol/PPA as solvent/catalyst
system (Scheme 2) afforded the trifluoromethyl quinolines.[15]
In most cases, increasing the bulk of the R group on the
diketone led to formation of the 2-CF3 quinoline product in larger
proportion than the 4-CF3 isomer. Exceptions to this tendency
were entry 4 (R ¼ t-Bu), in which the 4-CF3 isomer formed to the
exclusion of the 2-CF3 quinoline product, and entry 5 (R ¼ Ph)
where the major product was the 4-CF3 isomer. These cases seem
to suggest that steric demands during the initial nucleophilic
addition of aniline to the diketone may not be as great as those
which must be overcome in the annulation (electrophilic
aromatic substitution (EAS)) step in this reaction.
Quinoline formation kinetic studies
Reaction progress was conveniently monitored over a 3 h period
by 19F NMR using CFCl3 as an internal reference. Consumption of
the diketone tautomers (d ꢀ ꢁ77–78 ppm), imine, and enamine
intermediates (d ꢀ ꢁ75–83 ppm) as well as formation of the two
The effect of aniline substituents on quinoline product
distribution in this work is in accord with that of previous
investigators,[12,29] for example, o-substituted anilines give
8-quinolines, m-substituted anilines give predominately
7-quinolines, and p-substituted anilines give 6-quinolines. In
only one case, Table 1 entry 18, was the 5-quinoline observed.
Figure 1. Cyclization positions for quinoline formation
www.interscience.wiley.com/journal/poc
Copyright ß 2008 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2009, 22 110–117

J. C. SLOOP
sizable for entry 2, is small for entries 6 and 16, indicating that
EWGs enhance the electrophilic nature of the carbonyl under-
going condensation in the rate-determining step through
induction for this series of reactions.
The complexity of the modified Combes condensation and the
acidic reaction conditions employed in this study make analysis
of the aniline substituent effect on quinoline formation rates
challenging. Each step in this reaction is subject to substituent
effects; these influences may oppose each other.[34,35] In the
initial nucleophilic addition, protonation of approximately 80% of
the aniline species undoubtedly slows this step. If this step were
rate limiting, o,p-EDGs would accelerate the reaction while EWGs
would have the opposite effect, with a correlation between log k
and so/m/p anticipated. Such a correlation of the data for Table 1
entries 1, 7–14, and 18 was attempted, but the data fit
unsatisfactorily to log k ¼ ꢁ0.12s ꢁ 6.94 (R ¼ 0.099, s ¼ 0.176).
For a rate-determining EAS step, through-resonance m-EDGs
on the intermediate imine or enamine would accelerate the
reaction, with a correlation between log k and sþp anticip-
ated.[36,37] Inductive o,p-EWGs would slow the reaction and a
correlation of log k with smþ would be expected.[38] Correlation of
log k and sþo=m=p for Table 1 entries 1, 7–14, and 18 gave a least
squares fit to the equation log k ¼ ꢁ0.31sþo=m=p ꢁ 6.98 (R ¼ 0.95,
s ¼ 0.075).
Figure 2. Trifluoromethyl-b-diketones
With respect to —CF3 substituent regiochemistry in the
quinoline products, two patterns are discernible. Methoxy-
substituted anilines (Table 1, entries 9, 10, 12, 18) gave
predominantly the 2-CF3 quinoline, whereas the chloro- and
fluoroanilines (Table 1, entries 8, 11, 13–14) gave a majority of the
4-CF3 regioisomer.
Effect of diketones and anilines on quinoline formation
rates
Bulky R groups on the diketone were found to slow the reaction
rate by as much as 80%. For the aliphatic examples, this is likely
due to increased VW of functional groups adjacent to the
annulation site of the transition state in the rate-determining
step.[32] Because of the similarities in electronic effect, log k values
of Table 1, entries 1, 3, 4, and 15 correlate well with VW to log
k ¼ ꢁ0.0031VW ꢁ 6.95 (R > 0.99, s ¼ 2.00).
