Rate constants for reactions of alkyl radicals with water and methanol complexes of triethylborane

Article abstract of DOI:10.1021/jo070336s

(Chemical Equation Presented) Reactions of secondary alkyl radicals with triethylborane and several of its complexes were studied. The H-atom transfer reactions from Et₃B-OH₂ and Et₃B-OD₂ were suppressed by addition of pyridine to the reaction mixture. Rate constants for reactions of secondary alkyl radicals with triethylborane and its complexes with water, deuterium oxide, methanol, and THF at ambient temperature were determined by radical clock methods. Cyclization of the 1-undecyl-5-hexenyl radical and ring opening of the 1-cyclobutyldodecyl radical were evaluated as clock reactions. The cyclobutylcarbinyl radical ring opening had the appropriate velocity for relatively precise determinations of the ratios of rate constants for H-atom transfer trapping and rearrangement, and these ratios combined with an estimated rate constant for the cyclobutylcarbinyl radical ring opening gave absolute values for the rate constants for the H-atom transfer reactions. For example, the triethylborane-water complex reacts with a secondary alkyl radical in benzene at 20°C with a rate constant of 2 × 10⁴ M ^-1 s^-1. Variable temperature studies with the Et ₃B-CH₃OH complex in toluene indicate that the hydrogen atom transfer reaction has unusually high entropic demand, which results in substantially more efficient hydrogen atom transfer trapping reactions in competition with radical ring opening and cyclization reactions at reduced temperatures.

Full text of DOI:10.1021/jo070336s

Rate Constants for Reactions of Alkyl Radicals with Water and
Methanol Complexes of Triethylborane
Jing Jin and Martin Newcomb*
Department of Chemistry, UniVersity of Illinois at Chicago, 845 West Taylor Street,
Chicago, Illinois 60607
men@uic.edu
ReceiVed February 18, 2007
Reactions of secondary alkyl radicals with triethylborane and several of its complexes were studied. The
H-atom transfer reactions from Et3B-OH2 and Et3B-OD2 were suppressed by addition of pyridine to
the reaction mixture. Rate constants for reactions of secondary alkyl radicals with triethylborane and its
complexes with water, deuterium oxide, methanol, and THF at ambient temperature were determined by
radical clock methods. Cyclization of the 1-undecyl-5-hexenyl radical and ring opening of the
1-cyclobutyldodecyl radical were evaluated as clock reactions. The cyclobutylcarbinyl radical ring opening
had the appropriate velocity for relatively precise determinations of the ratios of rate constants for H-atom
transfer trapping and rearrangement, and these ratios combined with an estimated rate constant for the
cyclobutylcarbinyl radical ring opening gave absolute values for the rate constants for the H-atom transfer
reactions. For example, the triethylborane-water complex reacts with a secondary alkyl radical in benzene
at 20 °C with a rate constant of 2 × 104 M-1 s-1. Variable temperature studies with the Et3B-CH3OH
complex in toluene indicate that the hydrogen atom transfer reaction has unusually high entropic demand,
which results in substantially more efficient hydrogen atom transfer trapping reactions in competition
with radical ring opening and cyclization reactions at reduced temperatures.
Introduction
reagents, alternative radical reducing, or hydrogen atom transfer
trapping, agents have been developed in recent years.
Radical chain reactions have become an integral part of
organic synthetic methodology.1-3 The most common final step
in a chain reaction sequence is radical reduction by hydrogen
atom transfer, and trialkyltin hydrides often have been used as
hydrogen atom donors. Tin hydrides are relatively inexpensive,
they react fast enough to permit their use in low concentrations,4
and the tin-centered radicals formed as intermediates propagate
chain sequences efficiently with many radical precursors.5
Despite these advantages, tin hydride reagents increasingly are
being avoided because tin compounds are toxic and tin-
containing byproducts are difficult to remove from product
mixtures. To address concerns about the use of tin hydride
Many useful applications of borane reagents in radical-based
synthetic methodology exist,6 and one of the more remarkable
recent discoveries was that simple hydroxylic compounds (water
and methanol) complexed with borane reagents can react with
alkyl radicals by H-atom donation. The R-hydrogens of trial-
kylboranes were known to be relatively reactive toward radicals,
with an estimated R-C-H bond dissociation energy (BDE) of
Et3B of only 80 kcal/mol,7,8 and complexation of boranes with
amines was known to activate B-H bonds toward reactions
with radicals.9 The recent findings demonstrated that hydroxylic
hydrogen atoms in the borane complexes also are reactive. Thus,
Wood and co-workers reported that the water complexes with
both Et3B and Me3B reacted with alkyl radicals by H-atom
(1) Giese, B. Radicals in Organic Synthesis: Formation of Carbon-
Carbon Bonds; Pergamon: Oxford, 1986.
