Palladium-catalysed synthesis of dibenzo[de,g]quinolines. A novel approach to the B-ring system of aporphine-related heterocycles

Article abstract of DOI:10.1002/(sici)1099-0690(199908)1999:8<1957::aid-ejoc1957>3.0.co;2-i

Dibenzo[de,g]quinolines 5 were formed by the palladium-catalysed heteroannulation of disubstituted alkynes and 1-iodo-10- (dimethylamino)phenanthrene (3c). Symmetric alkynes led to high levels of regioselectivity. These reactions constitute a new synthesis of the B-ring system of aporphine heterocycles.

Full text of DOI:10.1002/(sici)1099-0690(199908)1999:8<1957::aid-ejoc1957>3.0.co;2-i

FULL PAPER
Palladium-Catalysed Synthesis of Dibenzo[de,g]quinolines.
A Novel Approach to the B-Ring System of Aporphine-Related Heterocycles
Anne-Elisabeth Gies,[a] Michel Pfeffer,*[a] Claude Sirlin,[a] and John Spencer[a][°]
Keywords: Aporphines / Nitrogen heterocycles / Palladium / Homogeneous catalysis / Dibenzo[de,g]quinolines
Dibenzo[de,g]quinolines 5 were formed by the palladium-
catalysed heteroannulation of disubstituted alkynes and 1-
iodo-10-(dimethylamino)phenanthrene (3c). Symmetric alky-
nes led to high levels of regioselectivity. These reactions
constitute new synthesis of the B-ring system of
aporphine heterocycles.
a
served in the insertion of disymmetric alkynes and the in
situ N-demethylation leading to the neutral six-membered
heterocycles 2.
Aporphines are isoquinoline-related alkaloids with di-
verse therapeutic or physiological properties and occur in
numerous plants. For example, Boldine (Peumus boldus) has
choleretic and cholagogue properties;[5a] apomorphine, a
nonnatural aporphine, is used as an antiparkinsonism ag-
ent,[5b] while several others have potential uses in cancer
treatment[5c] (Scheme 2).
Introduction
The metallation of aromatic tertiary amines has been the
subject of intensive investigation, notably in palladium(II)
chemistry[1] and synthetic applications of the resulting
cyclopalladated complexes have been explored. In particu-
lar, reactions of these complexes with disubstituted alkynes
have been developed into attractive and powerful methods
for heterocycle synthesis.[2] Whereas intensive studies have
been directed towards the palladium-mediated synthesis of
five- and six-membered N-heterocycles starting from pri-
mary or secondary amines containing an ortho iodo sub-
stituent,[3] only a few examples involving tertiary amine
substrates have been reported.
Recently, we embarked on a study aimed at extending
the synthetic potential of the reaction of cyclopalladated
aromatic tertiary amines with alkynes with a view to pre-
paring heterocyclic compounds with potential biological ac-
tivity. We showed that the heteroannulation of 8-iodo-1-(di-
methylamino)naphthalene (1) with a number of disubsti-
tuted alkynes catalysed by the palladium complex 1b led to
the benzo[d,e]quinolines 2 (Scheme 1).[4]
Scheme 2
These alkaloids are usually prepared according to a mul-
tistep protocol employing a BischlerϪNapieralski conden-
sation followed by a Pschorr cyclisation[6a] or a non-phe-
nolic oxidative coupling[6b] of the intermediate benzyliso-
quinoline I (Scheme 3).
Scheme 1
Key features of this protocol, which involves the forma-
tion of both a CϪC and a CϪN bond, include the favour-
able yields obtained for alkynes containing aryl or electron-
withdrawing groups, the high level of regioselectivity ob-
Scheme 3
The benzo[d,e]quinolines, 2 depicted above effectively
constitute the ABC fragment of aporphines.[6c] Therefore
the route presented in Scheme 1, which involves the con-
struction of the B fragment, merited further investigation as
a potential synthetic strategy for the preparation of closely
related analogues of these alkaloids by the heteroannulation
[a]
`
´
Laboratoire de Syntheses Metallo-Induites, UMR 7513 CNRS
´
Universite Louis Pasteur, 4 rue Blaise Pascal, F-67000 Stras-
bourg, France
Fax: (internat.) ϩ33 3 8845 4667
E-mail: pfeffer@chimie.u-strasbg.fr
[°] Current address: The Thrombosis Research Institute
Emmanuel Kaye Building Chelsea, UK, SW3 6 LR
Eur. J. Org. Chem. 1999, 1957Ϫ1961
WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
1434Ϫ193X/99/0808Ϫ1957 $ 17.50ϩ.50/0
1957

A.-E. Gies, M. Pfeffer, C. Sirlin, J. Spencer
FULL PAPER
of 1-iodo-10-(dimethylamino)phenanthrene (3c) (Scheme reaction medium turned black, most likely because of the
4).
