A novel Birch reduction of aromatic compounds using aqueous titanium trichloride: Anions of trans-10b,10c-dimethyl-2,7,10b,10c-tetrahydropyrene

Article abstract of DOI:10.1016/S0040-4039(02)02785-5

The Birch reduction of 10b,10c-dimethyl-10b,10c-dihydropyrene 1 and anthracene could be predicted on the basis of their reduction potentials and achieved readily with aqueous titanium trichloride in near quantitative yields. Controlled reduction of a nitro group could be achieved under these conditions with the aromatic hydrocarbon remaining intact. The anion derived from the hexane obtained from reduction of 1 provided synthetic routes to derivatives of 1 inaccessible from direct substitution reactions of 1. Oxidative dimerization of the anion led to the formation of a series of interesting products.

Full text of DOI:10.1016/S0040-4039(02)02785-5

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 1271–1274
A novel Birch reduction of aromatic compounds using aqueous
titanium trichloride: anions of
trans-10b,10c-dimethyl-2,7,10b,10c-tetrahydropyrene
Jianping Jiang and Yee-Hing Lai*
Department of Chemistry, National University of Singapore, 3 Science Drive 3, Singapore 117543
Received 16 October 2002; revised 28 November 2002; accepted 6 December 2002
Abstract—The Birch reduction of 10b,10c-dimethyl-10b,10c-dihydropyrene 1 and anthracene could be predicted on the basis of
their reduction potentials and achieved readily with aqueous titanium trichloride in near quantitative yields. Controlled reduction
of a nitro group could be achieved under these conditions with the aromatic hydrocarbon remaining intact. The anion derived
from the hexane obtained from reduction of 1 provided synthetic routes to derivatives of 1 inaccessible from direct substitution
reactions of 1. Oxidative dimerization of the anion led to the formation of a series of interesting products. © 2003 Elsevier Science
Ltd. All rights reserved.
The applications of McMurry’s reagents involving low-
valent titanium species in reduction and reductive cou-
pling reactions of numerous organic functional groups
are well documented.1 Unlike the low-valent titanium
species which are usually generated in situ, the Ti(III)
ion is commercially available as titanium trichloride in
aqueous solution. The Ti(III) ion is a mild reducing
agent at a low pH, but the reducing power of the
Ti(III)/Ti(IV) system is strongly pH dependent and
increases markedly in strongly basic medium.2 In an
acidic solution, E=(0.100−0.1182pH+0.0591) log[TiO2+
/Ti3+] where Ti(IV) is present as soluble TiO2+. In a
basic medium, E=(0.029−0.2364pH−0.0591)log[Ti3+]
where insoluble TiO2 is formed, and the reducing power
of the Ti(III)/Ti(IV) system could in fact be more
negative than −2.5 V.3 10b,10c-Dimethyl-10b,10c-dihy-
dropyrene 1 and its derivatives form the most exten-
sively studied family of aromatic annulenes.4 Its
reduction to the hexaene 2 is generally effected by a
Birch reduction of 15 or reduction of the quinone 3 by
a mixture of lithium aluminum hydride and aluminum
chloride.6 These reactions however have to be carried
out in anhydrous conditions and at a low temperature
(−78°C). The yield of 2 is only moderate (about 50%).
It was reported that the first reduction potential of 1 is
−2.22 V in DMF.7 Thus reduction of 1 to give 2 could,
in principle, be achieved by Ti(III) under certain condi-
tions. Although hexaene 2 was used as a non-aromatic
model in the numerous studies of the diatropicity of 1
and its derivatives,4 its chemical properties have rarely
been investigated. In this paper we wish to report a
simple Birch-like reduction of 1 to 2 by titanium
trichloride and the chemical behavior of the monoanion
and dianion derived from 2.
Treatment of 1 with commercially available aqueous
titanium trichloride (about 15% in 10% HCl) in THF at
room temperature did not result in any reaction. Pro-
gressive addition of NaOH indicated a gradual increase
in the formation of 2 on the basis of TLC analyses.