(2) Jasperse, C. P.; Curran, D. P.; Fevig, T. L. Chem. ReV. 1991, 91,
1237-1286.
(3) Radicals in Organic Synthesis; Renaud, P., Sibi, M. P., Eds.; Wiley-
VCH: Weinheim, Germany, 2001.
(6) Darmency, V.; Renaud, P. Top. Curr. Chem. 2006, 263, 71-106.
(7) Grotewold, J.; Lissi, E. A.; Scaiano, J. C. J. Organomet. Chem. 1969,
19, 431-434.
(4) Chatgilialoglu, C.; Newcomb, M. AdV. Organomet. Chem. 1999, 44,
(8) Walton, J. C.; McCarroll, A. J.; Chen, Q.; Carboni, B.; Nziengui, R.
J. Am. Chem. Soc. 2000, 122, 5455-5463.
67-112.
(5) Carlsson, D. J.; Ingold, K. U. J. Am. Chem. Soc. 1968, 90, 7047-
7055.
(9) Paul, V.; Roberts, B. P.; Willis, C. R. J. Chem. Soc., Perkin Trans.
2 1989, 1953-1961.
10.1021/jo070336s CCC: $37.00 © 2007 American Chemical Society
Published on Web 06/19/2007
5098
J. Org. Chem. 2007, 72, 5098-5103

Rate Constants for Reactions of Alkyl Radicals
TABLE 1. Product Yields from Reactions of the 1-Methylnonyl
Radicala
SCHEME 1
equiv
of Et3Bb
additives (equiv)b
none
H2O (5)
D2O (5)
% yieldc
% Dd
5
5
10
5
5
5
83
73
82
71
e
61
50
24
53
80
74
62
33
1
D2O (5)
D2O (5), pyridine (1)
D2O (5), pyridine (3)
D2O (5), pyridine (5)
D2O (5), pyridine (10)
5
5
transfer of the hydroxylic hydrogen atoms,10 and Renaud and
co-workers found that methanol similarly reacted with radicals
in the presence of catechol boranes.11 The requisite feature of
these complexes is that Lewis acid-base interactions between
the boranes and hydroxylic compounds substantially reduce the
O-H homolytic BDEs in the complexes. Indeed, for the case
of water complexed with Me3B, the O-H BDE was computed
to be reduced by 30 kcal/mol from that of the O-H bond in
uncomplexed water.10
a Reactions in benzene at room temperature with O2 initiation. b Equiva-
lents of borane or additive relative to the radical precursor. c Percent yield
of decane. d Percentage of decane-d1. e Not determined.
SCHEME 2
The use of borane-complexed alcohols and water as radical
reducing agents is quite attractive due to the ease of removing
boron-containing byproducts. The full potential of these reagents
in radical chain reactions should become apparent as more
examples of their applications are reported and with increased
understanding of the mechanisms and rate constants for their
reactions. In the present work, we address the reactivity and
kinetics question in a radical clock12,13 study that provided rate
constants for reactions of an alkyl radical with triethylborane
and its complexes of water and methanol. The rate constants at
room temperature are approximately 2 orders of magnitude
smaller than that of the archetypal tin hydride reagent Bu3SnH,
and hydrogen atom abstraction from the R-position of the
boranes competes with abstraction of the hydroxylic H-atoms
of the complexes. At reduced reaction temperatures, H-atom
transfer from the Et3B-CH3OH complex was unexpectedly
efficient in comparison to radical rearrangements, reflecting
unusually high entropy demand in the transition state for the
H-atom transfer reaction of the borane complex.
incorporated into the decane. Eventually, the deuterium incor-
poration approached zero at high pyridine concentration.
Interestingly, the absolute yield of non-deuterated decane was
relatively constant in the studies with D2O and pyridine present;
this suggests that a background reaction involving Et3B reacting
as a H-atom donor persisted under all conditions.
