Mechanism of the reaction of tri-n-butylstannyl anion with n-butyl bromide and iodide as studied by the 119Sn CIDNP technique

Article abstract of DOI:10.1021/om990094p

The mechanism of the reaction of the tri-n-butylstannyl anion with n-butyl bromide and iodide in THF was studied by the chemically induced dynamic nuclear polarization (CIDNP) technique. Strong ¹¹⁹Sn CIDNP spectra were observed for the first ti

Full text of DOI:10.1021/om990094p

Organometallics 1999, 18, 2941-2943
2941
Mech a n ism of th e Rea ction of Tr i-n -bu tylsta n n yl An ion
w ith n -Bu tyl Br om id e a n d Iod id e As Stu d ied by th e 119Sn
CIDNP Tech n iqu e
Masanobu Wakasa*
Molecular Photochemistry Laboratory, The Institute of Physical and Chemical Research
(RIKEN), Wako, Saitama 351-0198, J apan
Tsuyoshi Kugita
Department of Environmental & Material Science, Teikyo University of Science and
Technology, Uenohara, Yamanashi 409-0193, J apan
Received February 12, 1999
Summary: The mechanism of the reaction of the tri-n-
butylstannyl anion with n-butyl bromide and iodide in
THF was studied by the chemically induced dynamic
nuclear polarization (CIDNP) technique. Strong 119Sn
CIDNP spectra were observed for the first time. The
emissive CIDNP phase of the tetra-n-butylstannane
proves that radical intermediates are involved in the
present reactions.
the reaction of tri-n-butylstannyl anion with s- and tert-
butyl bromides and iodides and succeeded in the direct
observation of the tri-n-butylstannyl radical for the first
time.8 However, in the case of the reactions with n-butyl
halides, the formation of a stannyl radical could not be
observed. Thus, to this day, a reaction intermediate of
the title reactions has not been observed directly, and
so the reaction mechanism is still unclear.
Chemically induced dynamic nuclear polarization
(CIDNP) is the most powerful technique for studying
radical reactions.9-16 The appearance of enhanced ab-
sorption or emission in the NMR spectra of products
generated through a radical pathway gives vital details
concerning the reaction mechanism. Therefore, exten-
The mechanism of the reactions of organostannyl
anions with alkyl halides has received considerable
attention over the past half-century.1-4
R3SnM + R′X f R3SnR′ + MX
M ) Li, Na, K; X ) Cl, Br, I
(1)
1
sive H and 13C CIDNP studies have been carried out
for many types of photochemical reactions.15,17 However,
such CIDNP techniques are inappropriate for fast
thermal reactions.18 This is so because it is necessary
to use a sample flow system, in which the reactants are
mixed in an NMR tube placed in an NMR spectrometer,
and to remove the reaction products continuously from
the NMR tube during signal accumulation. Since, in
such a flow system, the NMR tube cannot be rotated,
Because this reaction often has shown unexpected
behavior with respect to yields and stereochemistry,5,6
extensive studies have been carried out on the stereo-
chemical course of such reactions and on quantitative
analyses of product distributions.1-6 Three basic mecha-
nistic pathways have been proposed: (1) classical SN2
substitution, (2) a radical pathway (initial electron-
transfer process from the anion to the halide, leading
to radical intermediates), and (3) an ionic pathway
(initial halogen-metal exchange, leading to carbanion
intermediates).1-4 For reactions of the tri-n-butylstannyl
anion with n-butyl bromide and iodide, it was reported
that both the classical SN2 substitution and a radical
pathway are involved.2,7
1
the resolution and sensitivity of the H and 13C NMR
spectra are much lower than usual.19 We chose to apply
the 119Sn CIDNP technique20 and found strong CIDNP
spectra of the reaction products. Especially, the emissive
119Sn CIDNP phase of the main product, tetra-n-
(8) Wakasa, M.; Kugita, T. Organometallics 1998, 17, 1913.
(9) Bargon, J .; Fischer, H. Z. Naturforsch. 1967, A22, 1556.
(10) Ward, H. R.; Lawler, R. G. J . Am. Chem. Soc. 1967, 89, 5518.
(11) Closs, G. L. J . Am. Chem. Soc. 1969, 91, 4552.
(12) Kaptein, R. J . Chem. Soc., Chem. Commun. 1971, 732.
(13) Mitchell, T. N. Tetrahedron Lett. 1972, 2281.
(14) Kaptein, R.; Oosterhoff, L. J . Chem. Phys. Lett. 1969, 4, 195,
214.
