The rate retarding effect of alkyl groups on arene metalation quantified

Article abstract of DOI:10.1016/S0040-4039(99)02160-7

Metalation/trapping sequences applied to relatively inert, tert-alkyl substituted arenes tend to give poor yields unless the reagent is used in high concentrations. Under optimized conditions, even 1,4-bis(tert- butyl)benzene, 1,1,3,3-tetramethylindane an

Full text of DOI:10.1016/S0040-4039(99)02160-7

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 667–670
The rate retarding effect of alkyl groups on arene metalation
quantified
Eckhard Baston, Qian Wang and Manfred Schlosser ∗
Section de Chimie (BCh), Université, CH-1015 Lausanne, Switzerland
Received 21 October 1999; accepted 12 November 1999
Abstract
Metalation/trapping sequences applied to relatively inert, tert-alkyl substituted arenes tend to give poor yields
unless the reagent is used in high concentrations. Under optimized conditions, even 1,4-bis(tert-butyl)benzene,
1,1,3,3-tetramethylindane and 1,1,2,2,3,3-hexamethylindane can be smoothly converted into derivatives. Compe-
tition experiments enable the quantitative assessment of tert-alkyl substituent effects on the metalation rates at
aromatic ortho, meta and para positions. Unlike alkyl groups, hetero elements generally accelerate metalation
reactions. 1,1,3,3-Tetramethyl-1,3-dihydroisobenzofuran undergoes the hydrogen/metal exchange 3–6 times faster
than the indanes mentioned above, the reaction occurring at both the 5- and at the 4-position. © 2000 Published by
Elsevier Science Ltd. All rights reserved.
One decade ago, we reported the metalation and subsequent transformation of tert-butylbenzene and
related ‘spiny’ hydrocarbons.1 However, yields exceeding 50%, as reported,1 proved to be difficultly
reproducible. As it turned out, satisfactory yields can be achieved only when the substrate and the
superbasic mixed-metal reagent (LIC–KOR)2,3 are employed in the 1.5–2.0 M range (see Fig. 1).
Fig. 1. Consecutive reaction of 1,3-bis(tert-butyl)benzene with butyllithium potassium tert-butoxide (suspension in hexanes)
and dry ice: yields of isolated 3,5-bis(tert-butyl)benzoic acid as a function of the metalation time (t) and concentration
∗
Corresponding author. Fax: ++41-21-692-39-65.
0040-4039/00/$ - see front matter © 2000 Published by Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(99)02160-7
tetl 16088

668
Employing a high-concentration (1.5 M) protocol, it was possible for the first time to attack a
position ortho to a bulky tert-butyl group. When a suspension of potassium tert-butoxide in cyclohexane
containing equivalent amounts of butyllithium and 1,4-di(tert-butyl)benzene was stirred for 25 h at 25°C
before being poured on dry ice and neutralized, 10% of 2,5-di(tert-butyl)benzoic acid (1) were formed.
The yield raised to 24% when the reaction was conducted at 75°C under otherwise identical conditions.
In view of the difficulties to prepare 1,2-di(tert-butyl)benzene,4 the readily accessible 1,1,3,3-
tetramethylindane5 and 1,1,2,2,3,3-hexamethylindane6 were selected as surrogates. Superbase promoted
metalation proceeded smoothly (24 h at 25°C in hexanes) and, after carboxylation, the 5-(1,1,3,3-
tetramethyl)indanecarboxylic acid (2) and the 5-(1,1,2,2,3,3)-hexamethylindane carboxylic acid (3), were
isolated in 51 and 55% yield, respectively.
The oxa-analogous 1,1,3,3-tetramethyl-1,3-dihydroisobenzofuran7 afforded two regioisomers
side by side.8 The main product (23%) was the already described 5-(1,1,3,3-tetramethyl-
1,3-dihydroisobenzofuran)carboxylic acid (4a).9 The minor isomer 4-(1,1,3,3-tetramethyl-1,3-
dihydroisobenzofuran)carboxylic acid (4b; 12%) appears to have been previously overlooked.
Competition experiments1,10 were carried out to quantify the substituent effects on the metalation rates
(see Tables 1 and 2).11 All ‘spiny’ hydrocarbons were found to react more slowly than benzene itself,
the sterically shielded 1,4-di(tert-butyl)benzene being clearly the most inert substrate. As the partial rate
factors of simple substrates (e.g., tert-butylbenzene, see Table 1, or 1-methyl-1-phenylcyclopropane1)
reveal, alkyl substituents seem to exert a slightly stronger rate retarding effect when located at the meta
rather than the para position relative to the deprotonation site.
In comparison with the 1,1,3,3-tetramethylindane case, the additional pair of geminal methyl groups
present in 1,1,2,2,3,3-hexamethylindane impedes the metalation of the latter substrate by a factor of
two (see Table 2). In contrast, 1,1,3,3-tetramethyl-1,3-dihydroisobenzofuran reacts with the superbase
three times faster than the oxygen-free reference compound (Table 2). If one breaks down this overall
effect according to the competing regioisomeric channels, metalation at the positions 5 and 4 proves
to be accelerated twofold and at least 20-fold, respectively. At present, it is not clear whether the
inductive electron-withdrawing effect of the heteroatom is transmitted through the σ-bonded skeleton
or by polarization of the π-cloud of the aromatic electron sextet. As molecular models show, a direct
coordination of the oxygen atom to the proton abstracting reagent can be ruled out for reasons of
congestion. Therefore, no analogy should be drawn to the ortho-lithiation of methyl triphenylmethyl
ether12 which in a sequence of subsequent steps is converted into 9-phenyl-9-fluorenyllithium while
1,1,1-triphenylethane is totally unreactive under identical conditions.