The yield of decane in the studies in Table 1 was ca. 80%
under the best conditions and decreased to about 20% when a
large excess of pyridine was present. The absence of deuterium
in the decane in the presence of excess pyridine indicates that
ca. 20% of the alkyl radicals reacted in disproportionation
reactions and/or with the R-positions of (complexed) Et3B, and
these background reactions appeared to be relatively constant
in all studies. Several reaction channels are available to the
1-methylnonyl radicals to account for reduced yields when
efficient H-atom trapping agents are not present. Radical-radical
reactions (couplings and disproportionation reactions) are pos-
sible as is alkyl radical reaction with oxygen that is admitted
slowly into the reaction mixture to initiate chain reactions. Alkyl
radicals also are expected to add to borane-complexed pyridine
in reactions similar to those of alkyl radicals with protonated
pyridines.14,15 In addition, the alkyl radical can add to xanthate
precursor at the thione sulfur to give the self-trapped product
S-methyl-S′-(1-methylnonyl)dithiocarbonate (Scheme 2). The
self-trapping reaction of xanthate esters is a well-known and
important process in their radical chemistry,16 and it limits yields
in all reactions using xanthate esters as radical precursors.
The reaction in Scheme 2 was confirmed qualitatively by
NMR spectral analysis of the crude product mixtures from
reactions conducted in the presence of excess pyridine. In the
NMR spectrum of the starting xanthate ester, the methine proton
at C2 of the decyl group (adjacent to the oxygen atom) absorbed
as a multiplet at δ 5.66-5.72. In the NMR spectra of the crude
Results and Discussion
In a study patterned after that reported by Wood and co-
workers,10 the 1-methylnonyl radical (1) produced in chain
reactions from the xanthate ester of 2-decanol was allowed to
react in benzene in the presence of triethylborane with various
additives present (Scheme 1). The objective was to evaluate
whether the trialkylborane reagent was responsible for a portion
of the H-atom transfer reactions. The yield of decane was
determined by quantitative GC analysis, and, where appropriate,
the percent incorporation of deuterium was determined by GC-
mass spectrometry.
The results are in Table 1. Decane was formed in 70-80%
yield with Et3B alone or with a 1:1 mixture of Et3B and water.
For a 1:1 mixture of Et3B with D2O, the yield of decane was
unchanged, and the amount of deuterium incorporation in the
decane was ca. 80%. With D2O present, addition of pyridine as
a competitive Lewis base for Et3B resulted in a reduction in
both the yield of decane and the percentage of deuterium
(10) Spiegel, D. A.; Wiberg, K. B.; Schacherer, L. N.; Medeiros, M. R.;
Wood, J. L. J. Am. Chem. Soc. 2005, 127, 12513-12515.
(11) Pozzi, D.; Scanlan, E. M.; Renaud, P. J. Am. Chem. Soc. 2005,
127, 14204-14205.
(12) Griller, D.; Ingold, K. U. Acc. Chem. Res. 1980, 13, 317-323.
(13) Newcomb, M. Tetrahedron 1993, 49, 1151-1176.
(14) Minisci, F.; Vismara, E.; Fontana, F.; Morini, G.; Serravalle, M.;
Giordano, C. J. Org. Chem. 1986, 51, 4411-4417.
(15) Minisci, F.; Vismara, E.; Fontana, F. Heterocycles 1989, 28, 489-
519.
(16) Quiclet-Sire, B.; Zard, S. Z. Top. Curr. Chem. 2006, 264, 201-
236.
J. Org. Chem, Vol. 72, No. 14, 2007 5099

Jin and Newcomb
TABLE 2. Rate Constants for Reactions of Radical Trapping
Agents with Secondary Alkyl Radicals at Room Temperaturea
SCHEME 3
kH/kR
kH
b
radical
complex
(M-1
)
(M-1 s-1
)
3
Et3B-H2O
Et3B-D2O
Et3B-CH3OH
Et3B
Et3B-H2O
Et3B-D2O
Et3B-CH3OH
Et3B-THF
0.12 ( 0.02
0.04 ( 0.005
0.11 ( 0.02
4.2 ( 0.3
2 × 104
0.8 × 104
2 × 104
8
0.4 × 104
2.0 × 104
0.8 × 104
1.1 × 104
0.5 × 104
20.0 ( 1.0
8.2 ( 0.8
11.2 ( 1.4
5.2 ( 0.2
a For calculations of kH, the rate constant for cyclization of radical 3
was assumed to be 2 × 105 s-1, and the rate constant for ring opening of
the cyclobutylcarbinyl radical 8 was assumed to be 1 × 103 s-1
.
b The stated
errors in the relative rate constants are at one standard deviation for the
least-squares fits of the data.