An important approach to the question of the mech-
anism of such reactions is that of direct observation of
a reaction intermediate. However, there have been few
reports concerning intermediates in such reactions.
Recently, we carried out stopped-flow measurements of
(15) Muus, L. T., Atkins, P. W., McLauchlan, K. A., Pedersen, J .
B., Eds. Chemically Induced Magnetic Polarization; Reidel: Dordercht,
1977.
(1) Ashby, E. C. Acc. Chem. Res. 1988, 21, 414, and references
therein.
(2) Alnajjar, M. S.; Kuivila, H. G. J . Am. Chem. Soc. 1985, 107, 416,
and references therein.
(3) Newcomb, M.; Curran, D. P. Acc. Chem. Res. 1988, 21, 206, and
references therein.
(4) Save´ant, J .-M. Adv. Phys. Org. Chem. 1990, 26, 1, and references
(16) Lawler, R. G.; Barbara, P. F. J . Magn. Reson. 1980, 40, 135.
(17) Turro, N. J . Modern Molecular Photochemistry; Benjamin: CA,
1978.
(18) In the case of slow thermal reactions, 1H and 13C CIDNP spectra
have been easily observed.9,10
therein.
(19) In the present reactions, moreover, an insoluble salt (KX) was
generated in the NMR tube. Thus the resolution and sensitivity of the
1H and 13C NMR spectra were very low, and the 1H and 13C CIDNP
spectra could not be observed.
(5) Smith, G. F.; Kuivila, H. G.; Simon, R.; Sultan, L. J . Am. Chem.
Soc. 1981, 103, 833.
(6) San Filippo, J ., J r.; Silbermann, J . J . Am. Chem. Soc. 1981, 103,
5588.
(20) The chemical shift (∼600 ppm) and relative sensitivity (2.23
times larger than that of 13C) of119Sn NMR are much larger than those
of 1H and 13C NMR.
(7) Ashby, E. C.; Su, W.-Yang; Pham, T. N. Organometallics 1985,
4, 1493.
10.1021/om990094p CCC: $18.00 © 1999 American Chemical Society
Publication on Web 07/10/1999

2942 Organometallics, Vol. 18, No. 16, 1999
Communications
F igu r e 1. NMR spectra observed for the reaction of the tri-n-butylstannyl anion with (a) n-BuBr and (b) n-BuI in THF.
The signals denoted by * are Et4Sn (+2.2 ppm) and Me6Sn2 (-109 ppm).
butylstannane, proves that radical intermediates are
involved in the present reactions.
Tri-n-butylstannylpotassium, the stannyl anion re-
agent used, was prepared by reaction of hexa-n-butyl-
distannane (Bu3SnSnBu3, Aldrich Chemical Co.) with
potassium in dry THF under an argon atmosphere at 0
°C.21
The 119Sn NMR spectra of the reaction of the stannyl
anion with n-butyl halides in THF were measured at
room temperature. Although the J -coupling between tin
and proton would give the polarization by the multiplet
1
effects, to observe a simple 119Sn CIDNP spectrum, H
decoupling was employed. The 119Sn NMR spectra
obtained before, during, and after reactions of the
stannyl anion with n-BuBr and n-BuI are shown in
Figure 1. Several new signals were observed in the
spectra obtained during and after the reactions. On the
basis of NMR data reported in the literature,25 the
signals observed at -12 and -83 ppm can safely be
assigned to the 119Sn resonances of tetra-n-butylstan-
nane (Bu4Sn) and hexa-n-butyldistannane (Bu6Sn2),
respectively. The main signal due to Bu4Sn shows strong
emission (E). In addition, the signal due to Bu6Sn2,
observed in the reaction with n-BuI, shows enhanced
absorption (A).26
Although the SN2 pathway cannot be wholly excluded,
these polarized signals prove that the formation mech-
anisms of Bu4Sn and Bu6Sn2 involve radical pathways.
Therefore, the mechanism of the n-Bu3SnK/n-BuX reac-
tion can be rationalized as shown in Scheme 1. In the
primary process, the stannyl radical and the alkyl halide
anion radical are generated. Since electron transfer to
n-butyl bromide and iodide is dissociative,4 the n-butyl
radical and halide anion are generated immediately. In
the cage, the tri-n-butylstannyl and n-butyl radicals
couple to form Bu4Sn. However, the stannyl radicals can
partially escape from the cage. Those that escape may
couple with each other, resulting in formation of n-Bu6-
Sn2.
