Absolute Rate Constants for Reactions of Tributylstannyl Radicals with Bromoalkanes, Episulfides, and α-Halomethyl-Episulfides, -Cyclopropanes, and -Oxiranes: New Rate Expressions for Sulfur and Bromine Atom Abstraction

Article abstract of DOI:10.1021/jo035467h

Arrhenius rate expressions were determined for the abstraction of bromine atom from 2-phenethyl bromide by tri-n-butylstannyl radical (Bu ₃Sn^.) in benzene using transient absorption spectroscopy, (log(k_abs,Br/M^-1 s^-1) = (9.21 ± 0.20) - (2.23 ± 0.28)/θ, θ = 2.3RT kcal/mol, errors are 2σ) and for the abstraction of sulfur atom from propylene sulfide to form propylene, (log(k_s/M^-1 s^-1) = (8.75 ± 0.91) - (2.35 ±1.33)/θ). Rate constants for reactions of organic bromides, RBr, with Bu₃Sn^. were found to vary as R = benzyl (15.6) > thiiranylmethyl (6.2) > oxiranylmethyl (3.1) > cyclopropylmethyl (1.3) > 2-phenethyl (1.0), with k_abs,Br = 6.8 × 10 ⁷ M^-1 s^-1 at 353 K for 2-phenethyl bromide. Bromine abstraction from α-bromomethylthiirane is about 7-fold faster than sulfur atom abstraction and is comparable to the reactivity of a secondary alkyl bromide. The potential surface for the vinylthiomethyl → allylthiyl radical rearrangement at UB3LYP/6-31G(d) and UB3LYP/6-311+G-(2d,2p) levels of theory suggests that the thiiranylmethyl radical is produced about 9 kcal/mol above the allylthiyl radical on the rearrangement surface, consistent with the observed enhancement of the Br atom abstraction from the thiirane and with synchronous C-S bond scission of the thiirane ring. The selectivities reported in this work for S vs Cl and Br abstraction provide applications for radical-based synthesis and new competition basis rate expressions for trialkylstannyl radicals.

Full text of DOI:10.1021/jo035467h

Absolu te Ra te Con sta n ts for Rea ction s of Tr ibu tylsta n n yl Ra d ica ls
w ith Br om oa lk a n es, Ep isu lfid es, a n d r-Ha lom eth yl-Ep isu lfid es,
Cyclop r op a n es, a n d -Oxir a n es: New Ra te Exp r ession s for Su lfu r
-
a n d Br om in e Atom Abstr a ction
,
†
†
‡
†
†
J ames A. Franz,* Wendy J . Shaw, Claude N. Lamb, Tom Autrey, Douglas S. Kolwaite,
†
†
Donald M. Camaioni, and Mikhail S. Alnajjar
Pacific Northwest National Laboratory, 902 Battelle Boulevard, Richland, Washington 99352, and
Department of Chemistry, North Carolina A&T, 1601 East Market Street,
Greensboro, North Carolina 27411
james.franz@pnl.gov
Received October 6, 2003
Arrhenius rate expressions were determined for the abstraction of bromine atom from 2-phenethyl
•
bromide by tri-n-butylstannyl radical (Bu
3
Sn ) in benzene using transient absorption spectroscopy,
-
1
-1
(log(kabs,Br/M
s ) ) (9.21 ( 0.20) - (2.23 ( 0.28)/θ, θ ) 2.3RT kcal/mol, errors are 2σ) and for the
abstraction of sulfur atom from propylene sulfide to form propylene, (log(k
.91) - (2.35 (1.33)/θ). Rate constants for reactions of organic bromides, RBr, with Bu Sn were
found to vary as R ) benzyl (15.6) > thiiranylmethyl (6.2) > oxiranylmethyl (3.1) > cyclopropyl-
-
1
-1
s
/M
s ) ) (8.75 (
•
0
3
7
-1 -1
methyl (1.3) > 2-phenethyl (1.0), with kabs,Br ) 6.8 × 10 M
s
at 353 K for 2-phenethyl bromide.
Bromine abstraction from R-bromomethylthiirane is about 7-fold faster than sulfur atom abstraction
and is comparable to the reactivity of a secondary alkyl bromide. The potential surface for the
vinylthiomethyl f allylthiyl radical rearrangement at UB3LYP/6-31G(d) and UB3LYP/6-311+G-
(
2d,2p) levels of theory suggests that the thiiranylmethyl radical is produced about 9 kcal/mol above
the allylthiyl radical on the rearrangement surface, consistent with the observed enhancement of
the Br atom abstraction from the thiirane and with synchronous C-S bond scission of the thiirane
ring. The selectivities reported in this work for S vs Cl and Br abstraction provide applications for
radical-based synthesis and new competition basis rate expressions for trialkylstannyl radicals.
In tr od u ction
are lacking. The cis-2-aziridinyl-, oxiranyl-, and cyclo-
propylcarbinyl radicals are predicted to exhibit activation
The ring-opening reactions of the cyclopropylcarbinyl
radical and the nitrogen and oxygen analogues are the
subject of ongoing interest (eq 1, X ) NR, CH , O, S).
2
barriers (E , 298 K) of 4.0, 4.4, and 7.5 kcal/mol for C-N,
a
C-O, or C-C bond cleavage, respectively, and barriers
of 10.2, 12.2, and 7.5 kcal/mol for C-C cleavage of the
systems.1 Our interest in the thiiranylmethyl radical
system was inspired by semiempirical calculations show-
ing that the related phenyl-thiaspiro[2.5]octadienyl radi-
3
cal, 2, undergoes barrierless C-S scission (eq 2).
The focus of this work was to determine, particularly for
X ) S in eq 1, whether atom abstraction processes that
produce 1 may be significantly enhanced by the synchro-
nous, exothermic ring opening of 1 to form allylthiyl
radical. High accuracy theoretical studies of the cyclo-
2
propylcarbinyl radical (1, X ) CH ) and N- and O-
analogues have been carried out using the CBS-RAD
1
theoretical model, yielding predicted rates approaching
experimental kinetic accuracy. The cyclopropylcarbinyl
ring-opening reaction is very accurately known and
2
serves to calibrate the theoretical models. Comparably
accurate experimental rate expressions for the 2-aziri-
dinyl- and oxiranylmethyl radical ring-opening reactions
†
Pacific Northwest National Laboratory.
North Carolina A&T.
Work in this laboratory to characterize the vinylthio-
methyl (5) f allylthiyl (6) potential surface using ab
initio electronic structure calculations was underway
‡
(
1) Smith, D. M.; Nicolaides, A.; Golding, B. T.; Radom, L. J . Am.
Chem. Soc. 1998, 120, 10223-10233.
2) Newcomb, M. J . Org. Chem. 1999, 64, 1225-1231.