669
Table 1
Metalation of tert-alkyl substituted benzenes by means of the LIC–KOR superbase in hexanes
(HEX) or tetrahydrofuran (THF): relative rates and partial rate factors[a] as assessed by competition
experiments
Table 2
Metalation and subsequent carboxylation of 1,1,3,3-tetramethylindane, 1,1,2,2,3,3-hexamethylindane
and 1,1,3,3-tetramethyl-1,3-dihydroisobenzofuran by means of the LIC–KOR superbase: relative rates
and partial rate factors[a] as assessed by competition experiments

670
Acknowledgements
This work was supported by the Schweizerische Nationalfonds zur Förderung der wissenschaftlichen
Forschung, Bern (grant nos. 20-49’307-96 and 20-55’303-98).
References
1. Schlosser, M.; Choi, J. H.; Takagishi, S. Tetrahedron 1990, 46, 5633–5648.
2. Schlosser, M. J. Organomet. Chem. 1967, 8, 9–16; Schlosser, M.; Strunk, S. Tetrahedron Lett. 1984, 25, 741–744.
3. Schlosser, M. Modern Synthetic Methods 1992, 6, 227–271.
4. Hoogzand, C.; Hübel, W. Angew. Chem. 1961, 73, 680; Barclay, L. R. C.; Milligan, C. E.; Hall, N. D. Can. J. Chem. 1962,
40, 1664–1671; Burgstahler, A. W.; Chien, P.-L.; Abdel Rahman, M. O. J. Am. Chem. Soc. 1964, 86, 5281–5290; Nakayama,
J.; Yamaoka, S.; Nakanishi, T.; Hoshino, M. J. Am. Chem. Soc. 1988, 110, 6598–6599; Nakayama, J.; Hirashima, A. J. Am.
Chem. Soc. 1990, 112, 7648–7653.
5. Bogert, M. T.; Davidson, D. J. Am. Chem. Soc. 1934, 56, 185–190; Sikkar, R.; Martinson, P.; Nilsson, B. Acta Chem. Scand.
Ser. B 1978, 32, 257–263.
6. Baran, J.; Mayr, H. J. Org. Chem. 1988, 53, 4626–4628.
7. Bennett, G. M.; Wain, R. L. J. Chem. Soc. 1936, 1114–1120.
8. As expected (see: Schlosser, M.; Geneste, H. Chem. Eur. J. 1998, 4, 1969–1973), the superbasic mixture of tert-butyllithium
and potassium tert-butoxide deprotonates the ‘spiny’ isobenzofuran exclusively at the 5-position.
9. Shi, G.-Q.; Cottens, S.; Shiba, S. A.; Schlosser, M. Tetrahedron 1992, 48, 10560–10574.
10. Ingold, C. K.; Shaw, F. R. J. Chem. Soc. 1927, 2918–2929; Schlosser, M.; Ladenberger, V. Chem. Ber. 1967, 100,
3901–3915.
11. The consumption of starting materials as the formation of derivatives thereof was monitored by gas chromatography (30
m, silicon rubber SE-30, 120°C; 30 m, silicon rubber DB-1701, 130°C), the corresponding peak areas being integrated
relative to those of reference compounds (‘internal standards’) and being corrected by calibration factors. New products
were characterized by their NMR and mass spectral data and by correct elementary analyses.
12. Gilman, H.; Meikle, W. J.; Morton, J.W. J. Am. Chem. Soc. 1952, 74, 6282–6284.