partial formation of metallic palladium, then became ap-
parently more homogeneous as the black precipitate ap-
peared to redissolve after 0.1 h with the formation of a clear
yellow solution. Addition of 3c to this solution did not alter
its appearance and the catalytic formation of the heterocy-
cle took place without any visible change in the solution.
However, a decrease in temperature led to the irreversible
deposition of palladium black with a consequent loss of
catalytic activity. This observation is in marked contrast to
previous observations[4] where the formation of 2 occurred
in the presence of palladium black.
Scheme 4
In this paper we report the reaction between 3c and
This behaviour seems to indicate that the novel heterocy-
cle 5 obtained may interact with the palladium(0) formed,
thus preventing its precipitation as palladium metal. How-
ever, every attempt to characterise such species failed. Note
that the role of 3c is obviously to regenerate 4 from this
latter soluble Pd0 species, as we have shown that in the ab-
sence of 4, no heterocyclisation occurs (see below, Table 2,
entry 5).
alkynes in the presence of a palladium catalyst 4 (Scheme
5). We chose this system based upon our previous obser-
vations made with the related system 1.[4]
The catalytic heterocyclisation of 1 could be achieved
with alkynes bearing either electron-releasing or electron-
withdrawing substituents. No reaction was observed when
3c was treated with electron-rich alkynes such as 2,2-di-
methyl-4-pentyne or 3-hexyne. Even with diphenylacety-
lene, good yields of the expected heterocycle 5e could only
be obtained when 4 was treated with one equivalent of al-
kyne per palladium atom in a stoichiometric reaction. In
this case, carrying out the reaction as depicted in Scheme
5, i.e. with 10% of 4 did not lead to reproducible results:
we found in one case only a 50% conversion of 3c to 5e
after 21 h. However, this result was achieved on one oc-
casion only and thus the stoichiometric route for the pre-
paration of 5e was preferred.
Scheme 5
Results and Discussion
The starting material, 9-phenanthrylamine 3a, which is
Previously, we have noted that “electron-poor” alkynes
both commercially available and accessible by a multistep afford the best yields in the heterocyclisation of 1, this was
synthesis,[7] was obtained in 80% yield by lithiation of 9- also the case in our current study; dimethylacetylene dicarb-
bromophenanthrene followed by treatment with trimethyl- oxylate yielded 32% of 5d starting with 3c in the presence
silylmethyl azide (TMSMA)[8] and subsequent hydrolysis. of 4 (10%).
Dimethylation gave the known 3b,[9] which was converted
The best yields of compounds 5 were obtained with non-
into the iodo-bridged cyclopalladated complex 4[9] by a symmetrical internal alkynes in which one substituent is
CϪH activation process, or into 3c by a chelation con- electron-withdrawing. The comparison of three alkynes,
trolled lithiation followed by the addition of iodine.
ethyl phenylpropynoate, 4-phenyl-3-butyn-2-one, and phe-
Our initial attempts at catalysis focused on directly apply- nylpropynal (Table 1) showed that the best conversion was
ing the conditions developed earlier for the synthesis of 2 attained when the alkyne was substituted by the most elec-
(see Scheme 1) in which the best palladium catalyst was tron-withdrawing group of the series (i.e. ϪCHO).