4
(
when the work of Pasto appeared. Pasto demonstrated
1
0.1021/jo035467h CCC: $27.50 © 2004 American Chemical Society
Published on Web 01/16/2004
1
020
J . Org. Chem. 2004, 69, 1020-1027

Rate Expressions for S and Br Atom Abstraction
that the thiiranylmethyl radical, 3, is not an intermediate
on the vinylthiomethyl f allylthiyl potential surface and
TABLE 1. Com p etitive Rea ction s of Bu 3Sn H w ith
Ep isu lfid es a n d Br om id es in Ben zen e a t 80 °C
4
thus should undergo barrierless C-S scission to form the
allylthiyl radical (eq 3). DFT calculations of the vinylthio-
methyl f allylthiyl radical rearrangement, presented
here at the B3LYP/6-31G(d) and B3LYP/6-311+G-
(
2d,2p) levels of theory, confirm the findings of Pasto.4
The computational results suggest that homolytic path-
ways producing the thiiranylmethyl radical, such as
halogen atom abstraction from R-halomethylthiiranes,
should be enhanced by synchronous C-S scission pro-
ducing allylthiyl radical (eq 4). Similar considerations
suggest that oxiranylmethyl radicals should not exhibit
enhanced rates of formation in atom transfer reactions.
Oxiranylalkyl radicals (1, X ) O) have been experimen-
tally demonstrated to undergo rapid and, thermochem-
istry allowing, reversible C-O and C-C cleavage, leading
to products reflecting partial equilibration of allyloxyl and
vinyloxyalkyl radicals depending on trapping condi-
tions.5
,6-11
Electronic structure calculations of Pasto1,4
suggest C-O cleavage of oxiranylmethyl radical will
provide the kinetically favored alkoxy radical product.
C-C scission products are expected only if the substitu-
ent R (eq 4) is a group that provides stabilization to the
Thus, this study sought to determine the rate constants
for bromine atom abstraction from R-bromomethylthi-
irane, -oxiranes, and -cyclopropanes, with the reactivity
of the thiirane structure expected to be enhanced relative
to either the oxirane or cyclopropane analogues for
considerations cited above. In the course of the study, it
was discovered that sulfur atom was selectively and rap-
5
,6,12,13
resulting radical,
leading to the thermodynamically
The potential well for oxiranyl
favored product.1
,4,7-11
radicals has been demonstrated to be shallow. Results
of Krosley and Gleicher suggest that the oxiranylmethyl
radical will undergo ring opening with a rate constant
idly abstracted from R-chloromethylthiiranes and other
1
0
-1
of >10
s
. Recently, Krishnamurthy et al. and Grossi
•
thiirane systems by tri-n-butylstannyl radical (Bu
3
Sn )
et al.1
4,15
demonstrated the intermediacy of the oxiranyl
(eq 5). Thus, we report relative and absolute rates for
radical by, respectively, trapping the radical with thiophe-
nol and recording its ESR spectrum at low temperature.
The magnitude of the potential well for the oxiranyl-
methyl radical has been addressed in calculations per-
1
3,6
formed by Smith et al. and Pasto and was found to
agree with experiments that suggest a minimal well.
Finally, it has been appreciated for some years that the
cyclopropylmethyl C-H bond exhibits enhanced reactiv-
ity in radical-forming reactions, as a result of the stabi-
abstraction of halogen atoms from a variety of R-halom-
ethylcyclopropanes, oxiranes, and thiiranes, and Arrhe-
nius data for abstraction of bromine atom from a primary
alkyl bromide and sulfur atom from propylene sulfide.
Supporting computational results characterizing the
thiiranylmethyl ring-opening surface are presented.
lization of the radical π-orbital with the HOMO of the
adjacent cyclopropyl ring.1
6,17
However, no experimental
study has directly compared the reactivity of cyclopro-
pylcarbinyl, oxiranylmethyl, and thiiranylmethyl systems.
(
3) Alnajjar, M. S.; Franz, J . A. J . Am. Chem. Soc. 1992, 114, 1052-
Resu lts a n d Discu ssion
1
058.
(
(
(
4) Pasto, D. J . J . Org. Chem. 1996, 61, 252-256.
Rela tive Ra tes of Br om in e a n d Su lfu r Atom
Abstr a ction by tr i-n -Bu tylsta n n yl Ra d ica l. Relative
rate constants were determined by reaction of competing
pairs of halide or episulfide and 2-phenethylbromide with
5) Nonhebel, D. C. Chem. Soc. Rev. 1993, 347.
6) Ziegler, F. E.; Petersen, A. K. J . Org. Chem. 1995, 60, 2666-
2
1
667.
(7) Breen, A. P.; Murphy, J . A. J . Chem. Soc., Chem. Commun. 1993,
91-192.
limiting concentrations of Bu
3
SnH at 80 °C in benzene-
8) Davies, A. G.; Tse, M. J . Organomet. Chem. 1978, 155, 25-30.
d
6
. Table 1 presents relative and absolute rate constants
Soc., Perkin Trans. 2 1976, 1044-1047.
10) Marples, B. A.; Rudderham, J . A.; Slawin, A. M.; Edwards, A.
J .; Hird, N. W. Tetrahedron Lett. 1997, 38, 3599-3602.
11) Barton, D. H.; Motherwell, R. S.; Motherwell, W. B. J . Chem.
for reaction of the tributylstannyl radical with halides
and episulfides. The reaction products generated from the
loss of a sulfur atom from propylene sulfide, styrene
sulfide epithiobromohydrin, and epithiochlorohydrin (en-
tries B, G, F and A, respectively) are propylene, styrene,
allyl bromide, and allyl chloride. 2-Phenethyl bromide,
cyclopropylcarbinyl bromide, epibromohydrin, epithio-
bromohydrin, and benzyl bromide (entries C, D, E, F, and
H) all lose bromine to give, respectively, ethyl benzene,
butene, allyl alcohol, allyl thiol, and toluene. The rate
(
(
Soc., Perkin Trans. 1 1981, 2363-2367.
(
12) Kuivila, H. G. Acc. Chem. Res. 1968, 1, 299.
(13) Krosley, K. W.; Gleicher, G. J .; Clapp, G. E. J . Org. Chem. 1992,
5
7, 840-844.
14) Krishnamurthy, V.; Rawal, V. J . Org. Chem. 1997, 62, 1572-
573.
15) Grossi, L.; Strazzari, S.; Gilbert, B. C.; Whitwood, A. C. J . Org.
Chem. 1998, 63, 8366-8372.
16) Roberts, C.; Walton, J . C. J . Chem. Soc. Chem. Commun. 1984,
109-1111.
17) Roberts, C.; Walton, J . C. J . Chem. Soc., Perkin Trans. 2 1985,
41-846.
