Spectroscopic studies of tributylstannyl radical. Rates of formation, termination, and abstraction determined by transient absorption spectroscopy

Article abstract of DOI:10.1021/om049933k

Transient absorption spectroscopy (TAS) was used to measure the rate of formation and the rate of self-termination of the main group metal hydride, tri-n-butyltin hydride (Bu₃-SnH). Irradiation of di-tert-butyl peroxide in the presence of Bu

Full text of DOI:10.1021/om049933k

2
080
Organometallics 2004, 23, 2080-2086
Sp ectr oscop ic Stu d ies of Tr ibu tylsta n n yl Ra d ica l. Ra tes
of F or m a tion , Ter m in a tion , a n d Abstr a ction Deter m in ed
by Tr a n sien t Absor p tion Sp ectr oscop y
Wendy J . Shaw, Pramod Kandandarachchi, J ames A. Franz, and Tom Autrey*
Pacific Northwest National Laboratory, Fundamental Science Division,
POB 999 MS#K2-57, Richland, Washington 99352
Received J anuary 26, 2004
Transient absorption spectroscopy (TAS) was used to measure the rate of formation and
the rate of self-termination of the main group metal hydride, tri-n-butyltin hydride (Bu
SnH). Irradiation of di-tert-butyl peroxide in the presence of Bu SnH generates the tri-n-
Sn ) by hydrogen atom abstraction. Analysis of the growth of the
3
-
3
•
butylstannyl radical (Bu
3
•
8
-1
Bu
s
3
Sn at 400 nm, corrected for radical termination, yields kH-abs ) (3.5 ( 0.3) × 10 M
. The extinction coefficient of the Bu Sn (ꢀ ) 1620 ( 40 cm at 400 nm), measured
for the first time in this work, is 3.5 times greater than the literature estimate. The
-
1
•
-1
-1
3
M
•
3 3
experimental value for ꢀ(Bu Sn ) permits the quantitative measurement of the rate of Bu -
•
9
-1 -1
Sn self-termination (2k
t
) (3.6 ( 0.3) × 10 M
s ) in benzene at 296 K. This experimental
value is greater than previous literature estimates but less than predicted by the
Smoluchowski equation (2k
s
9
-1 -1
t
) (4.4 ( 0.2) × 10 M
s ) using our measured diffusion
-
5
2
-1
coefficient for tin hydride (D ) (1.34 ( 0.08) × 10 cm s at 296 K in benzene). This work
shows that the rates of hydrogen abstraction from Bu
3
SnH by tert-butoxyl radical and self-
Sn are both more rapid than previously reported. A mechanistic kinetic
model was developed with this new kinetic information to fit the time-dependent TAS signal
•
termination of Bu
3
•
3
of Bu Sn . This approach provides the optimum reaction conditions, both low hydride and
halide concentrations, to measure the rate of bromine atom abstraction from 2-phenethyl
7
-1 -1
bromide (kBr-abs ) (3.7 ( 0.3) × 10 M
s ).
In tr od u ction
experimental conditions favoring competing path-
4
,5
ways, and empirical corrections used to correct bond
Main group metal hydrides are important hydrogen
6
enthalpies prompted us to undertake a closer examina-
atom donors utilized in synthetic free radical strategies
and as hydrogen atom donors in fundamental studies
tion of the spectroscopic properties of the tri-n-butyl-
•
stannyl radical (Bu3Sn ). An improved understanding
1
of organic free radicals. Tin hydride and the corre-
of the spectroscopic, kinetic, and thermodynamic prop-
erties of the tin hydride and stannyl radical is necessary
to improve the accuracy of basis rate determinations for
competitive rate studies. In particular, an Arrhenius
rate expression for halide abstraction by the stannyl
radical from 2-phenethyl bromide would provide an
important basis reaction in our competitive rate stud-
sponding stannyl radical in particular have been used
as kinetic and thermodynamic standards for developing
basis rate reactions.2 However, discrepancies in the
literature regarding the self-termination rate due to
ambiguities reported for the optical properties of the
3
stannyl radical, kinetic measurements performed under
7
ies.
*
Corresponding author: E-mail: tom.autrey@pnl.gov.
(
1) (a) Chatgilialoglu, C. In Radicals in Organic Synthesis; Renaud,
Scheme 1 shows the detailed mechanistic pathway for
the radical chain reduction of alkyl halides (RX f RH).
Irradiation of a photoinitiator, di-tert-butyl peroxide,
with a pulsed laser leads to the prompt (within the 20
ns laser pulse) formation of tert-butoxyl radicals (t-
P., Sibi, M. P., Eds.; Wiley-VCH: Weinheim, 2001; pp 28-49. (b)
Gilbert, B. C.; Parsons, A. F. J . Chem. Soc., Perkin Trans 2 2002, 367.
c) Baker, S. R., Parsons, A. F.; Pons, J .-F.; Wilson, M. Tetrahedron
(
Lett. 1998, 39, 7197. (d) Ryu, I.; Araki, F.; Minakata, S.; Komatsu, M.
Tetrahedron Lett. 1998, 39, 6335. (e) Spanswick, J .; Ingold, K. U. Int.
J . Chem. Kinet. 1970, 2, 157. (f) Carlsson, D. J .; Ingold, K. U. J . Am.
Chem. Soc. 1968, 90, 7047. (g) McIlroy, S.; Moran, R. J .; Falvey, D. E.