found to be the iodo-bridged cyclopalladated (dimethylami-
no)naphthalene compound 1b in refluxing PhCl. The use
Table 1. Results for the heterocyclisation of 3c and alkynes substi-
tuted with different electron-withdrawing groups on use of 5% pal-
ladium
of other conventional palladium catalysts were (in general)
unsuccessful, although in the case of PhCϵCCOOEt the
corresponding heterocycle 2a could be synthesised in 70%
yield using the Pd(OAc)2/PPh3 catalytic system.[10]
Entry R1
R2
Product Reflux (h) Conversion (%)[a]
In the present case, the dimer
4 abbreviated as
{(dmpa)Pd(µ-I)}2, was also found to be the best catalyst
for the reaction depicted in Scheme 5, although the precise
reaction conditions had to be modified in order to achieve
the best results. Thus we found that 4 had to be added to a
1
2
3
COOEt Ph
5a
5b
5c
100
100
58
81
100
100
COMe
CHO
Ph
Ph
[a]
Measured by 1H NMR vs. Ph3CH as the internal reference Ϫ
solution of the alkyne in PhCl at reflux temperature. The see Experimental Section.
1958
Eur. J. Org. Chem. 1999, 1957Ϫ1961

Palladium-Catalysed Synthesis of Dibenzo[de,g]quinolines
FULL PAPER
High levels of regioselectivity were observed when sym- was noted and different products observed when 4 was ad-
metric alkynes were employed. In accord with previous re- ded to a reaction mixture containing the alkyne and 3c. In
1
sults, we conclude from H NMR that the phenyl group in this instance the reaction mixture darkened and remained
5a, 5b, and 5c is on the carbon adjacent to the nitrogen so throughout the reaction, and in addition to 5a (22%),
atom[11] i.e. C-5. In N-methyl-2,3-diphenyldibenzo[de,g]qui- two other products were characterised (Scheme 6). The first
noline 5e an important shielding of the H-3 proton was of them was easily identified as the N-methylated derivative
1
observed due to the adjacent Ph group, whereas in 5a, 5b, 6 (20%), and the second displayed a H-NMR spectrum
and 5c such a high-field shift was absent. Therefore we con- very similar to that of 5a with the exception that no NϪMe
cluded that the phenyl group was not attached to C-4 but group was observed. This product was deduced to be the
to C-5. Analysing the reaction between 3c and ethyl phenyl- carbocyclic derivative 7 (23%) on the basis of both a High-
propynoate in closer detail allowed us to obtain interesting Resolution FAB mass spectrum and a ROESY NMR
information about both the likely reaction pathway and the experiment, which allowed us to conclude that the phenyl
optimisation of the reaction conditions. For a given catalyst unit was indeed at the carbon atom adjacent to C-6.
concentration, the concentration of 5a increases linearly as
a function of time, over a given period the duration of
which is a function of the concentration of 4. The resulting
rate constant enabled us to calculate the corresponding
TOF; the concentration of 5a attained (at the end of this
period) is evidently related to the maximum conversion of
3c to 5a.
Scheme 6
Table 2. Catalytic heterocyclisation of 3c and PhCϵCCOOEt to 5a
Entry 4 (mmol-%)
Reaction
time (h)
Max conversion
(%)
TOF
(hϪ1
)
Conclusion
1
2
3
4
5
10
5
100
100
50
95
81
55
34
0
0.07
0.10
1.10
0.95
0
A palladium-catalysed route towards aporphine-related
dibenzo[de,g]quinolines has been developed. Although this
reaction is closely related to the protocol employed for the
synthesis of 2 (see Scheme 1), it is clearly not a simple ex-
tension of this work. The reaction is thought to proceed by
an initial stoichiometric-like insertion of the alkyne in the
PdϪC bond of 4, followed by reductive elimination with
CϪN bond formation to afford 5 and palladium(0). The
latter undergoes oxidative addition to 3c with the regener-
ation of 4 and continuation of the catalytic cycle. The de-
methylation reaction that occurs synchronously to the inser-
tion of the alkynes is a very important feature of this heter-
ocyclisation process. In the related reaction involving 1
(Scheme 1) we were able to show that MeI was produced,
whereas when the corresponding N,N-diethyl derivative of
1 was used, EtI was formed.[9] Since the alkyl halides
formed in both cases are primary alkyl iodides, we favour
an SN2 like mechanism for the dealkylation process.