(
1
(
(
1
8
7
-1 -1
constants range from 3.5 × 10 M
s
for abstraction
(
9
-1
of sulfur atom from propylene sulfide to 1.1 × 10 M
J . Org. Chem, Vol. 69, No. 4, 2004 1021

Franz et al.
s- for abstraction of bromine atom from benzyl bromide
1
the presence of allylic chlorides, in contrast to allylic
bromides where removal of the halide atom is dominant,
about 7-fold faster than sulfur atom abstraction.
at 353 K in benzene. The latter rate is approximately 16
times lower than the bimolecular diffusive encounter rate
for Bu
the R-bromomethylthiirane (F), bromomethyloxirane (E),
and other cyclopropylcarbinyl systems (entries B and D)
with Bu
•
3
Sn with 2-phenethyl bromide. The reactivity of
Absolu te Ra tes a n d Ar r h en iu s P a r a m eter s of
Br om in e Atom Abstr a ction fr om P h en eth yl Br o-
m id e by Tr i-n -bu tylsta n n yl Ra d ica l. Carlsson and
•
19
3
Sn relative to primary alkyl bromide invites
Ingold reported some of the first room-temperature
comparison with relative rates of reactivity of primary,
absolute rate constants for reaction of stannyl radicals
with halides. A decade later, Scaiano reported an
1
8
20
secondary, and tertiary alkyl bromides. Kuivila and
1
9
Carlsson and Ingold reported the relative rates of
ambient temperature rate constant for abstraction of
bromine atom from a primary alkyl bromide. However,
temperature-dependent rate constants for the halides
have not been reported. To place the relative rates of
Table 1 on an absolute basis and to determine temper-
ature-dependent data for abstraction of sulfur atom from
thiiranes, transient absorption spectroscopy was used to
observe the reaction of the tin radical with primary
bromide using the photolysis of di-tert-butyl peroxide in
the presence of tri-n-butylstannane and 2-phenethyl
bromide as depicted in eqs 6-12:
•
3
reaction of Bu Sn with primary alkyl, secondary alkyl,
and tertiary alkyl bromides to be 1:3:7 at 45 °C in
nonpolar solvent. Applying the experimental value of log-
-
1
-1
(
A/M
bromide along with the 318 K rate constant of 4.8 ×
0 M to the secondary and tertiary relative rates
s ) ) 9.21 found in this work for phenethyl
7
-1 -1
1
s
2
1
allows the published rates to be scaled to 353 K for
comparison to the results of Table 1. The scaled relative
rates at 353 K are thus predicted to be 1.0 (primary):2.7
(secondary):5.8 (tertiary). Examination of Table 1 reveals
the cyclopropylcarbinyl analogues to span the range of
primary, secondary, and tertiary alkyl rates, cyclopro-
pylcarbinyl being consistent with a primary bromide,
R-bromomethyloxirane similar to secondary bromide, and
R-bromomethylthiirane comparable to a tertiary alkyl
bromide.
hν
t
t
t
•
BuOO Bu
8 2 BuO
(6)
(7)
3
08 nm
k7
t
•
t
•
BuO + Bu SNH
98 BuOH + Bu3Sn
3
kBr
A narrow range of relative rates of the bromides is
expected as a result of high reactivity and the cor-
respondingly narrow range of relative enthalpy changes,
Bu SNH + PhCH CH Br
8
3
2
2
•
Bu SnBr + PhCH CH
2
(8)
3
2
∆
(∆H°) for bromine atom transfer to tin radical. Relative
k9
•
•
enthalpy changes reflect both ground-state effects on
alkyl bromides as well as product radical heats of
formation:
PhCH CH + Bu SnH 98 PhCH CH + Bu Sn (9)
2 2 3 2 2 3
kt
9
2
PhCH CH
8 non-radical products
(10)
2
2
∆
H1
o
1
•
•
1•
kct
9
•
•
R Br + Bu Sn
8 R + Bu SnBr
3
3
PhCH CH + Bu3Sn
8 non-radical products (11)
2
2
∆
H2
o
2
2•
kt′
9
•
R Br + Bu Sn
8 R + Bu SnBr
3
3
2Bu Sn
8 non-radical products
(12)
3
o
o
o
o
2
∆
∆H ) DH - ∆H1 ) -∆H (R Br) +
2 f
The time dependence of tin radical concentration
following the laser pulse is given by eq 13:
o
1
o
2•
o
1•
∆
H (R Br) + ∆H (R ) - ∆H (R )
f f f
Thus, the relative enthalpy change for abstraction of
•
d[Bu Sn ]
3
•
a secondary alkyl bromine atom versus abstraction of a
primary alkyl bromine atom is ∆∆H° ) -0.6 kcal/mol,
and for abstraction of a bromine atom from a primary
bromide versus abstraction of bromine atom from a
tertiary alkyl bromide, ∆∆H° ) -1.3 kcal/mol, resulting
in the narrow range of observed rates.
) k [BuO ][Bu SnH] -
7
3
dt
•
k [Bu Sn ][PhCH CH Br] +
Br
3
2
2
•
k [Bu SnH][PhCH CH ] -
9
3
2
2
•
•
• 2
k [PhCH CH ][Bu Sn ] - 2k ′[Bu3Sn ] (13)
ct
2
2
3
t
The pattern of selectivity in abstraction of sulfur atom
from thiiranes is consistent with the cleavage of the C-S
bond in the rate-determining step of S atom abstraction:
-d[Bu
dt
Sn•]
3
•
≈
k [Bu Sn ][PhCH CH Br] (14)
Br 3 2 2
ln([BuSn•]
t0
[
BuSn•]
t1
)
kobs ) k [PhCH CH Br] + C
(15)
Br
2
2
∆t
(
18) Kuivila, H. G. M., L. W.; Warner, C. R. J . Am. Chem. Soc. 1962,
The trend in relative rates, chloromethylthiirane (A) <
methylthiirane (B) < phenylthiirane (C), corresponds to
the stability of the forming radical center. An important
feature of Table 1 is that sulfur atoms can be removed
from the thiiranyl ring with near-complete selectively in
8
4, 4, 3584-3586.
(19) Carlsson, D. J .; Ingold, K. U. J . Am. Chem. Soc. 1968, 90, 7047-
7
055.
(
20) Scaiano, J . C. J . Am. Chem. Soc. 1980, 102, 7747-7753.
(21) Macey, R. I. a. O., G. F.; Kagi Shareware, 1442-A Walnut Street
#392-GO, Berkeley, CA 94709-1405; 2002.
1
022 J . Org. Chem., Vol. 69, No. 4, 2004

Rate Expressions for S and Br Atom Abstraction
F IGURE 2. Plots of kobs vs [PhCH
order analysis of transient absorption data between 1.5 and
.5 µs for abstraction of Br atom from PhCH CH Br at
temperatures indicated.
