Dimetalation: The acidity of monometalated arenes towards superbasic reagents

Article abstract of DOI:10.1002/1099-0690(200111)2001:21<3985::AID-EJOC3985>3.0.CO;2-Y

Twofold hydrogen/metal interconversions ("dimetalations") can be accomplished with "spiny" arenes (tert-butylbenzene, 1,4-di-tert-butylbenzene, 1,1,3,3-tetramethylindane and congeners) and N,N-crowded anilines (2,2,6,6-tetramethyl-1-phenylpiperidine and N

Full text of DOI:10.1002/1099-0690(200111)2001:21<3985::AID-EJOC3985>3.0.CO;2-Y

FULL PAPER
Dimetalation: The Acidity of Monometalated Arenes Towards Superbasic
Reagents
Eckhard Baston,[a] Raimondo Maggi,[a] Kirstin Friedrich,[a] and Manfred Schlosser*[a]
Keywords: Dimetalated arenes / Aggregation / Solvent effects / Steric hindrance / π-Cloud polarization
Twofold hydrogen/metal interconversions ("dimetalations")
can be accomplished with ‘‘spiny’’ arenes (tert-butylbenzene,
1,4-di-tert-butylbenzene, 1,1,3,3-tetramethylindane and con-
butoxide mixture in hydrocarbon media. In tetrahydrofuran,
no trace of a dimetalated species is found, although such spe-
cies, once generated in a different manner, are moderately
geners) and N,N-crowded anilines (2,2,6,6-tetramethyl-1- stable in this solvent at −75 °C. 1,4-Dimetalloarenes are fav-
phenylpiperidine and N,N-bis[triethylsilyl]aniline), albeit in-
ored over the 1,3-isomers; 1,2-isomers do not form at all. As
completely (average and optimum yields of trapping prod- found by competition kinetics, 2,2,6,6-tetramethyl-1-phenyl-
ucts 25% and 41%, respectively). All the evidence points at
an aggregate phenomenon. Dimetalation occurs only with an
excess of a highly concentrated butyllithium/potassium tert-
piperidine and N,N-bis(triethylsilyl)aniline undergo mono-
metalation at roughly the same rate as tert-butylbenzene and
much more slowly than benzene does.
Introduction
Results and Discussions
A principal objection had to be addressed first. Even if
dimetal species should indeed be less basic than monometal
analogs, their generation may be impeded by the notori-
ously poor solubility of the monometalated species in par-
affin-type hydrocarbons, the standard reaction media. This
argument no longer holds if one works in homogeneous
solution. To date, no trace of a dimetalated benzene deriva-
tive has ever been detected in tetrahydrofuran-rich mixtures,
which at a quick glance would appear to be irrefutable evid-
ence against dimetalation in ethereal solvents. But can one
really rule out the rapid generation of dimetalated species
and their even more rapid destruction by solvent attack un-
der such conditions?
To examine this possibility, we treated benzene with equi-
molar amounts of butyllithium and potassium tert-butox-
ide, dissolved or suspended in hexanes, as described previ-
ously.[10] The solvent was then removed and the residue was
dissolved in precooled tetrahydrofuran. After 2 h of storage
at Ϫ75 °C, the composition of the organometallic interme-
diates (as monitored by interception with chlorotrimethylsi-
lane) had not changed much. The quantity of monometal-
ated species had increased slightly (51 Ǟ 53%) at the ex-
pense of the 1,3- and 1,4-phenylenedimetals (12 Ǟ 6% and
16 Ǟ 9%, respectively). A major proportion of dimetalated
species, had such species ever been produced in tetrahy-
drofuran, would obviously have survived.
Organic dimetal and polymetal compounds[1] continue to
attract attention because of their intriguing structures and
special modes of generation. They had an additional spot-
light shone on them when recent computational work re-
vealed phenyllithium to be more acidic than benzene.[2] This
opened not just the frivolous perspective that aryl anion-
arylmetal adducts might be long-lived in the gas phase, but
also that dimetalated arenes might be thermodynamically
favored over monometalated ones even in condensed
phases. In continuation of previous investigations,[3] we
have studied this issue experimentally. As detailed below,
our findings clarify the problem in all major aspects:
Dimetalation can be achieved with a variety of aromatic
substrates bearing one or two tert-alkyl groups or hetero-
substituents. So far, dimetalation at the same aromatic ring
has been reported only with benzene itself, with variable
amounts (generally Ͻ 10%) of isophthalic and terephthalic
acid being identified after carboxylation.[4Ϫ9]
Dimetalation of monosubstituted benzenes occurs ex-
clusively at the meta positions. In this way, adverse steric
and electrostatic effects are obviously minimized.
Dimetalation is only kinetically, not thermodynamically
favored over monometalation. It is mediated by high-
olecular aggregates, the formation of which requires high
reagent concentrations and noncoordinating solvents.