If, however, the observed substituent effect for this reaction
system is a combination of the effects exerted in the two
aforementioned steps, a more satisfactory correlation may be
achievþed using a dual parameter approach such as log
k ¼ rso=m=p ꢁ r0so/m/p or log k ¼ rðsoþ=m=p ꢁ xso=m=pÞ . Figure 3
depicts a plot of this type for the entries above.
However, log k for entry 5 (R ¼ Ph) does not correlate with its
van der Waals volume (45.84 cm3/mol). While steric effects play a
role in formation of the aromatic substituted quinolines, the
predominant factor is electronic in nature. The extended
conjugation exhibited by aromatic b-diketones lowers their
ground state energy, decreasing their reactivity toward nucleo-
philic addition relative to alkyl b-diketones.[33]
The data were best fit to the equation log k ¼ ꢁ0.32
(sþo=m=p ꢁ 0.3so/m/p) ꢁ 6.99 (R ¼ 0.98, s ¼ 0.036).[39] This correlation
shows that while EDGs on the imine intermediate such as
m-OCH3 increase the rate of product formation through
resonance in the annulation step,[27] the dominance of inductive
electron withdrawing effects over resonance donation contri-
butions in the nucleophilic addition step offset this by 30%. This is
one contributing factor to the small magnitude of r, ꢁ0.32, which
is a composite of the reaction constants for the imine formation
and EAS steps.
The steric argument above failed for entries 2, 6, and 16 as well,
where a-fluoro substituents on the diketones were found to
accelerate the reaction by as much as 50%. This effect, while
Scheme 2. Condensation of trifluoromethyl-b-diketones with anilines
J. Phys. Org. Chem. 2009, 22 110–117
Copyright ß 2008 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc

QUINOLINE FORMATION VIA COMBES REACTION
Table 1. Quinoline formation kinetic dataa
O
O
R
CF3
X
kobs ꢃ 10ꢁ7d
Entry
Quinoline (%)b
R, X ¼ H,F
dPT/dtc (mM/s)
(mM s)ꢁ1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
2a(28)/3a(72)
2b (100)
2c(90)/3c(10)
3d(100)
2e(30)/3e(70)
2g(100)
4a(100)
5a(11)/6a(89)
7a(56)/8a(44)
9a(69)/10a(31)
11a(23)/12a(77)
13a(100)
14a(46)/15a(54)
16a(44)/17a(56)
14f(50)/15f(50)
16g(100)
18f(100)
19a(74)/20a(26)
CH3
CF3
C2H5
C(CH3)3
Ph
0.083
0.115
0.075
0.066
0.040
0.087
0.066
0.066
0.076
0.127
0.066
0.076
0.073
0.067
0.072
0.084
0.076
0.198
1.02
1.42
0.926
0.815
0.494
1.07
0.815
0.815
0.938
1.57
0.815
0.938
0.901
0.827
0.889
1.04
CH3, X ¼ F
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH(CH3)2
CH3, X ¼ F
CH(CH3)2
CH3
0.938
2.45
a (reactant) ¼ 0.900 M (ꢄ0.001), 343 K.
b Product distributions are those obtained prior to distillation or other purification.
c dPT/dt ¼ dP2/dt þ dP3/dt; accuracy ꢄ 0.001 m M/s.
d kobs determined from dPT/dt; accuracy ꢄ 0.001.
Other factors may also conspire to produce the small value for
r observed here. Since temperature and dielectric constant are
inversely proportional to log kobs, reaction conditions used in this
study would tend to drive down the value.[40,41] Additionally, the
temperature at which this investigation was carried out for this
reaction series may be near an isokinetic point and an attenuated
substituent effect is being observed.[34,42] Additional studies
would be needed to verify these effects.
