3}4.

PHYSICAL ORGANIC CHEMISTRY

starting material. In fact, esters that have only one a-hydrogen atom
do not undergo the acetoacetic ester condensation in the presence of
sodium alkoxidc catalysts, because the fl-keto ester formed would not be
sufficiently acidic to be transformed to its conjugate base to any signifi-
cant extent. The equilibrium in such reactions, however, may be forced
to the right by the use of more strongly basic catalysts. Triphenyl-
methylsodium, mesitylmagnesium bromide, sodium amide, and diiSO-
propylaminomagnesium bromide arc among the bases that have been
found capable of bringing about the formation of fl-keto esters that con-
tain no a-hydrogen atoms. u.u
13-3b. Alkaline Cleavage of fl-Diketones and Related Compounds. In
general, compounds of the type RCOX may be easily cleaved by alkali
if X
6
is a fairly stable anion (cf. the alkaline hydrolysis of esters). It is
therefore not unexpected that fl-diketones are fairly readily cleaved by
bases. Pearson and Sandy studied the cleavage of acetylacetone by
sodium ethoxide in ethanoL
17
They made the observation, surprising
at first glance, that the reaction is of the first order, its rate being propor-
tional to the concentration of the reactant present in the smaller con-
centration. This observation is quite rational, however, in vi ew of the
following reaction mechanism:
e e
CH,COCHsCOCHa + EtO CH,COCHCOCH, + EtOH
e
0 0 0 0
+ EtO
8


6Et
e
-> CH,CO,Et + CH,COCH,
e e
CH,COCHa + EtOH-> CH,COCH, + EtO
If we assume that the rate-controlling step is the decomposition of the
acetylacetone-ethoxide-ion adduct and that all the other steps are rela-
tively fast, we can derive the kinetic equation
v = k[CH3COCH.COCH,][EtO- ]
However, if the acid-base equilibrium of the first equation is established
rapidly,
[CHaCOCH.COCH,][EtO- ] =
Therefore v =
(13-2)
"C. R. Hauser et al., J . Am. Chern. Soc., 68,2647 (1946); 69, 2649 (1947); 71, 1350
(1949).
., R. G. Pearson and A. C. Sandy, J . Am. Chem. Soc., 73,931 (1951).
I
c-c BOND FORMATION AND CLEAVAGE IN RC02H DERivATIVES 315
Acetylacetone is an acid of such a strength that the concentration of its
anion present in a basic solution will be essentially equal to the concentra-
tion of sodium ethoxide or of acetylacetone added, depending on which is
the limiting reagent. The mechanism given, then, explains the observa-
tion of first-order kinetics. The alkaline ethanolysis of phenacylpyridin-
ium ions follows a similar kinetic equation and has the same type of
mechanism.
17
The alkaline hydrolysis of acetylacetone is a more complicated reac-
tion, the kinetic equation having the form of (13-2) with an additional
term involving hydroxide ion.
11
v = +
The first term corresponds to cleavage by a mechanism like that followed
in alcoholic solution. As Pearson and Mayerle point out, the second
term, involving an additional hydroxide ion, is probably due to cleavage
by the mechanism
e e
0 .0
I oH- I e e
===' CHaCCH,COCHa-+ CHaCO. + CH.COCHa
6H 6e
Although the doubly charged intermediate must be present at a much
smaller concentration than the singly charged one, its specific rate con-
stant for cleavage is probably much larger. This situation is somewhat
similar to that found in the Cannizzaro reaction (Sec. ll-4a).
Dimethylacetylacetone cannot' be "protected" by transformation to
its conjugate base and is therefore cleaved much more rapidly than the
unsubstitute<I compound or its monomethyl derivative.
0
II
CH,CC(CH,).COCH, + OH- ...... CH,CO.- + (CH,).CHCOCH,
The kinetics in this case are second-order, first-order in hydroxide ion
and first-order in ketone.
18
The study of the basic cleavage of chloral hydrate carried out by
Gustafsson and Johanson showed that this part of the haloform reaction
probably has a mechanism of the type described for the cleavage of
fl-diketones.
11
This mechanism should be modified to agree with the
observation of general base catalysis," and, indeed, it is possible that the
reactions studied by Pearson and coworkers are also subject to general
base catalysis.
13
R. G. Pearson and E. A. Maycrl c, J . A m. hem. Soc., 73,926 (1951) .
10
C. Gustaf880n and M. Johanson, A eta Chern . Scand., 2, 42 (1948).
316 PHYSICAL ORGANIC CHEMISTRY
. PROBLEMS
. I. Suggest a mechaniam to explain why potassium cyanide is a catal
m the tranaformatlon of RCOC0
2
H to RCHO. yst
2. a plau.sible mechanism for the reaction of threonine (CH,CHOHCHNH,_
CO,H) wtth and aluminum iona. The products include conaiderable
acetaldehyde and g)ycme and less a-ketobutyric acid and ammonia .
. 3. equilibrium c?nstants, as in the explanation in Sec. 13-3a, explain how the
mued Clll.lBen condensatiOn of ethyl benzoate with ethyl isobutyrate is driven to th
right by the use of triphenylmethylsodium. e
4. Suggest a reasonable mechaniam for the alkaline cleavage (general base-eata.-
lyzed) of chloral hydrate to chloroform and formate iona.
Chapter }4.
REARRANGEMENTS DUE
TO ELECTRON-DEFICIENT CARBON
Several types of rearrangement reactions have already been described
under various other headings. The Stevens rearrangement has been
mentioned as an example of an SNi reaction (Sec. 6-3); allylic rearrange-
ments by the SNl and SN2' mechanisms were discussed in Sec. 6-5; and
the Favorskii rearrangement in Sec. l0-3c. Other carbon-skeleton
rearrangements will be discussed here.
14.-1. Carbonium-ion Rearrangements. Of the very large number
of carbonium-ion rearrangements we shall discuss only a few, most of
them chosen because of the care with which they have been studied. It is
hoped, nevertheless, that certain generalizations pointed out in this and
subsequent sections will lead to an understanding of the nature of most
of the rearrangements not mentioned specifically.
14-la. Evidence that Carbonium Ions Are Intermediates. Meerwein
and van Ernster obtained some of the first strong evidence that car-
bonium ions may be intermediates in carbon-skeleton rearrangements
in their study of the rearrangement of camphene hydrochloride to iso-
bornyl chloride. They observed that the reaction rate increased with
the nature of the solvent in roughly the same order that had been found
for the ability of solvents to ionize triphenylmethyl chloride. They
further established that compounds such as HgCh, FeCI,, SnCI., etc.,
capable of coordinating with a chloride ion, were excellent catalysts for
the reaction. ft was later shown that the rearrangement of the chloride
was slower than that of the bromide or arylsulfonate but faster than that
of the trichloroacetate or m-nitrobenzoate.' This order of reactivity is
understandable, since the ease of ionization should increase with the
stability of the anion being formed.
The probability of a carbonium-ion mechanism for rearrangement
reactions was emphasized (in somewhat different language) by Whitmore
1
H. Meerwein and K. van Emster, Btr., 55, 2500 (1922) .
1
H. Meerwein, 0 . Hammel, A. Serini, nnd J . Vorster, Ann., 453, 16 (1927).
31 7