To gain further insights, competition experiments were
carried out. In this way, substituent effects on metalation
rates were quantified.
Solubility in hydrocarbon media does not appear to be a
crucial factor. The amounts of monometalated, para-di-
metalated, and meta-dimetalated species (M ϭ K ϩ Li), as
reflected by their carboxylation products (M ϭ COOH),
remain constant (approximately 60%, 15%, and 10%, re-
spectively) regardless of whether a 0.5 or 1.5 [3] solution
of benzene is treated with equimolar quantities of butylli-
thium and potassium tert-butoxide. In contrast, the propor-
[a]
´
Section de Chimie (BCh), Universite,
1015 Lausanne, Switzerland
Fax (internat.) ϩ 41-21/692-3965
Eur. J. Org. Chem. 2001, 3985Ϫ3989
WILEY-VCH Verlag GmbH, 69451 Weinheim, 2001
1434Ϫ193X/01/1121Ϫ3985 $ 17.50ϩ.50/0
3985

E. Baston, R. Maggi, K. Friedrich, M. Schlosser
FULL PAPER
tions shift considerably when the superbasic reagent mix-
ture is employed in threefold excess, which promotes the
formation of para-dimetalated and meta-dimetalated inter-
mediates (26 and 24%) at the expense of the monometalated
one (42%).
In order to examine heteroelement effects further, the in-
vestigation was extended to two aniline derivatives. Both
2,2,6,6-tetramethyl-1-phenylpiperidine and N,N-bis(tri-
methylsilyl)aniline showed high levels of dimetalation when
treated with an excess of the superbase. The former sub-
strate afforded doubly and singly substituted products in a
28:45 ratio (14 vs. 13a ϩ 13b, at a 1:5:5 stoichiometry), the
latter in a 41:40 ratio (16 vs. 15a ϩ 15b, at a 1:3:3 stoichi-
ometry).
Bulky substiuents should increase the solubility of or-
ganometallic intermediates in petroleum ether. If tert-butyl-
benzene is compared with benzene, however, the ratio of
dimetalated species to monometalated species (4 vs. 3a ϩ
3b) diminishes from 25:60 to 9:71 and from 50:42 to 19:69
at the 1:1:1 and 1:3:3 stoichiometries, respectively.
1,4-Di-tert-butylbenzene is particularly tough to crack,
all aromatic positions suffering from massive steric hind-
rance. Dimetalation generating the para- and meta-isomers
6a and 6b in significant yields occurs only if a fivefold ex-
cess of butyllithium and potassium tert-butoxide is em-
ployed at elevated temperatures (e.g., ϩ75 °C) and pro-
longed reaction times are used.
Despite their structural diversity, the substrates examined
exhibit common reactivity features. To push dimetalation
to synthetically useful levels, a threefold or fivefold excess of
the metalating reagent is required in all cases. If equimolar
amounts of the substrate and the superbase are allowed to
react with each other, only benzene affords substantial pro-
portions of dimetalated species (up to 15% of the para-iso-
mer and 10% of the meta-isomer). Under the same condi-
tions, tert-butylbenzene, 2,2,6,6-tetramethyl-1-phenylpiper-
idine, and N,N-bis(trimethylsilyl)aniline give only 9%, 3%,
and 2%, respectively, of the sterically uncongested meta-di-
metalated species. All this evidence suggests that the second
deprotonation is accomplished within a tightly packed
mixed aggregate, the depicted 1:2:2 and 1:4:4 adducts (17
and 18) being intended only as a nonexhaustive illustration
of a few possibilities.
Recently the superbase-promoted monometalation of
1,1,3,3-tetramethylindane,
1,1,2,2,3,3-hexamethylindane,
and 1,1,3,3-tetramethyl-1,3-dihydroisobenzofuran has been
reported.[10] These substrates can be considered as substi-
tutes for the less readily available 1,2-(di-tert-butyl)benzene.
Monosubstitution of an ortho-position does not occur with
the hydrocarbons (neither 7a nor 9a), but does, at least to
some extent, with the cyclic ether (5% of 11a), thus sug-
gesting the operation of a long-range inductive heteroele-
ment effect. Dimetalation implies proton abstraction from
one crowded position. This may explain why only moderate
yields of derivatives 8 and 10 can be achieved even when a
fivefold excess of the superbasic reagent is used. No more
than traces of the species 12 have been identified. This is
presumably a consequence of the fragility of the oxygen het-
erocycle, which tends to undergo ring scission by β-elimina-
tion.
3986
Eur. J. Org. Chem. 2001, 3985Ϫ3989

The Acidity of Monometalated Arenes Towards Superbasic Reagents
FULL PAPER
Electrostatics obviously dictate the regiochemical out- monometalated intermediate, 3-[N,N-bis(triethyl)silyl]amin-
come of dimetalations. Whereas ortho-dimetalations are rig- ophenyllithium, for example. The second metal may then
orously precluded, para-dimetalation dominates over meta- be introduced separately and regioselectively by treatment
dimetalation whenever this is possible, as with benzene with the superbase.