J . Phys. Chem. A 2000, 104, 11154. (h) Sibi, M. P.; Porter, N. A. Acc.
Chem. Res. 1999, 32, 163-171.
•
BuO ), eq 1. In the presence of tin hydride (Bu3SnH)
•
the t-BuO abstracts a H atom to yield t-BuOH and Bu3-
•
•
Sn , eq 2. In the absence of halide, the Bu3Sn will be
consumed by radical self-termination at diffusion-
(
2) (a) Chatgilialoglu, C.; Newcomb, M. Adv. Organomet. Chem.
999, 44, 67-112. (b) Galli, C.; Pau, T. Tetrahedron 1998, 54, 2893.
c) Crich, D.; Recupero, F. Chem. Commun. 1998, 189. (d) Franz, J .
1
(
A.; Suleman, N. K.; Alnajjar, M. S. J . Org. Chem. 1986, 51, 19. (e)
Musa, O. M.; Horner, J . H.; Shahin, H.; Newcomb, M. J . Am. Chem.
Soc. 1996, 118, 3862. (f) Walling, C.; Cioffari, A. J . Am. Chem. Soc.
(4) Scaiano, J . C. J . Am. Chem. Soc. 1980, 102, 5399.
(5) Ingold, K. U.; Lusztyk, J .; Scaiano, J . C. J . Am. Chem. Soc. 1984,
106, 343. Reaction conditions for bromide abstraction: tin hydride 10
wt %, propyl bromide 20-100 mM.
(6) Laarhoven, L. J . J .; Mulder, P.; Wayner, D. D. M. Acc. Chem.
Res. 1999, 32, 342.
1
5
972, 94, 6059. (g) Lemieux, R. P.; Beak, P. J . Org. Chem. 1990, 55,
454. (h) J ohnston, L. J .; Lusztyk, J .; Wayner, D. D. M.; Abeywick-
reyma, A. N.; Beckwith, A. L. J .; Scaiano, J . C.; Ingold, K. U. J . Am.
Chem. Soc. 1985, 107, 4594.
(
3) Nau, W. M.; Cozens, F. L.; Scaiano, J . C. J . Am. Chem. Soc. 1996,
(7) Shaw, W. J .; Lamb, C. N.; Autrey, T.; Kolwaite, D.; Camaioni,
D. M.; Alnajjar, M. S.; Franz, J . A. J . Org. Chem. 2004, 69, 1020.
1
18, 2275.
1
0.1021/om049933k CCC: $27.50 © 2004 American Chemical Society
Publication on Web 04/01/2004

Tributylstannyl Radical
Sch em e 1
Organometallics, Vol. 23, No. 9, 2004 2081
F igu r e 1. Transient absorption spectra of tri-n-butylstan-
nyl radical obtained from 308 nm pulsed irradiation of di-
tert-butyl peroxide in 0.02 M tri-n-butyltin hydride at 90
ns (O) and 150 µs (×) after photolysis.
controlled rates, eq 3. However, in the presence of an
alkyl halide (RX), halide abstraction competes with
radical termination to yield alkyl radical, eq 4. At low
radical concentrations, chain propagation, eq 5, com-
petes with alkyl radical termination, eq 6.
the literature to develop a mechanistic kinetic model
that accurately describes the time-dependent absorption
•
properties of Bu3Sn .
Although Scheme 1 details a multistep free radical
chain reaction, the rate of each step can be measured
by time-resolved spectroscopic methods under appropri-
ate experimental conditions. Fischer and co-workers
have measured the rates of self-termination of several
Resu lts a n d Discu ssion
Irradiation of di-tert-butyl peroxide in an argon-
purged benzene solution containing tin hydride yields
a transient intermediate with a strong absorbance
feature at 400 nm, as shown in Figure 1. The rate of
formation of the transient is dependent on the tin
hydride concentration and decays to baseline over a
period of several microseconds by second-order path-
ways. In the presence of a primary halide, phenethyl
bromide (PhCH2CH2Br), the lifetime of the 400 nm
transient is reduced and the observed decay is first-
order in bromide concentration at lower tin hydride
concentrations. These observations are consistent with
the previous assignment of the 400 nm transient as Bu3-
8
organic free radicals by EPR methods; however, the tri-
n-butylstannyl radical is not easily observable by EPR
9
•
in solution. Fortunately, Bu3Sn is reported to have a
characteristic absorption feature at 400 nm that is
sufficiently strongly absorbing10 to be monitored by
transient absorption spectroscopy (TAS), and the rate
of halide abstraction can be obtained by monitoring the
•
disappearance of Bu3Sn as a function of halide concen-
tration. However, the accuracy and precision of the
observed rate will be suspect if the experimental condi-
tions are not carefully controlled. Specifically, to obtain
an unambiguous bromide abstraction rate constant from
a measure of the radical decay as a function of halide
concentration, three experimental conditions must be
•
4
Sn . One notable anomaly is the decay behavior of the
transient at higher hydride concentrations; the absorp-
tion signal at 400 nm does not appear to come back to
baseline over several microseconds. The “persistent
absorption” is due to the chain reaction that regenerates
the tri-n-butylstannyl radical, not from the hydrogen
atom abstraction by tert-butoxyl radical, but from
hydrogen atom abstraction by the alkyl radical, eq 5.
At higher halide concentrations the chain reaction leads
to a persistent absorption on the microsecond time scale
that leads to an error in the measured abstraction rate.