SCHEME 4
product mixtures, this signal was lost, and a new multiplet at δ
5.02-5.08 was observed. We assume that the new signal is from
the methine proton at C2 in the decyl group, now adjacent to
the sulfur atom in the dithiocarbonate product.
For kinetic determinations, a chain reaction using the well-
established 5-exo cyclization of a 5-hexenyl radical as a
clock12,13 was designed (Scheme 3). The xanthate ester 2 was
employed as the precursor because xanthate esters from second-
ary alcohols are relatively efficient in radical generation.16
Details for the synthesis of 2 are in the Supporting Information.
The chain reaction sequence gave radical 3 that either cyclized
to give a mixture of radicals 4 or reacted with a H-atom transfer
agent X-H to give 1-heptadecene (5). Cyclic radicals 4 also
can react with the H-atom trapping agent to give cis- and trans-
1-methyl-2-undecylcyclopentane (6). The ratio of cis-6 to trans-6
was ca. 1:3, similar to the ratio of isomers formed in the
cyclization of the 1-methyl-5-hexenyl radical.17,18
The kinetically important reactions in the radical clock study
are the cyclization and the H-atom trapping reactions of radical
3. The kinetics are described by eq 1, where [5] and [6] are the
molar yields of the respective products, kH is the second-order
rate constant for H-atom transfer, kR is the first-order rate
constant for rearrangement (cyclization in this case), and [X-H]
is the concentration of trapping agent.13 A plot of acyclic product
5 to cyclic product 6 as a function of concentration of trapping
agent X-H will have a slope of kH/kR.
studies appear to be internally consistent with those obtained
with a second radical clock (see below).
A slower radical clock reaction was desired so that compa-
rable amounts of unrearranged and rearranged products would
be formed in reactions with the borane complexes. With such
a clock, the precision in the product ratios would be expected
to improve considerably. Ring openings of cyclobutylcarbinyl
radicals are known to have rate constants at ambient temperature
that are approximately 2 orders of magnitude smaller than the
rate constants for cyclizations of 5-hexenyl radicals.13,19,20 Thus,
the cyclobutylcarbinyl radical 8 from precursor 7 was envisioned
as a useful clock (Scheme 4). As in the previous study with 3,
radical 8 can be trapped by X-H with rate constant kH to give
the cyclobutane product 10 or can rearrange with rate constant
kR, and the acyclic radical(s) can then react with the H-atom
donor. In practice, we observed that trans-4-hexadecene (11)
was the only significant isomer formed in the radical ring
opening. The ratio of products 10/11 can be used to determine
the ratio of rate constants according to eq 2.
[5]/[6] ) (kH/kR) [X-H]
(1)
A series of reactions was conducted with precursor 2 and
varying concentrations of 1:1 mixtures of Et3B with water,
deuterium oxide, and methanol, and the product ratios were
determined by GC. The ratios of rate constants (kH/kR) from
these studies are listed in Table 2. Due to the relatively small
rate constant for the trapping reaction, radical 3 mainly cyclized,
giving cyclic to acyclic product ratios in the range of 15:1 to
50:1, and the error in the ratio of products determined from the
slopes of the plots according to eq 1 approached 20% of the
absolute value. Nonetheless, the values of kH determined in these
[10]/[11] ) (kH/kR) [X-H]
(2)
The synthesis of precursor 7 is described in the Supporting
Information. Xanthate 7 was allowed to react with H-atom
(19) Beckwith, A. L. J.; Moad, G. J. Chem. Soc., Perkin Trans. 2 1980,
1083-1092.
(20) Ingold, K. U.; Maillard, B.; Walton, J. C. J. Chem. Soc., Perkin
Trans. 2 1981, 970-974.
(17) Walling, C.; Cioffari, A. J. Am. Chem. Soc. 1972, 94, 6059-6064.
(18) Beckwith, A. L. J.; Easton, C. J.; Lawrence, T.; Serelis, A. K. Aust.
J. Chem. 1983, 36, 545-556.
5100 J. Org. Chem., Vol. 72, No. 14, 2007

Rate Constants for Reactions of Alkyl Radicals
plexation will accelerate a 5-exo radical cyclization such as that
in radical 3.21 The computational results suggest that an alkali
metal ion will accelerate the radical cyclization in the gas phase,
but little evidence for such an acceleration in condensed phase
is available, probably because the binding constants for forma-
tion of complexes between an alkene and an alkali metal ion
are small. For the ring opening reaction of radical 8, Lewis acids
should not affect the reaction because a complex will not be
present in condensed phase.