2 2
CH Br], M. Pseudo-first-
2
2
2
TABLE 2. Va lu es of kobs (s-1) (eq 15) of Abstr a ction of
Br Atom fr om P h en eth yl Br om id e by Tr i-n -bu tylsta n n yl
Ra d ica l (Con cen tr a tion s a n d Tem p er a tu r es In d ica ted )
T ((1)
(
K)
0.0050 M 0.0102 M 0.0146 M 0.0198 M 0.0249 Ma
5
5
5
5b
5
2
2
2
3
3
3
3
84
89
99
10
2.32 × 10 4.13 × 10 5.35 × 10
3.39 × 10 5.45 × 10 6.54 × 10
2.02 × 10 4.26 × 10 5.51 × 10
4.36 × 10 5.68 × 10 8.41 × 10
2.90 × 10 5.02 × 10 8.16 × 10
7.22 × 10
8.43 × 10
7.69 × 10
5
5
5
5
5
5
5
5
5
5
5
5
4
c
5
5
5c
5
18 1.03 × 10
25 2.30 × 10
33 4.17 × 10
5
5
4.09 × 10 6.71 × 10 9.27 × 10
5
5
5c
3.06 × 10 5.56 × 10 8.64 × 10
a
b
c
Concentrations of PhCH2CH2Br. Values of kobs. Average of
two data points. All other values based on average of five data
points.
the conditions utilized, termination reactions k
′ occur to negligible extent during the first 8 µs, and
the choices of k , kct, and k ′ have little effect on the
optimized values of k and k . Kinetic fitting is conve-
t
, kct, and
•
3
F IGURE 1. (a) Transient absorbance profiles of Bu Sn (0.02
k
t
M Bu
trace represents 300 transients at a laser repetition rate of 1
Hz, 2 mL/min flow, resolution of 1 ns/pt. Dividing by ꢀ400
3 2 2
SnH) at indicated PhCH CH Br concentrations. Each
t
t
7
9
)
-
1
-1
22
niently carried out using the Berkeley Madonna pro-
1
620 M cm yields data in concentration units. (b) Nu-
•
21
merical fit of decay of Bu
3
•
Sn . ([PhCH
0
2
CH
2
Br] ) 0.0198 M,
gram. The kinetic fit reveals that the last three terms
t
-5
[Bu
3
SnH] ) 0.02 M, [ BuO ]
) 1.15 × 10 M. Rate constants
of eq 13 are small compared to the halogen abstraction
term, prior to 2.5 µs, providing that suitably low concen-
trations of alkyl bromide and tin hydride are employed.
High concentrations of bromide and tin hydride lead to
the persistence of the tin radical due to the radical chain
pathway of eqs 8 and 9. The numerical integration of eq
13 (Figure 1b) confirms that a simple logarithmic treat-
ment of the data (Figure 1c,d) between 1.0 and 2.5 µs
k
of k
1
7
and k
8
were optimized to fit the observed data, with values
9
-1 -1
9
-1 -1
t
′ ) 1.8 × 10 M
s
, kct ) 2.0 × 10 M
s
, k
t
) 2.2 ×
9
-1 -1
6
-1 -1
0 M
s
and k
9
) 2.3 × 10 M
s , yielding the optimized
8
-1 -1
values of constants k
7
) 2.0 × 10 M
s
and k
8
) kBr) 3.0
7
-1
-1
•
×
10
M
s
for abstraction of Br atom by Bu
3
Sn . (c)
Treatment of data between 1.5 and 2.5 µs from Figure 1b. The
slope of this curve provides kobs ) kBr[PhCH CH
2
2
Br] ) 5.6 ×
5
-1
5
7
-1 -1
1
0 s , or kBr ) 5.6 × 10 /0.0198 ) 2.8 × 10 M
s .
(eq 15) yields suitably accurate values of kBr. Thus, simple
Under conditions of the low laser pulse intensity and low
halide and hydride concentrations used in this study,
numerical integration of the differential equations de-
logarithmic treatment of the concentration (or equiva-
•
lently, optical density) of the Bu
was carried out to yield straight lines with the slope kobs
) kBr[PhCH CH Br]. Plots of kobs against [PhCH CH Br]
3
Sn decay versus time
•
•
scribing the time dependence of tBuO , PhCH
2
CH
2
, and
2
2
2
2
•
Bu
3
Sn radical concentrations from eqs 6-12 reveals that
provide values of kBr at each temperature (Figure 2). The
resulting data is tabulated in Table 2, and temperature-
dependent rate data are plotted in Figure 3. As shown
below, the fit of data using data between 1.5 and 2.5 µs
agrees with detailed numerical integration to within
3-4% and requires no preassignment of rate constants.
Ar r h en iu s Da ta for Abstr a ction of Su lfu r Atom
by Tr i-n -bu tylsta n n yl Ra d ica l fr om P r op ylen e Su l-
fid e. Relative rates of abstraction of sulfur atom from
propylene sulfide vs abstraction for bromine atom from
tert-butoxyl radical is fully consumed by 1 µs (see, e.g.,
Figure 1b), simplifying the kinetic analysis by allowing
the first term of eq 13 to be neglected. Formation of the
stannyl radical by alkyl radical abstraction (k ) is slow
9
on the time scale of abstraction of Br atom by the tin
radical, and the time dependence of the evolution of the
•
Bu
3
Sn is due nearly exclusively to bromine atom ab-
straction, with negligible contributions from radical cross-
and self-termination reactions (eqs 11 and 12). Kinetic
•
fit of the observed decay of Bu
3
Sn is carried out by
6
2-phenylethyl bromide in benzene-d by the stannyl
numerical integration of eq 3, based on eqs 7-12, with
optimization of rate constant kBr (eq 8) while holding
radical were carried out over a 60 °C temperature range
using GC and NMR product quantitation. Samples were
sealed in Pyrex tubes following three freeze-pump-thaw
7 9 t t
values of k , k , k , kct, and k
′ of eq 13 constant.21 Under
J . Org. Chem, Vol. 69, No. 4, 2004 1023

Franz et al.
kinetic experiments, allylthiotributylstannane was not
formed. The absence of this product at lower reagent
concentrations suggests that it does not result from
•
attack of Bu
3
Sn at the sulfur atom of R-bromomethylthi-
irane followed by Br loss, but rather that allylthiotribu-
•
tylstannane results from the cross-termination of Bu
3
Sn
S radicals. Control experiments re-
vealed that concentrations of propylene sulfide as low as
•
2 2
and CH dCHCH
-
4
3
× 10 M would have been detectable, giving a lower
limit of [allylthiol]/[propylene sulfide] > 3000. At an
average [Bu
3
SnH] ) 0.5 M and a rate constant of 2.4 ×
6
-1 -1
1
0 M
s
for abstraction of hydrogen atom by a primary
F IGURE 3. Temperature-dependent rate data for the ab-
straction of bromine atom from phenethyl bromide by the
tributylstannyl radical. The experimental data result in an
23
radical, the lower limit for the rate constant for opening
6
-1
of the thiiranylmethyl radical is kre g 3000(2.4 × 10 M
-
1
9
-1
s
)(0.5) g 3 × 10 s
.