•
met. First, the Bu3Sn is not still being formed; that is,
all the tert-butoxyl radical has completely reacted, eq
•
1
. Second, self-termination of Bu3Sn , i.e., disappearance
by a competing second-order reaction, does not dominate
the transient decay, eq 3.11 Third, the subsequent
reaction of the alkyl radical with tin hydride must not
regenerate significant concentrations of stannyl radical
on the time scale of our measurement, i.e., eq 5.
Alternatively, knowledge of an accurate extinction coef-
1
2
Beckwith and co-workers have recently suggested
that the tri-n-butylstannyl radical may add to benzene
at 70 °C, as inferred from rate retardation in the
reduction of benzoate esters relative to cyclohexane. We
did not observed any absorbance in the room-tempera-
ture TAS in the wavelength region where cyclohexadi-
enyl radical derivatives have been observed as formed
•
ficient for n-Bu3Sn also allows quantitiative kinetic
modeling of the suite of eqs 1-6. The goal of this work
is to improve the accuracy of time-resolved spectroscopic
measurements, by discovery of the optimum experimen-
tal reaction conditions, to obtain a quantitative descrip-
tion of the chemical and physical properties of the
1
3
by other pathways.
•
Ra te of Bu 3Sn Self-Ter m in a tion (2kt). Assuming
radical-radical reactions are diffusion controlled, the
rate of self-termination can be estimated from the
Smoluchowski equation, eq 7, where σ is a spin statisti-
•
Bu3Sn species. The work below describes a more
•
accurate measurement of the rate of Bu3Sn formation,
•
eq 2, the rate of Bu3Sn self-termination, eq 3, and the
rate of halide abstraction, eq 4, by using time-resolved
transient absorption spectroscopy. The rate data deter-
mined in this work are combined with rate data from
s
2kt ) (8π/1000)σF DABN
(7)
Α
cal factor, F is the reaction diameter of the radical
(
(
8) Fischer, H.; Paul, H. Acc. Chem. Res. 1987, 20, 200.
9) Mochida, K.; Wakasa, M.; Sakagunchi, Y.; Hayashi, H. J . Am.
Chem. Soc. 1987, 109, 7942.
(12) Beckwith, A. L. J .; Bowry, V. W.; Bowman, W. R.; Mann, E.;
Parr, J .; Storey, J . M. D. Angew. Chem., Int. Ed. 2004, 43, 95. We thank
a reviewer for bringing this reference to our attention.
(13) Effio, A.; Griller, D.; Ingold, K. U.; Scaiano, J . C.; Sheng, S. J .
J . Am. Chem. Soc. 1980, 102, 6063.
(
10) Wakasa, M.; Kugita, T. Organometallics 1998, 17, 1913.
(11) Cross-termination reactions of stannyl radical with alkyl radi-
cal, reaction 6b, can also decrease the observed absorbance. See:
R u¨ egge, D.; Fischer, H. Int. J . Chem. Kinet. 1986, 18, 145.

2
082 Organometallics, Vol. 23, No. 9, 2004
Shaw et al.
species, DAB is the diffusion coefficient of the radical
solute A in the solvent B, and N is Avogadro’s number.
For several organic radical species this empirical ap-
proach has been demonstrated to be accurate to within
8
2
0% of the corresponding experimental values.
Assuming that only solvent-caged singlet radical pairs
1
react (i.e., σ ) /4), all that is needed to calculate the
s
Smoluchowski termination rate, 2kt , are the radical
reaction diameter and diffusion coefficient. It is gener-
•
ally assumed that both F and D of the radical (R ) are
1
4
well modeled by the corresponding hydrocarbon (RH).
These two assumptions have been shown to be valid for
several organic radicals in nonassociating organic sol-
vents, and we believe it is a valid approximation for tri-
n-butylstannyl radical in benzene. Specifically, it is
unlikely that addition of hydrogen atom to the tri-n-
butylstannyl radical will significantly change the reac-
F igu r e 2. Decay of 400 nm transient generated from 308
nm pulsed irradiation of di-tert-butyl peroxide in 0.02 M
tri-n-butyltin hydride. Inset is observed second-order decay
1
5
tion diameter. Since the reaction diameter is twice the
radius of the reacting species, we calculate the radius
of Bu3SnH from the known molar density to obtain an
estimate of the reaction diameter of the radical assum-
ing close packing spheres.16 To be confident that the
6
-1
at 400 nm, 2k
t
/ꢀ ) 2.16 × 10 s
.
of Bu3SnH in benzene. Comparison of our experimen-
-5
2
-1
tally measured DTD ()(1.34 ( 0.08) × 10 cm s at
•
assumption is valid for DAB (i.e., DBu3SnH ) DBu3Sn ),
2
96 K) coefficient obtained from the dispersion method
at 296 K) shows
excellent agreement between the experimental method
there must be no significant interactions between the
radical and the solvent. This assumption appears rea-
sonable given the recent demonstration that D values
for aryl alcohols are comparable to D of the correspond-
ing ketyl radicals in hydroxylic solvents.17 With our
confidence in the above pair of assumptions two paths
were used to determine D of Bu3SnH to be used as a
-5
2
-1
with DSW ()1.30 × 10
cm
s
and the empirical calculation for tri-n-butyltin hydride
1
in benzene. Setting σ ) /4 and using the average of DTD
-5
2
-1
and DSW ((1.32 ( 0.08) × 10 cm s ) combined with
-8
16
F ()8.73 × 10 cm) obtained from the molar volume
of tin hydride we calculate the rate of self-termination
•
representation for D of Bu3Sn in benzene.