When the reaction was carried out in a strongly basic
Keywords: titanium trichloride; Birch reduction; dihydropyrene;
anion; oxidative coupling.
* Corresponding author. Tel.: +65-6874-2914; fax: +65-6779-1691;
e-mail: chmlaiyh@nus.edu.sg
0040-4039/03/$ - see front matter © 2003 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)02785-5

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
1272
J. Jiang, Y.-H. Lai / Tetrahedron Letters 44 (2003) 1271–1274
reaction, the aminohexaene 6 was isolated in a 95%
medium using THF/CH3OH (1:1 v/v) as a mixed sol-
vent system, a quantitative yield of 2 was isolated after
30 min of reaction time.† To the best of our knowledge
this is the first example to show that titanium chloride
reduces an aromatic hydrocarbon directly in a Birch-
like manner. The presence of CH3OH seemed to
improve the solubility of NaOH in the solvent system
and also serve as a good hydrogen donor.
yield.§
The reduction potential of anthracene was reported to
be −2.50 V.10 Treatment of anthracene with Ti(III)
under similar conditions afforded 9,10-dihydroan-
thracene in a 95% isolated yield. The Ti(III) system
may thus be a useful reducing reagent for Birch-type
reduction of selected aromatic compounds under mild
conditions, in particular the relatively more reactive
(less stable) non-benzenoid annulenes.
The reaction mechanism of a classical Birch reduction
involves a direct electron transfer from the metal.11 It is
known that aqueous titanium trichloride can also act as
an electron transfer reagent.12 In an aqueous solution,
the Ti(III) ion may exist as different complexed cations.
3+
The cation Ti(H2O)6 is believed to be responsible for
electron-transfer reactions.13 When fluoride ions were
added to an aqueous solution of titanium trichloride,
the initially observed purple color due to the presence
3+
The reaction was exothermic and proceeded rapidly at
room temperature but at 0°C product formation was,
however, very slow. This allowed a qualitative study of
the relative ease of reducing the dihyropyrene and other
organic functional groups. When 2-bromo- or 2,7-
dibromodihydropyrene (4a and 4b, respectively) was
reacted with Ti(III) under similar conditions, the bromo
function was removed8 cleanly in addition to the reduc-
tion of the dihydropyrene even at 0°C. The hexaene 2
was isolated in near quantitative yield in both cases. At
0°C selective reduction of the nitro group9 in com-
pound 5a could however be achieved to afford the
amine 5b in an 87% yield.‡ When an excess of titanium
trichloride was used and the reaction mixture was
allowed to warm up by the exothermic nature of the
of Ti(H2O)6 was replaced by a persistent green color
attributed to the formation of the TiF6 anion.14 In
3−
this work, there was no product formation when the
reaction was carried out in the absence of NaOH.
When an excess of NaOH was present, a colorless
solution was observed. It may be proposed that in a
3+
strongly basic medium the initial Ti(H2O)6 ions could
3-x
be converted to a series of Ti(H2O)6-x(OH)x complex
ions which are responsible for the reduction of 1 to 2
via electron transfer processes.
Treatment of 2 with an excess of LDA at −70°C
resulted in a dark blue solution which was quenched
instantly by addition of an excess of methyl iodide.
Chromatography on neutral alumina led to the isola-
tion of two major fractions. Eluted first was believed to
be a mixture of 1 and 2-methylhexane 7a which could
not be further separated by chromatography. When
treated with DDQ this mixture however, afforded 1 and
2-formyldihydropyrene 8a.5 The second major fraction
isolated from the reaction of 2 and excess LDA was a
mixture of the diastereoisomers of 2,7-dimethylhexane
9. Conversion of a mixture of 9 to 2,7-diformyldihy-
dropyrene 1015 could also be achieved by oxidation
with DDQ. The above observations clearly indicated
the initial formation of the mono anion and dianion
from 2. Similar results were obtained when 2 was
treated with n-BuLi followed by quenching with methyl
iodide. It was however impossible to drive the initial
reaction to form the dianion alone even when a large
excess of base was used. Decomposition reactions
seemed to compete and resulted in a complicated mix-
ture of products.