The negative value of r, nonetheless, is in accord with
rate-determining EAS. However, correlation of log kPT with
(sþp ꢁ 0.3sm) and ((spþ þ so) ꢁ 0.3 sm)) for Table 1 entries 10 and
18, respectively, are suggestive of a multi-step process in which
the inductive substituent effects exerted during the nucleophilic
addition step are significant and must be accounted for.
Proposed mechanisms of quinoline formation
These results enable us to provide some mechanistic detail of the
modified Combes condensation leading to the 2-CF3 quinoline.
The following scheme summarizes reactions of the diketo form,
enol 2, and Michael-fashion attack of enol 1 with aniline. An
analogous mechanism involving the other carbonyl of the diketo
form, enol 1, and Michael-fashion attack of enol 2 leads to the
4-CF3 quinoline.
The first stage of the reaction likely commences with formation
of a PPA ester (PPE), accompanied by some aniline protonation
due to the acidity of the medium.[30] Approximately 200 mM of
the free base is available for nucleophilic addition.
Table 2. Initial rates for quinoline formation determined by
the isolation method
Stage 2 follows with a preassociative complex consisting of the
aniline/anilinium, diketone, and PPE. The initial nucleophilic
addition is assisted by the PPE. While the protonation of aniline
likely slows the nucleophilic addition step, the diketone
consumption rate for Table 1, entry 1 {ꢁ[d(1a)/dt]} was measured
at 0.134 mM/s, and is not rate limiting. Both the diketo and enol
tautomeric forms of the trifluoromethyl-b-diketones likely
undergo nucleophilic attack by aniline[33] since its borderline
hardness makes both 1,2- and 1,4-carbonyl additions possible.
Intermediates in this stage of the pathway are likely the initial
amino-phosphate ester adduct, I and the amino-alcohol adduct,
[PhNH2]
(mM)
[1a]
(mM)
dPT/dt
(Mm/s)
nRel
(PhNH2)
nRel
(1a)
450
900
900
1800
900
900
900
450
900
0.041
0.083
0.041
0.165
0.165
0.494
1.00
1.00
0.494
1.99
1800
1.99
www.interscience.wiley.com/journal/poc
Copyright ß 2008 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2009, 22 110–117

J. C. SLOOP
Figure 3. Log(kPT/k0) versus sx for 1a and substituted quinolines
Figure 4. Phenylimine and phenylenamine intermediates
I0. Intermediate I undergoes dehydration to the imine inter-
mediate II rapidly, while elimination of water from I0 leads to the
enamine intermediate II0. Tautomerization of II0F II takes place
during this stage.
Imines II-2a and II-3a shown in Figure 4 leading to quinoline
products 2a and 3a were isolated and characterized. The
enamine intermediate II-3a0 was identified by 19F NMR and
1H NMR, but constituted only 5% of the reaction mixture.
Attempts to separate the intermediates led to tautomerization of
II-3a0 to II-3a. When the intermediates were resubjected to
reaction conditions, the consumption rate (ꢁd[IIT]/
dt ¼ 0.083 mM/s) was clearly comparable to the quinoline
formation rate shown in Table 1, entry 1.
Rate-determining annulation occurs in the third stage and a
rapid proton transfer leads to the dihydroquinoline phosphate
ester intermediate III. Dihydroquinoline formation is consistent
with the observation that EWGs on the diketones and
m-substituted EDGs on aniline enhance the rate of intermediate
II disappearance.
The final stage of the mechanism contains the dehydration of
intermediate III to the heterocyclic products. At 70 8C and in the
presence of PPA, this step is fast.
by an interplay of steric and electronic effects; these factors
generally lead to a preference for 2-CF3 quinolines.
The reaction follows 2nd order kinetics, 1st order in the
diketone and 1st order in aniline. Diketones containing alkyl
groups, multiple —CF3 substituents or those with a 2-fluoro and a
—CF3 group react more rapidly than aryl diketones while anilines
containing EDGs at the meta position accelerate quinoline
formation.
