Checked by William E. Parham and Siemen Groen.
1. Procedure
2. Notes
1. The empirical formula of
n-butylmagnesium chloride prepared in
methylcyclohexane cannot readily be determined because of the virtual insolubility of the reagent in this medium. The reagent is somewhat more soluble in aromatic media such as
toluene or
isopropylbenzene, and, although the empirical formula of the solute may initially approach C
4H
9MgCl, there is a tendency for precipitation of
magnesium chloride from solution. This process appears to be catalyzed by traces of alkoxides, which are liable to be formed after contact of
oxygen with the solution. In practice, products will tend to contain less halogen than is required by the simple formula C
4H
9MgCl. The reagents are associated (see reference
2 for a fuller discussion).
2. The
nitrogen used (British Oxygen Co., White Spot) contained about 10 p.p.m. of
oxygen and was dried by passage through a glass spiral cooled in
acetone and solid
carbon dioxide. For the most precise work, the submitters reduced the proportion of
oxygen to about 0.1 p.p.m. by scrubbing the
nitrogen with
chromous chloride solution in a
Nilox apparatus (Southern Analytical Ltd., Camberley, Surrey, England).
3. The apparatus should preferable be baked at 120° for several hours immediately before use. The uppermost region of condensing
methylcyclohexane should not be cloudy; if it is, a few milliters should be allowed to distil.
5.
Magnesium powder (grade 4, Magnesium Elektron Ltd., Manchester, England) was used within 6 months of its grinding by the manufacturer, and was sieved to the stated particle size. The use of unsieved material often gives results nearly as good, but exact reproducibility is more difficult because of variations of the particle size distribution from sample to sample. In general, the more freshly ground the
magnesium, the shorter are the induction periods before reaction and, to a limited extent, the higher are the yields of organomagnesium product.
6. Fresh
magnesium turnings for Grignard reaction can be used if suitable powder is unavailable, but initiation of reaction is likely to be prolonged, and the subsequent addition of the halide solution should occupy at least 30 minutes, longer if possible.
7.
Methylcyclohexane is purified by shaking with 3 portions of concentrated
sulfuric acid, washing successively with water,
sodium carbonate solution, and water, drying over
calcium sulfate (Drierite), and distilling. The material boiling at
100–101° is used. Other nonsolvating media which can be used are
toluene, xylenes,
cumene,
tetralin,
light petroleum (b.p.
80°),
decalin, and kerosene; aliphatic media are preferred, for reasons given in references
2,
3, and
4.
8. The rate of flow of
nitrogen should be just sufficient to maintain a positive pressure in the apparatus. Too rapid a flow leads to loss of
1-chlorobutane.
10. The yield of product is increased to
81% (analyzed by evolution of
n-butane) if
0.67 g. (0.0033 mole) of aluminum isopropoxide is added to the suspension of
magnesium before addition of the halide solution. Alternatively, an equivalent amount of
2-propanol and
iodine (giving 0.01 mole of C
3H
7OMgI) may be added. These modified procedures (particularly the second) also shorten the induction periods and render unnecessary any special drying of the reagents and apparatus and the use of fresh
magnesium.
The products in such cases contain complexes between
n-butylmagnesium chloride and the particular alkoxide employed. With the stated low proportions of alkoxides, these complexes broadly resemble the alkoxide-free materials, but increased proportions of the alkoxide component give complexes having generally decreased chemical reactivity (see references
3 and
4).
11. The reaction generally starts without addition of
iodine as an initiator, but the use of a crystal of
iodine (no stirring) may occasionally be necessary with "old"
magnesium or insufficiently dried materials or apparatus. A slower rate of addition of
1-chlorobutane gives slightly higher yields; for example, addition over a period of 60 minutes gave yields of
82–87%.
12. The reflux condenser was connected by an
adaptor and Teflon tube to a trap of known weight which was cooled by a mixture of
acetone and solid
carbon dioxide. The flow of
nitrogen was stopped, and an excess of water (about 15 ml.) was added dropwise through the dropping funnel to the stirred reaction product. The resulting mixture was heated at the reflux temperature, and the
butane was collected in the trap. The weight of
butane, b.p.
−1° to 0°, was
4.23–4.35 g. (
73–76% yield).
14. Sufficient dry
ether (approximately 100 ml.) is added to bring the organomagnesium products into solution. Aliquot portions of the solution are then added to a known volume of standard
hydrochloric acid, and the excess acid is determined by titration with standard base. Yields determined in this way tend to be a few percent higher than those determined by collection of
n-butane (Note
12).
3. Discussion
The method is an extension of the well-known Grignard synthesis in ethers to the use of nonsolvating media, and is a development of procedures previously reported.
2,3,4,5,6 A version of it has been employed with straight-chain primary alkyl chlorides, bromides, and iodides from C
2 to C
14,
5,6,7 and in solvents (or an excess of the halide) which permit reaction temperatures above 120°, with simple aryl halides such as
chlorobenzene and
1-chloronaphthalene. Branched-chain primary, secondary, and tertiary alkyl halides, allyl, vinyl, and benzyl halides either fail to react or give extensive side reactions. Better results are reported to be obtained in such cases with the use of catalytic quantities of a mixture of an alkoxide and an
ether such as
diethyl ether or
tetrahydrofuran in a hydrocarbon medium, but the products are not, of course, completely unsolvated.
4
4. Merits of the Preparation
Copyright © 1921-2002, Organic Syntheses, Inc. All Rights Reserved