† A mixture of NaOH (2.2 g, 54 mmol) and 1 (300 mg, 1.3 mmol) in
THF/CH3OH (50 mL each) was stirred for 5 min and then aqueous
titanium trichloride (8.0 mL, 7.8 mmol; 15% in 10% hydrochloric
acid) was added. The mixture became warm and after 0.5 h the
solution was poured into ice/water (250 mL). The mixture was
extracted with CH2Cl2 (3×60 mL). The organic layer was washed
with water, dried, and evaporated. The residue was chro-
matographed on neutral alumina using hexane as eluant to afford
compound 2 (300 mg, 99%); mp 168–170°C (lit.5 168–170°C).
‡ A mixture of 5a (20 mg, 0.057 mmol) and NaOH (0.8 g, 20 mmol)
in THF/CH3OH (12 mL each) was stirred until all NaOH dissolved,
and then cooled in an ice bath. Aqueous titanium trichloride (0.6
mL, 0.58 mmol; 15% in 10% hydrochloric acid) was added dropwise
to the mixture. After 0.5 h the solution was poured into ice-water
and extracted with CH2Cl2 (3×30 mL). The organic layer was
washed with water, dried, and evaporated. The residue was chro-
matographed on a neutral alumina column using hexane/CH2Cl2
§ The procedure was similar to that used for reduction of 1. Com-
pound 6 was isolated as a mixture of two isomers: mp 190–194°C;
1H NMR l 7.05–7.09 (m, 2H), 6.72–6.75 (m, 1H), 6.65 (d, 1H,
J=7.3 Hz), 5.98, 6.05 (AB, 4H, J=9.6 Hz), 5.75 (t, 2H, J=3.8 Hz),
5.72 (d, 2H, J=2.0 Hz), 4.27 (br s, 1H), 3.84 (br s, 2H), 2.98–3.00
(m, 2H), 1.056, 1.054 (s, 6H); MS m/z: 325 (M+, 30), 308 (100), 293
(64); Mr calcd for C24H23N: 325.1830. Found: 325.1817. Anal. calcd
for C24H23N: C, 88.57; H, 7.12. Found: C, 88.35; H, 7.06.
1
(1:1 v/v) as eluant to give 5b as a yellow oil (25 mg, 87%): H NMR
l 8.86 (s, 2H), 8.62, 8.58 (AB, 4H, J=7.7 Hz), 8.52 (d, 2H, J=7.7
Hz), 8.01 (t, 1H, J=7.6 Hz), 7.78–7.98 (m, 2H), 6.99–7.13 (m, 2H),
3.86 (br s, 2H), −3.99, −4.02 (s, 6H); umax (hexane) 496 (m 8500), 388
(33800), 346 (66600) nm; MS m/z: 323 (M+, 30), 308 (100), 293 (90);
Mr calcd for C24H21N: 323.1674. Found: 323.1681. Anal. calcd for
C24H21N: C, 89.12; H, 6.54. Found: C, 88.76; H, 6.59.