(p/m ratio of approximately 2.5 after statistical correction)
and 1,4-di-tert-butylbenzene (p/m ratio 4.0) as the sub-
strates.
Benzenes carrying a single bulky substituent undergo
Experimental Section
For standard laboratory practice and abbreviations, see recent pub-
lications from this laboratory.[11Ϫ12] 1H NMR spectra were re-
corded using solutions in deuteriochloroform or, if marked with an
asterisk, in perdeuterated dimethyl sulfoxide at 400 MHz.
monometalation at both the m- and the p-positions, but di-
metalation only at m,mЈ-positions. In other words, the di-
metalated intermediate is generated exclusively at the ex-
pense of the m-monometalated one. As a consequence, the
p/[mϩm,mЈ] ratios (all statistically corrected) do not vary
excessively as a function of the substrate/reagent stoichi-
ometry: between 1.3 and 3.2 with tert-butylbenzene, be-
tween 1.1 and 1.3 with 2,2,6,6-tetramethyl-1-phenylpiperid-
ine, and between 1.3 and 1.6 with N,N-bis(triethylsilyl)ani-
line.
tert-Butylbenzene as the Substrate: Potassium tert-butoxide (1.1 g,
10 mmol) was added to a solution of tert-butylbenzene (1.6 mL,
1.3 g, 10 mmol) and butyllithium (10 mmol) in hexanes (35 mL).
The suspension was vigorously stirred for 2 h at ϩ25 °C before
being poured on an excess of freshly crushed dry ice covered with
a layer of tetrahydrofuran. The mixture was acidified with 2  hy-
drochloric acid (50 mL) and extracted with diethyl ether (2 ϫ
25 mL). One twentieth of the solution was withdrawn and treated
with ethereal diazomethane until the yellow color persisted. A
known amount (approx. 10 mg) of tridecane was added as an ‘‘in-
Regioisomeric ratios reflect the reactivity at various sites
that compete with each other intramolecularly. They do not
tell us anything about how rapidly or slowly a given sub-
strate becomes metalated in comparison with a reference ternal standard’’ for the gas chromatographic analysis (2 m, 5% C-
20M, 50 °C Ǟ 200 °C; 2 m, 5% Ap-L, 50 °C Ǟ 220 °C). The peaks
of methyl 4-tert-butyl-benzoate,[13] methyl 3-tert-butylbenzoate,[14]
and dimethyl 5-tert-butylisophthalate[15] (3a, 3b, and 4, respect-
ively; M ϭ COOCH3) were identified by comparison of their reten-
tion times with those of authentic samples and their areas relative
to the reference compound were corrected using separately deter-
mined calibration factors. The products 3a, 3b, and 4 had formed
in 43%, 28%, and 9% yields. The remainder of the combined or-
ganic layers was washed with brine (2 ϫ 25 mL), dried and evapor-
ated to dryness. Recrystallization of the residue from acetic acid
gave a small amount (3%) of 5-tert-butylisophthalic acid; m.p. (de-
comp.) 340Ϫ342 °C. Ϫ When three equivalents (30 mmol) of butyl-
lithium and potassium tert-butoxide were employed under other-
wise identical conditions, the methyl esters 3a, 3b, and 4 were found
to be present in 54%, 15% and 19% yields.
compound. The easiest way to collect information on rela-
tive intermolecular reactivities is to carry out competition
kinetic experiments. This has already been accomplished for
a series of ‘‘spiny’’ arenes such as tert-butylbenzene and di-
tert-butylbenzenes, relative to benzene.[3,10] This study has
now been extended to the bulky aniline derivatives 2,2,6,6-
tetramethyl-1-phenylpiperidine and N,N-bis(triethylsilyl)-
aniline (see Table 1).
