Facile halogen exchange reactions: Chloroform with bromoform and carbon tetrachloride with carbon tetrabromide

Article abstract of DOI:10.1021/jo951692q

Both of the title systems undergo rapid halogen exchange (half-life ca. 1-2 min) in N-methylpyrolidinone with catalytic sodium hydroxide at room temperature. Yet they differ markedly in response to added p-dinitrobenzene. The rate of the haloform exchange is unaffected, whereas the rate of the carbon tetrahalide exchange is severely retarded. The known base-induced halogen exchange reaction between chloroform and bromoform is shown not to proceed through a reversible carbene intermediate as claimed in the literature. It appears to be best described in terms of the so-called RARP mechanism (radical anion-radical pair). The mechanism proposed for the rapid exchange between carbon tetrachloride and carbon tetrabromide is initial electron transfer, halide ion loss, and ensuing radical chain scrambling of halogen atoms. The acronym RARC, standing for radical anion-radical chain, is proposed.

Full text of DOI:10.1021/jo951692q

J . Org. Chem. 1996, 61, 4933-4936
4933
F a cile Ha logen Exch a n ge Rea ction s: Ch lor ofor m w ith Br om ofor m
a n d Ca r bon Tetr a ch lor id e w ith Ca r bon Tetr a br om id e
J on A. Orvik
Agricultural Chemicals Process Research, Dow Chemical, USA, Midland, Michigan 48674
Received September 14, 1995X
Both of the title systems undergo rapid halogen exchange (half-life ca. 1-2 min) in N-
methylpyrolidinone with catalytic sodium hydroxide at room temperature. Yet they differ markedly
in response to added p-dinitrobenzene. The rate of the haloform exchange is unaffected, whereas
the rate of the carbon tetrahalide exchange is severely retarded. The known base-induced halogen
exchange reaction between chloroform and bromoform is shown not to proceed through a reversible
carbene intermediate as claimed in the literature. It appears to be best described in terms of the
so-called RARP mechanism (radical anion-radical pair). The mechanism proposed for the rapid
exchange between carbon tetrachloride and carbon tetrabromide is initial electron transfer, halide
ion loss, and ensuing radical chain scrambling of halogen atoms. The acronym RARC, standing
for radical anion-radical chain, is proposed.
In tr od u ction
(d) A more recent report10 emphasized the role of single
electron transfer in the halogen exchange. These authors
claim partial formation of disproportionation products
Ch lor ofor m Exch an ge with Br om ofor m . The halo-
gen exchange in haloforms, such as that between chlo-
roform and bromoform, under basic conditions is well
2 2 4
(CH X and CX ) along with the exchange. They also
known.1
-5
discount the role of halide ion reversal of carbene
formation. They cite ESR support for their interpreta-
tion, but with no experimental details as to how the
reactions or analyses were performed.
This redistribution reaction has served pre-
parative purposes.1 The mechanism of exchange has
been interpreted differently by various workers.
(
a) Earlier reports attributed the exchange to reversible
1
,3
carbene formation in which an opposite halide ion is
encountered by a carbene, which then reverts to a
haloform anion and thence to a haloform with mixed
halogen atoms. This mechanism was based on products
formed and the known reaction of a carbene with halide
ion in solution, originally reported by Hine and co-
Ca r b on Tet r a ch lor id e E xch a n ge w it h Ca r b on
9
Tetr a br om id e. Since we first reported this reaction,
Sasson et al.11 have observed tetrabutylammonium fluo-
ride trihydrate (TBAF)-catalyzed scrambling of halogen
in BrCCl , Br CCl , and Br CCl, each individually. They
3
2
2
3
also observed exchange between haloforms and carbon
tetrahalides, but specifically reported that there was no
exchange between bromoform and chloroform under their
conditions (neat halocarbons and catalytic TBAF at room
temperature). The mechanism proposed for these ex-
changes is polar attack of a haloform anion on a carbon
tetrahalide. When only carbon tetrahalides were present
initially, small amounts of haloform formed in the
mixtures were presumed to be the active agent.
workers.6
,7
4
(b) Another explanation was advanced that a carbene
reacts directly with another haloform molecule to produce
the mixed haloform and a different carbene. This
explanation was based on the fact that phenyl trichlo-
romethylmercury in refluxing bromoform reacted to give
a mixture of chloroform, bromodichloromethane, and
dibromochloromethane (3:1:8 ratio). The authors saw
this process as occurring in addition to the carbene-
halide ion reversal process. A difficulty with this expla-
nation is that the primary process in such a system would
Much of the data on halogen exchange in haloforms
have been gathered in conjunction primarily with studies
of carbene reactions, where the halogen exchange was
essentially a side phenomenon. The studies reported
here on exchange in haloforms and tetrahalides have
focused on the more rapid exchange reaction. In par-
ticular, we were interested in probing the effect on
kinetics of additives, especially p-dinitrobenzene, a typi-
cal radical anion scavenger.12 Like Sasson et al., we too
observed rapid exchange between haloforms and tetra-
halide under our conditions. However, we found the
additive effects (mechanistic probes) difficult to interpret
and therefore focused our efforts on the individual
systems. We believe this approach was fruitful.
be expected to be carbene insertion to give C
2
com-
pounds.8
(
c) The carbene reversal mechanism was rejected on
9
two findings. First, soluble halide added to the system
under exchange conditions showed no halide incorpora-
tion (Br /CHCl and Cl /CHBr ), and second, no cyclo-
3 3
-
-
propane product could be detected when an alkene (1-
octene) was added.
X
Abstract published in Advance ACS Abstracts, J uly 1, 1996.
(
1) Fedorynski, M.; Polawska, M.; Nitchske, K.; Kowalski, W.;
Makosza, M. Synth. Commun. 1979, 7(4), 287.
2) Dehmlow, E. V.; Slopianka, M. Liebigs Ann. Chem. 1979, 10,
465.
(
1
3
(
3) Dehmlow, E. V.; Lissel, M.; Heider, J . Tetrahedron 1977, 33(3),
63.
(10) Xu, L.-X.; Tao, F.-G.; Yu, T.-Y.; Ge, M. T.; Chen, S.-M. Acta
Chim. Sinica 1987, 45, 415-417.
(11) Sasson, Y.; Kitson, F.; Webster, O. W. Bull. Soc. Chim. Fr. 1993,
130, 599-600.
(12) (a) Bunnett, J . F. Acc. Chem. Res. 1978, 11, 413. (b) Bowman,
W. R. Photoinduced Nucleophilic Substitution at sp3 Carbon. In
Photoinduced Electron Transfer; Fox, M. A., Chanon, M., Eds.;
Elsevier: Amsterdam, 1988; Part C, Chapter 4.8, pp 487-552. (c)
Todres, Z. V. Ion-Radical Organic Reactions, Tetrahedron Report No.
187. Tetrahedron 1987, 41(14), 2771-2823.
(
(
(
(
(
(
4) Kimpenhaus, W.; Buddrus, J . Chem. Ber. 1977, 110(4), 1304.
5) Dehmlow, E. V.; Broda, W. Chem. Ber. 1982, 115, 3894-3897.
6) Hine, J .; Dowell, A. M., J r. J . Am. Chem. Soc. 1950, 2438.
7) Hine, J .; Prosser, F. P. J . Am. Chem. Soc. 1958, 80, 4282.
8) Franzen, V. Liebigs Ann. Chem. 1959, 627, 22-27.
9) Orvik, J . A. A Facile Halogen Exchange Reaction; Seventh
IUPAC Conference on Physical Organic Chemistry, Auckland, New
Zealand, 20-24 Aug 1984. Parts of this work were presented at this
meeting.
S0022-3263(95)01692-6 CCC: $12.00 © 1996 American Chemical Society