•
s
9
of Bu3Sn in benzene at 296 K (2kt ) (4.4 ( 0.2) × 10
).
Alternatively, the rate of a radical self-termination
Our first approach uses the Stokes-Einstein (SE)
equation, eq 8, which predicts that the diffusion coef-
ficient of the solute is proportional to the ratio of the
temperature (T) to viscosity (η) multiplied by a numer-
ical constant (kB/6πra), where kB is the Boltzman
constant and ra is the radius of the solute. The Spernol-
-1 -1
M
s
reaction, 2kt, can be determined experimentally from
the observed second-order decay of the transient absor-
bance, A(t), shown in Figure 2. Least-squares analysis
of 1/A(t) verses time yields the absorbance termination
Wirtz (SW) modification of the SE equation, eq 9, uses
6
-1
rate constant, 2kt/ꢀ ) (2.2 ( 0.2) × 10 s at 400 nm
a microfriction factor (f)1
8,19
to correct the diffusion
21
for our experimental apparatus. Given an accurate
coefficient for differences in solute-solvent molecular
size and shape.
•
value for the extinction coefficient of ꢀ(Bu3Sn ) at 400
nm, the rate constant for tri-n-butylstannyl radical self-
termination can be obtained. The cited literature value
DSE ) (T/η)(k /6πr )
(8)
(9)
-1
-1
3
B
a
(ꢀlit ) 450 M cm ) at 400 nm yields 2kt ) (9.9 ( 0.2)
8
-1 -1
×
10 M
s
in benzene at 296 K, which is significantly
DSW ) DSE/f
slower than expected for a diffusion-controlled reaction
pathway.
Our second approach, the Taylor’s dispersion (TD),20
method was used to measure the diffusion coefficient
Our experimental termination rate constant deter-
mined from the published extinction coefficient and our
decay data is well outside the typical uncertainty of the
Smoluchowski equation. This is a surprising result
given that Fischer has demonstrated that the Smolu-
chowski equation is generally very close to experiment
(
(
14) Shuh, H.-H.; Fischer, H. Int. J . Chem. Kinet. 1976, 8, 341.
15) MW n-Bu SnH ) 290.7, MW n-Bu Sn radical ) 289.7. On the
3
3
other hand for a lighter molecule the molecular volume of HO radical
14.4 mL/mol, MW 17 versus 18 mL/mol for H O, MW ) 18). This
O). Autrey, T.;
(
2
amounts to a 93% smaller F(HO) compared to F(H
2
Brown, A. K.; Camaioni, D. M.; Dupuis, M.; Foster, N. S.; Getty, A. J .
Am. Chem. Soc. 2004, 126, 3680.
(within (20%) for conventional organic radicals in
-8
nonassociating organic solvents, such as benzene. There-
fore, either the Smoluchowski method fails badly for tri-
n-butylstannyl radical in benzene or the literature
report of ꢀ400 nm is in error by as much as 400%.
(
16) F, the reaction diameter, 8.73 × 10 cm, is twice the radius r
/3 -22 3
x x
(
r ) (3 xV
N/4π)1 ). V
is the molecular volume 4.47 × 10
cm /
molecule calculated from molar density, x ) 0.74, the volume fraction
for cubic closest-packing spheres, and N is Avogadro’s number.
(
17) Autrey, T.; Kandanarachchi, P.; Franz, J . A. J . Phys. Chem. A
2
001, 105, 5948.
(
(
Resolution of this discrepancy is important. If the
18) Shuh, H.-H.; Fischer, H. Helv. Chim. Acta 1978, 91, 2130.
19) Spernol-Wirtz microfriction factor f ) (0.16 + 0.4r /r )(0.9 +
r - 0.25T r). Reduced temperatures, T , are given by, Txr ) (T -
xf)/(Txb - Txf), where Txf and Txb are freezing and boiling points of
species x ) n-Bu SnH (A) and benzene (B). The radii in the micro-
friction factor term are given by r
Smoluchowski equation fails to predict Bu3Sn• self-
A
B
0
T
.4T
A
B
r
termination in benzene, this may indicate unique radi-
cal termination dynamics occurring with Bu3Sn• in
3
1
/3
A
A
) (3V (c)/4FN) , where c ) 0.74,
the volume fraction for cubic closest-packed spheres. Andrade viscosity
t
(21) The decay is proportional to 2k /ꢀl, were ꢀ is the extinction
of benzene is given by the expression ln(η,cP) ) -4.117 + 2093.7/RT.
coefficient of the transient and l is the path length, in our experiment
) 1 cm.
(
20) Taylor, S. G. Proc. R. Soc. London A 1953, 219, 186.