Table 1. Metalation of tert-butylbenzene, 2,2,6,6-tetramethyl-1-
phenylpiperidine, and N,N-bis(triethylsilyl)aniline with butylli-
thium in the presence of potassium tert-butoxide in hexanes at
25 °C: relative rates krel [a] and partial rate factors kf
[b]
1,4-Di-tert-butylbenzene as the Substrate: A vigorously stirred mix-
ture containing 1,4-di-tert-butylbenzene (19 g, 0.10 mol), butylli-
thium (0.50 mol), and potassium tert-butoxide (0.50 mol) in hex-
anes (0.35 L) was heated under reflux (approx. 75 °C) for 25 h. It
was cooled to Ϫ75 °C and diluted with precooled tetrahydrofuran
(0.10 L) before being poured on an excess of freshly crushed dry
ice covered with a layer of tetrahydrofuran (50 mL). Diethyl ether
(0.25 L) was added and the products were extracted with a 10%
aqueous solution (5 ϫ 50 mL) of sodium hydroxide and, after acid-
ification of this alkaline phase to pH 2, with dichloromethane (3 ϫ
50 mL). After drying and evaporation of the solvent, methyl iodide
(12 mL, 28 g, 0.20 mol), potassium carbonate (28 g, 0.20 mol), and
acetone (0.25 L) were added and the mixture was stirred for 25 h
at 25 °C.[16] The methyl 2,5-di-tert-butylbenzoate[17] (5: M ϭ
COOCH3) was isolated by distillation (bp 148Ϫ150 °C/7 Torr). The
residue was absorbed on silica gel (0.25 L). Elution with a 1:19
(v/v) mixture of diethyl ether and hexanes resulted in the separation
of dimethyl 2,5-di-tert-butylterephthalate [6a: M ϭ COOCH3; m.p.
[a]
[b]
Reaction rates relative to that of benzene ϵ 6.0. Ϫ
Relative
rates multiplied by the fraction of reaction occurring at the given
site (ortho, meta, or para) and divided by the number of equivalent
exchange-active positions (2 or 1).
Are dimetalations useful practically? Without doubt, if
one is aiming at the preparation of a dicarboxylic acid, a
dialdehyde, or a dialcohol by reaction with carbon dioxide,
a formamide, or formaldehyde, respectively. Such highly po-
lar derivatives can easily be separated and purified even in
cases of moderate yields. Moreover, improvements are pos-
sible. To boost yields and to avoid the formation of re-
gioisomers, one may first perform a permutational halogen/
1
141Ϫ142 °C. Ϫ H NMR: δ ϭ 7.34 (s, 2 H), 3.91 (s, 6 H), 1.37 (s,
18 H)] and dimethyl 2,5-di-tert-butylisophthalate [6b:
M ϭ
COOCH3; m.p. 99Ϫ100 °C. Ϫ 1H NMR: δ ϭ 7.33 (s, 2 H),
3.90 (s, 6 H), 1.43 (s, 9 H), 1.29 (s, 9 H)]. Ϫ Hydrolysis of each
component under standard conditions (acetic acid, boron
metal or metalloid/metal exchange to generate a specific trifluorideϪdiethyl ether, 15 h 25 °C) afforded 2,5-di-tert-butyl-
Eur. J. Org. Chem. 2001, 3985Ϫ3989 3987

E. Baston, R. Maggi, K. Friedrich, M. Schlosser
FULL PAPER
benzoic acid[18,19] (5: R ϭ COOH; m.p. 162Ϫ163 °C), 2,5-di-tert-
were again characterized by their spectra and combustion analysis
butylterephthalic acid [6a: R ϭ COOH; m.p. Ͼ 250 °C. Ϫ as 1.3-dihydro-1,1,3,3-tetramethyl-4-isobenzofurancarboxylic acid
C16H22O4 (278.35): calcd. C 69.04, H 7.97; found C 69.08, H 8.13] (11a) and 1,3-dihydro-1,1,3,3-tetramethyl-5-isobenzofurancarbox-
and 2,5-di-tert-butylisophthalic acid [6b: R ϭ COOH; m.p. Ͼ 250
°C. Ϫ C16H22O4 (278.35): calcd. C 69.04, H 7.97; found C 69.15,
H 7.92]. Ϫ To determine the exact yields of the esters 5, 6a, and
6b (M ϭ COOCH3) obtained, a suspension containing potassium
ylic acid[23] (11b) and 1,3-dihydro-1,1,3,3-tetramethyl-4,6-isobenzo-
furandicarboxylic acid (12). Ϫ 1.3-Dihydro-1,1,3,3-tetramethyl-4-
isobenzofurancarboxylic Acid (11a): M ϭ COOH; m.p. 185Ϫ187
1
°C. Ϫ H NMR: δ ϭ 8.07 (dd, J ϭ 7.6, 1.3 Hz, 1 H), 7.41 (t, J ϭ
tert-butoxide (2.8 g, 25 mmol), butyllithium (25 mmol), and 1,4-di- 7.5 Hz, 1 H), 7.34 (dd, J ϭ 7.5, 1.3 Hz, 1 H), 1.76 (s, 6 H), 1.54 (s,
tert-butylbenzene (5.0 or 25 mmol) was vigorously stirred for 25 h 6 H). Ϫ C13H16O3 (220.67): calcd. C 70.87, H 7.33; found C 70.68,
at 75 °C. Treatment with carbon dioxide and diazomethane fol-
lowed as described above. The product concentrations were deter-
H 7.12]. Ϫ 1,3-Dihydro-1,1,3,3-tetramethyl-5-isobenzofurancarbox-
ylic Acid (11b):[23] M ϭ COOH; m.p. 202Ϫ204 °C. Ϫ 1H NMR:
mined by gas chromatographic analysis (30 m, DB-210, 100 Ǟ 200 δ ϭ 8.08 (dd, J ϭ 7.8, 1.2 Hz, 1 H), 7.65 (d, J ϭ 1.2 Hz, 1 H), 7.26
°C; 30 m DB-1701, 110 Ǟ 200 °C), using 4,4Ј-di-tert-butylbiphenyl (d, J ϭ 7.8 Hz, 1 H), 1.58 (s, 6 H), 1.57 (s, 6 H)]. Ϫ 1,3-Dihydro-
as an internal calibrated standard.