The modified-Combes condensation investigated under the
specified experimental conditions and pH was not found to be as
sensitive to substituent effects as expected. This can be
attributed, at least in part, to the composite nature of r for
this reaction. Otherwise the negative value of r, the isolation of
imine and enamine intermediates, the consumption rate ꢁd[int]/
dt < ꢁd[diketone]/dt, as well as the observation that ꢁd[int]/
dt ꢀ d[quinoline]/dt are consistent with other rate-limiting EAS
processes.
Scheme 3 provides a picture of quinoline formation from
diketone and aniline that is consistent with the kinetics, observed
substituent effects, and presence of identified intermediates.
EXPERIMENTAL
CONCLUSIONS
Chemicals and instrumentation
This kinetics study of a modified Combes trifluoromethylquino-
line synthesis has yielded information regarding how substitu-
ents influence product regioselectivity and reaction rate for this
important condensation. Quinoline regioselectivity is influenced
All chemicals used in the following procedures were obtained
from the Aldrich Chemical Company. Solvents used were of
spectrophotometric grade. Melting points were obtained on a
Mel-Temp melting point apparatus and are uncorrected. NMR
J. Phys. Org. Chem. 2009, 22 110–117
Copyright ß 2008 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc

QUINOLINE FORMATION VIA COMBES REACTION
Scheme 3. Proposed aniline formation mechanism
data were collected using a Varian VXR-200 spectrometer using
CDCl3 as an internal standard unless otherwise noted
(19F: 168 MHz in ethanol w/CFCl3 as an external standard,
Type J temperature controller w/thermocouple. Combustion
analysis was performed with a Perkin–Elmer CHN analyzer.
1
13C: 50.2 MHz, H: 200 MHz). Solution pH was determined using
General procedures for the preparation of quinolines[15]
an Orion SA520 pH meter with a polymer body sealed
combination reference electrode and calibrated using an HCl/
NaCl buffer solution in ethanol at pH 1. Temperatures for the
kinetic runs were thermostatically maintained by an Ace Glass
NMR kinetics studies
To a 100 ml volumetric flask equipped with a magnetic stirrer was
added approximately 60.00 ml 95% ethanol via buret, followed by
www.interscience.wiley.com/journal/poc
Copyright ß 2008 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2009, 22 110–117

J. C. SLOOP
addition of 15.00 g PPA. The mixture was stirred and warmed to
70 8C (ꢄ1 8C) to solubilize the PPA. The pH of the solution was
measured at 0.15 (ꢄ0.01). Then, approximately 90.00 mmol of the
1,3-diketone and an equimolar amount of the aniline were added
in rapid succession and the solution diluted with warm ethanol to
100.00 ml with a buret. The reaction mixture was transferred to a
two-necked flask equipped with a reflux condenser and
thermostatically heated at 70 8C (ꢄ1 8C) under a CaCl2 drying
tube. A 90.00 mmol solution of the 1,3-diketone in 95% ethanol
was prepared, and a 19F NMR taken to serve as a t ¼ 0 s sample.
Aliquots from the reaction flask were then extracted at 180.0 s
intervals, transferred to an NMR tube, quenched in an ice bath,
diluted with CDCl3, and 19F NMR data collected. Samples were
drawn over a 3 h period. After 12 h, all reaction mixtures were
neutralized with saturated NaHCO3, extracted with CH2Cl2, dried
over MgSO4, the solvent removed under reduced pressure, and
then distilled or recrystallized from EtOH.
7-Methoxy-4-methyl-2-trifluoromethylquinoline (9a) and
7-methoxy-2-methyl-4-trifluoromethylquinoline (10a)
These compounds were obtained as a tan solid (EtOH), a 69:31
mixture of 9a:10a, mp 65–68 8C. NMR:9a: 1H: d 2.7 (3H, s), 6.5 (d,
J ¼ 6.6 Hz, 1H), 6.7 (lH, s), 6.9 (lH, d, J ¼ 6.7 Hz), 7.2 (lH, d, J ¼ 6.9 Hz).