1
0.6
0
60
127
TOF and maximum conversion are reported vs. the
amount of palladium used (mmol-%) and Figure 1 indicates
that maximum catalytic activity should be observed for ca.
2.5% of 4 leading to an optimum yield of 70% of the hetero-
cycle 5a. Conversely, with higher amounts of 4 a lower ac-
tivity was observed, although an almost quantitative yield
of 5a resulted. This suggests that when higher concen-
trations of 4 are employed, a small amount of catalyst is
required for the reaction, but it may be replaced by the
“stock” of 4 present in solution thus leading to higher con-
versions.
The total synthesis of true aporphines with this method-
ology as a key step would necessitate (i) the preliminary
synthesis of phenanthrylamines in which the phenanthrene
unit has a closer structural analogy to those found in the
natural alkaloids,[12] and (ii) to establish the conditions for
the removal of the substituents at C-4 and C-5. From this
latter point of view, the diester 5d is an ideal candidate.
Experimental Section
General: Reactions were carried out under a dry nitrogen atmos-
phere. All solvents were purified by conventional distillation tech-
niques with the exception of commercial grade chlorobenzene
(PhCl) which was used as received. Palladium reagents were pre-
Figure 1
The reaction conditions proved to be crucial in the case
of PhCϵCCOOEt, since a totally different reaction course pared from PdCl2 according to the literature.[13] Ϫ NMR experi-
Eur. J. Org. Chem. 1999, 1957Ϫ1961
1959

A.-E. Gies, M. Pfeffer, C. Sirlin, J. Spencer
FULL PAPER
ments were performed on Bruker SY 200, AC 300, AM 400, and
of the methine proton of Ph3CH (δ ϭ 5.50) employed as internal
DRX 500 spectrometers. For the 13C-NMR spectra, the nature of standard. Ϫ “Standard conditions” refer to a Schlenk tube
the different carbon atoms was specified using a Dept 135 experi-
equipped with a magnetic stirrer and a condenser in a thermostated
ment when possible. The spectra were measured in CDCl3 and oil bath at 150°C.
chemical shifts are given in ppm relative to TMS. Ϫ Mass spectra
Ethyl N-Methyl-2-phenyldibenzo[de,g]quinoline-3-carboxylate
and microanalysis were performed by the Service Commun de Mic-
(5a): Ethyl phenylpropynoate (124 mg, 0.71 mmol) in PhCl (20 mL)
was heated to reflux under the standard conditions (see above). At
reflux, a solution of 4 (27 mg) in PhCl (20 mL) was rapidly added,
causing the colour of the reaction mixture to turn from light yellow
to black. The iodo compound 3c (204 mg, 0.59 mmol) in PhCl
(20 mL) was then added. A few minutes later, the colour of the
reaction mixture became bright yellow and remained so for the
remainder of the reaction (7.5 h). Removal of the solvent in vacuo,
followed by chromatography of the resulting black oily residue
(Et2O:hexane, 10:90) gave bands of unchanged 3c followed by the
heterocycle 5a (Et2O/hexane, 30:70), as golden flakes (115 mg, 47%)
´
roanalyse and the Laboratoire de Spectrometrie de Masse de Stras-
bourg. Ϫ The chromatographic separations were performed on
Merck silica gel Si 60 (40Ϫ63 µm) unless otherwise specified. Ϫ
Ph3CH (used as internal reference) and n-butyllithium (nBuLi,
1.6  in hexanes) were commercial products. Alkynes were ob-
tained from commercial sources, trimethylsilylmethylazide
(TMSMA),[8] 3b[9] and 4[9] were synthesised according to known
procedures.
9-Phenanthrylamine (3a) (CAUTION: Aromatic amines are known
to be mutagenic and carcinogenic agents. Therefore the following
synthesis should only be undertaken in well-ventilated room with ap-
propriate safety clothes): nBuLi (31.3 mmol, 19.5 mL) was added
dropwise to a solution of 9-bromophenanthrene (6.7 g, 26.1 mmol)
in a 1:9 Et2O/n-hexane mixture (50 mL) in a 500 mL round-bot-
tomed Schlenk flask equipped with a large magnetic stirrer. After
1.5 h, the white precipitate was collected by filtration, washed with
n-hexane (2 ϫ 10 mL), dried in vacuo, and suspended in Et2O.