Tributylstannyl Radical
Organometallics, Vol. 23, No. 9, 2004 2083
F igu r e 3. Extinction coefficient measurement of tri-n-
butylstannyl radical measured at 400 nm (ꢀ ) 1620 ( 40
-
1
-1
M
cm ) using Aberchrome 540. The absorbances of
F igu r e 4. Irradiation of a benzene solution of di-tert-butyl
fulgide and tin radical were measured as a function of laser
power (308 nm) just after the laser was fired. The extinc-
tion coefficient of the stannyl radical is proportional to the
slope of the actinometer (mact) (]) and to the slope of tin
radical (mSn) (0).
peroxide containing 0.02 M Bu
3
SnH. Comparison of 400
Sn ) before (red O) and after (black +)
termination correction. Inset shows first-order analysis of
•
nm transient (n-Bu
3
8
-1 -1
rise time with correction, kabs ) 3.5 × 10
M
s
, and
8
-1
s-1, for radical
without correction, kabs ) 4.0 × 10
M
termination.
benzene, e.g., σ * 1/4 and/or DBu3Sn• * DBu3SnH. The
likelihood that the quoted literature value for the
extinction coefficient was in serious error convinced us
subsequently taken as a direct experimental measure.3
Thus, our value provides the first experimental measure
•
•
to revisit the determination of ꢀ of the Bu3Sn radical.
of the ꢀ(Bu3Sn ). Using our experimentally determined
•
•
Extin ction Coefficien t of Bu 3Sn . Pulsed irradia-
extinction coefficient of Bu3Sn we obtain the rate of self-
tion of Aberchrome 540 at 308 nm generates a deep red
termination from our experimental decay data at 400
exp
9
-1 -1
(
λmax ) 540 nm) cyclized isomer within nanoseconds
nm (2kt
) (3.6 ( 0.3) × 10 M
s ). The difference
2
2
s
with a quantum efficiency of 0.20. In the same sample
cell a solution of di-tert-butyl peroxide solution with
matching absorbance, containing 20 mM Bu3SnH, was
between the empirical Smoluchowski estimate of 2kt
9
-1 -1
((4.4 ( 0.3) × 10 M
s ) and our experimental value
is almost within the 20% error bars typically observed
•
8
irradiated under identical conditions. Since t-BuO is
in this type of comparison as noted by Fischer. If the
•
23
s
exp
rapidly and quantitatively converted to Bu3Sn under
the reaction conditions, the quantum yield for tri-n-
butylstannyl radical can be determined from the litera-
ture quantum yield of t-BuO , ΦSn ) ΦtBuO . A plot of
∆
difference between 2kt and 2kt is real, the difference
may be explained by solvent cage dynamics dictating
competition between cage-escape versus cage-recombi-
•
•
• 24
26
nation. Lui and Scott observed the opposite effect for
•
25
OD at time ) 0 for both the Bu3Sn (λ ) 400 nm)
phenylthiyl radical termination: the observed radical
termination was faster than expected on the basis of
-
1
and the actinometer (λmax ) 540 nm, ꢀact ) 8200 cm
M
-
1
1
) as a function of incident laser power is shown in
Smoluchowski. This is most likely due to σ > /4 due to
Figure 3. Least-squares fit of the data provides the ratio
the fast intersystem crossing of triplet radical pairs in
2
7
•
•
the solvent cage. In the present case one would have
of ꢀSn /ꢀact by direct comparison of the slopes (mSn ) 4.32
-
4
-4
1
×
10 , F 0.996; mact ) 2.62 × 10 , F 0.996). Using eq
to argue that σ < /4 for tri-n-butylstannyl radical
•
1
0 the extinction coefficient for Bu3Sn was determined
termination in benzene. Further work over an extended
temperature range is planned with a goal to elucidate
the origin of this difference.
•
-1
-1
(
ꢀSn )1620 ( 40 M cm at 400 nm).
•
ꢀSn ) ꢀact(Φact/ΦSn /mact
•
•
) × (mSn
•
)
(10)
R a t e of F or m a t ion of Bu 3Sn . At the radical
-
6
concentrations generated by a pulsed laser, 10 M <
•
-5
The discrepancy between our experimental extinction
[R ] < 5 × 10 M, radical termination will occur during
the rise time of the experiment, leading to reduction in
the transient absorbance on the time scale of the
coefficient, ꢀexp ) 1620 ( 40 M-1 cm at 400 nm, and
-1
-
1
-1
the literature extinction coefficient, ꢀlit ) 450 M cm
at 400 nm, appears due to a previous misunderstanding.
Close examination of the original citation shows that
there was no direct experimental measure of the extinc-
tion coefficient. It appears that the ꢀlit at 400 nm was
experiment. Therefore, A and consequently A∞ are
t
reduced in amplitude by radical self-termination. This
leads to errors in the measurement of kabs (ca. 5-90%).28
To obtain an accurate measure of kabs, the quantity of
4
1e
•
estimated from an estimate of 2kt in cyclohexane and
Bu Sn consumed by self-termination pathways occur-
3
•
ring during the growth of Bu3Sn can be accounted for
with a termination correction value ()2kt/ꢀ)(A(t) ∆t), as
(
22) Wintgens, V. J ., L. J .; Scaiano, J . C. J . Am. Chem. Soc. 1988,
10, 511.
23) Millimolar concentrations of tin hydride are sufficient to trap
all the tert-butoxyl radical. The rate of â-scission of tert-butoxyl radical
2
1
(
shown in eq 11. Substitution of *At for At in the
conventional first-order analysis yields eq 12. The
results of this approach are illustrated in Figure 4,
4
-1
is 1.4 × 10
N.; Yurkovskaya, A. V. J . Phys. Chem. A 1998, 102, 7975.
24) The quantum yield for t-BuO radical formation from irradiation
s
in benzene. Tsentalovich, Y. P.; Kulik, L. V.; Gristan,
(
of di-tert-butyl peroxide in benzene at ambient temperature was 1.66.