2
l3C: d 24, 114, 121 (CF3, q, JC—F ¼ 270 Hz), 143 (C—CF3, q,
2JC—F ¼ 31 Hz, C—CF3). 19F: d ꢁ67.2 (3F, s). 10a: 1H: d 2.6 (3H, s), 6.4
(lH, d, J ¼ 6.8 Hz), 6.6 (lH, s), 6.8 (lH, d, J ¼ 6.9 Hz), 7.1 (lH, d,
J ¼ 6.7 Hz). l3C: 6 24, 114, 124 (CF3, q, 1JC—F ¼ 272 Hz), 133 (C—CF3,
2
q, JC—F ¼ 33 Hz). 19F:d ꢁ62.3 (3F, s). Analysis: calculated
for C13H12F3NO: C, 59.76, H, 4.18, N, 5.81. Found: C, 59.80, H,
4.15, N, 5.82.
6-Chloro-4-methyl-2-trifluoromethylquinoline (11a) and
6-chloro-2-methyl-4-trifluoromethylquinoline (12a)
These compounds were obtained as a brown solid (EtOH), a 23:77
mixture of 11a:12a, mp 106–109 8C, lit[15] mp 105–108 8C.
4-Methyl-2-trifluoromethylquinoline (2a) and 2-methyl-4-
trifluoromethylquinoline (3a)
These compounds were obtained as a yellow–brown liquid, a
28:72 mixture of 2a:3a, bp 116–120 8C (20 mmHg), lit[15] 116–120
8C (20 mmHg).
6-Methoxy-4-methyl-2-trifluoromethylquinoline (13a)
This compound was obtained as a yellowish-brown solid (EtOH),
13a, mp 84–86 8C, lit[15] mp 83–87 8C.
7-Ch1oro-4-methyl-2-trifluoromethy1quinoline (14a) and 7-ch1oro-
2-methyl-4-trifluoromethy1quino1ine (15a)
These compounds were obtained as a yellowish-brown liquid, a
46:54 mixture of 14a:15a, bp 151–155 8C (4 mmHg), lit[15] bp
151–154 8C (4 mmHg).
2,4-Ditrifluoromethylquinoline (2b)
This compound was obtained as a yellow–green liquid, 2b, bp
162–163 8C (20 mmHg), lit[15] bp 160–162 8C (20 mmHg).
4-Ethyl-2-trifluoromethylquinoline (2c) and 2-ethyl-4-
trifluoromethylquinoline (3c)
These compounds were obtained as a yellow liquid, a 90:10
mixture of 2c:3c, bp 159–162 8C (20 mmHg), lit[15] bp 160–162 8C
(20 mmHg).
7-Ch1oro-4-isopropyl-2-trifluoromethylquinoline (14f) and
7-chloro-2-isopropyl-4-trifluoromethylquinoline (15f)
These compounds were obtained as a yellow–brown liquid, a 1:1
mixture of 14f:15f, bp 163–165 8C (25 mmHg), lit[15] bp 165 8C
(25 mmHg).
2-t-Butyl-4-trifluoromethylquinoline (3d)
This compound was obtained as a tan solid (EtOH), 3c, mp
142–144 8C, lit[15] mp 142–143 8C.
6-Fluoro-4-methyl-2-trifluoromethy1quinoline (16a) and
6-fluoro-2-methyl-4-trifluoromethylquinoline (17a)
These compounds were obtained as a yellowish-brown solid
(EtOH), a 44:56 mixture of 16a:17a, mp 95–97 8C, lit[15] mp 95–
97 8C.
4-Phenyl-2-trifluoromethylquinoline (2e) and 2-phenyl-4-
trifluoromethylquinoline (3e)
These compounds were obtained as a yellow–brown liquid, a
30:70 mixture of 2e:3e, bp 192–193 8C (20 mmHg), lit[15] bp
193 8C (20 mmHg).