4
934 J . Org. Chem., Vol. 61, No. 15, 1996
Orvik
F igu r e 2. Effects of additives, DTBN and p-DNB, on the rate
of halogen exchange between CHCl
3 3
and CHBr . Only CH-
BrCl is shown for clarity. Conditions are the same as for
2
Figure 1 except for the additives, expressed as mole percent-
age.
F igu r e 1. Halogen exchange between CHCl
CHBr (16.8 mmol). Conditions: NMP solvent, 5.0 g; 50%
sodium hydroxide catalyst, 0.5 g; rt. At equilibrium the mole
ratios were CHCl ) CHBr ) 1.0; CHBrCl ) CHBr Cl ) 2.5.
3
(16.7 mmol) and
3
area % of the total halocarbons; in a standard reaction
with no additives, such as in Figure 1, this number was
less than 1%). In addition, products of substitution of
one of the nitro groups in p-DNB, apparently by a
trihalomethyl anion, were detected by GC-MS. The
products detected were correct in mass for I and II, with
I being present in much greater abundance, especially
early in the reaction (the para configuration is inferred
from the original isomer configuration). Approximately
3
3
2
2
Resu lts a n d Discu ssion
3 3
Kin etics of Exch a n ge betw een CHCl a n d CHBr .
The exchanges were very rapid in N-methylpyrrolidinone
NMP) with 50% sodium hydroxide at room temperature.
(
Kinetic measurements under the usual conditions of
running the reaction (see Experimental Section) showed
a half-life for exchange of about 1 min. The GC area
percentages for the disappearance of starting materials
and formation of products are shown as Figure 1. The
product ratios are those expected for a statistical distri-
bution from mixing of the halogen atoms.
6
0% of the p-DNB was converted.
Previous workers explained halogen exchange in ha-
loforms as occurring by a reverse reaction of dihalocar-
bene with a halide ion, the halide ion being bromide if
the carbene is dichlorocarbene and vice versa. Reversal
is a known phenomenon,6 but seemed unlikely here
since there is no halide ion of either kind in solution at
the start, not until there is extensive haloform hydrolysis.
Attempts were made to see whether inorganic halide ion
would be incorporated into the haloform. Both bromide
ions with chloroform and chloride ions with bromoform
were tested for incorporation under the normal condi-
tions. In these reaction mixtures the quaternary am-
monium halides dissolved completely. The results were
that in neither case could a trace of product be detected
that would indicate incorporation of external ionic halide.
After a sufficiently long period of time passed for detec-
tion of exchange in each reaction (30-60 min), the
opposite haloform was added to each mixture, resulting
in rapid exchange.
Lack of cyclopropane products when an alkene (1-
octene) was added also indicates that carbenes are not
in evidence during the rapid halogen exchange. The
kinetic effects of both di-tert-butyl nitroxide (DTBN) and
p-dinitrobenzene (p-DNB) as mechanistic probes were
examined. Plots of these results are shown as Figure 2.
Effect of p-DNB. Even at a fairly high concentration,
p-DNB has no effect on the rate of exchange. The
maximum amount of exchange is decreased, however.
Also, an unusually large amount of mixed carbon tetra-
halides was present at the end of the reaction (12.6 GC
The following evidence supports this assignment:
(a) The nitro group on an appropriately deactivated
,7
aromatic ring can often be a better leaving group than
fluoride;13 (b) trihalomethyl anions can substitute on
deactivated aromatic rings;1
4,15
and (c) an example of a
reversible reduction of a trichloromethyl side chain to a
dichloromethyl with chloroform and base has been re-
ported.16 Scheme 1 is proposed as the origin of I and II.
3 3
Mech a n ism of CHCl /CHBr Ha logen Atom Ex-
ch a n ge. The evidence above shows a rapid, base-
catalyzed halogen atom exchange, with lack of kinetic
inhibition by p-DNB. In addition, there were observed
neither two-carbon species nor methylene dihalide spe-
cies. These observations are consistent with a direct
nucleophilic attack of a trihalomethyl anion on a halogen
atom of a haloform. However, the radical anion-radical
pair (RARP) mechanism developed by Meyers and co-
workers17 offers an alternative mechanistic view that is
(13) Beck, J . R. Nucleophilic Displacement of Aromatic Nitro
Groups, Tetrahedron Report No. 54. Tetrahedron 1978, 34, 2057-
2
068.
(14) Orvik, J . A. U.S. Patent 4,739,070, 19; Chem. Abstr. 1988, 108,
04500w.
15) Atkinson, P. J .; Gold, V. J . Chem. Soc., Chem. Commun. 1983,
140.
2
(
(
16) Orvik, J . A. J . Org. Chem. 1994, 59, 12-17.
(17) Meyers, C. Y.; Matthews, W. S.; Ho, L. L.; Kolb, V. M.; Parody,
T. E. Catalysis in Organic Synthesis, 6th ed.; Smith, G. J . V. J ., Ed.;
Academic: New York, 1977; pp 197-278.