Wayner, D. D. M.; Lusztyk, E.; Page, D.; Ingold, K. U.; Mulder, P.;
Laarhoven, L. J . J .; Aldrich, H. S. J . Am. Chem. Soc. 1995, 117, 8737.
(26) Lui, S. N.; Scott, T. W. J . Phys. Chem. 1989, 93, 1393.
(27) Autrey, T.; Devadoss, C.; Sauerweun, B.; Franz, J . A.; Schuster,
G. B. J . Phys. Chem. 1995, 99, 869.
(28) Franz, J . A.; Bushaw, B. A.; Alnajjar, M. S. J . Am. Chem. Soc.
1989, 111, 268.
(25) The absorbance at time ) 0 is calculate by fitting the second-
order decay and extrapolating back to time ) zero.

2
084 Organometallics, Vol. 23, No. 9, 2004
Shaw et al.
Ta ble 1. Kin etic a n d Op tica l P r op er ty Da ta for
Tr i-n -bu tyltin Hyd r id e a n d th e Cor r esp on d in g
Sta n n yl Ra d ica l, Com p a r ison of Exp er im en ta l a n d
a
Liter a tu r e Da ta
pathway
this work
literature
citation
Rate (M- s-1
(3.5 ( 0.3) × 10
1
)
•
8
8
t-BuO +
1.9 × 10
b
c
d
n-Bu
Sn +
n-Bu
3
SnH
•
9
7
9
n-Bu
3
(3.6 ( 0.3) × 10
(3.7 ( 0.3) × 10
ND
2.1 × 10
Sn•
3
•
7
n-Bu
3
Sn +
2.6 ( 0.4 × 10
PhCH
PhCH
CH •
n-Bu SnH
2 2
CH Br
9
2
2
+
2.3 × 10
3
Extinction Coefficient
-
1
-1
(
cm
M
at 400 nm)
450
n-Bu
3
3
Sn•
1620 ( 40
f
Diffusion Coefficient
F igu r e 5. Comparison of experimental 400 nm transient
2
-1
(cm
s
)
•
(n-Bu
3
Sn ) growth and decay with the mechanistic kinetic
-5
-5
n-Bu
SnH
(1.34 ( 0.08) × 10
1.30 × 10
g
model prediction using data determined in this study, ꢀ )
a
b
c
-1
-1
8
-1 -1
Reported errors are 1 standard deviation. Ref 4. Ref 1f in
1620 M cm , kH-abs ) 3.5 × 10 M
s
, and 2k
t
) 3.6
9
-1
9
-1 -1
cyclohexane, 1.4 × 10
s
, corrected for differences in viscosity
× 10 M
s . The solid red line through the absorbance
d
η(cyclohexane)/η(benzene) ) 0.898/0.603. Ref 5, n-propyl bromide.
is model prediction.
e
ND, not determined in this work; ref 26, ethyl radical + tin
f
g
hydride was used for kinetic model. Refs 3 and 4 Calculated at
for primary alkyl radical abstraction from tin hydride
2
96 K by the Spernol-Wertz method as discussed in ref 18.
(
eq 5), allow us to develop a mechanistic kinetic model
•
(MKM) that predicts the concentration of [Bu3Sn ] as a
which compares the observed absorption of tri-n-butyl-
stannyl radical to the absorption corrected for radical
termination occurring during the rise time of the
observed signal.
function of both halide and hydride concentration that
we can directly compare with our experimental TAS
data. One of the goals in the present work was to
discover a TAS “time window” to measure the rate of
•
2
bromide abstraction by Bu3Sn from 2-phenethyl bro-
*
Α(t) ) A(t) + (2k /ꢀ)(A(t) ∆t)
(11)
(12)
∑
t
mide, eq 4. A comparison of the MKM developed from
Scheme 1 using rate data from Table 1 with the
kabs [Bu SnH] ) ln[A /(A - *A )]
3
∞
∞
t
•
experimental absorbance of Bu3Sn as a function of time
•
is shown in Figure 5. It is apparent from examination
of Figure 5 that there exists a finite window in time
where Bu3Sn has completely formed and has not begun
The rate of formation of Bu3Sn generated by H atom
•
abstraction from Bu3SnH by t-BuO can be obtained
•
from the rise time of the termination-corrected absorp-
tion data shown in Figure 4. By correcting the observed
experimental At measured at 400 nm using this ap-
to undergo self-termination to a significant extent.