TMSMA (1.2 equiv.) was added, and after 1.5 h, treated with 6 
HCl (100 mL). The sticky pale pink precipitate formed was rapidly
collected (light-sensitive), and ether (200 mL) and a saturated solu-
tion of KOH (50 mL) was added. The mixture was stirred vigor-
ously, the organic phase separated, and the aqueous phase re-ex-
tracted with several 50 mL portions of ether until the orange amine
colouration disappeared. The original acidic aqueous layer was
neutralised and extracted with ether as described above. The com-
bined organic phases were concentrated in vacuo, redissolved in
CH2Cl2, dried with CaCl2, and treated with activated charcoal. Fil-
tration through Celite and concentration of the filtrate afforded
4.0 g of 3a (80%), yellowish pink solid. Spectral data (1H and 13C
NMR) were consistent with those from a commercial sample (Ald-
rich).
1
and H-NMR data in accord with previous observations[9]. Ϫ 13C
NMR (125 MHz): δ ϭ 167.7, 146.0, 139.9, 135.8, 133.8, 132.0,
130.8, 129.1, 125.2, 110.2 (quaternary carbons), 129.1 (2C), 129.0,
128.7 (2C), 128.2, 127.1, 126.9, 123.3, 122.6, 118.3, 116.8, 100.0
(CH), 60.3 (CH2), 37.3, 13.5 (CH3).
3-Acetyl-N-methyl-2-phenyldibenzo[de,g]quinoline (5b): Same con-
ditions as for 5a. 24% yield (28 mg) from 4-phenyl-3-butyn-2-one
(49 mg, 0.34 mmol), 4 (13 mg), and 3c (101 mg, 0.29 mmol), reac-
tion time 6 h. 5b was obtained as yellow crystals from CH2Cl2/n-
hexane (99:1) at Ϫ20°C. Ϫ 1H NMR (500 MHz): δ ϭ 8.36 (d, 1
H, J ϭ 8.2 Hz), 8.19 (d, 1 H, J ϭ 8.2 Hz), 7.62 (d, 1 H, J ϭ 8.0 Hz),
7.50Ϫ7.34 (m, 8 H), 7.17 (d, 1 H, J ϭ 7.7 Hz), 6.53 (s, H7), 3.06
(s, 3 H, NMe), 1.77 (s, 3 H, COMe). Ϫ 13C NMR (125 MHz): δ ϭ
203.2, 143.5, 140.2, 135.0, 133.8, 132.3, 130.6, 129.9 (2C), 129.8,
129.2 (2C), 128.3, 127.2, 126.9, 125.4, 125.1, 123.4, 122.7, 119.4,
118.5, 116.6, 100.1, 37.6, 31.6. Ϫ C25H19NO (349.15): C 85.93, H
5.48, N 4.01; found C 85.81, H 5.50, N 3.97.
Dimethyl N-Methyldibenzo[de,g]quinoline-2,3-dicarboxylate (5d):
Same conditions as for 5a. 32% yield (35 mg) from dimethyl ace-
tylenedicarboxylate (66 mg, 0.47 mmol), 4 (13 mg), and 3c (101 mg,
0.29 mmol), reaction time 3.5 h. 5d was obtained as orange crystals
from CH2Cl2/n-hexane (99:1) at Ϫ20 °C overnight, in with 1H-
NMR data in accord with those previously reported[9]. Ϫ 13C
NMR (50 MHz): δ ϭ 166.2, 165.1, 144.4, 138.0, 133.0, 131.8,
129.1, 128.3, 127.4, 127.1, 125.9, 125.4, 124.3, 122.6, 119.5, 119.5,
104.9, 101.3, 53.1, 51.9, 37.3.