Bu3SnSnBu3 + 2K f 2Bu3Sn- K+
(2)
The CIDNP spectra measurements were carried out
with a 500 MHz FT-NMR instrument (J EOL R-500).22
THF-d8 was used as a lock substance. Tetraethylstan-
nane (Et4Sn; δ +2.2 ppm) and hexamethyldistannane
(Me6Sn2; δ -109 ppm) were used as internal standards
for the chemical shifts23 and the signal phase. THF-d8,
Et4Sn, and Me6Sn2 were placed in a capillary (1 mm),
which was fixed in the NMR tube (5 mm). The THF
solutions of the stannyl anion and the alkyl halide were
separately charged into the NMR tube and mixed in the
NMR spectrometer.24 The concentrations of tri-n-butyl-
stannyl anion and alkyl halide in the employed THF
solutions were (2.5-2.9) × 10-1 and 5.4 × 10-1 mol
dm-3, respectively.
(21) Tri-n-butylstannylpotassium was prepared similarly to the
preparation of trimethylstannylsodium as described in ref 5. In a
typical reaction, 5.30 g (136 mmol) of finely cut potassium was added
to 300 mL of dry THF. Hexa-n-butyldistannane, 25.0 g (43.1 mmol),
was added, and the mixture was stirred at 0 °C for 3-4 h. All
procedures were carried out under an argon atmosphere.
(22) 119Sn spectra were taken with an observed frequency of 186.4
MHz, a scan range of 26109.66 Hz, and an acquisition time of 1.255 s.
The total interval of one scan including the signal acquisition was 5.055
s. The pulse width was 5.6 µs (as a 90° pulse). Typically, the signal
was accumulated 64 times. Thus, the total time required to record a
spectrum was about 5.4 min. The peak width of Bu4Sn was 6.6 Hz.
(23) The chemical shifts are given as δ values with tetramethyl-
stannane as standard.
(25) Lehnig, M.; Neumann, W. P.; Seifert, P. J . Organomet. Chem.
1978, 162, 145.
(26) In the case of the reaction with n-BuBr, we could not observe
a clear CIDNP spectrum of n-Bu6Sn2. This may be due to the low yield
of n-Bu6Sn2.
(24) THF solutions of the stannyl anion and alkyl halide flowed
through an NMR tube with a flow rate of 0.5-1.0 mL/s.

Communications
Organometallics, Vol. 18, No. 16, 1999 2943
34
Sch em e 1
formation of n-Bu6Sn2 and butane is consistent with
the intermediacy of the tri-n-butylstannyl and n-butyl
radicals. In the presence of dicyclohexylphosphine (2.4
M), which is a well-known radical trap,2,7 the yields of
n-Bu4Sn dropped to 74% (with n-BuBr) and 55% (with
n-BuI).35 From these results, the yield of n-Bu4Sn
generated through the radical pathway is at least 20%,
but it is considered to be much larger.36
Finally, the coupling rate constant of the geminate
radical pair is considered. From the experiments with
a cyclizable radical probe, in which both cyclic and
unrearranged products were generated, Ashby and co-
workers reported that the reaction of alkyl halide with
trimethylstannyl anion proceeded by an electron-
transfer process involving radical intermediates.7 This
means that the rate of the coupling reaction of the
geminate radical pair is comparable to that of cyclization
of the radical probe. San Filippo and Silbermann,
however, suggested that the rate of cyclization (105 s-1
)
is 104-5 times smaller than that of the diffusion-
controlled coupling reaction (109-10 s-1).6 Furthermore,
Newcomb and Curran reported that the observation of
rearranged products in a cyclizable radical probe experi-
ment does not provide conclusive evidence for an
electron-transfer process involving radical intermedi-
ates, because its intermediates may not always be on
the direct pathway between reactants and products.3
Thus, the geminate coupling process is considered to be
much faster than that reported by Ashby and co-
workers. In the present study, the rate constant of the
coupling reaction was considered to be as fast as 109-10
s-1, because it is necessary for the appearance of the
CIDNP that the rate constant of the coupling reaction
of the radical pair is comparable to that of the singlet-
triplet spin conversion due to the hyperfine coupling
mechanism (109-10 s-1).15 It is noteworthy that the
CIDNP technique is much more powerful for determi-
nation of the radical pathway than the cyclizable radical
probe.