With the caveat that the rate constant for cyclization of radical
3 might be affected by the Lewis acid Et3B, we used estimated
rate constants for cyclization of 3 and ring opening of 8 to
calculate second-order rate constants for the H-atom transfer
reactions. Both the 5-hexenyl radical and the 1-methyl-5-hexenyl
radical cyclize at 20 °C with rate constants of kR ) 2 × 105
FIGURE 1. Product ratios from reactions of radical 8 in the presence
of Et3B with water (b), with deuterium oxide (9), or with methanol
(2). U/R is the ratio of the cyclobutyl product 10 to ring opened product
11. The lines are least-squares fits for each data set.
s-1 17,18,22,23
and we assumed that this would be equal to the
,
rate constant for cyclization of 3. For the ring opening of radical
8, we assumed that the rate constant would be equal to that for
ring opening of the cyclobutylcarbinyl radical at 20 °C, kR )
trapping agents at varying concentrations, and the products were
determined by GC analysis. Hydrocarbon products 10 and 11
were identified by comparison to authentic samples. The yields
of hydrocarbons 10 and 11 were high (>70%) with Et3B-H2O,
consistent with expectations that the generation of 8 from a
radical chain reaction of 7 would be relatively efficient. Also
as expected, the ratios of unrearranged to rearranged products
were considerably larger than those from reactions of radical 3.
Figure 1 shows representative results for the reactions of
radical 8 in the presence of Et3B with additives. The ratio of
10 to 11 is plotted against the concentration of the Et3B complex,
and the slope of these plots is the ratio of rate constants kH/kR.
The near-zero intercepts of these plots are noteworthy because
they show that the radical reaction is not reversible (see below).
Complete results for the studies with radical 8 are given in the
Supporting Information, and the ratios of rate constants obtained
in these studies are listed in Table 2. In most of the experiments,
the relative amount of unrearranged product was in the range
of 20-80%, and good precision was obtained in the relative
rate constants.
1000 s-1 4,19,24
These values for kR and the experimental ratios
.
of rate constants gave the values for kH that are listed in Table 2.
The values of the absolute rate constants kH might have
systematic errors due to the assumed rate constants for the
rearrangements, but those errors are not likely to be as large as
a factor of 2. More importantly, the relative rate constants for
two trapping agents reacting with the same radical clock will
have good accuracy. Thus, triethylborane complexed with water
reacted with radical 8 five times as fast as the uncomplexed
borane reagent, and the presence of the Lewis basic solvent THF
as an additive had little effect on the reactivity of the
R-hydrogens of Et3B. Assuming that the R-hydrogen atoms in
Et3B complexed with H2O and D2O have the same reactivity
as in the uncomplexed borane, one can factor out an equal
contribution to the kinetics of reactions of the borane R-hydro-
gens from the H2O and D2O reactions to give a kinetic isotope
effect for reaction of the water complexes with radical 8 of kH/
kD ) (2.0 - 0.4)/(0.8 - 0.4) ) 4.
The rate constants we determined are for reactions at room
temperature, but one can make qualitative predictions about the
rate constants at other temperatures because various second-
order hydrogen atom transfer reactions have relatively consistent
entropy demands. More importantly for synthetic considerations,
one can predict that the hydrogen atom transfer reactions from
the triethylborane-water and -methanol complexes should
become increasingly competitive with radical rearrangements
such as we used for the radical clocks in this work. For example,
the log A values for many hydrogen atom transfer reactions to
carbon-centered radicals are in the range of 8-9.4,13 The 5-exo
radical cyclization reactions, such as the cyclization of radical
clock 3, have entropy of activation terms that are less negative
than those for second-order bimolecular H-atom transfer reac-
tions; one can use an approximate value of log A ) 10 for such
When a radical clock study is performed with a relatively
slow trapping agent, one should determine whether the clock
reaction is reversible under the reaction conditions. That can
be established in the type of study performed here because the
intercepts of the plots of product ratios versus concentration of
trapping agent would be non-zero if the radical reaction was
reversible.13 As indicated in Figure 1 and listed in the Supporting
Information, we found zero intercepts within experimental error
for all reactions with radical 8, thus ensuring that the reactions
were not reversible.