-
1
-1
activation energy of 2.23 ( 0.28 kcal/mol and log(A/M
s )
2
Ra tes of Br om in e Atom Abstr a ction fr om Br o-
of 9.21 ( 0.20 (errors are 2σ, R ) 0.98), 11-60 °C, kBr (298
7
-1 -1
K) ) 3.8 ( 0.5 × 10 M
s
. The temperature range is defined
m om eth yloxir a n es a n d -Th iir a n es by Tr i-n -bu tyl-
sta n n yl Ra d ica l: En h a n cem en t of Br om in e Atom
Abstr a ction by th e Cyclop r op ylca r bin yl, Oxir a n yl,
a n d Th iir a n e Rin g. The measured rate constant of
by the melting point of benzene and the thermal stability of
the peroxide/tin hydride/bromide mixture.
bromine atom abstraction from PhCH
temperature, kBr ) 3.8 ( 0.5 × 10 M
2
CH
2
Br (C) at room
7
-1 -1
s , is comparable
to the rate of abstraction from n-propyl bromide previ-
ously reported by Ingold and co-workers.24 Bromine atom
abstraction from R-bromomethylcyclopropane (D) is only
slightly faster than from simple primary alkyl bromide.
The modest effect of the cyclopropyl group on the reactiv-
ity of primary alkyl bromide is consistent with the small
stabilization of a primary alkyl radical center by the
cyclopropyl group, estimated in early theoretical work to
2
5
be 1-1.8 kcal/mol compared to a primary alkane. The
relative rate constant for abstraction of bromine atom
from R-bromomethylthiirane (entry F of Table 1) of 6.2
reflects a significant rate enhancement of bromine atom
abstraction compared to R-bromomethylcyclopropane
(entry D of Table 1) of 1.3 and 2-phenylethyl bromide
(entry C of Table 1) of 1.0. We interpret this rate
enhancement to reflect synchronous opening of the thi-
irane ring with bromine atom abstraction. While rate
F IGURE 4. Arrhenius data for abstraction of sulfur atom
from propylene sulfide by tri-n-butylstannyl radical.
cycles. Solutions of 2-phenethyl bromide, propylene sul-
fide, azo-2-isobutyronitrile (AIBN), and internal GC/NMR
standard (p-xylene) in benzene-d
6
were heated in a
thermostated oil bath. Ethylbenzene was measured by
GC relative to internal p-xylene, and propylene sul-
fide was measured by comparing the vinyl hydrogen
constants for atom and group displacement reactions by
1
•
atoms of propylene to internal p-xylene standard by H
Bu
3
Sn reflect primarily enthalpic effects, polar effects
NMR integration to yield the relative rate expression:
and electron-transfer pathways, and perhaps formation
of transient adducts involving the stannyl radical and
the bromine atom in the displacement reactions may
contribute to the description of the transition state
properties of Br abstraction.13 The enthalpy changes
related to the observed rate enhancements can be esti-
mated from the thermochemistry of the vinylthiomethyl
f thiiranylmethyl f allylthiyl system (eq 16 and Figure
log(kBr/k
S
) ) (0.46 ( 0.90) - (0.115 ( 1.3)/θ, θ ) 2.3RT
kcal/mol. Combining the relative rate expression kBr/k
S
with the absolute rate expression for bromide abstraction
from 2-phenethyl bromide, kBr, yields an Arrhenius rate
expression for abstraction of sulfur atom from propylene
-
1
-1
sulfide: log(k
S
/M
s
) ) (8.75 ( 0.92) - (2.35 ( 1.32)/
7
-1 -1
θ, θ ) 2.3RT kcal/mol, k
The results are depicted in Figure 4.
S
(298 K) ) 1.1 × 10 M
s .
Rea ction of Ep ith iobr om oh yd r in w ith 1 M Tr i-
n -bu t ylst a n n a n e. At 80 °C in benzene-d , R-bromo-
methylthiirane was converted to allylthiol, and allyl
bromide with concentrations of Bu SnH and bromide of
ca. 0.05 M, with no observed autoinitiation of the reduc-
tion reaction as demonstrated by monitoring the stock
6
3
5
). The heat of formation of allylthiol, 15 kcal/mol, leads
to an estimate of the heat of formation of vinylthiomethyl
radical of 51 kcal/mol by assigning an S-H bond strength
of 88 kcal/mol. The heat of formation of thiiranylmethyl
solution by NMR. When 1 M Bu
momethylthiirane were combined in benzene-d
3
SnH and 1 M R-bro-
, the
6
bromide was consumed (93% conversion, 7% residual
starting bromide) at room temperature in less than 30
min forming allylthiol (78%) and allyl bromide and
propylene (13%) and allylthiotributylstannane (9%), with
no propylene sulfide detected. At lower concentrations
(
22) Autrey, T.; Shaw, W. J .; Franz, J . A. Unpublished work.
(23) Chatgilialoglu, C.; Ingold, K. U.; Scaiano, J . C. J . Am. Chem.
Soc. 1981, 103, 7739-7742.
24) Ingold, K. U.; Lusztyk, J .; Scaiano, J . C. J . Am. Chem. Soc.
984, 106, 343-348.
25) Radom, L.; Paviot, J .; Pople, J . A.; Schleyer, P. V. R. J . Chem.
Soc. Chem. Comm 1974, 58-60.
(
1
(
of Bu
3
SnH and bromides (e.g., 0.05 M) employed in
1
024 J . Org. Chem., Vol. 69, No. 4, 2004

Rate Expressions for S and Br Atom Abstraction
TABLE 3. Electr on ic En er gies a n d Tota l Th er m a l
Cor r ection s to En th a lp y, 298 K, for Sp ecies on th e
Vin ylth iom eth yl f Allylth iyl Su r fa cea
electronic energy, hartree/particle
UB3LYP/
6-311+G(2d,2p)
UB3LYP/
6-31G(d)
enthalpy
correction
species
allylthiyl
TS1
TS2
-515.52197
-515.47235
-515.49605
-515.50488
-515.45165
-515.40032
-515.42747
-515.42957
0.074684
0.070495
0.072178
0.070723
vinylthiomethyl
a
Structures optimized at the UB3LYP/6-311+G(2d,2p) and
UB3LYP/6-31G(d) levels of theory. Frequencies calculated at
UB3LYP/6-311+G(2d,2p).
a) and by 0.9989 for the vibrational term (x ) b).