3,6-Difluoro-4-methyl-2-trifluoromethylquinoline (16g)
This compound was obtained as a yellowish-brown liquid, 16g,
bp 152–155 8C, lit[15] bp 152–155 8C.
3-Fluoro-4-methyl-2-trifluoromethylquinoline (2g)
This compound was obtained as a yellowish–brown liquid, 2g, bp
157–159 8C, lit[15] bp 158–160 8C.
8-Amino-4-isopropyl-2-trifluoromethylquinoline (18f)
This compound was obtained as a brownish-black solid (EtOH),
18f, mp 161–163 8C, lit[15] mp 160–162 8C.
8-Bromo-4-methyl-2-trifluoromethylquinoline (4a)
This compound was obtained as a brown solid (EtOH), 4a, mp
75–77 8C, lit[15] mp 75–77 8C.
5,7-Dimethoxy-4-methyl-2-trifluoromethylquinoline (19a) and
5,7-dimethoxy-2-methyl-4-trifluoromethylquinoline (20a)
These compounds were obtained as a yellowish-brown solid
(EtOH), a 74:26 mixture of 19a:20a, mp 95–97 8C. NMR:19a: 1H: d
2.7 (3H, s), 6.5 (d, J ¼ 6.6 Hz, 1H), 6.7 (lH, s), 6.9 (lH, d, J ¼ 6.7 Hz), 7.2
(lH, d, J ¼ 6.9 Hz). l3C: d 24, 114, 121 (CF3, q, 2JC—F ¼ 270 Hz), 143
(C—CF3, q, 2JC—F ¼ 31 Hz, C—CF3). 19F: d ꢁ67.2 (3F, s). 20a: 1H: d
2.6 (3H, s), 6.4 (lH, d, J ¼ 6.8 Hz), 6.6 (lH, s), 6.8 (lH, d, J ¼ 6.9 Hz), 7.1
(lH, d, J ¼ 6.7 Hz). l3C: d 24, 114, 124 (CF3, q, 1JC—F ¼ 272 Hz), 133
(C—CF3, q, 2JC—F ¼ 33 Hz). 19F:d ꢁ62.3 (3F, s). Analysis: calculated
for C13H12F3NO: C, 57.57, H, 4.46, N, 5.16. Found: C, 57.60, H, 4.42,
N, 5.15.
8-Chloro-4-methyl-2-trifluoromethylquinoline (5a) and
8-chloro-2-methyl-4-triftuoromethylquinoline (6a)
These compounds were obtained as a brown solid (EtOH), a 11:89
mixture of 5a:6a, mp 122–125 8C, lit[15] mp 122–124 8C.
8-Methoxy-4-methyl-2-trifluoromethylquinoline (7a) and
8-methoxy-2-methyl-4-trifluoromethylquinoline (8a)
These compounds were obtained as a yellowish solid (EtOH), a
56:44 mixture of 7a:8a, mp 75–77 8C, lit[15] mp 75–77 8C.
J. Phys. Org. Chem. 2009, 22 110–117
Copyright ß 2008 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc

QUINOLINE FORMATION VIA COMBES REACTION
Preparation of intermediates
Yamashkin, E. A. Oreshkina, Chem. Heterocycl. Compos. 2006, 42(6),
2006.
[10] J. A. Knight, H. K. Porter, P. K. Calaway, J. Am. Chem. Soc. 1944, 66,
1893.
[11] W. S. Johnson, E. Woroch, F. J. Mathews, J. Am. Chem. Soc. 1947, 69,
566.
4-Phenylimino-5,5,5-trifluoro-2-pentanone (II-2a),
4-phenylimino-1,1,1-trifluoro-2-pentanone (II-3a),
4-anilino-1,1,1-trifluoro-3-penten-2-one (II-3a0)
To a solution of 3.08 g (20.0 mmol) 1a in 100 ml absolute ethanol
is added 1.86 g (20.0 mmol) aniline and heated to 70 8C with
stirring for 4 h. The solvent was removed under reduced pressure
giving 4.03 g (88%) of a 30:65:5 mixture (determined by
[12] F. Misani, M. T. Bogert, J. Am. Chem. Soc. 1945, 67, 347.