The ratios of rate constants determined experimentally can
be used to calculate absolute rate constants kH if the rate constant
for the rearrangement reaction is known. This “radical clock”
method is commonly employed with the assumption that the
rate constants for the rearrangement reactions are not affected
by any special reaction conditions. The rate constants for simple
radical cyclization and ring opening reactions such as those used
here are generally not subject to noticeable solvent effects.13
Nonetheless, it is possible, in principle, that Lewis acids will
affect the kinetics of radical reactions, and Horn and Clark
explored computationally the possibility that Lewis acid com-
(23) Lusztyk, J.; Maillard, B.; Deycard, S.; Lindsay, D. A.; Ingold, K.
U. J. Org. Chem. 1987, 52, 3509-3514.
(24) In the original study of ring openings of cyclobutylcarbinyl radical
ring openings, the value obtained for the rate constant for cyclobutylcarbinyl
ring opening was k ) 500 s-1 at ambient temperature, but the value assumed
for the rate constant for the competing reaction, Bu3SnH trapping of an
alkyl radical, was 1 × 106 M-1 s-1 (ref 19). The currently accepted value
for the rate constant for the Bu3SnH reaction is 2 × 106 M-1 s-1 (ref 4).
(25) For a reaction with ∆Sq ) 0 eu at room temperature, the equivalent
Arrhenius function term is log A ) 13.1.
(21) Horn, A. H. C.; Clark, T. J. Am. Chem. Soc. 2003, 125, 2809-
2816.
(22) Chatgilialoglu, C.; Ingold, K. U.; Scaiano, J. C. J. Am. Chem. Soc.
1981, 103, 7739-7742.
(26) Wilt, J. W.; Lusztyk, J.; Peeran, M.; Ingold, K. U. J. Am. Chem.
Soc. 1988, 110, 281-287.
J. Org. Chem, Vol. 72, No. 14, 2007 5101

Jin and Newcomb
TABLE 3. Rate Constants for Reactions of H-Atom Donors with
Alkyl Radicals at Room Temperature and Bond Dissociation
Energiesa
BDE
donor
kH (M-1 s-1
)
ref
(kcal/mol)b
ref
32
7
10
33
32
34
38
Et3SiHc
3.2 × 102
4
90
80
86e
83
79
74
89
83
78
Et3Bd
4 × 103
this work
this work
26
27
4, 22
28
Et3B-H2O
Bu3GeH
((CH3)3Si)3SiH
Bu3SnH
t-BuSH
2.0 × 104
1.0 × 105
3.9 × 105
2.2 × 106
6 × 106
9 × 107
29, 30
30, 31
29, 35
39
FIGURE 2. Logarithm of ratio of products 10 to 11 as a function of
temperature for reactions in toluene in the presence of 0.26 M Et3B-
CH3OH. Note that log([10]/[11]) ) log(kH/kR).
PhSH
PhSeH
1.2 × 109
a Rate constants for reactions with primary and/or secondary alkyl
radicals. b BDE ) homolytic bond dissociation energy. The BDE values
listed are from original literature, but re-evaluation of the data for group
14 hydrides indicates that the scale should be shifted upward (see ref 4).
c Rate constant for reaction of the Si-H bond only; the R-hydrogens of the
ethyl groups also react. d Rate constant for reaction of Et3B is assumed to
involve only the R-hydrogen atoms. e Computed BDE for the O-H bond
in the Me3B-H2O complex.
a cyclization.13 Radical ring opening reactions, such as ring
opening of radical clock 8, have even less negative or slightly
positive entropies of activation,13 and one can estimate the log
A term for ring opening of 8 as log A ≈ 13.25 The result is that
one would predict that the hydrogen atom transfer reaction from
Et3B-OH2 or Et3B-CH3OH will become increasingly competi-
tive with a 5-exo radical cyclization or a radical ring opening
as the reaction temperature is decreased.
alkyl radicals are listed in Table 3.4,22,26-31 Some Group 16
H-atom donors also are listed in Table 3. Note that the reactivity
of group 14 metal hydrides is mainly a function of the difference
in bond dissociation energy (BDE) for the donor4,32-34 and a
C-H bond of an alkyl group; the BDE values for primary and
secondary C-H bonds are 101.1 and 98.6 kcal/mol, respec-
tively.35,36 In the case of group 16 donors, high reactivity results
from both differences in BDEs29,35,37-39 and favorable polar
effects for reactions of electron-rich carbon radicals with an
electron-deficient hydrogen atom of the donor.
The above prediction was tested in the following study.