1,1,3,3-tetramethyl-4,6-isobenzofurandicarboxylic Acid (12): M ϭ
COOH; m.p. Ͼ 250 °C. Ϫ H NMR*: δ ϭ 8.34 (d, J ϭ 1.5 Hz, 1
1
1,1,3,3-Tetramethylindane as the Substrate: A suspension con-
taining potassium tert-butoxide (5.6 g, 50 mmol), butyllithium
(50 mmol), and 1,1,3,3-tetramethylindane[20Ϫ21] (1.7 g, 10 mmol) in
hexanes (50 mL) was vigorously stirred for 25 h at ϩ25 °C. The
mixture was worked up as described in the preceding section to
afford the esters 7b and 8 (M ϭ COOCH3), which could readily be
separated by elution from silica gel, using a 1:4 (v/v) mixture of
ethyl acetate and hexanes. Hydrolysis gave 7b and 8. Ϫ 1,1,3,3-
Tetramethylindane-5-carboxylic Acid (7b): M ϭ COOH; m.p.
H), 7.97 (d, J ϭ 1.5 Hz, 1 H), 1.62 (s, 6 H), 1.45 (s, 6 H). Ϫ
C15H18O4 (262.30): calcd. C 68.69, H 6.92; found C 68.53, H 7.22].
ϪGas chromatography (1 m, 5% SE-30, 90 °C; 30 m, DB-1701, 120
°C; methyl tert-butylbenzoate as the internal standard), showed
that the esters (M ϭ COOCH3) 11a, 11b, and 12 were formed in
5%, 10%, and 4% yields or in 12%, 23%, and 0% yields, respectively,
depending on whether a 1:5:5 or a 1:1:1 substrate/bases stoichi-
ometry was chosen. Ϫ When tert-butyllithium (1 equiv.) was used
to replace butyllithium as a component of the superbasic reagent
mixture, isomer 11b was formed exclusively (52%).
1
174Ϫ175 °C. Ϫ H NMR*: δ ϭ 7.79 (dd, J ϭ 7.8, 1.6 Hz, 1 H),
7.72 (d, J ϭ 1.6 Hz, 1 H), 7.27 (d, J ϭ 7.8 Hz, 1 H), 1.91 (s, 2 H),
1.28 (s, 12 H). Ϫ C14H18O2 (218.29): calcd. C 77.03, H 8.31; found
C 76.92, H 8.25. Ϫ 1,1,3,3-Tetramethylindane-4,6-dicarboxylic Acid
2,2,6,6-Tetramethyl-1-phenylpiperidine as the Substrate: A suspen-
sion containing potassium tert-butoxide (10 mmol), butyllithium
(10 mmol), and 2,2,6,6-tetramethyl-1-phenylpiperidine (0.43 g,
2.0 mmol; see below) in hexanes (10 mL) was vigorously stirred for
25 h at 25 °C. Methyl iodide (0.62 mL, 1.4 g, 10 mmol) in tetrahy-
drofuran (10 mL) was added at Ϫ75 °C. After having reached 25
°C, the mixture was filtered and its composition was examined by
gas chromatography (1 m, 5% SE-30, 100 °C; 30 m, DB-1701, 120
°C; 1,4-di-tert-butylbenzene as an internal standard). The products
were identified by comparison of their retention times with those
of authentic samples (see below). 2,2,6,6-Tetramethyl-1-p-tolyl-
piperidine (13a: R ϭ CH3), 2,2,6,6-tetramethyl-1-m-tolylpiperidine
(13b: R ϭ CH3) and 1-(3,5-dimethylphenyl)-2,2,6,6-tetramethylpi-
peridine (14: R ϭ CH3) were found to be present in concentrations
corresponding to yields of 29%, 16%, and 28%. When a 1:1:1 sub-
strate (2.2 g, 10 mmol)/bases (2 ϫ 10 mmol) stoichiometry was ap-
plied and the metalation time was simultaneously shortened to 2 h
(at 25 °C), the products 13a, 13b, and 14 were formed in 10%,
16%, and 3% yields. Ϫ 2,2,6,6-Tetramethyl-1-phenylpiperidine was
prepared by sequentially adding 2,2,6,6-tetramethylpiperidine
(84 mL, 71 g, 0.50 mol) and bromobenzene (53 mL, 79 g, 0.50 mol)
to butyllithium (0.50 mol) in tetrahydrofuran (0.50 L) and by keep-
ing this solution for 25 h at 25 °C. The product was isolated by
distillation; bp 115Ϫ117 °C/12 Torr; n2D0 ϭ 1.5236; 46%. Ϫ 1H
NMR: δ ϭ 7.2 (m, 5 H), 1.7 (m, 2 H), 1.6 (m, 4 H), 1.01 (s, 12 H).