Furthermore, this time window is dependent on the
concentration of Bu3SnH. For the reaction conditions
8
-1
proach the rate constant for kabs is (3.5 ( 0.3) × 10 M
•
-1
shown in Figure 5 ([Bu3SnH] ) 0.02 M) Bu3Sn has
s
. Without correction, the apparent rate of hydrogen
8
-1 -1
reached a maximum after ca. 500 ns. After about 2.5
atom abstraction (kabs ) (4.0 ( 0.3) × 10 M
s ) is
•
µs Bu3Sn has decayed to ca. 85% of the initial concen-
overestimated, even though the radical concentration
-
5
tration. These results suggest that a time window
between 0.5 µs < t < 2.5 µs should provide the best
opportunity to measure the bromide abstraction rate by
transient absorption spectroscopy. One might be tempted
to use higher tin hydride concentrations to reduce the
is kept relatively low, about 1.9 × 10 M total radical
concentration, by attenuation of the excitation pulse (<2
mJ ). Although this termination-correction approach
does not lead to an increase in precision of the measure-
ment, it does result in an improvement of the experi-
mental accuracy. In the present example, at low laser
power, the experimental accuracy is enhanced by ∼10%;
however, at higher laser power (increased radical con-
centrations) obviously the error in rate accuracy can
become quite significant. Our measured value for kabs
is well outside the error of the previous literature
determination (see Table 1). The source of this discrep-
ancy reflects the challenge inherent in accurate mea-
surements of fast pseudo-first-order kinetics using
nanosecond TAS. In a subsequent paper we will describe
our measurements of this abstraction reaction by time-
resolved photoacoustic calorimetry, which is better
•
lifetime of t-BuO to enable the measurement of the
kinetic experiment earlier; however this leads to greater
interference from subsequent chain reaction. At higher
halide and/or tin hydride concentrations the alkyl
radical formed by abstraction (eq 4) subsequently at-
•
tacks the tin hydride to yield a “second” Bu3Sn (eq 5).
This series of chain reactions will occur at high tin
hydride concentrations and result in a persistent ab-
sorbance on the TAS time scale.
The persistent absorbance or “offset” created by the
chain reaction may explain the apparent low rate of
abstraction reported previously for n-propyl bromide
7
-1 -1
5
2
9
((2.6 ( 0.4) × 10 M
s
at 295 K). At higher halide
suited to measure kinetics on this time scale.
P r im a r y Ha lid e Abstr a ction by Bu 3Sn . The ex-
perimental data obtained in this work, ꢀBu3Sn , kabs (eq
), and 2kt (eq 3), combined with the literature value
•
concentrations, the chain reactions occur on the ob-
served time scale of the TAS experiment to regenerate
the tri-n-butylstannyl radical and consequently lead to
•
e
30
2
(
29) Autrey, T.; Foster, N. S.; Klepzig, K.; Amonette, J . A.; Dasch-
(30) Chatgilialoglu, C.; Ingold, K. U.; Scaiano, J . C. J . Am. Chem.
Soc. 1981, 103, 7739.
bach, J . L. Rev. Sci. Instrum. 1998, 69, 2246.

Tributylstannyl Radical
Organometallics, Vol. 23, No. 9, 2004 2085
tert-butyl peroxide, tri-n-butyltin hydride, and phenethyl
bromide (Aldrich) were used as received. Benzene was distilled
and recrystallized from the melt to minimize interference from
fluorescent impurities. AcuChem,31 used in the kinetic model-
ing, uses a variable step integrator to solve stiff integration
problems.
Tr a n sien t Absor p tion Sp ectr oscop y. The laser flash
1
0
photolysis system has been described previously. The at-
tenuated output (<2 mJ ) of an excimer laser (Lambda Physik
Compex-100, XeCl, 308 nm) was directed onto a fluorescent
sample cell perpendicular 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 minimized, 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
F igu r e 6. Observed rate of transient decay at 400 nm (n-
•
Bu
3
Sn ) as a function of 2-phenethyl bromide concentration
at 299 K. Least-squares analysis of the slope provides the
•
rate of bromine atom abstraction by (n-Bu
primary alkyl bromide, (3.7 ( 0.3) × 10 M
3
Sn ) from a
7
-1 -1
nm to detect the absorbance maximum of the visible Bu
3
Sn•
s
.
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 a housing
attached to the exit slit of the MC. The output of the PMT
tube was digitized on a Lecroy 534 transient digitizer, and the
data were transferred to a PC for analysis.
underestimating the halide abstraction rate; thus while
the initial tri-n-butylstannyl radical is consumed by
halide abstraction, the tri-n-butylstannyl radical is
regenerated by the subsequent chain reactions! How-
ever, at lower halide and hydride concentrations the
offset can be minimized. Kinetic modeling shows that
at a tin concentration of 20 mM the halide concentration
should be kept < 25 mM to reduce the formation of an
offset arising from the subsequent chain reactions. The
Extin ction Coefficien t of Tr i-n -bu tylsta n n yl Ra d ica l.
15
The methods demonstrated by Wintgens et al. were used to
measure the extinction coefficient of the stannyl radical. A
solution of the Fulgide actinometer, Aberchrome 540, was
-1
prepared in toluene with an absorbance of 0.24 cm . Irradia-
tion of the pale yellow fulgide actinometer with a pulsed UV
light source (308 nm) generates the deep red cyclized isomer
with a quantum yield of 0.2. The extinction coefficient of the
red cyclized isomer has an absorbance maximum at 540 nm
•
rate of bromide atom abstraction by Bu3Sn measured
under optimum conditions, low halide (10-25 mM) and
hydride concentrations (20 mM), fit between 1.5 and 2.5
µs, yields the rate for primary bromide abstraction ((3.7
-
1
-1
(ꢀ ) 8200 cm
M ). The solution was pumped through the
7
-1 -1
(
0.3) × 10 M
s
at 299 K), as derived from Figure
fluorescence cell (20 mL/min), and the ∆OD(t) of the sample
was measured at 540 nm as a function of excitation pulse
energy. A second solution containing tin hydride (0.02 M) and
di-tert-butyl peroxide (ca. 2 vol %) providing an absorbance of
6
.