10-(Dimethylamino)-1-iodophenanthrene (3c): nBuLi (15.5 mL,
25 mmol) was added to 1:1 Et2O/n-hexane (20 mL) solution of 3b
(3.6 g, 16.4 mmol) and stirred for 4 days at room temperature. The
off-white precipitate was collected by filtration, washed with n-pen-
tane (2 ϫ 10 mL) then suspended in ether (20 mL). An ethereal
solution (40 mL) of iodine (4.5 g, 17.7 mmol) was slowly added and
the reaction mixture stirred for one hour. After washing with a
saturated Na2S2O3 solution (2 ϫ 50 mL), the organic phase was
dried with MgSO4 then treated with activated charcoal. Filtration
and concentration of the filtrate afforded 3.45 g of 3c (61%) as a
brown oil. Yellow needles were obtained from CH2Cl2/n-hexane
(99:1). Ϫ 1H NMR (300 MHz): δ ϭ 8.72Ϫ8.69 (m, 1 H), 8.55Ϫ8.52
(m, 1 H), 8.36Ϫ8.34 (m, 1 H), 7.80Ϫ7.78 (m, 1 H), 7.57Ϫ7.52 (m,
2 H), 7.47 (s, 1 H), 7.24Ϫ7.19 (m, 1 H), 2.74 (s, 6 H). Ϫ 13C NMR
(75 MHz): δ ϭ 148.0, 141.9, 133.6, 132.4, 128.2, 127.5, 127.3, 127.1,
125.2, 123.6, 122.6, 115.7, 89.5, 44.5. Ϫ C16H14IN (315.11): C
55.35, H 4.06, N 4.03; found C 55.91, H 4.20, N 3.86.
N-Methyl-2,3-diphenyldibenzo[de,g]quinoline (5e): Compound 5e
was obtained from diphenylacetylene (83 mg, 0.46 mmol) by treat-
ment with an approximately stoichiometric amount of 4 (200 mg,
0.44 mmol) in PhCl (30 mL) at reflux for 3 h. 5e was obtained as
greenish yellow crystals from hot CH2Cl2 (261 mg, 77%). Ϫ 1H
NMR (400 MHz): δ ϭ 8.37 (d, 1 H, J ϭ 8.3 Hz), 8.14 (d, 1 H, J ϭ
7.8 Hz), 7.65 (d, 1 H, J ϭ 8.1 Hz), 7.47Ϫ7.43 (m, 1 H), 7.36Ϫ7.32
(m, 1 H), 7.26Ϫ7.07 (m, 11 H), 6.61 (d, 1 H, J ϭ 7.5 Hz), 6.50 (s,
1 H), 3.03 (s, 3 H). Ϫ 13C NMR (100 MHz): δ ϭ 142.3, 141.0,
138.0, 136.2, 135.6, 134.5, 132.2, 131.8, 130.2 (2C), 128.3 (2C),
128.2 (3C), 127.9, 127.8, 127.0, 126.7, 126.3, 125.6, 124.7, 122.7,
122.5, 117.8, 117.7, 116.9, 98.1, 37.4. Ϫ C29H21N (383.17): C 90.83,
H 5.52, N 3.65; found C 90.23, H 5.51, N 3.58.
Catalysis: The amount of 3c indicated was calculated to take into
account the corresponding tertiary amine present in the catalyst 4.
The data reported in Table 2 were obtained by taking several
samples (2 mL) of the reaction mixture during the reaction time.
Each sample was dried in vacuo and dissolved in CDCl3 prior to
NMR analysis. The concentration of the various species was calcu-
lated from their normalised integrals (for 3c, δ ϭ 2.74, 6 H, NMe2;
2-Formyl-N-methyl-3-phenyldibenzo[de,g]quinoline (5c): Phenylpro-
pynal (33.2 mg, 0.3 mmol) was added to a 0.04  solution of
Ph3CH in PhCl (25 mL) and it was heated to reflux in the standard
conditions. The volume was adjusted to 40 mL with PhCl, 2.5 mL
of a 0.0018  solution of 4 in PhCl and 5 mL of a 0.03  solution
for 5, δ ഠ 3, 3 H, NMe and δ ഠ 6.5, 1 H, H7) vs. the integral of 3c in PhCl. After 58 h at reflux, the same isolation procedure as
1960 Eur. J. Org. Chem. 1999, 1957Ϫ1961

Palladium-Catalysed Synthesis of Dibenzo[de,g]quinolines
FULL PAPER
before gave the pure heterocycle 5c (45 mg, 81%) as red crystals
ane, 30:70).gave yellow crystals (257 mg, 71%) on concentration.