Ϫ C15H23N (217.35): calcd. C 82.89, H 10.66; found C 83.01, H
10.55. Ϫ A 55:45 mixture of 1-(4-methylphenyl)- and 1-(3-methyl-
phenyl)-2,2,6,6-tetramethylpiperidine (13a and 13b: M ϭ CH3) was
obtained analogously, using 4-bromotoluene (62 mL, 86 g, 0.50
mol) instead of bromobenzene; bp 123Ϫ125 °C/110 Torr; 47%. Ϫ
C16H25N (231.38): calcd. C 83.06, H 10.89; found C 83.42, H 10.61.
Ϫ 1-(3,5-Dimethylphenyl)-2,2,6,6-tetramethylpiperidine (14): (M ϭ
CH3) was made by heating a mixture of 1-fluoro-3,5-dimethylben-
zene (1.9 g, 15 mmol), 2,2,6,6-tetramethylpiperidine (5.0 mL, 4.2 g,
30 mmol), and sodium amide (2.3 g, 60 mmol) in hexamethylphos-
1
(8): M ϭ COOH; m.p. Ͼ 250 °C. Ϫ H NMR*: δ ϭ 8.07 (d, J ϭ
1.6 Hz, 1 H), 7.87 (d, J ϭ 1.6 Hz, 1 H), 1.94 (s, 2 H), 1.45 (s, 6 H),
1.30 (s, 6 H). Ϫ C15H18O4 (262.30): calcd. C 68.69, H 6.92; found
C 68.53, H 7.22. Ϫ
The exact yields of esters were determined by gas chromatographic
analysis (1 m, 5% SE-30, 90 °C; 30 m, DB-1701, 120 °C) using
methyl tert-butylbenzoate as an internal and calibrated standard.
When the hydrocarbon substrate and the basic reagents were em-
ployed in a stoichiometric 1:1:1 ratio, ester 7b (M ϭ COOCH3)
was obtained in 51% yield, but no diester 8 was formed at all.
1,1,2,2,3,3-Hexamethylindane as the Substrate: The reaction was
performed with 1,1,2,2,3,3-hexamethylindane[22] (2.0 g, 10 mL) ex-
actly as described above for 1,1,3,3-tetramethylindane. The stand-
ard workup procedure (see the 1,4-di-tert-butylbenzene section)
provided 1,1,2,2,3,3-hexamethylindane-5-carboxylic acid (9b) and
1,1,2,2,3,3-hexamethylindane-4,6-dicarboxylic acid (10).
؊
1,1,2,2,3,3-Hexamethylindane-5-carboxylic Acid (9b): M ϭ COOH;
m.p. 179Ϫ180 °C. Ϫ 1H NMR: δ ϭ 7.97 (d, J ϭ 8.0 Hz, 1 H), 7.87
(s, 1 H), 7.22 (d, J ϭ 8.0 Hz, 1 H), 1.25 (s, 6 H), 1.23 (s, 6 H), 0.89
(s, 6 H). Ϫ C16H22O2 (246.35): calcd. C 78.01, H 9.00; found C
78.10, H 9.10]. Ϫ 1,1,2,2,3,3-Hexamethylindane-4,6-dicarboxylic
1
Acid (10): M ϭ COOH; m.p. Ͼ 250 °C. Ϫ H NMR*: δ ϭ 13.14
(s, broad, 2 H), 7.79 (d, J ϭ 1.6 Hz, 1 H), 7.78 (d, J ϭ 1.6 Hz, 1
H), 1.32 (s, 6 H), 1.21 (s, 6 H), 0.81 (s, 6 H). Ϫ C17H22O4 (290.34):
calcd. C 70.32, H 7.64; found C 70.12, H 7.64]. Ϫ Gas chromato-
graphic analysis (1 m, 5% SE-30, 90 °C; 30 m, DB-1701, 120 °C,
methyl tert-butylbenzoate as the internal standard) of the methyl
esters (M ϭ COOCH3), identified 9b and 10 in 3% and 26% yields,
respectively, but 9b appeared as the sole product in 52% yield if a
1:1:1 substrate/superbase mixture was used in place of one of 1:5:5.