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
J. Jiang, Y.-H. Lai / Tetrahedron Letters 44 (2003) 1271–1274
1273
The dihydropyrene 1 is known to react readily with
electrophiles to yield a series of 2-substituted deriva-
tives.5 There are however some functional groups, for
example, the trimethylsilyl group¶ that cannot be intro-
duced directly via electrophilic substitutions. Knowing
that oxidation of the hexane 2 to 1 could be achieved
cleanly,5,15 the controlled formation of the mono anion
from 2 could be a useful and complementary synthetic
route to mono-substituted derivatives of 1. When the
hexaene 2 was treated with 1 equiv. of n-BuLi at −78°C
followed by quenching the very intensely blue solution
with an excess of (CH3)3SiCl, the 2-trimethylsilylhexane
7b was isolated as light yellow crystals in an 85% yield
after chromatography.ꢀꢀ Oxidation of 7b by DDQ in dry
benzene at 0°C afforded the desired 2-trimethylsilyldi-
hydropyrene 8b in a 97% yield. Compound 8b was
isolated as dark green crystals and its electronic spec-
trum is very similar to that of the parent 1.18 This
suggests that the trimethylsilyl group does not induce
any appreciable conjugation effect on the macroring.
The chemical shifts of the internal methyl protons in
8b, observed at l −4.18 and −4.21, respectively, are also
similar to that of the parent 1 (l −4.25).6 Thus the
bulky trimethylsilyl group does not seem to impose any
significant steric or electronic effect on the diatropicity
(ring current) of the dihydropyrene.
2 with NaOH in thoroughly degassed THF/CH3OH for
24 h at room temperature merely returned the starting
material. Repeating the reaction with exposure to air
for 10 h, however, resulted in the isolation of dihy-
dropyrene 1 (34%), bis(dihydropyrenyl) 11 (24%) and
bis(dihydropyrenyl)dione 12 (10%).** When a similar
reaction was carried out by bubbling oxygen into the
solution, a similar mixture of 1 (40%), 11 (8%) and 12
(48%) was obtained within 1 h. The above results
clearly indicate an initial oxidative coupling of the
anion of 2 to yield the intermediate 13 although all our
attempts in the isolation of 13 failed. Autoxidation
(dimerization) of organic anions is well documented.20
Two subsequent, consecutive oxidation reactions would
have led to the formation of 11 followed by dione 12.
This was supported by the fact that doing the reaction
under oxygen decreased the yield of 11 with a corre-
sponding increase in that of 12. Due to restricted
rotation at the central bridging carbonꢀcarbon double
bond in 12, the molecule exists as two rigid geometric
isomers 12a and 12b. This was clearly supported by its
1H NMR spectrum H1,3,10,12 were observed as two
singlets at l 7.06 and 7.07 while the internal methyl
protons appeared as two sets of singlets at l 1.26/1.27
and 1.09/1.10, respectively. Chemically, compound 12
was reduced by a mixture of lithium aluminum hydride
and aluminum chloride to afford 11 in a 76% yield.††
An attempt was made to study the stability of the anion
of 2 under prolonged reaction times. Thus treatment of
¶ The trimethylsilyl group could undergo iododesilylation16 and the
aryl iodide could be used in aryl–aryl coupling reactions in the
synthesis of novel aromatic compounds.17
** A mixture of 2 (120 mg, 0.43 mmol) and NaOH (3 g, 75 mmol) in
THF/CH3OH (40 mL each) was stirred for 24 h with exposure to
air. The solution was then poured into ice-water and extracted with