Radical clock 8 was allowed to react in a series of reactions at
various temperatures and with an equal concentration of Et3B-
CH3OH (0.26 M) in each reaction. The relative ratio of
unrearranged to rearranged product ([10]/[11]) was 0.7 at 53
°C but increased at lower temperatures to give, ultimately, a
ratio of 80 at -30 °C. These results are qualitatively in line
with the predicted temperature-dependent behavior.
A quantitative evaluation of the variable temperature results
indicates that the hydrogen atom transfer reaction from Et3B-
CH3OH is more entropically demanding than initially assumed.
The product ratios versus temperature are presented in Figure
2 in the form of a relative Arrhenius plot. For these data, the
relative Arrhenius function in eq 3 would apply, where the
subscripts H and R represent the hydrogen atom transfer reaction
and ring opening reaction of 8, respectively. The plot in Figure
2 gives ∆log A ) (-6 ( 1) and ∆Ea ) (-8.8 ( 1.2) kcal/mol,
where the errors are at 2σ.
Relatively unreactive H-atom donors, such as Et3SiH, are not
useful in synthesis because they react too slowly with carbon
radicals, and radical-radical coupling reactions dominate when
chain reactions are attempted. The germanium hydride Bu3GeH
reacts fast enough to be used in synthesis,26 and the Et3B-
H2O complex has similar reactivity. Thus, the inherent reactivity
of the borane-water complex appears to be adequate for some
synthetic conversions.
The borane-water complex should have a similar polar
acceleration effect as observed in thiols, and this factor in
(27) Chatgilialoglu, C.; Dickhaut, J.; Giese, B. J. Org. Chem. 1991, 56,
6399-6403.
log(kH/kR) ) (log AH - log AR) - (EaH - EaR)/2.3RT
(3)
(28) Ha, C.; Horner, J. H.; Newcomb, M.; Varick, T. R.; Arnold, B. R.;
Lusztyk, J. J. Org. Chem. 1993, 58, 1194-1198.
(29) Franz, J. A.; Bushaw, B. A.; Alnajjar, M. S. J. Am. Chem. Soc.
1989, 111, 268-275.
Another qualitative evaluation of the temperature dependence
of hydrogen atom transfer from triethylborane complexes was
made using the 5-exo cyclization of radical 3. The entropic
demand of a 5-exo radical cyclization is more negative than
that for ring opening of a cyclobutylcarbinyl radical but not as
negative as that of the bimolecular hydrogen atom transfer
reaction. Thus, a reduction of reaction temperature was expected
to result in more efficient trapping of the acyclic radical 3. At
20 °C with 0.1 M Et3B-CH3OH, acyclic product 5 was only
2% of the product mixture. By increasing the concentration of
Et3B-CH3OH to 0.3 M and reducing the reaction temperature
to -30 °C, the relative yield of acyclic product 5 was increased
to 20%.
(30) Newcomb, M.; Choi, S. Y.; Horner, J. H. J. Org. Chem. 1999, 64,
1225-1231.
(31) Newcomb, M.; Varick, T. R.; Ha, C.; Manek, M. B.; Yue, X. J.
Am. Chem. Soc. 1992, 114, 8158-8163.
(32) Kanabus-Kaminska, J. M.; Hawari, J. A.; Griller, D.; Chatgilialoglu,
C. J. Am. Chem. Soc. 1987, 109, 5267-5268.
(33) Clark, K. B.; Griller, D. Organometallics 1991, 10, 746-750.
(34) Burkey, T. J.; Majewski, M.; Griller, D. J. Am. Chem. Soc. 1986,
108, 2218-2221.
(35) McMillen, D. F.; Golden, D. M. Annu. ReV. Phys. Chem. 1982, 33,
493-532.
(36) Berkowitz, J.; Ellison, G. B.; Gutman, D. J. Phys. Chem. 1994, 98,
2744-2765.
(37) Griller, D.; Kanabus-Kaminska, J. M.; Maccoll, A. THEOCHEM
1988, 40, 125-131.
For perspective, it is useful to compare the reactivity of the
Et3B-OH2 complex as a H-atom donor for alkyl radicals with
the reactivities of other H-atom transfer trapping agents. Group
14 metal hydrides are commonly used in synthesis, and rate
constants for archetypal members of this group reacting with
(38) Janousek, B. K.; Reed, K. J.; Brauman, J. I. J. Am. Chem. Soc.
1980, 102, 3125-3129.
(39) Leeck, D. T.; Li, R. M.; Chyall, L. J.; Kenttamaa, H. I. J. Phys.
Chem. 1996, 100, 6608-6611.