1,3-Dihydro-1,1,3,3-tetramethylisobenzofuran as the Substrate: 1,3-
Dihydro-1,1,3,3-tetramethylisobenzofuran[23] (1.8 g, 10 mmol) was phorus triamide (30 mL) and tetrahydrofuran (10 mL) at 50 °C for
submitted to the same reaction and workup conditions as described
for 1,4-di-tert-butylbenzene (see above). The three products found 2 Torr; 21%. Ϫ H NMR: δ ϭ 7.21 (s, 2 H), 6.93 (s, 1 H), 2.37 (s,
25 h. The product was collected by distillation; bp 93Ϫ94 °C/
1
3988
Eur. J. Org. Chem. 2001, 3985Ϫ3989

The Acidity of Monometalated Arenes Towards Superbasic Reagents
FULL PAPER
6 H), 1.7 (m, 2 H), 1.6 (m, 4 H), 1.04 (s, 12 H,). Ϫ C17H27N 12 H). Ϫ C20H39NSi2 (349.70): calcd. C 68.69, H 11.24; found C
(245.41): calcd. C 83.20, H 11.10, N 5.71; found C 83.21, H 10.79, 68.52, H 11.31].
N 6.05.
Acknowledgments
N,N-Bis(triethylsilyl)aniline as the Substrate: A suspension con-
taining potassium tert-butoxide (3.4 g, 30 mmol), butyllithium
(30 mmol), and N,N-bis(trimethylsilyl)aniline (3.2 g, 10 mmol; see
below) in hexanes (20 mL) was vigorously stirred for 2 h at 25 °C.
Methyl iodide (1.9 mL, 4.3 g, 30 mmol) in tetrahydrofuran (20 mL)
and a known amount (approx. 0.5 g) of octadecane (as an internal
reference compound) were added at Ϫ75 °C. When it had warmed
up to 25 °C, the mixture was washed with brine (5 ϫ 10 mL) and
analyzed by gas chromatography (30 m, DB-FFAP, 130 °C; 30 m,
DB-1701, 165 °C). The new products 15a (M ϭ CH3, 36%), 15b
(M ϭ CH3, 4%), and 16 (M ϭ CH3, 41%) were identified by com-
parison of their retention times with those of authentic samples
(see below). When a 1:1:1 substrate stoichiometry (9.6 g, 30 mmol)/
bases (2 ϫ 30 mmol) was observed, the yields of 15a, 15b, and 16
were 24%, 36%, and 2%, respectively. Ϫ N,N-Bis(triethylsilyl)ani-
line: The compound was prepared by sequential addition, at Ϫ75
°C, of aniline (9.1 mL, 9.8 g, 0.10 mol) and chlorotriethylsilane
(17 mL, 15 g, 0.10 mol) to butyllithium (0.10 mol) in tetrahydrofu-
ran (0.20 L) and hexanes (60 mL). After 1 h at 25 °C, the mixture
was cooled to Ϫ75 °C and again treated sequentially with butyl-
lithium (0.10 mol) in hexanes (60 mL) and chlorotriethylsilane
(17 mL, 15 g, 0.10 mol) before being allowed to stand overnight
(18 h) at 25 °C. The fine precipitate was removed by filtration with
suction through a pad of kieselguhr and the solvents were evapor-
ated. Distillation of the residue under reduced pressure afforded a
colorless liquid; bp 109Ϫ111 °C/0.5 Torr; n2D0 ϭ 1.5075; 18 g (57%).
This work was financially supported by the Schweizerische Na-
tionalfonds zur Förderung der wissenschaftlichen Forschung, Bern
(20Ϫ55Ј303Ϫ98). The authors are also indebted to Callary Chem.
Co. (Mr. S. McCalmont), Pittsburg, for a gift of chemicals.
[1]
A. Maercker in Houben-Weyl: Methoden der organischen
Chemie (Ed.: M. Hanack), Vol. E19d, pp, 448Ϫ566; Thieme,
Stuttgart 1993.
[2]
S. M. Bachrach, M. Hare, S. R. Kass, J. Am. Chem. Soc. 1998,
120, 12646Ϫ12649.
M. Schlosser, J. H. Choi, S. Takagishi, Tetrahedron 1990, 46,
5633Ϫ5648.
A. A. Morton, I. Hechenbleikner, J. Am. Chem. Soc. 1936,
58, 1024Ϫ1027.
A. A. Morton, I. Hechenbleikner, J. Am. Chem. Soc. 1936,
58, 2599Ϫ2605.
[3]
[4]
[5]
[6]
A. A. Morton, F. Fallwell, J. Am. Chem. Soc. 1938, 60,
1924Ϫ1927.
[7]
A. A. Morton, G. M. Richardson, A. T. Hallowell, J. Am.
Chem. Soc. 1941, 63, 327Ϫ330.
[8]
A. A. Morton, E. L. Little, W. O. Strong, J. Am. Chem. Soc.
1943, 65, 1339Ϫ1346.
