Rate constant for the ring opening of the 2,2-difluorocyclopropylcarbinyl radical

Article abstract of DOI:10.1021/ol005578s

(figure presented) The rate constant for the unimolecular ring opening of the 2,2-difluorocyclopropylcarbinyl radical was determined via its competitive bimolecular trapping by TEMPO. The value of this rate constant (3.4 × 10¹¹ s^-1 a

Full text of DOI:10.1021/ol005578s

ORGANIC
LETTERS
2000
Vol. 2, No. 6
835-837
Rate Constant for the Ring Opening of
the 2,2-Difluorocyclopropylcarbinyl
Radical
Feng Tian and William R. Dolbier, Jr.*
Department of Chemistry, UniVersity of Florida, GainesVille, Florida 32611-7200
wrd@chem.ufl.edu
Received January 24, 2000
ABSTRACT
The rate constant for the unimolecular ring opening of the 2,2-difluorocyclopropylcarbinyl radical was determined via its competitive bimolecular
trapping by TEMPO. The value of this rate constant (3.4 × 1011 s-1 at 99.3 °C) is about 500 times larger than that of the parent, unfluorinated
radical and about 5 times smaller than that of the trans-2-phenylcyclopropylcarbinyl radical.
The experimental determination of rate constants of unimo-
lecular rearrangements of radicals, so-called “radical clocks”,
has been an area of significant research interest for the past
20 years.1 The cyclopropylcarbinyl to allylcarbinyl radical
rearrangement2 has attracted particular attention in recent
years because its family of derivatives provides ultrafast
mechanistic probes of radical intermediacy and lifetime,3 with
rate constants ranging from 1.2 × 108 s-1 for the parent to
3 × 1011 s-1 for the 2-phenylcyclopropylcarbinyl radical.4
The 2,2-difluorocyclopropylcarbinyl radical, 1, also un-
dergoes an extraordinarily fast and regiospecific unimolecular
ring opening distal to the geminal fluorine substituents to
form the 2,2-difluoro-3-butenyl radical, 2 (Scheme 1).
parameters for this rearrangement (∆H‡) 1.6 kcal/mol;
∆S‡ ) -2 cal/deg), a rate constant of 1.5 × 1011 s-1 for this
ring opening at 25 °C has been calculated.5 Previous attempts
to measure this rate constant experimentally, using neat
n-Bu3SnH6 and PhSeH7 as H-transfer radical traps, led to
no observable amount of non-ring-opened products in either
case, and therefore such results only provided measures of
the minimum rate constant (>108 s-1 and >3 × 1010 s-1,
respectively).
Nitroxyl radicals, such as 2,2,6,6-tetramethylpiperidin-1-
oxyl (TEMPO),8 are very efficient radical traps that, because
of their large rate constant for trapping (kT ) 1.2 × 109 M-1
s-1 at 25 °C)8,9 and because of one’s ability to detect very
low concentrations of their products of radical trapping (by
HPLC/ESI-MS), provided us with perhaps our last op-
portunity to observe an operational bimolecular competition
involving the trapping of 1 in competition with its ring
opening to 2 (Scheme 2).
Scheme 1
On the basis of earlier work by Beckwith dealing with
competitive nitroxyl radical trapping,10 tert-butyl perester 5
(5) Tian, F.; Bartberger, M. D.; Dolbier, W. R., Jr. J. Org. Chem. 1999,
64, 540.
Although experimental calibration of this process has not
been accomplished until now, using computed activation
(6) Dolbier, W. R., Jr.; Sellers, S. F. J. Am. Chem. Soc. 1982, 104, 2494.
(7) Tian, F. Ph.D. Dissertation, University of Florida, Gainesville, FL,
1999.
(1) Griller, D.; Ingold, K. U. Acc. Chem. Res. 1980, 13, 317.
(2) Newcomb, M.; Glenn, A. G. J. Am. Chem. Soc. 1989, 111, 275.
(3) Nonhebel, D. C. Chem. Soc. ReV. 1993, 347-359.
(4) Newcomb, M.; Johnson, C. C.; Manek, M. B.; Varick, T. R. J. J.
Am. Chem. Soc. 1992, 114, 10915.
(8) Chateauneuf, J.; Lusztyk, J.; Ingold, K. U. J. Org. Chem. 1988, 53,
1629.
(9) Beckwith, A. L. J.; Bowry, V. W. J. Org. Chem. 1988, 53, 1632.
(10) Beckwith, A. L. J.; Bowry, V. W. J. Am. Chem. Soc. 1994, 116,
2710.
10.1021/ol005578s CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/01/2000