Facile Halogen Exchange Reactions
J . Org. Chem., Vol. 61, No. 15, 1996 4935
Sch em e 1
Sch em e 2
F igu r e 3. Halogen exchange between CCl
CBr (3.0 mmol). Conditions: NMP solvent, 5.0 g; 50% sodium
hydroxide catalyst, 0.5 g; rt. CCl not shown for reasons of
used due to limited solubility
4
(26 mmol) and
4
4
scale. Smaller mole ratio of CBr
in the system.
4
also consistent with the observed facts. Scheme 2
outlines the two mechanisms as applied to the halogen
exchange in haloforms. The essential difference in these
two proposals is the electron transfer characteristics of
the bracketed pair that exchanges halogens. The first
reactive species is presumed to be the tribromomethyl
anion (bromoform deprotonates much more rapidly than
chloroform),18 which is assumed to be at a steady state
concentration. This attacks a chloroform molecule with
a net transfer of a chlorine radical to the bromoform
anion within a cage. This combination rapidly exchanges
halogens to a more or less statistical distribution before
diffusing apart as a new haloform-haloform anion pair.
Direct anion attack (the bottom bracketed pair) would
be favored by the principle of Ockham’s razor but appears
to be more highly energetic than electron transfer.
Calculations from thermodynamic data estimate an
uphill energy of about 20 kcal/mol for the ionic transfer
versus about 11-12 kcal/mol uphill for the electron
transfer.19 While such calculations are approximate, they
are supportive of the RARP mechanism. The key ex-
change reaction occurring within a solvent cage accounts
for the lack of inhibition by p-DNB. Our best explanation
of the rate-enhancing effect of DTBN is that it is able,
by electron acceptance from a tribromomethyl anion, to
initiate a few radical chains via the tribromomethyl
radical.20
F igu r e 4. Effect of p-DNB on the rate of halogen exchange
betweeen CCl and CBr Conditions are the same as for
Figure 3, except for the indicated mol % of p-DNB.
4
4
.
Sch em e 3
Kin etics of Exch a n ge betw een CBr
4 4
a n d CCl .
This reaction shows a decidedly different pattern of
additive effects on the kinetics of exchange. As Figure 4
shows, p-DNB is a strong inhibitor. Both Ionol and
DTBN (not shown) were found to accelerate the reaction.
Reliance on the effect of DTBN as a mechanistic probe
seems doubtful due to its different possible modes of
reaction.20 However, the pattern of Ionol (rate enhanc-
ing) and p-DNB (rate retarding) is the same as reported
for the effect on the halogen exchange between chloro-
16
form and 2-chloro-6-trichloromethylpyridine. Thus the
probability is strong for a similar mechanism (Scheme
(
18) Hine, J . Divalent Carbon; Ronald Press: New York, 1964; p
3
). Electron transfer from hydroxide (or Ionol anion) to
3
9.
(
a carbon tetrahalide followed by halide ion loss to
generate a trihalomethyl radical constitutes initiation.
Propagation is through a series of reversible reactions
with a carbon tetrahalide. Termination reactions are
speculative, the only evidence for which may be the
19) Lias, S. J .; Bartmess, J . E.; Liebman, J . F.; Holmes, J . L.; Levin,
R. D.; Mallard, W. G. J . Phys. Chem. Ref. Data 1988, 17 (Suppl. 1).
20) Known ability of DTBN to accept electrons: (a) Forrester, A.
(
R.; Hay, J . M.; Thomson, R. H. Organic Chemistry of Stable Free
Radicals; Academic Press: New York, 1968; Chapter 7. (b) Rossi, R.
A.; Bunnett, J . F. J . Org. Chem. 1973, 38, 1407.