CH2Cl2 (3×50 mL). The organic layer was washed with water,
dried, and evaporated. The residue was chromatographed on silica
gel using hexane as eluant. Eluted first was 1 (40 mg, 34%). Eluted
next was 11 (28 mg, 24%): mp 196-198°C (lit.19 195–199°C). Chang-
ing the eluant to CH2Cl2/CH3COCH3 (10:1 v/v) eluted 12 (12 mg,
ꢀꢀ A solution of 2 (50 mg, 0.214 mmol) in THF (25 mL) was cooled to
−78°C. n-BuLi (5 mL, 1.5 M solution in hexane) was added slowly
under nitrogen. The mixture was stirred for 45 min and (CH3)3SiCl
(1 mL) was added dropwise. The solution was brought to 0°C over
2 h and ice-water (20 mL) was added. The organic layer was
separated, washed with water, dried, and concentrated. The residue
was chromatographed on silica gel using hexane as eluant to give 7b
(55 mg, 85%): mp 172–176°C; 1H NMR l 5.93, 5.87 (AB, 4H,
J=9.6 Hz), 5.6 (m, 2H), 5.58 (d, 2H, J=2.8 Hz), 3.0 (m, 2H), 2.6
(m, 1H), 1.04, 1.01 (s, 6H), 0.06 (s, 9H); MS m/z 306 (M+, 42), 259
(45), 217 (78), 202 (90), 73 (100). A solution of 7b (31 mg, 0.1
mmol) in benzene (40 mL) was cooled to 0°C and DDQ (65 mg)
was added. The mixture was stirred for 20 min and ice-water (40
mL) was added. The organic layer was washed, dried, and evapo-
rated. The residue was chromatographed on silica gel using hexane
as eluant to give 8b (29 mg, 97%): mp 187–189°C; 1H NMR l 8.73
(s, 2H), 8.62 (s, 4H), 8.57 (d, 2H, J=7.7 Hz), 8.10 (t, 1H, J=7.7
Hz), 0.57 (s, 9H), −4.18, −4.21 (s, 6H); umax (hexane) 478 (m 15900),
380 (57300), 338 (94800) nm; MS m/z 304 (M+, 35), 259 (40), 217
(55), 202 (66), 73 (100). Anal. calcd for C21H24Si: C, 82.83; H, 7.94.
Found: C, 82.53; H, 7.81.
1
10%) as a mixture of two isomers: mp 196–204°C; H NMR l 7.06,
7.07 (s, 4 H), 6.39, 6.59 (AB, 8H, J=9.4 Hz), 6.27 (br s, 4H), 1.26,
1.27 (s, 6H), 1.09, 1.10 (s, 6H); umax (hexane) 560 (m 85400), 526
(59300), 304 (40900), 268 (46700); MS m/z: 492 (M+, 56), 477 (78),
462 (100), 376 (40), 187 (65); Anal. calcd for C36H28O2: C, 87.78; H,
5.73. Found: C, 87.33; H, 5.50.
†† LiAlH4 (10 mg, 0.26 mmol) was added to a solution of 12 (20 mg,
0.04 mmol) in THF (40 mL) followed by AlCl3 (ca. 6 mg). The
mixture was heated at reflux for 1 h, and then 10% HCl (10 mL)
was added to quench the reaction and the solvent was evaporated.
The solid residue was extracted with CH2Cl2 (3×30 mL). The
organic layer was washed with water, 10% aqueous NaHCO3 and
water, dried, and evaporated. The residue was chromatographed on
silica gel using hexane as eluant to give 11 (14 mg, 76%).

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
1274
J. Jiang, Y.-H. Lai / Tetrahedron Letters 44 (2003) 1271–1274
In conclusion, a very efficient Birch-like reduction for
the conversion of dihydropyrene 1 to bis-triene 2 was
achieved with Ti(III). This could be applicable to other
aromatic systems within the appropriate range of
reduction potentials. The mono-anion and dianion
derived from bis-triene 2 could serve as synthetic pre-
cursors to substituted derivatives of 1 inaccessible
through normal electrophilic substitutions.
R. H.; Dingle, T. W.; Williams, R. V.; Mahedevan, R.
Acta Chem. Scand. B. 1983, 37, 445–450.
8. Ho, T. L.; Wong, C. M. Synth. Commun. 1973, 3, 237–
239.
9. (a) Somei, M.; Kato, K.; Inoue, S. Chem. Pharm. Bull.
1980, 28, 2515–2518; (b) McMurry, J. E.; Silvestri, M. J.
Org. Chem. 1975, 40, 1502–1504; (c) Ho, T. L.; Wong, C.
M. Synthesis 1974, 45.
10. Cheng, E.; Sun, T. C.; Su, Y. O. J. Chin. Chem. Soc.
1993, 40, 551–555.