Vibrations identified as internal torsions of less than 250
-1
28
cm were replaced by 1/2RT. Geometries of the species
f
F IGURE 5. Values of ∆H ° (in parentheses) of allylthiyl,
are included in Supporting Information. Electronic struc-
vinylthiomethyl, and thiiranylmethyl radicals are thermo-
chemical estimates in kcal/mol based on values of the parent
hydrocarbons allylthiol (15 ( 2), propylene sulfide (11 ( 0.5)
and methylvinylthioether (18 ( 0.2) (Lias, S. G.; Bartmess, J .
E.; Liebman, J . F.; Holmes, J . L.; Levin, R. D.; Mallard, W. G.
Gas-Pase Ion and Neutral Thermochemistry. J . Phys. Chem.
Ref. Data 1988, 17, Suppl. 1 using BDE values for S-H (88
2
kcal/mol), C-H (100 kcal/mol) and SCH -H (93.4 kcal/mol),
respectively. The solid line is the potential surface for the
rearrangement at B3LYP/6-31G(d), 0 K. TS1 corresponds to
C2-C4 bond elongation. TS2 corresponds to torsional motion
ture calculations at the B3LYP/6-31G(d) and B3LYP/6-
3
11+G(2d,2p)/B3LYP/6-31G(d) levels of theory illustrate
the barrierless C-S bond scission to form the allylthiyl
radical. Figure 5 (connected line) shows the B3LYP/
6-31G(d) potential surface for the vinylthiomethyl to
allylthiyl radical rearrangement. The potential surface
confirms the report of Pasto for the thiiranylmethyl
radical in which full geometry optimization calculations
resulted in the spontaneous ring opening to the allylthiyl
2 2
of the CH as the plane of the CH group is aligned perpen-
dicularly to the C2-S3 bond for scission. Note that TS2 is not
a saddle point on the ∠ C2-S3-C4 surface since TS2 occurs
2
in the CH torsional space.
radical is estimated to be 60 kcal/mol, arrived at by
assigning the C-H bond dissociation energy of 101 kcal/
mol to the methyl C-H bond of propylene sulfide. This
model predicts that synchronous ring opening during
bromine atom abstraction would be 8-10 kcal/mol more
exothermic than Br abstraction from 2-phenethylbro-
mide, consistent with the experimental observations.
Electr on ic Str u ctu r e Ca lcu la tion s. Geometries and
frequencies for the allylthiyl f vinylthiomethyl surface
were optimized at UB3LYP/6-31G(d) and UB3LYP/
4
radical. The inflection point observed for the thiiranyl-
methyl radical (indicated by the arrow, Figure 5) is
consistent with the CdC bond forming concurrently with
C-S bond cleavage and is strong evidence that thiira-
nylalkyl radical (unlike cyclopropylcarbinyl and oxira-
nylmethyl analogue) does not exist as a stable interme-
1,4
diate. These results are consistent with our previous
observation, using semiempirical calculations, that thi-
aspirooctadienyl radicals such as 2 are not intermediate
structures. In addition, Figure 5 shows the enthalpies
6
-311+G(2d,2p) levels of theory using the Gaussian 98
26
set of programs. The connected line of Figure 5 depicts
UB3LYP/6-31G(d) energies at 0 K. Frequency calcula-
tions revealed a single negative eigenvalue for the
transition structures shown in Figure 5, and all positive
eigenvalues for the two ground state molecules. Enthalpy
corrections to ambient temperature to UB3LYP/6-311+G-
(
inside parentheses) of allylthiyl and vinylthiomethyl
radicals and the transition states. The UB3LYP/6-31G-
d) calculations predict that the thiiranylmethyl (TS2)
(
radical lies 13.6 kcal/mol above the allylthiyl radical. The
UB3LYP/6-311+G(2d,2p) calculations predict TS2 to lie
even higher, 14.7 kcal/mol above allylthiyl radical. These
values are rather substantially higher than the value
estimated from thermochemical kinetics, 8 ( 2 kcal/mol.
Thus, although the qualitative shape of the potential
surface predicted by the DFT approach appears correct,
the rather large deviation of the DFT energies for the
transitional structures is a matter for concern.
(
2d,2p) energies were made using scaled UB3LYP/
2
7
6
-311+G(2d,2p) frequencies following eq 17. Each vi-
, was scaled by 0.9806 for the ZPE term (x )
bration, λ
k
(26) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J . R.; Zakrzewski, V. G.; Montgomery, J . A.,
J r.; Stratmann, R. E.; Burant, J . C.; Dapprich, S.; Millam, J . M.;
Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J .;
Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo,
C.; Clifford, S.; Ochterski, J .; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J . B.; Cioslowski, J .; Ortiz, J . V.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.;
Fox, D. J .; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; J ohnson, B. G.; Chen,
W.; Wong, M. W.; Andres, J . L.; Head-Gordon, M.; Replogle, E. S.;
Pople, J . A. Gaussian 98; Gaussian, Inc.: Pittsburgh, PA, 1998.
Finally, we note that experiments were carried out to
establish a lower limit of the rate ring opening of
thiiranylmethyl radical. In experiments utilizing 1 M
1
Bu
3
SnH and R-bromomethylthiirane, H NMR measure-
ments detected no propylene sulfide product. This obser-
vation provided a lower limit of the rate constant for
(
27) McQuarrie, D. A. Statistical Mechanics; Harper and Row: New
York, 1976.
(28) Scott, A. P.; Radom, L. J . Phys. Chem. A 1996, 100, 16502.
J . Org. Chem, Vol. 69, No. 4, 2004 1025

Franz et al.
9
opening of the thiiranylmethyl radical of kre > 3 × 10
, consistent with barrierless ring opening.
Ab st r a ct ion of Br om in e At om fr om r-Br om o-
m eth yloxir a n e. The ring opening of the oxiranylmethyl
radical (10) was found to have a small rate increase as
compared to cyclopropylcarbinyl radical. Rearrangement
radical is as follows: R-bromomethylcyclopropane (0.85
-
1
8
8
s
× 10 ), R-bromomethyloxirane (2.1 × 10 ), and R-bro-
8
momethylthiirane (4.2 × 10 ). The increase in rate of
bromine atom abstraction by tin radical from the thi-
iranylmethyl bromide provides evidence for synchronous
thiirane ring opening providing an enthalpic enhance-
ment of the bromine atom abstraction rate. Electronic
structure calculations and thermochemical kinetic esti-
mates show that the thiiranylmethyl radical is produced
about 9 kcal/mol above the allylthiyl radical, leading to
reactivity of the thiiranyl bromide comparable to second-
ary alkyl bromides. When halogen atom abstraction is
slow, as with R-chloromethylthiirane, sulfur atom ab-
straction by tin radical becomes exclusive. The oxirany-
lalkyl counterparts show only a small rate increase in
halogen atom abstraction compared to that of the cyclo-
propyl analogue.
of oxiranylmethyl radical yielded only allyl alcohol from
allyloxy radical (11), consistent with other literature
reports indicating that the C-O bond cleavage is the
kinetically favored product.1
,4-11
In addition, oxiranyl-
1
methyl radical was found by reliable calculations to have
very little enthalpic gain in forming the allyloxy radical.