process that competes favorably with the desired unimo-
lecular homolytic decomposition process of 5.
Scheme 2
Although the intervention of this polar process lowers the
yield for the desired homolytic process, it should not affect
the diagnostic efficacy of the kinetically controlled ratio of
trapped products 3 and 4, which would derive strictly from
the homolytic process. Indeed, careful HPLC/ESI-MS
analysis of the aforementioned product mixture indicated the
presence of two products (ratio, ∼100:1) that had a molecular
weight consistent with their deriving from TEMPO trapping
of the radical species 1 and 2. To unambiguously identify
the minor (presumably non-ring-opened) product, 3 along
with presumed major product 4 were synthesized by alterna-
tive routes, the products were fully characterized,11 and
mixtures of the two authentic materials were subjected to
HPLC/ESI-MS analysis. Their retention times were coin-
cident with and their mass spectra were identical in all ways
with those of the two products from the TEMPO trapping
experiments of 5. Therefore, it appears that a successful
competitive trapping of 1 prior to its ring opening to 2 has
occurred.
To relate the experimental total ion count integration of 3
and 4 to their relative concentrations, it was necessary to
determine their relative sensitivities to the analytical method.
A relative response coefficient for 3 versus 4 of 0.71 ( 0.10
was obtained using pure samples of 2 and 3 under analytical
conditions that approximated those used in the competition
study.19 The actual competition study was carried out by
decomposing perester 5 (0.012 M in cyclohexane) at 99.3
°C in the presence of varying concentrations of TEMPO
(0.5-1.1 M) and observing the ratio of products 3 and 4 (as
measured by HPLC/ESI-MS) as a function of TEMPO
concentration.20 The obtained ratios of 3:4, after correction
by the response coefficient, ranged from 0.0069 at 0.5 M
TEMPO to 0.0139 at 1.1 M TEMPO. A plot of these ratios
versus [TEMPO] yielded a slope of 1.2 ( 0.2 × 10-2 (r2 )
0.945), which according to eq 1 (Scheme 2) is equal to the
ratio of kT/kr.
appeared to be the precursor of choice for this competition
study of 1 f 2. Perester 511 was synthesized readily from
3-butenyl benzoate via a novel difluorocyclopropanation
procedure12 followed by standard hydrolytic, oxidative, and
functionalization methodologies. When heated to 99.3 °C
in cyclohexane, in the absence of TEMPO, 5 underwent
smooth, first-order decomposition to give a product mixture
that, by 19F analysis, consisted only of ring-opened prod-
ucts.13 The mechanism of the thermal decomposition of
peresters (RCO3-tBu), in cases where R is a primary alkyl
group, is likely a two-step process involving reversible O-O
•
bond homolysis, followed by loss of CO2 by the RCO2
radical to form R•.14,15 The fact that the rate constant for
decomposition of perester 5 (kd ) 2.1 (( 0.2) × 10-5 s-1)
is only 5.5 times larger than that of tert-butyl peracetate16 is
consistent with the 2,2-difluorocyclopropylcarbinyl radical
not being a significantly stabilized radical, relative to a simple
primary system.
A 0.01 M solution of 5 in cyclohexane, containing a large
excess (0.6 M) of TEMPO, was observed to decompose with
enhanced rate,17 and two products could be detected by 19F
NMR, the ring-opened, TEMPO-trapped product 4 and the
ring-closed carboxylic acid 7 (ratios of 7:4 ) 2.3 and 3.5
for 0.6 and 1.0 M TEMPO, respectively) (Scheme 3).
Beckwith’s determined activation parameters9 for kT [log
kT ) 9.7 - 0.9θ] allow one to calculate the rate constant kT
Scheme 3
(11) Products are fully characterized by 1H, 13C, and 19F NMR
spectroscopy and by HRMS and/or elemental analysis.
(12) Tian, F.; Kruger, V.; Bautista, O.; Duan, J.-X.; Li, A.-R.; Dolbier,
W. R., Jr.; Chen, Q.-Y. Org. Lett. 2000, 2, 563-4.
(13) Ring-opened and non-ring-opened products are readily distinguished
by their distinctive 19F NMR signals: the former always appears as a single
fluorine signal, whereas the latter always appears as an AB pattern (two
fluorine signals).
(14) Koenig, T.; Huntington, J.; Cruthoff J. Am. Chem. Soc. 1970, 92,
5413.
(15) Neuman, R. C.; Behar, J. V. J. Am. Chem. Soc. 1969, 91, 6024.
(16) Bartlett, P. D.; Hiatt, R. R. J. Am. Chem. Soc. 1958, 68, 1398.
(17) The overall rates of decomposition of 5 were found to be dependent
on [TEMPO], with pseudo-first-order kobs ) 4.0 and 4.9 × 10-5 s-1 for
[TEMPO] ) 0.6 and 1.0 M, respectively.
(18) Moad, G.; Rizzardo, E.; Solomon, D. H. Tetrahedron Lett. 1981,
22, 1165.
(19) Obtained by comparing ratios of experimental integrations of total
ion current for the respective products from the kinetic experiments with
the ratios for samples containing the two products in authentic ratios between
1:1 to 1:300, where the concentrations (total ion current) of the larger
component, 4, always approximated those in the kinetic study.
(20) To minimize chromatographic ambiguities by largely quenching the
t-BuO• radicals, 5 vol % of 1,4-cyclohexadiene was added to the mixture.
Carboxylic acid product 7 appears to have been formed via
the previously reported polar mechanism involving electron
transfer from TEMPO,18 a bimolecular, TEMPO-induced
836
Org. Lett., Vol. 2, No. 6, 2000