4
936 J . Org. Chem., Vol. 61, No. 15, 1996
Orvik
appearance of a small amount of chloroform indicating
reduction. However, this is extremely small. In the
standard reaction (Figure 3), only 0.05% was detected
by 9 min, with nothing before that.
A convenient shorthand notation to the pathway
described in Scheme 3 would be RARC, for radical anion-
radical chain.
reaction. As mentioned above, no traces of haloforms
were detected in the carbon tetrachloride-carbon tetra-
bromide exchange (half-life of 2-3 min) until equilibrium
was nearly complete. This is difficult to account for by
a direct polar reaction.
Exp er im en ta l Section
Can the mechanism described by Sasson11 account for
our results? A simple anionic attack of a trace of
bromoform anion on a chlorine of carbon tetrachloride,
for example, cannot account for the strong inhibition by
p-DNB or the rapid rate of reaction. An exchange
between bromoform and carbon tetrachloride where both
were substantially equimolar had a half-life of about 1
min. This would mean that vanishingly small amounts
of adventitious haloform should lead to extremely slow
The general experimental procedure is the same as that
given previously.16 Solvent NMP was “Gold Label” from
Aldrich. Commercial chemicals were used without further
purification. All identifications were made by GC-MS.
Ack n ow led gm en t. I thank Mr. Al Friedle for the
GC-MS results.
J O951692Q