(40) Allen, R. P.; Roberts, B. P.; Willis, C. R. J. Chem. Soc., Chem.
Commun. 1989, 1387-1388.
5102 J. Org. Chem., Vol. 72, No. 14, 2007

Rate Constants for Reactions of Alkyl Radicals
combination with the computed small BDE10 for the Me3B-
H2O complex of 86 kcal/mol would lead to the prediction that
hydrogen atom transfer trapping would be relatively efficient.
In fact, the Et3B-ROH complexes react less rapidly at room
temperature than one might expect, apparently reflecting the
high entropy demand of these reactions indicated in the variable
temperature studies. The positive aspect of this reactivity effect
is that one can expect large relative changes in product
distributions as a function of reaction temperature, introducing
a subtle method for controlling product yield.
Experimental Section
The syntheses of the xanthate ester from 2-decanol and radical
precursors 2 and 7 and details for identifications of the products
from their reactions are described in the Supporting Information.
Yields were determined by analytical GC with thermal conductivity
detection. In studies involving measurements of the percentage of
deuterium in a product, GC-mass spectrometry with EI ionization
and SIM mode data collection was employed.
Procedure for Kinetic Studies. In a typical reaction, xanthate
precursor 7 (20 mg, 0.06 mmol) was dissolved in 3.4 mL of
benzene, and 21.6 µL of D2O (1.1 mmol) was added. The mixture
was degassed by bubbling with nitrogen gas for 20 min. Et3B (1.0
M in hexane, 1.21 mL, 1.2 mmol) was added to the mixture. The
reaction mixture was stirred, and oxygen was introduced slowly
via a bubbler. The reaction was continued until no starting material
was observed by TLC analysis (ca. 6 h). The reaction mixture was
passed through a small column containing silica gel with hexane
elution. Tetradecane was added as an internal standard, and the
mixture was analyzed by GC. The yields of products 10 and 11
were determined using measured response factors.
Variable Temperature Study. Xanthate radical precursor 7 (40
mg, 0.12 mmol) was dissolved in dry toluene (3.4 mL), and 49 µL
of MeOH (1.2 mmol) was added. The mixture was degassed by
bubbling with nitrogen gas for 20 min and equilibrated in a bath at
the desired temperature. Et3B (1.0 M in hexane, 1.21 mL, 1.2 mmol)
was added to the mixture. The reaction mixture was stirred, and
oxygen was introduced slowly via a bubbler. The reaction mixture
was stirred, and the progress of the reaction was followed by TLC.
Reactions at reduced temperature required extended reaction times
(up to 24 h). The reaction mixture was passed through a small
column containing silica gel with hexane elution, and the eluent
was analyzed by GC to determine the ratio of acyclic product 10
to rearranged product 11.
One essential requirement for synthetic applications is that
each step in the radical chain sequence is efficient, and that
would limit the use of the borane-water complex to xanthates
and related radical precursors and possibly alkyl iodides because
an alkyl chloride or alkyl bromide will not react efficiently with
an ethyl radical that is formed from the Et3B-OH radical after
H-atom reaction of the complex.10 Such a limitation also exists
for thiols because a thiyl radical will not abstract halogen from
an alkyl halide. Nonetheless, thiols can be used with alkyl
halides in a mixed reagent system comprised of a thiol as the
H-atom donor and Et3SiH, which reacts with the thiyl radical
to give a silyl radical that propagates the chain with alkyl
halides.40 It is interesting to speculate that a similar mixed
reagent system might permit the use of the borane-water
complex as a H-atom donor when alkyl halide radical precursors
are employed.
In summary, the room temperature rate constants for reactions
of triethylborane complexes of water and methanol as H-atom
donors to alkyl radicals are 2 orders of magnitude smaller than
those for reactions of tin hydride reagents. These reactions are
adequately fast at room temperature for some synthetic applica-
tions, and they are increasingly competitive with radical
rearrangements at reduced temperatures. One anticipates in-
creasing numbers of applications of borane complexes as radical
reducing agents given the ease of removal of the boron-
containing byproducts, especially if a mixed reagent system can
be developed that permits the use of common alkyl halides as
radical precursors.
Acknowledgment. This work was supported by a grant from
the National Science Foundation (CHE-0601857).
Supporting Information Available: Synthetic details, detailed
kinetic results, and NMR spectra. This material is available free of
charge via the Internet at http://pubs.acs.org.
JO070336S
J. Org. Chem, Vol. 72, No. 14, 2007 5103