at 99.3 °C to be 1.5 × 109 M-1 s-1, which yields the value
of 1.3 × 1011 s-1 for rate constant kr.21 This value of kr is
consistent with that (3.4 × 1011 s-1 at 99.3 °C) calculated
from the earlier-mentioned computed activation parameters
for the rearrangement of 1 to 2, which were derived using
conventional transition state theory at the UB3LYP/6-
311+G(2df,2p)//UB3LYP/6-31G(d) level of theory.5 Rate
constant kr should exhibit little temperature dependence, and
assuming the preexponential factor of the computed param-
eters, a value of 6 × 1010 s-1 can be considered valid for kr
at room temperature.
Thus, a careful kinetic study involving competitive bimo-
lecular trapping of the 2,2-difluorocyclopropylcarbinyl radical
by TEMPO in competition with its unimolecular rearrange-
ment to the 2,2-difluoro-3-butenyl radical has resulted in the
first determination of the rate constant for this rearrangement,
the value of which is about 500 times larger than that of
cyclopropylcarbinyl itself and 5 times smaller than that of
the 2-phenylcyclopropylcarbinyl radical. Considering its
modest steric demand, this radical clock should prove a
useful addition to the growing list of ultrafast radical clocks
that can be used to probe the mechanisms of reactions that
potentially involve radical intermediates. As noted in our
earlier communication, this probe is also capable of clearly
distinguishing between radical and carbocation intermediates
because of their distinctive ring-opening regiochemistries.22
Moreover neither radical nor carbocation formation receives
significant kinetic preference as a result of the presence of
this probe. Therefore, the 2,2-difluorocyclopropylcarbinyl
system, now calibrated, should be an ideal mechanistic probe
of radical or carbocation intermediacy.
Acknowledgment. Support of this research in part by the
National Science Foundation is acknowledged with thanks.
Continuing discussions with Professor Merle A. Battiste have
provided immeasurable assistance.
Supporting Information Available: Full details regard-
ing the synthesis and characterization of products 3-5,
procedures for kinetic experiments, and tables of kinetic data.
This material is available free of charge via the Internet at
http://pubs.acs.org.
OL005578S
(21) This value has a 2σ error of 0.2 that does not take into account the
error in Beckwith’s data (not provided).
(22) Tian, F.; Battiste, M. A.; Dolbier, W. R., Jr. Org. Lett. 1999, 1,
193-196.
Org. Lett., Vol. 2, No. 6, 2000
837