11. Birch, A. J.; Nasipuri, D. Tetrahedron 1959, 6, 148–153.
12. (a) Clerici, A.; Clerici, L.; Porta, O. Tetrahedron Lett.
1996, 37, 3035–3038; (b) Araneo, S.; Clerici, A.; Porta, O.
Tetrahedron Lett. 1994, 35, 2213–2216; (c) Clerici, A.;
Porta, O.; Arnone, A. J. Org. Chem. 1990, 55, 1240–1248;
(d) Clerici, A.; Porta, O. J. Org. Chem. 1987, 52, 5099–
5104; (e) Clerici, A.; Porta, O. Tetrahedron 1982, 38,
1293–1297; (f) Clerici, A.; Porta, O.; Riva, M. Tetra-
hedron Lett. 1981, 22, 1043–1046.
Acknowledgements
This work was supported by the National University of
Singapore (NUS). The authors thank the staff in the
Chemical and Molecular Analysis Centre, Department
of Chemistry, NUS, for their technical assistance.
13. (a) Taube, H. Chem. Rev. 1952, 50, 69–126; (b) Haim, A.
Acc. Chem. Rev. 1975, 8, 264–272.
References
14. Fenske, R. F.; Caulton, K. G.; Radtke, D. D.; Sweeney,
1. (a) McMurry, J. E. Chem. Rev. 1989, 89, 1513–1524; (b)
Lenoir, D. Synthesis 1989, 883–897; (c) Betschart, C.;
Seebach, D. Chimia 1989, 43, 39–49; (d) McMurry, J. E.
Acc. Chem. Res. 1983, 16, 405–411; (e) Welzel, P. Nachr.
Chem. Tech. Lab. 1983, 31, 814–816; (f) Lai, Y.-H. Org.
Prep. Proced. Int. 1980, 12, 361–391.
2. Clerici, A.; Porta, O. Tetrahedron Lett. 1982, 23, 3517–
3520.
3. Pourbaix, M. Atlas of Electrochemical Equilibria in
Aqueous Solution, translated from Atlas d’Equilibres Elec-
trochimiques by J. A. Franklin; Pergamon: New York,
1966.
4. Mitchell, R. H. Chem. Rev. 2001, 101, 1301–1315 and
references cited therein.
5. Phillips, J. B.; Molyneux, R. J.; Sturm, E.; Boekelheide,
V. J. Am. Chem. Soc. 1967, 89, 1704–1709.
6. Boekelheide, V.; Phillips, J. B. J. Am. Chem. Soc. 1967,
89, 1695–1704.
C. C. Inorg. Chem. 1966, 5, 951–960.
15. Mitchell, R. H.; Calder, I.; Huisman, H.; Boekelheide, V.
Tetrahedron 1975, 31, 1109–1111.
16. (a) Colvin, E. Silicon in Organic Synthesis; Butterworths:
London, 1981; (b) Vollhardt, K. P. C. Angew. Chem., Int.
Ed. Engl. 1984, 23, 539–556.
17. Chaumeil, H.; Le Drian, C.; Defoin, A. Synthesis 2002,
757–760.
18. Mitchell, R. H.; Boekelheide, V. J. Chem. Soc., Chem.
Commun. 1970, 1555–1556.
19. Mitchell, R. H.; Chaudhary, M.; Dingle, T. W.; Williams,
R. V. J. Am. Chem. Soc. 1984, 106, 7776–7779.
20. (a) Organic Peroxides; Swern, D. Ed.; Interscience: New
York, 1970; Vol. 2; (b) Barton, D. H. R.; Jones, D. W. J.
Chem. Soc. 1965, 3563–3570; (c) Russell, G. A.; Bemis,
A. G. J. Am. Chem. Soc. 1966, 88, 5491–5497; (d) Gers-
mann, H. R.; Bickel, A. F. J. Chem. Soc. B 1971,
2231–2237.
7. Fry, A. J.; Simon, J.; Tashiro, M.; Yamato, T.; Mitchell,