The increase in rate for the ring opening is thought to
result from the slight enthalpic gain or possibly from
polar effects in the transition state, as discussed above.
This interpretation, along with our results, is consistent
with the experimental evidence1 that suggests that the
oxiranylmethyl radical is a very short-lived intermediate
in the interconversion of vinyloxymethyl radical to allyl-
oxy radical rather than a synchronous ring opening.
Desu lfu r iza tion of Ep isu lfid es w ith Tr i-n -bu tyl-
sta n n a n e: Syn th etic Utility. In contrast to the highly
reactive bromine atom in R-bromomethylthiirane, the
reactivity of the chlorine atom in R-chloromethylthiirane
toward tri-n-butylstannyl radical is much slower than
that of the sulfur atom in the thiirane ring. The selective
abstraction of sulfur atom from the thiirane ring allows
the formation of allyl chloride as the only observed
product. Sulfur atom was also abstracted very efficiently
from both styrene sulfide and propylene sulfide. Episul-
fides are ordinarily converted to the corresponding olefins
Exp er im en ta l Section
P r ep a r a tion of Rea gen ts. Except as noted, all reagents
were commercially available. R-Bromomethylthiirane, R-chlo-
romethylthiirane, and epithiostyrene were prepared using
4,15
33
literature procedures as follows. The corresponding epoxides
and a slight molar equivalent excess of triphenylphosphine
sulfide were combined in methylene chloride and chilled in a
dry ice bath. Dry trifluoroacetic acid (2 equiv) was added
dropwise. The suspension was allowed to warm to room
temperature, stirred for several hours, and neutralized by
addition of solid sodium bicarbonate. The mixture was filtered
and concentrated by rotary evaporator. R-Bromomethylthi-
irane and R-chloromethylthiirane were vacuum distilled at
3
5-36 °C/∼0.2 Torr. Epithiostyrene was isolated by column
chromatography on a silica column and further purified using
rotating-plate chromatography on silica. R-Bromomethylthi-
1
irane: H NMR (300 MHz, benzene-d
9
6
): 4.26 (1H, ddd, J )
.9, 4.2, and 1.4 Hz), 3.74 (1H, m, J ) 9.9, 5.1, 4.2 Hz), 3.55
(
(
1H, t, J ) 9.85 Hz), 3.02 (1H, dt, J ) 5.9 and 1.4 Hz), 2.69
1H, dd, J ) 5.1 and 1.5 Hz). 1 C NMR (75 MHz, benzene-d
3
):
6
3
6.3, 33.5, 26.4 R-Chloromethylthiirane: 1H NMR (300 MHz,
by reaction with triphenyl phosphine or triethyl phos-
benzene-d ): 3.2 (1H, ddd, J ) 6.3, 4.4, and 1.3 Hz), 2.65 (1H,t,
6
phite.2
9-31
This work demonstrates that episulfides can
SnH and
a suitable radical initiator. The high rates of reaction of
J ) 9.6 Hz), 2.57 (1H,m), 1.87 (1H, dt, J ) 5.8 and 1.3 Hz),
1
3
also be readily desulfurized by reaction with Bu
3
1.58 (1H, dd, J ) 4.9 and 1.4 Hz). C NMR (75 MHz, benzene-
1
6
d ): 48.6, 33.5, 25.2 Epithiostyrene: H NMR (300 MHz,
•
7
8
-1 -1
benzene-d ): 7.09-6.92 (5H, m), 3.40 (1H,t, J ) 6.0 Hz), 2.28
Bu
3
Sn with episulfides (10 -10 M
s
), 5 orders of
6
3
2
(1H, dd, J ) 6.6 and 1.4 Hz), 2.15 (1H, dd, J ) 5.5 and 1.2
Hz).
magnitude more reactive than simple dialkyl sulfides,
guarantees that the episulfide group can be converted
to the olefin in the presence of a variety of reactive
functional groups. Thus, reaction of R-chloromethylthi-
Allylth iyl-tr i-(n -bu tyl)sta n n a n e. To a solution of 2.35 g
(102.2 mmol) of potassium metal in 100 mL of methanol under
nitrogen was added dropwise a solution of 7.36 g (99.2 mmol)
of allylmercaptan in 50 mL of methanol. After 30 min of
3
irane (0.05 M) with a slight excess of Bu SnH and
initiator (AIBN) in benzene at 80 °C leads to the clean,
quantitative formation of allyl chloride.
stirring a solution of 32.55 g (100 mmol) of Bu
of dry THF was added. After 1 h of stirring, 200 mL of water
was added and the mixture was extracted with CH Cl , dried
over MgSO , concentrated to a yellow oil, and distilled (0.5
mmHg, 108-110 °C) to give pure allylthiyl-tri-(n-butyl)-
3
SnCl in 50 mL
2
2
4
Con clu sion s
1
stannane (20 g, 56%). H NMR (benzene-d
6
): 0.9 (t, 9H), 1.05-
.12 (m, 6H), 1.26-1.38 (m, 6H), 1.53-1.65 (m, 6H), 3.20-
.23 (m,2H), 4.85-4.90 (m, 1H), 5.05-5.12 (m, 1H), 5.9-6.02
New Arrhenius rate expressions have been determined
for the abstraction of bromine atom from phenethyl
bromide and sulfur atom from propylene sulfide by
tributylstannyl radical, along with relative and absolute
rates of sulfur and bromine atom abstraction from
episulfide, oxirane, and cyclopropylalkane systems. Rate
enhancement of bromine atom abstraction from cyclo-
propylcarbinyl, oxiranylmethyl, and by tri-n-butylstannyl
1
3
2
119
117
(
m, 2H). Sn-H couplings were: J ( Sn and Sn-CH) ∼50 Hz,
3
119 117 3 119 117
J ( Sn and SnCCH) ∼40 Hz, and J ( Sn and SnCCCH)
∼31 Hz. 13C NMR(benzene-d ): 14.08, 14.38, 27.92, 29.54,
6
30.53, 114.86, 140.24. Sn-C couplings observed were as fol-
1
119
117
2
119
lows: J ( Sn and
Sn-C) 315, 330 Hz, J ( Sn and
1
17
2
119
117
4
Sn-C-C) 59, 61.8 Hz, J ( Sn-S-C and
Sn-S-C) 5 Hz,
3J ( Sn-C-C-C and
Hz. Anal. Calcd for C15
2.68. Found: C, 49.85; H, 9.12; S, 9.12; Sn, 32.54.
119
117
SnC-C-C) 21 Hz, and J (SnCCCC) ∼5
H
32SnS: C,49.61; H, 8.88; S, 8.83; Sn,
(
(
(
(
29) Davis, R. E. J . Org. Chem. 1958, 23, 1767.