[13] E. A. Steck, L. L. Hallock, A. J. Holland, L. T. Fletcher, J. Am. Chem. Soc.
1948, 70, 1012.
[14] E. A. Steck, L. L. Hallock, A. J. Holland, J. Am. Chem. Soc. 1946, 68,
pp. 129, 132, 380.
[15] J. C. Sloop, C. L. Bumgardner, W. D. Loehle, J. Fluor. Chem. 2002, 118,
135.
[16] W. S. Johnson, F. J. Mathews, J. Am. Chem. Soc. 1944, 66, 210.
[17] E. Roberts, E. E. Turnur, J. Chem. Soc. 1927, 1832.
[18] R. Huisgen, Annales 1949, 564, 16.
[19] J. L. Born, J. Org. Chem. 1972, 37, 3952.
[20] J. C. Sloop, C. L. Bumgardner, G. Washington, W. D. Loehle, S. S. Sankar,
A. B. Lewis, J. Fluor. Chem. 2006, 127, 780.
[21] C. Reichardt, Solvents and Solvent Effects in Organic Chemistry, 3rd
edn, (pp. 107, 329, 375) Wiley-VCH Verlag GmbH & Co., KGaA,
Weinheim, 2003.
[22] J. W. Bunting, J. P. Kanter, J. Am. Chem. Soc. 1993, 115, 11705.
[23] Y. Chiang, A. J. Kresge, P. A. Walsh, J. Am. Chem. Soc. 1982, 104, 6122.
[24] A. Gero, J. Org. Chem. 1954, 19, 1960.
[25] J. Reid, M. Calvin, J. Am. Chem. Soc. 1950, 72, 2948.
[26] J. C. Sloop, C. L. Bumgardner, J. Fluor. Chem. 1992, 56, 141.
[27] T. H. Lowry, K. S. Richardson, Mechanism and Theory in Organic
Chemistry, 3rd edn, (pp. 151, 318–320; 702–720; 726) Harper &
Row Publishers, New York, 1987.
19
integration of F NMR resonances) of II-2a:II-3a:II-3a0 as a
brownish oil. Chromatography of the crude reaction mixture
(20% EtOAC/hexane) resulted in isomerization of II-3a0 to II-3a.
NMR: II-2a: 1H: d 2.22 (s, 3H), 2.99 (s, 2H), 7.00–7.12 (m, 3H), 7.46 (t,
J ¼ 7.42 Hz, 2H). 13C:
d
22.0, 30.6, 118.3 (—CF3,
q, 1JC—F ¼ 275.1 Hz), 123.5, 128.6, 130.2, 152.1, 166.2 (C—CF3,
q, 2JC—F ¼ 34.2 Hz), 208.1. 19F: d: ꢁ81.5 (—CF3, s, 3F). II-3a: 1H: d
2.01 (s, 3H), 2.86 (s, 2H), 7.00–7.12 (m, 3H), 7.41 (t, J ¼ 7.40 Hz, 2H).
13C: d 22.2, 27.9, 117.8 (ꢁCF3, q, 1JC—F ¼ 269.2 Hz), 120.7, 126.9,
128.7, 150.1, 162.1 (C—CF3, q, 2JC—F ¼ 35.4 Hz), 208.1. 19F: d:
ꢁ83.1 (—CF3, s, 3F). II-3a0 (crude reaction mixture): 1H: d 5.7 (s,
1H), 11.4, (—NH, bs, 1H). 19F: d: ꢁ75.2 (—CF3, s, 3F). Analysis:
calulated for C11H10F3NO: C: 57.64%, H: 4.40%, N: 6.11%. Found:
C: 57.69%, H: 4.42%, N: 6.16%.