[9]
H. Gilman, R. H. Kirby, J. Am. Chem. Soc. 1936, 58,
2074Ϫ2075.
[10]
E. Baston, Q. Wang, M. Schlosser, Tetrahedron Lett. 2000,
41, 667Ϫ670.
M. Schlosser, J. Porwisiak, F. Mongin, Tetrahedron 1998, 54,
895Ϫ900.
Q. Wang, H.-x. Wei, M. Schlosser, Eur. J. Org. Chem. 1999,
3263Ϫ3268.
[11]
[12]
1
[13]
Ϫ H NMR: δ ϭ 7.17 (t, J ϭ 7.9 Hz, 2 H), 7.03 (t, J ϭ 7.7 Hz, 1
I. Nishiguchi, T. Irashima, J. Org. Chem. 1985, 50, 539Ϫ541.
[14]
H), 6.9 (m, 2 H,), 0.89 (t, J ϭ 7.9 Hz, 18 H), 0.54 (q, J ϭ 7.9 Hz,
12 H). Ϫ C18H35NSi2 (321.66): calcd. C 67.21, H 10.97; found C
67.14, H 10.82. Ϫ The same procedure was used to obtain 15a, 15b
and 16. ؊ N,N-Bis(triethylsilyl)-p-toluidine (15a): (M ϭ CH3); bp
W. J. Houlihan, P. G. Munder, D. H. Handley, S. H. Cheon, V.
A. Parrino, J. Med. Chem. 1995, 234Ϫ240.
[15]
D. Nightingale, H. D. Radford, O. G. Shanholtzer, J. Am.
Chem. Soc. 1942, 64, 1662Ϫ1665.
[16]
S. C. Welch, C. Y. Chou, J. M. Gruber, J. M. Asserq, J. Org.
1
130Ϫ135 °C/0.3 Torr; n2D0 ϭ 1.5122; 28 g (84%). Ϫ H NMR: δ ϭ
Chem. 1985, 50, 2668Ϫ2676.
[17]
7.00 (d, J ϭ 8.0 Hz, 2 H), 6.86 (d, J ϭ 8.0 Hz, 2 H), 2.31 (s, 3 H),
0.92 (t, J ϭ 7.8 Hz, 12 H], 0.55 (q, J ϭ 7.8 Hz, 18 H). Ϫ
C19H37NSi2 (335.68): calcd. C 67.98, H 11.11; found C 67.59, H
11.14]. Ϫ N,N-Bis(triethylsilyl)-m-toluidine (15b): (M ϭ CH3); bp
L. R. C. Barclay, N. D. Hall, G. A. Cooke, Can. J. Chem. 1962,
40, 1981Ϫ1985.
[18]
M. G. J. Beets, W. Meersburg, H. Van Essen, Recl. Trav. Chim.
Pays-Bas 1959, 7, 570 Ϫ585.
[19]
J. M. A. Baas, B. M. Wepster, Recl. Trav. Chim. Pays-Bas 1967,
1
148Ϫ150 °C/0.3 Torr; n2D0 ϭ 1.5850; 27 g (84%). Ϫ H NMR: δ ϭ
86, 69Ϫ79.
[20]
7.05 (t, J ϭ 7.5 Hz, 1 H), 6.85 (d, J ϭ 7.5 Hz, 1 H), 6.76 (s, 1 H),
6.75 (d, J ϭ 7.2 Hz, 1 H), 2.27 (s, 3 H), 0.89, (t, J ϭ 7.9 Hz, 18
H), 0.52 (q, J ϭ 7.9 Hz, 12 H). Ϫ C19H37NSi2 (335.68): calcd. C
67.98, H 11.11; found C 68.03, H 11.25]. Ϫ N,N-Bis(triethylsilyl)-
3,5-dimethylaniline (16): (M ϭ CH3); bp 130Ϫ137 °C/0.5 Torr;
M. T. Bogert, D. Davidson, J. Am. Chem. Soc. 1934, 56,
185Ϫ190.
[21]
R. Sikkar, P. Martinson, B. Nilsson, Acta Chem. Scand., Ser.
B 1978, 32, 257Ϫ263.
[22]
J. Baran, H. Mayr, J. Org. Chem. 1988, 53, 4626Ϫ4628.
[23]
G. M. Bennett, R. L. Wain, J. Chem. Soc. 1936, 1114Ϫ1120.
1
n2D0 ϭ 1.5061; 23 g (66%). Ϫ H NMR: δ ϭ 6.69 (s, 1 H), 6.58 (s,
Received November 23, 2000
[O00599]
2 H), 2.25 (s, 6 H), 0.91 (t, J ϭ 7.8 Hz, 18 H), 0.54 (q, J ϭ 7.8 Hz,
Eur. J. Org. Chem. 2001, 3985Ϫ3989
3989