from CH2Cl2 after storage at Ϫ20°C overnight. Ϫ 1H NMR 1H-NMR data were in agreement with those previously reported.[4]
(500 MHz): δ ϭ 9.15 (s, 1 H), 9.02 (d, 1 H, J ϭ 7.8 Hz), 8.44 (d, 1
[1]
H, J ϭ 7.9 Hz), 8.33 (d, 1 H, J ϭ 8.3 Hz), 7.68Ϫ7.39 (m, 9 H),
6.79 (s, 1 H), 3.11 (s, 3 H). Ϫ 13C NMR (125 MHz): δ ϭ 190.0,
161.1, 137.9, 133.0, 132.6, 131.2, 129.9, 129.5 (2C), 129.5, 129.4,
129.3, 128.7, 127.5, 127.2, 126.5, 124.8, 124.7, 122.6, 119.6, 119.3,
112.9, 103.5, 37.5. Ϫ C24H17NO (335.13): C 85.95, H 5.11, N 4.18;
found C 85.55, H 5.07, N 4.10.
A. D. Ryabov, Chem. Rev. 1990, 90, 403Ϫ424.
[2]
[2a] M. Pfeffer, Recl. Trav. Chem. Pays-Bas 1990, 109, 567Ϫ576.
[2b]
Ϫ
M. Pfeffer, Pure & Applied Chem. 1992, 64, 335Ϫ342. Ϫ
[2c] J. Spencer, M. Pfeffer in: Advances in Metal-Organic Chemis-
try (Ed.: L. S. Liebeskind), JAI Press, London, 1998, vol. 6,
103Ϫ144.
[3] [3a]
R. C. Larock, E. K. Yum, J. Am. Chem. Soc. 1991, 113,
[3b]
6689Ϫ6690. Ϫ
K. R. Roesch, R. C. Larock, J. Org. Chem.
[3c]
Ethyl 5-Phenylacephenanthrylene-4-carboxylate (7): The same pro-
cedure as for 5c was applied with ethyl phenylpropynoate (60 mg,
0.3 mmol) except that 3c (85.4 mg, 0.25 mmol) was added prior to
4 [1 mL of a 11.5 mg solution in PhCl (10 mL)]. After 124 h reflux,
the solvent was removed in vacuo and the residue chromatographed
(200 mL Et2O/n-hexane, 10:90 then 200 mL Et2O/n-hexane, 15:85)
to afford 7 (Rf ϭ 0.43, yellow solid, 20 mg, 23%) and 6 (Rf ϭ 0.32,
brown wax, 10 mg, 20%) and finally 5a (Rf ϭ 0.32, 21 mg, 22%).
1998, 63, 5306Ϫ5307. Ϫ
R. C. Larock, E. K. Yum, M. D.
Refvik, J. Org. Chem. 1998, 63, 7652Ϫ7662 and refs. cited ther-
ein.
[4]
M. Pfeffer, N. Beydoun, Synthesis 1990, 729Ϫ732.
[5] [5a]
R. Hansel, Phytopharmaka, 2nd edn., Springer-Verlag,
[5b]
Berlin, 1991, p.186Ϫ191. Ϫ
G. L. Gessa, G. U. Corsini in:
Apomorphine and Other Dopaminomimetics, Raven Press, New
York, 1981, Vol. 1 Basic Pharmacology, Vol. 2 Clinical Pharma-
cology. Ϫ [5c] K. Likhitwitaywuid, C. K. Angerhofer, H. Y. Chai,
J. M. Pezzuto, G. A. Cordell, N. Ruangrungsi, Lloydia 1993,
56, 1468Ϫ1478.