The polarization observed in the present reaction
cannot be explained by Kaptein’s CIDNP phase rule
with the high-field approximation, in which only S-T0
transition is considered.27 This is so because the 119Sn
hyperfine coupling (hfc) constant in the trialkylstannyl
radicals (a(119Sn) ) -0.161 T)28 is not small compared
to the electronic interaction in the radical pairs (∆gB )
(g(Bu3Sn•)29 - g(Bu•)30) × B ) (2.016-2.0027) × 11.74
T ) 0.156 T), and the high-field approximation breaks
119
down in the rule. For a quantitative analysis of the
-
Sn CIDNP spectra, one should use the stochastic
Liouville equation including S-T-1 and S-T+1 transi-
tions and all necessary interactions,15,31 but this is
beyond the scope of the present paper. However, from
the experimental and theoretical results reported previ-
ously,32 the qualitative features of the 119Sn CIDNP
spectra for the reactions of trialkylstannyl radicals (∆g
> 0, a < 0) can be summarized as follows: (1) in-cage
and out-of-cage products show the same polarization,
(2) the products generated from a triplet radical pair or
a free pair show enhanced absorption, and (3) the
distannane generated from a free pair shows enhanced
absorption. In the present reactions, if the geminate
radical pair of the stannyl and n-butyl radicals is a
triplet one, the in-cage product of n-Bu4Sn is expected
to show enhanced absorption. As shown in Figure 1,
n-Bu4Sn, however, shows emission. Therefore, the gemi-
nate radical pair is considered to be a singlet radical
pair. On the other hand, n-Bu6Sn2 is considered to be
formed from a free pair of two stannyl radicals.
According to the CIDNP technique, a small amount
of polarized species can give a strong CIDNP spectrum.
To avoid such over interpretation of 119Sn CIDNP,
quantitative analyses of products also were carried out.
The reactions of the tri-n-butylstannyl anion with
n-butyl bromide and iodide gave n-Bu4Sn (94% (with
n-BuBr) and 69% (with n-BuI)), n-Bu6Sn2 (2% (with
n-BuBr) and 30% (with n-BuI)), and butane.33 The
In the present study, we found the strong emissive
CIDNP signal of the main product of n-Bu4Sn. This
result proves that the reactions of the tri-n-butylstannyl
anion with n-butyl bromide and iodide involve a radical
pathway. It is expected that the 119Sn CIDNP technique
can systematically clarify the mechanisms of reactions
of organostannyl anions with many other halides.
Ack n ow led gm en t. We thank Prof. Hisaharu Ha-
yashi (The Institute of Physical and Chemical Research)
for his encouragement. M.W. thanks a partial support
from the President’s Special Research Grant of The
Institute of Physical and Chemical Research.
OM990094P
(34) Although product analyses have been made extensively in
previous studies, the formation of hexaalkyldistannane was not
reported. In the present study, an aliquot was treated with excess
water and unreacted hexa-n-butyldistannane was found to be present
in only trace yield (<1%) by GLC using a Rtx-1 capillary column
(27) Kaptein, R. J . Am. Chem. Soc. 1972, 94, 6251.
(28) Bennett, J . E.; Howard, J . A. Chem. Phys. Lett. 1972, 15, 322.
(29) Lehnig, M. Tetrahedron Lett. 1977, 3663.
(30) Cookson, P. G.; Davies, A. G.; Fazal, N. A.; Roberts, B. P. J .
Am. Chem. Soc. 1976, 98, 616.
(31) Freed, J . H.; Pedersen, J . B. Adv. Magn. Reson. 1976, 8, 1.
(32) Lehnig, M. Chem. Phys. 1975, 8, 419. Lehnig, M. Chem. Phys.
1981, 54, 323. Kruppa, A. I.; Taraban, M. B.; Shokhirev, N. V.;
Svarovsky, S. A.; Leshina, T. V. Chem. Phys. Lett. 1996, 258, 316.
Kruppa, A. I.; Taraban, M. B.; Svarovsky, S. A.; Leshina, T. V. J . Chem.
Soc., Perkin Trans. 2 1996, 2151.
(Restek Co., 15
m
× 0.25 mm). Thus, it is clear that hexa-n-
butyldistannane is the product derived from coupling of the stannyl
radicals.
(35) Similar decreases in the yields of substituted products were
already reported in refs 2 and 7.
(36) Since the coupling of the geminate radical pair is very fast
(109-10 s-1), DCPH could not trap all n-butyl radicals in the pair. If
the coupling and trapping rate constants are 1 × 1010 s-1and 1 × 109
s-1 M-1, respectively, four-fifths of the radical pairs cannot be trapped
by DCPH (2.4 M).
(33) Not quantified.
(37) Lee, K.-W.; San Filippo, J ., J r. Organometallics 1983, 2, 906.