Acknowledgements
[28] E. B. Mulloch, R. Searby, R. Suschitzky, J. Chem. Soc. C 1970,
829.
[29] O. Kentaro, T. Adachi, M. Tomie, K. Kondo, I. Inoue, J. Chem. Soc. Perkin
Trans 1 1972, 173.
[30] K. A. Clayton, D. Black, J. B. Harper, Tetrahedron 2008, 64, 3183.
[31] R. Linderman, K. Kirollos, Tetrahedron Lett. 1989, 31, 2689.
[32] A. Bondi, J. Phys. Chem. 1964, 68, 441.
The author is indebted to Dr Gary Washington for the useful
kinetics discussions. The author thanks the USMA Faculty
Research Fund for providing financial support for this work.
[33] J. C. Sloop, B. Lechner, G. Washington, C. L. Bumgardner, W. D. Loehle,
W. Creasy, Int. J. Chem. Kin. 2008, 40(7), 370.
REFERENCES
[34] J. A. Hirsch, Concepts in Theoretical Organic Chemistry, (pp. 112–123;
143–144) Allyn & Bacon Inc., Boston, 1974.
[35] J. Hine, Physical Organic Chemistry, 2nd edn, (pp. 255–257.) McGraw-
Hill Book Co., Inc., New York, 1962.
[36] L. P. Hammett, Physical Organic Chemistry, (pp. 365–367), McGraw-Hill
book Co., Inc., New York, 1970.
[37] M. T. Tribble, J. G. Traynham, J. Am. Chem. Soc. 1969, 91, 379.
[38] H. C. Brown, Y. Okamoto, J. Am. Chem. Soc. 1958, 79, 4979.
[39] Correlation of log(k/k0) for Table 1, entry 18 was effected with
sþp þ so. This is consistent with the presence of two EDGs (1 ortho,
1 para) capable of simultaneous mesomeric interaction with a
growing positive charge in the rate determining annulation tran-
sition state. Correlation of log(k/k0) for o,p-anilines was effected
with sþm.
[1] G. S. Dow, T. N. Heady, A. K. Bhattacharjee, D. Caridha, L. Gerena, M.
Gettayacamin, C. A. Lanteri, N. Obaldia, III, N. Roncal, T. Shearer, P. L.
Smith, A. Tungtaeng, L. Wolf, M. Cabezas, D. Yourick, K. S. Smith,
Antimicrob. Agents Chemother. 2006, 50(12), 4132–4143.
[2] Y. Katayama, T. Ootsubo, S. Tsuda, Jpn. Kokai Tokkyo Koho 1993, 3.
[3] K. M. Muraleedharan, M. A. Avery, Eds, J. B. Taylor, D. J. Triggle,
Comprehensive Medicinal Chemistry II, (Vol. 7, p. 765) Elsevier Ltd.,
Oxford, 2006.
[4] Eds.: R. Filler, Y. Kobayashi, Biomedical Aspects of Fluorine Chemistry,
Kodansha, Tokyo, 1982.
[5] M. Conrad, L. Limpach, Bericht 1887, 20, 944 L-948.
[6] L. Knorr, Annales 1886, 236, 69.
[7] L. Knorr, Annales 1888, 245, 357.
[8] C. Combes, Chem. Ber. 1896, 29, 2456.
¨
[40] E. S. Amis, Solvent Effects on Reaction Rates and Mechanisms,
(pp. 65–67) Academic Press, New York, 1966.
[41] L. P. Hammett, J. Am. Chem. Soc. 1937, 59, 96.
[42] J. E. Leffler, J. Org. Chem. 1955, 20, 1202.
[9] The Friedlander and Pfitzinger methods use an o-carbonyl or
o-carboxy aniline, while the Skraup and Doebner-Miller syntheses
use an a,b-unsaturated carbonyl electrophile with aniline. See S. A.
www.interscience.wiley.com/journal/poc
Copyright ß 2008 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2009, 22 110–117