1
Ϫ H NMR (500 MHz): δ ϭ 8.70 (d, 1 H, J ϭ 8.1 Hz), 8.47 (d, 1
[6] [6a]
´
F. Roblot, R. Hocquemiller, A. Cave, Bull. Soc. Chim. Fr.
H, J ϭ 8.1 Hz), 8.27 (d, 1 H, J ϭ 6.9 Hz), 8.03 (s, 1 H), 8.00 (ddd,
1 H, J ϭ 8.0, 0.6, 0.6 Hz), 7.80Ϫ7.74 (m, 2 H), 7.66Ϫ7.62 (m, 3
H), 7.56Ϫ7.49 (m, 3 H), 4.33 (q, 2 H), 1.27 (t, 3 H). Ϫ HRMS
(FAB); m/z: 350.1312. Ϫ C25H18O2 (350.4170).
[6b]
1990, 127, 258Ϫ267. Ϫ
Y. Landais, J. P. Robin, Tetrahedron
1992, 48, 7185Ϫ7196. Ϫ [6c] I. W. Southon, J. Buckingham, Dic-
tionary of Alkaloids, Chapman and Hall, New-York, 1989.
[7] [7a]
M. A. Goldberg, E. P. Ordas, G. Carsh, J. Am. Chem. Soc.
[7b]
1947, 69, 260Ϫ262. Ϫ
L. F. Fieser, R. P. Jacobsen, C. C.
1
N-Methylphenanthrylamine (6): Ϫ H NMR (300 MHz): δ ϭ 8.71
Price, J. Am. Chem. Soc. 1936, 58, 2163Ϫ2166.
K. Nishiyama, N. Tanaka, J. Chem. Soc. Chem. Commun.
1983, 1322Ϫ1323.
[8]
[9]
(d, 1 H, J ϭ 7.7 Hz), 8.54 (d, 1 H, J ϭ 8.2 Hz), 7.89 (d, 1 H, J ϭ
8.5 Hz), 7.74Ϫ7.61 (m, 3 H), 7.52Ϫ7.37 (m, 2 H), 6.78 (s, 1 H),
4.43 (br s, 1 H), 3.11 (s, 3 H). Ϫ HRMS (FAB); m/z: 207.1049. Ϫ
C15H13N (207.1048).
M. Pfeffer, N. Sutter-Beydoun, A. de Cian, J. Fischer, J. Or-
ganomet. Chem. 1993, 453, 139Ϫ146.
[10]
Under the same conditions, however, 3c only afforded a 1:1
mixture of both starting material and product 6a after 48 h in
refluxing N-methylpyrrolidine.
Alternative Synthesis of Ethyl N-Methyl-2-phenylbenzo[d,e]quino-
line-3-carboxylate (2a): An NMP suspension (40 mL) of 1 (320 mg,
1.1 mmol), Pd(OAc)2 (13.4 mg, 0.06 mmol), ethyl phenylpropyno-
ate (222 mg, 1.3 mmol), Na2CO3 (552 mg, 5.2 mmol), LiCl (44 mg,
1.0 mmol), and PPh3 (14 mg, 0.05 mmol) was heated for 21 h at
130°C. After cooling to room temperature, water was added
(50 mL) and the product extracted with ether (5 ϫ 25 mL). The
combined organic phases were washed with water (5 ϫ 25 mL),
dried with MgSO4, filtered, and concentrated to yield a brown oil.
Purification by column chromatography (alumina: CH2Cl2/n-hex-
[11] [11a]
J. Spencer, M. Pfeffer, A. DeCian, J. Fischer, J. Org. Chem.
1995, 60, 1005Ϫ1012. Ϫ [11b] M. Pfeffer, M. A. Rotteveel, G. Le
[11c]
Borgne, J. Fischer, J. Org. Chem. 1992, 57, 2147Ϫ2154. Ϫ
J. Spencer, M. Pfeffer, N. Kyritsakas, J. Fischer, Organometallics,
1995,14, 2214Ϫ2224.
[12]
[13]
A.-E. Gies, M. Pfeffer, J. Org. Chem. 1999.
H. M. Colqhoun, J. Holton, D. J. Thompson, M. V. Twigg in
New Pathways for Organic Synthesis: Practical Applications of
Transition Metals, Plenum Press, New York & London, 1984.
Received February 12, 1999
[O99095]
Eur. J. Org. Chem. 1999, 1957Ϫ1961
1961