A narrow window of reagent concentrations is thus
dictated for these determinations; however, at higher
bromide concentrations, the generated alkyl radical
reacts with tin hydride to regenerate the tri-n-butyl-
stannyl radical, and at lower halide concentrations, tri-
n-butylstannyl radical termination begins to contribute
to the decay at 400 nm.
-
1
0.24 cm at 308 nm was prepared in benzene. The solution
containing the tin hydride was pumped through the quartz
fluorescence cell (2 mL/min), and the ∆OD(t) was measured
at 400 nm. Analysis of the time-dependent decay at 400 nm
followed expected second-order behavior. Care must be exer-
cised in using TAS to measure second-order kinetics due to
radical concentration gradients that can affect the precision
of the measurement. For 1 cm TAS cells the perpendicular
pump/probe geometry is preferred for measuring second-order
Su m m a r y a n d Con clu sion s
In this work we have used transient absorption
spectroscopy (TAS) to show that the rates of hydrogen
abstraction from Bu3SnH by tert-butoxyl radical, bro-
kinetics to minimize concentration gradients across the diam-
-1
eter of the probe beam. In our experiments, OD ) 0.24 cm
,
the concentration gradient across the probe beam, front side
to back side, is ca. 5%. Alternative, short path cells, e.g., 1
•
mine atom abstraction by Bu3Sn , and self-termination
•
32
of Bu3Sn are all more rapid than previously reported.
mm, have been used to minimize the concentration gradient.
•
The extinction coefficient of Bu3Sn is shown to be 3.5
Br om in e Atom Abstr a ction . 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 (10.2,
14, 19.8, and 24.9 mM). The solutions were purged with
solvent-saturated argon and charged into a gastight syringe.
The samples were pumped (Sage syringe pump, model 355)
through a low-volume quartz fluorescent flow cell (NGS
T59FL) to avoid the formation of photoproducts and to ensure
the bromide concentration did not change significantly under
the reaction conditions. Control experiments showed that a
times larger than the literature estimate. This new
datum was used to develop a kinetic model to predict
optimal experimental conditions to measure the rate of
•
bromine atom abstraction by Bu3Sn from a primary
alkyl bromide. These experimental results and the
kinetic modeling will permit a more accurate evaluation
of absolute rate of bromine atom abstraction to be used
in competitive rate studies.
Exp er im en ta l Section
(
31) Braun, B.; Herron, J . T.; Kahaner, D. K. Int. J . Chem. Kinet.
988, 20, 51.
32) Fischer, H.; Paul, H.; M u¨ nger, K.; Dschen, T. J . Chem. Soc.,
Perkin Trans. 2 1985, 213.
1
Gen er a l P r oced u r es. All reagents were purchased from
commercial vendors. Aberchrome 540 (Aberchromics Ltd.), di-
(

2
086 Organometallics, Vol. 23, No. 9, 2004
Shaw et al.
2
long pass filter (LP-365) was necessary to attenuate the
ultraviolet output from the probe beam (150 W Xe lamp) to
prevent decomposition of the primary bromide.
tube (r), the retention time (t), and the variance (σ ) by the
2
2
equation D ) r t/24σ . The D was calculated using the
Spernol-Wirtz treatment; DAB/T ) k/(6πηr f) where f is the
A
Diffu sion Coefficien t of Bu
3
Sn H in Ben zen e, Exp er i-
microfriction factor.
m en t a l Disp er sion , a n d Ca lcu la t ed (Sp er n ol-Wir t z)
Meth od s. The diffusion coefficients of tri-n-butyltin hydride
in benzene were measured with Taylor’s dispersion method
using a coil of PEEK tubing (length 22.8 m; internal radius
Ack n ow led gm en t. Support for this work from the
Office of Science, Office of Basic Energy Sciences,
Chemical Sciences Division, is gratefully acknowledged.
The work was conducted at Pacific Northwest Labora-
tory, which is operated for the U.S. Department of
Energy by Battelle under Contract DE-ACO6-76RL0
0
.0247 ( 0.0004 cm; coil diameter 21 cm) held at constant
temperature in a Neslab RTE-211 temperature bath. A
detailed description of our apparatus used to measure diffusion
coefficients is described elsewhere.33 Briefly, 10 µL of dilute
1
830.
Bu
3
SnH solutions (0.06-0.15 M) were injected and loaded
(Rheodyne on the PEEK tubing coil containing helium-purged
solvent pumped through the coil at a flow rate of 0.2 mL/min
using Waters associate HPLC pump with Waters 410 dif-
ferential refractometer or Waters 2410 refractive index detec-
Su p p or tin g In for m a tion Ava ila ble: Input file for MKM
of chain reaction shown in Scheme 1 to predict time-dependent
behavior of tri-n-butylstannyl radical. Output of MKM for high
halide (0.1 M RBr and tin hydride (0.06 M). The MKM input
file can be used to predict time-dependent behavior of tri-n-
butylstannyl radical as a function of initial tin hydride and
alkylbromide concentrations. This material is available free
of charge via the Internet at http://pubs.acs.org.
2
tor). The variance (σ ) of the tri-n-butyltin hydride peak was
determined by the least-squares analysis of a Gaussian curve
of the refractive index data. The experimental diffusion
coefficient, D, is related to the internal radius of the coiled
(33) Autrey, T.; Alnajjar, M. S.; Nelson, D. A.; Franz, J . A. J . Org.
Chem. 1991, 56, 2197.
OM049933K