3
30) Schuetz, R. D.; J acobs, R. L. J . Org. Chem. 1958, 23, 1799.
31) Nereiter, N.; Bordwell, F. G. J . Am. Chem. Soc. 1959, 81, 578.
32) Beckwith, A. L. J .; Pigou, P. E. Aust. J . Chem. 1986, 39, 77-
(33) Chan, T. H.; Finkenbine, J . R. J . Am. Chem. Soc. 1972, 94,
2880-2882.
8
7.
1
026 J . Org. Chem., Vol. 69, No. 4, 2004

Rate Expressions for S and Br Atom Abstraction
Tr a n sien t Absor ba n ce Sp ectr oscop y (TAS). TAS meth-
ods were used to measure the absolute rate of halide abstrac-
tion by tri-n-butylstannyl radical. The laser flash photolysis
discharge, a range where the tert-butoxy radicals are en-
tirely consumed and stannyl radical self-termination is neg-
ligible. At temperatures above 300 K, the growth of alkyl
radical also contributed to the pure first-order decay, and thus
2
0
system has been described previously. The attenuated output
5 mJ ) of an excimer laser (Lambda Physik Compex-100, XeCl,
08 nm) was directed onto a fluorescent sample cell perpen-
(
3
in this temperature range, 2k
to 1.9 µs.
t
/ꢀ represents data fit from 1.5
dicular to the direction of probe beam (150-W Xe lamp, model
XMN-150, Optical Radiation Corp., enclosed within a Spectral
Energy lamp housing, LH 150, Universal arc lamp power
supply, LPS 251). The filtered output of the probe lamp was
focused (L1, 250-mm cylindrical lens) at the center of the
fluorescence cell. Luminescence from the solvent was mini-
mized but not completely eliminated by placing an iris (2 mm)
in the path of the probe beam placed between the flow cell
and the monochromator (MC, Oriel model 77250). A second
lens (L2, 50 mm) was placed before the entrance slit to the
MC. In the kinetic experiments the MC was set to 400 nm to
detect the absorbance maximum of the visible tributylstannyl
radical absorption. A PMT tube (RCA 6199 wired in a 4-dynode
chain, powered by a negative high voltage power supply,
Pacific Photometric Instruments, model 226) was fixed in PMT
housing attached to the exit slit of the MC. The output of the
PMT tube was digitized on a Lecroy 534 transient digitizer at
a resolution of 1 ns/pt and the data transferred to a PC for
analysis.
Deter m in a tion of Rela tive Ra tes. Stock solutions were
prepared by combining Bu
3
SnH (0.02-0.07 M), azobisisobu-
3
-
tyronitrile (AIBN, (3-5) × 10 M), xylene (internal GC and
NMR standard), 2-phenylethyl bromide (0.05-0.5 M), and
competing reagent (0.05-0.1 M) in argon-purged benzene-d
6
under argon. Samples (500-600 µL) of the solutions were
transferred to argon-purged Pyrex tubes and sealed under
vacuum after three freeze-thaw degassing cycles. The samples
were heated in a Braun Thermomix 1480 oil bath for 10-12
min at 80 °C. Product and reactant concentrations were
1
measured before and after reaction by careful H NMR
integration and/or capillary GC measurements of products and
starting materials using calibrated internal standards. The
relative rates for competition of initial concentrations of
reagents A
0 0 t t
and B resulting in final concentrations A and B
is given by34
kA
ln At - ln A0
)
kB ln B - ln B0
t
Stock solutions of the tri-n-butyltin hydride (0.02 M) and
di-tert-butyl peroxide (0.2 M) were prepared in benzene
containing phenethyl bromide (0, 5, 10.2, 19.8, and 24.9 mM)
and had an absorbance at 308 of 0.30 across 1 cm. The
solutions were deaerated by purging with argon and charged
into a gastight syringe prior to kinetic analysis. The sample
was flowed through a quartz fluorescent cell (NGS T59FL)
clamped within a heated (or cooled) C-shaped anodized
aluminum cell block. Measurement of the temperature of the
sample, monitored at the exit port of the flow cell, was
facilitated with the use of a Digisense temperature controller
and thermocouple. Measurements were made at 284, 289, 299,
In practice, the products from A and B are measured. The
•
relative rate constant k
A
/k
B
for reaction of the Bu
3
Sn with
reagent A forming product P
A
vs reaction with B forming
product P
B
is given by the familiar relation
PA
ln 1 -
(
)
)
kA
kB
A0
PB
B0
)
ln 1 -
(
3
10, 318, 325, and 333 ( 1 K.
Relative rates are presented in Table 2.
The samples were flowed (Sage syringe pump, model 355)
through a low-volume quartz fluorescent flow cell (NGS
T59FL) to avoid the accumulation of photoproducts and to
ensure constant reagent concentrations. Three control experi-
ments demonstrated that the solutions were thermally and
photochemically stable under the reaction conditions with flow
rates of 2 mL/min and a laser repetition rate of 1 Hz: (1)
Photolysis of the flowing solution was shown to be negligible
after the UV output from the probe beam (150-W Xe lamp)
Ack n ow led gm en t. This work was supported by the
Office of Science, Office of Basic Energy Sciences,
Division of Chemical Sciences, U. S. Department of
Energy at the Pacific Northwest National Laboratory
(PNNL). PNNL is operated by Battelle Memorial Insti-
tute under contract DE-ACO6-76RLO 1830.
was suitably attenuated with a long pass glass filter (LP-365).
Su p p or tin g In for m a tion Ava ila ble: Geometries and
analytical frequencies from density functional calculations.
This material is available free of charge via the Internet at
http://pubs.acs.org.
1
(2) A measure of the reaction products by H NMR before and
after photolysis (flow rate of 1.0 mL and 0.5 Hz repetition rate)
assured that the bromide concentration can be assumed
1
constant under the reaction conditions. (3) H NMR experi-
J O035467H
ments demonstrated the stability of bromide/tin hydride
solutions at 60 °C. 2k
t
/ꢀ was determined by linearizing the data
(1/∆OD) and fitting the data from 1.5 to 2.5 µs from the laser
(34) Walling, C.; J acknow, B. B. J . Am. Chem. Soc. 1960, 82, 6108.
J . Org. Chem, Vol. 69, No. 4, 2004 1027